diff --git a/runs/single_internvl/context/L01_s13/adapter/README.md b/runs/single_internvl/context/L01_s13/adapter/README.md new file mode 100644 index 0000000000000000000000000000000000000000..4f9f856cf0e1531b47e25d266833b2977a19d3ac --- /dev/null +++ b/runs/single_internvl/context/L01_s13/adapter/README.md @@ -0,0 +1,207 @@ +--- +base_model: /data/OPSD/assets/models/InternVL3_5-8B-HF +library_name: peft +pipeline_tag: text-generation +tags: +- base_model:adapter:/data/OPSD/assets/models/InternVL3_5-8B-HF +- lora +- transformers +--- + +# Model Card for Model ID + + + + + +## Model Details + +### Model Description + + + + + +- **Developed by:** [More Information Needed] +- **Funded by [optional]:** [More Information Needed] +- **Shared by [optional]:** [More Information Needed] +- **Model type:** [More Information Needed] +- **Language(s) (NLP):** [More Information Needed] +- **License:** [More Information Needed] +- **Finetuned from model [optional]:** [More Information Needed] + +### Model Sources [optional] + + + +- **Repository:** [More Information Needed] +- **Paper [optional]:** [More Information Needed] +- **Demo [optional]:** [More Information Needed] + +## Uses + + + +### Direct Use + + + +[More Information Needed] + +### Downstream Use [optional] + + + +[More Information Needed] + +### Out-of-Scope Use + + + +[More Information Needed] + +## Bias, Risks, and Limitations + + + +[More Information Needed] + +### Recommendations + + + +Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. More information needed for further recommendations. + +## How to Get Started with the Model + +Use the code below to get started with the model. + +[More Information Needed] + +## Training Details + +### Training Data + + + +[More Information Needed] + +### Training Procedure + + + +#### Preprocessing [optional] + +[More Information Needed] + + +#### Training Hyperparameters + +- **Training regime:** [More Information Needed] + +#### Speeds, Sizes, Times [optional] + + + +[More Information Needed] + +## Evaluation + + + +### Testing Data, Factors & Metrics + +#### Testing Data + + + +[More Information Needed] + +#### Factors + + + +[More Information Needed] + +#### Metrics + + + +[More Information Needed] + +### Results + +[More Information Needed] + +#### Summary + + + +## Model Examination [optional] + + + +[More Information Needed] + +## Environmental Impact + + + +Carbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700). + +- **Hardware Type:** [More Information Needed] +- **Hours used:** [More Information Needed] +- **Cloud Provider:** [More Information Needed] +- **Compute Region:** [More Information Needed] +- **Carbon Emitted:** [More Information Needed] + +## Technical Specifications [optional] + +### Model Architecture and Objective + +[More Information Needed] + +### Compute Infrastructure + +[More Information Needed] + +#### Hardware + +[More Information Needed] + +#### Software + +[More Information Needed] + +## Citation [optional] + + + +**BibTeX:** + +[More Information Needed] + +**APA:** + +[More Information Needed] + +## Glossary [optional] + + + +[More Information Needed] + +## More Information [optional] + +[More Information Needed] + +## Model Card Authors [optional] + +[More Information Needed] + +## Model Card Contact + +[More Information Needed] +### Framework versions + +- PEFT 0.20.0 \ No newline at end of file diff --git a/runs/single_internvl/context/L01_s13/adapter/adapter_config.json b/runs/single_internvl/context/L01_s13/adapter/adapter_config.json new file mode 100644 index 0000000000000000000000000000000000000000..937d5bffbfc4488c98e496fd35a81b77cc0d5bbf --- /dev/null +++ b/runs/single_internvl/context/L01_s13/adapter/adapter_config.json @@ -0,0 +1,50 @@ +{ + "alora_invocation_tokens": null, + "alpha_pattern": {}, + "arrow_config": null, + "auto_mapping": null, + "base_model_name_or_path": "/data/OPSD/assets/models/InternVL3_5-8B-HF", + "bias": "none", + "corda_config": null, + "ensure_weight_tying": false, + "eva_config": null, + "exclude_modules": null, + "fan_in_fan_out": false, + "inference_mode": true, + "init_lora_weights": true, + "layer_replication": null, + "layers_pattern": null, + "layers_to_transform": null, + "loftq_config": {}, + "lora_alpha": 32, + "lora_bias": false, + "lora_dropout": 0.05, + "lora_ga_config": null, + "megatron_config": null, + "megatron_core": "megatron.core", + "modules_to_save": null, + "monteclora_config": null, + "peft_type": "LORA", + "peft_version": "0.20.0", + "qalora_group_size": 16, + "r": 16, + "rank_pattern": {}, + "revision": null, + "target_modules": [ + "model.language_model.layers.1.self_attn.k_proj", + "model.language_model.layers.1.mlp.down_proj", + "model.language_model.layers.1.mlp.up_proj", + "model.language_model.layers.1.self_attn.q_proj", + "model.language_model.layers.1.mlp.gate_proj", + "model.language_model.layers.1.self_attn.v_proj", + "model.language_model.layers.1.self_attn.o_proj" + ], + "target_parameters": null, + "task_type": "CAUSAL_LM", + "trainable_token_indices": null, + "use_bdlora": null, + "use_dora": false, + "use_qalora": false, + "use_rslora": false, + "velora_config": null +} \ No newline at end of file diff --git a/runs/single_internvl/context/L01_s13/config.json b/runs/single_internvl/context/L01_s13/config.json new file mode 100644 index 0000000000000000000000000000000000000000..9c5fdb33c2adb31adaef8d13d71b877b6bf5b655 --- /dev/null +++ b/runs/single_internvl/context/L01_s13/config.json @@ -0,0 +1,24 @@ +{ + "behavior": "context", + "seed": 13, + "model": "/data/OPSD/assets/models/InternVL3_5-8B-HF", + "backbone": "internvl", + "train": "/data/OPSD/assets/v3/context.jsonl", + "items": 1000, + "steps": 250, + "epochs": 1, + "steps_per_epoch": 250, + "groups_per_step": 4, + "lr": 5e-05, + "scheduler": "cosine", + "warmup_steps": 16, + "final_fraction": 0.1, + "weight_decay": 0.0, + "clip": 1.0, + "initial_loss_scale": 4.0, + "adapter": { + "layer": 1, + "rank": 16, + "trainable_parameters": 1212416 + } +} \ No newline at end of file diff --git a/runs/single_internvl/context/L01_s13/steps.jsonl b/runs/single_internvl/context/L01_s13/steps.jsonl new file mode 100644 index 0000000000000000000000000000000000000000..3b37abe9442b45af14c3245037cef7605fb6c91a --- /dev/null +++ b/runs/single_internvl/context/L01_s13/steps.jsonl @@ -0,0 +1,250 @@ +{"step": 1, "steps": 250, "lr": 3.125e-06, "grad_norm": 0.41529902815818787, "loss_scale": 4.0, "retries": 0, "seconds": 1.5, "max_memory_gb": 18.94, "ce": 0.5227766726748087, "kl": 0.08654405153356493, "context": 0.5660487064160407} +{"step": 2, "steps": 250, "lr": 6.25e-06, "grad_norm": 0.630935549736023, "loss_scale": 4.0, "retries": 0, "seconds": 2.0, "max_memory_gb": 18.95, "ce": 0.6404551289961091, "kl": 0.4672839170342513, "context": 0.8740971157312742} +{"step": 3, "steps": 250, "lr": 9.375000000000001e-06, "grad_norm": 1.6668903827667236, "loss_scale": 4.0, "retries": 0, "seconds": 2.5, "max_memory_gb": 18.95, "ce": 1.096772275865078, "kl": 0.06840587593615055, "context": 1.1309751942753792} +{"step": 4, "steps": 250, "lr": 1.25e-05, "grad_norm": 0.8812512755393982, "loss_scale": 4.0, "retries": 0, "seconds": 3.0, "max_memory_gb": 18.95, "ce": 0.45131903726723976, "kl": 0.42920650115434, "context": 0.6659222840607981} +{"step": 5, "steps": 250, "lr": 1.5625e-05, "grad_norm": 0.017900390550494194, "loss_scale": 4.0, "retries": 0, "seconds": 3.5, "max_memory_gb": 18.95, "ce": 0.009307772110332735, "kl": 0.008224977031204617, "context": 0.013420260773273185} +{"step": 6, "steps": 250, "lr": 1.8750000000000002e-05, "grad_norm": 0.4201781153678894, "loss_scale": 4.0, "retries": 0, "seconds": 4.0, "max_memory_gb": 18.95, "ce": 0.2907808331801789, "kl": 0.07358878948252823, "context": 0.3275752314075362} +{"step": 7, "steps": 250, "lr": 2.1875e-05, "grad_norm": 1.3890644311904907, "loss_scale": 4.0, "retries": 0, "seconds": 4.5, "max_memory_gb": 18.95, "ce": 0.8263224340080342, "kl": 0.12455667414030813, "context": 0.8886007743385562} +{"step": 8, "steps": 250, "lr": 2.5e-05, "grad_norm": 1.3712068796157837, "loss_scale": 4.0, "retries": 0, "seconds": 5.0, "max_memory_gb": 18.96, "ce": 0.7885623220354319, "kl": 0.28962204651907086, "context": 0.9333733469247818} +{"step": 9, "steps": 250, "lr": 2.8125000000000003e-05, "grad_norm": 0.9882053732872009, "loss_scale": 4.0, "retries": 0, "seconds": 5.5, "max_memory_gb": 18.96, "ce": 0.5176718682050705, "kl": 0.13548408180940896, "context": 0.5854139216244221} +{"step": 10, "steps": 250, "lr": 3.125e-05, "grad_norm": 0.17280255258083344, "loss_scale": 4.0, "retries": 0, "seconds": 6.0, "max_memory_gb": 18.96, "ce": 0.05230281721014762, "kl": 0.04618364811176434, "context": 0.07539463957073167} +{"step": 11, "steps": 250, "lr": 3.4375e-05, "grad_norm": 1.0913310050964355, "loss_scale": 4.0, "retries": 0, "seconds": 6.5, "max_memory_gb": 18.96, "ce": 0.6889468172521447, "kl": 0.004747790688952591, "context": 0.6913207115358091} +{"step": 12, "steps": 250, "lr": 3.7500000000000003e-05, "grad_norm": 0.8883460760116577, "loss_scale": 4.0, "retries": 0, "seconds": 7.0, "max_memory_gb": 18.96, "ce": 0.4559845983703781, "kl": 0.0027409624003666977, "context": 0.45735508319012297} +{"step": 13, "steps": 250, "lr": 4.0625000000000005e-05, "grad_norm": 0.5871704816818237, "loss_scale": 4.0, "retries": 0, "seconds": 7.5, "max_memory_gb": 18.96, "ce": 0.41896960307713016, "kl": 0.019890877421858022, "context": 0.4289150576769316} +{"step": 14, "steps": 250, "lr": 4.375e-05, "grad_norm": 1.4011739492416382, "loss_scale": 4.0, "retries": 0, "seconds": 8.0, "max_memory_gb": 18.96, "ce": 0.7459917217493057, "kl": 0.9444645934272557, "context": 1.2182240188121796} +{"step": 15, "steps": 250, "lr": 4.6875e-05, "grad_norm": 1.2687474489212036, "loss_scale": 4.0, "retries": 0, "seconds": 8.5, "max_memory_gb": 18.96, "ce": 0.4896892928145462, "kl": 0.46457273134967636, "context": 0.7219756473061807} +{"step": 16, "steps": 250, "lr": 5e-05, "grad_norm": 0.40850716829299927, "loss_scale": 4.0, "retries": 0, "seconds": 9.0, "max_memory_gb": 18.96, "ce": 0.20205964230262907, "kl": 0.00092566255057136, "context": 0.20252247389726108} +{"step": 17, "steps": 250, "lr": 5e-05, "grad_norm": 0.5144745707511902, "loss_scale": 4.0, "retries": 0, "seconds": 9.5, "max_memory_gb": 18.96, "ce": 0.3119380276184529, "kl": 0.8303875376586802, "context": 0.7271318020066246} +{"step": 18, "steps": 250, "lr": 4.999797225177623e-05, "grad_norm": 1.3989078998565674, "loss_scale": 4.0, "retries": 0, "seconds": 10.0, "max_memory_gb": 18.96, "ce": 0.505332664353773, "kl": 0.20895072488929145, "context": 0.6098080199299147} +{"step": 19, "steps": 250, "lr": 4.999188937259492e-05, "grad_norm": 2.441981315612793, "loss_scale": 4.0, "retries": 0, "seconds": 10.5, "max_memory_gb": 18.96, "ce": 0.7192528173327446, "kl": 0.16943835164420307, "context": 0.8039719834923744} +{"step": 20, "steps": 250, "lr": 4.998175245886032e-05, "grad_norm": 0.03257448598742485, "loss_scale": 4.0, "retries": 0, "seconds": 11.1, "max_memory_gb": 18.96, "ce": 0.0024086263565550325, "kl": 0.020944038364177686, "context": 0.012880645335826557} +{"step": 21, "steps": 250, "lr": 4.996756333769319e-05, "grad_norm": 1.2985447645187378, "loss_scale": 4.0, "retries": 0, "seconds": 11.6, "max_memory_gb": 18.96, "ce": 0.33468019276460836, "kl": 0.04821694786005537, "context": 0.3587886760069523} +{"step": 22, "steps": 250, "lr": 4.994932456660157e-05, "grad_norm": 1.2551078796386719, "loss_scale": 4.0, "retries": 0, "seconds": 12.1, "max_memory_gb": 18.96, "ce": 0.35539993245038204, "kl": 0.03179829720102134, "context": 0.37129907045164146} +{"step": 23, "steps": 250, "lr": 4.992703943301973e-05, "grad_norm": 1.625778317451477, "loss_scale": 4.0, "retries": 0, "seconds": 12.6, "max_memory_gb": 18.96, "ce": 0.5366190668974014, "kl": 0.9957966441995723, "context": 1.0345173653067832} +{"step": 24, "steps": 250, "lr": 4.9900711953715675e-05, "grad_norm": 1.2230366468429565, "loss_scale": 4.0, "retries": 0, "seconds": 13.1, "max_memory_gb": 18.96, "ce": 0.2796066291630268, "kl": 0.30150755820795894, "context": 0.43036041036248207} +{"step": 25, "steps": 250, "lr": 4.987034687406713e-05, "grad_norm": 1.581318974494934, "loss_scale": 4.0, "retries": 0, "seconds": 13.6, "max_memory_gb": 18.96, "ce": 0.6545408715210215, "kl": 0.20354427956044674, "context": 0.7563130296766758} +{"step": 26, "steps": 250, "lr": 4.9835949667206215e-05, "grad_norm": 1.482092022895813, "loss_scale": 4.0, "retries": 0, "seconds": 14.1, "max_memory_gb": 18.96, "ce": 0.2845962792634964, "kl": 0.7069850598927587, "context": 0.6380888223648071} +{"step": 27, "steps": 250, "lr": 4.9797526533032945e-05, "grad_norm": 1.6222131252288818, "loss_scale": 4.0, "retries": 0, "seconds": 14.6, "max_memory_gb": 18.96, "ce": 0.37228576313646045, "kl": 0.4234901958025148, "context": 0.5840308593469672} +{"step": 28, "steps": 250, "lr": 4.9755084397097716e-05, "grad_norm": 0.95625901222229, "loss_scale": 4.0, "retries": 0, "seconds": 15.1, "max_memory_gb": 18.96, "ce": 0.6020690612494946, "kl": 0.9375541990157217, "context": 1.0708461254835129} +{"step": 29, "steps": 250, "lr": 4.970863090935304e-05, "grad_norm": 1.3546438217163086, "loss_scale": 4.0, "retries": 0, "seconds": 15.6, "max_memory_gb": 18.96, "ce": 0.2546321381814778, "kl": 0.3638458698987961, "context": 0.43655507266521454} +{"step": 30, "steps": 250, "lr": 4.965817444277468e-05, "grad_norm": 1.3015401363372803, "loss_scale": 4.0, "retries": 0, "seconds": 16.2, "max_memory_gb": 18.96, "ce": 0.3217579457723332, "kl": 0.011077957261477422, "context": 0.3272969345034653} +{"step": 31, "steps": 250, "lr": 4.9603724091852456e-05, "grad_norm": 2.178178548812866, "loss_scale": 4.0, "retries": 0, "seconds": 16.7, "max_memory_gb": 18.96, "ce": 0.6041071526706219, "kl": 0.15398449823260307, "context": 0.6810994036495686} +{"step": 32, "steps": 250, "lr": 4.9545289670951025e-05, "grad_norm": 2.1632025241851807, "loss_scale": 4.0, "retries": 0, "seconds": 17.2, "max_memory_gb": 18.96, "ce": 0.3851197846233845, "kl": 0.017631475086091086, "context": 0.39393552392721176} +{"step": 33, "steps": 250, "lr": 4.948288171254089e-05, "grad_norm": 0.823045551776886, "loss_scale": 4.0, "retries": 0, "seconds": 17.7, "max_memory_gb": 18.96, "ce": 0.21530713699758053, "kl": 0.06403093424160033, "context": 0.2473226049914956} +{"step": 34, "steps": 250, "lr": 4.941651146529998e-05, "grad_norm": 0.9138999581336975, "loss_scale": 4.0, "retries": 0, "seconds": 18.2, "max_memory_gb": 18.96, "ce": 0.15053801928115718, "kl": 0.768636640906152, "context": 0.5348563246589038} +{"step": 35, "steps": 250, "lr": 4.9346190892086174e-05, "grad_norm": 0.9930412769317627, "loss_scale": 4.0, "retries": 0, "seconds": 18.7, "max_memory_gb": 18.96, "ce": 0.17215012093947735, "kl": 0.003027120965271024, "context": 0.17366368159127887} +{"step": 36, "steps": 250, "lr": 4.927193266778101e-05, "grad_norm": 0.7445875406265259, "loss_scale": 4.0, "retries": 0, "seconds": 19.2, "max_memory_gb": 18.96, "ce": 0.21996188519551652, "kl": 0.056002757817623205, "context": 0.24796326718933415} +{"step": 37, "steps": 250, "lr": 4.919375017700515e-05, "grad_norm": 0.8231956958770752, "loss_scale": 4.0, "retries": 0, "seconds": 19.7, "max_memory_gb": 18.96, "ce": 0.31887743176775984, "kl": 0.7800079201115295, "context": 0.708881382481195} +{"step": 38, "steps": 250, "lr": 4.911165751170588e-05, "grad_norm": 0.8759904503822327, "loss_scale": 4.0, "retries": 0, "seconds": 20.2, "max_memory_gb": 18.96, "ce": 0.13414412178099155, "kl": 0.5065826356876642, "context": 0.3874354362487793} +{"step": 39, "steps": 250, "lr": 4.902566946861708e-05, "grad_norm": 0.35606151819229126, "loss_scale": 4.0, "retries": 0, "seconds": 20.7, "max_memory_gb": 18.96, "ce": 0.17632671473256778, "kl": 0.012248083497979678, "context": 0.18245075378217734} +{"step": 40, "steps": 250, "lr": 4.89358015465922e-05, "grad_norm": 0.6511844992637634, "loss_scale": 4.0, "retries": 0, "seconds": 21.2, "max_memory_gb": 18.96, "ce": 0.08877622777072247, "kl": 0.012029766716295853, "context": 0.09479111022665165} +{"step": 41, "steps": 250, "lr": 4.884206994381078e-05, "grad_norm": 0.8662262558937073, "loss_scale": 4.0, "retries": 0, "seconds": 21.7, "max_memory_gb": 18.96, "ce": 0.40725398319773376, "kl": 0.12376161874271929, "context": 0.46913477778434753} +{"step": 42, "steps": 250, "lr": 4.874449155485866e-05, "grad_norm": 0.36961913108825684, "loss_scale": 4.0, "retries": 0, "seconds": 22.2, "max_memory_gb": 18.96, "ce": 0.03265774571627844, "kl": 0.0012094670764781768, "context": 0.033262478609685786} +{"step": 43, "steps": 250, "lr": 4.864308396768294e-05, "grad_norm": 0.653836190700531, "loss_scale": 4.0, "retries": 0, "seconds": 22.7, "max_memory_gb": 18.96, "ce": 0.0630049918036093, "kl": 0.0900785189223825, "context": 0.10804425216338132} +{"step": 44, "steps": 250, "lr": 4.853786546042184e-05, "grad_norm": 1.9988741874694824, "loss_scale": 4.0, "retries": 0, "seconds": 23.2, "max_memory_gb": 18.96, "ce": 0.6567837692273315, "kl": 1.406298268258979, "context": 1.3599329182470683} +{"step": 45, "steps": 250, "lr": 4.842885499811012e-05, "grad_norm": 0.7796401977539062, "loss_scale": 4.0, "retries": 0, "seconds": 23.7, "max_memory_gb": 18.96, "ce": 0.9091887520044111, "kl": 0.24430093786213547, "context": 1.0313391968375072} +{"step": 46, "steps": 250, "lr": 4.831607222926078e-05, "grad_norm": 4.715503692626953, "loss_scale": 4.0, "retries": 0, "seconds": 24.2, "max_memory_gb": 18.96, "ce": 0.1277937963604927, "kl": 0.2929844274185598, "context": 0.2742860112339258} +{"step": 47, "steps": 250, "lr": 4.8199537482323545e-05, "grad_norm": 0.527878999710083, "loss_scale": 4.0, "retries": 0, "seconds": 24.7, "max_memory_gb": 18.96, "ce": 0.04322355194017291, "kl": 0.05331542179919779, "context": 0.06988126377109438} +{"step": 48, "steps": 250, "lr": 4.807927176202073e-05, "grad_norm": 0.8197498917579651, "loss_scale": 4.0, "retries": 0, "seconds": 25.2, "max_memory_gb": 18.96, "ce": 0.12411393629736267, "kl": 0.009469441041801474, "context": 0.1288486560806632} +{"step": 49, "steps": 250, "lr": 4.795529674556129e-05, "grad_norm": 0.6398574709892273, "loss_scale": 4.0, "retries": 0, "seconds": 25.7, "max_memory_gb": 18.96, "ce": 0.06697774384883814, "kl": 0.37099982965503386, "context": 0.2524776642640063} +{"step": 50, "steps": 250, "lr": 4.782763477873361e-05, "grad_norm": 0.8370528817176819, "loss_scale": 4.0, "retries": 0, "seconds": 26.2, "max_memory_gb": 18.96, "ce": 0.1866778254516248, "kl": 0.008382773223274853, "context": 0.1908692189899739} +{"step": 51, "steps": 250, "lr": 4.769630887187782e-05, "grad_norm": 1.5910791158676147, "loss_scale": 4.0, "retries": 0, "seconds": 26.7, "max_memory_gb": 18.96, "ce": 0.6366300064655661, "kl": 0.021767326145891275, "context": 0.6475136970075255} +{"step": 52, "steps": 250, "lr": 4.756134269573832e-05, "grad_norm": 0.9754644632339478, "loss_scale": 4.0, "retries": 0, "seconds": 27.2, "max_memory_gb": 18.96, "ce": 0.18621333510964178, "kl": 0.2591856258222833, "context": 0.31580613972619176} +{"step": 53, "steps": 250, "lr": 4.7422760577197226e-05, "grad_norm": 0.4058437645435333, "loss_scale": 4.0, "retries": 0, "seconds": 27.7, "max_memory_gb": 18.96, "ce": 0.1761016531381756, "kl": 0.4517017435282469, "context": 0.40195252373814583} +{"step": 54, "steps": 250, "lr": 4.728058749488963e-05, "grad_norm": 0.9106970429420471, "loss_scale": 4.0, "retries": 0, "seconds": 28.2, "max_memory_gb": 18.96, "ce": 0.05733488214900717, "kl": 0.3693843043874949, "context": 0.24202703847549856} +{"step": 55, "steps": 250, "lr": 4.713484907470133e-05, "grad_norm": 0.054107196629047394, "loss_scale": 4.0, "retries": 0, "seconds": 28.7, "max_memory_gb": 18.96, "ce": 0.016388643824029714, "kl": 0.015054533443276341, "context": 0.02391591010382399} +{"step": 56, "steps": 250, "lr": 4.6985571585149876e-05, "grad_norm": 0.33712777495384216, "loss_scale": 4.0, "retries": 0, "seconds": 29.2, "max_memory_gb": 18.96, "ce": 0.04439991433173418, "kl": 0.30272785876877606, "context": 0.19576384127140045} +{"step": 57, "steps": 250, "lr": 4.6832781932649884e-05, "grad_norm": 0.16109968721866608, "loss_scale": 4.0, "retries": 0, "seconds": 29.7, "max_memory_gb": 18.96, "ce": 0.016743099025916308, "kl": 0.019652200004202314, "context": 0.026569199966615997} +{"step": 58, "steps": 250, "lr": 4.667650765666325e-05, "grad_norm": 0.30084240436553955, "loss_scale": 4.0, "retries": 0, "seconds": 30.2, "max_memory_gb": 18.96, "ce": 0.05128277537005488, "kl": 0.00283143303526856, "context": 0.05269849352771416} +{"step": 59, "steps": 250, "lr": 4.651677692473538e-05, "grad_norm": 0.1316792070865631, "loss_scale": 4.0, "retries": 0, "seconds": 30.7, "max_memory_gb": 18.96, "ce": 0.02450391925231088, "kl": 0.04635546883218922, "context": 0.04768165361019783} +{"step": 60, "steps": 250, "lr": 4.635361852741808e-05, "grad_norm": 0.1019187793135643, "loss_scale": 4.0, "retries": 0, "seconds": 31.2, "max_memory_gb": 18.96, "ce": 0.0720350428018719, "kl": 0.061443222453817725, "context": 0.10275665618246421} +{"step": 61, "steps": 250, "lr": 4.6187061873080264e-05, "grad_norm": 0.6655155420303345, "loss_scale": 4.0, "retries": 0, "seconds": 31.7, "max_memory_gb": 18.96, "ce": 0.043357341550290585, "kl": 0.9950835090130568, "context": 0.5408990979194641} +{"step": 62, "steps": 250, "lr": 4.601713698260727e-05, "grad_norm": 0.22032321989536285, "loss_scale": 4.0, "retries": 0, "seconds": 32.2, "max_memory_gb": 18.96, "ce": 0.04638195445295423, "kl": 0.009682625415734947, "context": 0.051223268295871094} +{"step": 63, "steps": 250, "lr": 4.5843874483989746e-05, "grad_norm": 0.14040741324424744, "loss_scale": 4.0, "retries": 0, "seconds": 32.7, "max_memory_gb": 18.96, "ce": 0.03994716424494982, "kl": 0.1237264961891924, "context": 0.10181041387841105} +{"step": 64, "steps": 250, "lr": 4.566730560680316e-05, "grad_norm": 0.1458318680524826, "loss_scale": 4.0, "retries": 0, "seconds": 33.2, "max_memory_gb": 18.96, "ce": 0.026001896942034364, "kl": 0.014457380457315594, "context": 0.0332305864430964} +{"step": 65, "steps": 250, "lr": 4.548746217657878e-05, "grad_norm": 0.5951931476593018, "loss_scale": 4.0, "retries": 0, "seconds": 33.7, "max_memory_gb": 18.96, "ce": 0.2044749795459211, "kl": 0.1739400397054851, "context": 0.2914449982345104} +{"step": 66, "steps": 250, "lr": 4.5304376609067396e-05, "grad_norm": 0.4100831151008606, "loss_scale": 4.0, "retries": 0, "seconds": 34.3, "max_memory_gb": 18.96, "ce": 0.1610533402120211, "kl": 0.07454902027794219, "context": 0.1983278503364545} +{"step": 67, "steps": 250, "lr": 4.5118081904396504e-05, "grad_norm": 0.4895205497741699, "loss_scale": 4.0, "retries": 0, "seconds": 34.8, "max_memory_gb": 18.96, "ce": 0.04390105959464563, "kl": 0.04947941959835589, "context": 0.0686407700413838} +{"step": 68, "steps": 250, "lr": 4.4928611641122225e-05, "grad_norm": 0.20726370811462402, "loss_scale": 4.0, "retries": 0, "seconds": 35.3, "max_memory_gb": 18.96, "ce": 0.01710079710392165, "kl": 0.054022367559809936, "context": 0.04411198027264618} +{"step": 69, "steps": 250, "lr": 4.473599997017701e-05, "grad_norm": 0.5213683843612671, "loss_scale": 4.0, "retries": 0, "seconds": 35.8, "max_memory_gb": 18.96, "ce": 0.2848783084191382, "kl": 0.04515450727194548, "context": 0.3074555713683367} +{"step": 70, "steps": 250, "lr": 4.4540281608714053e-05, "grad_norm": 0.20511581003665924, "loss_scale": 4.0, "retries": 0, "seconds": 36.3, "max_memory_gb": 18.96, "ce": 0.014576786779798567, "kl": 0.001354803407593863, "context": 0.015254188503604382} +{"step": 71, "steps": 250, "lr": 4.434149183384977e-05, "grad_norm": 0.4987012445926666, "loss_scale": 4.0, "retries": 0, "seconds": 36.8, "max_memory_gb": 18.96, "ce": 0.04872734340278839, "kl": 0.016553360444959253, "context": 0.05700402316142572} +{"step": 72, "steps": 250, "lr": 4.413966647630534e-05, "grad_norm": 0.29828161001205444, "loss_scale": 4.0, "retries": 0, "seconds": 37.3, "max_memory_gb": 18.96, "ce": 0.03794885822571814, "kl": 0.43808947689831257, "context": 0.25699360109865665} +{"step": 73, "steps": 250, "lr": 4.3934841913948323e-05, "grad_norm": 1.4907500743865967, "loss_scale": 4.0, "retries": 0, "seconds": 37.8, "max_memory_gb": 18.96, "ce": 0.4175184343002911, "kl": 0.15271727762228693, "context": 0.4938770731241675} +{"step": 74, "steps": 250, "lr": 4.372705506523583e-05, "grad_norm": 0.6462470889091492, "loss_scale": 4.0, "retries": 0, "seconds": 38.3, "max_memory_gb": 18.96, "ce": 0.10638844376308043, "kl": 0.12937032300396822, "context": 0.17107360503359814} +{"step": 75, "steps": 250, "lr": 4.351634338256017e-05, "grad_norm": 0.8233071565628052, "loss_scale": 4.0, "retries": 0, "seconds": 38.8, "max_memory_gb": 18.96, "ce": 0.4431042771029752, "kl": 1.0514792064204812, "context": 0.9688438987359405} +{"step": 76, "steps": 250, "lr": 4.330274484549825e-05, "grad_norm": 0.43098267912864685, "loss_scale": 4.0, "retries": 0, "seconds": 39.3, "max_memory_gb": 18.96, "ce": 0.021766084362752736, "kl": 1.8231912070768885, "context": 0.9333617084193975} +{"step": 77, "steps": 250, "lr": 4.308629795396599e-05, "grad_norm": 0.056420572102069855, "loss_scale": 4.0, "retries": 0, "seconds": 39.8, "max_memory_gb": 18.96, "ce": 0.009407816745806485, "kl": 0.013229883918029373, "context": 0.016022759169572964} +{"step": 78, "steps": 250, "lr": 4.286704172127894e-05, "grad_norm": 0.13427695631980896, "loss_scale": 4.0, "retries": 0, "seconds": 40.3, "max_memory_gb": 18.96, "ce": 0.0035367312957532704, "kl": 0.0742490008706227, "context": 0.04066123062511906} +{"step": 79, "steps": 250, "lr": 4.264501566712033e-05, "grad_norm": 0.5610014200210571, "loss_scale": 4.0, "retries": 0, "seconds": 40.8, "max_memory_gb": 18.96, "ce": 0.05136837621466839, "kl": 0.06963474837357353, "context": 0.08618575041600707} +{"step": 80, "steps": 250, "lr": 4.2420259810417894e-05, "grad_norm": 0.18953104317188263, "loss_scale": 4.0, "retries": 0, "seconds": 41.3, "max_memory_gb": 18.96, "ce": 0.011239749489504902, "kl": 0.08980649053273737, "context": 0.056142995454138145} +{"step": 81, "steps": 250, "lr": 4.219281466213066e-05, "grad_norm": 0.5339422821998596, "loss_scale": 4.0, "retries": 0, "seconds": 41.8, "max_memory_gb": 18.96, "ce": 0.11582899279892445, "kl": 0.09279448608867824, "context": 0.16222624038346112} +{"step": 82, "steps": 250, "lr": 4.1962721217947135e-05, "grad_norm": 0.30350780487060547, "loss_scale": 4.0, "retries": 0, "seconds": 42.3, "max_memory_gb": 18.96, "ce": 0.03838304668897763, "kl": 0.010404451764770783, "context": 0.04358527215663344} +{"step": 83, "steps": 250, "lr": 4.173002095089599e-05, "grad_norm": 0.3309577703475952, "loss_scale": 4.0, "retries": 0, "seconds": 42.8, "max_memory_gb": 18.96, "ce": 0.17120755137875676, "kl": 0.8258629122865386, "context": 0.5841390072600916} +{"step": 84, "steps": 250, "lr": 4.149475580387084e-05, "grad_norm": 0.22325891256332397, "loss_scale": 4.0, "retries": 0, "seconds": 43.3, "max_memory_gb": 18.96, "ce": 0.021560225426128454, "kl": 0.05866225290355942, "context": 0.050891352274447854} +{"step": 85, "steps": 250, "lr": 4.1256968182070275e-05, "grad_norm": 2.0988142490386963, "loss_scale": 4.0, "retries": 0, "seconds": 43.8, "max_memory_gb": 18.96, "ce": 0.09791610882223267, "kl": 0.08249838096307371, "context": 0.13916529930486377} +{"step": 86, "steps": 250, "lr": 4.101670094535453e-05, "grad_norm": 1.6013784408569336, "loss_scale": 4.0, "retries": 0, "seconds": 44.3, "max_memory_gb": 18.96, "ce": 0.26567687664646655, "kl": 0.5634722396498546, "context": 0.5474129817448556} +{"step": 87, "steps": 250, "lr": 4.077399740052027e-05, "grad_norm": 0.78244948387146, "loss_scale": 4.0, "retries": 0, "seconds": 44.8, "max_memory_gb": 18.96, "ce": 0.01142694233567454, "kl": 0.7967089731246233, "context": 0.4097814252600074} +{"step": 88, "steps": 250, "lr": 4.0528901293494775e-05, "grad_norm": 0.08671758323907852, "loss_scale": 4.0, "retries": 0, "seconds": 45.3, "max_memory_gb": 18.96, "ce": 0.007394502040824591, "kl": 0.033042527527868515, "context": 0.02391576668514972} +{"step": 89, "steps": 250, "lr": 4.0281456801451004e-05, "grad_norm": 0.10365752130746841, "loss_scale": 4.0, "retries": 0, "seconds": 45.8, "max_memory_gb": 18.96, "ce": 0.010549725739792848, "kl": 0.00832796507893363, "context": 0.014713708106683043} +{"step": 90, "steps": 250, "lr": 4.0031708524844884e-05, "grad_norm": 0.20420098304748535, "loss_scale": 4.0, "retries": 0, "seconds": 46.3, "max_memory_gb": 18.96, "ce": 0.022955902233661618, "kl": 0.012992301140911877, "context": 0.029452052694978192} +{"step": 91, "steps": 250, "lr": 3.977970147937635e-05, "grad_norm": 0.4648401141166687, "loss_scale": 4.0, "retries": 0, "seconds": 46.8, "max_memory_gb": 18.96, "ce": 0.17000147222279338, "kl": 0.6714539211006922, "context": 0.5057284409231215} +{"step": 92, "steps": 250, "lr": 3.952548108787553e-05, "grad_norm": 1.5578516721725464, "loss_scale": 4.0, "retries": 0, "seconds": 47.4, "max_memory_gb": 18.96, "ce": 0.4453744730872131, "kl": 0.5221160057762972, "context": 0.7064324468356062} +{"step": 93, "steps": 250, "lr": 3.9269093172115496e-05, "grad_norm": 1.2088490724563599, "loss_scale": 4.0, "retries": 0, "seconds": 47.9, "max_memory_gb": 18.96, "ce": 0.0703731972607784, "kl": 0.5181256934301928, "context": 0.3294360414147377} +{"step": 94, "steps": 250, "lr": 3.901058394455323e-05, "grad_norm": 1.6625189781188965, "loss_scale": 4.0, "retries": 0, "seconds": 48.4, "max_memory_gb": 18.96, "ce": 0.32796623674221337, "kl": 0.15993681442341767, "context": 0.40793465985916555} +{"step": 95, "steps": 250, "lr": 3.875e-05, "grad_norm": 1.8752552270889282, "loss_scale": 4.0, "retries": 0, "seconds": 48.9, "max_memory_gb": 18.96, "ce": 0.6420130477054045, "kl": 0.029452618095092475, "context": 0.6567393448203802} +{"step": 96, "steps": 250, "lr": 3.8487388307223005e-05, "grad_norm": 0.7073447704315186, "loss_scale": 4.0, "retries": 0, "seconds": 49.4, "max_memory_gb": 18.96, "ce": 0.534109107684344, "kl": 0.00568621646380052, "context": 0.5369522353867069} +{"step": 97, "steps": 250, "lr": 3.822279620047943e-05, "grad_norm": 2.61953067779541, "loss_scale": 4.0, "retries": 0, "seconds": 49.9, "max_memory_gb": 18.96, "ce": 0.02887815331632737, "kl": 0.4844901212491095, "context": 0.2711232090368867} +{"step": 98, "steps": 250, "lr": 3.795627137098479e-05, "grad_norm": 0.08754952251911163, "loss_scale": 4.0, "retries": 0, "seconds": 50.4, "max_memory_gb": 18.96, "ce": 0.0049501672610858805, "kl": 0.0988186548856902, "context": 0.05435949303864618} +{"step": 99, "steps": 250, "lr": 3.7687861858316844e-05, "grad_norm": 0.3461315333843231, "loss_scale": 4.0, "retries": 0, "seconds": 50.9, "max_memory_gb": 18.96, "ce": 0.03453727294208875, "kl": 0.013974352174045634, "context": 0.041524449679855024} +{"step": 100, "steps": 250, "lr": 3.741761604175668e-05, "grad_norm": 0.6663182973861694, "loss_scale": 4.0, "retries": 0, "seconds": 51.4, "max_memory_gb": 18.96, "ce": 0.028162088012322783, "kl": 0.8004364707667264, "context": 0.42838032252620906} +{"step": 101, "steps": 250, "lr": 3.714558263156872e-05, "grad_norm": 0.28902667760849, "loss_scale": 4.0, "retries": 0, "seconds": 51.9, "max_memory_gb": 18.96, "ce": 0.006148882530396804, "kl": 0.0995752906310372, "context": 0.055936527147423476} +{"step": 102, "steps": 250, "lr": 3.6871810660220904e-05, "grad_norm": 0.11631490290164948, "loss_scale": 4.0, "retries": 0, "seconds": 52.4, "max_memory_gb": 18.96, "ce": 0.04690119909355417, "kl": 0.039179587038233876, "context": 0.06649099406786263} +{"step": 103, "steps": 250, "lr": 3.659634947354686e-05, "grad_norm": 0.1967485398054123, "loss_scale": 4.0, "retries": 0, "seconds": 52.9, "max_memory_gb": 18.96, "ce": 0.019680983417856623, "kl": 0.02022741342352674, "context": 0.02979468982448452} +{"step": 104, "steps": 250, "lr": 3.631924872185169e-05, "grad_norm": 0.046305056661367416, "loss_scale": 4.0, "retries": 0, "seconds": 53.4, "max_memory_gb": 18.96, "ce": 0.003638441958173644, "kl": 0.0056495059111227874, "context": 0.006463194695243146} +{"step": 105, "steps": 250, "lr": 3.6040558350962674e-05, "grad_norm": 0.21142996847629547, "loss_scale": 4.0, "retries": 0, "seconds": 54.0, "max_memory_gb": 18.96, "ce": 0.06746011739596725, "kl": 0.38595246884506196, "context": 0.2604363556019962} +{"step": 106, "steps": 250, "lr": 3.576032859322687e-05, "grad_norm": 0.16882707178592682, "loss_scale": 4.0, "retries": 0, "seconds": 54.5, "max_memory_gb": 18.96, "ce": 0.009285319996706676, "kl": 0.009808907114347676, "context": 0.014189773464750033} +{"step": 107, "steps": 250, "lr": 3.5478609958457056e-05, "grad_norm": 1.6739987134933472, "loss_scale": 4.0, "retries": 0, "seconds": 55.0, "max_memory_gb": 18.96, "ce": 0.8426965724029287, "kl": 0.08854798271886466, "context": 0.8869705698743928} +{"step": 108, "steps": 250, "lr": 3.519545322482755e-05, "grad_norm": 0.17834429442882538, "loss_scale": 4.0, "retries": 0, "seconds": 55.5, "max_memory_gb": 18.96, "ce": 0.0137188092667202, "kl": 0.002809380345581758, "context": 0.015123499149922282} +{"step": 109, "steps": 250, "lr": 3.491090942972179e-05, "grad_norm": 0.3872220516204834, "loss_scale": 4.0, "retries": 0, "seconds": 56.0, "max_memory_gb": 18.96, "ce": 0.046501971752149984, "kl": 0.38637590780854225, "context": 0.23968991823494434} +{"step": 110, "steps": 250, "lr": 3.4625029860533135e-05, "grad_norm": 2.0720765590667725, "loss_scale": 4.0, "retries": 0, "seconds": 56.5, "max_memory_gb": 18.96, "ce": 0.32240709103643894, "kl": 0.33162558800540864, "context": 0.4882198730483651} +{"step": 111, "steps": 250, "lr": 3.433786604542056e-05, "grad_norm": 0.01806967705488205, "loss_scale": 4.0, "retries": 0, "seconds": 57.0, "max_memory_gb": 18.96, "ce": 0.0014339681265482795, "kl": 0.0018688087229747907, "context": 0.0023683724357397296} +{"step": 112, "steps": 250, "lr": 3.404946974402104e-05, "grad_norm": 0.025271175429224968, "loss_scale": 4.0, "retries": 0, "seconds": 57.6, "max_memory_gb": 18.96, "ce": 0.0034598659549374133, "kl": 0.0005149464554961014, "context": 0.003717339161084965} +{"step": 113, "steps": 250, "lr": 3.375989293812019e-05, "grad_norm": 0.6220430135726929, "loss_scale": 4.0, "retries": 0, "seconds": 58.1, "max_memory_gb": 18.96, "ce": 0.008134944480843842, "kl": 0.5000790535705164, "context": 0.2581744643393904} +{"step": 114, "steps": 250, "lr": 3.3469187822282766e-05, "grad_norm": 1.2349562644958496, "loss_scale": 4.0, "retries": 0, "seconds": 58.6, "max_memory_gb": 18.96, "ce": 0.016232651192694902, "kl": 1.240225177258253, "context": 0.6363452402874827} +{"step": 115, "steps": 250, "lr": 3.317740679444499e-05, "grad_norm": 0.19357512891292572, "loss_scale": 4.0, "retries": 0, "seconds": 59.1, "max_memory_gb": 18.96, "ce": 0.01220725635994313, "kl": 0.054612674117379356, "context": 0.03951359330312698} +{"step": 116, "steps": 250, "lr": 3.2884602446470056e-05, "grad_norm": 1.244156002998352, "loss_scale": 4.0, "retries": 0, "seconds": 59.6, "max_memory_gb": 18.96, "ce": 0.22801323342901014, "kl": 0.6681981122463299, "context": 0.562112319355947} +{"step": 117, "steps": 250, "lr": 3.259082755466876e-05, "grad_norm": 2.222731828689575, "loss_scale": 4.0, "retries": 0, "seconds": 60.1, "max_memory_gb": 18.97, "ce": 0.17652968525726465, "kl": 0.29329155880986946, "context": 0.32317546463309554} +{"step": 118, "steps": 250, "lr": 3.229613507028686e-05, "grad_norm": 0.3917120397090912, "loss_scale": 4.0, "retries": 0, "seconds": 60.6, "max_memory_gb": 18.97, "ce": 0.18515902590843325, "kl": 0.01907473486392064, "context": 0.19469639776070835} +{"step": 119, "steps": 250, "lr": 3.2000578109961004e-05, "grad_norm": 0.23290924727916718, "loss_scale": 4.0, "retries": 0, "seconds": 61.1, "max_memory_gb": 18.97, "ce": 0.014112862441834295, "kl": 0.017130525485612225, "context": 0.022678125471429667} +{"step": 120, "steps": 250, "lr": 3.1704209946144616e-05, "grad_norm": 1.1884337663650513, "loss_scale": 4.0, "retries": 0, "seconds": 61.6, "max_memory_gb": 18.97, "ce": 0.8004684394982178, "kl": 0.20221318372932728, "context": 0.9015750506951008} +{"step": 121, "steps": 250, "lr": 3.140708399750594e-05, "grad_norm": 1.1067469120025635, "loss_scale": 4.0, "retries": 0, "seconds": 62.1, "max_memory_gb": 18.97, "ce": 0.055410970862794784, "kl": 0.049679361924063414, "context": 0.08025065185393032} +{"step": 122, "steps": 250, "lr": 3.1109253819299606e-05, "grad_norm": 0.11171060055494308, "loss_scale": 4.0, "retries": 0, "seconds": 62.6, "max_memory_gb": 18.97, "ce": 0.03568339610501425, "kl": 0.01274519205071556, "context": 0.04205599157648976} +{"step": 123, "steps": 250, "lr": 3.081077309371361e-05, "grad_norm": 0.2115929275751114, "loss_scale": 4.0, "retries": 0, "seconds": 63.1, "max_memory_gb": 18.97, "ce": 0.029075018013827503, "kl": 0.025210241845343262, "context": 0.04168013937305659} +{"step": 124, "steps": 250, "lr": 3.051169562019339e-05, "grad_norm": 1.367719054222107, "loss_scale": 4.0, "retries": 0, "seconds": 63.6, "max_memory_gb": 18.97, "ce": 0.7136793443060014, "kl": 0.7366017602616921, "context": 1.0819802535697818} +{"step": 125, "steps": 250, "lr": 3.021207530574477e-05, "grad_norm": 1.0015259981155396, "loss_scale": 4.0, "retries": 0, "seconds": 64.1, "max_memory_gb": 18.97, "ce": 0.11642471965751611, "kl": 0.13906041852897033, "context": 0.18595492333406582} +{"step": 126, "steps": 250, "lr": 2.991196615521758e-05, "grad_norm": 1.2558355331420898, "loss_scale": 4.0, "retries": 0, "seconds": 64.6, "max_memory_gb": 18.97, "ce": 0.4884591866284609, "kl": 0.15004092128947377, "context": 0.5634796358644962} +{"step": 127, "steps": 250, "lr": 2.9611422261571536e-05, "grad_norm": 0.049282222986221313, "loss_scale": 4.0, "retries": 0, "seconds": 65.1, "max_memory_gb": 18.97, "ce": 0.012633966615794634, "kl": 0.0015113023400772363, "context": 0.013389617790380726} +{"step": 128, "steps": 250, "lr": 2.931049779612634e-05, "grad_norm": 1.8383291959762573, "loss_scale": 4.0, "retries": 0, "seconds": 65.6, "max_memory_gb": 18.97, "ce": 0.47038151627930347, "kl": 1.6573186521563912, "context": 1.2990408520272467} +{"step": 129, "steps": 250, "lr": 2.9009246998797557e-05, "grad_norm": 0.18006093800067902, "loss_scale": 4.0, "retries": 0, "seconds": 66.2, "max_memory_gb": 18.97, "ce": 0.08355676825158298, "kl": 0.0368889095261693, "context": 0.10200122557580471} +{"step": 130, "steps": 250, "lr": 2.8707724168320238e-05, "grad_norm": 0.14378507435321808, "loss_scale": 4.0, "retries": 0, "seconds": 66.7, "max_memory_gb": 18.97, "ce": 0.04114120258600451, "kl": 0.31942464257190295, "context": 0.20085352833848447} +{"step": 131, "steps": 250, "lr": 2.840598365246184e-05, "grad_norm": 2.077575445175171, "loss_scale": 4.0, "retries": 0, "seconds": 67.1, "max_memory_gb": 18.97, "ce": 0.13041764221998164, "kl": 0.30488520092330873, "context": 0.28286024287808686} +{"step": 132, "steps": 250, "lr": 2.8104079838226383e-05, "grad_norm": 0.2610035538673401, "loss_scale": 4.0, "retries": 0, "seconds": 67.7, "max_memory_gb": 18.97, "ce": 0.18431325217170524, "kl": 0.03508622491699498, "context": 0.20185636113819783} +{"step": 133, "steps": 250, "lr": 2.7802067142051446e-05, "grad_norm": 0.48632341623306274, "loss_scale": 4.0, "retries": 0, "seconds": 68.2, "max_memory_gb": 18.97, "ce": 0.34639231563778594, "kl": 0.20117653533816338, "context": 0.44698057882487774} +{"step": 134, "steps": 250, "lr": 2.7500000000000004e-05, "grad_norm": 0.74669349193573, "loss_scale": 4.0, "retries": 0, "seconds": 68.7, "max_memory_gb": 18.97, "ce": 0.24432065858673013, "kl": 0.005657983643043707, "context": 0.24714965692328406} +{"step": 135, "steps": 250, "lr": 2.7197932857948566e-05, "grad_norm": 0.12908843159675598, "loss_scale": 4.0, "retries": 0, "seconds": 69.2, "max_memory_gb": 18.97, "ce": 0.03785982300178148, "kl": 0.09848007370601408, "context": 0.08709985774476081} +{"step": 136, "steps": 250, "lr": 2.689592016177363e-05, "grad_norm": 4.125553131103516, "loss_scale": 4.0, "retries": 0, "seconds": 69.6, "max_memory_gb": 18.97, "ce": 0.20316723825817462, "kl": 0.27714152506086975, "context": 0.34173798582924064} +{"step": 137, "steps": 250, "lr": 2.659401634753817e-05, "grad_norm": 0.06894055753946304, "loss_scale": 4.0, "retries": 0, "seconds": 70.2, "max_memory_gb": 18.97, "ce": 0.008769909807597287, "kl": 0.007171174991526641, "context": 0.012355497688986361} +{"step": 138, "steps": 250, "lr": 2.6292275831679764e-05, "grad_norm": 3.325916051864624, "loss_scale": 4.0, "retries": 0, "seconds": 70.7, "max_memory_gb": 18.97, "ce": 0.11350376787595451, "kl": 0.2521299282088876, "context": 0.23956872429698706} +{"step": 139, "steps": 250, "lr": 2.5990753001202445e-05, "grad_norm": 0.7467379570007324, "loss_scale": 4.0, "retries": 0, "seconds": 71.2, "max_memory_gb": 18.97, "ce": 0.0034422428943798877, "kl": 0.4290576507046353, "context": 0.2179710645796149} +{"step": 140, "steps": 250, "lr": 2.5689502203873672e-05, "grad_norm": 0.042760975658893585, "loss_scale": 4.0, "retries": 0, "seconds": 71.7, "max_memory_gb": 18.97, "ce": 0.005663720869051758, "kl": 0.0054394764883909374, "context": 0.008383459062315524} +{"step": 141, "steps": 250, "lr": 2.538857773842847e-05, "grad_norm": 0.014881729148328304, "loss_scale": 4.0, "retries": 0, "seconds": 72.2, "max_memory_gb": 18.97, "ce": 0.003131317724182736, "kl": 0.005968341569769109, "context": 0.006115488562500104} +{"step": 142, "steps": 250, "lr": 2.5088033844782426e-05, "grad_norm": 0.9214199185371399, "loss_scale": 4.0, "retries": 0, "seconds": 72.7, "max_memory_gb": 18.97, "ce": 0.05730243215839437, "kl": 0.04072082648053765, "context": 0.07766284584067762} +{"step": 143, "steps": 250, "lr": 2.4787924694255237e-05, "grad_norm": 1.2487913370132446, "loss_scale": 4.0, "retries": 0, "seconds": 73.2, "max_memory_gb": 18.97, "ce": 0.7206597110762232, "kl": 0.05686473673085857, "context": 0.7490920969030412} +{"step": 144, "steps": 250, "lr": 2.4488304379806627e-05, "grad_norm": 0.6882110238075256, "loss_scale": 4.0, "retries": 0, "seconds": 73.7, "max_memory_gb": 18.97, "ce": 0.09586014179512858, "kl": 0.23392554430756718, "context": 0.21282292273826897} +{"step": 145, "steps": 250, "lr": 2.4189226906286388e-05, "grad_norm": 0.05322660505771637, "loss_scale": 4.0, "retries": 0, "seconds": 74.3, "max_memory_gb": 18.97, "ce": 0.0034693200141191483, "kl": 0.011830216032251428, "context": 0.00938442822007346} +{"step": 146, "steps": 250, "lr": 2.3890746180700395e-05, "grad_norm": 0.06664061546325684, "loss_scale": 4.0, "retries": 0, "seconds": 74.8, "max_memory_gb": 18.97, "ce": 0.0067826874437741935, "kl": 0.011876699116328382, "context": 0.012721037084702402} +{"step": 147, "steps": 250, "lr": 2.359291600249407e-05, "grad_norm": 0.6190969944000244, "loss_scale": 4.0, "retries": 0, "seconds": 75.3, "max_memory_gb": 18.97, "ce": 0.4830112859453948, "kl": 0.6172582402141416, "context": 0.7916404074494494} +{"step": 148, "steps": 250, "lr": 2.3295790053855392e-05, "grad_norm": 0.05152052640914917, "loss_scale": 4.0, "retries": 0, "seconds": 75.8, "max_memory_gb": 18.97, "ce": 0.004413216029433897, "kl": 0.008239653379860101, "context": 0.00853304282463796} +{"step": 149, "steps": 250, "lr": 2.299942189003901e-05, "grad_norm": 0.2330605387687683, "loss_scale": 4.0, "retries": 0, "seconds": 76.3, "max_memory_gb": 18.97, "ce": 0.030890114372596145, "kl": 0.01181194577657152, "context": 0.03679608725360595} +{"step": 150, "steps": 250, "lr": 2.2703864929713144e-05, "grad_norm": 0.17089878022670746, "loss_scale": 4.0, "retries": 0, "seconds": 76.8, "max_memory_gb": 18.97, "ce": 0.050796293071471155, "kl": 0.03019681735895574, "context": 0.0658947005867958} +{"step": 151, "steps": 250, "lr": 2.2409172445331257e-05, "grad_norm": 0.09490539878606796, "loss_scale": 4.0, "retries": 0, "seconds": 77.3, "max_memory_gb": 18.97, "ce": 0.010387602242190042, "kl": 0.016476459031252944, "context": 0.018625831757162814} +{"step": 152, "steps": 250, "lr": 2.211539755352996e-05, "grad_norm": 0.2693023383617401, "loss_scale": 4.0, "retries": 0, "seconds": 77.8, "max_memory_gb": 18.97, "ce": 0.01531800904103875, "kl": 0.7253150598451867, "context": 0.3779755333753201} +{"step": 153, "steps": 250, "lr": 2.1822593205555016e-05, "grad_norm": 0.2959217131137848, "loss_scale": 4.0, "retries": 0, "seconds": 78.3, "max_memory_gb": 18.97, "ce": 0.010297489092408796, "kl": 0.4974973120974937, "context": 0.2590461394365775} +{"step": 154, "steps": 250, "lr": 2.1530812177717236e-05, "grad_norm": 2.8318052291870117, "loss_scale": 4.0, "retries": 0, "seconds": 78.8, "max_memory_gb": 18.97, "ce": 0.12412031344138086, "kl": 0.1376490427646786, "context": 0.19294483866542578} +{"step": 155, "steps": 250, "lr": 2.1240107061879815e-05, "grad_norm": 0.2852438688278198, "loss_scale": 4.0, "retries": 0, "seconds": 79.3, "max_memory_gb": 18.97, "ce": 0.20988452534356838, "kl": 0.011278326216142887, "context": 0.21552368473021488} +{"step": 156, "steps": 250, "lr": 2.095053025597896e-05, "grad_norm": 4.3682074546813965, "loss_scale": 4.0, "retries": 0, "seconds": 79.8, "max_memory_gb": 18.97, "ce": 0.011769015895424673, "kl": 0.4405403404152821, "context": 0.23203918860053818} +{"step": 157, "steps": 250, "lr": 2.066213395457945e-05, "grad_norm": 1.9768022298812866, "loss_scale": 4.0, "retries": 0, "seconds": 80.3, "max_memory_gb": 18.97, "ce": 0.1435489461873658, "kl": 0.10951289893273497, "context": 0.19830539188114926} +{"step": 158, "steps": 250, "lr": 2.0374970139466877e-05, "grad_norm": 0.335369348526001, "loss_scale": 4.0, "retries": 0, "seconds": 80.8, "max_memory_gb": 18.97, "ce": 0.16255593898040388, "kl": 0.0873932357353624, "context": 0.20625255120830843} +{"step": 159, "steps": 250, "lr": 2.0089090570278213e-05, "grad_norm": 0.3600740134716034, "loss_scale": 4.0, "retries": 0, "seconds": 81.4, "max_memory_gb": 18.98, "ce": 0.0032806728304421995, "kl": 0.04949203215073794, "context": 0.028026688465615734} +{"step": 160, "steps": 250, "lr": 1.9804546775172457e-05, "grad_norm": 0.8391955494880676, "loss_scale": 4.0, "retries": 0, "seconds": 81.9, "max_memory_gb": 18.98, "ce": 0.01995205710409209, "kl": 0.5038218500558287, "context": 0.2718629811424762} +{"step": 161, "steps": 250, "lr": 1.9521390041542946e-05, "grad_norm": 1.2608169317245483, "loss_scale": 4.0, "retries": 0, "seconds": 82.4, "max_memory_gb": 18.98, "ce": 0.09343571961653652, "kl": 0.11857361241595754, "context": 0.15272252769500483} +{"step": 162, "steps": 250, "lr": 1.923967140677313e-05, "grad_norm": 1.1041406393051147, "loss_scale": 4.0, "retries": 0, "seconds": 82.8, "max_memory_gb": 18.98, "ce": 0.15397512686104164, "kl": 0.29293865807994734, "context": 0.30044445589737734} +{"step": 163, "steps": 250, "lr": 1.8959441649037335e-05, "grad_norm": 1.6480916738510132, "loss_scale": 4.0, "retries": 0, "seconds": 83.3, "max_memory_gb": 18.98, "ce": 0.2856952587608248, "kl": 0.953682441264391, "context": 0.7625364817213267} +{"step": 164, "steps": 250, "lr": 1.868075127814832e-05, "grad_norm": 5.773058891296387, "loss_scale": 4.0, "retries": 0, "seconds": 83.9, "max_memory_gb": 18.98, "ce": 0.1491083505388815, "kl": 0.27798689348855987, "context": 0.2881017976324074} +{"step": 165, "steps": 250, "lr": 1.8403650526453143e-05, "grad_norm": 0.05439922958612442, "loss_scale": 4.0, "retries": 0, "seconds": 84.4, "max_memory_gb": 18.98, "ce": 0.0022571998515559244, "kl": 0.01891255639930023, "context": 0.011713477804732975} +{"step": 166, "steps": 250, "lr": 1.812818933977912e-05, "grad_norm": 1.7474507093429565, "loss_scale": 4.0, "retries": 0, "seconds": 84.9, "max_memory_gb": 18.98, "ce": 0.2565801796299638, "kl": 0.13496075454168022, "context": 0.3240605624887394} +{"step": 167, "steps": 250, "lr": 1.785441736843128e-05, "grad_norm": 0.1972493678331375, "loss_scale": 4.0, "retries": 0, "seconds": 85.4, "max_memory_gb": 18.98, "ce": 0.0354742988711223, "kl": 0.005083370189822745, "context": 0.038015984115190804} +{"step": 168, "steps": 250, "lr": 1.7582383958243314e-05, "grad_norm": 0.11330270022153854, "loss_scale": 4.0, "retries": 0, "seconds": 85.9, "max_memory_gb": 18.98, "ce": 0.03145776747260243, "kl": 0.004042641943669878, "context": 0.033479088931926526} +{"step": 169, "steps": 250, "lr": 1.7312138141683164e-05, "grad_norm": 0.9190447926521301, "loss_scale": 4.0, "retries": 0, "seconds": 86.4, "max_memory_gb": 18.98, "ce": 0.1113847597152926, "kl": 1.6336406469345093, "context": 0.9282051101326942} +{"step": 170, "steps": 250, "lr": 1.704372862901521e-05, "grad_norm": 0.11079658567905426, "loss_scale": 4.0, "retries": 0, "seconds": 86.9, "max_memory_gb": 18.98, "ce": 0.004615597268639249, "kl": 0.0032201120775425807, "context": 0.006225653298315592} +{"step": 171, "steps": 250, "lr": 1.6777203799520573e-05, "grad_norm": 0.510127067565918, "loss_scale": 4.0, "retries": 0, "seconds": 87.5, "max_memory_gb": 18.98, "ce": 0.03507966661209139, "kl": 0.31855554642970674, "context": 0.19435744192742277} +{"step": 172, "steps": 250, "lr": 1.6512611692777008e-05, "grad_norm": 0.08791660517454147, "loss_scale": 4.0, "retries": 0, "seconds": 88.0, "max_memory_gb": 18.98, "ce": 0.01645446929614991, "kl": 0.01520232018083334, "context": 0.024055629735812545} +{"step": 173, "steps": 250, "lr": 1.6250000000000005e-05, "grad_norm": 0.15108653903007507, "loss_scale": 4.0, "retries": 0, "seconds": 88.5, "max_memory_gb": 18.98, "ce": 0.05284354658215307, "kl": 0.03461645451898221, "context": 0.07015177491120994} +{"step": 174, "steps": 250, "lr": 1.598941605544678e-05, "grad_norm": 0.23246392607688904, "loss_scale": 4.0, "retries": 0, "seconds": 89.0, "max_memory_gb": 18.98, "ce": 0.05626359941197734, "kl": 0.030253342460639487, "context": 0.07139027250195795} +{"step": 175, "steps": 250, "lr": 1.5730906827884506e-05, "grad_norm": 2.5122969150543213, "loss_scale": 4.0, "retries": 0, "seconds": 89.6, "max_memory_gb": 18.98, "ce": 0.6937809803057462, "kl": 0.10485951835289598, "context": 0.7462107352912426} +{"step": 176, "steps": 250, "lr": 1.5474518912124476e-05, "grad_norm": 0.08984461426734924, "loss_scale": 4.0, "retries": 0, "seconds": 90.1, "max_memory_gb": 18.98, "ce": 0.009840656304731965, "kl": 0.002888279983380926, "context": 0.011284796230029315} +{"step": 177, "steps": 250, "lr": 1.5220298520623646e-05, "grad_norm": 0.37388190627098083, "loss_scale": 4.0, "retries": 0, "seconds": 90.6, "max_memory_gb": 18.98, "ce": 0.008172553854819853, "kl": 0.9434828018347616, "context": 0.479913941497216} +{"step": 178, "steps": 250, "lr": 1.4968291475155121e-05, "grad_norm": 0.9274818897247314, "loss_scale": 4.0, "retries": 0, "seconds": 91.1, "max_memory_gb": 18.98, "ce": 0.03602658032468753, "kl": 0.2058732916120789, "context": 0.13896322798973415} +{"step": 179, "steps": 250, "lr": 1.4718543198548998e-05, "grad_norm": 0.6005780696868896, "loss_scale": 4.0, "retries": 0, "seconds": 91.6, "max_memory_gb": 18.98, "ce": 0.3447389024331642, "kl": 0.039891827409519465, "context": 0.3646848067837709} +{"step": 180, "steps": 250, "lr": 1.4471098706505227e-05, "grad_norm": 0.08659005910158157, "loss_scale": 4.0, "retries": 0, "seconds": 92.1, "max_memory_gb": 18.98, "ce": 0.030035919276997447, "kl": 0.021924684406258166, "context": 0.04099826142191887} +{"step": 181, "steps": 250, "lr": 1.4226002599479743e-05, "grad_norm": 0.15513546764850616, "loss_scale": 4.0, "retries": 0, "seconds": 92.6, "max_memory_gb": 18.98, "ce": 0.009498732746578753, "kl": 0.11192346836469369, "context": 0.0654604674782604} +{"step": 182, "steps": 250, "lr": 1.3983299054645471e-05, "grad_norm": 3.5124008655548096, "loss_scale": 4.0, "retries": 0, "seconds": 93.1, "max_memory_gb": 18.98, "ce": 0.046218769435654394, "kl": 0.4246017233999737, "context": 0.25851962869637646} +{"step": 183, "steps": 250, "lr": 1.3743031817929724e-05, "grad_norm": 0.023386266082525253, "loss_scale": 4.0, "retries": 0, "seconds": 93.6, "max_memory_gb": 18.98, "ce": 0.0022230041836337477, "kl": 0.0006404414093594823, "context": 0.0025432248003198765} +{"step": 184, "steps": 250, "lr": 1.3505244196129164e-05, "grad_norm": 0.07818631827831268, "loss_scale": 4.0, "retries": 0, "seconds": 94.2, "max_memory_gb": 18.98, "ce": 0.004802599323738832, "kl": 0.0029455789299390744, "context": 0.006275388892390765} +{"step": 185, "steps": 250, "lr": 1.3269979049104014e-05, "grad_norm": 0.29358386993408203, "loss_scale": 4.0, "retries": 0, "seconds": 94.7, "max_memory_gb": 18.98, "ce": 0.028919396398123354, "kl": 0.128206392750144, "context": 0.09302259725518525} +{"step": 186, "steps": 250, "lr": 1.3037278782052863e-05, "grad_norm": 4.325638771057129, "loss_scale": 4.0, "retries": 0, "seconds": 95.2, "max_memory_gb": 18.98, "ce": 0.2607893957210763, "kl": 0.5764876204520988, "context": 0.5490332220997516} +{"step": 187, "steps": 250, "lr": 1.2807185337869343e-05, "grad_norm": 0.8024309277534485, "loss_scale": 4.0, "retries": 0, "seconds": 95.7, "max_memory_gb": 18.98, "ce": 0.1421125393389957, "kl": 0.566490052267909, "context": 0.4253575773909688} +{"step": 188, "steps": 250, "lr": 1.2579740189582115e-05, "grad_norm": 1.2086436748504639, "loss_scale": 4.0, "retries": 0, "seconds": 96.2, "max_memory_gb": 18.98, "ce": 0.3504050384990478, "kl": 0.015541225068773201, "context": 0.35817564893841336} +{"step": 189, "steps": 250, "lr": 1.2354984332879668e-05, "grad_norm": 1.52878999710083, "loss_scale": 4.0, "retries": 0, "seconds": 96.7, "max_memory_gb": 18.98, "ce": 0.3562849891022779, "kl": 0.4399126132193487, "context": 0.5762412956682965} +{"step": 190, "steps": 250, "lr": 1.2132958278721064e-05, "grad_norm": 0.06981436163187027, "loss_scale": 4.0, "retries": 0, "seconds": 97.2, "max_memory_gb": 18.98, "ce": 0.004208537167869508, "kl": 0.036032077245181426, "context": 0.02222457667812705} +{"step": 191, "steps": 250, "lr": 1.1913702046034016e-05, "grad_norm": 7.076426982879639, "loss_scale": 4.0, "retries": 0, "seconds": 97.7, "max_memory_gb": 18.98, "ce": 0.19781881657218037, "kl": 0.30148935728857396, "context": 0.34856347851200553} +{"step": 192, "steps": 250, "lr": 1.1697255154501756e-05, "grad_norm": 1.276957631111145, "loss_scale": 4.0, "retries": 0, "seconds": 98.2, "max_memory_gb": 18.98, "ce": 0.5102504066308029, "kl": 0.11476722898078151, "context": 0.5676340211066417} +{"step": 193, "steps": 250, "lr": 1.1483656617439842e-05, "grad_norm": 0.10395245999097824, "loss_scale": 4.0, "retries": 0, "seconds": 98.7, "max_memory_gb": 18.98, "ce": 0.010168041440920206, "kl": 0.018802635536758316, "context": 0.019569358394619485} +{"step": 194, "steps": 250, "lr": 1.1272944934764177e-05, "grad_norm": 0.2493802309036255, "loss_scale": 4.0, "retries": 0, "seconds": 99.2, "max_memory_gb": 18.98, "ce": 0.021816864471475128, "kl": 0.024713891791179776, "context": 0.03417381035978906} +{"step": 195, "steps": 250, "lr": 1.1065158086051684e-05, "grad_norm": 0.02485785260796547, "loss_scale": 4.0, "retries": 0, "seconds": 99.7, "max_memory_gb": 18.98, "ce": 0.00778044734033756, "kl": 0.006314644044323359, "context": 0.01093776899506338} +{"step": 196, "steps": 250, "lr": 1.086033352369467e-05, "grad_norm": 0.47263216972351074, "loss_scale": 4.0, "retries": 0, "seconds": 100.2, "max_memory_gb": 18.98, "ce": 0.0022769295824218716, "kl": 0.6611237913830124, "context": 0.3328388311820163} +{"step": 197, "steps": 250, "lr": 1.0658508166150224e-05, "grad_norm": 0.9383494257926941, "loss_scale": 4.0, "retries": 0, "seconds": 100.8, "max_memory_gb": 18.98, "ce": 0.0022877223964314908, "kl": 0.9108470051141921, "context": 0.4577112279366702} +{"step": 198, "steps": 250, "lr": 1.0459718391285954e-05, "grad_norm": 0.48888057470321655, "loss_scale": 4.0, "retries": 0, "seconds": 101.3, "max_memory_gb": 18.98, "ce": 0.05491276651082444, "kl": 0.01774496415600879, "context": 0.0637852475992986} +{"step": 199, "steps": 250, "lr": 1.0264000029822999e-05, "grad_norm": 0.5311357975006104, "loss_scale": 4.0, "retries": 0, "seconds": 101.8, "max_memory_gb": 18.98, "ce": 0.1997921514630434, "kl": 0.22646693600700019, "context": 0.3130256208605715} +{"step": 200, "steps": 250, "lr": 1.0071388358877775e-05, "grad_norm": 1.9523428678512573, "loss_scale": 4.0, "retries": 0, "seconds": 102.3, "max_memory_gb": 18.98, "ce": 0.432535801679478, "kl": 0.05552561691365554, "context": 0.4602986061363481} +{"step": 201, "steps": 250, "lr": 9.881918095603501e-06, "grad_norm": 1.075562596321106, "loss_scale": 4.0, "retries": 0, "seconds": 102.8, "max_memory_gb": 18.98, "ce": 0.0707029315381078, "kl": 0.16471292808273574, "context": 0.15305939561221749} +{"step": 202, "steps": 250, "lr": 9.695623390932612e-06, "grad_norm": 0.029294639825820923, "loss_scale": 4.0, "retries": 0, "seconds": 103.3, "max_memory_gb": 18.98, "ce": 0.0019277989536021778, "kl": 0.002745613201625474, "context": 0.003300605658296263} +{"step": 203, "steps": 250, "lr": 9.51253782342123e-06, "grad_norm": 0.4088200628757477, "loss_scale": 4.0, "retries": 0, "seconds": 103.8, "max_memory_gb": 18.98, "ce": 0.012283415868296288, "kl": 0.026889100729022175, "context": 0.025727966480189934} +{"step": 204, "steps": 250, "lr": 9.332694393196846e-06, "grad_norm": 0.11881604045629501, "loss_scale": 4.0, "retries": 0, "seconds": 104.3, "max_memory_gb": 18.98, "ce": 0.032972912420518696, "kl": 0.014392861485248432, "context": 0.04016934451647103} +{"step": 205, "steps": 250, "lr": 9.156125516010256e-06, "grad_norm": 0.2215890884399414, "loss_scale": 4.0, "retries": 0, "seconds": 104.8, "max_memory_gb": 18.98, "ce": 0.0015036104305181652, "kl": 0.6460757015665877, "context": 0.3245414472767152} +{"step": 206, "steps": 250, "lr": 8.982863017392733e-06, "grad_norm": 0.056184571236371994, "loss_scale": 4.0, "retries": 0, "seconds": 105.3, "max_memory_gb": 18.98, "ce": 0.00740640157391681, "kl": 0.003182952027373176, "context": 0.00899787732578261} +{"step": 207, "steps": 250, "lr": 8.81293812691974e-06, "grad_norm": 0.12145908921957016, "loss_scale": 4.0, "retries": 0, "seconds": 105.8, "max_memory_gb": 18.98, "ce": 0.051357803680730285, "kl": 0.07206586823303951, "context": 0.08739074153709225} +{"step": 208, "steps": 250, "lr": 8.646381472581927e-06, "grad_norm": 0.07247952371835709, "loss_scale": 4.0, "retries": 0, "seconds": 106.3, "max_memory_gb": 18.98, "ce": 0.005362880976463202, "kl": 0.0021831555059179664, "context": 0.006454458802181762} +{"step": 209, "steps": 250, "lr": 8.483223075264622e-06, "grad_norm": 0.6271920204162598, "loss_scale": 4.0, "retries": 0, "seconds": 106.8, "max_memory_gb": 18.98, "ce": 0.1379451782850083, "kl": 0.19017858827282907, "context": 0.2330344673100626} +{"step": 210, "steps": 250, "lr": 8.323492343336747e-06, "grad_norm": 1.682898759841919, "loss_scale": 4.0, "retries": 0, "seconds": 107.3, "max_memory_gb": 18.98, "ce": 0.5158633854266554, "kl": 0.18077340282798104, "context": 0.6062500830867066} +{"step": 211, "steps": 250, "lr": 8.167218067350122e-06, "grad_norm": 0.35970282554626465, "loss_scale": 4.0, "retries": 0, "seconds": 107.8, "max_memory_gb": 18.98, "ce": 0.06818124884739518, "kl": 0.2185630047461018, "context": 0.17746275896206498} +{"step": 212, "steps": 250, "lr": 8.01442841485013e-06, "grad_norm": 0.37219077348709106, "loss_scale": 4.0, "retries": 0, "seconds": 108.4, "max_memory_gb": 18.98, "ce": 0.015250517521053553, "kl": 1.1745796625036746, "context": 0.60254037741106} +{"step": 213, "steps": 250, "lr": 7.865150925298678e-06, "grad_norm": 0.45927923917770386, "loss_scale": 4.0, "retries": 0, "seconds": 108.9, "max_memory_gb": 18.98, "ce": 0.23562622417375678, "kl": 0.06291248667457694, "context": 0.2670824637862097} +{"step": 214, "steps": 250, "lr": 7.719412505110375e-06, "grad_norm": 0.5675139427185059, "loss_scale": 4.0, "retries": 0, "seconds": 109.4, "max_memory_gb": 18.98, "ce": 0.4940532047985471, "kl": 0.20468345152039547, "context": 0.5963949164724909} +{"step": 215, "steps": 250, "lr": 7.57723942280278e-06, "grad_norm": 1.1753995418548584, "loss_scale": 4.0, "retries": 0, "seconds": 109.9, "max_memory_gb": 18.98, "ce": 0.6518336723920584, "kl": 0.02018178930728709, "context": 0.6619245837946437} +{"step": 216, "steps": 250, "lr": 7.438657304261687e-06, "grad_norm": 0.9219538569450378, "loss_scale": 4.0, "retries": 0, "seconds": 110.4, "max_memory_gb": 18.98, "ce": 0.022916375561180757, "kl": 0.17131841169612017, "context": 0.10857557861527312} +{"step": 217, "steps": 250, "lr": 7.30369112812219e-06, "grad_norm": 0.014599842950701714, "loss_scale": 4.0, "retries": 0, "seconds": 110.9, "max_memory_gb": 18.98, "ce": 0.0007626533042639494, "kl": 0.003919585065887077, "context": 0.0027224457371630706} +{"step": 218, "steps": 250, "lr": 7.172365221266403e-06, "grad_norm": 0.8641185164451599, "loss_scale": 4.0, "retries": 0, "seconds": 111.4, "max_memory_gb": 18.98, "ce": 0.127925387583673, "kl": 0.3007251173257828, "context": 0.27828794717788696} +{"step": 219, "steps": 250, "lr": 7.0447032544387154e-06, "grad_norm": 1.3180837631225586, "loss_scale": 4.0, "retries": 0, "seconds": 111.9, "max_memory_gb": 18.98, "ce": 0.005148557836946566, "kl": 0.2747845639823936, "context": 0.14254083238483872} +{"step": 220, "steps": 250, "lr": 6.920728237979274e-06, "grad_norm": 0.11102703958749771, "loss_scale": 4.0, "retries": 0, "seconds": 112.4, "max_memory_gb": 18.98, "ce": 0.053502118522374076, "kl": 0.007500465055272798, "context": 0.05725234941201052} +{"step": 221, "steps": 250, "lr": 6.800462517676456e-06, "grad_norm": 0.29060178995132446, "loss_scale": 4.0, "retries": 0, "seconds": 112.9, "max_memory_gb": 18.98, "ce": 0.003925756609532982, "kl": 0.6514936300227419, "context": 0.32967255695257336} +{"step": 222, "steps": 250, "lr": 6.683927770739222e-06, "grad_norm": 0.15971165895462036, "loss_scale": 4.0, "retries": 0, "seconds": 113.5, "max_memory_gb": 18.98, "ce": 0.03683381643713801, "kl": 0.07616139100715458, "context": 0.07491451168243657} +{"step": 223, "steps": 250, "lr": 6.571145001889892e-06, "grad_norm": 0.1708502620458603, "loss_scale": 4.0, "retries": 0, "seconds": 114.0, "max_memory_gb": 18.98, "ce": 0.006482746022811625, "kl": 0.00042421620310051367, "context": 0.006694854204397416} +{"step": 224, "steps": 250, "lr": 6.46213453957817e-06, "grad_norm": 0.3116990923881531, "loss_scale": 4.0, "retries": 0, "seconds": 114.5, "max_memory_gb": 18.98, "ce": 0.008104746186290868, "kl": 0.2523290235449167, "context": 0.13426925854582805} +{"step": 225, "steps": 250, "lr": 6.356916032317063e-06, "grad_norm": 0.1807195395231247, "loss_scale": 4.0, "retries": 0, "seconds": 115.0, "max_memory_gb": 18.98, "ce": 0.0029644981004821602, "kl": 0.29927087222495175, "context": 0.15259992852224968} +{"step": 226, "steps": 250, "lr": 6.255508445141349e-06, "grad_norm": 2.1209359169006348, "loss_scale": 4.0, "retries": 0, "seconds": 115.5, "max_memory_gb": 18.98, "ce": 0.15223212714772671, "kl": 0.0973063075216487, "context": 0.20088528087944724} +{"step": 227, "steps": 250, "lr": 6.157930056189228e-06, "grad_norm": 0.245442733168602, "loss_scale": 4.0, "retries": 0, "seconds": 116.0, "max_memory_gb": 18.98, "ce": 0.02513497412110155, "kl": 0.003610693325754255, "context": 0.026940320180074195} +{"step": 228, "steps": 250, "lr": 6.064198453407796e-06, "grad_norm": 0.4937261641025543, "loss_scale": 4.0, "retries": 0, "seconds": 116.5, "max_memory_gb": 18.98, "ce": 0.0016577949172642548, "kl": 0.5495242518372834, "context": 0.2764199276425643} +{"step": 229, "steps": 250, "lr": 5.974330531382935e-06, "grad_norm": 0.3317440450191498, "loss_scale": 4.0, "retries": 0, "seconds": 117.0, "max_memory_gb": 18.98, "ce": 0.06311184691548988, "kl": 0.016503276320236182, "context": 0.07136348300264217} +{"step": 230, "steps": 250, "lr": 5.888342488294125e-06, "grad_norm": 2.5097782611846924, "loss_scale": 4.0, "retries": 0, "seconds": 117.5, "max_memory_gb": 18.98, "ce": 0.536767880046682, "kl": 0.6664086254313588, "context": 0.86997215935844} +{"step": 231, "steps": 250, "lr": 5.806249822994849e-06, "grad_norm": 0.15092229843139648, "loss_scale": 4.0, "retries": 0, "seconds": 118.0, "max_memory_gb": 18.98, "ce": 0.02031589940088452, "kl": 0.14670809423842002, "context": 0.09366994768424775} +{"step": 232, "steps": 250, "lr": 5.728067332218994e-06, "grad_norm": 0.11961091309785843, "loss_scale": 4.0, "retries": 0, "seconds": 118.5, "max_memory_gb": 18.98, "ce": 0.021503361640498042, "kl": 0.02094500692328438, "context": 0.03197586559690535} +{"step": 233, "steps": 250, "lr": 5.65380910791383e-06, "grad_norm": 0.45496058464050293, "loss_scale": 4.0, "retries": 0, "seconds": 119.0, "max_memory_gb": 18.98, "ce": 0.008016634732484818, "kl": 0.13776437286287546, "context": 0.07689882308477536} +{"step": 234, "steps": 250, "lr": 5.583488534700022e-06, "grad_norm": 0.6353839635848999, "loss_scale": 4.0, "retries": 0, "seconds": 119.5, "max_memory_gb": 18.98, "ce": 0.008714014787983615, "kl": 0.16091385892650578, "context": 0.08917094324715436} +{"step": 235, "steps": 250, "lr": 5.517118287459118e-06, "grad_norm": 1.327990174293518, "loss_scale": 4.0, "retries": 0, "seconds": 120.0, "max_memory_gb": 18.98, "ce": 0.44797027437380166, "kl": 0.059361370111219, "context": 0.4776509631774388} +{"step": 236, "steps": 250, "lr": 5.454710329048978e-06, "grad_norm": 0.032696157693862915, "loss_scale": 4.0, "retries": 0, "seconds": 120.5, "max_memory_gb": 18.98, "ce": 0.005291718014632352, "kl": 0.012696591322310269, "context": 0.011640013733995147} +{"step": 237, "steps": 250, "lr": 5.3962759081475436e-06, "grad_norm": 0.6939123868942261, "loss_scale": 4.0, "retries": 0, "seconds": 121.0, "max_memory_gb": 18.98, "ce": 0.07645469328781473, "kl": 0.05248076940102919, "context": 0.10269507807606715} +{"step": 238, "steps": 250, "lr": 5.341825557225321e-06, "grad_norm": 0.3137120306491852, "loss_scale": 4.0, "retries": 0, "seconds": 121.5, "max_memory_gb": 18.98, "ce": 0.0264911656267941, "kl": 0.018280836005942547, "context": 0.03563158237375319} +{"step": 239, "steps": 250, "lr": 5.291369090646961e-06, "grad_norm": 1.514756202697754, "loss_scale": 4.0, "retries": 0, "seconds": 122.0, "max_memory_gb": 18.98, "ce": 0.10683096142020077, "kl": 0.04932699473283719, "context": 0.131494456960354} +{"step": 240, "steps": 250, "lr": 5.244915602902289e-06, "grad_norm": 0.20388109982013702, "loss_scale": 4.0, "retries": 0, "seconds": 122.5, "max_memory_gb": 18.98, "ce": 0.11825199065060588, "kl": 0.04540035521131358, "context": 0.14095216906571295} +{"step": 241, "steps": 250, "lr": 5.20247346696706e-06, "grad_norm": 0.6219296455383301, "loss_scale": 4.0, "retries": 0, "seconds": 123.0, "max_memory_gb": 18.98, "ce": 0.08633769121661317, "kl": 0.029027049142314354, "context": 0.10085121357587923} +{"step": 242, "steps": 250, "lr": 5.164050332793787e-06, "grad_norm": 0.10016396641731262, "loss_scale": 4.0, "retries": 0, "seconds": 123.6, "max_memory_gb": 18.98, "ce": 0.002635877138345677, "kl": 0.07778649407327975, "context": 0.04152912359131733} +{"step": 243, "steps": 250, "lr": 5.129653125932873e-06, "grad_norm": 0.14145655930042267, "loss_scale": 4.0, "retries": 0, "seconds": 124.2, "max_memory_gb": 18.98, "ce": 0.011882355953275692, "kl": 0.012073329289705725, "context": 0.017919021178386174} +{"step": 244, "steps": 250, "lr": 5.099288046284329e-06, "grad_norm": 1.0349540710449219, "loss_scale": 4.0, "retries": 0, "seconds": 124.8, "max_memory_gb": 18.98, "ce": 0.03583223115128931, "kl": 0.20815578964538872, "context": 0.13991012645419687} +{"step": 245, "steps": 250, "lr": 5.072960566980277e-06, "grad_norm": 1.4871704578399658, "loss_scale": 4.0, "retries": 0, "seconds": 125.3, "max_memory_gb": 18.98, "ce": 0.3452576547861099, "kl": 0.10355844802688807, "context": 0.3970368728041649} +{"step": 246, "steps": 250, "lr": 5.050675433398435e-06, "grad_norm": 0.024403948336839676, "loss_scale": 4.0, "retries": 0, "seconds": 125.8, "max_memory_gb": 18.98, "ce": 0.0008483973306283588, "kl": 0.003287613219981722, "context": 0.0024922038828663062} +{"step": 247, "steps": 250, "lr": 5.032436662306811e-06, "grad_norm": 0.9192941188812256, "loss_scale": 4.0, "retries": 0, "seconds": 126.3, "max_memory_gb": 18.98, "ce": 0.02335078642136068, "kl": 0.6352659323683838, "context": 0.3409837462604628} +{"step": 248, "steps": 250, "lr": 5.018247541139684e-06, "grad_norm": 0.3144189119338989, "loss_scale": 4.0, "retries": 0, "seconds": 126.8, "max_memory_gb": 18.98, "ce": 0.028346118662739173, "kl": 0.025070299510844052, "context": 0.040881268214434385} +{"step": 249, "steps": 250, "lr": 5.00811062740508e-06, "grad_norm": 0.16786018013954163, "loss_scale": 4.0, "retries": 0, "seconds": 127.3, "max_memory_gb": 18.98, "ce": 0.027028330136090517, "kl": 0.0296074912766926, "context": 0.04183207580354065} +{"step": 250, "steps": 250, "lr": 5.002027748223778e-06, "grad_norm": 0.03469132259488106, "loss_scale": 4.0, "retries": 0, "seconds": 127.8, "max_memory_gb": 18.98, "ce": 0.004101354010344949, "kl": 0.00305124836449977, "context": 0.0056269781780429184} diff --git a/runs/single_internvl/context/L02_s13/adapter/README.md b/runs/single_internvl/context/L02_s13/adapter/README.md new file mode 100644 index 0000000000000000000000000000000000000000..4f9f856cf0e1531b47e25d266833b2977a19d3ac --- /dev/null +++ b/runs/single_internvl/context/L02_s13/adapter/README.md @@ -0,0 +1,207 @@ +--- +base_model: /data/OPSD/assets/models/InternVL3_5-8B-HF +library_name: peft +pipeline_tag: text-generation +tags: +- base_model:adapter:/data/OPSD/assets/models/InternVL3_5-8B-HF +- lora +- transformers +--- + +# Model Card for Model ID + + + + + +## Model Details + +### Model Description + + + + + +- **Developed by:** [More Information Needed] +- **Funded by [optional]:** [More Information Needed] +- **Shared by [optional]:** [More Information Needed] +- **Model type:** [More Information Needed] +- **Language(s) (NLP):** [More Information Needed] +- **License:** [More Information Needed] +- **Finetuned from model [optional]:** [More Information Needed] + +### Model Sources [optional] + + + +- **Repository:** [More Information Needed] +- **Paper [optional]:** [More Information Needed] +- **Demo [optional]:** [More Information Needed] + +## Uses + + + +### Direct Use + + + +[More Information Needed] + +### Downstream Use [optional] + + + +[More Information Needed] + +### Out-of-Scope Use + + + +[More Information Needed] + +## Bias, Risks, and Limitations + + + +[More Information Needed] + +### Recommendations + + + +Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. More information needed for further recommendations. + +## How to Get Started with the Model + +Use the code below to get started with the model. + +[More Information Needed] + +## Training Details + +### Training Data + + + +[More Information Needed] + +### Training Procedure + + + +#### Preprocessing [optional] + +[More Information Needed] + + +#### Training Hyperparameters + +- **Training regime:** [More Information Needed] + +#### Speeds, Sizes, Times [optional] + + + +[More Information Needed] + +## Evaluation + + + +### Testing Data, Factors & Metrics + +#### Testing Data + + + +[More Information Needed] + +#### Factors + + + +[More Information Needed] + +#### Metrics + + + +[More Information Needed] + +### Results + +[More Information Needed] + +#### Summary + + + +## Model Examination [optional] + + + +[More Information Needed] + +## Environmental Impact + + + +Carbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700). + +- **Hardware Type:** [More Information Needed] +- **Hours used:** [More Information Needed] +- **Cloud Provider:** [More Information Needed] +- **Compute Region:** [More Information Needed] +- **Carbon Emitted:** [More Information Needed] + +## Technical Specifications [optional] + +### Model Architecture and Objective + +[More Information Needed] + +### Compute Infrastructure + +[More Information Needed] + +#### Hardware + +[More Information Needed] + +#### Software + +[More Information Needed] + +## Citation [optional] + + + +**BibTeX:** + +[More Information Needed] + +**APA:** + +[More Information Needed] + +## Glossary [optional] + + + +[More Information Needed] + +## More Information [optional] + +[More Information Needed] + +## Model Card Authors [optional] + +[More Information Needed] + +## Model Card Contact + +[More Information Needed] +### Framework versions + +- PEFT 0.20.0 \ No newline at end of file diff --git a/runs/single_internvl/context/L02_s13/adapter/adapter_config.json b/runs/single_internvl/context/L02_s13/adapter/adapter_config.json new file mode 100644 index 0000000000000000000000000000000000000000..0e69ff92920e3c40b58eccfb4e2da9a79baa360d --- /dev/null +++ b/runs/single_internvl/context/L02_s13/adapter/adapter_config.json @@ -0,0 +1,50 @@ +{ + "alora_invocation_tokens": null, + "alpha_pattern": {}, + "arrow_config": null, + "auto_mapping": null, + "base_model_name_or_path": "/data/OPSD/assets/models/InternVL3_5-8B-HF", + "bias": "none", + "corda_config": null, + "ensure_weight_tying": false, + "eva_config": null, + "exclude_modules": null, + "fan_in_fan_out": false, + "inference_mode": true, + "init_lora_weights": true, + "layer_replication": null, + "layers_pattern": null, + "layers_to_transform": null, + "loftq_config": {}, + "lora_alpha": 32, + "lora_bias": false, + "lora_dropout": 0.05, + "lora_ga_config": null, + "megatron_config": null, + "megatron_core": "megatron.core", + "modules_to_save": null, + "monteclora_config": null, + "peft_type": "LORA", + "peft_version": "0.20.0", + "qalora_group_size": 16, + "r": 16, + "rank_pattern": {}, + "revision": null, + "target_modules": [ + "model.language_model.layers.2.self_attn.o_proj", + "model.language_model.layers.2.mlp.up_proj", + "model.language_model.layers.2.self_attn.q_proj", + "model.language_model.layers.2.mlp.gate_proj", + "model.language_model.layers.2.self_attn.k_proj", + "model.language_model.layers.2.mlp.down_proj", + "model.language_model.layers.2.self_attn.v_proj" + ], + "target_parameters": null, + "task_type": "CAUSAL_LM", + "trainable_token_indices": null, + "use_bdlora": null, + "use_dora": false, + "use_qalora": false, + "use_rslora": false, + "velora_config": null +} \ No newline at end of file diff --git a/runs/single_internvl/context/L02_s13/config.json b/runs/single_internvl/context/L02_s13/config.json new file mode 100644 index 0000000000000000000000000000000000000000..dd09e7b0fe7c7d22ee4fa310138e6cecfdf64450 --- /dev/null +++ b/runs/single_internvl/context/L02_s13/config.json @@ -0,0 +1,24 @@ +{ + "behavior": "context", + "seed": 13, + "model": "/data/OPSD/assets/models/InternVL3_5-8B-HF", + "backbone": "internvl", + "train": "/data/OPSD/assets/v3/context.jsonl", + "items": 1000, + "steps": 250, + "epochs": 1, + "steps_per_epoch": 250, + "groups_per_step": 4, + "lr": 5e-05, + "scheduler": "cosine", + "warmup_steps": 16, + "final_fraction": 0.1, + "weight_decay": 0.0, + "clip": 1.0, + "initial_loss_scale": 4.0, + "adapter": { + "layer": 2, + "rank": 16, + "trainable_parameters": 1212416 + } +} \ No newline at end of file diff --git a/runs/single_internvl/context/L02_s13/steps.jsonl b/runs/single_internvl/context/L02_s13/steps.jsonl new file mode 100644 index 0000000000000000000000000000000000000000..8b7935d5b66d7c298e2f89496268d2411b98e21b --- /dev/null +++ b/runs/single_internvl/context/L02_s13/steps.jsonl @@ -0,0 +1,250 @@ +{"step": 1, "steps": 250, "lr": 3.125e-06, "grad_norm": 0.5113595724105835, "loss_scale": 4.0, "retries": 0, "seconds": 1.5, "max_memory_gb": 18.86, "ce": 0.5227766726748087, "kl": 0.08654405153356493, "context": 0.5660487064160407} +{"step": 2, "steps": 250, "lr": 6.25e-06, "grad_norm": 0.8831896185874939, "loss_scale": 4.0, "retries": 0, "seconds": 2.0, "max_memory_gb": 18.87, "ce": 0.6385493846828467, "kl": 0.4661056873178495, "context": 0.8716021942018415} +{"step": 3, "steps": 250, "lr": 9.375000000000001e-06, "grad_norm": 2.078836441040039, "loss_scale": 4.0, "retries": 0, "seconds": 2.5, "max_memory_gb": 18.87, "ce": 1.098431184887886, "kl": 0.06860820762813091, "context": 1.1327353119850159} +{"step": 4, "steps": 250, "lr": 1.25e-05, "grad_norm": 1.0982985496520996, "loss_scale": 4.0, "retries": 0, "seconds": 3.1, "max_memory_gb": 18.87, "ce": 0.45131311725708656, "kl": 0.42913408672393416, "context": 0.6658801622470492} +{"step": 5, "steps": 250, "lr": 1.5625e-05, "grad_norm": 0.02149192802608013, "loss_scale": 4.0, "retries": 0, "seconds": 3.6, "max_memory_gb": 18.87, "ce": 0.009262646888601012, "kl": 0.008212978355004452, "context": 0.013369136417168193} +{"step": 6, "steps": 250, "lr": 1.8750000000000002e-05, "grad_norm": 0.48654404282569885, "loss_scale": 4.0, "retries": 0, "seconds": 4.1, "max_memory_gb": 18.87, "ce": 0.2871151866565924, "kl": 0.07454044656697079, "context": 0.32438541192095727} +{"step": 7, "steps": 250, "lr": 2.1875e-05, "grad_norm": 1.766396403312683, "loss_scale": 4.0, "retries": 0, "seconds": 4.6, "max_memory_gb": 18.87, "ce": 0.8232347047942312, "kl": 0.12348244012480336, "context": 0.8849759378954332} +{"step": 8, "steps": 250, "lr": 2.5e-05, "grad_norm": 1.636108160018921, "loss_scale": 4.0, "retries": 0, "seconds": 5.1, "max_memory_gb": 18.88, "ce": 0.7816616557538509, "kl": 0.28901919117197394, "context": 0.9261712692677975} +{"step": 9, "steps": 250, "lr": 2.8125000000000003e-05, "grad_norm": 1.1591521501541138, "loss_scale": 4.0, "retries": 0, "seconds": 5.6, "max_memory_gb": 18.88, "ce": 0.5146011114120483, "kl": 0.13601952316821553, "context": 0.582610871642828} +{"step": 10, "steps": 250, "lr": 3.125e-05, "grad_norm": 0.19274839758872986, "loss_scale": 4.0, "retries": 0, "seconds": 6.1, "max_memory_gb": 18.88, "ce": 0.05225200984568801, "kl": 0.04544100305065513, "context": 0.07497251301538199} +{"step": 11, "steps": 250, "lr": 3.4375e-05, "grad_norm": 1.297598958015442, "loss_scale": 4.0, "retries": 0, "seconds": 6.7, "max_memory_gb": 18.88, "ce": 0.6870616157175391, "kl": 0.004987321702628833, "context": 0.6895552827118081} +{"step": 12, "steps": 250, "lr": 3.7500000000000003e-05, "grad_norm": 0.9506701231002808, "loss_scale": 4.0, "retries": 0, "seconds": 7.2, "max_memory_gb": 18.88, "ce": 0.4598000148243955, "kl": 0.002664682920112682, "context": 0.4611323652370629} +{"step": 13, "steps": 250, "lr": 4.0625000000000005e-05, "grad_norm": 0.72252357006073, "loss_scale": 4.0, "retries": 0, "seconds": 7.7, "max_memory_gb": 18.88, "ce": 0.4186271364305867, "kl": 0.019770268265347113, "context": 0.42851228077779524} +{"step": 14, "steps": 250, "lr": 4.375e-05, "grad_norm": 1.6273829936981201, "loss_scale": 4.0, "retries": 0, "seconds": 8.2, "max_memory_gb": 18.88, "ce": 0.7322311699390411, "kl": 0.9434448080137372, "context": 1.203953579068184} +{"step": 15, "steps": 250, "lr": 4.6875e-05, "grad_norm": 1.5432528257369995, "loss_scale": 4.0, "retries": 0, "seconds": 8.7, "max_memory_gb": 18.88, "ce": 0.4826182944880202, "kl": 0.46426316538975243, "context": 0.7147498846334202} +{"step": 16, "steps": 250, "lr": 5e-05, "grad_norm": 0.47789692878723145, "loss_scale": 4.0, "retries": 0, "seconds": 9.2, "max_memory_gb": 18.88, "ce": 0.1989589036165853, "kl": 0.0009021912501339102, "context": 0.19941000494145555} +{"step": 17, "steps": 250, "lr": 5e-05, "grad_norm": 0.595705509185791, "loss_scale": 4.0, "retries": 0, "seconds": 9.7, "max_memory_gb": 18.88, "ce": 0.30816595719079487, "kl": 0.8294909682299476, "context": 0.7229114273213781} +{"step": 18, "steps": 250, "lr": 4.999797225177623e-05, "grad_norm": 1.5658928155899048, "loss_scale": 4.0, "retries": 0, "seconds": 10.2, "max_memory_gb": 18.88, "ce": 0.49361010815482587, "kl": 0.2102842673921259, "context": 0.5987522543000523} +{"step": 19, "steps": 250, "lr": 4.999188937259492e-05, "grad_norm": 2.74367094039917, "loss_scale": 4.0, "retries": 0, "seconds": 10.7, "max_memory_gb": 18.88, "ce": 0.6955608278512955, "kl": 0.17086793365888298, "context": 0.7809948027133942} +{"step": 20, "steps": 250, "lr": 4.998175245886032e-05, "grad_norm": 0.027533747255802155, "loss_scale": 4.0, "retries": 0, "seconds": 11.3, "max_memory_gb": 18.88, "ce": 0.0023556412934340187, "kl": 0.020668803106673295, "context": 0.012690042600297602} +{"step": 21, "steps": 250, "lr": 4.996756333769319e-05, "grad_norm": 1.542750597000122, "loss_scale": 4.0, "retries": 0, "seconds": 11.8, "max_memory_gb": 18.88, "ce": 0.31264629607085226, "kl": 0.04866100795334205, "context": 0.33697679630859056} +{"step": 22, "steps": 250, "lr": 4.994932456660157e-05, "grad_norm": 1.4042367935180664, "loss_scale": 4.0, "retries": 0, "seconds": 12.3, "max_memory_gb": 18.88, "ce": 0.3388922617014032, "kl": 0.032960911088594, "context": 0.35537270466738846} +{"step": 23, "steps": 250, "lr": 4.992703943301973e-05, "grad_norm": 1.8449172973632812, "loss_scale": 4.0, "retries": 0, "seconds": 12.8, "max_memory_gb": 18.88, "ce": 0.5174039897365219, "kl": 0.98827925608839, "context": 1.011543623280886} +{"step": 24, "steps": 250, "lr": 4.9900711953715675e-05, "grad_norm": 1.3545864820480347, "loss_scale": 4.0, "retries": 0, "seconds": 13.3, "max_memory_gb": 18.88, "ce": 0.2655661329627037, "kl": 0.30270162876695395, "context": 0.4169169422239065} +{"step": 25, "steps": 250, "lr": 4.987034687406713e-05, "grad_norm": 1.7266093492507935, "loss_scale": 4.0, "retries": 0, "seconds": 13.8, "max_memory_gb": 18.88, "ce": 0.6325299715535948, "kl": 0.20507778623141348, "context": 0.7350688688457012} +{"step": 26, "steps": 250, "lr": 4.9835949667206215e-05, "grad_norm": 1.6026840209960938, "loss_scale": 4.0, "retries": 0, "seconds": 14.4, "max_memory_gb": 18.88, "ce": 0.2632291540503502, "kl": 0.7033011235762388, "context": 0.6148797050118446} +{"step": 27, "steps": 250, "lr": 4.9797526533032945e-05, "grad_norm": 1.7736295461654663, "loss_scale": 4.0, "retries": 0, "seconds": 14.9, "max_memory_gb": 18.88, "ce": 0.3415953044386697, "kl": 0.41885100388753926, "context": 0.5510207996849203} +{"step": 28, "steps": 250, "lr": 4.9755084397097716e-05, "grad_norm": 0.8158969283103943, "loss_scale": 4.0, "retries": 0, "seconds": 15.4, "max_memory_gb": 18.88, "ce": 0.5915569439530373, "kl": 0.9430697424104437, "context": 1.063091829419136} +{"step": 29, "steps": 250, "lr": 4.970863090935304e-05, "grad_norm": 1.4336439371109009, "loss_scale": 4.0, "retries": 0, "seconds": 15.9, "max_memory_gb": 18.88, "ce": 0.22879723040387034, "kl": 0.3679667663818691, "context": 0.41278060898184776} +{"step": 30, "steps": 250, "lr": 4.965817444277468e-05, "grad_norm": 1.4258280992507935, "loss_scale": 4.0, "retries": 0, "seconds": 16.4, "max_memory_gb": 18.88, "ce": 0.29597065763300634, "kl": 0.012094407606127788, "context": 0.3020178636688797} +{"step": 31, "steps": 250, "lr": 4.9603724091852456e-05, "grad_norm": 2.368364095687866, "loss_scale": 4.0, "retries": 0, "seconds": 16.9, "max_memory_gb": 18.88, "ce": 0.5646668076515198, "kl": 0.15558304451406002, "context": 0.6424583308398724} +{"step": 32, "steps": 250, "lr": 4.9545289670951025e-05, "grad_norm": 2.2849440574645996, "loss_scale": 4.0, "retries": 0, "seconds": 17.5, "max_memory_gb": 18.88, "ce": 0.33852413296699524, "kl": 0.0174285288376268, "context": 0.347238402813673} +{"step": 33, "steps": 250, "lr": 4.948288171254089e-05, "grad_norm": 0.8515433073043823, "loss_scale": 4.0, "retries": 0, "seconds": 18.0, "max_memory_gb": 18.88, "ce": 0.19782700389623642, "kl": 0.06652530818246305, "context": 0.23108965856954455} +{"step": 34, "steps": 250, "lr": 4.941651146529998e-05, "grad_norm": 0.9592751860618591, "loss_scale": 4.0, "retries": 0, "seconds": 18.5, "max_memory_gb": 18.88, "ce": 0.12500090804132924, "kl": 0.7711627234875778, "context": 0.5105822618534148} +{"step": 35, "steps": 250, "lr": 4.9346190892086174e-05, "grad_norm": 0.9967995285987854, "loss_scale": 4.0, "retries": 0, "seconds": 19.0, "max_memory_gb": 18.88, "ce": 0.1461316129862098, "kl": 0.0040863378962967545, "context": 0.1481747828074731} +{"step": 36, "steps": 250, "lr": 4.927193266778101e-05, "grad_norm": 0.7660173177719116, "loss_scale": 4.0, "retries": 0, "seconds": 19.5, "max_memory_gb": 18.88, "ce": 0.19938988986541517, "kl": 0.057358146746992134, "context": 0.22806896406109445} +{"step": 37, "steps": 250, "lr": 4.919375017700515e-05, "grad_norm": 0.8075708150863647, "loss_scale": 4.0, "retries": 0, "seconds": 20.0, "max_memory_gb": 18.88, "ce": 0.30030008978792466, "kl": 0.7799924541031942, "context": 0.6902963075553998} +{"step": 38, "steps": 250, "lr": 4.911165751170588e-05, "grad_norm": 0.810559093952179, "loss_scale": 4.0, "retries": 0, "seconds": 20.5, "max_memory_gb": 18.88, "ce": 0.10551806539297104, "kl": 0.5072835236787796, "context": 0.359159829095006} +{"step": 39, "steps": 250, "lr": 4.902566946861708e-05, "grad_norm": 0.356835275888443, "loss_scale": 4.0, "retries": 0, "seconds": 21.0, "max_memory_gb": 18.88, "ce": 0.16724744548264425, "kl": 0.012548611019155942, "context": 0.17352174752159044} +{"step": 40, "steps": 250, "lr": 4.89358015465922e-05, "grad_norm": 0.6043879985809326, "loss_scale": 4.0, "retries": 0, "seconds": 21.5, "max_memory_gb": 18.88, "ce": 0.07092333998298272, "kl": 0.012480068719014525, "context": 0.0771633745462168} +{"step": 41, "steps": 250, "lr": 4.884206994381078e-05, "grad_norm": 1.1334683895111084, "loss_scale": 4.0, "retries": 0, "seconds": 22.0, "max_memory_gb": 18.88, "ce": 0.40875118284020573, "kl": 0.12902081792708486, "context": 0.4732615896500647} +{"step": 42, "steps": 250, "lr": 4.874449155485866e-05, "grad_norm": 0.27106907963752747, "loss_scale": 4.0, "retries": 0, "seconds": 22.5, "max_memory_gb": 18.88, "ce": 0.021656896347849397, "kl": 0.0006079156755731674, "context": 0.021960854315693723} +{"step": 43, "steps": 250, "lr": 4.864308396768294e-05, "grad_norm": 0.5822651386260986, "loss_scale": 4.0, "retries": 0, "seconds": 23.0, "max_memory_gb": 18.88, "ce": 0.03808628905972, "kl": 0.09920682742449571, "context": 0.08768970459641423} +{"step": 44, "steps": 250, "lr": 4.853786546042184e-05, "grad_norm": 1.9198886156082153, "loss_scale": 4.0, "retries": 0, "seconds": 23.6, "max_memory_gb": 18.88, "ce": 0.6349793334593414, "kl": 1.403952966342331, "context": 1.3369558203557972} +{"step": 45, "steps": 250, "lr": 4.842885499811012e-05, "grad_norm": 0.8576054573059082, "loss_scale": 4.0, "retries": 0, "seconds": 24.1, "max_memory_gb": 18.88, "ce": 0.9093665395630524, "kl": 0.24546256712346803, "context": 1.032097838964546} +{"step": 46, "steps": 250, "lr": 4.831607222926078e-05, "grad_norm": 0.4399333596229553, "loss_scale": 4.0, "retries": 0, "seconds": 24.6, "max_memory_gb": 18.88, "ce": 0.11865309439599514, "kl": 0.289541226811707, "context": 0.26342371106147766} +{"step": 47, "steps": 250, "lr": 4.8199537482323545e-05, "grad_norm": 0.4626289904117584, "loss_scale": 4.0, "retries": 0, "seconds": 25.1, "max_memory_gb": 18.88, "ce": 0.02978264380362816, "kl": 0.047433964093215764, "context": 0.053499626548727974} +{"step": 48, "steps": 250, "lr": 4.807927176202073e-05, "grad_norm": 0.7467386722564697, "loss_scale": 4.0, "retries": 0, "seconds": 25.6, "max_memory_gb": 18.88, "ce": 0.09227684864890762, "kl": 0.009112413903494598, "context": 0.09683305257931352} +{"step": 49, "steps": 250, "lr": 4.795529674556129e-05, "grad_norm": 0.4085307717323303, "loss_scale": 4.0, "retries": 0, "seconds": 26.1, "max_memory_gb": 18.88, "ce": 0.041734760539839044, "kl": 0.3851880495541309, "context": 0.23432878904350218} +{"step": 50, "steps": 250, "lr": 4.782763477873361e-05, "grad_norm": 0.6423295736312866, "loss_scale": 4.0, "retries": 0, "seconds": 26.6, "max_memory_gb": 18.88, "ce": 0.15438325429568067, "kl": 0.008005480543943122, "context": 0.15838599305425305} +{"step": 51, "steps": 250, "lr": 4.769630887187782e-05, "grad_norm": 1.9470500946044922, "loss_scale": 4.0, "retries": 0, "seconds": 27.1, "max_memory_gb": 18.88, "ce": 0.6095561306447053, "kl": 0.019245454194560807, "context": 0.6191788721771445} +{"step": 52, "steps": 250, "lr": 4.756134269573832e-05, "grad_norm": 1.2939679622650146, "loss_scale": 4.0, "retries": 0, "seconds": 27.6, "max_memory_gb": 18.88, "ce": 0.15465629933169112, "kl": 0.27574429637752473, "context": 0.29252844490110874} +{"step": 53, "steps": 250, "lr": 4.7422760577197226e-05, "grad_norm": 0.47400233149528503, "loss_scale": 4.0, "retries": 0, "seconds": 28.1, "max_memory_gb": 18.88, "ce": 0.1832674810430035, "kl": 0.4304545698687434, "context": 0.39849476516246796} +{"step": 54, "steps": 250, "lr": 4.728058749488963e-05, "grad_norm": 0.5721714496612549, "loss_scale": 4.0, "retries": 0, "seconds": 28.6, "max_memory_gb": 18.88, "ce": 0.05620607116725296, "kl": 0.3530711789499037, "context": 0.23274165973998606} +{"step": 55, "steps": 250, "lr": 4.713484907470133e-05, "grad_norm": 0.05784311145544052, "loss_scale": 4.0, "retries": 0, "seconds": 29.1, "max_memory_gb": 18.88, "ce": 0.015214839891996235, "kl": 0.015100616104973597, "context": 0.022765147834434174} +{"step": 56, "steps": 250, "lr": 4.6985571585149876e-05, "grad_norm": 0.28254464268684387, "loss_scale": 4.0, "retries": 0, "seconds": 29.6, "max_memory_gb": 18.88, "ce": 0.03510688792448491, "kl": 0.30697250948287547, "context": 0.18859314545989037} +{"step": 57, "steps": 250, "lr": 4.6832781932649884e-05, "grad_norm": 0.07182689756155014, "loss_scale": 4.0, "retries": 0, "seconds": 30.1, "max_memory_gb": 18.88, "ce": 0.011962271517404588, "kl": 0.0193986533640782, "context": 0.021661598206264898} +{"step": 58, "steps": 250, "lr": 4.667650765666325e-05, "grad_norm": 0.3119354248046875, "loss_scale": 4.0, "retries": 0, "seconds": 30.6, "max_memory_gb": 18.88, "ce": 0.040557609987445176, "kl": 0.0023375425448080023, "context": 0.04172637957526604} +{"step": 59, "steps": 250, "lr": 4.651677692473538e-05, "grad_norm": 0.17477233707904816, "loss_scale": 4.0, "retries": 0, "seconds": 31.2, "max_memory_gb": 18.88, "ce": 0.024076633897493593, "kl": 0.06579507562855724, "context": 0.05697416889597662} +{"step": 60, "steps": 250, "lr": 4.635361852741808e-05, "grad_norm": 0.11249559372663498, "loss_scale": 4.0, "retries": 0, "seconds": 31.7, "max_memory_gb": 18.88, "ce": 0.07213514648901764, "kl": 0.06269878198509105, "context": 0.10348453384358436} +{"step": 61, "steps": 250, "lr": 4.6187061873080264e-05, "grad_norm": 0.5968500375747681, "loss_scale": 4.0, "retries": 0, "seconds": 32.2, "max_memory_gb": 18.88, "ce": 0.029185163090005517, "kl": 1.0186271648854017, "context": 0.5384987406432629} +{"step": 62, "steps": 250, "lr": 4.601713698260727e-05, "grad_norm": 0.33136048913002014, "loss_scale": 4.0, "retries": 0, "seconds": 32.7, "max_memory_gb": 18.88, "ce": 0.039533545554149896, "kl": 0.010085488698678091, "context": 0.04457628936506808} +{"step": 63, "steps": 250, "lr": 4.5843874483989746e-05, "grad_norm": 0.14737972617149353, "loss_scale": 4.0, "retries": 0, "seconds": 33.2, "max_memory_gb": 18.88, "ce": 0.035687374882400036, "kl": 0.12787054923683172, "context": 0.09962264867499471} +{"step": 64, "steps": 250, "lr": 4.566730560680316e-05, "grad_norm": 0.12392888218164444, "loss_scale": 4.0, "retries": 0, "seconds": 33.7, "max_memory_gb": 18.88, "ce": 0.020047694677487016, "kl": 0.01259436592226848, "context": 0.026344877667725086} +{"step": 65, "steps": 250, "lr": 4.548746217657878e-05, "grad_norm": 0.5253763198852539, "loss_scale": 4.0, "retries": 0, "seconds": 34.2, "max_memory_gb": 18.88, "ce": 0.19899121019989252, "kl": 0.17438765824772418, "context": 0.2861850350163877} +{"step": 66, "steps": 250, "lr": 4.5304376609067396e-05, "grad_norm": 0.44296810030937195, "loss_scale": 4.0, "retries": 0, "seconds": 34.7, "max_memory_gb": 18.88, "ce": 0.15271974743484407, "kl": 0.06921341986736707, "context": 0.187326457397603} +{"step": 67, "steps": 250, "lr": 4.5118081904396504e-05, "grad_norm": 0.4484291672706604, "loss_scale": 4.0, "retries": 0, "seconds": 35.3, "max_memory_gb": 18.88, "ce": 0.027913753787288442, "kl": 0.04339889367111027, "context": 0.04961320152506232} +{"step": 68, "steps": 250, "lr": 4.4928611641122225e-05, "grad_norm": 0.19014142453670502, "loss_scale": 4.0, "retries": 0, "seconds": 35.8, "max_memory_gb": 18.88, "ce": 0.012002645266875334, "kl": 0.06207210784123163, "context": 0.0430386992302374} +{"step": 69, "steps": 250, "lr": 4.473599997017701e-05, "grad_norm": 0.5069074034690857, "loss_scale": 4.0, "retries": 0, "seconds": 36.3, "max_memory_gb": 18.88, "ce": 0.2735484786098823, "kl": 0.044543919852003455, "context": 0.2958204501774162} +{"step": 70, "steps": 250, "lr": 4.4540281608714053e-05, "grad_norm": 0.10731574147939682, "loss_scale": 4.0, "retries": 0, "seconds": 36.8, "max_memory_gb": 18.88, "ce": 0.006269634410273284, "kl": 0.00037294940011634026, "context": 0.006456108967540786} +{"step": 71, "steps": 250, "lr": 4.434149183384977e-05, "grad_norm": 0.29614412784576416, "loss_scale": 4.0, "retries": 0, "seconds": 37.3, "max_memory_gb": 18.88, "ce": 0.02368951761036442, "kl": 0.013877169491024688, "context": 0.030628101598267676} +{"step": 72, "steps": 250, "lr": 4.413966647630534e-05, "grad_norm": 0.23012785613536835, "loss_scale": 4.0, "retries": 0, "seconds": 37.8, "max_memory_gb": 18.88, "ce": 0.024218989070504904, "kl": 0.4495849722297862, "context": 0.2490114774554968} +{"step": 73, "steps": 250, "lr": 4.3934841913948323e-05, "grad_norm": 1.1915099620819092, "loss_scale": 4.0, "retries": 0, "seconds": 38.3, "max_memory_gb": 18.88, "ce": 0.41515193930536043, "kl": 0.12604301941610174, "context": 0.4781734415482788} +{"step": 74, "steps": 250, "lr": 4.372705506523583e-05, "grad_norm": 0.4188486337661743, "loss_scale": 4.0, "retries": 0, "seconds": 38.8, "max_memory_gb": 18.88, "ce": 0.09296812855336611, "kl": 0.1373348102424643, "context": 0.1616355327432757} +{"step": 75, "steps": 250, "lr": 4.351634338256017e-05, "grad_norm": 0.7426994442939758, "loss_scale": 4.0, "retries": 0, "seconds": 39.4, "max_memory_gb": 18.88, "ce": 0.4447729145613266, "kl": 1.0575919058173895, "context": 0.9735688953660429} +{"step": 76, "steps": 250, "lr": 4.330274484549825e-05, "grad_norm": 0.4913827180862427, "loss_scale": 4.0, "retries": 0, "seconds": 39.9, "max_memory_gb": 18.88, "ce": 0.01899548279470764, "kl": 1.8341394618619233, "context": 0.9360651975148357} +{"step": 77, "steps": 250, "lr": 4.308629795396599e-05, "grad_norm": 0.04694395512342453, "loss_scale": 4.0, "retries": 0, "seconds": 40.4, "max_memory_gb": 18.88, "ce": 0.007853632338083116, "kl": 0.012030034918552701, "context": 0.013868649446521886} +{"step": 78, "steps": 250, "lr": 4.286704172127894e-05, "grad_norm": 0.08969906717538834, "loss_scale": 4.0, "retries": 0, "seconds": 40.9, "max_memory_gb": 18.88, "ce": 0.002572662808233872, "kl": 0.0740471308818087, "context": 0.039596229209564626} +{"step": 79, "steps": 250, "lr": 4.264501566712033e-05, "grad_norm": 0.5547751188278198, "loss_scale": 4.0, "retries": 0, "seconds": 41.4, "max_memory_gb": 18.88, "ce": 0.042697395448158204, "kl": 0.07000628994273939, "context": 0.07770054228967638} +{"step": 80, "steps": 250, "lr": 4.2420259810417894e-05, "grad_norm": 0.15122023224830627, "loss_scale": 4.0, "retries": 0, "seconds": 41.9, "max_memory_gb": 18.88, "ce": 0.005733423635319923, "kl": 0.08658641122467259, "context": 0.04902662778886224} +{"step": 81, "steps": 250, "lr": 4.219281466213066e-05, "grad_norm": 0.4814637005329132, "loss_scale": 4.0, "retries": 0, "seconds": 42.4, "max_memory_gb": 18.88, "ce": 0.10485120955854654, "kl": 0.09545669483486563, "context": 0.1525795531924814} +{"step": 82, "steps": 250, "lr": 4.1962721217947135e-05, "grad_norm": 0.27430132031440735, "loss_scale": 4.0, "retries": 0, "seconds": 42.9, "max_memory_gb": 18.88, "ce": 0.02691927074920386, "kl": 0.005224459422606742, "context": 0.02953150009852834} +{"step": 83, "steps": 250, "lr": 4.173002095089599e-05, "grad_norm": 0.3892926573753357, "loss_scale": 4.0, "retries": 0, "seconds": 43.4, "max_memory_gb": 18.88, "ce": 0.17205384958651848, "kl": 0.8328409044479486, "context": 0.5884743120404892} +{"step": 84, "steps": 250, "lr": 4.149475580387084e-05, "grad_norm": 0.21825510263442993, "loss_scale": 4.0, "retries": 0, "seconds": 43.9, "max_memory_gb": 18.88, "ce": 0.020734932255436433, "kl": 0.05396751138755462, "context": 0.04771868899888432} +{"step": 85, "steps": 250, "lr": 4.1256968182070275e-05, "grad_norm": 1.7392094135284424, "loss_scale": 4.0, "retries": 0, "seconds": 44.4, "max_memory_gb": 18.88, "ce": 0.047781276643490855, "kl": 0.061957826385492965, "context": 0.07876018797298912} +{"step": 86, "steps": 250, "lr": 4.101670094535453e-05, "grad_norm": 1.484867811203003, "loss_scale": 4.0, "retries": 0, "seconds": 44.9, "max_memory_gb": 18.88, "ce": 0.28635678469436243, "kl": 0.5904164458625019, "context": 0.5815649775322527} +{"step": 87, "steps": 250, "lr": 4.077399740052027e-05, "grad_norm": 0.742273211479187, "loss_scale": 4.0, "retries": 0, "seconds": 45.4, "max_memory_gb": 18.88, "ce": 0.008402338848100044, "kl": 0.770933729596436, "context": 0.393869211897254} +{"step": 88, "steps": 250, "lr": 4.0528901293494775e-05, "grad_norm": 0.11680491268634796, "loss_scale": 4.0, "retries": 0, "seconds": 46.0, "max_memory_gb": 18.88, "ce": 0.005680304045426965, "kl": 0.03254066309538928, "context": 0.021950635334860635} +{"step": 89, "steps": 250, "lr": 4.0281456801451004e-05, "grad_norm": 0.07942890375852585, "loss_scale": 4.0, "retries": 0, "seconds": 46.5, "max_memory_gb": 18.88, "ce": 0.010610836661726353, "kl": 0.008340989138559962, "context": 0.014781331576159573} +{"step": 90, "steps": 250, "lr": 4.0031708524844884e-05, "grad_norm": 0.17913760244846344, "loss_scale": 4.0, "retries": 0, "seconds": 47.0, "max_memory_gb": 18.88, "ce": 0.014799913231399842, "kl": 0.016041817216319032, "context": 0.02282082161400467} +{"step": 91, "steps": 250, "lr": 3.977970147937635e-05, "grad_norm": 0.43138787150382996, "loss_scale": 4.0, "retries": 0, "seconds": 47.5, "max_memory_gb": 18.88, "ce": 0.15117375550653378, "kl": 0.7199798083784117, "context": 0.5111636657493364} +{"step": 92, "steps": 250, "lr": 3.952548108787553e-05, "grad_norm": 1.5078376531600952, "loss_scale": 4.0, "retries": 0, "seconds": 48.0, "max_memory_gb": 18.88, "ce": 0.47582230185435037, "kl": 0.5463971026929357, "context": 0.7490208822891873} +{"step": 93, "steps": 250, "lr": 3.9269093172115496e-05, "grad_norm": 1.043196439743042, "loss_scale": 4.0, "retries": 0, "seconds": 48.5, "max_memory_gb": 18.88, "ce": 0.059661963663529605, "kl": 0.5290146787010599, "context": 0.3241693072486669} +{"step": 94, "steps": 250, "lr": 3.901058394455323e-05, "grad_norm": 2.3947033882141113, "loss_scale": 4.0, "retries": 0, "seconds": 49.0, "max_memory_gb": 18.88, "ce": 0.343331994721666, "kl": 0.1709398437233176, "context": 0.42880193097516894} +{"step": 95, "steps": 250, "lr": 3.875e-05, "grad_norm": 1.1919069290161133, "loss_scale": 4.0, "retries": 0, "seconds": 49.5, "max_memory_gb": 18.88, "ce": 0.6912346010212786, "kl": 0.026654368499293923, "context": 0.7045617704279721} +{"step": 96, "steps": 250, "lr": 3.8487388307223005e-05, "grad_norm": 0.6814757585525513, "loss_scale": 4.0, "retries": 0, "seconds": 50.1, "max_memory_gb": 18.88, "ce": 0.5363397034088848, "kl": 0.005619933130219579, "context": 0.5391496940283105} +{"step": 97, "steps": 250, "lr": 3.822279620047943e-05, "grad_norm": 2.214322328567505, "loss_scale": 4.0, "retries": 0, "seconds": 50.6, "max_memory_gb": 18.88, "ce": 0.028740245397784747, "kl": 0.5820843671681359, "context": 0.31978243216872215} +{"step": 98, "steps": 250, "lr": 3.795627137098479e-05, "grad_norm": 0.08354225009679794, "loss_scale": 4.0, "retries": 0, "seconds": 51.1, "max_memory_gb": 18.88, "ce": 0.0035890029794245493, "kl": 0.09831708979345422, "context": 0.05274754625679634} +{"step": 99, "steps": 250, "lr": 3.7687861858316844e-05, "grad_norm": 0.589219331741333, "loss_scale": 4.0, "retries": 0, "seconds": 51.6, "max_memory_gb": 18.88, "ce": 0.03543608151994704, "kl": 0.01903488566949818, "context": 0.04495352364938299} +{"step": 100, "steps": 250, "lr": 3.741761604175668e-05, "grad_norm": 0.5695346593856812, "loss_scale": 4.0, "retries": 0, "seconds": 52.1, "max_memory_gb": 18.88, "ce": 0.02914485803921707, "kl": 0.8091576422593789, "context": 0.43372367951087654} +{"step": 101, "steps": 250, "lr": 3.714558263156872e-05, "grad_norm": 0.2582665979862213, "loss_scale": 4.0, "retries": 0, "seconds": 52.6, "max_memory_gb": 18.88, "ce": 0.0042288092081435025, "kl": 0.09200364985736087, "context": 0.050230632623424754} +{"step": 102, "steps": 250, "lr": 3.6871810660220904e-05, "grad_norm": 0.12736155092716217, "loss_scale": 4.0, "retries": 0, "seconds": 53.1, "max_memory_gb": 18.88, "ce": 0.04879177262773737, "kl": 0.03769864980131388, "context": 0.0676410966552794} +{"step": 103, "steps": 250, "lr": 3.659634947354686e-05, "grad_norm": 0.20410892367362976, "loss_scale": 4.0, "retries": 0, "seconds": 53.6, "max_memory_gb": 18.88, "ce": 0.01742412327803322, "kl": 0.02163885850313818, "context": 0.028243553460924886} +{"step": 104, "steps": 250, "lr": 3.631924872185169e-05, "grad_norm": 0.030742142349481583, "loss_scale": 4.0, "retries": 0, "seconds": 54.1, "max_memory_gb": 18.88, "ce": 0.002433383921015775, "kl": 0.0050562516059926566, "context": 0.00496150970866438} +{"step": 105, "steps": 250, "lr": 3.6040558350962674e-05, "grad_norm": 0.2778078317642212, "loss_scale": 4.0, "retries": 0, "seconds": 54.6, "max_memory_gb": 18.88, "ce": 0.06916242942679673, "kl": 0.3825884227408096, "context": 0.26045664260163903} +{"step": 106, "steps": 250, "lr": 3.576032859322687e-05, "grad_norm": 0.08988302946090698, "loss_scale": 4.0, "retries": 0, "seconds": 55.1, "max_memory_gb": 18.88, "ce": 0.005508097314304905, "kl": 0.005412839488599275, "context": 0.008214517118176445} +{"step": 107, "steps": 250, "lr": 3.5478609958457056e-05, "grad_norm": 1.699154257774353, "loss_scale": 4.0, "retries": 0, "seconds": 55.7, "max_memory_gb": 18.88, "ce": 0.8679920309878071, "kl": 0.08564963402000103, "context": 0.9108168372658838} +{"step": 108, "steps": 250, "lr": 3.519545322482755e-05, "grad_norm": 0.10021886229515076, "loss_scale": 4.0, "retries": 0, "seconds": 56.2, "max_memory_gb": 18.88, "ce": 0.009820480889175087, "kl": 0.0014367111162982837, "context": 0.010538836653722683} +{"step": 109, "steps": 250, "lr": 3.491090942972179e-05, "grad_norm": 0.32888296246528625, "loss_scale": 4.0, "retries": 0, "seconds": 56.7, "max_memory_gb": 18.88, "ce": 0.04322106213658117, "kl": 0.37913041561841965, "context": 0.23278627637773752} +{"step": 110, "steps": 250, "lr": 3.4625029860533135e-05, "grad_norm": 4.989300727844238, "loss_scale": 4.0, "retries": 0, "seconds": 57.2, "max_memory_gb": 18.88, "ce": 0.29900934058241546, "kl": 0.2963506575906649, "context": 0.44718467025086284} +{"step": 111, "steps": 250, "lr": 3.433786604542056e-05, "grad_norm": 0.01362613309174776, "loss_scale": 4.0, "retries": 0, "seconds": 57.7, "max_memory_gb": 18.88, "ce": 0.001107942804310369, "kl": 0.0013606439406430582, "context": 0.001788264784408966} +{"step": 112, "steps": 250, "lr": 3.404946974402104e-05, "grad_norm": 0.022301629185676575, "loss_scale": 4.0, "retries": 0, "seconds": 58.2, "max_memory_gb": 18.88, "ce": 0.0026211283693555743, "kl": 0.0004548519204945478, "context": 0.0028485543953138404} +{"step": 113, "steps": 250, "lr": 3.375989293812019e-05, "grad_norm": 0.8071959614753723, "loss_scale": 4.0, "retries": 0, "seconds": 58.7, "max_memory_gb": 18.88, "ce": 0.007810546609107405, "kl": 0.5211830346379429, "context": 0.26840206095948815} +{"step": 114, "steps": 250, "lr": 3.3469187822282766e-05, "grad_norm": 1.183284878730774, "loss_scale": 4.0, "retries": 0, "seconds": 59.2, "max_memory_gb": 18.88, "ce": 0.015037076198495924, "kl": 1.2496597010176629, "context": 0.639866933459416} +{"step": 115, "steps": 250, "lr": 3.317740679444499e-05, "grad_norm": 0.1717205047607422, "loss_scale": 4.0, "retries": 0, "seconds": 59.7, "max_memory_gb": 18.88, "ce": 0.011683077862471691, "kl": 0.054788958295830525, "context": 0.0390775568648678} +{"step": 116, "steps": 250, "lr": 3.2884602446470056e-05, "grad_norm": 1.1080090999603271, "loss_scale": 4.0, "retries": 0, "seconds": 60.2, "max_memory_gb": 18.88, "ce": 0.21326091537957836, "kl": 0.6330828563598487, "context": 0.529802343560732} +{"step": 117, "steps": 250, "lr": 3.259082755466876e-05, "grad_norm": 2.0682930946350098, "loss_scale": 4.0, "retries": 0, "seconds": 60.7, "max_memory_gb": 18.89, "ce": 0.18685158744483488, "kl": 0.3030004665142769, "context": 0.3383518207047018} +{"step": 118, "steps": 250, "lr": 3.229613507028686e-05, "grad_norm": 0.3343595266342163, "loss_scale": 4.0, "retries": 0, "seconds": 61.3, "max_memory_gb": 18.89, "ce": 0.17128222900282708, "kl": 0.020991614033619044, "context": 0.1817780334586132} +{"step": 119, "steps": 250, "lr": 3.2000578109961004e-05, "grad_norm": 0.16238318383693695, "loss_scale": 4.0, "retries": 0, "seconds": 61.8, "max_memory_gb": 18.89, "ce": 0.015038180366900633, "kl": 0.015255734262950682, "context": 0.02266604714350251} +{"step": 120, "steps": 250, "lr": 3.1704209946144616e-05, "grad_norm": 1.0973008871078491, "loss_scale": 4.0, "retries": 0, "seconds": 62.3, "max_memory_gb": 18.89, "ce": 0.8185217813006602, "kl": 0.20451467654493172, "context": 0.9207791136577725} +{"step": 121, "steps": 250, "lr": 3.140708399750594e-05, "grad_norm": 0.2579769194126129, "loss_scale": 4.0, "retries": 0, "seconds": 62.8, "max_memory_gb": 18.89, "ce": 0.05070473282785315, "kl": 0.04336546685863141, "context": 0.07238746857092337} +{"step": 122, "steps": 250, "lr": 3.1109253819299606e-05, "grad_norm": 0.13044314086437225, "loss_scale": 4.0, "retries": 0, "seconds": 63.3, "max_memory_gb": 18.89, "ce": 0.03751759706847224, "kl": 0.013418399934380432, "context": 0.04422679770505056} +{"step": 123, "steps": 250, "lr": 3.081077309371361e-05, "grad_norm": 0.2901575565338135, "loss_scale": 4.0, "retries": 0, "seconds": 63.8, "max_memory_gb": 18.89, "ce": 0.02908824064070359, "kl": 0.022619874769588932, "context": 0.040398177632596344} +{"step": 124, "steps": 250, "lr": 3.051169562019339e-05, "grad_norm": 1.3625659942626953, "loss_scale": 4.0, "retries": 0, "seconds": 64.4, "max_memory_gb": 18.89, "ce": 0.6997214895673096, "kl": 0.7350226845228462, "context": 1.0672328538494185} +{"step": 125, "steps": 250, "lr": 3.021207530574477e-05, "grad_norm": 1.279762625694275, "loss_scale": 4.0, "retries": 0, "seconds": 64.8, "max_memory_gb": 18.89, "ce": 0.13248213832412148, "kl": 0.15677112736739218, "context": 0.21086771271075122} +{"step": 126, "steps": 250, "lr": 2.991196615521758e-05, "grad_norm": 1.3904930353164673, "loss_scale": 4.0, "retries": 0, "seconds": 65.4, "max_memory_gb": 18.89, "ce": 0.46761762350797653, "kl": 0.1625472162850201, "context": 0.548891231417656} +{"step": 127, "steps": 250, "lr": 2.9611422261571536e-05, "grad_norm": 0.06018422171473503, "loss_scale": 4.0, "retries": 0, "seconds": 65.9, "max_memory_gb": 18.89, "ce": 0.012815514479370904, "kl": 0.001472092917538248, "context": 0.01355156117824663} +{"step": 128, "steps": 250, "lr": 2.931049779612634e-05, "grad_norm": 1.556426763534546, "loss_scale": 4.0, "retries": 0, "seconds": 66.4, "max_memory_gb": 18.89, "ce": 0.4877530694066081, "kl": 1.625829963650176, "context": 1.3006680069956928} +{"step": 129, "steps": 250, "lr": 2.9009246998797557e-05, "grad_norm": 0.17675979435443878, "loss_scale": 4.0, "retries": 0, "seconds": 66.9, "max_memory_gb": 18.89, "ce": 0.07872130069881678, "kl": 0.0365673495689407, "context": 0.09700497658923268} +{"step": 130, "steps": 250, "lr": 2.8707724168320238e-05, "grad_norm": 0.1440631002187729, "loss_scale": 4.0, "retries": 0, "seconds": 67.4, "max_memory_gb": 18.89, "ce": 0.039983502560062334, "kl": 0.3169354952069625, "context": 0.19845124921994284} +{"step": 131, "steps": 250, "lr": 2.840598365246184e-05, "grad_norm": 1.761919617652893, "loss_scale": 4.0, "retries": 0, "seconds": 67.9, "max_memory_gb": 18.89, "ce": 0.11020715062477393, "kl": 0.26966858439845964, "context": 0.24504144294769503} +{"step": 132, "steps": 250, "lr": 2.8104079838226383e-05, "grad_norm": 0.25378644466400146, "loss_scale": 4.0, "retries": 0, "seconds": 68.5, "max_memory_gb": 18.89, "ce": 0.17155381355405552, "kl": 0.03408046506137907, "context": 0.18859405097464332} +{"step": 133, "steps": 250, "lr": 2.7802067142051446e-05, "grad_norm": 0.9350085258483887, "loss_scale": 4.0, "retries": 0, "seconds": 69.1, "max_memory_gb": 18.89, "ce": 0.3463568896113429, "kl": 0.19533413369208574, "context": 0.44402395375072956} +{"step": 134, "steps": 250, "lr": 2.7500000000000004e-05, "grad_norm": 0.7425650358200073, "loss_scale": 4.0, "retries": 0, "seconds": 69.7, "max_memory_gb": 18.89, "ce": 0.2550849492827183, "kl": 0.005537800359299183, "context": 0.257853858778617} +{"step": 135, "steps": 250, "lr": 2.7197932857948566e-05, "grad_norm": 0.11762218922376633, "loss_scale": 4.0, "retries": 0, "seconds": 70.2, "max_memory_gb": 18.89, "ce": 0.034445243974914774, "kl": 0.10000827729527373, "context": 0.08444938459433615} +{"step": 136, "steps": 250, "lr": 2.689592016177363e-05, "grad_norm": 4.4566874504089355, "loss_scale": 4.0, "retries": 0, "seconds": 70.7, "max_memory_gb": 18.89, "ce": 0.29240538328303955, "kl": 0.3567829923995305, "context": 0.4707968789880397} +{"step": 137, "steps": 250, "lr": 2.659401634753817e-05, "grad_norm": 0.05169380456209183, "loss_scale": 4.0, "retries": 0, "seconds": 71.2, "max_memory_gb": 18.89, "ce": 0.0073819988683681, "kl": 0.00775223936943803, "context": 0.011258118320256472} +{"step": 138, "steps": 250, "lr": 2.6292275831679764e-05, "grad_norm": 2.0606894493103027, "loss_scale": 4.0, "retries": 0, "seconds": 71.7, "max_memory_gb": 18.89, "ce": 0.10334185883402824, "kl": 0.23173288628458977, "context": 0.2192082954570651} +{"step": 139, "steps": 250, "lr": 2.5990753001202445e-05, "grad_norm": 0.7088186144828796, "loss_scale": 4.0, "retries": 0, "seconds": 72.2, "max_memory_gb": 18.89, "ce": 0.0025176682211167645, "kl": 0.4668662438343745, "context": 0.23595079587539658} +{"step": 140, "steps": 250, "lr": 2.5689502203873672e-05, "grad_norm": 0.0392739437520504, "loss_scale": 4.0, "retries": 0, "seconds": 72.8, "max_memory_gb": 18.89, "ce": 0.0050070770284946775, "kl": 0.0048849135637283325, "context": 0.007449533601175062} +{"step": 141, "steps": 250, "lr": 2.538857773842847e-05, "grad_norm": 0.028299445286393166, "loss_scale": 4.0, "retries": 0, "seconds": 73.3, "max_memory_gb": 18.89, "ce": 0.0031285111108445562, "kl": 0.005875022747204639, "context": 0.0060660225281026214} +{"step": 142, "steps": 250, "lr": 2.5088033844782426e-05, "grad_norm": 0.9832183122634888, "loss_scale": 4.0, "retries": 0, "seconds": 73.8, "max_memory_gb": 18.89, "ce": 0.0557661214043037, "kl": 0.03547204367350787, "context": 0.07350214314647019} +{"step": 143, "steps": 250, "lr": 2.4787924694255237e-05, "grad_norm": 1.2569880485534668, "loss_scale": 4.0, "retries": 0, "seconds": 74.3, "max_memory_gb": 18.89, "ce": 0.7052900893959304, "kl": 0.04767127914510638, "context": 0.7291257363212935} +{"step": 144, "steps": 250, "lr": 2.4488304379806627e-05, "grad_norm": 0.8552082180976868, "loss_scale": 4.0, "retries": 0, "seconds": 74.8, "max_memory_gb": 18.89, "ce": 0.1334397722966969, "kl": 0.20779064309317619, "context": 0.2373350909911096} +{"step": 145, "steps": 250, "lr": 2.4189226906286388e-05, "grad_norm": 0.04527544230222702, "loss_scale": 4.0, "retries": 0, "seconds": 75.3, "max_memory_gb": 18.89, "ce": 0.0023920859275676776, "kl": 0.010452494388744071, "context": 0.007618332867423305} +{"step": 146, "steps": 250, "lr": 2.3890746180700395e-05, "grad_norm": 0.06806769967079163, "loss_scale": 4.0, "retries": 0, "seconds": 75.8, "max_memory_gb": 18.89, "ce": 0.00571758730802685, "kl": 0.013143622018105816, "context": 0.012289398582652211} +{"step": 147, "steps": 250, "lr": 2.359291600249407e-05, "grad_norm": 0.5965805053710938, "loss_scale": 4.0, "retries": 0, "seconds": 76.4, "max_memory_gb": 18.89, "ce": 0.46920313287046156, "kl": 0.6272097787295934, "context": 0.7828080110011797} +{"step": 148, "steps": 250, "lr": 2.3295790053855392e-05, "grad_norm": 0.07706161588430405, "loss_scale": 4.0, "retries": 0, "seconds": 76.9, "max_memory_gb": 18.89, "ce": 0.0059839392050662354, "kl": 0.012459494504582835, "context": 0.012213686055474682} +{"step": 149, "steps": 250, "lr": 2.299942189003901e-05, "grad_norm": 0.2682195007801056, "loss_scale": 4.0, "retries": 0, "seconds": 77.4, "max_memory_gb": 18.89, "ce": 0.03231970823253505, "kl": 0.013301085869898088, "context": 0.038970251160208136} +{"step": 150, "steps": 250, "lr": 2.2703864929713144e-05, "grad_norm": 0.13815590739250183, "loss_scale": 4.0, "retries": 0, "seconds": 77.9, "max_memory_gb": 18.89, "ce": 0.051556661317590624, "kl": 0.023378448211587965, "context": 0.06324588553979993} +{"step": 151, "steps": 250, "lr": 2.2409172445331257e-05, "grad_norm": 0.07738377153873444, "loss_scale": 4.0, "retries": 0, "seconds": 78.4, "max_memory_gb": 18.89, "ce": 0.009928006597874628, "kl": 0.015562560093371758, "context": 0.01770928711266606} +{"step": 152, "steps": 250, "lr": 2.211539755352996e-05, "grad_norm": 0.26473984122276306, "loss_scale": 4.0, "retries": 0, "seconds": 78.9, "max_memory_gb": 18.89, "ce": 0.012553656335512642, "kl": 0.7146060263295269, "context": 0.3698566610892158} +{"step": 153, "steps": 250, "lr": 2.1822593205555016e-05, "grad_norm": 0.29259851574897766, "loss_scale": 4.0, "retries": 0, "seconds": 79.5, "max_memory_gb": 18.89, "ce": 0.009743491694280237, "kl": 0.49015998849472453, "context": 0.25482348175046354} +{"step": 154, "steps": 250, "lr": 2.1530812177717236e-05, "grad_norm": 5.170650482177734, "loss_scale": 4.0, "retries": 0, "seconds": 80.0, "max_memory_gb": 18.89, "ce": 0.22635695175267756, "kl": 0.10096408089157194, "context": 0.2768389950506389} +{"step": 155, "steps": 250, "lr": 2.1240107061879815e-05, "grad_norm": 0.32605645060539246, "loss_scale": 4.0, "retries": 0, "seconds": 80.5, "max_memory_gb": 18.89, "ce": 0.20195462819719978, "kl": 0.010813109495302342, "context": 0.20736118391050695} +{"step": 156, "steps": 250, "lr": 2.095053025597896e-05, "grad_norm": 4.149329662322998, "loss_scale": 4.0, "retries": 0, "seconds": 81.0, "max_memory_gb": 18.89, "ce": 0.011585809259486268, "kl": 0.5309517379698718, "context": 0.2770616656735001} +{"step": 157, "steps": 250, "lr": 2.066213395457945e-05, "grad_norm": 1.8956719636917114, "loss_scale": 4.0, "retries": 0, "seconds": 81.5, "max_memory_gb": 18.89, "ce": 0.17430788991623558, "kl": 0.1103549388317333, "context": 0.22948536317562684} +{"step": 158, "steps": 250, "lr": 2.0374970139466877e-05, "grad_norm": 0.3056911528110504, "loss_scale": 4.0, "retries": 0, "seconds": 82.1, "max_memory_gb": 18.89, "ce": 0.16214629114074341, "kl": 0.0978023222196498, "context": 0.21104745879711118} +{"step": 159, "steps": 250, "lr": 2.0089090570278213e-05, "grad_norm": 0.4842010736465454, "loss_scale": 4.0, "retries": 0, "seconds": 82.6, "max_memory_gb": 18.9, "ce": 0.003474015389656415, "kl": 0.08684656226250809, "context": 0.04689729785604868} +{"step": 160, "steps": 250, "lr": 1.9804546775172457e-05, "grad_norm": 0.9468673467636108, "loss_scale": 4.0, "retries": 0, "seconds": 83.1, "max_memory_gb": 18.9, "ce": 0.023742396791931242, "kl": 0.401237522426527, "context": 0.22436116728931665} +{"step": 161, "steps": 250, "lr": 1.9521390041542946e-05, "grad_norm": 1.2423341274261475, "loss_scale": 4.0, "retries": 0, "seconds": 83.6, "max_memory_gb": 18.9, "ce": 0.0988069017566886, "kl": 0.11994909667328102, "context": 0.15878144637099467} +{"step": 162, "steps": 250, "lr": 1.923967140677313e-05, "grad_norm": 1.0983953475952148, "loss_scale": 4.0, "retries": 0, "seconds": 84.1, "max_memory_gb": 18.9, "ce": 0.1415381482383964, "kl": 0.27228345389448805, "context": 0.27767989011408645} +{"step": 163, "steps": 250, "lr": 1.8959441649037335e-05, "grad_norm": 1.5998307466506958, "loss_scale": 4.0, "retries": 0, "seconds": 84.7, "max_memory_gb": 18.9, "ce": 0.2556429873802699, "kl": 0.9354742137948051, "context": 0.7233801156980917} +{"step": 164, "steps": 250, "lr": 1.868075127814832e-05, "grad_norm": 4.494747161865234, "loss_scale": 4.0, "retries": 0, "seconds": 85.2, "max_memory_gb": 18.9, "ce": 0.1696689094096655, "kl": 0.3397353502805345, "context": 0.3395365693140775} +{"step": 165, "steps": 250, "lr": 1.8403650526453143e-05, "grad_norm": 0.05098913237452507, "loss_scale": 4.0, "retries": 0, "seconds": 85.7, "max_memory_gb": 18.9, "ce": 0.0022457891191152157, "kl": 0.021904388286202447, "context": 0.013197982971178135} +{"step": 166, "steps": 250, "lr": 1.812818933977912e-05, "grad_norm": 1.7256860733032227, "loss_scale": 4.0, "retries": 0, "seconds": 86.2, "max_memory_gb": 18.9, "ce": 0.24846602516481653, "kl": 0.14364255111286184, "context": 0.32028729512239806} +{"step": 167, "steps": 250, "lr": 1.785441736843128e-05, "grad_norm": 0.21737374365329742, "loss_scale": 4.0, "retries": 0, "seconds": 86.8, "max_memory_gb": 18.9, "ce": 0.03116546181263402, "kl": 0.007315540962736122, "context": 0.03482323215575889} +{"step": 168, "steps": 250, "lr": 1.7582383958243314e-05, "grad_norm": 0.09733942151069641, "loss_scale": 4.0, "retries": 0, "seconds": 87.3, "max_memory_gb": 18.9, "ce": 0.032030245653004386, "kl": 0.0036252559220883995, "context": 0.033842874778201804} +{"step": 169, "steps": 250, "lr": 1.7312138141683164e-05, "grad_norm": 1.1581867933273315, "loss_scale": 4.0, "retries": 0, "seconds": 87.8, "max_memory_gb": 18.9, "ce": 0.10644135042093694, "kl": 1.629457637667656, "context": 0.9211701676249504} +{"step": 170, "steps": 250, "lr": 1.704372862901521e-05, "grad_norm": 0.06776829808950424, "loss_scale": 4.0, "retries": 0, "seconds": 88.3, "max_memory_gb": 18.9, "ce": 0.0030756278301851125, "kl": 0.001478969948948361, "context": 0.003815112911979668} +{"step": 171, "steps": 250, "lr": 1.6777203799520573e-05, "grad_norm": 0.4686625599861145, "loss_scale": 4.0, "retries": 0, "seconds": 88.8, "max_memory_gb": 18.9, "ce": 0.03458639364998817, "kl": 0.30187852478411514, "context": 0.18552565843128832} +{"step": 172, "steps": 250, "lr": 1.6512611692777008e-05, "grad_norm": 0.10762770473957062, "loss_scale": 4.0, "retries": 0, "seconds": 89.3, "max_memory_gb": 18.9, "ce": 0.015806299401447177, "kl": 0.014514032460283488, "context": 0.02306331554427743} +{"step": 173, "steps": 250, "lr": 1.6250000000000005e-05, "grad_norm": 0.12612314522266388, "loss_scale": 4.0, "retries": 0, "seconds": 89.8, "max_memory_gb": 18.9, "ce": 0.05629992007743567, "kl": 0.03400942051666789, "context": 0.07330462976824492} +{"step": 174, "steps": 250, "lr": 1.598941605544678e-05, "grad_norm": 0.23436784744262695, "loss_scale": 4.0, "retries": 0, "seconds": 90.3, "max_memory_gb": 18.9, "ce": 0.053333914202085, "kl": 0.032538042353408514, "context": 0.0696029325863492} +{"step": 175, "steps": 250, "lr": 1.5730906827884506e-05, "grad_norm": 3.0537896156311035, "loss_scale": 4.0, "retries": 0, "seconds": 90.8, "max_memory_gb": 18.9, "ce": 0.688330648932606, "kl": 0.10564704734133556, "context": 0.7411541873589158} +{"step": 176, "steps": 250, "lr": 1.5474518912124476e-05, "grad_norm": 0.03643615543842316, "loss_scale": 4.0, "retries": 0, "seconds": 91.4, "max_memory_gb": 18.9, "ce": 0.009367422142531723, "kl": 0.003070872124226298, "context": 0.010902858019107953} +{"step": 177, "steps": 250, "lr": 1.5220298520623646e-05, "grad_norm": 0.38081833720207214, "loss_scale": 4.0, "retries": 0, "seconds": 91.9, "max_memory_gb": 18.9, "ce": 0.008503638699039584, "kl": 0.9456400641356595, "context": 0.48132365263882093} +{"step": 178, "steps": 250, "lr": 1.4968291475155121e-05, "grad_norm": 0.8678735494613647, "loss_scale": 4.0, "retries": 0, "seconds": 92.4, "max_memory_gb": 18.9, "ce": 0.10888481841539033, "kl": 0.13936850534082623, "context": 0.17856907201348804} +{"step": 179, "steps": 250, "lr": 1.4718543198548998e-05, "grad_norm": 0.5732279419898987, "loss_scale": 4.0, "retries": 0, "seconds": 92.9, "max_memory_gb": 18.9, "ce": 0.3259128004319791, "kl": 0.040466211015882436, "context": 0.34614590035562287} +{"step": 180, "steps": 250, "lr": 1.4471098706505227e-05, "grad_norm": 0.08089287579059601, "loss_scale": 4.0, "retries": 0, "seconds": 93.4, "max_memory_gb": 18.9, "ce": 0.02950503909960389, "kl": 0.022444151458330452, "context": 0.040727114712353796} +{"step": 181, "steps": 250, "lr": 1.4226002599479743e-05, "grad_norm": 0.1227366179227829, "loss_scale": 4.0, "retries": 0, "seconds": 93.9, "max_memory_gb": 18.9, "ce": 0.0079719195200596, "kl": 0.11091881444735918, "context": 0.06343132734764367} +{"step": 182, "steps": 250, "lr": 1.3983299054645471e-05, "grad_norm": 2.447655200958252, "loss_scale": 4.0, "retries": 0, "seconds": 94.4, "max_memory_gb": 18.9, "ce": 0.04582215138361789, "kl": 0.2883826795132336, "context": 0.19001348927849904} +{"step": 183, "steps": 250, "lr": 1.3743031817929724e-05, "grad_norm": 0.016019307076931, "loss_scale": 4.0, "retries": 0, "seconds": 94.9, "max_memory_gb": 18.9, "ce": 0.001701760187643231, "kl": 0.00050809513936656, "context": 0.0019558077274268726} +{"step": 184, "steps": 250, "lr": 1.3505244196129164e-05, "grad_norm": 0.06497994065284729, "loss_scale": 4.0, "retries": 0, "seconds": 95.4, "max_memory_gb": 18.9, "ce": 0.003004587681061821, "kl": 0.0027335571285220794, "context": 0.004371366310806479} +{"step": 185, "steps": 250, "lr": 1.3269979049104014e-05, "grad_norm": 0.19403710961341858, "loss_scale": 4.0, "retries": 0, "seconds": 95.9, "max_memory_gb": 18.9, "ce": 0.020152282435446978, "kl": 0.12106990907341242, "context": 0.08068723953329027} +{"step": 186, "steps": 250, "lr": 1.3037278782052863e-05, "grad_norm": 6.1361565589904785, "loss_scale": 4.0, "retries": 0, "seconds": 96.5, "max_memory_gb": 18.9, "ce": 0.24550153430027422, "kl": 0.5575770901123178, "context": 0.5242900644843758} +{"step": 187, "steps": 250, "lr": 1.2807185337869343e-05, "grad_norm": 0.7373682260513306, "loss_scale": 4.0, "retries": 0, "seconds": 97.0, "max_memory_gb": 18.9, "ce": 0.1279454934410751, "kl": 0.5467671644873917, "context": 0.4013290833681822} +{"step": 188, "steps": 250, "lr": 1.2579740189582115e-05, "grad_norm": 1.1781491041183472, "loss_scale": 4.0, "retries": 0, "seconds": 97.5, "max_memory_gb": 18.9, "ce": 0.40258247221663623, "kl": 0.018301024581319325, "context": 0.4117329886994412} +{"step": 189, "steps": 250, "lr": 1.2354984332879668e-05, "grad_norm": 3.4264349937438965, "loss_scale": 4.0, "retries": 0, "seconds": 98.0, "max_memory_gb": 18.9, "ce": 0.5200550878071226, "kl": 0.15553018060745671, "context": 0.5978201744146645} +{"step": 190, "steps": 250, "lr": 1.2132958278721064e-05, "grad_norm": 0.0687575563788414, "loss_scale": 4.0, "retries": 0, "seconds": 98.5, "max_memory_gb": 18.9, "ce": 0.0036271603748900816, "kl": 0.03679524958715774, "context": 0.022024784877430648} +{"step": 191, "steps": 250, "lr": 1.1913702046034016e-05, "grad_norm": 3.6814522743225098, "loss_scale": 4.0, "retries": 0, "seconds": 99.0, "max_memory_gb": 18.9, "ce": 0.18557510995196935, "kl": 0.3009194073524668, "context": 0.3360348075748334} +{"step": 192, "steps": 250, "lr": 1.1697255154501756e-05, "grad_norm": 1.4721180200576782, "loss_scale": 4.0, "retries": 0, "seconds": 99.5, "max_memory_gb": 18.9, "ce": 0.49999095776001923, "kl": 0.14598525171459187, "context": 0.5729835733654909} +{"step": 193, "steps": 250, "lr": 1.1483656617439842e-05, "grad_norm": 0.1059667095541954, "loss_scale": 4.0, "retries": 0, "seconds": 100.0, "max_memory_gb": 18.9, "ce": 0.007939679305309255, "kl": 0.017605403892730465, "context": 0.01674238123723626} +{"step": 194, "steps": 250, "lr": 1.1272944934764177e-05, "grad_norm": 0.2537337839603424, "loss_scale": 4.0, "retries": 0, "seconds": 100.5, "max_memory_gb": 18.9, "ce": 0.02568903347128071, "kl": 0.020205609431286575, "context": 0.035791838192380965} +{"step": 195, "steps": 250, "lr": 1.1065158086051684e-05, "grad_norm": 0.02434499002993107, "loss_scale": 4.0, "retries": 0, "seconds": 101.0, "max_memory_gb": 18.9, "ce": 0.00746212423837278, "kl": 0.0058623973054636735, "context": 0.010393322969321162} +{"step": 196, "steps": 250, "lr": 1.086033352369467e-05, "grad_norm": 0.5026471614837646, "loss_scale": 4.0, "retries": 0, "seconds": 101.6, "max_memory_gb": 18.9, "ce": 0.0011579650708881672, "kl": 0.63811155817627, "context": 0.3202137384255366} +{"step": 197, "steps": 250, "lr": 1.0658508166150224e-05, "grad_norm": 0.8039165735244751, "loss_scale": 4.0, "retries": 0, "seconds": 102.1, "max_memory_gb": 18.9, "ce": 0.0016732350341044366, "kl": 0.8866619597756653, "context": 0.44500421508564614} +{"step": 198, "steps": 250, "lr": 1.0459718391285954e-05, "grad_norm": 1.338873267173767, "loss_scale": 4.0, "retries": 0, "seconds": 102.6, "max_memory_gb": 18.9, "ce": 0.06287257609074004, "kl": 0.020249730499926955, "context": 0.07299743941985071} +{"step": 199, "steps": 250, "lr": 1.0264000029822999e-05, "grad_norm": 0.60829097032547, "loss_scale": 4.0, "retries": 0, "seconds": 103.1, "max_memory_gb": 18.9, "ce": 0.1887878956113127, "kl": 0.22932188454615243, "context": 0.3034488447374315} +{"step": 200, "steps": 250, "lr": 1.0071388358877775e-05, "grad_norm": 1.6656337976455688, "loss_scale": 4.0, "retries": 0, "seconds": 103.6, "max_memory_gb": 18.9, "ce": 0.4980893843603553, "kl": 0.04889349005679833, "context": 0.5225361396296648} +{"step": 201, "steps": 250, "lr": 9.881918095603501e-06, "grad_norm": 0.7379074096679688, "loss_scale": 4.0, "retries": 0, "seconds": 104.1, "max_memory_gb": 18.9, "ce": 0.06446563342615264, "kl": 0.15137608818986337, "context": 0.14015367755200714} +{"step": 202, "steps": 250, "lr": 9.695623390932612e-06, "grad_norm": 0.02618861012160778, "loss_scale": 4.0, "retries": 0, "seconds": 104.6, "max_memory_gb": 18.9, "ce": 0.001129636897076125, "kl": 0.002711456433701187, "context": 0.002485365228039882} +{"step": 203, "steps": 250, "lr": 9.51253782342123e-06, "grad_norm": 0.5212881565093994, "loss_scale": 4.0, "retries": 0, "seconds": 105.1, "max_memory_gb": 18.9, "ce": 0.015413295579492114, "kl": 0.036313691933173686, "context": 0.03357014202629216} +{"step": 204, "steps": 250, "lr": 9.332694393196846e-06, "grad_norm": 0.10553821176290512, "loss_scale": 4.0, "retries": 0, "seconds": 105.6, "max_memory_gb": 18.9, "ce": 0.029062301968224347, "kl": 0.013561733008828014, "context": 0.03584316815249622} +{"step": 205, "steps": 250, "lr": 9.156125516010256e-06, "grad_norm": 0.19816187024116516, "loss_scale": 4.0, "retries": 0, "seconds": 106.1, "max_memory_gb": 18.9, "ce": 0.0012313841652940027, "kl": 0.6471413736217073, "context": 0.3248020609753439} +{"step": 206, "steps": 250, "lr": 8.982863017392733e-06, "grad_norm": 0.03645668178796768, "loss_scale": 4.0, "retries": 0, "seconds": 106.7, "max_memory_gb": 18.9, "ce": 0.005637440283408068, "kl": 0.0027462861227149915, "context": 0.00701058359209128} +{"step": 207, "steps": 250, "lr": 8.81293812691974e-06, "grad_norm": 0.10849815607070923, "loss_scale": 4.0, "retries": 0, "seconds": 107.2, "max_memory_gb": 18.9, "ce": 0.05235523470400949, "kl": 0.07746154823689722, "context": 0.09108600884792395} +{"step": 208, "steps": 250, "lr": 8.646381472581927e-06, "grad_norm": 0.04621731862425804, "loss_scale": 4.0, "retries": 0, "seconds": 107.7, "max_memory_gb": 18.9, "ce": 0.003396777119633043, "kl": 0.0019310610587126575, "context": 0.004362307648989372} +{"step": 209, "steps": 250, "lr": 8.483223075264622e-06, "grad_norm": 0.5562772154808044, "loss_scale": 4.0, "retries": 0, "seconds": 108.2, "max_memory_gb": 18.9, "ce": 0.13017253625730518, "kl": 0.17936423387072864, "context": 0.2198546524887206} +{"step": 210, "steps": 250, "lr": 8.323492343336747e-06, "grad_norm": 0.833697497844696, "loss_scale": 4.0, "retries": 0, "seconds": 108.7, "max_memory_gb": 18.9, "ce": 0.49371422678950694, "kl": 0.1720695651265487, "context": 0.5797490223621935} +{"step": 211, "steps": 250, "lr": 8.167218067350122e-06, "grad_norm": 0.26130619645118713, "loss_scale": 4.0, "retries": 0, "seconds": 109.2, "max_memory_gb": 18.9, "ce": 0.06848285777959973, "kl": 0.21473319749929942, "context": 0.17584945261478424} +{"step": 212, "steps": 250, "lr": 8.01442841485013e-06, "grad_norm": 0.30888688564300537, "loss_scale": 4.0, "retries": 0, "seconds": 109.7, "max_memory_gb": 18.9, "ce": 0.0130164111033082, "kl": 1.1525760027579963, "context": 0.5893044057302177} +{"step": 213, "steps": 250, "lr": 7.865150925298678e-06, "grad_norm": 0.38497477769851685, "loss_scale": 4.0, "retries": 0, "seconds": 110.2, "max_memory_gb": 18.9, "ce": 0.21829724741110113, "kl": 0.07414165265799966, "context": 0.255368069956603} +{"step": 214, "steps": 250, "lr": 7.719412505110375e-06, "grad_norm": 0.5274122357368469, "loss_scale": 4.0, "retries": 0, "seconds": 110.8, "max_memory_gb": 18.9, "ce": 0.48676197799795773, "kl": 0.20962539844913408, "context": 0.5915746574319201} +{"step": 215, "steps": 250, "lr": 7.57723942280278e-06, "grad_norm": 1.0787267684936523, "loss_scale": 4.0, "retries": 0, "seconds": 111.3, "max_memory_gb": 18.9, "ce": 0.6619613019174722, "kl": 0.02875064543411554, "context": 0.6763366320847126} +{"step": 216, "steps": 250, "lr": 7.438657304261687e-06, "grad_norm": 0.5038290619850159, "loss_scale": 4.0, "retries": 0, "seconds": 111.8, "max_memory_gb": 18.9, "ce": 0.023289493927222793, "kl": 0.1702736260012898, "context": 0.10842630879415083} +{"step": 217, "steps": 250, "lr": 7.30369112812219e-06, "grad_norm": 0.015605837106704712, "loss_scale": 4.0, "retries": 0, "seconds": 112.3, "max_memory_gb": 18.9, "ce": 0.0007389109377982095, "kl": 0.0036901253242831444, "context": 0.0025839737318165135} +{"step": 218, "steps": 250, "lr": 7.172365221266403e-06, "grad_norm": 0.6746407747268677, "loss_scale": 4.0, "retries": 0, "seconds": 112.8, "max_memory_gb": 18.9, "ce": 0.1232754229567945, "kl": 0.316683204844594, "context": 0.28161703050136566} +{"step": 219, "steps": 250, "lr": 7.0447032544387154e-06, "grad_norm": 0.7316732406616211, "loss_scale": 4.0, "retries": 0, "seconds": 113.3, "max_memory_gb": 18.9, "ce": 0.00603720545223041, "kl": 0.18420029794651782, "context": 0.09813735759234987} +{"step": 220, "steps": 250, "lr": 6.920728237979274e-06, "grad_norm": 0.09516873955726624, "loss_scale": 4.0, "retries": 0, "seconds": 113.8, "max_memory_gb": 18.9, "ce": 0.053841171986277914, "kl": 0.006749897277245509, "context": 0.05721612249362806} +{"step": 221, "steps": 250, "lr": 6.800462517676456e-06, "grad_norm": 0.27868616580963135, "loss_scale": 4.0, "retries": 0, "seconds": 114.3, "max_memory_gb": 18.9, "ce": 0.0035211716894991696, "kl": 0.6531330618890934, "context": 0.33008770167361945} +{"step": 222, "steps": 250, "lr": 6.683927770739222e-06, "grad_norm": 0.1526394784450531, "loss_scale": 4.0, "retries": 0, "seconds": 114.8, "max_memory_gb": 18.9, "ce": 0.03262870054459199, "kl": 0.07715457774060042, "context": 0.07120599130576011} +{"step": 223, "steps": 250, "lr": 6.571145001889892e-06, "grad_norm": 0.050615373998880386, "loss_scale": 4.0, "retries": 0, "seconds": 115.4, "max_memory_gb": 18.9, "ce": 0.0049331340997014195, "kl": 0.0004086665067006834, "context": 0.005137467393069528} +{"step": 224, "steps": 250, "lr": 6.46213453957817e-06, "grad_norm": 0.2917841970920563, "loss_scale": 4.0, "retries": 0, "seconds": 115.9, "max_memory_gb": 18.9, "ce": 0.007510998490033671, "kl": 0.27060433051701693, "context": 0.14281316456617787} +{"step": 225, "steps": 250, "lr": 6.356916032317063e-06, "grad_norm": 0.16860336065292358, "loss_scale": 4.0, "retries": 0, "seconds": 116.4, "max_memory_gb": 18.9, "ce": 0.0029891380545450374, "kl": 0.29247627810400445, "context": 0.14922727211524034} +{"step": 226, "steps": 250, "lr": 6.255508445141349e-06, "grad_norm": 1.7404640913009644, "loss_scale": 4.0, "retries": 0, "seconds": 116.9, "max_memory_gb": 18.9, "ce": 0.20475996096502058, "kl": 0.09697077388409525, "context": 0.2532453438907396} +{"step": 227, "steps": 250, "lr": 6.157930056189228e-06, "grad_norm": 0.32006022334098816, "loss_scale": 4.0, "retries": 0, "seconds": 117.4, "max_memory_gb": 18.9, "ce": 0.03414975168470846, "kl": 0.0064719965139374835, "context": 0.03738574889393931} +{"step": 228, "steps": 250, "lr": 6.064198453407796e-06, "grad_norm": 0.5106392502784729, "loss_scale": 4.0, "retries": 0, "seconds": 117.9, "max_memory_gb": 18.9, "ce": 0.001849978321843082, "kl": 0.5540874825383071, "context": 0.27889372278878} +{"step": 229, "steps": 250, "lr": 5.974330531382935e-06, "grad_norm": 0.36453601717948914, "loss_scale": 4.0, "retries": 0, "seconds": 118.4, "max_memory_gb": 18.9, "ce": 0.059651681813193136, "kl": 0.010095420939023825, "context": 0.06469939041198813} +{"step": 230, "steps": 250, "lr": 5.888342488294125e-06, "grad_norm": 2.380955934524536, "loss_scale": 4.0, "retries": 0, "seconds": 118.9, "max_memory_gb": 18.9, "ce": 0.46124749607406557, "kl": 0.6756113603769336, "context": 0.7990531483228551} +{"step": 231, "steps": 250, "lr": 5.806249822994849e-06, "grad_norm": 0.1453208476305008, "loss_scale": 4.0, "retries": 0, "seconds": 119.4, "max_memory_gb": 18.9, "ce": 0.021606644202620373, "kl": 0.14553453597181942, "context": 0.09437391134270001} +{"step": 232, "steps": 250, "lr": 5.728067332218994e-06, "grad_norm": 0.11891420185565948, "loss_scale": 4.0, "retries": 0, "seconds": 119.9, "max_memory_gb": 18.9, "ce": 0.018341325980145484, "kl": 0.020755715551786125, "context": 0.02871918270830065} +{"step": 233, "steps": 250, "lr": 5.65380910791383e-06, "grad_norm": 0.4520490765571594, "loss_scale": 4.0, "retries": 0, "seconds": 120.4, "max_memory_gb": 18.9, "ce": 0.008189158223103732, "kl": 0.17566660372540355, "context": 0.09602246090071276} +{"step": 234, "steps": 250, "lr": 5.583488534700022e-06, "grad_norm": 0.46901869773864746, "loss_scale": 4.0, "retries": 0, "seconds": 120.9, "max_memory_gb": 18.9, "ce": 0.015337582713982556, "kl": 0.15264665048744064, "context": 0.09166090885992162} +{"step": 235, "steps": 250, "lr": 5.517118287459118e-06, "grad_norm": 1.251568078994751, "loss_scale": 4.0, "retries": 0, "seconds": 121.4, "max_memory_gb": 18.9, "ce": 0.36951275294995867, "kl": 0.0682515601401974, "context": 0.4036385478611919} +{"step": 236, "steps": 250, "lr": 5.454710329048978e-06, "grad_norm": 0.038255076855421066, "loss_scale": 4.0, "retries": 0, "seconds": 121.9, "max_memory_gb": 18.9, "ce": 0.005662986498464306, "kl": 0.014127179521892685, "context": 0.012726576289423974} +{"step": 237, "steps": 250, "lr": 5.3962759081475436e-06, "grad_norm": 0.5643749833106995, "loss_scale": 4.0, "retries": 0, "seconds": 122.4, "max_memory_gb": 18.9, "ce": 0.0852058190612297, "kl": 0.053516092220434075, "context": 0.11196386168012396} +{"step": 238, "steps": 250, "lr": 5.341825557225321e-06, "grad_norm": 0.18103115260601044, "loss_scale": 4.0, "retries": 0, "seconds": 122.9, "max_memory_gb": 18.9, "ce": 0.024069887993391603, "kl": 0.0132474588790501, "context": 0.030693617125507444} +{"step": 239, "steps": 250, "lr": 5.291369090646961e-06, "grad_norm": 0.6106793880462646, "loss_scale": 4.0, "retries": 0, "seconds": 123.4, "max_memory_gb": 18.9, "ce": 0.10470721342426259, "kl": 0.03572518959117588, "context": 0.12256980442907661} +{"step": 240, "steps": 250, "lr": 5.244915602902289e-06, "grad_norm": 0.19120323657989502, "loss_scale": 4.0, "retries": 0, "seconds": 123.9, "max_memory_gb": 18.9, "ce": 0.1144380296318559, "kl": 0.045730036867098534, "context": 0.13730305003991816} +{"step": 241, "steps": 250, "lr": 5.20247346696706e-06, "grad_norm": 0.4426332414150238, "loss_scale": 4.0, "retries": 0, "seconds": 124.5, "max_memory_gb": 18.9, "ce": 0.07245563185279025, "kl": 0.034429699491738575, "context": 0.08967048206432082} +{"step": 242, "steps": 250, "lr": 5.164050332793787e-06, "grad_norm": 0.09602326899766922, "loss_scale": 4.0, "retries": 0, "seconds": 125.0, "max_memory_gb": 18.9, "ce": 0.0025339525327581214, "kl": 0.08439960592886564, "context": 0.04473375596171536} +{"step": 243, "steps": 250, "lr": 5.129653125932873e-06, "grad_norm": 0.11480620503425598, "loss_scale": 4.0, "retries": 0, "seconds": 125.5, "max_memory_gb": 18.9, "ce": 0.008107792506052647, "kl": 0.009992014744057087, "context": 0.01310379985079635} +{"step": 244, "steps": 250, "lr": 5.099288046284329e-06, "grad_norm": 1.725233554840088, "loss_scale": 4.0, "retries": 0, "seconds": 126.0, "max_memory_gb": 18.9, "ce": 0.05714576208265498, "kl": 0.20412181282154052, "context": 0.15920666477177292} +{"step": 245, "steps": 250, "lr": 5.072960566980277e-06, "grad_norm": 1.9835269451141357, "loss_scale": 4.0, "retries": 0, "seconds": 126.5, "max_memory_gb": 18.9, "ce": 0.36059736367315054, "kl": 0.06511531746946275, "context": 0.3931550197303295} +{"step": 246, "steps": 250, "lr": 5.050675433398435e-06, "grad_norm": 0.023404503241181374, "loss_scale": 4.0, "retries": 0, "seconds": 127.0, "max_memory_gb": 18.9, "ce": 0.0006721935096720699, "kl": 0.0025010072049553855, "context": 0.0019226971271564253} +{"step": 247, "steps": 250, "lr": 5.032436662306811e-06, "grad_norm": 1.845417857170105, "loss_scale": 4.0, "retries": 0, "seconds": 127.5, "max_memory_gb": 18.9, "ce": 0.05182857988165779, "kl": 0.6563408559773052, "context": 0.3799989995895885} +{"step": 248, "steps": 250, "lr": 5.018247541139684e-06, "grad_norm": 0.7408720850944519, "loss_scale": 4.0, "retries": 0, "seconds": 128.0, "max_memory_gb": 18.9, "ce": 0.027412584131525364, "kl": 0.06665971304755658, "context": 0.06074243946932256} +{"step": 249, "steps": 250, "lr": 5.00811062740508e-06, "grad_norm": 0.17499426007270813, "loss_scale": 4.0, "retries": 0, "seconds": 128.5, "max_memory_gb": 18.9, "ce": 0.0254914466640912, "kl": 0.031207602529320866, "context": 0.04109524900559336} +{"step": 250, "steps": 250, "lr": 5.002027748223778e-06, "grad_norm": 0.03326103463768959, "loss_scale": 4.0, "retries": 0, "seconds": 129.0, "max_memory_gb": 18.9, "ce": 0.0037128091644262895, "kl": 0.0027201432258152636, "context": 0.005072880630905274} diff --git a/runs/single_internvl/context/L07_s13/adapter/README.md b/runs/single_internvl/context/L07_s13/adapter/README.md new file mode 100644 index 0000000000000000000000000000000000000000..4f9f856cf0e1531b47e25d266833b2977a19d3ac --- /dev/null +++ b/runs/single_internvl/context/L07_s13/adapter/README.md @@ -0,0 +1,207 @@ +--- +base_model: /data/OPSD/assets/models/InternVL3_5-8B-HF +library_name: peft +pipeline_tag: text-generation +tags: +- base_model:adapter:/data/OPSD/assets/models/InternVL3_5-8B-HF +- lora +- transformers +--- + +# Model Card for Model ID + + + + + +## Model Details + +### Model Description + + + + + +- **Developed by:** [More Information Needed] +- **Funded by [optional]:** [More Information Needed] +- **Shared by [optional]:** [More Information Needed] +- **Model type:** [More Information Needed] +- **Language(s) (NLP):** [More Information Needed] +- **License:** [More Information Needed] +- **Finetuned from model [optional]:** [More Information Needed] + +### Model Sources [optional] + + + +- **Repository:** [More Information Needed] +- **Paper [optional]:** [More Information Needed] +- **Demo [optional]:** [More Information Needed] + +## Uses + + + +### Direct Use + + + +[More Information Needed] + +### Downstream Use [optional] + + + +[More Information Needed] + +### Out-of-Scope Use + + + +[More Information Needed] + +## Bias, Risks, and Limitations + + + +[More Information Needed] + +### Recommendations + + + +Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. More information needed for further recommendations. + +## How to Get Started with the Model + +Use the code below to get started with the model. + +[More Information Needed] + +## Training Details + +### Training Data + + + +[More Information Needed] + +### Training Procedure + + + +#### Preprocessing [optional] + +[More Information Needed] + + +#### Training Hyperparameters + +- **Training regime:** [More Information Needed] + +#### Speeds, Sizes, Times [optional] + + + +[More Information Needed] + +## Evaluation + + + +### Testing Data, Factors & Metrics + +#### Testing Data + + + +[More Information Needed] + +#### Factors + + + +[More Information Needed] + +#### Metrics + + + +[More Information Needed] + +### Results + +[More Information Needed] + +#### Summary + + + +## Model Examination [optional] + + + +[More Information Needed] + +## Environmental Impact + + + +Carbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700). + +- **Hardware Type:** [More Information Needed] +- **Hours used:** [More Information Needed] +- **Cloud Provider:** [More Information Needed] +- **Compute Region:** [More Information Needed] +- **Carbon Emitted:** [More Information Needed] + +## Technical Specifications [optional] + +### Model Architecture and Objective + +[More Information Needed] + +### Compute Infrastructure + +[More Information Needed] + +#### Hardware + +[More Information Needed] + +#### Software + +[More Information Needed] + +## Citation [optional] + + + +**BibTeX:** + +[More Information Needed] + +**APA:** + +[More Information Needed] + +## Glossary [optional] + + + +[More Information Needed] + +## More Information [optional] + +[More Information Needed] + +## Model Card Authors [optional] + +[More Information Needed] + +## Model Card Contact + +[More Information Needed] +### Framework versions + +- PEFT 0.20.0 \ No newline at end of file diff --git a/runs/single_internvl/context/L07_s13/adapter/adapter_config.json b/runs/single_internvl/context/L07_s13/adapter/adapter_config.json new file mode 100644 index 0000000000000000000000000000000000000000..7412b0ac72b7f39979e7cd1b4d5f34f09a0201c7 --- /dev/null +++ b/runs/single_internvl/context/L07_s13/adapter/adapter_config.json @@ -0,0 +1,50 @@ +{ + "alora_invocation_tokens": null, + "alpha_pattern": {}, + "arrow_config": null, + "auto_mapping": null, + "base_model_name_or_path": "/data/OPSD/assets/models/InternVL3_5-8B-HF", + "bias": "none", + "corda_config": null, + "ensure_weight_tying": false, + "eva_config": null, + "exclude_modules": null, + "fan_in_fan_out": false, + "inference_mode": true, + "init_lora_weights": true, + "layer_replication": null, + "layers_pattern": null, + "layers_to_transform": null, + "loftq_config": {}, + "lora_alpha": 32, + "lora_bias": false, + "lora_dropout": 0.05, + "lora_ga_config": null, + "megatron_config": null, + "megatron_core": "megatron.core", + "modules_to_save": null, + "monteclora_config": null, + "peft_type": "LORA", + "peft_version": "0.20.0", + "qalora_group_size": 16, + "r": 16, + "rank_pattern": {}, + "revision": null, + "target_modules": [ + "model.language_model.layers.7.self_attn.q_proj", + "model.language_model.layers.7.self_attn.o_proj", + "model.language_model.layers.7.mlp.gate_proj", + "model.language_model.layers.7.mlp.up_proj", + "model.language_model.layers.7.self_attn.v_proj", + "model.language_model.layers.7.mlp.down_proj", + "model.language_model.layers.7.self_attn.k_proj" + ], + "target_parameters": null, + "task_type": "CAUSAL_LM", + "trainable_token_indices": null, + "use_bdlora": null, + "use_dora": false, + "use_qalora": false, + "use_rslora": false, + "velora_config": null +} \ No newline at end of file diff --git a/runs/single_internvl/context/L07_s13/config.json b/runs/single_internvl/context/L07_s13/config.json new file mode 100644 index 0000000000000000000000000000000000000000..ba0a23f65f0782f403bfc7bd636abb34aa8b0596 --- /dev/null +++ b/runs/single_internvl/context/L07_s13/config.json @@ -0,0 +1,24 @@ +{ + "behavior": "context", + "seed": 13, + "model": "/data/OPSD/assets/models/InternVL3_5-8B-HF", + "backbone": "internvl", + "train": "/data/OPSD/assets/v3/context.jsonl", + "items": 1000, + "steps": 250, + "epochs": 1, + "steps_per_epoch": 250, + "groups_per_step": 4, + "lr": 5e-05, + "scheduler": "cosine", + "warmup_steps": 16, + "final_fraction": 0.1, + "weight_decay": 0.0, + "clip": 1.0, + "initial_loss_scale": 4.0, + "adapter": { + "layer": 7, + "rank": 16, + "trainable_parameters": 1212416 + } +} \ No newline at end of file diff --git a/runs/single_internvl/context/L07_s13/steps.jsonl b/runs/single_internvl/context/L07_s13/steps.jsonl new file mode 100644 index 0000000000000000000000000000000000000000..655debdfad970a5b5d88131f684b9e3fdbf286bd --- /dev/null +++ b/runs/single_internvl/context/L07_s13/steps.jsonl @@ -0,0 +1,250 @@ +{"step": 1, "steps": 250, "lr": 3.125e-06, "grad_norm": 0.8093783259391785, "loss_scale": 4.0, "retries": 0, "seconds": 1.6, "max_memory_gb": 18.45, "ce": 0.5227766726748087, "kl": 0.08654405153356493, "context": 0.5660487064160407} +{"step": 2, "steps": 250, "lr": 6.25e-06, "grad_norm": 1.1118857860565186, "loss_scale": 4.0, "retries": 0, "seconds": 2.1, "max_memory_gb": 18.46, "ce": 0.6394089035893558, "kl": 0.46732023550248414, "context": 0.8730690349329961} +{"step": 3, "steps": 250, "lr": 9.375000000000001e-06, "grad_norm": 3.2022171020507812, "loss_scale": 4.0, "retries": 0, "seconds": 2.5, "max_memory_gb": 18.46, "ce": 1.0957384780049324, "kl": 0.06678311806172132, "context": 1.129130057990551} +{"step": 4, "steps": 250, "lr": 1.25e-05, "grad_norm": 1.6975140571594238, "loss_scale": 4.0, "retries": 0, "seconds": 3.0, "max_memory_gb": 18.46, "ce": 0.44870465776330093, "kl": 0.4310474635644823, "context": 0.6642283752880758} +{"step": 5, "steps": 250, "lr": 1.5625e-05, "grad_norm": 0.03777274489402771, "loss_scale": 4.0, "retries": 0, "seconds": 3.5, "max_memory_gb": 18.46, "ce": 0.009266586348530836, "kl": 0.00829822193554719, "context": 0.01341569719807012} +{"step": 6, "steps": 250, "lr": 1.8750000000000002e-05, "grad_norm": 0.8053039312362671, "loss_scale": 4.0, "retries": 0, "seconds": 4.0, "max_memory_gb": 18.46, "ce": 0.28601379091560375, "kl": 0.07441438977548387, "context": 0.3232209905690979} +{"step": 7, "steps": 250, "lr": 2.1875e-05, "grad_norm": 2.581650733947754, "loss_scale": 4.0, "retries": 0, "seconds": 4.5, "max_memory_gb": 18.46, "ce": 0.7955253474065103, "kl": 0.12379652759568671, "context": 0.8574236180720618} +{"step": 8, "steps": 250, "lr": 2.5e-05, "grad_norm": 2.575434446334839, "loss_scale": 4.0, "retries": 0, "seconds": 5.1, "max_memory_gb": 18.47, "ce": 0.7383018843829632, "kl": 0.2866892742458731, "context": 0.8816465176641941} +{"step": 9, "steps": 250, "lr": 2.8125000000000003e-05, "grad_norm": 1.8268576860427856, "loss_scale": 4.0, "retries": 0, "seconds": 5.6, "max_memory_gb": 18.47, "ce": 0.48064903542399406, "kl": 0.13802096381550655, "context": 0.5496595203876495} +{"step": 10, "steps": 250, "lr": 3.125e-05, "grad_norm": 0.31950241327285767, "loss_scale": 4.0, "retries": 0, "seconds": 6.1, "max_memory_gb": 18.47, "ce": 0.04567332015722059, "kl": 0.04456141625996679, "context": 0.06795402627903968} +{"step": 11, "steps": 250, "lr": 3.4375e-05, "grad_norm": 2.13383150100708, "loss_scale": 4.0, "retries": 0, "seconds": 6.7, "max_memory_gb": 18.47, "ce": 0.6065322450667736, "kl": 0.004701573754118726, "context": 0.6088830335138482} +{"step": 12, "steps": 250, "lr": 3.7500000000000003e-05, "grad_norm": 1.6697543859481812, "loss_scale": 4.0, "retries": 0, "seconds": 7.2, "max_memory_gb": 18.47, "ce": 0.3901472469569853, "kl": 0.0030975250083429273, "context": 0.39169600673812965} +{"step": 13, "steps": 250, "lr": 4.0625000000000005e-05, "grad_norm": 1.2018824815750122, "loss_scale": 4.0, "retries": 0, "seconds": 7.8, "max_memory_gb": 18.47, "ce": 0.35798608777986374, "kl": 0.018559971811555442, "context": 0.36726607569653424} +{"step": 14, "steps": 250, "lr": 4.375e-05, "grad_norm": 2.4499354362487793, "loss_scale": 4.0, "retries": 0, "seconds": 8.3, "max_memory_gb": 18.47, "ce": 0.5771607309579849, "kl": 0.9421110753901303, "context": 1.0482162982225418} +{"step": 15, "steps": 250, "lr": 4.6875e-05, "grad_norm": 2.056537628173828, "loss_scale": 4.0, "retries": 0, "seconds": 8.8, "max_memory_gb": 18.47, "ce": 0.33719271530935657, "kl": 0.44883409085105086, "context": 0.5616097756433192} +{"step": 16, "steps": 250, "lr": 5e-05, "grad_norm": 0.7096481919288635, "loss_scale": 4.0, "retries": 0, "seconds": 9.3, "max_memory_gb": 18.47, "ce": 0.145615695269953, "kl": 0.001062145550349669, "context": 0.1461467708068085} +{"step": 17, "steps": 250, "lr": 5e-05, "grad_norm": 0.8859085440635681, "loss_scale": 4.0, "retries": 0, "seconds": 9.8, "max_memory_gb": 18.47, "ce": 0.2287517465883866, "kl": 0.8331494017038494, "context": 0.6453264351002872} +{"step": 18, "steps": 250, "lr": 4.999797225177623e-05, "grad_norm": 1.9893585443496704, "loss_scale": 4.0, "retries": 0, "seconds": 10.3, "max_memory_gb": 18.47, "ce": 0.24913179549912456, "kl": 0.2164157670995337, "context": 0.3573396782448981} +{"step": 19, "steps": 250, "lr": 4.999188937259492e-05, "grad_norm": 2.642822504043579, "loss_scale": 4.0, "retries": 0, "seconds": 10.8, "max_memory_gb": 18.47, "ce": 0.26016734540462494, "kl": 0.18805995909497142, "context": 0.35419732332229614} +{"step": 20, "steps": 250, "lr": 4.998175245886032e-05, "grad_norm": 0.05733498930931091, "loss_scale": 4.0, "retries": 0, "seconds": 11.3, "max_memory_gb": 18.47, "ce": 0.0029466383766703075, "kl": 0.020194954404360033, "context": 0.013044115505181253} +{"step": 21, "steps": 250, "lr": 4.996756333769319e-05, "grad_norm": 1.172355055809021, "loss_scale": 4.0, "retries": 0, "seconds": 11.8, "max_memory_gb": 18.47, "ce": 0.07792943335152813, "kl": 0.02476819167713984, "context": 0.0903135278076661} +{"step": 22, "steps": 250, "lr": 4.994932456660157e-05, "grad_norm": 0.9743797779083252, "loss_scale": 4.0, "retries": 0, "seconds": 12.3, "max_memory_gb": 18.47, "ce": 0.1020221168291755, "kl": 0.03452502120126155, "context": 0.11928463150979951} +{"step": 23, "steps": 250, "lr": 4.992703943301973e-05, "grad_norm": 1.8847167491912842, "loss_scale": 4.0, "retries": 0, "seconds": 12.9, "max_memory_gb": 18.47, "ce": 0.2852041699597976, "kl": 0.9630412208393864, "context": 0.7667247889357895} +{"step": 24, "steps": 250, "lr": 4.9900711953715675e-05, "grad_norm": 1.0312761068344116, "loss_scale": 4.0, "retries": 0, "seconds": 13.4, "max_memory_gb": 18.47, "ce": 0.07200741674751043, "kl": 0.1308149395044893, "context": 0.1374148833565414} +{"step": 25, "steps": 250, "lr": 4.987034687406713e-05, "grad_norm": 1.0321141481399536, "loss_scale": 4.0, "retries": 0, "seconds": 13.9, "max_memory_gb": 18.47, "ce": 0.3655803151195869, "kl": 0.14621682092547417, "context": 0.4386887401342392} +{"step": 26, "steps": 250, "lr": 4.9835949667206215e-05, "grad_norm": 0.45271408557891846, "loss_scale": 4.0, "retries": 0, "seconds": 14.4, "max_memory_gb": 18.47, "ce": 0.04713961202651262, "kl": 0.6653379362542182, "context": 0.37980858981609344} +{"step": 27, "steps": 250, "lr": 4.9797526533032945e-05, "grad_norm": 0.5400599241256714, "loss_scale": 4.0, "retries": 0, "seconds": 14.9, "max_memory_gb": 18.47, "ce": 0.10073327025747858, "kl": 0.34753570626708097, "context": 0.2745011308143148} +{"step": 28, "steps": 250, "lr": 4.9755084397097716e-05, "grad_norm": 0.7054867744445801, "loss_scale": 4.0, "retries": 0, "seconds": 15.4, "max_memory_gb": 18.47, "ce": 0.49164121877402067, "kl": 0.9566515039186925, "context": 0.969966946169734} +{"step": 29, "steps": 250, "lr": 4.970863090935304e-05, "grad_norm": 0.37614408135414124, "loss_scale": 4.0, "retries": 0, "seconds": 15.9, "max_memory_gb": 18.47, "ce": 0.04852468287572265, "kl": 0.39422000414924696, "context": 0.24563468131236732} +{"step": 30, "steps": 250, "lr": 4.965817444277468e-05, "grad_norm": 0.5022361874580383, "loss_scale": 4.0, "retries": 0, "seconds": 16.4, "max_memory_gb": 18.47, "ce": 0.057547370855900226, "kl": 0.0017324871332675684, "context": 0.05841361268721812} +{"step": 31, "steps": 250, "lr": 4.9603724091852456e-05, "grad_norm": 0.7347604036331177, "loss_scale": 4.0, "retries": 0, "seconds": 17.0, "max_memory_gb": 18.47, "ce": 0.21313571278005838, "kl": 0.09969231765717268, "context": 0.2629818692803383} +{"step": 32, "steps": 250, "lr": 4.9545289670951025e-05, "grad_norm": 0.3916105628013611, "loss_scale": 4.0, "retries": 0, "seconds": 17.5, "max_memory_gb": 18.47, "ce": 0.027991937240585685, "kl": 0.012173057330073789, "context": 0.03407846577465534} +{"step": 33, "steps": 250, "lr": 4.948288171254089e-05, "grad_norm": 0.36052554845809937, "loss_scale": 4.0, "retries": 0, "seconds": 18.0, "max_memory_gb": 18.47, "ce": 0.04952889960259199, "kl": 0.026749102980829775, "context": 0.06290345080196857} +{"step": 34, "steps": 250, "lr": 4.941651146529998e-05, "grad_norm": 0.47559359669685364, "loss_scale": 4.0, "retries": 0, "seconds": 18.5, "max_memory_gb": 18.47, "ce": 0.009776634899026249, "kl": 0.7948973570796625, "context": 0.4072253269441717} +{"step": 35, "steps": 250, "lr": 4.9346190892086174e-05, "grad_norm": 0.15000222623348236, "loss_scale": 4.0, "retries": 0, "seconds": 19.0, "max_memory_gb": 18.47, "ce": 0.009216489386744797, "kl": 0.002456795466969197, "context": 0.010444886982440948} +{"step": 36, "steps": 250, "lr": 4.927193266778101e-05, "grad_norm": 0.2670605480670929, "loss_scale": 4.0, "retries": 0, "seconds": 19.5, "max_memory_gb": 18.47, "ce": 0.09199556682142429, "kl": 0.05818757432280108, "context": 0.12108935706783086} +{"step": 37, "steps": 250, "lr": 4.919375017700515e-05, "grad_norm": 0.500725507736206, "loss_scale": 4.0, "retries": 0, "seconds": 20.1, "max_memory_gb": 18.47, "ce": 0.14659166455385275, "kl": 0.802193233394064, "context": 0.5476882910588756} +{"step": 38, "steps": 250, "lr": 4.911165751170588e-05, "grad_norm": 0.3101888597011566, "loss_scale": 4.0, "retries": 0, "seconds": 20.6, "max_memory_gb": 18.47, "ce": 0.023721180390566587, "kl": 0.5037924367934465, "context": 0.27561740204691887} +{"step": 39, "steps": 250, "lr": 4.902566946861708e-05, "grad_norm": 0.4437806308269501, "loss_scale": 4.0, "retries": 0, "seconds": 21.1, "max_memory_gb": 18.47, "ce": 0.11287622080999427, "kl": 0.011158273009641562, "context": 0.11845535911561456} +{"step": 40, "steps": 250, "lr": 4.89358015465922e-05, "grad_norm": 0.13431020081043243, "loss_scale": 4.0, "retries": 0, "seconds": 21.6, "max_memory_gb": 18.47, "ce": 0.010678554186597466, "kl": 0.008047391718719155, "context": 0.014702249580295756} +{"step": 41, "steps": 250, "lr": 4.884206994381078e-05, "grad_norm": 1.1641845703125, "loss_scale": 4.0, "retries": 0, "seconds": 22.0, "max_memory_gb": 18.47, "ce": 0.392167154350318, "kl": 0.12804345809854567, "context": 0.45618888223543763} +{"step": 42, "steps": 250, "lr": 4.874449155485866e-05, "grad_norm": 0.02337644062936306, "loss_scale": 4.0, "retries": 0, "seconds": 22.5, "max_memory_gb": 18.47, "ce": 0.00304308521663188, "kl": 0.0004328434697526973, "context": 0.003259506876929663} +{"step": 43, "steps": 250, "lr": 4.864308396768294e-05, "grad_norm": 0.04946286231279373, "loss_scale": 4.0, "retries": 0, "seconds": 23.0, "max_memory_gb": 18.47, "ce": 0.0026147825774387456, "kl": 0.0032583025604253635, "context": 0.004243933784891851} +{"step": 44, "steps": 250, "lr": 4.853786546042184e-05, "grad_norm": 2.9122231006622314, "loss_scale": 4.0, "retries": 0, "seconds": 23.5, "max_memory_gb": 18.47, "ce": 0.5569813381443964, "kl": 1.4045178340747952, "context": 1.259240238883649} +{"step": 45, "steps": 250, "lr": 4.842885499811012e-05, "grad_norm": 1.198068618774414, "loss_scale": 4.0, "retries": 0, "seconds": 24.0, "max_memory_gb": 18.47, "ce": 0.8597203533281572, "kl": 0.2382812823052518, "context": 0.9788610105460975} +{"step": 46, "steps": 250, "lr": 4.831607222926078e-05, "grad_norm": 0.5408453941345215, "loss_scale": 4.0, "retries": 0, "seconds": 24.5, "max_memory_gb": 18.47, "ce": 0.08630128490040079, "kl": 0.2700957402703352, "context": 0.22134915925562382} +{"step": 47, "steps": 250, "lr": 4.8199537482323545e-05, "grad_norm": 0.06763318181037903, "loss_scale": 4.0, "retries": 0, "seconds": 24.9, "max_memory_gb": 18.47, "ce": 0.005557872689678334, "kl": 0.014055873092729598, "context": 0.012585808843141422} +{"step": 48, "steps": 250, "lr": 4.807927176202073e-05, "grad_norm": 0.11076442152261734, "loss_scale": 4.0, "retries": 0, "seconds": 25.4, "max_memory_gb": 18.47, "ce": 0.005141554705915041, "kl": 0.0031191364396363497, "context": 0.006701123085804284} +{"step": 49, "steps": 250, "lr": 4.795529674556129e-05, "grad_norm": 0.3132251501083374, "loss_scale": 4.0, "retries": 0, "seconds": 25.8, "max_memory_gb": 18.47, "ce": 0.01125720732670743, "kl": 0.3336299625659649, "context": 0.1780721851173439} +{"step": 50, "steps": 250, "lr": 4.782763477873361e-05, "grad_norm": 0.4029938280582428, "loss_scale": 4.0, "retries": 0, "seconds": 26.3, "max_memory_gb": 18.47, "ce": 0.09431529464927735, "kl": 0.008365330199012533, "context": 0.09849795841000741} +{"step": 51, "steps": 250, "lr": 4.769630887187782e-05, "grad_norm": 46.715660095214844, "loss_scale": 4.0, "retries": 0, "seconds": 26.8, "max_memory_gb": 18.47, "ce": 0.6076984116989479, "kl": 0.021586551603377302, "context": 0.6184916753936704} +{"step": 52, "steps": 250, "lr": 4.756134269573832e-05, "grad_norm": 0.6578145027160645, "loss_scale": 4.0, "retries": 0, "seconds": 27.3, "max_memory_gb": 18.47, "ce": 0.02442845405312255, "kl": 0.26572012179531157, "context": 0.15728851174935699} +{"step": 53, "steps": 250, "lr": 4.7422760577197226e-05, "grad_norm": 0.7715049982070923, "loss_scale": 4.0, "retries": 0, "seconds": 27.8, "max_memory_gb": 18.47, "ce": 0.17113778955535963, "kl": 0.3719246219843626, "context": 0.3571001123636961} +{"step": 54, "steps": 250, "lr": 4.728058749488963e-05, "grad_norm": 0.5414842963218689, "loss_scale": 4.0, "retries": 0, "seconds": 28.3, "max_memory_gb": 18.47, "ce": 0.04047188819095027, "kl": 0.336333505998482, "context": 0.2086386407027021} +{"step": 55, "steps": 250, "lr": 4.713484907470133e-05, "grad_norm": 0.09809395670890808, "loss_scale": 4.0, "retries": 0, "seconds": 28.8, "max_memory_gb": 18.47, "ce": 0.016492908267537132, "kl": 0.013809049303119991, "context": 0.02339743234915659} +{"step": 56, "steps": 250, "lr": 4.6985571585149876e-05, "grad_norm": 0.2216639518737793, "loss_scale": 4.0, "retries": 0, "seconds": 29.3, "max_memory_gb": 18.47, "ce": 0.017594198463484645, "kl": 0.27544746291823685, "context": 0.15531793469563127} +{"step": 57, "steps": 250, "lr": 4.6832781932649884e-05, "grad_norm": 0.0746966153383255, "loss_scale": 4.0, "retries": 0, "seconds": 29.8, "max_memory_gb": 18.47, "ce": 0.009144820862275083, "kl": 0.018905126868048683, "context": 0.01859738428174751} +{"step": 58, "steps": 250, "lr": 4.667650765666325e-05, "grad_norm": 0.11966653168201447, "loss_scale": 4.0, "retries": 0, "seconds": 30.3, "max_memory_gb": 18.47, "ce": 0.005958992751402548, "kl": 0.0012491545649027103, "context": 0.006583570098882774} +{"step": 59, "steps": 250, "lr": 4.651677692473538e-05, "grad_norm": 0.22818322479724884, "loss_scale": 4.0, "retries": 0, "seconds": 30.8, "max_memory_gb": 18.47, "ce": 0.03305682296922896, "kl": 0.058545295119984075, "context": 0.062329469627002254} +{"step": 60, "steps": 250, "lr": 4.635361852741808e-05, "grad_norm": 0.2093343436717987, "loss_scale": 4.0, "retries": 0, "seconds": 31.3, "max_memory_gb": 18.47, "ce": 0.08145129279000685, "kl": 0.07202067828620784, "context": 0.11746162798954174} +{"step": 61, "steps": 250, "lr": 4.6187061873080264e-05, "grad_norm": 0.6857801675796509, "loss_scale": 4.0, "retries": 0, "seconds": 31.8, "max_memory_gb": 18.47, "ce": 0.019649143650894985, "kl": 0.9021352854178986, "context": 0.470716788462596} +{"step": 62, "steps": 250, "lr": 4.601713698260727e-05, "grad_norm": 0.16536346077919006, "loss_scale": 4.0, "retries": 0, "seconds": 32.3, "max_memory_gb": 18.47, "ce": 0.026928312188829295, "kl": 0.00955107274967304, "context": 0.03170384804252535} +{"step": 63, "steps": 250, "lr": 4.5843874483989746e-05, "grad_norm": 0.19693627953529358, "loss_scale": 4.0, "retries": 0, "seconds": 32.8, "max_memory_gb": 18.47, "ce": 0.0353465077932924, "kl": 0.12439606052066665, "context": 0.0975445406511426} +{"step": 64, "steps": 250, "lr": 4.566730560680316e-05, "grad_norm": 0.10437505692243576, "loss_scale": 4.0, "retries": 0, "seconds": 33.3, "max_memory_gb": 18.47, "ce": 0.013355272996705025, "kl": 0.01088431925745681, "context": 0.018797432770952582} +{"step": 65, "steps": 250, "lr": 4.548746217657878e-05, "grad_norm": 0.5225779414176941, "loss_scale": 4.0, "retries": 0, "seconds": 33.8, "max_memory_gb": 18.47, "ce": 0.18457683152519166, "kl": 0.15613333188230172, "context": 0.2626434895209968} +{"step": 66, "steps": 250, "lr": 4.5304376609067396e-05, "grad_norm": 0.7622052431106567, "loss_scale": 4.0, "retries": 0, "seconds": 34.4, "max_memory_gb": 18.47, "ce": 0.14555010496314935, "kl": 0.0686090927062537, "context": 0.1798546513745123} +{"step": 67, "steps": 250, "lr": 4.5118081904396504e-05, "grad_norm": 0.0734722763299942, "loss_scale": 4.0, "retries": 0, "seconds": 34.9, "max_memory_gb": 18.47, "ce": 0.004862985544605181, "kl": 0.008131194816087373, "context": 0.008928582887165248} +{"step": 68, "steps": 250, "lr": 4.4928611641122225e-05, "grad_norm": 0.15053582191467285, "loss_scale": 4.0, "retries": 0, "seconds": 35.4, "max_memory_gb": 18.47, "ce": 0.005566675354202744, "kl": 0.0426467159340973, "context": 0.02689003272462287} +{"step": 69, "steps": 250, "lr": 4.473599997017701e-05, "grad_norm": 0.6685784459114075, "loss_scale": 4.0, "retries": 0, "seconds": 35.9, "max_memory_gb": 18.47, "ce": 0.2949503767013084, "kl": 0.02357760170707479, "context": 0.3067391673102975} +{"step": 70, "steps": 250, "lr": 4.4540281608714053e-05, "grad_norm": 0.022214923053979874, "loss_scale": 4.0, "retries": 0, "seconds": 36.5, "max_memory_gb": 18.47, "ce": 0.0012319293891778216, "kl": 0.00026808765596797457, "context": 0.0013659731921507046} +{"step": 71, "steps": 250, "lr": 4.434149183384977e-05, "grad_norm": 0.053524039685726166, "loss_scale": 4.0, "retries": 0, "seconds": 37.0, "max_memory_gb": 18.47, "ce": 0.007356039413934923, "kl": 0.013959219300886616, "context": 0.014335648818814661} +{"step": 72, "steps": 250, "lr": 4.413966647630534e-05, "grad_norm": 0.2875429689884186, "loss_scale": 4.0, "retries": 0, "seconds": 37.5, "max_memory_gb": 18.47, "ce": 0.013828091352479532, "kl": 0.42454172583529726, "context": 0.2260989473434165} +{"step": 73, "steps": 250, "lr": 4.3934841913948323e-05, "grad_norm": 1.151853322982788, "loss_scale": 4.0, "retries": 0, "seconds": 37.9, "max_memory_gb": 18.47, "ce": 0.3751508732111688, "kl": 0.10019476104025671, "context": 0.4252482630145096} +{"step": 74, "steps": 250, "lr": 4.372705506523583e-05, "grad_norm": 0.3047366738319397, "loss_scale": 4.0, "retries": 0, "seconds": 38.4, "max_memory_gb": 18.47, "ce": 0.06375961776620898, "kl": 0.1182545273550204, "context": 0.12288688155240379} +{"step": 75, "steps": 250, "lr": 4.351634338256017e-05, "grad_norm": 1.0176622867584229, "loss_scale": 4.0, "retries": 0, "seconds": 39.0, "max_memory_gb": 18.47, "ce": 0.4315510602755239, "kl": 0.9882559906691313, "context": 0.9256790648214519} +{"step": 76, "steps": 250, "lr": 4.330274484549825e-05, "grad_norm": 0.8547210693359375, "loss_scale": 4.0, "retries": 0, "seconds": 39.5, "max_memory_gb": 18.47, "ce": 0.017175776963995304, "kl": 1.752604336681543, "context": 0.893477970326785} +{"step": 77, "steps": 250, "lr": 4.308629795396599e-05, "grad_norm": 0.05206604301929474, "loss_scale": 4.0, "retries": 0, "seconds": 39.9, "max_memory_gb": 18.47, "ce": 0.007572426351543982, "kl": 0.010777594541423241, "context": 0.012961223503225483} +{"step": 78, "steps": 250, "lr": 4.286704172127894e-05, "grad_norm": 0.13143490254878998, "loss_scale": 4.0, "retries": 0, "seconds": 40.4, "max_memory_gb": 18.47, "ce": 0.0015796128136571497, "kl": 0.08042647333058994, "context": 0.0417928489332553} +{"step": 79, "steps": 250, "lr": 4.264501566712033e-05, "grad_norm": 0.22757183015346527, "loss_scale": 4.0, "retries": 0, "seconds": 40.9, "max_memory_gb": 18.47, "ce": 0.015348420693044318, "kl": 0.02774356715144677, "context": 0.029220205171441194} +{"step": 80, "steps": 250, "lr": 4.2420259810417894e-05, "grad_norm": 0.15620863437652588, "loss_scale": 4.0, "retries": 0, "seconds": 41.4, "max_memory_gb": 18.47, "ce": 0.00182208668502426, "kl": 0.08379256839454285, "context": 0.04371837242888432} +{"step": 81, "steps": 250, "lr": 4.219281466213066e-05, "grad_norm": 0.47642049193382263, "loss_scale": 4.0, "retries": 0, "seconds": 41.9, "max_memory_gb": 18.47, "ce": 0.09069060790352523, "kl": 0.08515562966931611, "context": 0.13326842186506838} +{"step": 82, "steps": 250, "lr": 4.1962721217947135e-05, "grad_norm": 0.11674942821264267, "loss_scale": 4.0, "retries": 0, "seconds": 42.4, "max_memory_gb": 18.47, "ce": 0.009795729944016784, "kl": 0.002212842434346385, "context": 0.01090215120348148} +{"step": 83, "steps": 250, "lr": 4.173002095089599e-05, "grad_norm": 0.5172737240791321, "loss_scale": 4.0, "retries": 0, "seconds": 42.9, "max_memory_gb": 18.47, "ce": 0.15927229617955163, "kl": 0.7932196388283046, "context": 0.555882131884573} +{"step": 84, "steps": 250, "lr": 4.149475580387084e-05, "grad_norm": 0.3327857553958893, "loss_scale": 4.0, "retries": 0, "seconds": 43.4, "max_memory_gb": 18.47, "ce": 0.015116443001716107, "kl": 0.05847907246197792, "context": 0.04435597935025726} +{"step": 85, "steps": 250, "lr": 4.1256968182070275e-05, "grad_norm": 1.302587628364563, "loss_scale": 4.0, "retries": 0, "seconds": 43.9, "max_memory_gb": 18.47, "ce": 0.04602062771846249, "kl": 0.03839159014345839, "context": 0.06521642465327204} +{"step": 86, "steps": 250, "lr": 4.101670094535453e-05, "grad_norm": 1.5218003988265991, "loss_scale": 4.0, "retries": 0, "seconds": 44.4, "max_memory_gb": 18.47, "ce": 0.23213459562975913, "kl": 0.48972089996095747, "context": 0.4769950299523771} +{"step": 87, "steps": 250, "lr": 4.077399740052027e-05, "grad_norm": 0.7985764145851135, "loss_scale": 4.0, "retries": 0, "seconds": 44.9, "max_memory_gb": 18.47, "ce": 0.00636492388730403, "kl": 0.6809392673894763, "context": 0.34683455154299736} +{"step": 88, "steps": 250, "lr": 4.0528901293494775e-05, "grad_norm": 0.08496019244194031, "loss_scale": 4.0, "retries": 0, "seconds": 45.4, "max_memory_gb": 18.47, "ce": 0.00858509776298888, "kl": 0.02513904649524079, "context": 0.02115462124311307} +{"step": 89, "steps": 250, "lr": 4.0281456801451004e-05, "grad_norm": 0.08434867113828659, "loss_scale": 4.0, "retries": 0, "seconds": 45.9, "max_memory_gb": 18.47, "ce": 0.014910582203583544, "kl": 0.011859096877742559, "context": 0.020840131162913167} +{"step": 90, "steps": 250, "lr": 4.0031708524844884e-05, "grad_norm": 0.11765921860933304, "loss_scale": 4.0, "retries": 0, "seconds": 46.4, "max_memory_gb": 18.47, "ce": 0.009347681741928682, "kl": 0.008190628082957119, "context": 0.013442995608784258} +{"step": 91, "steps": 250, "lr": 3.977970147937635e-05, "grad_norm": 0.8421795964241028, "loss_scale": 4.0, "retries": 0, "seconds": 46.9, "max_memory_gb": 18.47, "ce": 0.2099076027952833, "kl": 0.5645861868761131, "context": 0.4922006857523229} +{"step": 92, "steps": 250, "lr": 3.952548108787553e-05, "grad_norm": 1.673797369003296, "loss_scale": 4.0, "retries": 0, "seconds": 47.4, "max_memory_gb": 18.47, "ce": 0.34511200603083125, "kl": 0.45970759821739193, "context": 0.5749658352906408} +{"step": 93, "steps": 250, "lr": 3.9269093172115496e-05, "grad_norm": 1.110667109489441, "loss_scale": 4.0, "retries": 0, "seconds": 47.9, "max_memory_gb": 18.47, "ce": 0.0614320739696268, "kl": 0.4623624675696192, "context": 0.2926133073924575} +{"step": 94, "steps": 250, "lr": 3.901058394455323e-05, "grad_norm": 1.5186080932617188, "loss_scale": 4.0, "retries": 0, "seconds": 48.4, "max_memory_gb": 18.47, "ce": 0.19886438973480836, "kl": 0.09673322566231946, "context": 0.24723100251867436} +{"step": 95, "steps": 250, "lr": 3.875e-05, "grad_norm": 1.5014947652816772, "loss_scale": 4.0, "retries": 0, "seconds": 48.9, "max_memory_gb": 18.47, "ce": 0.6098197237733984, "kl": 0.024554834235459566, "context": 0.6220971364527941} +{"step": 96, "steps": 250, "lr": 3.8487388307223005e-05, "grad_norm": 0.9440236687660217, "loss_scale": 4.0, "retries": 0, "seconds": 49.4, "max_memory_gb": 18.47, "ce": 0.47235715232091025, "kl": 0.006093075862736441, "context": 0.47540369984926656} +{"step": 97, "steps": 250, "lr": 3.822279620047943e-05, "grad_norm": 2.8847146034240723, "loss_scale": 4.0, "retries": 0, "seconds": 49.9, "max_memory_gb": 18.47, "ce": 0.020778329839231446, "kl": 0.3843791548279114, "context": 0.2129679098725319} +{"step": 98, "steps": 250, "lr": 3.795627137098479e-05, "grad_norm": 0.11759790033102036, "loss_scale": 4.0, "retries": 0, "seconds": 50.5, "max_memory_gb": 18.47, "ce": 0.0030730943217349704, "kl": 0.09931073142433888, "context": 0.05272845933723147} +{"step": 99, "steps": 250, "lr": 3.7687861858316844e-05, "grad_norm": 0.22539950907230377, "loss_scale": 4.0, "retries": 0, "seconds": 51.0, "max_memory_gb": 18.47, "ce": 0.021890117584007385, "kl": 0.012798676445527235, "context": 0.028289454650803236} +{"step": 100, "steps": 250, "lr": 3.741761604175668e-05, "grad_norm": 0.5116224884986877, "loss_scale": 4.0, "retries": 0, "seconds": 51.5, "max_memory_gb": 18.47, "ce": 0.03713291484746151, "kl": 0.6546516873713699, "context": 0.3644587595190387} +{"step": 101, "steps": 250, "lr": 3.714558263156872e-05, "grad_norm": 0.31528812646865845, "loss_scale": 4.0, "retries": 0, "seconds": 52.0, "max_memory_gb": 18.47, "ce": 0.005323405712260865, "kl": 0.08077613069326617, "context": 0.04571147006936371} +{"step": 102, "steps": 250, "lr": 3.6871810660220904e-05, "grad_norm": 0.21977216005325317, "loss_scale": 4.0, "retries": 0, "seconds": 52.5, "max_memory_gb": 18.47, "ce": 0.06077445950359106, "kl": 0.03138377610594034, "context": 0.07646634825505316} +{"step": 103, "steps": 250, "lr": 3.659634947354686e-05, "grad_norm": 0.13578081130981445, "loss_scale": 4.0, "retries": 0, "seconds": 53.0, "max_memory_gb": 18.47, "ce": 0.01157485730436747, "kl": 0.014721977823683119, "context": 0.018935846332169604} +{"step": 104, "steps": 250, "lr": 3.631924872185169e-05, "grad_norm": 0.0392204225063324, "loss_scale": 4.0, "retries": 0, "seconds": 53.4, "max_memory_gb": 18.47, "ce": 0.0033053398510674015, "kl": 0.0033978837018366903, "context": 0.005004281585570425} +{"step": 105, "steps": 250, "lr": 3.6040558350962674e-05, "grad_norm": 0.28268158435821533, "loss_scale": 4.0, "retries": 0, "seconds": 53.9, "max_memory_gb": 18.47, "ce": 0.07328182342462242, "kl": 0.36673070245888084, "context": 0.2566471677273512} +{"step": 106, "steps": 250, "lr": 3.576032859322687e-05, "grad_norm": 0.03147194907069206, "loss_scale": 4.0, "retries": 0, "seconds": 54.4, "max_memory_gb": 18.47, "ce": 0.003556175288395025, "kl": 0.001976060142624192, "context": 0.004544205381534994} +{"step": 107, "steps": 250, "lr": 3.5478609958457056e-05, "grad_norm": 2.3548102378845215, "loss_scale": 4.0, "retries": 0, "seconds": 54.9, "max_memory_gb": 18.47, "ce": 0.5499194997137238, "kl": 0.07555806817140365, "context": 0.5876985245740798} +{"step": 108, "steps": 250, "lr": 3.519545322482755e-05, "grad_norm": 0.4284462332725525, "loss_scale": 4.0, "retries": 0, "seconds": 55.4, "max_memory_gb": 18.47, "ce": 0.025452820671489462, "kl": 0.00640662715079543, "context": 0.028656133574259002} +{"step": 109, "steps": 250, "lr": 3.491090942972179e-05, "grad_norm": 0.7668432593345642, "loss_scale": 4.0, "retries": 0, "seconds": 55.9, "max_memory_gb": 18.47, "ce": 0.0587937074014917, "kl": 0.3448252473026514, "context": 0.23120632953941822} +{"step": 110, "steps": 250, "lr": 3.4625029860533135e-05, "grad_norm": 1.824815034866333, "loss_scale": 4.0, "retries": 0, "seconds": 56.4, "max_memory_gb": 18.47, "ce": 0.1815209488850087, "kl": 0.3166869154665619, "context": 0.33986440440639853} +{"step": 111, "steps": 250, "lr": 3.433786604542056e-05, "grad_norm": 0.019714536145329475, "loss_scale": 4.0, "retries": 0, "seconds": 57.0, "max_memory_gb": 18.47, "ce": 0.0016304849705193192, "kl": 0.0017540326689413632, "context": 0.002507501296349801} +{"step": 112, "steps": 250, "lr": 3.404946974402104e-05, "grad_norm": 0.07419957220554352, "loss_scale": 4.0, "retries": 0, "seconds": 57.5, "max_memory_gb": 18.47, "ce": 0.005536388081964105, "kl": 0.0005434619197330903, "context": 0.0058081190509255975} +{"step": 113, "steps": 250, "lr": 3.375989293812019e-05, "grad_norm": 0.3960195779800415, "loss_scale": 4.0, "retries": 0, "seconds": 58.1, "max_memory_gb": 18.47, "ce": 0.011603174498304725, "kl": 0.3744211355224252, "context": 0.19881374877877533} +{"step": 114, "steps": 250, "lr": 3.3469187822282766e-05, "grad_norm": 1.6573652029037476, "loss_scale": 4.0, "retries": 0, "seconds": 58.6, "max_memory_gb": 18.47, "ce": 0.018109203840140253, "kl": 1.2804102139198221, "context": 0.658314295753371} +{"step": 115, "steps": 250, "lr": 3.317740679444499e-05, "grad_norm": 0.12562158703804016, "loss_scale": 4.0, "retries": 0, "seconds": 59.1, "max_memory_gb": 18.47, "ce": 0.009458713509957306, "kl": 0.04864099810947664, "context": 0.03377921204082668} +{"step": 116, "steps": 250, "lr": 3.2884602446470056e-05, "grad_norm": 1.1433476209640503, "loss_scale": 4.0, "retries": 0, "seconds": 59.6, "max_memory_gb": 18.47, "ce": 0.15349731870082906, "kl": 0.5192712670450419, "context": 0.41313296713269665} +{"step": 117, "steps": 250, "lr": 3.259082755466876e-05, "grad_norm": 1.641991376876831, "loss_scale": 4.0, "retries": 0, "seconds": 60.1, "max_memory_gb": 18.48, "ce": 0.076166234590346, "kl": 0.12657512477017008, "context": 0.13945379697543103} +{"step": 118, "steps": 250, "lr": 3.229613507028686e-05, "grad_norm": 0.42933139204978943, "loss_scale": 4.0, "retries": 0, "seconds": 60.6, "max_memory_gb": 18.48, "ce": 0.1761109207291156, "kl": 0.011590738147788215, "context": 0.18190628502634354} +{"step": 119, "steps": 250, "lr": 3.2000578109961004e-05, "grad_norm": 0.1631118804216385, "loss_scale": 4.0, "retries": 0, "seconds": 61.1, "max_memory_gb": 18.48, "ce": 0.01579053325986024, "kl": 0.016972878896012844, "context": 0.024276972137158737} +{"step": 120, "steps": 250, "lr": 3.1704209946144616e-05, "grad_norm": 1.520583152770996, "loss_scale": 4.0, "retries": 0, "seconds": 61.7, "max_memory_gb": 18.48, "ce": 0.6038629712711554, "kl": 0.19555335966288112, "context": 0.701639641833026} +{"step": 121, "steps": 250, "lr": 3.140708399750594e-05, "grad_norm": 0.25486665964126587, "loss_scale": 4.0, "retries": 0, "seconds": 62.2, "max_memory_gb": 18.48, "ce": 0.051018117263083695, "kl": 0.039202583764563315, "context": 0.0706194086742471} +{"step": 122, "steps": 250, "lr": 3.1109253819299606e-05, "grad_norm": 0.3011314570903778, "loss_scale": 4.0, "retries": 0, "seconds": 62.7, "max_memory_gb": 18.48, "ce": 0.06282250296862912, "kl": 0.01569112250035687, "context": 0.07066806503280532} +{"step": 123, "steps": 250, "lr": 3.081077309371361e-05, "grad_norm": 0.20321784913539886, "loss_scale": 4.0, "retries": 0, "seconds": 63.2, "max_memory_gb": 18.48, "ce": 0.03803774400148541, "kl": 0.022566229396034032, "context": 0.04932085843756795} +{"step": 124, "steps": 250, "lr": 3.051169562019339e-05, "grad_norm": 1.5511012077331543, "loss_scale": 4.0, "retries": 0, "seconds": 63.7, "max_memory_gb": 18.48, "ce": 0.67026973614702, "kl": 0.6315400077619415, "context": 0.9860397577576805} +{"step": 125, "steps": 250, "lr": 3.021207530574477e-05, "grad_norm": 0.3709392845630646, "loss_scale": 4.0, "retries": 0, "seconds": 64.2, "max_memory_gb": 18.48, "ce": 0.10706060426309705, "kl": 0.09530479708337225, "context": 0.15471300238277763} +{"step": 126, "steps": 250, "lr": 2.991196615521758e-05, "grad_norm": 1.909188985824585, "loss_scale": 4.0, "retries": 0, "seconds": 64.7, "max_memory_gb": 18.48, "ce": 0.5171339027583599, "kl": 0.1567954858765006, "context": 0.5955316387116909} +{"step": 127, "steps": 250, "lr": 2.9611422261571536e-05, "grad_norm": 0.11333930492401123, "loss_scale": 4.0, "retries": 0, "seconds": 65.2, "max_memory_gb": 18.48, "ce": 0.015565007590339519, "kl": 0.0018970690962305525, "context": 0.016513541813765187} +{"step": 128, "steps": 250, "lr": 2.931049779612634e-05, "grad_norm": 2.200598955154419, "loss_scale": 4.0, "retries": 0, "seconds": 65.7, "max_memory_gb": 18.48, "ce": 0.29963945248164237, "kl": 1.237276355837821, "context": 0.9182776233647019} +{"step": 129, "steps": 250, "lr": 2.9009246998797557e-05, "grad_norm": 0.23478715121746063, "loss_scale": 4.0, "retries": 0, "seconds": 66.2, "max_memory_gb": 18.48, "ce": 0.07120763976126909, "kl": 0.03165793813968776, "context": 0.08703660877654329} +{"step": 130, "steps": 250, "lr": 2.8707724168320238e-05, "grad_norm": 0.2094045728445053, "loss_scale": 4.0, "retries": 0, "seconds": 66.7, "max_memory_gb": 18.48, "ce": 0.04055800673086196, "kl": 0.2729901996863191, "context": 0.17705310211749747} +{"step": 131, "steps": 250, "lr": 2.840598365246184e-05, "grad_norm": 2.056617498397827, "loss_scale": 4.0, "retries": 0, "seconds": 67.2, "max_memory_gb": 18.48, "ce": 0.21293925555073656, "kl": 0.360022222797852, "context": 0.3929503664840013} +{"step": 132, "steps": 250, "lr": 2.8104079838226383e-05, "grad_norm": 0.3060912489891052, "loss_scale": 4.0, "retries": 0, "seconds": 67.7, "max_memory_gb": 18.48, "ce": 0.16004177899594652, "kl": 0.029432747036480578, "context": 0.1747581571689807} +{"step": 133, "steps": 250, "lr": 2.7802067142051446e-05, "grad_norm": 0.5598915219306946, "loss_scale": 4.0, "retries": 0, "seconds": 68.3, "max_memory_gb": 18.48, "ce": 0.3408675689424854, "kl": 0.20099297957494855, "context": 0.44136404106393456} +{"step": 134, "steps": 250, "lr": 2.7500000000000004e-05, "grad_norm": 0.5206038951873779, "loss_scale": 4.0, "retries": 0, "seconds": 68.8, "max_memory_gb": 18.48, "ce": 0.22546121165032673, "kl": 0.005678557866843903, "context": 0.22830048639661982} +{"step": 135, "steps": 250, "lr": 2.7197932857948566e-05, "grad_norm": 0.11897749453783035, "loss_scale": 4.0, "retries": 0, "seconds": 69.2, "max_memory_gb": 18.48, "ce": 0.02147461156710051, "kl": 0.10277448336637462, "context": 0.07286185125121847} +{"step": 136, "steps": 250, "lr": 2.689592016177363e-05, "grad_norm": 2.968893527984619, "loss_scale": 4.0, "retries": 0, "seconds": 69.7, "max_memory_gb": 18.48, "ce": 0.15398396574892104, "kl": 0.1928658450488001, "context": 0.2504168879240751} +{"step": 137, "steps": 250, "lr": 2.659401634753817e-05, "grad_norm": 0.13729828596115112, "loss_scale": 4.0, "retries": 0, "seconds": 70.2, "max_memory_gb": 18.48, "ce": 0.0073618115711724386, "kl": 0.019591158416005783, "context": 0.01715739065548405} +{"step": 138, "steps": 250, "lr": 2.6292275831679764e-05, "grad_norm": 1.714805006980896, "loss_scale": 4.0, "retries": 0, "seconds": 70.7, "max_memory_gb": 18.48, "ce": 0.07896457915194333, "kl": 0.1552924821153283, "context": 0.15661082230508327} +{"step": 139, "steps": 250, "lr": 2.5990753001202445e-05, "grad_norm": 0.901220977306366, "loss_scale": 4.0, "retries": 0, "seconds": 71.2, "max_memory_gb": 18.48, "ce": 0.004033021301438566, "kl": 0.33815488692016515, "context": 0.17311046359100146} +{"step": 140, "steps": 250, "lr": 2.5689502203873672e-05, "grad_norm": 0.05156850442290306, "loss_scale": 4.0, "retries": 0, "seconds": 71.7, "max_memory_gb": 18.48, "ce": 0.004324786365032196, "kl": 0.0025793560926103964, "context": 0.005614464404061437} +{"step": 141, "steps": 250, "lr": 2.538857773842847e-05, "grad_norm": 0.03313900902867317, "loss_scale": 4.0, "retries": 0, "seconds": 72.2, "max_memory_gb": 18.48, "ce": 0.004514145766734146, "kl": 0.005085815952043049, "context": 0.007057053619064391} +{"step": 142, "steps": 250, "lr": 2.5088033844782426e-05, "grad_norm": 0.7064884901046753, "loss_scale": 4.0, "retries": 0, "seconds": 72.7, "max_memory_gb": 18.48, "ce": 0.048353448015404865, "kl": 0.02033207262866199, "context": 0.058519483427517116} +{"step": 143, "steps": 250, "lr": 2.4787924694255237e-05, "grad_norm": 1.9200443029403687, "loss_scale": 4.0, "retries": 0, "seconds": 73.3, "max_memory_gb": 18.48, "ce": 0.5190599969537288, "kl": 0.06423202747100731, "context": 0.5511759969667764} +{"step": 144, "steps": 250, "lr": 2.4488304379806627e-05, "grad_norm": 0.5954647660255432, "loss_scale": 4.0, "retries": 0, "seconds": 73.8, "max_memory_gb": 18.48, "ce": 0.08246875752229244, "kl": 0.20241073536453769, "context": 0.18367412383668125} +{"step": 145, "steps": 250, "lr": 2.4189226906286388e-05, "grad_norm": 0.04557415470480919, "loss_scale": 4.0, "retries": 0, "seconds": 74.3, "max_memory_gb": 18.48, "ce": 0.0018670876743271947, "kl": 0.009922206142164214, "context": 0.006828190489613917} +{"step": 146, "steps": 250, "lr": 2.3890746180700395e-05, "grad_norm": 0.08847998827695847, "loss_scale": 4.0, "retries": 0, "seconds": 74.8, "max_memory_gb": 18.48, "ce": 0.0075811108108609915, "kl": 0.023227424737342517, "context": 0.019194823398720473} +{"step": 147, "steps": 250, "lr": 2.359291600249407e-05, "grad_norm": 0.8320990204811096, "loss_scale": 4.0, "retries": 0, "seconds": 75.3, "max_memory_gb": 18.48, "ce": 0.4815437728902907, "kl": 0.5561051784679876, "context": 0.7595963802887127} +{"step": 148, "steps": 250, "lr": 2.3295790053855392e-05, "grad_norm": 0.26666635274887085, "loss_scale": 4.0, "retries": 0, "seconds": 75.9, "max_memory_gb": 18.48, "ce": 0.007391088784970634, "kl": 0.011212533107027411, "context": 0.012997355366678676} +{"step": 149, "steps": 250, "lr": 2.299942189003901e-05, "grad_norm": 0.19573387503623962, "loss_scale": 4.0, "retries": 0, "seconds": 76.4, "max_memory_gb": 18.48, "ce": 0.01642787962919101, "kl": 0.007548341367510147, "context": 0.020202050800435245} +{"step": 150, "steps": 250, "lr": 2.2703864929713144e-05, "grad_norm": 0.24380525946617126, "loss_scale": 4.0, "retries": 0, "seconds": 76.9, "max_memory_gb": 18.48, "ce": 0.06791000009980053, "kl": 0.02571185422129929, "context": 0.08076592744328082} +{"step": 151, "steps": 250, "lr": 2.2409172445331257e-05, "grad_norm": 0.2544405162334442, "loss_scale": 4.0, "retries": 0, "seconds": 77.4, "max_memory_gb": 18.48, "ce": 0.017481092712841928, "kl": 0.028456589043344138, "context": 0.03170938816765556} +{"step": 152, "steps": 250, "lr": 2.211539755352996e-05, "grad_norm": 0.40415701270103455, "loss_scale": 4.0, "retries": 0, "seconds": 77.9, "max_memory_gb": 18.48, "ce": 0.007578434218885377, "kl": 0.5964575107809651, "context": 0.30580719927093014} +{"step": 153, "steps": 250, "lr": 2.1822593205555016e-05, "grad_norm": 0.33012881875038147, "loss_scale": 4.0, "retries": 0, "seconds": 78.4, "max_memory_gb": 18.48, "ce": 0.01234734195531928, "kl": 0.38990626876875467, "context": 0.20730047494362225} +{"step": 154, "steps": 250, "lr": 2.1530812177717236e-05, "grad_norm": 1.8082482814788818, "loss_scale": 4.0, "retries": 0, "seconds": 78.9, "max_memory_gb": 18.48, "ce": 0.10562248883070424, "kl": 0.09251302533084527, "context": 0.15187900385353714} +{"step": 155, "steps": 250, "lr": 2.1240107061879815e-05, "grad_norm": 0.35463082790374756, "loss_scale": 4.0, "retries": 0, "seconds": 79.4, "max_memory_gb": 18.48, "ce": 0.19704864605773764, "kl": 0.010845071366929915, "context": 0.2024711868380109} +{"step": 156, "steps": 250, "lr": 2.095053025597896e-05, "grad_norm": 0.882461667060852, "loss_scale": 4.0, "retries": 0, "seconds": 80.0, "max_memory_gb": 18.48, "ce": 0.012594512467330787, "kl": 0.05573157912112947, "context": 0.04046030106837861} +{"step": 157, "steps": 250, "lr": 2.066213395457945e-05, "grad_norm": 0.7909230589866638, "loss_scale": 4.0, "retries": 0, "seconds": 80.5, "max_memory_gb": 18.48, "ce": 0.04473133571445942, "kl": 0.04146143932302948, "context": 0.06546205515041947} +{"step": 158, "steps": 250, "lr": 2.0374970139466877e-05, "grad_norm": 0.38760679960250854, "loss_scale": 4.0, "retries": 0, "seconds": 81.0, "max_memory_gb": 18.48, "ce": 0.1724336081861111, "kl": 0.07214322595245903, "context": 0.20850522216642275} +{"step": 159, "steps": 250, "lr": 2.0089090570278213e-05, "grad_norm": 0.6634434461593628, "loss_scale": 4.0, "retries": 0, "seconds": 81.5, "max_memory_gb": 18.48, "ce": 0.00503391501842998, "kl": 0.09123549549258314, "context": 0.05065166356507689} +{"step": 160, "steps": 250, "lr": 1.9804546775172457e-05, "grad_norm": 0.7087337374687195, "loss_scale": 4.0, "retries": 0, "seconds": 82.0, "max_memory_gb": 18.48, "ce": 0.024078057322185487, "kl": 0.3684059272054583, "context": 0.20828101428924128} +{"step": 161, "steps": 250, "lr": 1.9521390041542946e-05, "grad_norm": 0.39991867542266846, "loss_scale": 4.0, "retries": 0, "seconds": 82.5, "max_memory_gb": 18.48, "ce": 0.07022868914646097, "kl": 0.08408656355823041, "context": 0.11227197095286101} +{"step": 162, "steps": 250, "lr": 1.923967140677313e-05, "grad_norm": 1.1059890985488892, "loss_scale": 4.0, "retries": 0, "seconds": 83.0, "max_memory_gb": 18.48, "ce": 0.08983841762528755, "kl": 0.16737416217802092, "context": 0.17352549868519418} +{"step": 163, "steps": 250, "lr": 1.8959441649037335e-05, "grad_norm": 2.0623507499694824, "loss_scale": 4.0, "retries": 0, "seconds": 83.5, "max_memory_gb": 18.48, "ce": 0.23777518869610503, "kl": 0.7693347124441061, "context": 0.6224425446707755} +{"step": 164, "steps": 250, "lr": 1.868075127814832e-05, "grad_norm": 5.481932163238525, "loss_scale": 4.0, "retries": 0, "seconds": 84.0, "max_memory_gb": 18.48, "ce": 0.07998892768227961, "kl": 0.14725943698431365, "context": 0.15361865211161785} +{"step": 165, "steps": 250, "lr": 1.8403650526453143e-05, "grad_norm": 0.0833820179104805, "loss_scale": 4.0, "retries": 0, "seconds": 84.5, "max_memory_gb": 18.48, "ce": 0.004915788544167299, "kl": 0.02803207685610687, "context": 0.018931827318738215} +{"step": 166, "steps": 250, "lr": 1.812818933977912e-05, "grad_norm": 1.5972102880477905, "loss_scale": 4.0, "retries": 0, "seconds": 85.0, "max_memory_gb": 18.48, "ce": 0.4007855278905481, "kl": 0.14100711117498577, "context": 0.4712890723021701} +{"step": 167, "steps": 250, "lr": 1.785441736843128e-05, "grad_norm": 0.19399553537368774, "loss_scale": 4.0, "retries": 0, "seconds": 85.6, "max_memory_gb": 18.48, "ce": 0.024289147928357124, "kl": 0.0030636059236712754, "context": 0.025820950977504253} +{"step": 168, "steps": 250, "lr": 1.7582383958243314e-05, "grad_norm": 0.1141756922006607, "loss_scale": 4.0, "retries": 0, "seconds": 86.1, "max_memory_gb": 18.48, "ce": 0.028090385341783985, "kl": 0.002329969407583121, "context": 0.029255369110614993} +{"step": 169, "steps": 250, "lr": 1.7312138141683164e-05, "grad_norm": 1.0624289512634277, "loss_scale": 4.0, "retries": 0, "seconds": 86.6, "max_memory_gb": 18.48, "ce": 0.08839079475728795, "kl": 1.4561438858509064, "context": 0.8164627403020859} +{"step": 170, "steps": 250, "lr": 1.704372862901521e-05, "grad_norm": 0.02815074287354946, "loss_scale": 4.0, "retries": 0, "seconds": 87.1, "max_memory_gb": 18.48, "ce": 0.0017046406064764597, "kl": 0.0006126953485363629, "context": 0.0020109883043915033} +{"step": 171, "steps": 250, "lr": 1.6777203799520573e-05, "grad_norm": 0.7870268821716309, "loss_scale": 4.0, "retries": 0, "seconds": 87.6, "max_memory_gb": 18.48, "ce": 0.023741157552649383, "kl": 0.26435705681797117, "context": 0.15591968037188053} +{"step": 172, "steps": 250, "lr": 1.6512611692777008e-05, "grad_norm": 0.11187440901994705, "loss_scale": 4.0, "retries": 0, "seconds": 88.1, "max_memory_gb": 18.48, "ce": 0.016224606661126018, "kl": 0.017374750575982034, "context": 0.0249119809595868} +{"step": 173, "steps": 250, "lr": 1.6250000000000005e-05, "grad_norm": 0.16961349546909332, "loss_scale": 4.0, "retries": 0, "seconds": 88.6, "max_memory_gb": 18.48, "ce": 0.04984113515820354, "kl": 0.02850122284507961, "context": 0.06409174669533968} +{"step": 174, "steps": 250, "lr": 1.598941605544678e-05, "grad_norm": 0.21724176406860352, "loss_scale": 4.0, "retries": 0, "seconds": 89.1, "max_memory_gb": 18.48, "ce": 0.03695670250090188, "kl": 0.03445829745533047, "context": 0.05418585029838141} +{"step": 175, "steps": 250, "lr": 1.5730906827884506e-05, "grad_norm": 2.2885775566101074, "loss_scale": 4.0, "retries": 0, "seconds": 89.6, "max_memory_gb": 18.48, "ce": 0.3626390206627548, "kl": 0.08505502610933036, "context": 0.405166526325047} +{"step": 176, "steps": 250, "lr": 1.5474518912124476e-05, "grad_norm": 0.05461809039115906, "loss_scale": 4.0, "retries": 0, "seconds": 90.1, "max_memory_gb": 18.48, "ce": 0.009475256942096166, "kl": 0.002786136445138254, "context": 0.010868325014598668} +{"step": 177, "steps": 250, "lr": 1.5220298520623646e-05, "grad_norm": 0.6214926838874817, "loss_scale": 4.0, "retries": 0, "seconds": 90.6, "max_memory_gb": 18.48, "ce": 0.015435282271937467, "kl": 0.8021249046396406, "context": 0.41649773040262517} +{"step": 178, "steps": 250, "lr": 1.4968291475155121e-05, "grad_norm": 0.8385277390480042, "loss_scale": 4.0, "retries": 0, "seconds": 91.1, "max_memory_gb": 18.48, "ce": 0.07241985358996317, "kl": 0.1369177869783016, "context": 0.14087874698452652} +{"step": 179, "steps": 250, "lr": 1.4718543198548998e-05, "grad_norm": 0.7279931902885437, "loss_scale": 4.0, "retries": 0, "seconds": 91.6, "max_memory_gb": 18.48, "ce": 0.34662664375355234, "kl": 0.025103338151893695, "context": 0.3591783100782777} +{"step": 180, "steps": 250, "lr": 1.4471098706505227e-05, "grad_norm": 0.1678895652294159, "loss_scale": 4.0, "retries": 0, "seconds": 92.1, "max_memory_gb": 18.48, "ce": 0.039256186079001054, "kl": 0.023873618727520807, "context": 0.05119299543730449} +{"step": 181, "steps": 250, "lr": 1.4226002599479743e-05, "grad_norm": 0.1422039419412613, "loss_scale": 4.0, "retries": 0, "seconds": 92.6, "max_memory_gb": 18.48, "ce": 0.00827909167855978, "kl": 0.08995465171756223, "context": 0.05325641774106771} +{"step": 182, "steps": 250, "lr": 1.3983299054645471e-05, "grad_norm": 1.289318323135376, "loss_scale": 4.0, "retries": 0, "seconds": 93.1, "max_memory_gb": 18.48, "ce": 0.0227274437493179, "kl": 0.3774454868016619, "context": 0.21145019121468067} +{"step": 183, "steps": 250, "lr": 1.3743031817929724e-05, "grad_norm": 0.025013310834765434, "loss_scale": 4.0, "retries": 0, "seconds": 93.6, "max_memory_gb": 18.48, "ce": 0.002547097467868298, "kl": 0.000693658585078083, "context": 0.002893926764954813} +{"step": 184, "steps": 250, "lr": 1.3505244196129164e-05, "grad_norm": 0.03234568610787392, "loss_scale": 4.0, "retries": 0, "seconds": 94.1, "max_memory_gb": 18.48, "ce": 0.00232083832815988, "kl": 0.003070948931053863, "context": 0.0038563127018278465} +{"step": 185, "steps": 250, "lr": 1.3269979049104014e-05, "grad_norm": 0.15873751044273376, "loss_scale": 4.0, "retries": 0, "seconds": 94.8, "max_memory_gb": 18.48, "ce": 0.009548407047986984, "kl": 0.09272348387821694, "context": 0.055910148948896676} +{"step": 186, "steps": 250, "lr": 1.3037278782052863e-05, "grad_norm": 2.3909823894500732, "loss_scale": 4.0, "retries": 0, "seconds": 95.3, "max_memory_gb": 18.48, "ce": 0.1382546922777692, "kl": 0.33601366374387, "context": 0.30626152304466814} +{"step": 187, "steps": 250, "lr": 1.2807185337869343e-05, "grad_norm": 0.3072265684604645, "loss_scale": 4.0, "retries": 0, "seconds": 95.9, "max_memory_gb": 18.48, "ce": 0.08731738434289582, "kl": 0.48500519804656506, "context": 0.3298199992859736} +{"step": 188, "steps": 250, "lr": 1.2579740189582115e-05, "grad_norm": 1.5507251024246216, "loss_scale": 4.0, "retries": 0, "seconds": 96.5, "max_memory_gb": 18.48, "ce": 0.27495803195779445, "kl": 0.014098167657721206, "context": 0.28200710740929935} +{"step": 189, "steps": 250, "lr": 1.2354984332879668e-05, "grad_norm": 1.4446018934249878, "loss_scale": 4.0, "retries": 0, "seconds": 97.0, "max_memory_gb": 18.48, "ce": 0.3674656613729894, "kl": 0.29938407510053366, "context": 0.5171577092260122} +{"step": 190, "steps": 250, "lr": 1.2132958278721064e-05, "grad_norm": 0.098702073097229, "loss_scale": 4.0, "retries": 0, "seconds": 97.5, "max_memory_gb": 18.48, "ce": 0.004331600328441709, "kl": 0.04101794329471886, "context": 0.024840571684762836} +{"step": 191, "steps": 250, "lr": 1.1913702046034016e-05, "grad_norm": 4.276054859161377, "loss_scale": 4.0, "retries": 0, "seconds": 98.1, "max_memory_gb": 18.48, "ce": 0.27299035679607186, "kl": 0.2906342129949735, "context": 0.41830746359482873} +{"step": 192, "steps": 250, "lr": 1.1697255154501756e-05, "grad_norm": 1.0149080753326416, "loss_scale": 4.0, "retries": 0, "seconds": 98.7, "max_memory_gb": 18.48, "ce": 0.4552980269945692, "kl": 0.09031485590094235, "context": 0.5004554530896712} +{"step": 193, "steps": 250, "lr": 1.1483656617439842e-05, "grad_norm": 0.07238470762968063, "loss_scale": 4.0, "retries": 0, "seconds": 99.3, "max_memory_gb": 18.48, "ce": 0.004788070593349403, "kl": 0.007698922227064031, "context": 0.008637531707790913} +{"step": 194, "steps": 250, "lr": 1.1272944934764177e-05, "grad_norm": 0.3743046820163727, "loss_scale": 4.0, "retries": 0, "seconds": 99.9, "max_memory_gb": 18.48, "ce": 0.034316368168219924, "kl": 0.033495755167678, "context": 0.05106424575205892} +{"step": 195, "steps": 250, "lr": 1.1065158086051684e-05, "grad_norm": 0.06260156631469727, "loss_scale": 4.0, "retries": 0, "seconds": 100.5, "max_memory_gb": 18.48, "ce": 0.010247645775962155, "kl": 0.00505516659359273, "context": 0.012775229159160517} +{"step": 196, "steps": 250, "lr": 1.086033352369467e-05, "grad_norm": 0.4795375466346741, "loss_scale": 4.0, "retries": 0, "seconds": 101.2, "max_memory_gb": 18.48, "ce": 0.001172544683868182, "kl": 0.46261500919285936, "context": 0.232480055963606} +{"step": 197, "steps": 250, "lr": 1.0658508166150224e-05, "grad_norm": 1.0205079317092896, "loss_scale": 4.0, "retries": 0, "seconds": 101.8, "max_memory_gb": 18.48, "ce": 0.00388504151487723, "kl": 0.8052141382363516, "context": 0.4064920967211947} +{"step": 198, "steps": 250, "lr": 1.0459718391285954e-05, "grad_norm": 0.3217482268810272, "loss_scale": 4.0, "retries": 0, "seconds": 102.2, "max_memory_gb": 18.48, "ce": 0.052459574602835346, "kl": 0.0095133573977364, "context": 0.05721625275327824} +{"step": 199, "steps": 250, "lr": 1.0264000029822999e-05, "grad_norm": 0.5000986456871033, "loss_scale": 4.0, "retries": 0, "seconds": 102.7, "max_memory_gb": 18.48, "ce": 0.1528206932125613, "kl": 0.18004974109499017, "context": 0.2428455641202163} +{"step": 200, "steps": 250, "lr": 1.0071388358877775e-05, "grad_norm": 1.2796674966812134, "loss_scale": 4.0, "retries": 0, "seconds": 103.3, "max_memory_gb": 18.48, "ce": 0.3462075819261372, "kl": 0.043599960874416865, "context": 0.368007569340989} +{"step": 201, "steps": 250, "lr": 9.881918095603501e-06, "grad_norm": 0.8780553340911865, "loss_scale": 4.0, "retries": 0, "seconds": 103.8, "max_memory_gb": 18.48, "ce": 0.0834197456279071, "kl": 0.14039808964298572, "context": 0.153618790442124} +{"step": 202, "steps": 250, "lr": 9.695623390932612e-06, "grad_norm": 0.040122877806425095, "loss_scale": 4.0, "retries": 0, "seconds": 104.4, "max_memory_gb": 18.48, "ce": 0.0016855299609233043, "kl": 0.002875573779760998, "context": 0.0031233167655955185} +{"step": 203, "steps": 250, "lr": 9.51253782342123e-06, "grad_norm": 1.0846788883209229, "loss_scale": 4.0, "retries": 0, "seconds": 104.9, "max_memory_gb": 18.48, "ce": 0.044176101495395415, "kl": 0.08100443438161165, "context": 0.08467832248425111} +{"step": 204, "steps": 250, "lr": 9.332694393196846e-06, "grad_norm": 0.1762850433588028, "loss_scale": 4.0, "retries": 0, "seconds": 105.4, "max_memory_gb": 18.48, "ce": 0.03783526920597069, "kl": 0.013019267586059868, "context": 0.04434490104904398} +{"step": 205, "steps": 250, "lr": 9.156125516010256e-06, "grad_norm": 0.327686071395874, "loss_scale": 4.0, "retries": 0, "seconds": 105.9, "max_memory_gb": 18.48, "ce": 0.00420107546960935, "kl": 0.5703759284806438, "context": 0.28938904823735356} +{"step": 206, "steps": 250, "lr": 8.982863017392733e-06, "grad_norm": 0.05591459199786186, "loss_scale": 4.0, "retries": 0, "seconds": 106.5, "max_memory_gb": 18.48, "ce": 0.007303096277610166, "kl": 0.0035363587658139295, "context": 0.009071275864698691} +{"step": 207, "steps": 250, "lr": 8.81293812691974e-06, "grad_norm": 0.16052646934986115, "loss_scale": 4.0, "retries": 0, "seconds": 107.2, "max_memory_gb": 18.48, "ce": 0.06158375780796632, "kl": 0.06271295726764947, "context": 0.0929402363835834} +{"step": 208, "steps": 250, "lr": 8.646381472581927e-06, "grad_norm": 0.04686427116394043, "loss_scale": 4.0, "retries": 0, "seconds": 107.8, "max_memory_gb": 18.48, "ce": 0.004879402556980494, "kl": 0.0019712028515641578, "context": 0.005865004161023535} +{"step": 209, "steps": 250, "lr": 8.483223075264622e-06, "grad_norm": 0.7455464005470276, "loss_scale": 4.0, "retries": 0, "seconds": 108.4, "max_memory_gb": 18.48, "ce": 0.15362401772290468, "kl": 0.12469581266850582, "context": 0.2159719312330708} +{"step": 210, "steps": 250, "lr": 8.323492343336747e-06, "grad_norm": 2.718151092529297, "loss_scale": 4.0, "retries": 0, "seconds": 109.0, "max_memory_gb": 18.48, "ce": 0.5118235645477398, "kl": 0.1541785546214669, "context": 0.5889128260387224} +{"step": 211, "steps": 250, "lr": 8.167218067350122e-06, "grad_norm": 0.28620991110801697, "loss_scale": 4.0, "retries": 0, "seconds": 109.7, "max_memory_gb": 18.48, "ce": 0.12642978789517656, "kl": 0.13243705336935818, "context": 0.192648321040906} +{"step": 212, "steps": 250, "lr": 8.01442841485013e-06, "grad_norm": 0.64717698097229, "loss_scale": 4.0, "retries": 0, "seconds": 110.4, "max_memory_gb": 18.48, "ce": 0.037527224281802773, "kl": 1.0371684014244238, "context": 0.5561114144511521} +{"step": 213, "steps": 250, "lr": 7.865150925298678e-06, "grad_norm": 0.5169382095336914, "loss_scale": 4.0, "retries": 0, "seconds": 110.9, "max_memory_gb": 18.48, "ce": 0.21109288508887403, "kl": 0.05257501650339691, "context": 0.23738039321324322} +{"step": 214, "steps": 250, "lr": 7.719412505110375e-06, "grad_norm": 0.6128838062286377, "loss_scale": 4.0, "retries": 0, "seconds": 111.5, "max_memory_gb": 18.48, "ce": 0.41018887985046604, "kl": 0.18282414553686976, "context": 0.5016009469109122} +{"step": 215, "steps": 250, "lr": 7.57723942280278e-06, "grad_norm": 1.3910348415374756, "loss_scale": 4.0, "retries": 0, "seconds": 112.0, "max_memory_gb": 18.48, "ce": 0.647135672665172, "kl": 0.025282936079577212, "context": 0.6597771421293146} +{"step": 216, "steps": 250, "lr": 7.438657304261687e-06, "grad_norm": 0.2637961208820343, "loss_scale": 4.0, "retries": 0, "seconds": 112.5, "max_memory_gb": 18.48, "ce": 0.013972432992886752, "kl": 0.12051725946366787, "context": 0.0742310638888739} +{"step": 217, "steps": 250, "lr": 7.30369112812219e-06, "grad_norm": 0.028820227831602097, "loss_scale": 4.0, "retries": 0, "seconds": 113.1, "max_memory_gb": 18.48, "ce": 0.0014809537678956985, "kl": 0.0027204033694943064, "context": 0.002841155437636189} +{"step": 218, "steps": 250, "lr": 7.172365221266403e-06, "grad_norm": 0.6101322770118713, "loss_scale": 4.0, "retries": 0, "seconds": 113.7, "max_memory_gb": 18.48, "ce": 0.12720654136501253, "kl": 0.32248619198799133, "context": 0.28844963386654854} +{"step": 219, "steps": 250, "lr": 7.0447032544387154e-06, "grad_norm": 0.6405938267707825, "loss_scale": 4.0, "retries": 0, "seconds": 114.3, "max_memory_gb": 18.48, "ce": 0.0088790397712728, "kl": 0.1375268844421953, "context": 0.07764248494640924} +{"step": 220, "steps": 250, "lr": 6.920728237979274e-06, "grad_norm": 0.1925240159034729, "loss_scale": 4.0, "retries": 0, "seconds": 114.9, "max_memory_gb": 18.48, "ce": 0.06276771388365887, "kl": 0.008363311424545827, "context": 0.06694937051361194} +{"step": 221, "steps": 250, "lr": 6.800462517676456e-06, "grad_norm": 0.3475606143474579, "loss_scale": 4.0, "retries": 0, "seconds": 115.5, "max_memory_gb": 18.48, "ce": 0.006520731549244374, "kl": 0.5460272276650358, "context": 0.27953434400842525} +{"step": 222, "steps": 250, "lr": 6.683927770739222e-06, "grad_norm": 0.2101351022720337, "loss_scale": 4.0, "retries": 0, "seconds": 116.1, "max_memory_gb": 18.48, "ce": 0.02677522636804497, "kl": 0.08141629529404781, "context": 0.06748337263707072} +{"step": 223, "steps": 250, "lr": 6.571145001889892e-06, "grad_norm": 0.6011906862258911, "loss_scale": 4.0, "retries": 0, "seconds": 116.7, "max_memory_gb": 18.48, "ce": 0.03204498174454784, "kl": 0.004513380124990363, "context": 0.034301672043511644} +{"step": 224, "steps": 250, "lr": 6.46213453957817e-06, "grad_norm": 0.28707122802734375, "loss_scale": 4.0, "retries": 0, "seconds": 117.3, "max_memory_gb": 18.48, "ce": 0.016778119636001065, "kl": 0.21726807267077675, "context": 0.12541215957025997} +{"step": 225, "steps": 250, "lr": 6.356916032317063e-06, "grad_norm": 0.22778065502643585, "loss_scale": 4.0, "retries": 0, "seconds": 117.9, "max_memory_gb": 18.48, "ce": 0.004564971182844602, "kl": 0.23401332221692428, "context": 0.1215716328297276} +{"step": 226, "steps": 250, "lr": 6.255508445141349e-06, "grad_norm": 1.1402720212936401, "loss_scale": 4.0, "retries": 0, "seconds": 118.5, "max_memory_gb": 18.48, "ce": 0.060270661633694544, "kl": 0.04909426836820785, "context": 0.08481779383146204} +{"step": 227, "steps": 250, "lr": 6.157930056189228e-06, "grad_norm": 0.24906586110591888, "loss_scale": 4.0, "retries": 0, "seconds": 119.1, "max_memory_gb": 18.48, "ce": 0.021511676959562465, "kl": 0.005170211436052341, "context": 0.024096781791740796} +{"step": 228, "steps": 250, "lr": 6.064198453407796e-06, "grad_norm": 0.4489878714084625, "loss_scale": 4.0, "retries": 0, "seconds": 119.7, "max_memory_gb": 18.48, "ce": 0.00347842117480468, "kl": 0.3321473251271527, "context": 0.16955207925639115} +{"step": 229, "steps": 250, "lr": 5.974330531382935e-06, "grad_norm": 0.35518115758895874, "loss_scale": 4.0, "retries": 0, "seconds": 120.3, "max_memory_gb": 18.48, "ce": 0.0560305784092634, "kl": 0.01641975612119495, "context": 0.06424045750463847} +{"step": 230, "steps": 250, "lr": 5.888342488294125e-06, "grad_norm": 1.6344941854476929, "loss_scale": 4.0, "retries": 0, "seconds": 120.8, "max_memory_gb": 18.48, "ce": 0.5366228933780803, "kl": 0.4324381945771165, "context": 0.7528420169837773} +{"step": 231, "steps": 250, "lr": 5.806249822994849e-06, "grad_norm": 0.1975729912519455, "loss_scale": 4.0, "retries": 0, "seconds": 121.4, "max_memory_gb": 18.48, "ce": 0.0290237508379505, "kl": 0.13796918481239118, "context": 0.09800834184716223} +{"step": 232, "steps": 250, "lr": 5.728067332218994e-06, "grad_norm": 0.12636221945285797, "loss_scale": 4.0, "retries": 0, "seconds": 122.0, "max_memory_gb": 18.48, "ce": 0.02408856590045616, "kl": 0.011263537511695176, "context": 0.029720335558522493} +{"step": 233, "steps": 250, "lr": 5.65380910791383e-06, "grad_norm": 0.3503357470035553, "loss_scale": 4.0, "retries": 0, "seconds": 122.6, "max_memory_gb": 18.48, "ce": 0.008281099260784686, "kl": 0.12123468739446253, "context": 0.06889844150282443} +{"step": 234, "steps": 250, "lr": 5.583488534700022e-06, "grad_norm": 0.4642844796180725, "loss_scale": 4.0, "retries": 0, "seconds": 123.2, "max_memory_gb": 18.48, "ce": 0.011010012807673775, "kl": 0.09874151970871026, "context": 0.06038077351695392} +{"step": 235, "steps": 250, "lr": 5.517118287459118e-06, "grad_norm": 1.3751190900802612, "loss_scale": 4.0, "retries": 0, "seconds": 123.8, "max_memory_gb": 18.48, "ce": 0.3610643131978577, "kl": 0.0269700711742189, "context": 0.3745493347378215} +{"step": 236, "steps": 250, "lr": 5.454710329048978e-06, "grad_norm": 0.06759192794561386, "loss_scale": 4.0, "retries": 0, "seconds": 124.4, "max_memory_gb": 18.48, "ce": 0.008868780307238922, "kl": 0.017959263150260085, "context": 0.017848411764134653} +{"step": 237, "steps": 250, "lr": 5.3962759081475436e-06, "grad_norm": 0.44072625041007996, "loss_scale": 4.0, "retries": 0, "seconds": 125.0, "max_memory_gb": 18.48, "ce": 0.07132689845457207, "kl": 0.030052232012167224, "context": 0.08635301428148523} +{"step": 238, "steps": 250, "lr": 5.341825557225321e-06, "grad_norm": 0.3275434374809265, "loss_scale": 4.0, "retries": 0, "seconds": 125.6, "max_memory_gb": 18.48, "ce": 0.04027975129429251, "kl": 0.011146640716106049, "context": 0.04585307210800238} +{"step": 239, "steps": 250, "lr": 5.291369090646961e-06, "grad_norm": 1.0911821126937866, "loss_scale": 4.0, "retries": 0, "seconds": 126.2, "max_memory_gb": 18.48, "ce": 0.10632484682719223, "kl": 0.03748423696379177, "context": 0.12506696159834974} +{"step": 240, "steps": 250, "lr": 5.244915602902289e-06, "grad_norm": 0.22269682586193085, "loss_scale": 4.0, "retries": 0, "seconds": 126.8, "max_memory_gb": 18.48, "ce": 0.09390964270278346, "kl": 0.042945090663124574, "context": 0.11538219268550165} +{"step": 241, "steps": 250, "lr": 5.20247346696706e-06, "grad_norm": 0.944218635559082, "loss_scale": 4.0, "retries": 0, "seconds": 127.4, "max_memory_gb": 18.48, "ce": 0.10326486550911795, "kl": 0.0236517277371604, "context": 0.11509073310298845} +{"step": 242, "steps": 250, "lr": 5.164050332793787e-06, "grad_norm": 0.11738599836826324, "loss_scale": 4.0, "retries": 0, "seconds": 128.0, "max_memory_gb": 18.48, "ce": 0.00514639164248365, "kl": 0.06342997194406053, "context": 0.03686137877230067} +{"step": 243, "steps": 250, "lr": 5.129653125932873e-06, "grad_norm": 0.06359083205461502, "loss_scale": 4.0, "retries": 0, "seconds": 128.5, "max_memory_gb": 18.48, "ce": 0.004894513578619808, "kl": 0.005397318949690089, "context": 0.007593173126224428} +{"step": 244, "steps": 250, "lr": 5.099288046284329e-06, "grad_norm": 0.5815021991729736, "loss_scale": 4.0, "retries": 0, "seconds": 129.0, "max_memory_gb": 18.48, "ce": 0.04306828533299267, "kl": 0.18768466060282663, "context": 0.13691061505232938} +{"step": 245, "steps": 250, "lr": 5.072960566980277e-06, "grad_norm": 1.6499794721603394, "loss_scale": 4.0, "retries": 0, "seconds": 129.5, "max_memory_gb": 18.48, "ce": 0.3805263619287871, "kl": 0.09261178353335708, "context": 0.4268322443822399} +{"step": 246, "steps": 250, "lr": 5.050675433398435e-06, "grad_norm": 0.021813513711094856, "loss_scale": 4.0, "retries": 0, "seconds": 130.0, "max_memory_gb": 18.48, "ce": 0.0012082168614142574, "kl": 0.0008224753655667882, "context": 0.001619454546016641} +{"step": 247, "steps": 250, "lr": 5.032436662306811e-06, "grad_norm": 0.40018129348754883, "loss_scale": 4.0, "retries": 0, "seconds": 130.5, "max_memory_gb": 18.48, "ce": 0.006180022421176545, "kl": 0.4863664470394724, "context": 0.24936325060843956} +{"step": 248, "steps": 250, "lr": 5.018247541139684e-06, "grad_norm": 0.1977994441986084, "loss_scale": 4.0, "retries": 0, "seconds": 131.0, "max_memory_gb": 18.48, "ce": 0.031648120551835746, "kl": 0.015182808972895145, "context": 0.039239524165168405} +{"step": 249, "steps": 250, "lr": 5.00811062740508e-06, "grad_norm": 0.29526418447494507, "loss_scale": 4.0, "retries": 0, "seconds": 131.5, "max_memory_gb": 18.48, "ce": 0.03710743598639965, "kl": 0.044983799307374284, "context": 0.05959933763369918} +{"step": 250, "steps": 250, "lr": 5.002027748223778e-06, "grad_norm": 0.05052896961569786, "loss_scale": 4.0, "retries": 0, "seconds": 132.0, "max_memory_gb": 18.48, "ce": 0.005944758144323714, "kl": 0.003194412376615219, "context": 0.007541964034317061} diff --git a/runs/single_internvl/context/L09_s13/adapter/README.md b/runs/single_internvl/context/L09_s13/adapter/README.md new file mode 100644 index 0000000000000000000000000000000000000000..4f9f856cf0e1531b47e25d266833b2977a19d3ac --- /dev/null +++ b/runs/single_internvl/context/L09_s13/adapter/README.md @@ -0,0 +1,207 @@ +--- +base_model: /data/OPSD/assets/models/InternVL3_5-8B-HF +library_name: peft +pipeline_tag: text-generation +tags: +- base_model:adapter:/data/OPSD/assets/models/InternVL3_5-8B-HF +- lora +- transformers +--- + +# Model Card for Model ID + + + + + +## Model Details + +### Model Description + + + + + +- **Developed by:** [More Information Needed] +- **Funded by [optional]:** [More Information Needed] +- **Shared by [optional]:** [More Information Needed] +- **Model type:** [More Information Needed] +- **Language(s) (NLP):** [More Information Needed] +- **License:** [More Information Needed] +- **Finetuned from model [optional]:** [More Information Needed] + +### Model Sources [optional] + + + +- **Repository:** [More Information Needed] +- **Paper [optional]:** [More Information Needed] +- **Demo [optional]:** [More Information Needed] + +## Uses + + + +### Direct Use + + + +[More Information Needed] + +### Downstream Use [optional] + + + +[More Information Needed] + +### Out-of-Scope Use + + + +[More Information Needed] + +## Bias, Risks, and Limitations + + + +[More Information Needed] + +### Recommendations + + + +Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. More information needed for further recommendations. + +## How to Get Started with the Model + +Use the code below to get started with the model. + +[More Information Needed] + +## Training Details + +### Training Data + + + +[More Information Needed] + +### Training Procedure + + + +#### Preprocessing [optional] + +[More Information Needed] + + +#### Training Hyperparameters + +- **Training regime:** [More Information Needed] + +#### Speeds, Sizes, Times [optional] + + + +[More Information Needed] + +## Evaluation + + + +### Testing Data, Factors & Metrics + +#### Testing Data + + + +[More Information Needed] + +#### Factors + + + +[More Information Needed] + +#### Metrics + + + +[More Information Needed] + +### Results + +[More Information Needed] + +#### Summary + + + +## Model Examination [optional] + + + +[More Information Needed] + +## Environmental Impact + + + +Carbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700). + +- **Hardware Type:** [More Information Needed] +- **Hours used:** [More Information Needed] +- **Cloud Provider:** [More Information Needed] +- **Compute Region:** [More Information Needed] +- **Carbon Emitted:** [More Information Needed] + +## Technical Specifications [optional] + +### Model Architecture and Objective + +[More Information Needed] + +### Compute Infrastructure + +[More Information Needed] + +#### Hardware + +[More Information Needed] + +#### Software + +[More Information Needed] + +## Citation [optional] + + + +**BibTeX:** + +[More Information Needed] + +**APA:** + +[More Information Needed] + +## Glossary [optional] + + + +[More Information Needed] + +## More Information [optional] + +[More Information Needed] + +## Model Card Authors [optional] + +[More Information Needed] + +## Model Card Contact + +[More Information Needed] +### Framework versions + +- PEFT 0.20.0 \ No newline at end of file diff --git a/runs/single_internvl/context/L09_s13/adapter/adapter_config.json b/runs/single_internvl/context/L09_s13/adapter/adapter_config.json new file mode 100644 index 0000000000000000000000000000000000000000..daa3c1ee1f0e70fb02c9ff12668affc61a3869ce --- /dev/null +++ b/runs/single_internvl/context/L09_s13/adapter/adapter_config.json @@ -0,0 +1,50 @@ +{ + "alora_invocation_tokens": null, + "alpha_pattern": {}, + "arrow_config": null, + "auto_mapping": null, + "base_model_name_or_path": "/data/OPSD/assets/models/InternVL3_5-8B-HF", + "bias": "none", + "corda_config": null, + "ensure_weight_tying": false, + "eva_config": null, + "exclude_modules": null, + "fan_in_fan_out": false, + "inference_mode": true, + "init_lora_weights": true, + "layer_replication": null, + "layers_pattern": null, + "layers_to_transform": null, + "loftq_config": {}, + "lora_alpha": 32, + "lora_bias": false, + "lora_dropout": 0.05, + "lora_ga_config": null, + "megatron_config": null, + "megatron_core": "megatron.core", + "modules_to_save": null, + "monteclora_config": null, + "peft_type": "LORA", + "peft_version": "0.20.0", + "qalora_group_size": 16, + "r": 16, + "rank_pattern": {}, + "revision": null, + "target_modules": [ + "model.language_model.layers.9.self_attn.k_proj", + "model.language_model.layers.9.mlp.up_proj", + "model.language_model.layers.9.self_attn.q_proj", + "model.language_model.layers.9.self_attn.v_proj", + "model.language_model.layers.9.mlp.down_proj", + "model.language_model.layers.9.self_attn.o_proj", + "model.language_model.layers.9.mlp.gate_proj" + ], + "target_parameters": null, + "task_type": "CAUSAL_LM", + "trainable_token_indices": null, + "use_bdlora": null, + "use_dora": false, + "use_qalora": false, + "use_rslora": false, + "velora_config": null +} \ No newline at end of file diff --git a/runs/single_internvl/context/L09_s13/config.json b/runs/single_internvl/context/L09_s13/config.json new file mode 100644 index 0000000000000000000000000000000000000000..c88ca78d8c23c6214113a14376620128b8aac0b9 --- /dev/null +++ b/runs/single_internvl/context/L09_s13/config.json @@ -0,0 +1,24 @@ +{ + "behavior": "context", + "seed": 13, + "model": "/data/OPSD/assets/models/InternVL3_5-8B-HF", + "backbone": "internvl", + "train": "/data/OPSD/assets/v3/context.jsonl", + "items": 1000, + "steps": 250, + "epochs": 1, + "steps_per_epoch": 250, + "groups_per_step": 4, + "lr": 5e-05, + "scheduler": "cosine", + "warmup_steps": 16, + "final_fraction": 0.1, + "weight_decay": 0.0, + "clip": 1.0, + "initial_loss_scale": 4.0, + "adapter": { + "layer": 9, + "rank": 16, + "trainable_parameters": 1212416 + } +} \ No newline at end of file diff --git a/runs/single_internvl/context/L09_s13/steps.jsonl b/runs/single_internvl/context/L09_s13/steps.jsonl new file mode 100644 index 0000000000000000000000000000000000000000..2baa37b8a471d6d9919af79f359b956d1640eafa --- /dev/null +++ b/runs/single_internvl/context/L09_s13/steps.jsonl @@ -0,0 +1,250 @@ +{"step": 1, "steps": 250, "lr": 3.125e-06, "grad_norm": 0.8271442651748657, "loss_scale": 4.0, "retries": 0, "seconds": 1.6, "max_memory_gb": 18.29, "ce": 0.5227766726748087, "kl": 0.08654405153356493, "context": 0.5660487064160407} +{"step": 2, "steps": 250, "lr": 6.25e-06, "grad_norm": 1.0194190740585327, "loss_scale": 4.0, "retries": 0, "seconds": 2.0, "max_memory_gb": 18.3, "ce": 0.6370677337035886, "kl": 0.4671201559073097, "context": 0.8706278188183205} +{"step": 3, "steps": 250, "lr": 9.375000000000001e-06, "grad_norm": 3.2324845790863037, "loss_scale": 4.0, "retries": 0, "seconds": 2.5, "max_memory_gb": 18.3, "ce": 1.0975014567375183, "kl": 0.0678083086386323, "context": 1.131405621767044} +{"step": 4, "steps": 250, "lr": 1.25e-05, "grad_norm": 1.7054495811462402, "loss_scale": 4.0, "retries": 0, "seconds": 3.1, "max_memory_gb": 18.3, "ce": 0.44897803019557614, "kl": 0.43064634570419, "context": 0.6643011873020441} +{"step": 5, "steps": 250, "lr": 1.5625e-05, "grad_norm": 0.03915682062506676, "loss_scale": 4.0, "retries": 0, "seconds": 3.6, "max_memory_gb": 18.3, "ce": 0.009200214844895527, "kl": 0.00823867807048373, "context": 0.013319553938345052} +{"step": 6, "steps": 250, "lr": 1.8750000000000002e-05, "grad_norm": 0.815216064453125, "loss_scale": 4.0, "retries": 0, "seconds": 4.1, "max_memory_gb": 18.3, "ce": 0.28358335938537493, "kl": 0.07536936150427209, "context": 0.3212680385040585} +{"step": 7, "steps": 250, "lr": 2.1875e-05, "grad_norm": 2.7183189392089844, "loss_scale": 4.0, "retries": 0, "seconds": 4.5, "max_memory_gb": 18.3, "ce": 0.8022173918434419, "kl": 0.12636772719349665, "context": 0.8654012361148489} +{"step": 8, "steps": 250, "lr": 2.5e-05, "grad_norm": 2.67185640335083, "loss_scale": 4.0, "retries": 0, "seconds": 5.1, "max_memory_gb": 18.31, "ce": 0.7379385884851217, "kl": 0.28558468469418585, "context": 0.8807309456169605} +{"step": 9, "steps": 250, "lr": 2.8125000000000003e-05, "grad_norm": 1.8745927810668945, "loss_scale": 4.0, "retries": 0, "seconds": 5.6, "max_memory_gb": 18.31, "ce": 0.4797918666154146, "kl": 0.13590747630223632, "context": 0.5477456115186214} +{"step": 10, "steps": 250, "lr": 3.125e-05, "grad_norm": 0.317951500415802, "loss_scale": 4.0, "retries": 0, "seconds": 6.1, "max_memory_gb": 18.31, "ce": 0.044975467906624544, "kl": 0.04510957468301058, "context": 0.06753025599755347} +{"step": 11, "steps": 250, "lr": 3.4375e-05, "grad_norm": 2.280679225921631, "loss_scale": 4.0, "retries": 0, "seconds": 6.6, "max_memory_gb": 18.31, "ce": 0.5959243912948295, "kl": 0.004643247368221637, "context": 0.5982459975057282} +{"step": 12, "steps": 250, "lr": 3.7500000000000003e-05, "grad_norm": 1.6529954671859741, "loss_scale": 4.0, "retries": 0, "seconds": 7.1, "max_memory_gb": 18.31, "ce": 0.3827545993444801, "kl": 0.00319813551095649, "context": 0.3843536680233228} +{"step": 13, "steps": 250, "lr": 4.0625000000000005e-05, "grad_norm": 1.248725414276123, "loss_scale": 4.0, "retries": 0, "seconds": 7.6, "max_memory_gb": 18.31, "ce": 0.3522583464255149, "kl": 0.017397199781044037, "context": 0.36095693572133314} +{"step": 14, "steps": 250, "lr": 4.375e-05, "grad_norm": 2.579228401184082, "loss_scale": 4.0, "retries": 0, "seconds": 8.0, "max_memory_gb": 18.31, "ce": 0.5509330034255981, "kl": 0.9350451952777803, "context": 1.018455609679222} +{"step": 15, "steps": 250, "lr": 4.6875e-05, "grad_norm": 2.0880556106567383, "loss_scale": 4.0, "retries": 0, "seconds": 8.5, "max_memory_gb": 18.31, "ce": 0.31610093341691936, "kl": 0.44536169442289975, "context": 0.5387817787511722} +{"step": 16, "steps": 250, "lr": 5e-05, "grad_norm": 0.759891927242279, "loss_scale": 4.0, "retries": 0, "seconds": 9.0, "max_memory_gb": 18.31, "ce": 0.12891674086131388, "kl": 0.0012284926854135847, "context": 0.12953098172874888} +{"step": 17, "steps": 250, "lr": 5e-05, "grad_norm": 0.8608388900756836, "loss_scale": 4.0, "retries": 0, "seconds": 9.5, "max_memory_gb": 18.31, "ce": 0.2140708527294919, "kl": 0.8308871428889688, "context": 0.6295144269824959} +{"step": 18, "steps": 250, "lr": 4.999797225177623e-05, "grad_norm": 1.788520336151123, "loss_scale": 4.0, "retries": 0, "seconds": 10.0, "max_memory_gb": 18.31, "ce": 0.19873233897669706, "kl": 0.21671955438068835, "context": 0.3070921176113188} +{"step": 19, "steps": 250, "lr": 4.999188937259492e-05, "grad_norm": 2.068955421447754, "loss_scale": 4.0, "retries": 0, "seconds": 10.5, "max_memory_gb": 18.31, "ce": 0.1890895813703537, "kl": 0.18551677279174328, "context": 0.2818479724228382} +{"step": 20, "steps": 250, "lr": 4.998175245886032e-05, "grad_norm": 0.0597178116440773, "loss_scale": 4.0, "retries": 0, "seconds": 11.0, "max_memory_gb": 18.31, "ce": 0.003075766884649056, "kl": 0.01989615909224085, "context": 0.013023846626310842} +{"step": 21, "steps": 250, "lr": 4.996756333769319e-05, "grad_norm": 0.6678013205528259, "loss_scale": 4.0, "retries": 0, "seconds": 11.5, "max_memory_gb": 18.31, "ce": 0.04003822491631581, "kl": 0.02181251803813211, "context": 0.05094448488125636} +{"step": 22, "steps": 250, "lr": 4.994932456660157e-05, "grad_norm": 0.7086774706840515, "loss_scale": 4.0, "retries": 0, "seconds": 12.0, "max_memory_gb": 18.31, "ce": 0.07449555552011589, "kl": 0.030170661382726394, "context": 0.08958088318468072} +{"step": 23, "steps": 250, "lr": 4.992703943301973e-05, "grad_norm": 1.7407026290893555, "loss_scale": 4.0, "retries": 0, "seconds": 12.5, "max_memory_gb": 18.31, "ce": 0.2526570201262075, "kl": 0.944913686374548, "context": 0.7251138626734246} +{"step": 24, "steps": 250, "lr": 4.9900711953715675e-05, "grad_norm": 0.7018343806266785, "loss_scale": 4.0, "retries": 0, "seconds": 13.0, "max_memory_gb": 18.31, "ce": 0.04862230969592929, "kl": 0.10348979581613094, "context": 0.10036720940843225} +{"step": 25, "steps": 250, "lr": 4.987034687406713e-05, "grad_norm": 0.8349769115447998, "loss_scale": 4.0, "retries": 0, "seconds": 13.5, "max_memory_gb": 18.31, "ce": 0.3307002711262612, "kl": 0.14427863294258714, "context": 0.4028395861387253} +{"step": 26, "steps": 250, "lr": 4.9835949667206215e-05, "grad_norm": 0.3110015094280243, "loss_scale": 4.0, "retries": 0, "seconds": 14.0, "max_memory_gb": 18.31, "ce": 0.0335920515935868, "kl": 0.660844745580107, "context": 0.3640144132077694} +{"step": 27, "steps": 250, "lr": 4.9797526533032945e-05, "grad_norm": 0.33323943614959717, "loss_scale": 4.0, "retries": 0, "seconds": 14.5, "max_memory_gb": 18.31, "ce": 0.08659846185764764, "kl": 0.3315321889349434, "context": 0.2523645634064451} +{"step": 28, "steps": 250, "lr": 4.9755084397097716e-05, "grad_norm": 0.7230067849159241, "loss_scale": 4.0, "retries": 0, "seconds": 15.0, "max_memory_gb": 18.31, "ce": 0.47781292255967855, "kl": 0.9552468643523753, "context": 0.9554363619536161} +{"step": 29, "steps": 250, "lr": 4.970863090935304e-05, "grad_norm": 0.49240514636039734, "loss_scale": 4.0, "retries": 0, "seconds": 15.5, "max_memory_gb": 18.31, "ce": 0.04400980763603002, "kl": 0.37875144119607285, "context": 0.2333855254109949} +{"step": 30, "steps": 250, "lr": 4.965817444277468e-05, "grad_norm": 0.357898086309433, "loss_scale": 4.0, "retries": 0, "seconds": 16.0, "max_memory_gb": 18.31, "ce": 0.03463703545094177, "kl": 0.0013297466712174355, "context": 0.0353019089470763} +{"step": 31, "steps": 250, "lr": 4.9603724091852456e-05, "grad_norm": 0.6178283095359802, "loss_scale": 4.0, "retries": 0, "seconds": 16.5, "max_memory_gb": 18.31, "ce": 0.19598457869142294, "kl": 0.09698183147702366, "context": 0.24447548761963844} +{"step": 32, "steps": 250, "lr": 4.9545289670951025e-05, "grad_norm": 0.2658081650733948, "loss_scale": 4.0, "retries": 0, "seconds": 17.0, "max_memory_gb": 18.31, "ce": 0.01752031990326941, "kl": 0.010145730506337713, "context": 0.022593185771256685} +{"step": 33, "steps": 250, "lr": 4.948288171254089e-05, "grad_norm": 0.2661970555782318, "loss_scale": 4.0, "retries": 0, "seconds": 17.5, "max_memory_gb": 18.31, "ce": 0.04412325518205762, "kl": 0.02139992080628872, "context": 0.05482321698218584} +{"step": 34, "steps": 250, "lr": 4.941651146529998e-05, "grad_norm": 0.4194871187210083, "loss_scale": 4.0, "retries": 0, "seconds": 18.0, "max_memory_gb": 18.31, "ce": 0.006565787218278274, "kl": 0.7858887769825742, "context": 0.39951018022111384} +{"step": 35, "steps": 250, "lr": 4.9346190892086174e-05, "grad_norm": 0.08731282502412796, "loss_scale": 4.0, "retries": 0, "seconds": 18.4, "max_memory_gb": 18.31, "ce": 0.005038796691223979, "kl": 0.0018180009569732647, "context": 0.005947797151748091} +{"step": 36, "steps": 250, "lr": 4.927193266778101e-05, "grad_norm": 0.25056618452072144, "loss_scale": 4.0, "retries": 0, "seconds": 18.9, "max_memory_gb": 18.31, "ce": 0.08494866860564798, "kl": 0.05330599605804309, "context": 0.11160166485933587} +{"step": 37, "steps": 250, "lr": 4.919375017700515e-05, "grad_norm": 0.442525714635849, "loss_scale": 4.0, "retries": 0, "seconds": 19.4, "max_memory_gb": 18.31, "ce": 0.1357577591552399, "kl": 0.7879719984484836, "context": 0.5297437568660825} +{"step": 38, "steps": 250, "lr": 4.911165751170588e-05, "grad_norm": 0.3238140344619751, "loss_scale": 4.0, "retries": 0, "seconds": 19.9, "max_memory_gb": 18.31, "ce": 0.02170454617589712, "kl": 0.48757938639028, "context": 0.26549423253163695} +{"step": 39, "steps": 250, "lr": 4.902566946861708e-05, "grad_norm": 0.509132444858551, "loss_scale": 4.0, "retries": 0, "seconds": 20.4, "max_memory_gb": 18.31, "ce": 0.11005407731136074, "kl": 0.010317796091840137, "context": 0.11521297291619703} +{"step": 40, "steps": 250, "lr": 4.89358015465922e-05, "grad_norm": 0.12750421464443207, "loss_scale": 4.0, "retries": 0, "seconds": 20.9, "max_memory_gb": 18.31, "ce": 0.010254015534883365, "kl": 0.008598726795753464, "context": 0.014553379034623504} +{"step": 41, "steps": 250, "lr": 4.884206994381078e-05, "grad_norm": 1.2191094160079956, "loss_scale": 4.0, "retries": 0, "seconds": 21.4, "max_memory_gb": 18.31, "ce": 0.38792001543333754, "kl": 0.12753031472675502, "context": 0.45168516458943486} +{"step": 42, "steps": 250, "lr": 4.874449155485866e-05, "grad_norm": 0.025241464376449585, "loss_scale": 4.0, "retries": 0, "seconds": 21.9, "max_memory_gb": 18.31, "ce": 0.0034998498886125162, "kl": 0.00039334110624622554, "context": 0.0036965205217711627} +{"step": 43, "steps": 250, "lr": 4.864308396768294e-05, "grad_norm": 0.049043942242860794, "loss_scale": 4.0, "retries": 0, "seconds": 22.4, "max_memory_gb": 18.31, "ce": 0.002619688202685211, "kl": 0.0026478303771000355, "context": 0.003943603434890974} +{"step": 44, "steps": 250, "lr": 4.853786546042184e-05, "grad_norm": 3.00893497467041, "loss_scale": 4.0, "retries": 0, "seconds": 22.9, "max_memory_gb": 18.31, "ce": 0.5377192363957874, "kl": 1.3693463196468656, "context": 1.22239238320617} +{"step": 45, "steps": 250, "lr": 4.842885499811012e-05, "grad_norm": 1.2965245246887207, "loss_scale": 4.0, "retries": 0, "seconds": 23.4, "max_memory_gb": 18.31, "ce": 0.8312955310684629, "kl": 0.23856686468207045, "context": 0.9505789534887299} +{"step": 46, "steps": 250, "lr": 4.831607222926078e-05, "grad_norm": 0.440765917301178, "loss_scale": 4.0, "retries": 0, "seconds": 23.9, "max_memory_gb": 18.31, "ce": 0.08233691944042221, "kl": 0.2625017975224182, "context": 0.21358781831804663} +{"step": 47, "steps": 250, "lr": 4.8199537482323545e-05, "grad_norm": 0.0677410289645195, "loss_scale": 4.0, "retries": 0, "seconds": 24.4, "max_memory_gb": 18.31, "ce": 0.0059955336328130215, "kl": 0.01629141572630033, "context": 0.014141241437755525} +{"step": 48, "steps": 250, "lr": 4.807927176202073e-05, "grad_norm": 0.08412934839725494, "loss_scale": 4.0, "retries": 0, "seconds": 24.9, "max_memory_gb": 18.31, "ce": 0.0035189492045901716, "kl": 0.0032775604850030504, "context": 0.0051577293779700994} +{"step": 49, "steps": 250, "lr": 4.795529674556129e-05, "grad_norm": 0.34749090671539307, "loss_scale": 4.0, "retries": 0, "seconds": 25.4, "max_memory_gb": 18.31, "ce": 0.01230293919070391, "kl": 0.3123348592353068, "context": 0.1684703636856284} +{"step": 50, "steps": 250, "lr": 4.782763477873361e-05, "grad_norm": 0.5234605669975281, "loss_scale": 4.0, "retries": 0, "seconds": 25.9, "max_memory_gb": 18.31, "ce": 0.09955064213863807, "kl": 0.00804259459982859, "context": 0.10357194223615807} +{"step": 51, "steps": 250, "lr": 4.769630887187782e-05, "grad_norm": 1.5784813165664673, "loss_scale": 4.0, "retries": 0, "seconds": 26.4, "max_memory_gb": 18.31, "ce": 0.5286470650771662, "kl": 0.018678359187106253, "context": 0.5379862242989475} +{"step": 52, "steps": 250, "lr": 4.756134269573832e-05, "grad_norm": 0.6477727890014648, "loss_scale": 4.0, "retries": 0, "seconds": 26.9, "max_memory_gb": 18.31, "ce": 0.022461388492956758, "kl": 0.24887905095238239, "context": 0.14690091158263385} +{"step": 53, "steps": 250, "lr": 4.7422760577197226e-05, "grad_norm": 0.6637927889823914, "loss_scale": 4.0, "retries": 0, "seconds": 27.4, "max_memory_gb": 18.31, "ce": 0.1681324343662709, "kl": 0.37186820479109883, "context": 0.3540665372274816} +{"step": 54, "steps": 250, "lr": 4.728058749488963e-05, "grad_norm": 0.5234547257423401, "loss_scale": 4.0, "retries": 0, "seconds": 27.9, "max_memory_gb": 18.31, "ce": 0.039591327440575697, "kl": 0.3379093054754776, "context": 0.20854597924335394} +{"step": 55, "steps": 250, "lr": 4.713484907470133e-05, "grad_norm": 0.11448168754577637, "loss_scale": 4.0, "retries": 0, "seconds": 28.4, "max_memory_gb": 18.31, "ce": 0.01837207085918635, "kl": 0.014111172952013362, "context": 0.025427657819818705} +{"step": 56, "steps": 250, "lr": 4.6985571585149876e-05, "grad_norm": 0.23100119829177856, "loss_scale": 4.0, "retries": 0, "seconds": 28.9, "max_memory_gb": 18.31, "ce": 0.020300200441852212, "kl": 0.25196727947331965, "context": 0.14628384495154023} +{"step": 57, "steps": 250, "lr": 4.6832781932649884e-05, "grad_norm": 0.09128088504076004, "loss_scale": 4.0, "retries": 0, "seconds": 29.4, "max_memory_gb": 18.31, "ce": 0.0099630608165171, "kl": 0.020286070295696845, "context": 0.02010609688295517} +{"step": 58, "steps": 250, "lr": 4.667650765666325e-05, "grad_norm": 0.08040177077054977, "loss_scale": 4.0, "retries": 0, "seconds": 30.0, "max_memory_gb": 18.31, "ce": 0.003375453467015177, "kl": 0.0008728118018552777, "context": 0.003811859409324825} +{"step": 59, "steps": 250, "lr": 4.651677692473538e-05, "grad_norm": 0.22107486426830292, "loss_scale": 4.0, "retries": 0, "seconds": 30.6, "max_memory_gb": 18.31, "ce": 0.03786852068151347, "kl": 0.061335934849921614, "context": 0.0685364871751517} +{"step": 60, "steps": 250, "lr": 4.635361852741808e-05, "grad_norm": 0.24085482954978943, "loss_scale": 4.0, "retries": 0, "seconds": 31.1, "max_memory_gb": 18.31, "ce": 0.08507500181440264, "kl": 0.07190882985014468, "context": 0.12102941703051329} +{"step": 61, "steps": 250, "lr": 4.6187061873080264e-05, "grad_norm": 0.6412568688392639, "loss_scale": 4.0, "retries": 0, "seconds": 31.6, "max_memory_gb": 18.31, "ce": 0.021548072196310386, "kl": 0.8686348678093054, "context": 0.45586551426094957} +{"step": 62, "steps": 250, "lr": 4.601713698260727e-05, "grad_norm": 0.1573086678981781, "loss_scale": 4.0, "retries": 0, "seconds": 32.1, "max_memory_gb": 18.31, "ce": 0.03023563709575683, "kl": 0.009415628908755025, "context": 0.03494345012586564} +{"step": 63, "steps": 250, "lr": 4.5843874483989746e-05, "grad_norm": 0.2326931655406952, "loss_scale": 4.0, "retries": 0, "seconds": 32.6, "max_memory_gb": 18.31, "ce": 0.04089385352563113, "kl": 0.12802209332585335, "context": 0.10490489937365055} +{"step": 64, "steps": 250, "lr": 4.566730560680316e-05, "grad_norm": 0.10524874925613403, "loss_scale": 4.0, "retries": 0, "seconds": 33.1, "max_memory_gb": 18.31, "ce": 0.014312720159068704, "kl": 0.010959225706756115, "context": 0.019792332721408457} +{"step": 65, "steps": 250, "lr": 4.548746217657878e-05, "grad_norm": 0.5247194170951843, "loss_scale": 4.0, "retries": 0, "seconds": 33.6, "max_memory_gb": 18.31, "ce": 0.1858377407770604, "kl": 0.157864173699636, "context": 0.2647698314394802} +{"step": 66, "steps": 250, "lr": 4.5304376609067396e-05, "grad_norm": 0.7164639234542847, "loss_scale": 4.0, "retries": 0, "seconds": 34.1, "max_memory_gb": 18.31, "ce": 0.15131554120659985, "kl": 0.06751369391258777, "context": 0.1850723881337899} +{"step": 67, "steps": 250, "lr": 4.5118081904396504e-05, "grad_norm": 0.05772680416703224, "loss_scale": 4.0, "retries": 0, "seconds": 34.6, "max_memory_gb": 18.31, "ce": 0.005335082896635868, "kl": 0.008159080247423844, "context": 0.009414623025804758} +{"step": 68, "steps": 250, "lr": 4.4928611641122225e-05, "grad_norm": 0.14815087616443634, "loss_scale": 4.0, "retries": 0, "seconds": 35.1, "max_memory_gb": 18.31, "ce": 0.0052886468401993625, "kl": 0.03838271252607228, "context": 0.024480002666678047} +{"step": 69, "steps": 250, "lr": 4.473599997017701e-05, "grad_norm": 0.7072883248329163, "loss_scale": 4.0, "retries": 0, "seconds": 35.6, "max_memory_gb": 18.31, "ce": 0.29753925022669137, "kl": 0.02242511764052324, "context": 0.30875180481234565} +{"step": 70, "steps": 250, "lr": 4.4540281608714053e-05, "grad_norm": 0.02482912503182888, "loss_scale": 4.0, "retries": 0, "seconds": 36.1, "max_memory_gb": 18.31, "ce": 0.0013508947158697993, "kl": 0.00032829580231918953, "context": 0.001515042626124341} +{"step": 71, "steps": 250, "lr": 4.434149183384977e-05, "grad_norm": 0.06355972588062286, "loss_scale": 4.0, "retries": 0, "seconds": 36.6, "max_memory_gb": 18.31, "ce": 0.008627928433270426, "kl": 0.015657501382520422, "context": 0.01645667853881605} +{"step": 72, "steps": 250, "lr": 4.413966647630534e-05, "grad_norm": 0.29173770546913147, "loss_scale": 4.0, "retries": 0, "seconds": 37.1, "max_memory_gb": 18.31, "ce": 0.015608827030519024, "kl": 0.40952062740689144, "context": 0.2203691330505535} +{"step": 73, "steps": 250, "lr": 4.3934841913948323e-05, "grad_norm": 1.103294849395752, "loss_scale": 4.0, "retries": 0, "seconds": 37.5, "max_memory_gb": 18.31, "ce": 0.35900196746661095, "kl": 0.09040127827938704, "context": 0.40420260293831234} +{"step": 74, "steps": 250, "lr": 4.372705506523583e-05, "grad_norm": 0.2852693498134613, "loss_scale": 4.0, "retries": 0, "seconds": 38.0, "max_memory_gb": 18.31, "ce": 0.06570670588280336, "kl": 0.11226970819006965, "context": 0.1218415599596483} +{"step": 75, "steps": 250, "lr": 4.351634338256017e-05, "grad_norm": 2.1051464080810547, "loss_scale": 4.0, "retries": 0, "seconds": 38.5, "max_memory_gb": 18.31, "ce": 0.4293172266188776, "kl": 0.9435356995090842, "context": 0.9010850894264877} +{"step": 76, "steps": 250, "lr": 4.330274484549825e-05, "grad_norm": 0.9508921504020691, "loss_scale": 4.0, "retries": 0, "seconds": 39.0, "max_memory_gb": 18.31, "ce": 0.017855786536529195, "kl": 1.6894398319127504, "context": 0.862575686420314} +{"step": 77, "steps": 250, "lr": 4.308629795396599e-05, "grad_norm": 0.061677172780036926, "loss_scale": 4.0, "retries": 0, "seconds": 39.6, "max_memory_gb": 18.31, "ce": 0.009079406954697333, "kl": 0.01153336325012333, "context": 0.01484608840837609} +{"step": 78, "steps": 250, "lr": 4.286704172127894e-05, "grad_norm": 0.1496879905462265, "loss_scale": 4.0, "retries": 0, "seconds": 40.1, "max_memory_gb": 18.31, "ce": 0.00187073097913526, "kl": 0.09037069733312819, "context": 0.047056080162292346} +{"step": 79, "steps": 250, "lr": 4.264501566712033e-05, "grad_norm": 0.22822611033916473, "loss_scale": 4.0, "retries": 0, "seconds": 40.6, "max_memory_gb": 18.31, "ce": 0.01495709349273966, "kl": 0.02559577004831226, "context": 0.027754977643780876} +{"step": 80, "steps": 250, "lr": 4.2420259810417894e-05, "grad_norm": 0.1653987318277359, "loss_scale": 4.0, "retries": 0, "seconds": 41.1, "max_memory_gb": 18.31, "ce": 0.0022222725056053605, "kl": 0.08628513733768273, "context": 0.04536484184427536} +{"step": 81, "steps": 250, "lr": 4.219281466213066e-05, "grad_norm": 0.4409562945365906, "loss_scale": 4.0, "retries": 0, "seconds": 41.6, "max_memory_gb": 18.31, "ce": 0.09439568291418254, "kl": 0.08000158457434736, "context": 0.13439647888299078} +{"step": 82, "steps": 250, "lr": 4.1962721217947135e-05, "grad_norm": 0.07800929248332977, "loss_scale": 4.0, "retries": 0, "seconds": 42.1, "max_memory_gb": 18.31, "ce": 0.008717981429072097, "kl": 0.0021005420185247203, "context": 0.009768252552021295} +{"step": 83, "steps": 250, "lr": 4.173002095089599e-05, "grad_norm": 0.5409626960754395, "loss_scale": 4.0, "retries": 0, "seconds": 42.7, "max_memory_gb": 18.31, "ce": 0.15981310090865009, "kl": 0.7408919536464964, "context": 0.5302590896026231} +{"step": 84, "steps": 250, "lr": 4.149475580387084e-05, "grad_norm": 0.37875252962112427, "loss_scale": 4.0, "retries": 0, "seconds": 43.2, "max_memory_gb": 18.31, "ce": 0.015677110078286205, "kl": 0.06023865819361163, "context": 0.045796439672812994} +{"step": 85, "steps": 250, "lr": 4.1256968182070275e-05, "grad_norm": 2.014613628387451, "loss_scale": 4.0, "retries": 0, "seconds": 43.7, "max_memory_gb": 18.31, "ce": 0.03627780273836834, "kl": 0.02656441303943069, "context": 0.04956001019172618} +{"step": 86, "steps": 250, "lr": 4.101670094535453e-05, "grad_norm": 1.3812342882156372, "loss_scale": 4.0, "retries": 0, "seconds": 44.3, "max_memory_gb": 18.31, "ce": 0.20767677365802228, "kl": 0.45914862665813416, "context": 0.4372510735411197} +{"step": 87, "steps": 250, "lr": 4.077399740052027e-05, "grad_norm": 0.8873803615570068, "loss_scale": 4.0, "retries": 0, "seconds": 44.8, "max_memory_gb": 18.31, "ce": 0.006786598765756935, "kl": 0.6768288677558303, "context": 0.3452010303735733} +{"step": 88, "steps": 250, "lr": 4.0528901293494775e-05, "grad_norm": 0.10907445847988129, "loss_scale": 4.0, "retries": 0, "seconds": 45.3, "max_memory_gb": 18.31, "ce": 0.011958211309320177, "kl": 0.024966539517322417, "context": 0.024441480980385677} +{"step": 89, "steps": 250, "lr": 4.0281456801451004e-05, "grad_norm": 0.11461703479290009, "loss_scale": 4.0, "retries": 0, "seconds": 45.8, "max_memory_gb": 18.31, "ce": 0.017589248155672976, "kl": 0.01573455924517475, "context": 0.025456526906054933} +{"step": 90, "steps": 250, "lr": 4.0031708524844884e-05, "grad_norm": 0.08990652859210968, "loss_scale": 4.0, "retries": 0, "seconds": 46.3, "max_memory_gb": 18.31, "ce": 0.010102794025442563, "kl": 0.005364817508962005, "context": 0.012785202823579311} +{"step": 91, "steps": 250, "lr": 3.977970147937635e-05, "grad_norm": 1.0030909776687622, "loss_scale": 4.0, "retries": 0, "seconds": 46.9, "max_memory_gb": 18.31, "ce": 0.2547361430770252, "kl": 0.4817892905557528, "context": 0.49563080046209507} +{"step": 92, "steps": 250, "lr": 3.952548108787553e-05, "grad_norm": 1.6664037704467773, "loss_scale": 4.0, "retries": 0, "seconds": 47.4, "max_memory_gb": 18.31, "ce": 0.3109652580715192, "kl": 0.4325068355174153, "context": 0.5272186764123035} +{"step": 93, "steps": 250, "lr": 3.9269093172115496e-05, "grad_norm": 1.1071345806121826, "loss_scale": 4.0, "retries": 0, "seconds": 48.0, "max_memory_gb": 18.31, "ce": 0.055781604372896254, "kl": 0.4202911021639011, "context": 0.2659271619631909} +{"step": 94, "steps": 250, "lr": 3.901058394455323e-05, "grad_norm": 1.35893976688385, "loss_scale": 4.0, "retries": 0, "seconds": 48.5, "max_memory_gb": 18.31, "ce": 0.16453051622374915, "kl": 0.08305572860626853, "context": 0.2060583880811464} +{"step": 95, "steps": 250, "lr": 3.875e-05, "grad_norm": 1.5679717063903809, "loss_scale": 4.0, "retries": 0, "seconds": 49.0, "max_memory_gb": 18.31, "ce": 0.5703151841880754, "kl": 0.02564563532359898, "context": 0.5831379895098507} +{"step": 96, "steps": 250, "lr": 3.8487388307223005e-05, "grad_norm": 0.895445704460144, "loss_scale": 4.0, "retries": 0, "seconds": 49.5, "max_memory_gb": 18.31, "ce": 0.42565906509844353, "kl": 0.007494138495530933, "context": 0.42940612958045676} +{"step": 97, "steps": 250, "lr": 3.822279620047943e-05, "grad_norm": 2.331935405731201, "loss_scale": 4.0, "retries": 0, "seconds": 50.0, "max_memory_gb": 18.31, "ce": 0.022300483309663832, "kl": 0.286358773795655, "context": 0.16547987703233957} +{"step": 98, "steps": 250, "lr": 3.795627137098479e-05, "grad_norm": 0.13080203533172607, "loss_scale": 4.0, "retries": 0, "seconds": 50.6, "max_memory_gb": 18.31, "ce": 0.00347155389317777, "kl": 0.09984395488572773, "context": 0.05339353226736421} +{"step": 99, "steps": 250, "lr": 3.7687861858316844e-05, "grad_norm": 0.17755316197872162, "loss_scale": 4.0, "retries": 0, "seconds": 51.1, "max_memory_gb": 18.31, "ce": 0.021960176831271383, "kl": 0.011929655222047586, "context": 0.02792500412033405} +{"step": 100, "steps": 250, "lr": 3.741761604175668e-05, "grad_norm": 0.5429468750953674, "loss_scale": 4.0, "retries": 0, "seconds": 51.6, "max_memory_gb": 18.31, "ce": 0.04133227617421653, "kl": 0.6180298188373854, "context": 0.3503471881558653} +{"step": 101, "steps": 250, "lr": 3.714558263156872e-05, "grad_norm": 0.30605149269104004, "loss_scale": 4.0, "retries": 0, "seconds": 52.1, "max_memory_gb": 18.31, "ce": 0.006698833196423948, "kl": 0.06649263732833788, "context": 0.039945151132997125} +{"step": 102, "steps": 250, "lr": 3.6871810660220904e-05, "grad_norm": 0.2510412931442261, "loss_scale": 4.0, "retries": 0, "seconds": 52.7, "max_memory_gb": 18.31, "ce": 0.06549578881822526, "kl": 0.028672949643805623, "context": 0.07983226329088211} +{"step": 103, "steps": 250, "lr": 3.659634947354686e-05, "grad_norm": 0.1410365253686905, "loss_scale": 4.0, "retries": 0, "seconds": 53.2, "max_memory_gb": 18.31, "ce": 0.011974562577961478, "kl": 0.014821669186858344, "context": 0.01938539757975377} +{"step": 104, "steps": 250, "lr": 3.631924872185169e-05, "grad_norm": 0.04679744318127632, "loss_scale": 4.0, "retries": 0, "seconds": 53.6, "max_memory_gb": 18.31, "ce": 0.004186011996353045, "kl": 0.0036669072233053157, "context": 0.006019465668941848} +{"step": 105, "steps": 250, "lr": 3.6040558350962674e-05, "grad_norm": 0.28803887963294983, "loss_scale": 4.0, "retries": 0, "seconds": 54.2, "max_memory_gb": 18.31, "ce": 0.06890927022323012, "kl": 0.35519483988173306, "context": 0.246506683412008} +{"step": 106, "steps": 250, "lr": 3.576032859322687e-05, "grad_norm": 0.03341598063707352, "loss_scale": 4.0, "retries": 0, "seconds": 54.7, "max_memory_gb": 18.31, "ce": 0.004076313154655509, "kl": 0.0020114476828894112, "context": 0.00508203697972931} +{"step": 107, "steps": 250, "lr": 3.5478609958457056e-05, "grad_norm": 2.1681859493255615, "loss_scale": 4.0, "retries": 0, "seconds": 55.2, "max_memory_gb": 18.31, "ce": 0.4417700678022811, "kl": 0.08634567627649403, "context": 0.4849428880115738} +{"step": 108, "steps": 250, "lr": 3.519545322482755e-05, "grad_norm": 0.48647066950798035, "loss_scale": 4.0, "retries": 0, "seconds": 55.7, "max_memory_gb": 18.31, "ce": 0.0343145829574496, "kl": 0.00820200209432187, "context": 0.03841558207932394} +{"step": 109, "steps": 250, "lr": 3.491090942972179e-05, "grad_norm": 0.6846002340316772, "loss_scale": 4.0, "retries": 0, "seconds": 56.2, "max_memory_gb": 18.31, "ce": 0.06137305760057643, "kl": 0.3228507824242115, "context": 0.22279844619333744} +{"step": 110, "steps": 250, "lr": 3.4625029860533135e-05, "grad_norm": 1.50822913646698, "loss_scale": 4.0, "retries": 0, "seconds": 56.7, "max_memory_gb": 18.31, "ce": 0.1374713177792728, "kl": 0.28585120174102485, "context": 0.28039692249149084} +{"step": 111, "steps": 250, "lr": 3.433786604542056e-05, "grad_norm": 0.024111775681376457, "loss_scale": 4.0, "retries": 0, "seconds": 57.2, "max_memory_gb": 18.31, "ce": 0.0021621047926601022, "kl": 0.0017083865950553445, "context": 0.003016298040165566} +{"step": 112, "steps": 250, "lr": 3.404946974402104e-05, "grad_norm": 0.10347596555948257, "loss_scale": 4.0, "retries": 0, "seconds": 57.7, "max_memory_gb": 18.31, "ce": 0.007018106174655259, "kl": 0.0007215518126031384, "context": 0.0073788820882327855} +{"step": 113, "steps": 250, "lr": 3.375989293812019e-05, "grad_norm": 0.38887935876846313, "loss_scale": 4.0, "retries": 0, "seconds": 58.2, "max_memory_gb": 18.31, "ce": 0.012900965521112084, "kl": 0.3454453523736447, "context": 0.18562364811077714} +{"step": 114, "steps": 250, "lr": 3.3469187822282766e-05, "grad_norm": 1.7630459070205688, "loss_scale": 4.0, "retries": 0, "seconds": 58.7, "max_memory_gb": 18.31, "ce": 0.018704996880842373, "kl": 1.1927293869084679, "context": 0.6150696892291307} +{"step": 115, "steps": 250, "lr": 3.317740679444499e-05, "grad_norm": 0.1101866215467453, "loss_scale": 4.0, "retries": 0, "seconds": 59.2, "max_memory_gb": 18.31, "ce": 0.008981350198155269, "kl": 0.045786924281856045, "context": 0.031874812266323715} +{"step": 116, "steps": 250, "lr": 3.2884602446470056e-05, "grad_norm": 0.9345053434371948, "loss_scale": 4.0, "retries": 0, "seconds": 59.7, "max_memory_gb": 18.31, "ce": 0.1414639625363634, "kl": 0.45878546524545527, "context": 0.37085671009117505} +{"step": 117, "steps": 250, "lr": 3.259082755466876e-05, "grad_norm": 1.257325530052185, "loss_scale": 4.0, "retries": 0, "seconds": 60.2, "max_memory_gb": 18.32, "ce": 0.05684905192174483, "kl": 0.09393464975437382, "context": 0.10381637676618993} +{"step": 118, "steps": 250, "lr": 3.229613507028686e-05, "grad_norm": 0.4413956105709076, "loss_scale": 4.0, "retries": 0, "seconds": 60.7, "max_memory_gb": 18.32, "ce": 0.17424805247719632, "kl": 0.00947758858819725, "context": 0.17898684306419455} +{"step": 119, "steps": 250, "lr": 3.2000578109961004e-05, "grad_norm": 0.14378707110881805, "loss_scale": 4.0, "retries": 0, "seconds": 61.3, "max_memory_gb": 18.32, "ce": 0.015232775811455213, "kl": 0.017918615620601486, "context": 0.024192083161324263} +{"step": 120, "steps": 250, "lr": 3.1704209946144616e-05, "grad_norm": 1.4914829730987549, "loss_scale": 4.0, "retries": 0, "seconds": 61.8, "max_memory_gb": 18.32, "ce": 0.5377920826722402, "kl": 0.19317668967414647, "context": 0.6343804246280342} +{"step": 121, "steps": 250, "lr": 3.140708399750594e-05, "grad_norm": 0.2904178202152252, "loss_scale": 4.0, "retries": 0, "seconds": 62.3, "max_memory_gb": 18.32, "ce": 0.053207829125312855, "kl": 0.038793847226770595, "context": 0.07260475229122676} +{"step": 122, "steps": 250, "lr": 3.1109253819299606e-05, "grad_norm": 0.33162903785705566, "loss_scale": 4.0, "retries": 0, "seconds": 62.9, "max_memory_gb": 18.32, "ce": 0.07311893119185697, "kl": 0.01699711628316436, "context": 0.08161749171995325} +{"step": 123, "steps": 250, "lr": 3.081077309371361e-05, "grad_norm": 0.1969805657863617, "loss_scale": 4.0, "retries": 0, "seconds": 63.4, "max_memory_gb": 18.32, "ce": 0.03933385293930769, "kl": 0.021546875475905836, "context": 0.050107293063774705} +{"step": 124, "steps": 250, "lr": 3.051169562019339e-05, "grad_norm": 1.7286057472229004, "loss_scale": 4.0, "retries": 0, "seconds": 63.9, "max_memory_gb": 18.32, "ce": 0.6492247777641751, "kl": 0.5645864425350737, "context": 0.9315179915865883} +{"step": 125, "steps": 250, "lr": 3.021207530574477e-05, "grad_norm": 0.28831401467323303, "loss_scale": 4.0, "retries": 0, "seconds": 64.4, "max_memory_gb": 18.32, "ce": 0.10890397147159092, "kl": 0.08470876279170625, "context": 0.15125835611252114} +{"step": 126, "steps": 250, "lr": 2.991196615521758e-05, "grad_norm": 2.0238516330718994, "loss_scale": 4.0, "retries": 0, "seconds": 65.0, "max_memory_gb": 18.32, "ce": 0.5274408552795649, "kl": 0.14252504706382751, "context": 0.5987033825367689} +{"step": 127, "steps": 250, "lr": 2.9611422261571536e-05, "grad_norm": 0.10245393961668015, "loss_scale": 4.0, "retries": 0, "seconds": 65.5, "max_memory_gb": 18.32, "ce": 0.014558817027136683, "kl": 0.0013071886532998178, "context": 0.01521241175942123} +{"step": 128, "steps": 250, "lr": 2.931049779612634e-05, "grad_norm": 1.8297501802444458, "loss_scale": 4.0, "retries": 0, "seconds": 66.0, "max_memory_gb": 18.32, "ce": 0.23247215754236095, "kl": 1.1092689763754606, "context": 0.7871066400548443} +{"step": 129, "steps": 250, "lr": 2.9009246998797557e-05, "grad_norm": 0.27146583795547485, "loss_scale": 4.0, "retries": 0, "seconds": 66.5, "max_memory_gb": 18.32, "ce": 0.07202746835537255, "kl": 0.03164801249658922, "context": 0.08785147545859218} +{"step": 130, "steps": 250, "lr": 2.8707724168320238e-05, "grad_norm": 0.22427256405353546, "loss_scale": 4.0, "retries": 0, "seconds": 67.0, "max_memory_gb": 18.32, "ce": 0.044771852204576135, "kl": 0.24904753312875982, "context": 0.16929561854340136} +{"step": 131, "steps": 250, "lr": 2.840598365246184e-05, "grad_norm": 2.076586961746216, "loss_scale": 4.0, "retries": 0, "seconds": 67.6, "max_memory_gb": 18.32, "ce": 0.2482953515718691, "kl": 0.37229548415052705, "context": 0.43444310809718445} +{"step": 132, "steps": 250, "lr": 2.8104079838226383e-05, "grad_norm": 0.34264352917671204, "loss_scale": 4.0, "retries": 0, "seconds": 68.1, "max_memory_gb": 18.32, "ce": 0.16285525303101167, "kl": 0.029845034678146476, "context": 0.17777776850562077} +{"step": 133, "steps": 250, "lr": 2.7802067142051446e-05, "grad_norm": 0.6427749991416931, "loss_scale": 4.0, "retries": 0, "seconds": 68.6, "max_memory_gb": 18.32, "ce": 0.3505528871319257, "kl": 0.18939941423013806, "context": 0.44525258988142014} +{"step": 134, "steps": 250, "lr": 2.7500000000000004e-05, "grad_norm": 0.45090335607528687, "loss_scale": 4.0, "retries": 0, "seconds": 69.2, "max_memory_gb": 18.32, "ce": 0.2123092158435611, "kl": 0.0054588370049373225, "context": 0.21503863155521685} +{"step": 135, "steps": 250, "lr": 2.7197932857948566e-05, "grad_norm": 0.11884354799985886, "loss_scale": 4.0, "retries": 0, "seconds": 69.7, "max_memory_gb": 18.32, "ce": 0.022731609467882663, "kl": 0.09974041191162542, "context": 0.07260181364836171} +{"step": 136, "steps": 250, "lr": 2.689592016177363e-05, "grad_norm": 2.12007737159729, "loss_scale": 4.0, "retries": 0, "seconds": 70.2, "max_memory_gb": 18.32, "ce": 0.12608389870729297, "kl": 0.1442023723211605, "context": 0.19818507798481733} +{"step": 137, "steps": 250, "lr": 2.659401634753817e-05, "grad_norm": 0.207941472530365, "loss_scale": 4.0, "retries": 0, "seconds": 70.7, "max_memory_gb": 18.32, "ce": 0.009792064927751198, "kl": 0.025216998779796995, "context": 0.022400564281269908} +{"step": 138, "steps": 250, "lr": 2.6292275831679764e-05, "grad_norm": 1.2699213027954102, "loss_scale": 4.0, "retries": 0, "seconds": 71.2, "max_memory_gb": 18.32, "ce": 0.07596478005871177, "kl": 0.14272892195731401, "context": 0.14732923917472363} +{"step": 139, "steps": 250, "lr": 2.5990753001202445e-05, "grad_norm": 1.096739649772644, "loss_scale": 4.0, "retries": 0, "seconds": 71.8, "max_memory_gb": 18.32, "ce": 0.004828946679481305, "kl": 0.26618798770596186, "context": 0.13792294048471376} +{"step": 140, "steps": 250, "lr": 2.5689502203873672e-05, "grad_norm": 0.054367389529943466, "loss_scale": 4.0, "retries": 0, "seconds": 72.3, "max_memory_gb": 18.32, "ce": 0.0044682776206173, "kl": 0.002020289517531637, "context": 0.005478422506712377} +{"step": 141, "steps": 250, "lr": 2.538857773842847e-05, "grad_norm": 0.04765349254012108, "loss_scale": 4.0, "retries": 0, "seconds": 72.9, "max_memory_gb": 18.32, "ce": 0.006385273736668751, "kl": 0.005924487952142954, "context": 0.00934751785825938} +{"step": 142, "steps": 250, "lr": 2.5088033844782426e-05, "grad_norm": 0.5509640574455261, "loss_scale": 4.0, "retries": 0, "seconds": 73.4, "max_memory_gb": 18.32, "ce": 0.04319983260938898, "kl": 0.016914468782488257, "context": 0.05165706668049097} +{"step": 143, "steps": 250, "lr": 2.4787924694255237e-05, "grad_norm": 1.706475853919983, "loss_scale": 4.0, "retries": 0, "seconds": 73.9, "max_memory_gb": 18.32, "ce": 0.4345039306936087, "kl": 0.06621464301861124, "context": 0.4676112487068167} +{"step": 144, "steps": 250, "lr": 2.4488304379806627e-05, "grad_norm": 0.5445912480354309, "loss_scale": 4.0, "retries": 0, "seconds": 74.4, "max_memory_gb": 18.32, "ce": 0.08396710082888603, "kl": 0.17749332924722694, "context": 0.17271377064753324} +{"step": 145, "steps": 250, "lr": 2.4189226906286388e-05, "grad_norm": 0.0486426055431366, "loss_scale": 4.0, "retries": 0, "seconds": 74.9, "max_memory_gb": 18.32, "ce": 0.0020827301923418418, "kl": 0.009160864201930963, "context": 0.006663162392214872} +{"step": 146, "steps": 250, "lr": 2.3890746180700395e-05, "grad_norm": 0.1099620908498764, "loss_scale": 4.0, "retries": 0, "seconds": 75.5, "max_memory_gb": 18.32, "ce": 0.008902106084860861, "kl": 0.025466759982009535, "context": 0.021635486104059964} +{"step": 147, "steps": 250, "lr": 2.359291600249407e-05, "grad_norm": 0.9082925319671631, "loss_scale": 4.0, "retries": 0, "seconds": 76.0, "max_memory_gb": 18.32, "ce": 0.47280303329898743, "kl": 0.5343655392789515, "context": 0.7399858174612746} +{"step": 148, "steps": 250, "lr": 2.3295790053855392e-05, "grad_norm": 0.06484975665807724, "loss_scale": 4.0, "retries": 0, "seconds": 76.5, "max_memory_gb": 18.32, "ce": 0.005693223254638724, "kl": 0.012625720228243154, "context": 0.012006083386950195} +{"step": 149, "steps": 250, "lr": 2.299942189003901e-05, "grad_norm": 0.15911516547203064, "loss_scale": 4.0, "retries": 0, "seconds": 77.0, "max_memory_gb": 18.32, "ce": 0.014988549461122602, "kl": 0.006503829874418443, "context": 0.018240464210975915} +{"step": 150, "steps": 250, "lr": 2.2703864929713144e-05, "grad_norm": 0.2762005031108856, "loss_scale": 4.0, "retries": 0, "seconds": 77.5, "max_memory_gb": 18.32, "ce": 0.07002074451884255, "kl": 0.021550355362705886, "context": 0.0807959190569818} +{"step": 151, "steps": 250, "lr": 2.2409172445331257e-05, "grad_norm": 0.22790715098381042, "loss_scale": 4.0, "retries": 0, "seconds": 78.0, "max_memory_gb": 18.32, "ce": 0.01807371100585442, "kl": 0.026572008966468275, "context": 0.031359715496364515} +{"step": 152, "steps": 250, "lr": 2.211539755352996e-05, "grad_norm": 0.46757057309150696, "loss_scale": 4.0, "retries": 0, "seconds": 78.5, "max_memory_gb": 18.32, "ce": 0.007226378686027601, "kl": 0.541598946353588, "context": 0.2780258469720138} +{"step": 153, "steps": 250, "lr": 2.1822593205555016e-05, "grad_norm": 0.36913588643074036, "loss_scale": 4.0, "retries": 0, "seconds": 79.0, "max_memory_gb": 18.32, "ce": 0.012728224461170612, "kl": 0.37946306485719106, "context": 0.2024597545605502} +{"step": 154, "steps": 250, "lr": 2.1530812177717236e-05, "grad_norm": 1.9851493835449219, "loss_scale": 4.0, "retries": 0, "seconds": 79.4, "max_memory_gb": 18.32, "ce": 0.08524354084511288, "kl": 0.08611329994164407, "context": 0.12830018845852464} +{"step": 155, "steps": 250, "lr": 2.1240107061879815e-05, "grad_norm": 0.39415082335472107, "loss_scale": 4.0, "retries": 0, "seconds": 79.9, "max_memory_gb": 18.32, "ce": 0.1896178545539442, "kl": 0.010229275863821385, "context": 0.19473249623115407} +{"step": 156, "steps": 250, "lr": 2.095053025597896e-05, "grad_norm": 0.8266386985778809, "loss_scale": 4.0, "retries": 0, "seconds": 80.4, "max_memory_gb": 18.32, "ce": 0.011402905936847674, "kl": 0.049649627452026834, "context": 0.036227720709575806} +{"step": 157, "steps": 250, "lr": 2.066213395457945e-05, "grad_norm": 0.48895263671875, "loss_scale": 4.0, "retries": 0, "seconds": 80.9, "max_memory_gb": 18.32, "ce": 0.03306111495476216, "kl": 0.03383333087549545, "context": 0.049977780785411596} +{"step": 158, "steps": 250, "lr": 2.0374970139466877e-05, "grad_norm": 0.3375145494937897, "loss_scale": 4.0, "retries": 0, "seconds": 81.4, "max_memory_gb": 18.32, "ce": 0.18081623836769722, "kl": 0.07093927668756805, "context": 0.21628588330349885} +{"step": 159, "steps": 250, "lr": 2.0089090570278213e-05, "grad_norm": 0.6929412484169006, "loss_scale": 4.0, "retries": 0, "seconds": 81.9, "max_memory_gb": 18.32, "ce": 0.006077875732444227, "kl": 0.09467799312551506, "context": 0.05341687402687967} +{"step": 160, "steps": 250, "lr": 1.9804546775172457e-05, "grad_norm": 0.558924674987793, "loss_scale": 4.0, "retries": 0, "seconds": 82.4, "max_memory_gb": 18.32, "ce": 0.030957703362219036, "kl": 0.3437402455601841, "context": 0.2028278261423111} +{"step": 161, "steps": 250, "lr": 1.9521390041542946e-05, "grad_norm": 0.3582874834537506, "loss_scale": 4.0, "retries": 0, "seconds": 82.9, "max_memory_gb": 18.32, "ce": 0.07374430830532219, "kl": 0.09094552803662737, "context": 0.11921707048895769} +{"step": 162, "steps": 250, "lr": 1.923967140677313e-05, "grad_norm": 0.9946955442428589, "loss_scale": 4.0, "retries": 0, "seconds": 83.4, "max_memory_gb": 18.32, "ce": 0.07356584181252401, "kl": 0.13267254547099583, "context": 0.1399021219694987} +{"step": 163, "steps": 250, "lr": 1.8959441649037335e-05, "grad_norm": 2.043882369995117, "loss_scale": 4.0, "retries": 0, "seconds": 83.9, "max_memory_gb": 18.32, "ce": 0.22046136611606926, "kl": 0.6851350664510392, "context": 0.5630288964603096} +{"step": 164, "steps": 250, "lr": 1.868075127814832e-05, "grad_norm": 4.664178371429443, "loss_scale": 4.0, "retries": 0, "seconds": 84.4, "max_memory_gb": 18.32, "ce": 0.08806918375194073, "kl": 0.12747575213506934, "context": 0.15180706145474687} +{"step": 165, "steps": 250, "lr": 1.8403650526453143e-05, "grad_norm": 0.08535894751548767, "loss_scale": 4.0, "retries": 0, "seconds": 85.0, "max_memory_gb": 18.32, "ce": 0.006539188834722154, "kl": 0.02726138438356429, "context": 0.02016988120158203} +{"step": 166, "steps": 250, "lr": 1.812818933977912e-05, "grad_norm": 1.4631235599517822, "loss_scale": 4.0, "retries": 0, "seconds": 85.6, "max_memory_gb": 18.32, "ce": 0.41539592354092747, "kl": 0.13450014259433374, "context": 0.4826460105832666} +{"step": 167, "steps": 250, "lr": 1.785441736843128e-05, "grad_norm": 0.15342864394187927, "loss_scale": 4.0, "retries": 0, "seconds": 86.3, "max_memory_gb": 18.32, "ce": 0.02108615287579596, "kl": 0.003318505856441334, "context": 0.022745405323803425} +{"step": 168, "steps": 250, "lr": 1.7582383958243314e-05, "grad_norm": 0.12498246878385544, "loss_scale": 4.0, "retries": 0, "seconds": 86.8, "max_memory_gb": 18.32, "ce": 0.02858746850688476, "kl": 0.0021921703591942787, "context": 0.02968355457414873} +{"step": 169, "steps": 250, "lr": 1.7312138141683164e-05, "grad_norm": 1.0954875946044922, "loss_scale": 4.0, "retries": 0, "seconds": 87.3, "max_memory_gb": 18.32, "ce": 0.08188906841678545, "kl": 1.406306192278862, "context": 0.7850421518087387} +{"step": 170, "steps": 250, "lr": 1.704372862901521e-05, "grad_norm": 0.03152351826429367, "loss_scale": 4.0, "retries": 0, "seconds": 87.8, "max_memory_gb": 18.32, "ce": 0.0021627194655593485, "kl": 0.0005765459645772353, "context": 0.002450992469675839} +{"step": 171, "steps": 250, "lr": 1.6777203799520573e-05, "grad_norm": 0.669542133808136, "loss_scale": 4.0, "retries": 0, "seconds": 88.4, "max_memory_gb": 18.32, "ce": 0.023814264928660123, "kl": 0.2440278367139399, "context": 0.1458281825762242} +{"step": 172, "steps": 250, "lr": 1.6512611692777008e-05, "grad_norm": 0.13192112743854523, "loss_scale": 4.0, "retries": 0, "seconds": 89.0, "max_memory_gb": 18.32, "ce": 0.01541077037109062, "kl": 0.0162338292284403, "context": 0.023527684912551194} +{"step": 173, "steps": 250, "lr": 1.6250000000000005e-05, "grad_norm": 0.2009027749300003, "loss_scale": 4.0, "retries": 0, "seconds": 89.5, "max_memory_gb": 18.32, "ce": 0.051115245500113815, "kl": 0.027026507697883062, "context": 0.06462850124808028} +{"step": 174, "steps": 250, "lr": 1.598941605544678e-05, "grad_norm": 0.19806121289730072, "loss_scale": 4.0, "retries": 0, "seconds": 90.1, "max_memory_gb": 18.32, "ce": 0.036475951070315205, "kl": 0.03471217744652222, "context": 0.05383204142708564} +{"step": 175, "steps": 250, "lr": 1.5730906827884506e-05, "grad_norm": 2.5790300369262695, "loss_scale": 4.0, "retries": 0, "seconds": 90.7, "max_memory_gb": 18.32, "ce": 0.2282873704098165, "kl": 0.0871333759278059, "context": 0.2718540634959936} +{"step": 176, "steps": 250, "lr": 1.5474518912124476e-05, "grad_norm": 0.06777597218751907, "loss_scale": 4.0, "retries": 0, "seconds": 91.3, "max_memory_gb": 18.32, "ce": 0.010046235911431722, "kl": 0.0027512679880601354, "context": 0.01142186984361615} +{"step": 177, "steps": 250, "lr": 1.5220298520623646e-05, "grad_norm": 0.6261573433876038, "loss_scale": 4.0, "retries": 0, "seconds": 91.9, "max_memory_gb": 18.32, "ce": 0.021006907903938554, "kl": 0.7123800847548409, "context": 0.3771969493391225} +{"step": 178, "steps": 250, "lr": 1.4968291475155121e-05, "grad_norm": 0.8276799917221069, "loss_scale": 4.0, "retries": 0, "seconds": 92.3, "max_memory_gb": 18.32, "ce": 0.0650369023787789, "kl": 0.13677083831862547, "context": 0.13342231791466475} +{"step": 179, "steps": 250, "lr": 1.4718543198548998e-05, "grad_norm": 0.7660202980041504, "loss_scale": 4.0, "retries": 0, "seconds": 92.8, "max_memory_gb": 18.32, "ce": 0.3503832175920252, "kl": 0.024003041777177714, "context": 0.3623847432609182} +{"step": 180, "steps": 250, "lr": 1.4471098706505227e-05, "grad_norm": 0.2262033224105835, "loss_scale": 4.0, "retries": 0, "seconds": 93.3, "max_memory_gb": 18.32, "ce": 0.042553066115942784, "kl": 0.027848524870933034, "context": 0.05647732809302397} +{"step": 181, "steps": 250, "lr": 1.4226002599479743e-05, "grad_norm": 0.1759469211101532, "loss_scale": 4.0, "retries": 0, "seconds": 93.9, "max_memory_gb": 18.32, "ce": 0.00865196471568197, "kl": 0.08995552315900568, "context": 0.05362972675357014} +{"step": 182, "steps": 250, "lr": 1.3983299054645471e-05, "grad_norm": 2.0340685844421387, "loss_scale": 4.0, "retries": 0, "seconds": 94.5, "max_memory_gb": 18.32, "ce": 0.02276845381129533, "kl": 0.3373914938838425, "context": 0.19146420565084554} +{"step": 183, "steps": 250, "lr": 1.3743031817929724e-05, "grad_norm": 0.028469692915678024, "loss_scale": 4.0, "retries": 0, "seconds": 95.0, "max_memory_gb": 18.32, "ce": 0.0028639693391596666, "kl": 0.0007628437815583311, "context": 0.0032453911990160123} +{"step": 184, "steps": 250, "lr": 1.3505244196129164e-05, "grad_norm": 0.043963946402072906, "loss_scale": 4.0, "retries": 0, "seconds": 95.5, "max_memory_gb": 18.32, "ce": 0.0029243247699923813, "kl": 0.0042506471945671365, "context": 0.005049648585554678} +{"step": 185, "steps": 250, "lr": 1.3269979049104014e-05, "grad_norm": 0.22163943946361542, "loss_scale": 4.0, "retries": 0, "seconds": 96.0, "max_memory_gb": 18.32, "ce": 0.009569005895173177, "kl": 0.0911978408730647, "context": 0.0551679263880942} +{"step": 186, "steps": 250, "lr": 1.3037278782052863e-05, "grad_norm": 1.892106533050537, "loss_scale": 4.0, "retries": 0, "seconds": 96.6, "max_memory_gb": 18.32, "ce": 0.09960941837562132, "kl": 0.25609112537495093, "context": 0.22765497320506256} +{"step": 187, "steps": 250, "lr": 1.2807185337869343e-05, "grad_norm": 0.3338058590888977, "loss_scale": 4.0, "retries": 0, "seconds": 97.2, "max_memory_gb": 18.32, "ce": 0.08485686097992584, "kl": 0.47198047052370384, "context": 0.32084711140487343} +{"step": 188, "steps": 250, "lr": 1.2579740189582115e-05, "grad_norm": 0.8469110727310181, "loss_scale": 4.0, "retries": 0, "seconds": 97.8, "max_memory_gb": 18.32, "ce": 0.2536181266805215, "kl": 0.013596369162314659, "context": 0.26041631288899225} +{"step": 189, "steps": 250, "lr": 1.2354984332879668e-05, "grad_norm": 2.0594820976257324, "loss_scale": 4.0, "retries": 0, "seconds": 98.4, "max_memory_gb": 18.32, "ce": 0.33921142714098096, "kl": 0.34005062258802354, "context": 0.5092367337783799} +{"step": 190, "steps": 250, "lr": 1.2132958278721064e-05, "grad_norm": 0.10661167651414871, "loss_scale": 4.0, "retries": 0, "seconds": 99.0, "max_memory_gb": 18.32, "ce": 0.004779007955221459, "kl": 0.03870864066993818, "context": 0.024133328115567565} +{"step": 191, "steps": 250, "lr": 1.1913702046034016e-05, "grad_norm": 3.748889923095703, "loss_scale": 4.0, "retries": 0, "seconds": 99.6, "max_memory_gb": 18.32, "ce": 0.281895003921818, "kl": 0.26668632698965666, "context": 0.41523816736298613} +{"step": 192, "steps": 250, "lr": 1.1697255154501756e-05, "grad_norm": 0.9218399524688721, "loss_scale": 4.0, "retries": 0, "seconds": 100.2, "max_memory_gb": 18.32, "ce": 0.4501953414583113, "kl": 0.07898770000974764, "context": 0.4896892016986385} +{"step": 193, "steps": 250, "lr": 1.1483656617439842e-05, "grad_norm": 0.06635238230228424, "loss_scale": 4.0, "retries": 0, "seconds": 100.7, "max_memory_gb": 18.32, "ce": 0.005328866038325941, "kl": 0.006209034671201152, "context": 0.008433383256488014} +{"step": 194, "steps": 250, "lr": 1.1272944934764177e-05, "grad_norm": 0.4182707965373993, "loss_scale": 4.0, "retries": 0, "seconds": 101.3, "max_memory_gb": 18.32, "ce": 0.044339890679111704, "kl": 0.03804808350105304, "context": 0.06336393236415461} +{"step": 195, "steps": 250, "lr": 1.1065158086051684e-05, "grad_norm": 0.0952487662434578, "loss_scale": 4.0, "retries": 0, "seconds": 101.9, "max_memory_gb": 18.32, "ce": 0.013598955716588534, "kl": 0.0048748231238278095, "context": 0.01603636769868899} +{"step": 196, "steps": 250, "lr": 1.086033352369467e-05, "grad_norm": 0.5404762029647827, "loss_scale": 4.0, "retries": 0, "seconds": 102.4, "max_memory_gb": 18.32, "ce": 0.0012784295322489925, "kl": 0.4532658958917182, "context": 0.22791137734020595} +{"step": 197, "steps": 250, "lr": 1.0658508166150224e-05, "grad_norm": 1.0653188228607178, "loss_scale": 4.0, "retries": 0, "seconds": 103.0, "max_memory_gb": 18.32, "ce": 0.007248063120641746, "kl": 0.7256390749753336, "context": 0.3700676026928704} +{"step": 198, "steps": 250, "lr": 1.0459718391285954e-05, "grad_norm": 0.31150034070014954, "loss_scale": 4.0, "retries": 0, "seconds": 103.6, "max_memory_gb": 18.32, "ce": 0.05658267832768615, "kl": 0.009387738668920065, "context": 0.061276546330191195} +{"step": 199, "steps": 250, "lr": 1.0264000029822999e-05, "grad_norm": 0.5542716383934021, "loss_scale": 4.0, "retries": 0, "seconds": 104.1, "max_memory_gb": 18.32, "ce": 0.14616044331341982, "kl": 0.168922874967393, "context": 0.23062187299365178} +{"step": 200, "steps": 250, "lr": 1.0071388358877775e-05, "grad_norm": 0.9514266848564148, "loss_scale": 4.0, "retries": 0, "seconds": 104.7, "max_memory_gb": 18.32, "ce": 0.3038477217196487, "kl": 0.04275200345000485, "context": 0.3252237191190943} +{"step": 201, "steps": 250, "lr": 9.881918095603501e-06, "grad_norm": 0.8363171815872192, "loss_scale": 4.0, "retries": 0, "seconds": 105.3, "max_memory_gb": 18.32, "ce": 0.08419196825707331, "kl": 0.12579877062671585, "context": 0.1470913535449654} +{"step": 202, "steps": 250, "lr": 9.695623390932612e-06, "grad_norm": 0.036321401596069336, "loss_scale": 4.0, "retries": 0, "seconds": 106.0, "max_memory_gb": 18.32, "ce": 0.0017547895604366204, "kl": 0.0034265310342789235, "context": 0.0034680550270422827} +{"step": 203, "steps": 250, "lr": 9.51253782342123e-06, "grad_norm": 1.1788201332092285, "loss_scale": 4.0, "retries": 0, "seconds": 106.6, "max_memory_gb": 18.32, "ce": 0.035094179707812145, "kl": 0.06069554144050926, "context": 0.06544194859452546} +{"step": 204, "steps": 250, "lr": 9.332694393196846e-06, "grad_norm": 0.18624043464660645, "loss_scale": 4.0, "retries": 0, "seconds": 107.2, "max_memory_gb": 18.32, "ce": 0.037583860917948186, "kl": 0.012563459225930274, "context": 0.043865590705536306} +{"step": 205, "steps": 250, "lr": 9.156125516010256e-06, "grad_norm": 0.37462764978408813, "loss_scale": 4.0, "retries": 0, "seconds": 107.7, "max_memory_gb": 18.32, "ce": 0.004491896863328293, "kl": 0.5346619561460102, "context": 0.27182287839241326} +{"step": 206, "steps": 250, "lr": 8.982863017392733e-06, "grad_norm": 0.07716964930295944, "loss_scale": 4.0, "retries": 0, "seconds": 108.3, "max_memory_gb": 18.32, "ce": 0.00826396469892643, "kl": 0.0035285149419905792, "context": 0.010028222197433934} +{"step": 207, "steps": 250, "lr": 8.81293812691974e-06, "grad_norm": 0.17314763367176056, "loss_scale": 4.0, "retries": 0, "seconds": 108.9, "max_memory_gb": 18.32, "ce": 0.06774423169554211, "kl": 0.0520109327044338, "context": 0.09374970174394548} +{"step": 208, "steps": 250, "lr": 8.646381472581927e-06, "grad_norm": 0.05145974084734917, "loss_scale": 4.0, "retries": 0, "seconds": 109.5, "max_memory_gb": 18.32, "ce": 0.0051718110698857345, "kl": 0.0016895994704100303, "context": 0.006016610859660432} +{"step": 209, "steps": 250, "lr": 8.483223075264622e-06, "grad_norm": 0.8586048483848572, "loss_scale": 4.0, "retries": 0, "seconds": 110.1, "max_memory_gb": 18.32, "ce": 0.17124988546129316, "kl": 0.11271541669702856, "context": 0.2276075899717398} +{"step": 210, "steps": 250, "lr": 8.323492343336747e-06, "grad_norm": 2.151212453842163, "loss_scale": 4.0, "retries": 0, "seconds": 110.7, "max_memory_gb": 18.32, "ce": 0.5012006575489067, "kl": 0.14344467690534657, "context": 0.5729230239521712} +{"step": 211, "steps": 250, "lr": 8.167218067350122e-06, "grad_norm": 0.33279332518577576, "loss_scale": 4.0, "retries": 0, "seconds": 111.3, "max_memory_gb": 18.32, "ce": 0.13705339113948867, "kl": 0.11417012475430965, "context": 0.19413844775408506} +{"step": 212, "steps": 250, "lr": 8.01442841485013e-06, "grad_norm": 2.7034249305725098, "loss_scale": 4.0, "retries": 0, "seconds": 111.9, "max_memory_gb": 18.32, "ce": 0.05815726285800338, "kl": 1.0388916685187723, "context": 0.5776030949200504} +{"step": 213, "steps": 250, "lr": 7.865150925298678e-06, "grad_norm": 0.5176781415939331, "loss_scale": 4.0, "retries": 0, "seconds": 112.5, "max_memory_gb": 18.32, "ce": 0.22414544023922645, "kl": 0.04793263312603813, "context": 0.24811175657669082} +{"step": 214, "steps": 250, "lr": 7.719412505110375e-06, "grad_norm": 0.7593203783035278, "loss_scale": 4.0, "retries": 0, "seconds": 113.1, "max_memory_gb": 18.32, "ce": 0.4173375059835962, "kl": 0.17540782832656987, "context": 0.5050414172874298} +{"step": 215, "steps": 250, "lr": 7.57723942280278e-06, "grad_norm": 1.616382360458374, "loss_scale": 4.0, "retries": 0, "seconds": 113.7, "max_memory_gb": 18.32, "ce": 0.6351635371902375, "kl": 0.02144004523302101, "context": 0.6458835719167837} +{"step": 216, "steps": 250, "lr": 7.438657304261687e-06, "grad_norm": 0.2195647954940796, "loss_scale": 4.0, "retries": 0, "seconds": 114.3, "max_memory_gb": 18.32, "ce": 0.012587543722474948, "kl": 0.11340446515532676, "context": 0.06928977745701559} +{"step": 217, "steps": 250, "lr": 7.30369112812219e-06, "grad_norm": 0.028939571231603622, "loss_scale": 4.0, "retries": 0, "seconds": 114.9, "max_memory_gb": 18.32, "ce": 0.0018326881545362994, "kl": 0.002521322116081137, "context": 0.003093349252594635} +{"step": 218, "steps": 250, "lr": 7.172365221266403e-06, "grad_norm": 0.6299047470092773, "loss_scale": 4.0, "retries": 0, "seconds": 115.5, "max_memory_gb": 18.32, "ce": 0.1379614882171154, "kl": 0.30430164001882076, "context": 0.29011230915784836} +{"step": 219, "steps": 250, "lr": 7.0447032544387154e-06, "grad_norm": 0.6244353652000427, "loss_scale": 4.0, "retries": 0, "seconds": 116.1, "max_memory_gb": 18.32, "ce": 0.008934030323871411, "kl": 0.12486064338008873, "context": 0.07136435052962042} +{"step": 220, "steps": 250, "lr": 6.920728237979274e-06, "grad_norm": 0.2068178802728653, "loss_scale": 4.0, "retries": 0, "seconds": 116.7, "max_memory_gb": 18.32, "ce": 0.05954356287838891, "kl": 0.007496145362893003, "context": 0.06329163374903146} +{"step": 221, "steps": 250, "lr": 6.800462517676456e-06, "grad_norm": 0.3665604293346405, "loss_scale": 4.0, "retries": 0, "seconds": 117.3, "max_memory_gb": 18.32, "ce": 0.006715113850077614, "kl": 0.5280895679934474, "context": 0.27075990397133864} +{"step": 222, "steps": 250, "lr": 6.683927770739222e-06, "grad_norm": 0.24672779440879822, "loss_scale": 4.0, "retries": 0, "seconds": 117.9, "max_memory_gb": 18.32, "ce": 0.029806770908180624, "kl": 0.0815273009189923, "context": 0.07057041925145313} +{"step": 223, "steps": 250, "lr": 6.571145001889892e-06, "grad_norm": 1.298209547996521, "loss_scale": 4.0, "retries": 0, "seconds": 118.4, "max_memory_gb": 18.32, "ce": 0.07232539333199384, "kl": 0.006806317574955756, "context": 0.07572855197213357} +{"step": 224, "steps": 250, "lr": 6.46213453957817e-06, "grad_norm": 0.49164706468582153, "loss_scale": 4.0, "retries": 0, "seconds": 118.9, "max_memory_gb": 18.32, "ce": 0.03061221557436511, "kl": 0.20563910141754604, "context": 0.13343176324269734} +{"step": 225, "steps": 250, "lr": 6.356916032317063e-06, "grad_norm": 0.26813822984695435, "loss_scale": 4.0, "retries": 0, "seconds": 119.4, "max_memory_gb": 18.32, "ce": 0.005213821874349378, "kl": 0.21708939332165755, "context": 0.1137585150427185} +{"step": 226, "steps": 250, "lr": 6.255508445141349e-06, "grad_norm": 0.6957666873931885, "loss_scale": 4.0, "retries": 0, "seconds": 119.8, "max_memory_gb": 18.32, "ce": 0.03985205735079944, "kl": 0.03757523198146373, "context": 0.05863967485493049} +{"step": 227, "steps": 250, "lr": 6.157930056189228e-06, "grad_norm": 0.20748257637023926, "loss_scale": 4.0, "retries": 0, "seconds": 120.3, "max_memory_gb": 18.32, "ce": 0.01817698341983487, "kl": 0.006372778468175966, "context": 0.021363373103667982} +{"step": 228, "steps": 250, "lr": 6.064198453407796e-06, "grad_norm": 0.4645305275917053, "loss_scale": 4.0, "retries": 0, "seconds": 120.8, "max_memory_gb": 18.32, "ce": 0.00376829403830925, "kl": 0.3009946657693945, "context": 0.15426562243374065} +{"step": 229, "steps": 250, "lr": 5.974330531382935e-06, "grad_norm": 0.41877979040145874, "loss_scale": 4.0, "retries": 0, "seconds": 121.3, "max_memory_gb": 18.32, "ce": 0.06019767601537751, "kl": 0.01832512508133277, "context": 0.06936023786693113} +{"step": 230, "steps": 250, "lr": 5.888342488294125e-06, "grad_norm": 1.691939115524292, "loss_scale": 4.0, "retries": 0, "seconds": 121.8, "max_memory_gb": 18.32, "ce": 0.5377778253605356, "kl": 0.35850373312132433, "context": 0.7170297135598958} +{"step": 231, "steps": 250, "lr": 5.806249822994849e-06, "grad_norm": 0.3646712899208069, "loss_scale": 4.0, "retries": 0, "seconds": 122.2, "max_memory_gb": 18.32, "ce": 0.037260949560732115, "kl": 0.14946302821772406, "context": 0.11199246181058697} +{"step": 232, "steps": 250, "lr": 5.728067332218994e-06, "grad_norm": 0.16631165146827698, "loss_scale": 4.0, "retries": 0, "seconds": 122.7, "max_memory_gb": 18.32, "ce": 0.02645563954138197, "kl": 0.011759177665226161, "context": 0.032335229276213795} +{"step": 233, "steps": 250, "lr": 5.65380910791383e-06, "grad_norm": 0.3173955976963043, "loss_scale": 4.0, "retries": 0, "seconds": 123.2, "max_memory_gb": 18.32, "ce": 0.009739138185977936, "kl": 0.10379538463894278, "context": 0.061636831145733595} +{"step": 234, "steps": 250, "lr": 5.583488534700022e-06, "grad_norm": 0.3355146646499634, "loss_scale": 4.0, "retries": 0, "seconds": 123.7, "max_memory_gb": 18.32, "ce": 0.011441822411143221, "kl": 0.055399888660758734, "context": 0.03914176487887744} +{"step": 235, "steps": 250, "lr": 5.517118287459118e-06, "grad_norm": 1.3060599565505981, "loss_scale": 4.0, "retries": 0, "seconds": 124.2, "max_memory_gb": 18.32, "ce": 0.3462743010313716, "kl": 0.01850422933603113, "context": 0.3555264195165364} +{"step": 236, "steps": 250, "lr": 5.454710329048978e-06, "grad_norm": 0.08857408165931702, "loss_scale": 4.0, "retries": 0, "seconds": 124.7, "max_memory_gb": 18.32, "ce": 0.009962266783986706, "kl": 0.018094234179443447, "context": 0.019009384093806148} +{"step": 237, "steps": 250, "lr": 5.3962759081475436e-06, "grad_norm": 0.382929265499115, "loss_scale": 4.0, "retries": 0, "seconds": 125.1, "max_memory_gb": 18.32, "ce": 0.07211616149288602, "kl": 0.024178749628845253, "context": 0.08420553717587609} +{"step": 238, "steps": 250, "lr": 5.341825557225321e-06, "grad_norm": 0.42208045721054077, "loss_scale": 4.0, "retries": 0, "seconds": 125.6, "max_memory_gb": 18.32, "ce": 0.03572762788098771, "kl": 0.009967416915969807, "context": 0.040711335284868255} +{"step": 239, "steps": 250, "lr": 5.291369090646961e-06, "grad_norm": 0.3444626033306122, "loss_scale": 4.0, "retries": 0, "seconds": 126.1, "max_memory_gb": 18.32, "ce": 0.1081127087236382, "kl": 0.026000415411544964, "context": 0.12111291458131745} +{"step": 240, "steps": 250, "lr": 5.244915602902289e-06, "grad_norm": 0.3190174698829651, "loss_scale": 4.0, "retries": 0, "seconds": 126.6, "max_memory_gb": 18.32, "ce": 0.09575906774261966, "kl": 0.04131793564738473, "context": 0.11641803581733257} +{"step": 241, "steps": 250, "lr": 5.20247346696706e-06, "grad_norm": 0.9160267114639282, "loss_scale": 4.0, "retries": 0, "seconds": 127.1, "max_memory_gb": 18.32, "ce": 0.10300391793862218, "kl": 0.022326273101498373, "context": 0.11416705274314154} +{"step": 242, "steps": 250, "lr": 5.164050332793787e-06, "grad_norm": 0.12464330345392227, "loss_scale": 4.0, "retries": 0, "seconds": 127.6, "max_memory_gb": 18.32, "ce": 0.005780153100204188, "kl": 0.060918145813047886, "context": 0.03623922460246831} +{"step": 243, "steps": 250, "lr": 5.129653125932873e-06, "grad_norm": 0.06001122295856476, "loss_scale": 4.0, "retries": 0, "seconds": 128.1, "max_memory_gb": 18.32, "ce": 0.005218150559812784, "kl": 0.00478324037976563, "context": 0.007609770400449634} +{"step": 244, "steps": 250, "lr": 5.099288046284329e-06, "grad_norm": 0.6015925407409668, "loss_scale": 4.0, "retries": 0, "seconds": 128.6, "max_memory_gb": 18.32, "ce": 0.04855867731384933, "kl": 0.19098267553818005, "context": 0.1440500175303896} +{"step": 245, "steps": 250, "lr": 5.072960566980277e-06, "grad_norm": 2.558736801147461, "loss_scale": 4.0, "retries": 0, "seconds": 129.1, "max_memory_gb": 18.32, "ce": 0.4200516671407968, "kl": 0.049737941357307136, "context": 0.44492062949575484} +{"step": 246, "steps": 250, "lr": 5.050675433398435e-06, "grad_norm": 0.019028982147574425, "loss_scale": 4.0, "retries": 0, "seconds": 129.6, "max_memory_gb": 18.32, "ce": 0.0012473551396396942, "kl": 0.000738473012461327, "context": 0.0016165915876626968} +{"step": 247, "steps": 250, "lr": 5.032436662306811e-06, "grad_norm": 0.38743194937705994, "loss_scale": 4.0, "retries": 0, "seconds": 130.1, "max_memory_gb": 18.32, "ce": 0.006482511409558356, "kl": 0.4753480465733446, "context": 0.24415653699543327} +{"step": 248, "steps": 250, "lr": 5.018247541139684e-06, "grad_norm": 0.20653633773326874, "loss_scale": 4.0, "retries": 0, "seconds": 130.6, "max_memory_gb": 18.32, "ce": 0.030540917068719864, "kl": 0.011063692392781377, "context": 0.03607276326511055} +{"step": 249, "steps": 250, "lr": 5.00811062740508e-06, "grad_norm": 0.3610062301158905, "loss_scale": 4.0, "retries": 0, "seconds": 131.1, "max_memory_gb": 18.32, "ce": 0.0423017586581409, "kl": 0.049997084570350125, "context": 0.06730029871687293} +{"step": 250, "steps": 250, "lr": 5.002027748223778e-06, "grad_norm": 0.047395769506692886, "loss_scale": 4.0, "retries": 0, "seconds": 131.6, "max_memory_gb": 18.32, "ce": 0.0061965717759449035, "kl": 0.00298009116522735, "context": 0.007686617551371455} diff --git a/runs/single_internvl/context/L11_s13/adapter/README.md b/runs/single_internvl/context/L11_s13/adapter/README.md new file mode 100644 index 0000000000000000000000000000000000000000..4f9f856cf0e1531b47e25d266833b2977a19d3ac --- /dev/null +++ b/runs/single_internvl/context/L11_s13/adapter/README.md @@ -0,0 +1,207 @@ +--- +base_model: /data/OPSD/assets/models/InternVL3_5-8B-HF +library_name: peft +pipeline_tag: text-generation +tags: +- base_model:adapter:/data/OPSD/assets/models/InternVL3_5-8B-HF +- lora +- transformers +--- + +# Model Card for Model ID + + + + + +## Model Details + +### Model Description + + + + + +- **Developed by:** [More Information Needed] +- **Funded by [optional]:** [More Information Needed] +- **Shared by [optional]:** [More Information Needed] +- **Model type:** [More Information Needed] +- **Language(s) (NLP):** [More Information Needed] +- **License:** [More Information Needed] +- **Finetuned from model [optional]:** [More Information Needed] + +### Model Sources [optional] + + + +- **Repository:** [More Information Needed] +- **Paper [optional]:** [More Information Needed] +- **Demo [optional]:** [More Information Needed] + +## Uses + + + +### Direct Use + + + +[More Information Needed] + +### Downstream Use [optional] + + + +[More Information Needed] + +### Out-of-Scope Use + + + +[More Information Needed] + +## Bias, Risks, and Limitations + + + +[More Information Needed] + +### Recommendations + + + +Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. More information needed for further recommendations. + +## How to Get Started with the Model + +Use the code below to get started with the model. + +[More Information Needed] + +## Training Details + +### Training Data + + + +[More Information Needed] + +### Training Procedure + + + +#### Preprocessing [optional] + +[More Information Needed] + + +#### Training Hyperparameters + +- **Training regime:** [More Information Needed] + +#### Speeds, Sizes, Times [optional] + + + +[More Information Needed] + +## Evaluation + + + +### Testing Data, Factors & Metrics + +#### Testing Data + + + +[More Information Needed] + +#### Factors + + + +[More Information Needed] + +#### Metrics + + + +[More Information Needed] + +### Results + +[More Information Needed] + +#### Summary + + + +## Model Examination [optional] + + + +[More Information Needed] + +## Environmental Impact + + + +Carbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700). + +- **Hardware Type:** [More Information Needed] +- **Hours used:** [More Information Needed] +- **Cloud Provider:** [More Information Needed] +- **Compute Region:** [More Information Needed] +- **Carbon Emitted:** [More Information Needed] + +## Technical Specifications [optional] + +### Model Architecture and Objective + +[More Information Needed] + +### Compute Infrastructure + +[More Information Needed] + +#### Hardware + +[More Information Needed] + +#### Software + +[More Information Needed] + +## Citation [optional] + + + +**BibTeX:** + +[More Information Needed] + +**APA:** + +[More Information Needed] + +## Glossary [optional] + + + +[More Information Needed] + +## More Information [optional] + +[More Information Needed] + +## Model Card Authors [optional] + +[More Information Needed] + +## Model Card Contact + +[More Information Needed] +### Framework versions + +- PEFT 0.20.0 \ No newline at end of file diff --git a/runs/single_internvl/context/L11_s13/adapter/adapter_config.json b/runs/single_internvl/context/L11_s13/adapter/adapter_config.json new file mode 100644 index 0000000000000000000000000000000000000000..ff69847132cdb995b31c67d2cb9ea5714c48a782 --- /dev/null +++ b/runs/single_internvl/context/L11_s13/adapter/adapter_config.json @@ -0,0 +1,50 @@ +{ + "alora_invocation_tokens": null, + "alpha_pattern": {}, + "arrow_config": null, + "auto_mapping": null, + "base_model_name_or_path": "/data/OPSD/assets/models/InternVL3_5-8B-HF", + "bias": "none", + "corda_config": null, + "ensure_weight_tying": false, + "eva_config": null, + "exclude_modules": null, + "fan_in_fan_out": false, + "inference_mode": true, + "init_lora_weights": true, + "layer_replication": null, + "layers_pattern": null, + "layers_to_transform": null, + "loftq_config": {}, + "lora_alpha": 32, + "lora_bias": false, + "lora_dropout": 0.05, + "lora_ga_config": null, + "megatron_config": null, + "megatron_core": "megatron.core", + "modules_to_save": null, + "monteclora_config": null, + "peft_type": "LORA", + "peft_version": "0.20.0", + "qalora_group_size": 16, + "r": 16, + "rank_pattern": {}, + "revision": null, + "target_modules": [ + "model.language_model.layers.11.self_attn.q_proj", + "model.language_model.layers.11.self_attn.o_proj", + "model.language_model.layers.11.self_attn.k_proj", + "model.language_model.layers.11.mlp.gate_proj", + "model.language_model.layers.11.mlp.up_proj", + "model.language_model.layers.11.mlp.down_proj", + "model.language_model.layers.11.self_attn.v_proj" + ], + "target_parameters": null, + "task_type": "CAUSAL_LM", + "trainable_token_indices": null, + "use_bdlora": null, + "use_dora": false, + "use_qalora": false, + "use_rslora": false, + "velora_config": null +} \ No newline at end of file diff --git a/runs/single_internvl/context/L11_s13/config.json b/runs/single_internvl/context/L11_s13/config.json new file mode 100644 index 0000000000000000000000000000000000000000..ba5f1a762a9b273e0db42f10386b6ea48c3cf751 --- /dev/null +++ b/runs/single_internvl/context/L11_s13/config.json @@ -0,0 +1,24 @@ +{ + "behavior": "context", + "seed": 13, + "model": "/data/OPSD/assets/models/InternVL3_5-8B-HF", + "backbone": "internvl", + "train": "/data/OPSD/assets/v3/context.jsonl", + "items": 1000, + "steps": 250, + "epochs": 1, + "steps_per_epoch": 250, + "groups_per_step": 4, + "lr": 5e-05, + "scheduler": "cosine", + "warmup_steps": 16, + "final_fraction": 0.1, + "weight_decay": 0.0, + "clip": 1.0, + "initial_loss_scale": 4.0, + "adapter": { + "layer": 11, + "rank": 16, + "trainable_parameters": 1212416 + } +} \ No newline at end of file diff --git a/runs/single_internvl/context/L11_s13/steps.jsonl b/runs/single_internvl/context/L11_s13/steps.jsonl new file mode 100644 index 0000000000000000000000000000000000000000..cc1392dbc96afa7ba6e5e632ae6430dd83c08b00 --- /dev/null +++ b/runs/single_internvl/context/L11_s13/steps.jsonl @@ -0,0 +1,250 @@ +{"step": 1, "steps": 250, "lr": 3.125e-06, "grad_norm": 0.9291117787361145, "loss_scale": 4.0, "retries": 0, "seconds": 1.5, "max_memory_gb": 18.13, "ce": 0.5227766726748087, "kl": 0.08654405153356493, "context": 0.5660487064160407} +{"step": 2, "steps": 250, "lr": 6.25e-06, "grad_norm": 1.2146095037460327, "loss_scale": 4.0, "retries": 0, "seconds": 1.9, "max_memory_gb": 18.14, "ce": 0.6385369354175054, "kl": 0.46720077286204287, "context": 0.872137291567924} +{"step": 3, "steps": 250, "lr": 9.375000000000001e-06, "grad_norm": 3.2859768867492676, "loss_scale": 4.0, "retries": 0, "seconds": 2.4, "max_memory_gb": 18.14, "ce": 1.0982384979724884, "kl": 0.06742501817643642, "context": 1.1319509968161583} +{"step": 4, "steps": 250, "lr": 1.25e-05, "grad_norm": 1.7314800024032593, "loss_scale": 4.0, "retries": 0, "seconds": 2.9, "max_memory_gb": 18.14, "ce": 0.4504057520243805, "kl": 0.4305011790897879, "context": 0.6656563331562211} +{"step": 5, "steps": 250, "lr": 1.5625e-05, "grad_norm": 0.04778214544057846, "loss_scale": 4.0, "retries": 0, "seconds": 3.4, "max_memory_gb": 18.14, "ce": 0.0093093293489801, "kl": 0.008332225967023987, "context": 0.013475442319759168} +{"step": 6, "steps": 250, "lr": 1.8750000000000002e-05, "grad_norm": 0.9157899618148804, "loss_scale": 4.0, "retries": 0, "seconds": 3.9, "max_memory_gb": 18.14, "ce": 0.28440476044488605, "kl": 0.07494310860420228, "context": 0.3218763086333638} +{"step": 7, "steps": 250, "lr": 2.1875e-05, "grad_norm": 2.833906412124634, "loss_scale": 4.0, "retries": 0, "seconds": 4.4, "max_memory_gb": 18.14, "ce": 0.7944834506288316, "kl": 0.12474629241989987, "context": 0.8568565983514418} +{"step": 8, "steps": 250, "lr": 2.5e-05, "grad_norm": 2.822571039199829, "loss_scale": 4.0, "retries": 0, "seconds": 4.9, "max_memory_gb": 18.15, "ce": 0.7360810004174709, "kl": 0.2855896190740168, "context": 0.8788758236914873} +{"step": 9, "steps": 250, "lr": 2.8125000000000003e-05, "grad_norm": 2.015159845352173, "loss_scale": 4.0, "retries": 0, "seconds": 5.4, "max_memory_gb": 18.15, "ce": 0.473635358735919, "kl": 0.13433880507363938, "context": 0.5408047698438168} +{"step": 10, "steps": 250, "lr": 3.125e-05, "grad_norm": 0.361081063747406, "loss_scale": 4.0, "retries": 0, "seconds": 5.9, "max_memory_gb": 18.15, "ce": 0.04405079077696428, "kl": 0.04561266698874533, "context": 0.06685712636681274} +{"step": 11, "steps": 250, "lr": 3.4375e-05, "grad_norm": 2.348710536956787, "loss_scale": 4.0, "retries": 0, "seconds": 6.4, "max_memory_gb": 18.15, "ce": 0.5854048133624019, "kl": 0.004641342862669262, "context": 0.5877254874285427} +{"step": 12, "steps": 250, "lr": 3.7500000000000003e-05, "grad_norm": 1.7136592864990234, "loss_scale": 4.0, "retries": 0, "seconds": 6.9, "max_memory_gb": 18.15, "ce": 0.3754042784912599, "kl": 0.003411983462228818, "context": 0.37711027907607786} +{"step": 13, "steps": 250, "lr": 4.0625000000000005e-05, "grad_norm": 1.2995357513427734, "loss_scale": 4.0, "retries": 0, "seconds": 7.4, "max_memory_gb": 18.15, "ce": 0.3426686481689103, "kl": 0.01761364178128133, "context": 0.35147545517247636} +{"step": 14, "steps": 250, "lr": 4.375e-05, "grad_norm": 2.693561315536499, "loss_scale": 4.0, "retries": 0, "seconds": 7.8, "max_memory_gb": 18.15, "ce": 0.5339736342430115, "kl": 0.936545180156827, "context": 1.002246230840683} +{"step": 15, "steps": 250, "lr": 4.6875e-05, "grad_norm": 2.100045919418335, "loss_scale": 4.0, "retries": 0, "seconds": 8.3, "max_memory_gb": 18.15, "ce": 0.29800998223231545, "kl": 0.4435098171679783, "context": 0.5197649094427561} +{"step": 16, "steps": 250, "lr": 5e-05, "grad_norm": 0.7714201211929321, "loss_scale": 4.0, "retries": 0, "seconds": 8.8, "max_memory_gb": 18.15, "ce": 0.11922081660304684, "kl": 0.0014665123312624928, "context": 0.1199540735615301} +{"step": 17, "steps": 250, "lr": 5e-05, "grad_norm": 0.9302612543106079, "loss_scale": 4.0, "retries": 0, "seconds": 9.3, "max_memory_gb": 18.15, "ce": 0.2026217080419883, "kl": 0.8302952987141907, "context": 0.6177693508798257} +{"step": 18, "steps": 250, "lr": 4.999797225177623e-05, "grad_norm": 1.6706310510635376, "loss_scale": 4.0, "retries": 0, "seconds": 9.8, "max_memory_gb": 18.15, "ce": 0.16845112724695355, "kl": 0.21751208881323691, "context": 0.2772071774234064} +{"step": 19, "steps": 250, "lr": 4.999188937259492e-05, "grad_norm": 1.8701218366622925, "loss_scale": 4.0, "retries": 0, "seconds": 10.3, "max_memory_gb": 18.15, "ce": 0.1619601845741272, "kl": 0.18363978248089552, "context": 0.2537800818681717} +{"step": 20, "steps": 250, "lr": 4.998175245886032e-05, "grad_norm": 0.06947007775306702, "loss_scale": 4.0, "retries": 0, "seconds": 10.8, "max_memory_gb": 18.15, "ce": 0.0028635507915169, "kl": 0.02024099533446133, "context": 0.012984048546059057} +{"step": 21, "steps": 250, "lr": 4.996756333769319e-05, "grad_norm": 0.4905403256416321, "loss_scale": 4.0, "retries": 0, "seconds": 11.3, "max_memory_gb": 18.15, "ce": 0.02794297413129243, "kl": 0.019889475164745818, "context": 0.037887711714574834} +{"step": 22, "steps": 250, "lr": 4.994932456660157e-05, "grad_norm": 0.5875646471977234, "loss_scale": 4.0, "retries": 0, "seconds": 11.8, "max_memory_gb": 18.15, "ce": 0.06109267122519668, "kl": 0.024305802968228818, "context": 0.07324557026731782} +{"step": 23, "steps": 250, "lr": 4.992703943301973e-05, "grad_norm": 2.360948324203491, "loss_scale": 4.0, "retries": 0, "seconds": 12.3, "max_memory_gb": 18.15, "ce": 0.23952790781004296, "kl": 0.9458142951457376, "context": 0.7124350688873164} +{"step": 24, "steps": 250, "lr": 4.9900711953715675e-05, "grad_norm": 0.5826642513275146, "loss_scale": 4.0, "retries": 0, "seconds": 12.8, "max_memory_gb": 18.15, "ce": 0.03765900246798992, "kl": 0.09360127756372094, "context": 0.0844596407841891} +{"step": 25, "steps": 250, "lr": 4.987034687406713e-05, "grad_norm": 0.88252854347229, "loss_scale": 4.0, "retries": 0, "seconds": 13.3, "max_memory_gb": 18.15, "ce": 0.3231272671473562, "kl": 0.14252867689356208, "context": 0.394391605630517} +{"step": 26, "steps": 250, "lr": 4.9835949667206215e-05, "grad_norm": 0.29429858922958374, "loss_scale": 4.0, "retries": 0, "seconds": 13.8, "max_memory_gb": 18.15, "ce": 0.028894161339849234, "kl": 0.6534270849078894, "context": 0.3556076902896166} +{"step": 27, "steps": 250, "lr": 4.9797526533032945e-05, "grad_norm": 0.29467660188674927, "loss_scale": 4.0, "retries": 0, "seconds": 14.3, "max_memory_gb": 18.15, "ce": 0.08123791154503124, "kl": 0.330468225374716, "context": 0.24647202234336874} +{"step": 28, "steps": 250, "lr": 4.9755084397097716e-05, "grad_norm": 0.8321999311447144, "loss_scale": 4.0, "retries": 0, "seconds": 14.8, "max_memory_gb": 18.15, "ce": 0.4775619732681662, "kl": 0.9533117769751698, "context": 0.954217866063118} +{"step": 29, "steps": 250, "lr": 4.970863090935304e-05, "grad_norm": 0.49203595519065857, "loss_scale": 4.0, "retries": 0, "seconds": 15.3, "max_memory_gb": 18.15, "ce": 0.04384113394189626, "kl": 0.3799553078133613, "context": 0.23381878877989948} +{"step": 30, "steps": 250, "lr": 4.965817444277468e-05, "grad_norm": 0.324317991733551, "loss_scale": 4.0, "retries": 0, "seconds": 15.8, "max_memory_gb": 18.15, "ce": 0.027416471697506495, "kl": 0.001247662873538502, "context": 0.02804030220977438} +{"step": 31, "steps": 250, "lr": 4.9603724091852456e-05, "grad_norm": 0.6024863123893738, "loss_scale": 4.0, "retries": 0, "seconds": 16.3, "max_memory_gb": 18.15, "ce": 0.18535410449840128, "kl": 0.09620195958996192, "context": 0.23345509357750416} +{"step": 32, "steps": 250, "lr": 4.9545289670951025e-05, "grad_norm": 0.2149583101272583, "loss_scale": 4.0, "retries": 0, "seconds": 16.8, "max_memory_gb": 18.15, "ce": 0.013136480120010674, "kl": 0.009218828759912867, "context": 0.017745894379913807} +{"step": 33, "steps": 250, "lr": 4.948288171254089e-05, "grad_norm": 0.2947569489479065, "loss_scale": 4.0, "retries": 0, "seconds": 17.3, "max_memory_gb": 18.15, "ce": 0.04429748607799411, "kl": 0.020478466991335154, "context": 0.054536718875169754} +{"step": 34, "steps": 250, "lr": 4.941651146529998e-05, "grad_norm": 0.5242428779602051, "loss_scale": 4.0, "retries": 0, "seconds": 17.8, "max_memory_gb": 18.15, "ce": 0.005159375308721792, "kl": 0.7852372446882327, "context": 0.39777801007949165} +{"step": 35, "steps": 250, "lr": 4.9346190892086174e-05, "grad_norm": 0.06784414499998093, "loss_scale": 4.0, "retries": 0, "seconds": 18.3, "max_memory_gb": 18.15, "ce": 0.0043062665208708495, "kl": 0.001698021695119678, "context": 0.005155277482117526} +{"step": 36, "steps": 250, "lr": 4.927193266778101e-05, "grad_norm": 0.2684260606765747, "loss_scale": 4.0, "retries": 0, "seconds": 18.8, "max_memory_gb": 18.15, "ce": 0.08695755233929958, "kl": 0.052084985247347504, "context": 0.11300004489021376} +{"step": 37, "steps": 250, "lr": 4.919375017700515e-05, "grad_norm": 0.4864974617958069, "loss_scale": 4.0, "retries": 0, "seconds": 19.3, "max_memory_gb": 18.15, "ce": 0.13671289055491798, "kl": 0.779988172929734, "context": 0.5267069581896067} +{"step": 38, "steps": 250, "lr": 4.911165751170588e-05, "grad_norm": 0.30756163597106934, "loss_scale": 4.0, "retries": 0, "seconds": 19.7, "max_memory_gb": 18.15, "ce": 0.020800873870030046, "kl": 0.4772206577472389, "context": 0.2594112087972462} +{"step": 39, "steps": 250, "lr": 4.902566946861708e-05, "grad_norm": 0.48471468687057495, "loss_scale": 4.0, "retries": 0, "seconds": 20.2, "max_memory_gb": 18.15, "ce": 0.1070868434762815, "kl": 0.010766851526568644, "context": 0.1124702668748796} +{"step": 40, "steps": 250, "lr": 4.89358015465922e-05, "grad_norm": 0.11484257131814957, "loss_scale": 4.0, "retries": 0, "seconds": 20.7, "max_memory_gb": 18.15, "ce": 0.008355108293471858, "kl": 0.008493409113725647, "context": 0.012601812661159784} +{"step": 41, "steps": 250, "lr": 4.884206994381078e-05, "grad_norm": 1.4829702377319336, "loss_scale": 4.0, "retries": 0, "seconds": 21.2, "max_memory_gb": 18.15, "ce": 0.385015873296652, "kl": 0.12365006655454636, "context": 0.44684090465307236} +{"step": 42, "steps": 250, "lr": 4.874449155485866e-05, "grad_norm": 0.021626770496368408, "loss_scale": 4.0, "retries": 0, "seconds": 21.8, "max_memory_gb": 18.15, "ce": 0.003242572522140108, "kl": 0.00036962772628612584, "context": 0.0034273864584974945} +{"step": 43, "steps": 250, "lr": 4.864308396768294e-05, "grad_norm": 0.03350977972149849, "loss_scale": 4.0, "retries": 0, "seconds": 22.2, "max_memory_gb": 18.15, "ce": 0.0020211758310324512, "kl": 0.0022315212354442338, "context": 0.003136936473310925} +{"step": 44, "steps": 250, "lr": 4.853786546042184e-05, "grad_norm": 3.443107843399048, "loss_scale": 4.0, "retries": 0, "seconds": 22.7, "max_memory_gb": 18.15, "ce": 0.5276720335605205, "kl": 1.3351157978031551, "context": 1.1952299510594457} +{"step": 45, "steps": 250, "lr": 4.842885499811012e-05, "grad_norm": 1.4581031799316406, "loss_scale": 4.0, "retries": 0, "seconds": 23.2, "max_memory_gb": 18.15, "ce": 0.8125936535361689, "kl": 0.24667916005273582, "context": 0.9359332337917294} +{"step": 46, "steps": 250, "lr": 4.831607222926078e-05, "grad_norm": 0.5341800451278687, "loss_scale": 4.0, "retries": 0, "seconds": 23.9, "max_memory_gb": 18.15, "ce": 0.08185041529941373, "kl": 0.2538272278616205, "context": 0.2087640311801806} +{"step": 47, "steps": 250, "lr": 4.8199537482323545e-05, "grad_norm": 0.06719125807285309, "loss_scale": 4.0, "retries": 0, "seconds": 24.4, "max_memory_gb": 18.15, "ce": 0.005469122686918126, "kl": 0.01582615589722991, "context": 0.013382200675550848} +{"step": 48, "steps": 250, "lr": 4.807927176202073e-05, "grad_norm": 0.05683750659227371, "loss_scale": 4.0, "retries": 0, "seconds": 24.9, "max_memory_gb": 18.15, "ce": 0.0023746024089632556, "kl": 0.0033435368050049874, "context": 0.004046370799187571} +{"step": 49, "steps": 250, "lr": 4.795529674556129e-05, "grad_norm": 0.33464810252189636, "loss_scale": 4.0, "retries": 0, "seconds": 25.4, "max_memory_gb": 18.15, "ce": 0.013000300896237604, "kl": 0.3063874185476152, "context": 0.16619400266063167} +{"step": 50, "steps": 250, "lr": 4.782763477873361e-05, "grad_norm": 0.4262702763080597, "loss_scale": 4.0, "retries": 0, "seconds": 25.9, "max_memory_gb": 18.15, "ce": 0.08958555921708466, "kl": 0.008024155657039955, "context": 0.09359763995598769} +{"step": 51, "steps": 250, "lr": 4.769630887187782e-05, "grad_norm": 1.7017297744750977, "loss_scale": 4.0, "retries": 0, "seconds": 26.4, "max_memory_gb": 18.15, "ce": 0.5240702132905426, "kl": 0.020803862390152972, "context": 0.5344721331930486} +{"step": 52, "steps": 250, "lr": 4.756134269573832e-05, "grad_norm": 0.65752774477005, "loss_scale": 4.0, "retries": 0, "seconds": 26.9, "max_memory_gb": 18.15, "ce": 0.020795864576939493, "kl": 0.22530695877503604, "context": 0.13344934210181236} +{"step": 53, "steps": 250, "lr": 4.7422760577197226e-05, "grad_norm": 0.5713489651679993, "loss_scale": 4.0, "retries": 0, "seconds": 27.4, "max_memory_gb": 18.15, "ce": 0.14422618283424526, "kl": 0.4011994069442153, "context": 0.3448258969001472} +{"step": 54, "steps": 250, "lr": 4.728058749488963e-05, "grad_norm": 0.5661740303039551, "loss_scale": 4.0, "retries": 0, "seconds": 27.8, "max_memory_gb": 18.15, "ce": 0.0381371616240358, "kl": 0.32378645128483186, "context": 0.20003039049333893} +{"step": 55, "steps": 250, "lr": 4.713484907470133e-05, "grad_norm": 0.1165347546339035, "loss_scale": 4.0, "retries": 0, "seconds": 28.3, "max_memory_gb": 18.15, "ce": 0.017925121792359278, "kl": 0.013536558214468641, "context": 0.024693400191608816} +{"step": 56, "steps": 250, "lr": 4.6985571585149876e-05, "grad_norm": 0.24399791657924652, "loss_scale": 4.0, "retries": 0, "seconds": 28.8, "max_memory_gb": 18.15, "ce": 0.021400877623818815, "kl": 0.25242811581119895, "context": 0.14761492959223688} +{"step": 57, "steps": 250, "lr": 4.6832781932649884e-05, "grad_norm": 0.08965124189853668, "loss_scale": 4.0, "retries": 0, "seconds": 29.3, "max_memory_gb": 18.15, "ce": 0.009936314447259065, "kl": 0.020183196211291943, "context": 0.02002791162522044} +{"step": 58, "steps": 250, "lr": 4.667650765666325e-05, "grad_norm": 0.05846288055181503, "loss_scale": 4.0, "retries": 0, "seconds": 29.8, "max_memory_gb": 18.15, "ce": 0.0022479108465631725, "kl": 0.0010213859621899246, "context": 0.0027586038049776107} +{"step": 59, "steps": 250, "lr": 4.651677692473538e-05, "grad_norm": 0.23902785778045654, "loss_scale": 4.0, "retries": 0, "seconds": 30.4, "max_memory_gb": 18.15, "ce": 0.03808850413770415, "kl": 0.06277236915775575, "context": 0.06947468873113394} +{"step": 60, "steps": 250, "lr": 4.635361852741808e-05, "grad_norm": 0.2298155277967453, "loss_scale": 4.0, "retries": 0, "seconds": 30.8, "max_memory_gb": 18.15, "ce": 0.0831354790716432, "kl": 0.06896891753422096, "context": 0.11761993798427284} +{"step": 61, "steps": 250, "lr": 4.6187061873080264e-05, "grad_norm": 0.6572760343551636, "loss_scale": 4.0, "retries": 0, "seconds": 31.3, "max_memory_gb": 18.15, "ce": 0.02319277881179005, "kl": 0.8475354436795897, "context": 0.44696049464982934} +{"step": 62, "steps": 250, "lr": 4.601713698260727e-05, "grad_norm": 0.311914324760437, "loss_scale": 4.0, "retries": 0, "seconds": 31.8, "max_memory_gb": 18.15, "ce": 0.029605419942527078, "kl": 0.009449814933759626, "context": 0.034330328286159784} +{"step": 63, "steps": 250, "lr": 4.5843874483989746e-05, "grad_norm": 0.2377753108739853, "loss_scale": 4.0, "retries": 0, "seconds": 32.3, "max_memory_gb": 18.15, "ce": 0.039536086027510464, "kl": 0.1262578350433614, "context": 0.10266500432044268} +{"step": 64, "steps": 250, "lr": 4.566730560680316e-05, "grad_norm": 0.09804918617010117, "loss_scale": 4.0, "retries": 0, "seconds": 32.8, "max_memory_gb": 18.15, "ce": 0.013999658462125808, "kl": 0.011660254938760772, "context": 0.01982978609157726} +{"step": 65, "steps": 250, "lr": 4.548746217657878e-05, "grad_norm": 0.6114555597305298, "loss_scale": 4.0, "retries": 0, "seconds": 33.3, "max_memory_gb": 18.15, "ce": 0.18790228059515357, "kl": 0.1603372874087654, "context": 0.2680709243286401} +{"step": 66, "steps": 250, "lr": 4.5304376609067396e-05, "grad_norm": 0.6632749438285828, "loss_scale": 4.0, "retries": 0, "seconds": 33.8, "max_memory_gb": 18.15, "ce": 0.15592659945207288, "kl": 0.06541057590266064, "context": 0.18863188364912276} +{"step": 67, "steps": 250, "lr": 4.5118081904396504e-05, "grad_norm": 0.056615717709064484, "loss_scale": 4.0, "retries": 0, "seconds": 34.3, "max_memory_gb": 18.15, "ce": 0.0051587143243523315, "kl": 0.008212809192627901, "context": 0.009265118889743462} +{"step": 68, "steps": 250, "lr": 4.4928611641122225e-05, "grad_norm": 0.17257235944271088, "loss_scale": 4.0, "retries": 0, "seconds": 34.8, "max_memory_gb": 18.15, "ce": 0.005041736258135643, "kl": 0.035829770255077165, "context": 0.02295662112373975} +{"step": 69, "steps": 250, "lr": 4.473599997017701e-05, "grad_norm": 0.7452378273010254, "loss_scale": 4.0, "retries": 0, "seconds": 35.3, "max_memory_gb": 18.15, "ce": 0.29562281811377034, "kl": 0.02361737284809351, "context": 0.30743150983471423} +{"step": 70, "steps": 250, "lr": 4.4540281608714053e-05, "grad_norm": 0.01745838113129139, "loss_scale": 4.0, "retries": 0, "seconds": 35.8, "max_memory_gb": 18.15, "ce": 0.0011864941770909354, "kl": 0.0002568953150330344, "context": 0.0013149418373359367} +{"step": 71, "steps": 250, "lr": 4.434149183384977e-05, "grad_norm": 0.06813450902700424, "loss_scale": 4.0, "retries": 0, "seconds": 36.3, "max_memory_gb": 18.15, "ce": 0.009014494378789095, "kl": 0.01571958819113206, "context": 0.01687428903824184} +{"step": 72, "steps": 250, "lr": 4.413966647630534e-05, "grad_norm": 0.29901936650276184, "loss_scale": 4.0, "retries": 0, "seconds": 36.8, "max_memory_gb": 18.15, "ce": 0.016919521964155138, "kl": 0.398119637160562, "context": 0.21597933681914583} +{"step": 73, "steps": 250, "lr": 4.3934841913948323e-05, "grad_norm": 1.3325587511062622, "loss_scale": 4.0, "retries": 0, "seconds": 37.3, "max_memory_gb": 18.15, "ce": 0.3510820066949236, "kl": 0.09526643438039173, "context": 0.39871521829627454} +{"step": 74, "steps": 250, "lr": 4.372705506523583e-05, "grad_norm": 0.3305263817310333, "loss_scale": 4.0, "retries": 0, "seconds": 37.7, "max_memory_gb": 18.15, "ce": 0.0654347650715863, "kl": 0.11083396457070194, "context": 0.12085174732055748} +{"step": 75, "steps": 250, "lr": 4.351634338256017e-05, "grad_norm": 3.949610948562622, "loss_scale": 4.0, "retries": 0, "seconds": 38.2, "max_memory_gb": 18.15, "ce": 0.4296244882280007, "kl": 0.9210941893979907, "context": 0.8901715772226453} +{"step": 76, "steps": 250, "lr": 4.330274484549825e-05, "grad_norm": 0.9806334972381592, "loss_scale": 4.0, "retries": 0, "seconds": 38.7, "max_memory_gb": 18.15, "ce": 0.017849509204097558, "kl": 1.692798393254634, "context": 0.8642487394390628} +{"step": 77, "steps": 250, "lr": 4.308629795396599e-05, "grad_norm": 0.06665092706680298, "loss_scale": 4.0, "retries": 0, "seconds": 39.2, "max_memory_gb": 18.15, "ce": 0.010502472199732438, "kl": 0.01202208995619003, "context": 0.01651351701002568} +{"step": 78, "steps": 250, "lr": 4.286704172127894e-05, "grad_norm": 0.1444871723651886, "loss_scale": 4.0, "retries": 0, "seconds": 39.7, "max_memory_gb": 18.15, "ce": 0.0019854423735523596, "kl": 0.08611758548795478, "context": 0.045044236598187126} +{"step": 79, "steps": 250, "lr": 4.264501566712033e-05, "grad_norm": 0.2625989615917206, "loss_scale": 4.0, "retries": 0, "seconds": 40.2, "max_memory_gb": 18.15, "ce": 0.013929876256952411, "kl": 0.022563430089576286, "context": 0.02521159130265005} +{"step": 80, "steps": 250, "lr": 4.2420259810417894e-05, "grad_norm": 0.17506003379821777, "loss_scale": 4.0, "retries": 0, "seconds": 40.7, "max_memory_gb": 18.15, "ce": 0.00242383754812181, "kl": 0.08551266308404593, "context": 0.045180167457147036} +{"step": 81, "steps": 250, "lr": 4.219281466213066e-05, "grad_norm": 0.4046684503555298, "loss_scale": 4.0, "retries": 0, "seconds": 41.2, "max_memory_gb": 18.15, "ce": 0.09080989542417228, "kl": 0.0695925914333202, "context": 0.1256061962340027} +{"step": 82, "steps": 250, "lr": 4.1962721217947135e-05, "grad_norm": 0.08161745220422745, "loss_scale": 4.0, "retries": 0, "seconds": 41.7, "max_memory_gb": 18.15, "ce": 0.008477979194140062, "kl": 0.0025808995123952627, "context": 0.009768428979441524} +{"step": 83, "steps": 250, "lr": 4.173002095089599e-05, "grad_norm": 0.5586838722229004, "loss_scale": 4.0, "retries": 0, "seconds": 42.2, "max_memory_gb": 18.15, "ce": 0.1554780245933216, "kl": 0.7350764669463388, "context": 0.5230162503721658} +{"step": 84, "steps": 250, "lr": 4.149475580387084e-05, "grad_norm": 0.3561215400695801, "loss_scale": 4.0, "retries": 0, "seconds": 42.8, "max_memory_gb": 18.15, "ce": 0.017690274498818326, "kl": 0.05537455174453498, "context": 0.04537754976627184} +{"step": 85, "steps": 250, "lr": 4.1256968182070275e-05, "grad_norm": 1.172987699508667, "loss_scale": 4.0, "retries": 0, "seconds": 43.3, "max_memory_gb": 18.15, "ce": 0.03344687441631322, "kl": 0.023527804153250287, "context": 0.045210775564100913} +{"step": 86, "steps": 250, "lr": 4.101670094535453e-05, "grad_norm": 1.8033097982406616, "loss_scale": 4.0, "retries": 0, "seconds": 43.8, "max_memory_gb": 18.15, "ce": 0.18197088374290615, "kl": 0.42709056031890213, "context": 0.3955161648336798} +{"step": 87, "steps": 250, "lr": 4.077399740052027e-05, "grad_norm": 0.8820270299911499, "loss_scale": 4.0, "retries": 0, "seconds": 44.3, "max_memory_gb": 18.15, "ce": 0.007240172766614705, "kl": 0.6601191888330504, "context": 0.33729976369068027} +{"step": 88, "steps": 250, "lr": 4.0528901293494775e-05, "grad_norm": 0.11193723976612091, "loss_scale": 4.0, "retries": 0, "seconds": 44.8, "max_memory_gb": 18.15, "ce": 0.012850095254179905, "kl": 0.02313915708532477, "context": 0.024419673323791358} +{"step": 89, "steps": 250, "lr": 4.0281456801451004e-05, "grad_norm": 0.1412448137998581, "loss_scale": 4.0, "retries": 0, "seconds": 45.3, "max_memory_gb": 18.15, "ce": 0.018527779569467384, "kl": 0.016373045604268555, "context": 0.02671430190639512} +{"step": 90, "steps": 250, "lr": 4.0031708524844884e-05, "grad_norm": 0.09433746337890625, "loss_scale": 4.0, "retries": 0, "seconds": 45.8, "max_memory_gb": 18.15, "ce": 0.009960487484931946, "kl": 0.005210027622524649, "context": 0.012565501208882779} +{"step": 91, "steps": 250, "lr": 3.977970147937635e-05, "grad_norm": 1.252737045288086, "loss_scale": 4.0, "retries": 0, "seconds": 46.3, "max_memory_gb": 18.15, "ce": 0.27399093215353787, "kl": 0.44835442002658965, "context": 0.4981681389035657} +{"step": 92, "steps": 250, "lr": 3.952548108787553e-05, "grad_norm": 2.0004143714904785, "loss_scale": 4.0, "retries": 0, "seconds": 46.8, "max_memory_gb": 18.15, "ce": 0.28440687876354787, "kl": 0.42301125864287314, "context": 0.495912508420588} +{"step": 93, "steps": 250, "lr": 3.9269093172115496e-05, "grad_norm": 1.710572600364685, "loss_scale": 4.0, "retries": 0, "seconds": 47.3, "max_memory_gb": 18.15, "ce": 0.05860068608308211, "kl": 0.42423395872174297, "context": 0.2707176732365042} +{"step": 94, "steps": 250, "lr": 3.901058394455323e-05, "grad_norm": 1.5209276676177979, "loss_scale": 4.0, "retries": 0, "seconds": 47.8, "max_memory_gb": 18.15, "ce": 0.12770467216614634, "kl": 0.061761005548760295, "context": 0.15858517307788134} +{"step": 95, "steps": 250, "lr": 3.875e-05, "grad_norm": 1.6523054838180542, "loss_scale": 4.0, "retries": 0, "seconds": 48.3, "max_memory_gb": 18.15, "ce": 0.5695412301283795, "kl": 0.025873608188703656, "context": 0.5824780543334782} +{"step": 96, "steps": 250, "lr": 3.8487388307223005e-05, "grad_norm": 1.0071903467178345, "loss_scale": 4.0, "retries": 0, "seconds": 48.8, "max_memory_gb": 18.15, "ce": 0.4149770545773208, "kl": 0.007090575378242647, "context": 0.4185223360545933} +{"step": 97, "steps": 250, "lr": 3.822279620047943e-05, "grad_norm": 2.480346441268921, "loss_scale": 4.0, "retries": 0, "seconds": 49.4, "max_memory_gb": 18.15, "ce": 0.021384341380326077, "kl": 0.19468086474807933, "context": 0.118724774569273} +{"step": 98, "steps": 250, "lr": 3.795627137098479e-05, "grad_norm": 0.14298419654369354, "loss_scale": 4.0, "retries": 0, "seconds": 49.9, "max_memory_gb": 18.15, "ce": 0.0036883839711663313, "kl": 0.10199777010711841, "context": 0.05468726947583491} +{"step": 99, "steps": 250, "lr": 3.7687861858316844e-05, "grad_norm": 0.17731617391109467, "loss_scale": 4.0, "retries": 0, "seconds": 50.4, "max_memory_gb": 18.15, "ce": 0.018784405532642268, "kl": 0.010044523733085953, "context": 0.02380666652607033} +{"step": 100, "steps": 250, "lr": 3.741761604175668e-05, "grad_norm": 0.6063717603683472, "loss_scale": 4.0, "retries": 0, "seconds": 50.9, "max_memory_gb": 18.15, "ce": 0.0412164260778809, "kl": 0.5852357146068243, "context": 0.33383429373498075} +{"step": 101, "steps": 250, "lr": 3.714558263156872e-05, "grad_norm": 0.34269091486930847, "loss_scale": 4.0, "retries": 0, "seconds": 51.4, "max_memory_gb": 18.15, "ce": 0.005318293129676022, "kl": 0.0685440749512054, "context": 0.039590332424268126} +{"step": 102, "steps": 250, "lr": 3.6871810660220904e-05, "grad_norm": 0.2769660949707031, "loss_scale": 4.0, "retries": 0, "seconds": 52.0, "max_memory_gb": 18.15, "ce": 0.06967411190271378, "kl": 0.02986148581840098, "context": 0.08460485562682152} +{"step": 103, "steps": 250, "lr": 3.659634947354686e-05, "grad_norm": 0.15261976420879364, "loss_scale": 4.0, "retries": 0, "seconds": 52.5, "max_memory_gb": 18.15, "ce": 0.011053122929297388, "kl": 0.01223748371558031, "context": 0.01717186473251786} +{"step": 104, "steps": 250, "lr": 3.631924872185169e-05, "grad_norm": 0.0587712898850441, "loss_scale": 4.0, "retries": 0, "seconds": 53.0, "max_memory_gb": 18.15, "ce": 0.005153313293703832, "kl": 0.0044940208390471525, "context": 0.007400323826004751} +{"step": 105, "steps": 250, "lr": 3.6040558350962674e-05, "grad_norm": 0.32417377829551697, "loss_scale": 4.0, "retries": 0, "seconds": 53.5, "max_memory_gb": 18.15, "ce": 0.058768728049471974, "kl": 0.36151379137299955, "context": 0.23952562594786286} +{"step": 106, "steps": 250, "lr": 3.576032859322687e-05, "grad_norm": 0.037826742976903915, "loss_scale": 4.0, "retries": 0, "seconds": 54.0, "max_memory_gb": 18.15, "ce": 0.004639766528271139, "kl": 0.00227292776980903, "context": 0.005776230507763103} +{"step": 107, "steps": 250, "lr": 3.5478609958457056e-05, "grad_norm": 2.572845458984375, "loss_scale": 4.0, "retries": 0, "seconds": 54.6, "max_memory_gb": 18.15, "ce": 0.3671512344662915, "kl": 0.09287650836887451, "context": 0.41358947593107587} +{"step": 108, "steps": 250, "lr": 3.519545322482755e-05, "grad_norm": 0.9179390668869019, "loss_scale": 4.0, "retries": 0, "seconds": 55.1, "max_memory_gb": 18.15, "ce": 0.05531371074175695, "kl": 0.014164375390009809, "context": 0.0623958965879865} +{"step": 109, "steps": 250, "lr": 3.491090942972179e-05, "grad_norm": 0.753691554069519, "loss_scale": 4.0, "retries": 0, "seconds": 55.6, "max_memory_gb": 18.15, "ce": 0.061365943402051926, "kl": 0.3247556295245886, "context": 0.22374376002699137} +{"step": 110, "steps": 250, "lr": 3.4625029860533135e-05, "grad_norm": 1.509399652481079, "loss_scale": 4.0, "retries": 0, "seconds": 56.1, "max_memory_gb": 18.15, "ce": 0.10851941769942641, "kl": 0.25708426523488015, "context": 0.23706154758110642} +{"step": 111, "steps": 250, "lr": 3.433786604542056e-05, "grad_norm": 0.026777809485793114, "loss_scale": 4.0, "retries": 0, "seconds": 56.6, "max_memory_gb": 18.15, "ce": 0.002313477758434601, "kl": 0.0018921472774309223, "context": 0.003259551405790262} +{"step": 112, "steps": 250, "lr": 3.404946974402104e-05, "grad_norm": 0.14706629514694214, "loss_scale": 4.0, "retries": 0, "seconds": 57.1, "max_memory_gb": 18.15, "ce": 0.00915152719244361, "kl": 0.001110734709072858, "context": 0.00970689469249919} +{"step": 113, "steps": 250, "lr": 3.375989293812019e-05, "grad_norm": 0.35810816287994385, "loss_scale": 4.0, "retries": 0, "seconds": 57.7, "max_memory_gb": 18.15, "ce": 0.01307546824682504, "kl": 0.3201808414887637, "context": 0.17316588945686817} +{"step": 114, "steps": 250, "lr": 3.3469187822282766e-05, "grad_norm": 1.7111715078353882, "loss_scale": 4.0, "retries": 0, "seconds": 58.2, "max_memory_gb": 18.15, "ce": 0.017358009878080338, "kl": 1.128528239321895, "context": 0.5816221159766428} +{"step": 115, "steps": 250, "lr": 3.317740679444499e-05, "grad_norm": 0.11062951385974884, "loss_scale": 4.0, "retries": 0, "seconds": 58.7, "max_memory_gb": 18.15, "ce": 0.008623579167760909, "kl": 0.044445188366808, "context": 0.030846173263853416} +{"step": 116, "steps": 250, "lr": 3.2884602446470056e-05, "grad_norm": 3.0539543628692627, "loss_scale": 4.0, "retries": 0, "seconds": 59.2, "max_memory_gb": 18.15, "ce": 0.11249925999436527, "kl": 0.4396751314643552, "context": 0.3323368331184611} +{"step": 117, "steps": 250, "lr": 3.259082755466876e-05, "grad_norm": 1.359500527381897, "loss_scale": 4.0, "retries": 0, "seconds": 59.7, "max_memory_gb": 18.16, "ce": 0.04754152748500928, "kl": 0.07621566977468319, "context": 0.08564936608308926} +{"step": 118, "steps": 250, "lr": 3.229613507028686e-05, "grad_norm": 0.5457756519317627, "loss_scale": 4.0, "retries": 0, "seconds": 60.3, "max_memory_gb": 18.16, "ce": 0.18849364257766865, "kl": 0.008232342828705441, "context": 0.19260980852413923} +{"step": 119, "steps": 250, "lr": 3.2000578109961004e-05, "grad_norm": 0.15300253033638, "loss_scale": 4.0, "retries": 0, "seconds": 60.8, "max_memory_gb": 18.16, "ce": 0.014027058728970587, "kl": 0.01912536602731052, "context": 0.02358974233357003} +{"step": 120, "steps": 250, "lr": 3.1704209946144616e-05, "grad_norm": 1.9748860597610474, "loss_scale": 4.0, "retries": 0, "seconds": 61.3, "max_memory_gb": 18.16, "ce": 0.5046571904676966, "kl": 0.1957237555179745, "context": 0.6025190713698976} +{"step": 121, "steps": 250, "lr": 3.140708399750594e-05, "grad_norm": 0.3285699784755707, "loss_scale": 4.0, "retries": 0, "seconds": 61.8, "max_memory_gb": 18.16, "ce": 0.053181276307441294, "kl": 0.03528011208982207, "context": 0.07082133329095086} +{"step": 122, "steps": 250, "lr": 3.1109253819299606e-05, "grad_norm": 0.47480788826942444, "loss_scale": 4.0, "retries": 0, "seconds": 62.3, "max_memory_gb": 18.16, "ce": 0.08675171832146589, "kl": 0.01777930565731367, "context": 0.09564137272536755} +{"step": 123, "steps": 250, "lr": 3.081077309371361e-05, "grad_norm": 0.1821429431438446, "loss_scale": 4.0, "retries": 0, "seconds": 62.9, "max_memory_gb": 18.16, "ce": 0.03881792118772864, "kl": 0.02050077880267054, "context": 0.04906830983236432} +{"step": 124, "steps": 250, "lr": 3.051169562019339e-05, "grad_norm": 1.9378947019577026, "loss_scale": 4.0, "retries": 0, "seconds": 63.4, "max_memory_gb": 18.16, "ce": 0.6690018452645745, "kl": 0.5693810492412013, "context": 0.9536923978012055} +{"step": 125, "steps": 250, "lr": 3.021207530574477e-05, "grad_norm": 0.35464945435523987, "loss_scale": 4.0, "retries": 0, "seconds": 63.9, "max_memory_gb": 18.16, "ce": 0.11567087130970322, "kl": 0.08054358960362151, "context": 0.15594266608241014} +{"step": 126, "steps": 250, "lr": 2.991196615521758e-05, "grad_norm": 1.8551528453826904, "loss_scale": 4.0, "retries": 0, "seconds": 64.4, "max_memory_gb": 18.16, "ce": 0.5670040715485811, "kl": 0.1377873681485653, "context": 0.6358977425843477} +{"step": 127, "steps": 250, "lr": 2.9611422261571536e-05, "grad_norm": 0.11988420784473419, "loss_scale": 4.0, "retries": 0, "seconds": 64.9, "max_memory_gb": 18.16, "ce": 0.016605394921498373, "kl": 0.001317637761530932, "context": 0.017264214053284377} +{"step": 128, "steps": 250, "lr": 2.931049779612634e-05, "grad_norm": 2.193629026412964, "loss_scale": 4.0, "retries": 0, "seconds": 65.4, "max_memory_gb": 18.16, "ce": 0.19219108087418135, "kl": 1.1111277856398374, "context": 0.7477549706818536} +{"step": 129, "steps": 250, "lr": 2.9009246998797557e-05, "grad_norm": 0.2647772431373596, "loss_scale": 4.0, "retries": 0, "seconds": 65.9, "max_memory_gb": 18.16, "ce": 0.0731520865811035, "kl": 0.031096065147721674, "context": 0.08870011731050909} +{"step": 130, "steps": 250, "lr": 2.8707724168320238e-05, "grad_norm": 0.192884624004364, "loss_scale": 4.0, "retries": 0, "seconds": 66.4, "max_memory_gb": 18.16, "ce": 0.044592786696739495, "kl": 0.2447200210444862, "context": 0.16695280553540215} +{"step": 131, "steps": 250, "lr": 2.840598365246184e-05, "grad_norm": 2.526207208633423, "loss_scale": 4.0, "retries": 0, "seconds": 67.0, "max_memory_gb": 18.16, "ce": 0.30523367496789433, "kl": 0.3977161234943196, "context": 0.504091736394912} +{"step": 132, "steps": 250, "lr": 2.8104079838226383e-05, "grad_norm": 0.34303018450737, "loss_scale": 4.0, "retries": 0, "seconds": 67.5, "max_memory_gb": 18.16, "ce": 0.16779401829990093, "kl": 0.03080620406581147, "context": 0.18319711716321763} +{"step": 133, "steps": 250, "lr": 2.7802067142051446e-05, "grad_norm": 0.6927597522735596, "loss_scale": 4.0, "retries": 0, "seconds": 68.0, "max_memory_gb": 18.16, "ce": 0.3564182014670223, "kl": 0.18897814163938165, "context": 0.4509072727523744} +{"step": 134, "steps": 250, "lr": 2.7500000000000004e-05, "grad_norm": 0.531990110874176, "loss_scale": 4.0, "retries": 0, "seconds": 68.6, "max_memory_gb": 18.16, "ce": 0.18832053714504582, "kl": 0.004820739694395115, "context": 0.19073090186429908} +{"step": 135, "steps": 250, "lr": 2.7197932857948566e-05, "grad_norm": 0.11950720846652985, "loss_scale": 4.0, "retries": 0, "seconds": 69.1, "max_memory_gb": 18.16, "ce": 0.023128781176637858, "kl": 0.09778304179781117, "context": 0.07202030124608427} +{"step": 136, "steps": 250, "lr": 2.689592016177363e-05, "grad_norm": 1.6024932861328125, "loss_scale": 4.0, "retries": 0, "seconds": 69.6, "max_memory_gb": 18.16, "ce": 0.07936490944121033, "kl": 0.09866528268321417, "context": 0.12869755108840764} +{"step": 137, "steps": 250, "lr": 2.659401634753817e-05, "grad_norm": 0.36228346824645996, "loss_scale": 4.0, "retries": 0, "seconds": 70.1, "max_memory_gb": 18.16, "ce": 0.011205258662812412, "kl": 0.031123968459723983, "context": 0.026767242423375137} +{"step": 138, "steps": 250, "lr": 2.6292275831679764e-05, "grad_norm": 1.187363862991333, "loss_scale": 4.0, "retries": 0, "seconds": 70.6, "max_memory_gb": 18.16, "ce": 0.07308550993911922, "kl": 0.1314606498926878, "context": 0.1388158341869712} +{"step": 139, "steps": 250, "lr": 2.5990753001202445e-05, "grad_norm": 1.2456223964691162, "loss_scale": 4.0, "retries": 0, "seconds": 71.1, "max_memory_gb": 18.16, "ce": 0.005093655534437858, "kl": 0.19726368515375725, "context": 0.10372549413295928} +{"step": 140, "steps": 250, "lr": 2.5689502203873672e-05, "grad_norm": 0.05562524497509003, "loss_scale": 4.0, "retries": 0, "seconds": 71.7, "max_memory_gb": 18.16, "ce": 0.004926590685499832, "kl": 0.0020456954516703263, "context": 0.005949438374955207} +{"step": 141, "steps": 250, "lr": 2.538857773842847e-05, "grad_norm": 0.05634591728448868, "loss_scale": 4.0, "retries": 0, "seconds": 72.2, "max_memory_gb": 18.16, "ce": 0.007282567326910794, "kl": 0.0065080213971668854, "context": 0.010536578192841262} +{"step": 142, "steps": 250, "lr": 2.5088033844782426e-05, "grad_norm": 0.5550774335861206, "loss_scale": 4.0, "retries": 0, "seconds": 72.7, "max_memory_gb": 18.16, "ce": 0.0407154590357095, "kl": 0.01619102736003697, "context": 0.04881097318138927} +{"step": 143, "steps": 250, "lr": 2.4787924694255237e-05, "grad_norm": 2.6214711666107178, "loss_scale": 4.0, "retries": 0, "seconds": 73.2, "max_memory_gb": 18.16, "ce": 0.4350802690969431, "kl": 0.0699265446382924, "context": 0.4700435439881403} +{"step": 144, "steps": 250, "lr": 2.4488304379806627e-05, "grad_norm": 0.5755144953727722, "loss_scale": 4.0, "retries": 0, "seconds": 73.8, "max_memory_gb": 18.16, "ce": 0.06141856103204191, "kl": 0.19674048433080316, "context": 0.15978880785405636} +{"step": 145, "steps": 250, "lr": 2.4189226906286388e-05, "grad_norm": 0.04520151764154434, "loss_scale": 4.0, "retries": 0, "seconds": 74.3, "max_memory_gb": 18.16, "ce": 0.00209605970303528, "kl": 0.008727176412776316, "context": 0.0064596478041494265} +{"step": 146, "steps": 250, "lr": 2.3890746180700395e-05, "grad_norm": 0.10646949708461761, "loss_scale": 4.0, "retries": 0, "seconds": 74.8, "max_memory_gb": 18.16, "ce": 0.00923600874375552, "kl": 0.0217980863271805, "context": 0.020135051454417408} +{"step": 147, "steps": 250, "lr": 2.359291600249407e-05, "grad_norm": 1.0268045663833618, "loss_scale": 4.0, "retries": 0, "seconds": 75.3, "max_memory_gb": 18.16, "ce": 0.4889645984439994, "kl": 0.5199243706592824, "context": 0.7489267792843748} +{"step": 148, "steps": 250, "lr": 2.3295790053855392e-05, "grad_norm": 0.10247103869915009, "loss_scale": 4.0, "retries": 0, "seconds": 75.8, "max_memory_gb": 18.16, "ce": 0.006134760156783159, "kl": 0.011507052535307594, "context": 0.011888286542671267} +{"step": 149, "steps": 250, "lr": 2.299942189003901e-05, "grad_norm": 0.15284842252731323, "loss_scale": 4.0, "retries": 0, "seconds": 76.4, "max_memory_gb": 18.16, "ce": 0.011864277592394501, "kl": 0.004658267709601205, "context": 0.014193410868756473} +{"step": 150, "steps": 250, "lr": 2.2703864929713144e-05, "grad_norm": 0.2614138424396515, "loss_scale": 4.0, "retries": 0, "seconds": 76.9, "max_memory_gb": 18.16, "ce": 0.06194893328938633, "kl": 0.024311499029863626, "context": 0.07410468289162964} +{"step": 151, "steps": 250, "lr": 2.2409172445331257e-05, "grad_norm": 0.2876218557357788, "loss_scale": 4.0, "retries": 0, "seconds": 77.4, "max_memory_gb": 18.16, "ce": 0.02105469828529749, "kl": 0.028783877311070682, "context": 0.035446635069092736} +{"step": 152, "steps": 250, "lr": 2.211539755352996e-05, "grad_norm": 0.48000994324684143, "loss_scale": 4.0, "retries": 0, "seconds": 77.9, "max_memory_gb": 18.16, "ce": 0.007566517670056783, "kl": 0.5246122007147278, "context": 0.2698726211092435} +{"step": 153, "steps": 250, "lr": 2.1822593205555016e-05, "grad_norm": 0.3766905963420868, "loss_scale": 4.0, "retries": 0, "seconds": 78.6, "max_memory_gb": 18.16, "ce": 0.01396619663137244, "kl": 0.3703451588480675, "context": 0.19913878164516063} +{"step": 154, "steps": 250, "lr": 2.1530812177717236e-05, "grad_norm": 0.6332608461380005, "loss_scale": 4.0, "retries": 0, "seconds": 79.2, "max_memory_gb": 18.16, "ce": 0.06149989413097501, "kl": 0.08032134745735675, "context": 0.10166057193418965} +{"step": 155, "steps": 250, "lr": 2.1240107061879815e-05, "grad_norm": 0.37203967571258545, "loss_scale": 4.0, "retries": 0, "seconds": 79.9, "max_memory_gb": 18.16, "ce": 0.1837999059644062, "kl": 0.009653122410782089, "context": 0.1886264675667917} +{"step": 156, "steps": 250, "lr": 2.095053025597896e-05, "grad_norm": 0.7445963025093079, "loss_scale": 4.0, "retries": 0, "seconds": 80.5, "max_memory_gb": 18.16, "ce": 0.012341526682575932, "kl": 0.03888283690639582, "context": 0.03178294519602787} +{"step": 157, "steps": 250, "lr": 2.066213395457945e-05, "grad_norm": 0.3904459476470947, "loss_scale": 4.0, "retries": 0, "seconds": 81.0, "max_memory_gb": 18.16, "ce": 0.02582500409334898, "kl": 0.028552964940899983, "context": 0.04010148660745472} +{"step": 158, "steps": 250, "lr": 2.0374970139466877e-05, "grad_norm": 0.43466249108314514, "loss_scale": 4.0, "retries": 0, "seconds": 81.5, "max_memory_gb": 18.16, "ce": 0.19226834762957878, "kl": 0.06561337105813436, "context": 0.22507502953521907} +{"step": 159, "steps": 250, "lr": 2.0089090570278213e-05, "grad_norm": 0.4691566526889801, "loss_scale": 4.0, "retries": 0, "seconds": 82.2, "max_memory_gb": 18.16, "ce": 0.006627468857914209, "kl": 0.05176516744541004, "context": 0.03251005246420391} +{"step": 160, "steps": 250, "lr": 1.9804546775172457e-05, "grad_norm": 0.6593745350837708, "loss_scale": 4.0, "retries": 0, "seconds": 82.9, "max_memory_gb": 18.16, "ce": 0.03675386687973514, "kl": 0.33973436518135713, "context": 0.2066210462944582} +{"step": 161, "steps": 250, "lr": 1.9521390041542946e-05, "grad_norm": 0.45579877495765686, "loss_scale": 4.0, "retries": 0, "seconds": 83.5, "max_memory_gb": 18.16, "ce": 0.07794625172391534, "kl": 0.09538840128516313, "context": 0.1256404486775864} +{"step": 162, "steps": 250, "lr": 1.923967140677313e-05, "grad_norm": 1.016558289527893, "loss_scale": 4.0, "retries": 0, "seconds": 84.1, "max_memory_gb": 18.16, "ce": 0.06463748517853674, "kl": 0.11178696555725764, "context": 0.1205309679498896} +{"step": 163, "steps": 250, "lr": 1.8959441649037335e-05, "grad_norm": 2.2336862087249756, "loss_scale": 4.0, "retries": 0, "seconds": 84.8, "max_memory_gb": 18.16, "ce": 0.2347833207459189, "kl": 0.754215543391183, "context": 0.6118910773075186} +{"step": 164, "steps": 250, "lr": 1.868075127814832e-05, "grad_norm": 3.158400297164917, "loss_scale": 4.0, "retries": 0, "seconds": 85.4, "max_memory_gb": 18.16, "ce": 0.0803072792768944, "kl": 0.08795851663307985, "context": 0.1242865391104715} +{"step": 165, "steps": 250, "lr": 1.8403650526453143e-05, "grad_norm": 0.09860288351774216, "loss_scale": 4.0, "retries": 0, "seconds": 85.9, "max_memory_gb": 18.16, "ce": 0.006746628729160875, "kl": 0.026913940677331993, "context": 0.020203599124215543} +{"step": 166, "steps": 250, "lr": 1.812818933977912e-05, "grad_norm": 1.7399766445159912, "loss_scale": 4.0, "retries": 0, "seconds": 86.5, "max_memory_gb": 18.16, "ce": 0.4488376588560641, "kl": 0.1265149204991758, "context": 0.5120951246935874} +{"step": 167, "steps": 250, "lr": 1.785441736843128e-05, "grad_norm": 0.165523961186409, "loss_scale": 4.0, "retries": 0, "seconds": 87.0, "max_memory_gb": 18.16, "ce": 0.022568637039512396, "kl": 0.0033477775141363963, "context": 0.02424252510536462} +{"step": 168, "steps": 250, "lr": 1.7582383958243314e-05, "grad_norm": 0.12527862191200256, "loss_scale": 4.0, "retries": 0, "seconds": 87.5, "max_memory_gb": 18.16, "ce": 0.030177966327755712, "kl": 0.0018693027668632567, "context": 0.031112618424231187} +{"step": 169, "steps": 250, "lr": 1.7312138141683164e-05, "grad_norm": 1.136484980583191, "loss_scale": 4.0, "retries": 0, "seconds": 88.1, "max_memory_gb": 18.16, "ce": 0.08755737525643781, "kl": 1.3892329186201096, "context": 0.7821738347411156} +{"step": 170, "steps": 250, "lr": 1.704372862901521e-05, "grad_norm": 0.06107250228524208, "loss_scale": 4.0, "retries": 0, "seconds": 88.6, "max_memory_gb": 18.16, "ce": 0.0030922925361664966, "kl": 0.0008444003215117846, "context": 0.0035144927678629756} +{"step": 171, "steps": 250, "lr": 1.6777203799520573e-05, "grad_norm": 0.6445450186729431, "loss_scale": 4.0, "retries": 0, "seconds": 89.1, "max_memory_gb": 18.16, "ce": 0.024922322660131613, "kl": 0.24403326166793704, "context": 0.14693895558593795} +{"step": 172, "steps": 250, "lr": 1.6512611692777008e-05, "grad_norm": 0.120402030646801, "loss_scale": 4.0, "retries": 0, "seconds": 89.7, "max_memory_gb": 18.16, "ce": 0.015395148424431682, "kl": 0.015193125087535009, "context": 0.022991711157374084} +{"step": 173, "steps": 250, "lr": 1.6250000000000005e-05, "grad_norm": 0.1905524879693985, "loss_scale": 4.0, "retries": 0, "seconds": 90.3, "max_memory_gb": 18.16, "ce": 0.053643432387616485, "kl": 0.025733816913998453, "context": 0.06651034147944301} +{"step": 174, "steps": 250, "lr": 1.598941605544678e-05, "grad_norm": 0.19833362102508545, "loss_scale": 4.0, "retries": 0, "seconds": 90.9, "max_memory_gb": 18.16, "ce": 0.03746494752704166, "kl": 0.03559879600175009, "context": 0.05526434366765898} +{"step": 175, "steps": 250, "lr": 1.5730906827884506e-05, "grad_norm": 2.7611629962921143, "loss_scale": 4.0, "retries": 0, "seconds": 91.5, "max_memory_gb": 18.16, "ce": 0.15539117623120546, "kl": 0.09876603540033102, "context": 0.20477419532835484} +{"step": 176, "steps": 250, "lr": 1.5474518912124476e-05, "grad_norm": 0.07303284108638763, "loss_scale": 4.0, "retries": 0, "seconds": 92.2, "max_memory_gb": 18.16, "ce": 0.01110946579137817, "kl": 0.0028300197154749185, "context": 0.012524475750979036} +{"step": 177, "steps": 250, "lr": 1.5220298520623646e-05, "grad_norm": 0.6768840551376343, "loss_scale": 4.0, "retries": 0, "seconds": 92.8, "max_memory_gb": 18.16, "ce": 0.023624871086212806, "kl": 0.6844284205035365, "context": 0.3658390766795492} +{"step": 178, "steps": 250, "lr": 1.4968291475155121e-05, "grad_norm": 0.9679599404335022, "loss_scale": 4.0, "retries": 0, "seconds": 93.4, "max_memory_gb": 18.16, "ce": 0.05498152179643512, "kl": 0.10347235793597065, "context": 0.10671770083718002} +{"step": 179, "steps": 250, "lr": 1.4718543198548998e-05, "grad_norm": 0.8080229759216309, "loss_scale": 4.0, "retries": 0, "seconds": 94.1, "max_memory_gb": 18.16, "ce": 0.38240435368788894, "kl": 0.020387090349686332, "context": 0.39259790029609576} +{"step": 180, "steps": 250, "lr": 1.4471098706505227e-05, "grad_norm": 0.245510071516037, "loss_scale": 4.0, "retries": 0, "seconds": 94.7, "max_memory_gb": 18.16, "ce": 0.044531080158776604, "kl": 0.028149226011009887, "context": 0.05860569467768073} +{"step": 181, "steps": 250, "lr": 1.4226002599479743e-05, "grad_norm": 0.1663549244403839, "loss_scale": 4.0, "retries": 0, "seconds": 95.3, "max_memory_gb": 18.16, "ce": 0.009159324341453612, "kl": 0.09257169294869527, "context": 0.05544516915688291} +{"step": 182, "steps": 250, "lr": 1.3983299054645471e-05, "grad_norm": 0.8649271130561829, "loss_scale": 4.0, "retries": 0, "seconds": 95.8, "max_memory_gb": 18.16, "ce": 0.0264545485842973, "kl": 0.3064566419343464, "context": 0.17968287094845437} +{"step": 183, "steps": 250, "lr": 1.3743031817929724e-05, "grad_norm": 0.03246263787150383, "loss_scale": 4.0, "retries": 0, "seconds": 96.5, "max_memory_gb": 18.16, "ce": 0.0034264913338120095, "kl": 0.0008851219317875803, "context": 0.0038690521687385626} +{"step": 184, "steps": 250, "lr": 1.3505244196129164e-05, "grad_norm": 0.05769718438386917, "loss_scale": 4.0, "retries": 0, "seconds": 97.1, "max_memory_gb": 18.16, "ce": 0.003669514393550344, "kl": 0.005448766547488049, "context": 0.006393897667294368} +{"step": 185, "steps": 250, "lr": 1.3269979049104014e-05, "grad_norm": 0.19875091314315796, "loss_scale": 4.0, "retries": 0, "seconds": 97.7, "max_memory_gb": 18.16, "ce": 0.012117310456233099, "kl": 0.08580444120525499, "context": 0.055019530234858394} +{"step": 186, "steps": 250, "lr": 1.3037278782052863e-05, "grad_norm": 2.2527573108673096, "loss_scale": 4.0, "retries": 0, "seconds": 98.4, "max_memory_gb": 18.16, "ce": 0.09539636109548155, "kl": 0.23689167276097578, "context": 0.21384219663013937} +{"step": 187, "steps": 250, "lr": 1.2807185337869343e-05, "grad_norm": 0.31531596183776855, "loss_scale": 4.0, "retries": 0, "seconds": 99.0, "max_memory_gb": 18.16, "ce": 0.0819029373669764, "kl": 0.4850091665284708, "context": 0.3244075206457637} +{"step": 188, "steps": 250, "lr": 1.2579740189582115e-05, "grad_norm": 1.1778465509414673, "loss_scale": 4.0, "retries": 0, "seconds": 99.7, "max_memory_gb": 18.16, "ce": 0.2268725860158156, "kl": 0.013335193555121805, "context": 0.2335401832642674} +{"step": 189, "steps": 250, "lr": 1.2354984332879668e-05, "grad_norm": 1.8151841163635254, "loss_scale": 4.0, "retries": 0, "seconds": 100.4, "max_memory_gb": 18.16, "ce": 0.3449337752535939, "kl": 0.38448736630380154, "context": 0.5371774567756802} +{"step": 190, "steps": 250, "lr": 1.2132958278721064e-05, "grad_norm": 0.09387233853340149, "loss_scale": 4.0, "retries": 0, "seconds": 101.0, "max_memory_gb": 18.16, "ce": 0.005078710353700444, "kl": 0.03916587773710489, "context": 0.02466164992074482} +{"step": 191, "steps": 250, "lr": 1.1913702046034016e-05, "grad_norm": 3.3795812129974365, "loss_scale": 4.0, "retries": 0, "seconds": 101.6, "max_memory_gb": 18.16, "ce": 0.3354569557122886, "kl": 0.2900545084357873, "context": 0.48048420215491205} +{"step": 192, "steps": 250, "lr": 1.1697255154501756e-05, "grad_norm": 0.8634588122367859, "loss_scale": 4.0, "retries": 0, "seconds": 102.3, "max_memory_gb": 18.16, "ce": 0.450477282050997, "kl": 0.05583619609387824, "context": 0.4783953856676817} +{"step": 193, "steps": 250, "lr": 1.1483656617439842e-05, "grad_norm": 0.06392166763544083, "loss_scale": 4.0, "retries": 0, "seconds": 102.9, "max_memory_gb": 18.16, "ce": 0.005502071631781291, "kl": 0.0054622546563223295, "context": 0.00823319896881003} +{"step": 194, "steps": 250, "lr": 1.1272944934764177e-05, "grad_norm": 0.625909686088562, "loss_scale": 4.0, "retries": 0, "seconds": 103.6, "max_memory_gb": 18.16, "ce": 0.04564471935736947, "kl": 0.04325247365341056, "context": 0.06727095623500645} +{"step": 195, "steps": 250, "lr": 1.1065158086051684e-05, "grad_norm": 0.10466702282428741, "loss_scale": 4.0, "retries": 0, "seconds": 104.3, "max_memory_gb": 18.16, "ce": 0.014867123682051897, "kl": 0.004983745981007814, "context": 0.01735899693449028} +{"step": 196, "steps": 250, "lr": 1.086033352369467e-05, "grad_norm": 0.5522909760475159, "loss_scale": 4.0, "retries": 0, "seconds": 104.9, "max_memory_gb": 18.16, "ce": 0.001393627953802934, "kl": 0.44018437605791405, "context": 0.22148580998327816} +{"step": 197, "steps": 250, "lr": 1.0658508166150224e-05, "grad_norm": 1.0854878425598145, "loss_scale": 4.0, "retries": 0, "seconds": 105.6, "max_memory_gb": 18.16, "ce": 0.01059409385197796, "kl": 0.7083078075502272, "context": 0.3647480113140773} +{"step": 198, "steps": 250, "lr": 1.0459718391285954e-05, "grad_norm": 0.330223947763443, "loss_scale": 4.0, "retries": 0, "seconds": 106.2, "max_memory_gb": 18.16, "ce": 0.059626264483085833, "kl": 0.010963610373437405, "context": 0.06510807009181008} +{"step": 199, "steps": 250, "lr": 1.0264000029822999e-05, "grad_norm": 0.5528402328491211, "loss_scale": 4.0, "retries": 0, "seconds": 106.9, "max_memory_gb": 18.16, "ce": 0.1474127100664191, "kl": 0.16077489008603152, "context": 0.22780015133321285} +{"step": 200, "steps": 250, "lr": 1.0071388358877775e-05, "grad_norm": 0.7919898629188538, "loss_scale": 4.0, "retries": 0, "seconds": 107.5, "max_memory_gb": 18.16, "ce": 0.28080491413129494, "kl": 0.040287258474563714, "context": 0.3009485367801972} +{"step": 201, "steps": 250, "lr": 9.881918095603501e-06, "grad_norm": 0.7278779745101929, "loss_scale": 4.0, "retries": 0, "seconds": 108.2, "max_memory_gb": 18.16, "ce": 0.08082071965327486, "kl": 0.11276446154079167, "context": 0.13720295784878545} +{"step": 202, "steps": 250, "lr": 9.695623390932612e-06, "grad_norm": 0.044589608907699585, "loss_scale": 4.0, "retries": 0, "seconds": 108.9, "max_memory_gb": 18.16, "ce": 0.002232952656413545, "kl": 0.003336050789641831, "context": 0.00390097815397894} +{"step": 203, "steps": 250, "lr": 9.51253782342123e-06, "grad_norm": 0.8003522157669067, "loss_scale": 4.0, "retries": 0, "seconds": 109.5, "max_memory_gb": 18.16, "ce": 0.022660467591776978, "kl": 0.03741606487892568, "context": 0.0413684991071932} +{"step": 204, "steps": 250, "lr": 9.332694393196846e-06, "grad_norm": 0.191671222448349, "loss_scale": 4.0, "retries": 0, "seconds": 110.2, "max_memory_gb": 18.16, "ce": 0.03895501440274529, "kl": 0.01185367233119905, "context": 0.04488185024820268} +{"step": 205, "steps": 250, "lr": 9.156125516010256e-06, "grad_norm": 0.3934735953807831, "loss_scale": 4.0, "retries": 0, "seconds": 110.8, "max_memory_gb": 18.16, "ce": 0.004632411524653435, "kl": 0.5347243833093671, "context": 0.2719945965218358} +{"step": 206, "steps": 250, "lr": 8.982863017392733e-06, "grad_norm": 0.08756950497627258, "loss_scale": 4.0, "retries": 0, "seconds": 111.4, "max_memory_gb": 18.16, "ce": 0.009319663367932662, "kl": 0.0038752531345380703, "context": 0.011257289992499864} +{"step": 207, "steps": 250, "lr": 8.81293812691974e-06, "grad_norm": 0.19778138399124146, "loss_scale": 4.0, "retries": 0, "seconds": 111.9, "max_memory_gb": 18.16, "ce": 0.06877758330665529, "kl": 0.04651077228481881, "context": 0.09203297318890691} +{"step": 208, "steps": 250, "lr": 8.646381472581927e-06, "grad_norm": 0.04080832377076149, "loss_scale": 4.0, "retries": 0, "seconds": 112.4, "max_memory_gb": 18.16, "ce": 0.00503996922634542, "kl": 0.0018084675866703037, "context": 0.005944203105173074} +{"step": 209, "steps": 250, "lr": 8.483223075264622e-06, "grad_norm": 0.8914674520492554, "loss_scale": 4.0, "retries": 0, "seconds": 113.0, "max_memory_gb": 18.16, "ce": 0.18576994113391265, "kl": 0.10597444418090163, "context": 0.23875715950271115} +{"step": 210, "steps": 250, "lr": 8.323492343336747e-06, "grad_norm": 0.9354951977729797, "loss_scale": 4.0, "retries": 0, "seconds": 113.5, "max_memory_gb": 18.16, "ce": 0.5194884926822851, "kl": 0.15485736144182738, "context": 0.596917161354213} +{"step": 211, "steps": 250, "lr": 8.167218067350122e-06, "grad_norm": 0.36138802766799927, "loss_scale": 4.0, "retries": 0, "seconds": 114.0, "max_memory_gb": 18.16, "ce": 0.14771630411269143, "kl": 0.10074471388361417, "context": 0.19808866071980447} +{"step": 212, "steps": 250, "lr": 8.01442841485013e-06, "grad_norm": 2.426981210708618, "loss_scale": 4.0, "retries": 0, "seconds": 114.5, "max_memory_gb": 18.16, "ce": 0.07346208690432832, "kl": 1.0592524367384613, "context": 0.6030883136554621} +{"step": 213, "steps": 250, "lr": 7.865150925298678e-06, "grad_norm": 0.5065577626228333, "loss_scale": 4.0, "retries": 0, "seconds": 115.0, "max_memory_gb": 18.16, "ce": 0.2241905099072028, "kl": 0.04365069052801118, "context": 0.24601586040807888} +{"step": 214, "steps": 250, "lr": 7.719412505110375e-06, "grad_norm": 0.7668130993843079, "loss_scale": 4.0, "retries": 0, "seconds": 115.6, "max_memory_gb": 18.16, "ce": 0.42778084040037356, "kl": 0.1731422187294811, "context": 0.5143519558187108} +{"step": 215, "steps": 250, "lr": 7.57723942280278e-06, "grad_norm": 1.6302787065505981, "loss_scale": 4.0, "retries": 0, "seconds": 116.1, "max_memory_gb": 18.16, "ce": 0.6303804616618436, "kl": 0.022339635801955637, "context": 0.64155027490051} +{"step": 216, "steps": 250, "lr": 7.438657304261687e-06, "grad_norm": 0.19681987166404724, "loss_scale": 4.0, "retries": 0, "seconds": 116.6, "max_memory_gb": 18.16, "ce": 0.01218225629418157, "kl": 0.1055387411997799, "context": 0.06495162527426146} +{"step": 217, "steps": 250, "lr": 7.30369112812219e-06, "grad_norm": 0.03386642783880234, "loss_scale": 4.0, "retries": 0, "seconds": 117.1, "max_memory_gb": 18.16, "ce": 0.0020038165530422702, "kl": 0.0033573467644600896, "context": 0.0036824898706981912} +{"step": 218, "steps": 250, "lr": 7.172365221266403e-06, "grad_norm": 0.7575856447219849, "loss_scale": 4.0, "retries": 0, "seconds": 117.6, "max_memory_gb": 18.16, "ce": 0.16211352008394897, "kl": 0.28187140449881554, "context": 0.3030492179095745} +{"step": 219, "steps": 250, "lr": 7.0447032544387154e-06, "grad_norm": 0.9681352972984314, "loss_scale": 4.0, "retries": 0, "seconds": 118.2, "max_memory_gb": 18.16, "ce": 0.009112019601161592, "kl": 0.17450792000454385, "context": 0.09636598275392316} +{"step": 220, "steps": 250, "lr": 6.920728237979274e-06, "grad_norm": 0.19180582463741302, "loss_scale": 4.0, "retries": 0, "seconds": 118.7, "max_memory_gb": 18.16, "ce": 0.060637763046543114, "kl": 0.007231757848785492, "context": 0.06425364503229503} +{"step": 221, "steps": 250, "lr": 6.800462517676456e-06, "grad_norm": 0.3820079565048218, "loss_scale": 4.0, "retries": 0, "seconds": 119.2, "max_memory_gb": 18.16, "ce": 0.008356048463610932, "kl": 0.518702984401898, "context": 0.2677075387327932} +{"step": 222, "steps": 250, "lr": 6.683927770739222e-06, "grad_norm": 0.24675315618515015, "loss_scale": 4.0, "retries": 0, "seconds": 119.7, "max_memory_gb": 18.16, "ce": 0.03466674132505432, "kl": 0.07627193564076151, "context": 0.07280270727642346} +{"step": 223, "steps": 250, "lr": 6.571145001889892e-06, "grad_norm": 1.3674052953720093, "loss_scale": 4.0, "retries": 0, "seconds": 120.2, "max_memory_gb": 18.16, "ce": 0.08323781201761449, "kl": 0.006250570939300815, "context": 0.08636309692519717} +{"step": 224, "steps": 250, "lr": 6.46213453957817e-06, "grad_norm": 0.7938270568847656, "loss_scale": 4.0, "retries": 0, "seconds": 120.7, "max_memory_gb": 18.16, "ce": 0.04413888382259756, "kl": 0.20205712973256595, "context": 0.14516744890715927} +{"step": 225, "steps": 250, "lr": 6.356916032317063e-06, "grad_norm": 0.2484951913356781, "loss_scale": 4.0, "retries": 0, "seconds": 121.3, "max_memory_gb": 18.16, "ce": 0.0065091080468846485, "kl": 0.2034016040415736, "context": 0.10820990955107845} +{"step": 226, "steps": 250, "lr": 6.255508445141349e-06, "grad_norm": 0.47446727752685547, "loss_scale": 4.0, "retries": 0, "seconds": 121.7, "max_memory_gb": 18.16, "ce": 0.023856698302552104, "kl": 0.028114961882238276, "context": 0.03791417842148803} +{"step": 227, "steps": 250, "lr": 6.157930056189228e-06, "grad_norm": 0.20403781533241272, "loss_scale": 4.0, "retries": 0, "seconds": 122.2, "max_memory_gb": 18.16, "ce": 0.016959247921477072, "kl": 0.007786294838297181, "context": 0.020852395296969917} +{"step": 228, "steps": 250, "lr": 6.064198453407796e-06, "grad_norm": 0.4342653155326843, "loss_scale": 4.0, "retries": 0, "seconds": 122.7, "max_memory_gb": 18.16, "ce": 0.003987133779446594, "kl": 0.29992785667127464, "context": 0.1539510638685897} +{"step": 229, "steps": 250, "lr": 5.974330531382935e-06, "grad_norm": 1.1929080486297607, "loss_scale": 4.0, "retries": 0, "seconds": 123.2, "max_memory_gb": 18.16, "ce": 0.07510731220827438, "kl": 0.043493661164575315, "context": 0.09685414208797738} +{"step": 230, "steps": 250, "lr": 5.888342488294125e-06, "grad_norm": 1.6187242269515991, "loss_scale": 4.0, "retries": 0, "seconds": 123.8, "max_memory_gb": 18.16, "ce": 0.5274349360406632, "kl": 0.3586580588016659, "context": 0.7067639703163877} +{"step": 231, "steps": 250, "lr": 5.806249822994849e-06, "grad_norm": 0.6230618953704834, "loss_scale": 4.0, "retries": 0, "seconds": 124.3, "max_memory_gb": 18.16, "ce": 0.04540930224175099, "kl": 0.15963640427798964, "context": 0.12522750344942324} +{"step": 232, "steps": 250, "lr": 5.728067332218994e-06, "grad_norm": 0.14963126182556152, "loss_scale": 4.0, "retries": 0, "seconds": 124.9, "max_memory_gb": 18.16, "ce": 0.027202629076782614, "kl": 0.011775925871916115, "context": 0.033090591023210436} +{"step": 233, "steps": 250, "lr": 5.65380910791383e-06, "grad_norm": 0.25389304757118225, "loss_scale": 4.0, "retries": 0, "seconds": 125.4, "max_memory_gb": 18.16, "ce": 0.010113343363627791, "kl": 0.08441372902598232, "context": 0.052320207469165325} +{"step": 234, "steps": 250, "lr": 5.583488534700022e-06, "grad_norm": 0.29434823989868164, "loss_scale": 4.0, "retries": 0, "seconds": 125.9, "max_memory_gb": 18.16, "ce": 0.011670611798763275, "kl": 0.03770634427564801, "context": 0.030523783920216374} +{"step": 235, "steps": 250, "lr": 5.517118287459118e-06, "grad_norm": 1.343923807144165, "loss_scale": 4.0, "retries": 0, "seconds": 126.4, "max_memory_gb": 18.16, "ce": 0.3803454174340004, "kl": 0.016355427927919663, "context": 0.388523144996725} +{"step": 236, "steps": 250, "lr": 5.454710329048978e-06, "grad_norm": 0.084680937230587, "loss_scale": 4.0, "retries": 0, "seconds": 126.9, "max_memory_gb": 18.16, "ce": 0.010279748217726592, "kl": 0.017215369567566086, "context": 0.018887432641349733} +{"step": 237, "steps": 250, "lr": 5.3962759081475436e-06, "grad_norm": 0.5091727375984192, "loss_scale": 4.0, "retries": 0, "seconds": 127.5, "max_memory_gb": 18.16, "ce": 0.06918482220498845, "kl": 0.023706546278162932, "context": 0.08103809168096632} +{"step": 238, "steps": 250, "lr": 5.341825557225321e-06, "grad_norm": 0.34352460503578186, "loss_scale": 4.0, "retries": 0, "seconds": 128.0, "max_memory_gb": 18.16, "ce": 0.031561302035697736, "kl": 0.009022316033224342, "context": 0.03607246121100616} +{"step": 239, "steps": 250, "lr": 5.291369090646961e-06, "grad_norm": 0.33480289578437805, "loss_scale": 4.0, "retries": 0, "seconds": 128.5, "max_memory_gb": 18.16, "ce": 0.10961479635443538, "kl": 0.021336240570235532, "context": 0.12028291652677581} +{"step": 240, "steps": 250, "lr": 5.244915602902289e-06, "grad_norm": 0.3301413655281067, "loss_scale": 4.0, "retries": 0, "seconds": 129.0, "max_memory_gb": 18.16, "ce": 0.10360276100982446, "kl": 0.03892290751355176, "context": 0.12306421615357976} +{"step": 241, "steps": 250, "lr": 5.20247346696706e-06, "grad_norm": 1.1262208223342896, "loss_scale": 4.0, "retries": 0, "seconds": 129.5, "max_memory_gb": 18.16, "ce": 0.1251646450473345, "kl": 0.026949096303724218, "context": 0.13863919592404272} +{"step": 242, "steps": 250, "lr": 5.164050332793787e-06, "grad_norm": 0.1120605543255806, "loss_scale": 4.0, "retries": 0, "seconds": 130.0, "max_memory_gb": 18.16, "ce": 0.007063552904583048, "kl": 0.06028173987942864, "context": 0.03720442167832516} +{"step": 243, "steps": 250, "lr": 5.129653125932873e-06, "grad_norm": 0.06022348627448082, "loss_scale": 4.0, "retries": 0, "seconds": 130.6, "max_memory_gb": 18.16, "ce": 0.005608267325442284, "kl": 0.004706614578026347, "context": 0.007961574476212263} +{"step": 244, "steps": 250, "lr": 5.099288046284329e-06, "grad_norm": 0.9655460119247437, "loss_scale": 4.0, "retries": 0, "seconds": 131.1, "max_memory_gb": 18.16, "ce": 0.04138711076666368, "kl": 0.21312253546420834, "context": 0.14794837896624813} +{"step": 245, "steps": 250, "lr": 5.072960566980277e-06, "grad_norm": 2.1830437183380127, "loss_scale": 4.0, "retries": 0, "seconds": 131.6, "max_memory_gb": 18.16, "ce": 0.43063946662005037, "kl": 0.04215784720145166, "context": 0.4517183913849294} +{"step": 246, "steps": 250, "lr": 5.050675433398435e-06, "grad_norm": 0.01686519756913185, "loss_scale": 4.0, "retries": 0, "seconds": 132.1, "max_memory_gb": 18.16, "ce": 0.0015149969622143544, "kl": 0.000849447240398149, "context": 0.0019397205614950508} +{"step": 247, "steps": 250, "lr": 5.032436662306811e-06, "grad_norm": 0.39668557047843933, "loss_scale": 4.0, "retries": 0, "seconds": 132.6, "max_memory_gb": 18.16, "ce": 0.006870272787637077, "kl": 0.4846877374511678, "context": 0.24921413451374974} +{"step": 248, "steps": 250, "lr": 5.018247541139684e-06, "grad_norm": 0.19301418960094452, "loss_scale": 4.0, "retries": 0, "seconds": 133.1, "max_memory_gb": 18.16, "ce": 0.031124230765271932, "kl": 0.008216290618292987, "context": 0.035232376772910357} +{"step": 249, "steps": 250, "lr": 5.00811062740508e-06, "grad_norm": 0.37099117040634155, "loss_scale": 4.0, "retries": 0, "seconds": 133.7, "max_memory_gb": 18.16, "ce": 0.04723994294181466, "kl": 0.05285527545493096, "context": 0.0736675807274878} +{"step": 250, "steps": 250, "lr": 5.002027748223778e-06, "grad_norm": 0.05412944778800011, "loss_scale": 4.0, "retries": 0, "seconds": 134.2, "max_memory_gb": 18.16, "ce": 0.006934169796295464, "kl": 0.0034225238050566986, "context": 0.008645431458717212} diff --git a/runs/single_internvl/context/L16_s13/adapter/README.md b/runs/single_internvl/context/L16_s13/adapter/README.md new file mode 100644 index 0000000000000000000000000000000000000000..4f9f856cf0e1531b47e25d266833b2977a19d3ac --- /dev/null +++ b/runs/single_internvl/context/L16_s13/adapter/README.md @@ -0,0 +1,207 @@ +--- +base_model: /data/OPSD/assets/models/InternVL3_5-8B-HF +library_name: peft +pipeline_tag: text-generation +tags: +- base_model:adapter:/data/OPSD/assets/models/InternVL3_5-8B-HF +- lora +- transformers +--- + +# Model Card for Model ID + + + + + +## Model Details + +### Model Description + + + + + +- **Developed by:** [More Information Needed] +- **Funded by [optional]:** [More Information Needed] +- **Shared by [optional]:** [More Information Needed] +- **Model type:** [More Information Needed] +- **Language(s) (NLP):** [More Information Needed] +- **License:** [More Information Needed] +- **Finetuned from model [optional]:** [More Information Needed] + +### Model Sources [optional] + + + +- **Repository:** [More Information Needed] +- **Paper [optional]:** [More Information Needed] +- **Demo [optional]:** [More Information Needed] + +## Uses + + + +### Direct Use + + + +[More Information Needed] + +### Downstream Use [optional] + + + +[More Information Needed] + +### Out-of-Scope Use + + + +[More Information Needed] + +## Bias, Risks, and Limitations + + + +[More Information Needed] + +### Recommendations + + + +Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. More information needed for further recommendations. + +## How to Get Started with the Model + +Use the code below to get started with the model. + +[More Information Needed] + +## Training Details + +### Training Data + + + +[More Information Needed] + +### Training Procedure + + + +#### Preprocessing [optional] + +[More Information Needed] + + +#### Training Hyperparameters + +- **Training regime:** [More Information Needed] + +#### Speeds, Sizes, Times [optional] + + + +[More Information Needed] + +## Evaluation + + + +### Testing Data, Factors & Metrics + +#### Testing Data + + + +[More Information Needed] + +#### Factors + + + +[More Information Needed] + +#### Metrics + + + +[More Information Needed] + +### Results + +[More Information Needed] + +#### Summary + + + +## Model Examination [optional] + + + +[More Information Needed] + +## Environmental Impact + + + +Carbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700). + +- **Hardware Type:** [More Information Needed] +- **Hours used:** [More Information Needed] +- **Cloud Provider:** [More Information Needed] +- **Compute Region:** [More Information Needed] +- **Carbon Emitted:** [More Information Needed] + +## Technical Specifications [optional] + +### Model Architecture and Objective + +[More Information Needed] + +### Compute Infrastructure + +[More Information Needed] + +#### Hardware + +[More Information Needed] + +#### Software + +[More Information Needed] + +## Citation [optional] + + + +**BibTeX:** + +[More Information Needed] + +**APA:** + +[More Information Needed] + +## Glossary [optional] + + + +[More Information Needed] + +## More Information [optional] + +[More Information Needed] + +## Model Card Authors [optional] + +[More Information Needed] + +## Model Card Contact + +[More Information Needed] +### Framework versions + +- PEFT 0.20.0 \ No newline at end of file diff --git a/runs/single_internvl/context/L16_s13/adapter/adapter_config.json b/runs/single_internvl/context/L16_s13/adapter/adapter_config.json new file mode 100644 index 0000000000000000000000000000000000000000..22e2412dbb6fb3dc183eaa84490892d69be91401 --- /dev/null +++ b/runs/single_internvl/context/L16_s13/adapter/adapter_config.json @@ -0,0 +1,50 @@ +{ + "alora_invocation_tokens": null, + "alpha_pattern": {}, + "arrow_config": null, + "auto_mapping": null, + "base_model_name_or_path": "/data/OPSD/assets/models/InternVL3_5-8B-HF", + "bias": "none", + "corda_config": null, + "ensure_weight_tying": false, + "eva_config": null, + "exclude_modules": null, + "fan_in_fan_out": false, + "inference_mode": true, + "init_lora_weights": true, + "layer_replication": null, + "layers_pattern": null, + "layers_to_transform": null, + "loftq_config": {}, + "lora_alpha": 32, + "lora_bias": false, + "lora_dropout": 0.05, + "lora_ga_config": null, + "megatron_config": null, + "megatron_core": "megatron.core", + "modules_to_save": null, + "monteclora_config": null, + "peft_type": "LORA", + "peft_version": "0.20.0", + "qalora_group_size": 16, + "r": 16, + "rank_pattern": {}, + "revision": null, + "target_modules": [ + "model.language_model.layers.16.self_attn.k_proj", + "model.language_model.layers.16.self_attn.q_proj", + "model.language_model.layers.16.mlp.up_proj", + "model.language_model.layers.16.self_attn.v_proj", + "model.language_model.layers.16.self_attn.o_proj", + "model.language_model.layers.16.mlp.down_proj", + "model.language_model.layers.16.mlp.gate_proj" + ], + "target_parameters": null, + "task_type": "CAUSAL_LM", + "trainable_token_indices": null, + "use_bdlora": null, + "use_dora": false, + "use_qalora": false, + "use_rslora": false, + "velora_config": null +} \ No newline at end of file diff --git a/runs/single_internvl/context/L16_s13/config.json b/runs/single_internvl/context/L16_s13/config.json new file mode 100644 index 0000000000000000000000000000000000000000..e496dad93237b1b56de62fee30e4cb3f1f673e36 --- /dev/null +++ b/runs/single_internvl/context/L16_s13/config.json @@ -0,0 +1,24 @@ +{ + "behavior": "context", + "seed": 13, + "model": "/data/OPSD/assets/models/InternVL3_5-8B-HF", + "backbone": "internvl", + "train": "/data/OPSD/assets/v3/context.jsonl", + "items": 1000, + "steps": 250, + "epochs": 1, + "steps_per_epoch": 250, + "groups_per_step": 4, + "lr": 5e-05, + "scheduler": "cosine", + "warmup_steps": 16, + "final_fraction": 0.1, + "weight_decay": 0.0, + "clip": 1.0, + "initial_loss_scale": 4.0, + "adapter": { + "layer": 16, + "rank": 16, + "trainable_parameters": 1212416 + } +} \ No newline at end of file diff --git a/runs/single_internvl/context/L16_s13/steps.jsonl b/runs/single_internvl/context/L16_s13/steps.jsonl new file mode 100644 index 0000000000000000000000000000000000000000..c7b08cbb8a9ac2075da8f496a25511eee23e7cd7 --- /dev/null +++ b/runs/single_internvl/context/L16_s13/steps.jsonl @@ -0,0 +1,250 @@ +{"step": 1, "steps": 250, "lr": 3.125e-06, "grad_norm": 1.2319722175598145, "loss_scale": 4.0, "retries": 0, "seconds": 1.5, "max_memory_gb": 17.73, "ce": 0.5227766726748087, "kl": 0.08654405153356493, "context": 0.5660487064160407} +{"step": 2, "steps": 250, "lr": 6.25e-06, "grad_norm": 1.5816937685012817, "loss_scale": 4.0, "retries": 0, "seconds": 1.9, "max_memory_gb": 17.74, "ce": 0.6389555425630533, "kl": 0.4673658170898989, "context": 0.8726384816254722} +{"step": 3, "steps": 250, "lr": 9.375000000000001e-06, "grad_norm": 4.529516696929932, "loss_scale": 4.0, "retries": 0, "seconds": 2.3, "max_memory_gb": 17.74, "ce": 1.0992508679628372, "kl": 0.06730219349265099, "context": 1.132901981472969} +{"step": 4, "steps": 250, "lr": 1.25e-05, "grad_norm": 2.5088653564453125, "loss_scale": 4.0, "retries": 0, "seconds": 2.8, "max_memory_gb": 17.74, "ce": 0.447088462926331, "kl": 0.430662757366008, "context": 0.662419846674311} +{"step": 5, "steps": 250, "lr": 1.5625e-05, "grad_norm": 0.04643377661705017, "loss_scale": 4.0, "retries": 0, "seconds": 3.3, "max_memory_gb": 17.74, "ce": 0.009248094203940127, "kl": 0.008221633281209506, "context": 0.013358911048271693} +{"step": 6, "steps": 250, "lr": 1.8750000000000002e-05, "grad_norm": 1.2712005376815796, "loss_scale": 4.0, "retries": 0, "seconds": 3.8, "max_memory_gb": 17.74, "ce": 0.28200617697439156, "kl": 0.07438646068840171, "context": 0.3191994063236052} +{"step": 7, "steps": 250, "lr": 2.1875e-05, "grad_norm": 3.959188461303711, "loss_scale": 4.0, "retries": 0, "seconds": 4.2, "max_memory_gb": 17.74, "ce": 0.7718550093177328, "kl": 0.12527147167156727, "context": 0.8344907407299615} +{"step": 8, "steps": 250, "lr": 2.5e-05, "grad_norm": 3.7620463371276855, "loss_scale": 4.0, "retries": 0, "seconds": 4.7, "max_memory_gb": 17.75, "ce": 0.7145880274474621, "kl": 0.28356534731574357, "context": 0.8563706986606121} +{"step": 9, "steps": 250, "lr": 2.8125000000000003e-05, "grad_norm": 2.8637924194335938, "loss_scale": 4.0, "retries": 0, "seconds": 5.2, "max_memory_gb": 17.75, "ce": 0.4437823947519064, "kl": 0.136436051921919, "context": 0.5120004303753376} +{"step": 10, "steps": 250, "lr": 3.125e-05, "grad_norm": 0.44045084714889526, "loss_scale": 4.0, "retries": 0, "seconds": 5.7, "max_memory_gb": 17.75, "ce": 0.04012025259726215, "kl": 0.04502242314629257, "context": 0.06263146200217307} +{"step": 11, "steps": 250, "lr": 3.4375e-05, "grad_norm": 3.2226321697235107, "loss_scale": 4.0, "retries": 0, "seconds": 6.1, "max_memory_gb": 17.75, "ce": 0.537233399154502, "kl": 0.004832660577449133, "context": 0.5396497370384168} +{"step": 12, "steps": 250, "lr": 3.7500000000000003e-05, "grad_norm": 2.238381862640381, "loss_scale": 4.0, "retries": 0, "seconds": 6.6, "max_memory_gb": 17.75, "ce": 0.32995051738180337, "kl": 0.0037641099802385725, "context": 0.3318325654927321} +{"step": 13, "steps": 250, "lr": 4.0625000000000005e-05, "grad_norm": 1.7204259634017944, "loss_scale": 4.0, "retries": 0, "seconds": 7.1, "max_memory_gb": 17.75, "ce": 0.30343703570542857, "kl": 0.014591269056836609, "context": 0.31073266006205813} +{"step": 14, "steps": 250, "lr": 4.375e-05, "grad_norm": 3.2855827808380127, "loss_scale": 4.0, "retries": 0, "seconds": 7.5, "max_memory_gb": 17.75, "ce": 0.42983679473400116, "kl": 0.9154488197527826, "context": 0.887561209499836} +{"step": 15, "steps": 250, "lr": 4.6875e-05, "grad_norm": 2.3335177898406982, "loss_scale": 4.0, "retries": 0, "seconds": 8.0, "max_memory_gb": 17.75, "ce": 0.21873645295613642, "kl": 0.4332627390160724, "context": 0.4353678224497344} +{"step": 16, "steps": 250, "lr": 5e-05, "grad_norm": 0.8139029145240784, "loss_scale": 4.0, "retries": 0, "seconds": 8.4, "max_memory_gb": 17.75, "ce": 0.08512376115686493, "kl": 0.0026352268607752194, "context": 0.08644137252122164} +{"step": 17, "steps": 250, "lr": 5e-05, "grad_norm": 0.9985685348510742, "loss_scale": 4.0, "retries": 0, "seconds": 8.8, "max_memory_gb": 17.75, "ce": 0.16359343999647535, "kl": 0.8323929276084527, "context": 0.5797899084864184} +{"step": 18, "steps": 250, "lr": 4.999797225177623e-05, "grad_norm": 1.0310770273208618, "loss_scale": 4.0, "retries": 0, "seconds": 9.3, "max_memory_gb": 17.75, "ce": 0.07354483184462879, "kl": 0.2281871897575911, "context": 0.18763842617045157} +{"step": 19, "steps": 250, "lr": 4.999188937259492e-05, "grad_norm": 1.0408846139907837, "loss_scale": 4.0, "retries": 0, "seconds": 9.7, "max_memory_gb": 17.75, "ce": 0.06581765785813332, "kl": 0.17552482686005533, "context": 0.15358007233589888} +{"step": 20, "steps": 250, "lr": 4.998175245886032e-05, "grad_norm": 0.08073427528142929, "loss_scale": 4.0, "retries": 0, "seconds": 10.1, "max_memory_gb": 17.75, "ce": 0.0025511375097266864, "kl": 0.01957241916352359, "context": 0.012337346859567333} +{"step": 21, "steps": 250, "lr": 4.996756333769319e-05, "grad_norm": 0.17704032361507416, "loss_scale": 4.0, "retries": 0, "seconds": 10.6, "max_memory_gb": 17.75, "ce": 0.007585319341160357, "kl": 0.01734199763268407, "context": 0.01625631831757346} +{"step": 22, "steps": 250, "lr": 4.994932456660157e-05, "grad_norm": 0.30719903111457825, "loss_scale": 4.0, "retries": 0, "seconds": 11.0, "max_memory_gb": 17.75, "ce": 0.0358782771800179, "kl": 0.017825100912887137, "context": 0.044790828455006704} +{"step": 23, "steps": 250, "lr": 4.992703943301973e-05, "grad_norm": 3.949448347091675, "loss_scale": 4.0, "retries": 0, "seconds": 11.5, "max_memory_gb": 17.75, "ce": 0.19682922237370803, "kl": 0.9719000103363555, "context": 0.6827792298699933} +{"step": 24, "steps": 250, "lr": 4.9900711953715675e-05, "grad_norm": 0.3523868918418884, "loss_scale": 4.0, "retries": 0, "seconds": 11.9, "max_memory_gb": 17.75, "ce": 0.014857353060506284, "kl": 0.07416656205896288, "context": 0.05194063461385667} +{"step": 25, "steps": 250, "lr": 4.987034687406713e-05, "grad_norm": 1.099323034286499, "loss_scale": 4.0, "retries": 0, "seconds": 12.3, "max_memory_gb": 17.75, "ce": 0.3129321330034145, "kl": 0.1373327684123069, "context": 0.38159852381795645} +{"step": 26, "steps": 250, "lr": 4.9835949667206215e-05, "grad_norm": 0.2828271985054016, "loss_scale": 4.0, "retries": 0, "seconds": 12.8, "max_memory_gb": 17.75, "ce": 0.020881773147266358, "kl": 0.647398776258342, "context": 0.34458114800509065} +{"step": 27, "steps": 250, "lr": 4.9797526533032945e-05, "grad_norm": 0.2982996106147766, "loss_scale": 4.0, "retries": 0, "seconds": 13.2, "max_memory_gb": 17.75, "ce": 0.08496072042544256, "kl": 0.3192923987780887, "context": 0.2446069226534746} +{"step": 28, "steps": 250, "lr": 4.9755084397097716e-05, "grad_norm": 1.2198131084442139, "loss_scale": 4.0, "retries": 0, "seconds": 13.6, "max_memory_gb": 17.75, "ce": 0.46289670816622674, "kl": 0.9748985916376114, "context": 0.950346008874476} +{"step": 29, "steps": 250, "lr": 4.970863090935304e-05, "grad_norm": 0.5356914401054382, "loss_scale": 4.0, "retries": 0, "seconds": 14.1, "max_memory_gb": 17.75, "ce": 0.04399340273812413, "kl": 0.3817104079644196, "context": 0.23484860768076032} +{"step": 30, "steps": 250, "lr": 4.965817444277468e-05, "grad_norm": 0.2844121754169464, "loss_scale": 4.0, "retries": 0, "seconds": 14.5, "max_memory_gb": 17.75, "ce": 0.015760199996293522, "kl": 0.0009044391872521373, "context": 0.016212419910516473} +{"step": 31, "steps": 250, "lr": 4.9603724091852456e-05, "grad_norm": 0.6890127062797546, "loss_scale": 4.0, "retries": 0, "seconds": 15.0, "max_memory_gb": 17.75, "ce": 0.17778790299780667, "kl": 0.10806185624096543, "context": 0.23181882360950112} +{"step": 32, "steps": 250, "lr": 4.9545289670951025e-05, "grad_norm": 0.16904953122138977, "loss_scale": 4.0, "retries": 0, "seconds": 15.4, "max_memory_gb": 17.75, "ce": 0.008160828787367791, "kl": 0.006549293422722258, "context": 0.011435475607868284} +{"step": 33, "steps": 250, "lr": 4.948288171254089e-05, "grad_norm": 0.33346301317214966, "loss_scale": 4.0, "retries": 0, "seconds": 15.8, "max_memory_gb": 17.75, "ce": 0.044160736724734306, "kl": 0.017306996160186827, "context": 0.05281423311680555} +{"step": 34, "steps": 250, "lr": 4.941651146529998e-05, "grad_norm": 0.6070664525032043, "loss_scale": 4.0, "retries": 0, "seconds": 16.3, "max_memory_gb": 17.75, "ce": 0.0037004765836172737, "kl": 0.7847345011796278, "context": 0.3960677367358585} +{"step": 35, "steps": 250, "lr": 4.9346190892086174e-05, "grad_norm": 0.04317091405391693, "loss_scale": 4.0, "retries": 0, "seconds": 16.7, "max_memory_gb": 17.75, "ce": 0.0024420841509709135, "kl": 0.0009611186792426452, "context": 0.002922643441706896} +{"step": 36, "steps": 250, "lr": 4.927193266778101e-05, "grad_norm": 0.30805668234825134, "loss_scale": 4.0, "retries": 0, "seconds": 17.2, "max_memory_gb": 17.75, "ce": 0.07952551599009894, "kl": 0.05343276564963162, "context": 0.10624190038652159} +{"step": 37, "steps": 250, "lr": 4.919375017700515e-05, "grad_norm": 0.635492205619812, "loss_scale": 4.0, "retries": 0, "seconds": 17.6, "max_memory_gb": 17.75, "ce": 0.12163863080786541, "kl": 0.793500219238922, "context": 0.5183887349558063} +{"step": 38, "steps": 250, "lr": 4.911165751170588e-05, "grad_norm": 0.37430837750434875, "loss_scale": 4.0, "retries": 0, "seconds": 18.0, "max_memory_gb": 17.75, "ce": 0.017454226268455386, "kl": 0.46769361472979654, "context": 0.25130103318952024} +{"step": 39, "steps": 250, "lr": 4.902566946861708e-05, "grad_norm": 0.5293652415275574, "loss_scale": 4.0, "retries": 0, "seconds": 18.5, "max_memory_gb": 17.75, "ce": 0.10709668917843373, "kl": 0.011555174729437567, "context": 0.11287427612114698} +{"step": 40, "steps": 250, "lr": 4.89358015465922e-05, "grad_norm": 0.12179223448038101, "loss_scale": 4.0, "retries": 0, "seconds": 18.9, "max_memory_gb": 17.75, "ce": 0.007602894795127213, "kl": 0.008062628083280288, "context": 0.01163420919328928} +{"step": 41, "steps": 250, "lr": 4.884206994381078e-05, "grad_norm": 1.4430935382843018, "loss_scale": 4.0, "retries": 0, "seconds": 19.3, "max_memory_gb": 17.75, "ce": 0.37909196471446194, "kl": 0.11942226230166852, "context": 0.43880308186635375} +{"step": 42, "steps": 250, "lr": 4.874449155485866e-05, "grad_norm": 0.02662881836295128, "loss_scale": 4.0, "retries": 0, "seconds": 19.8, "max_memory_gb": 17.75, "ce": 0.0034217201537103392, "kl": 0.00038804997711849865, "context": 0.003615745124989189} +{"step": 43, "steps": 250, "lr": 4.864308396768294e-05, "grad_norm": 0.045526858419179916, "loss_scale": 4.0, "retries": 0, "seconds": 20.2, "max_memory_gb": 17.75, "ce": 0.0022446137663791887, "kl": 0.0021288305142661557, "context": 0.003309029038064182} +{"step": 44, "steps": 250, "lr": 4.853786546042184e-05, "grad_norm": 5.104233264923096, "loss_scale": 4.0, "retries": 0, "seconds": 20.7, "max_memory_gb": 17.75, "ce": 0.46780399506678805, "kl": 1.2835047396292794, "context": 1.109556359529961} +{"step": 45, "steps": 250, "lr": 4.842885499811012e-05, "grad_norm": 1.8080356121063232, "loss_scale": 4.0, "retries": 0, "seconds": 21.1, "max_memory_gb": 17.75, "ce": 0.7670933515473735, "kl": 0.2691634900547797, "context": 0.9016750784066971} +{"step": 46, "steps": 250, "lr": 4.831607222926078e-05, "grad_norm": 0.6320818066596985, "loss_scale": 4.0, "retries": 0, "seconds": 21.5, "max_memory_gb": 17.75, "ce": 0.06590125848015305, "kl": 0.25917781243333593, "context": 0.19549016596283764} +{"step": 47, "steps": 250, "lr": 4.8199537482323545e-05, "grad_norm": 0.08844560384750366, "loss_scale": 4.0, "retries": 0, "seconds": 21.9, "max_memory_gb": 17.75, "ce": 0.005352779608074343, "kl": 0.015219655237160623, "context": 0.012962606706423685} +{"step": 48, "steps": 250, "lr": 4.807927176202073e-05, "grad_norm": 0.051266804337501526, "loss_scale": 4.0, "retries": 0, "seconds": 22.4, "max_memory_gb": 17.75, "ce": 0.0016162213360075839, "kl": 0.0031306592836699565, "context": 0.003181551001034677} +{"step": 49, "steps": 250, "lr": 4.795529674556129e-05, "grad_norm": 0.391717791557312, "loss_scale": 4.0, "retries": 0, "seconds": 22.8, "max_memory_gb": 17.75, "ce": 0.01352538014180027, "kl": 0.3026531023059533, "context": 0.16485193740663817} +{"step": 50, "steps": 250, "lr": 4.782763477873361e-05, "grad_norm": 0.5219753384590149, "loss_scale": 4.0, "retries": 0, "seconds": 23.2, "max_memory_gb": 17.75, "ce": 0.09080663228451158, "kl": 0.00802992016542703, "context": 0.09482159321487416} +{"step": 51, "steps": 250, "lr": 4.769630887187782e-05, "grad_norm": 2.3101372718811035, "loss_scale": 4.0, "retries": 0, "seconds": 23.7, "max_memory_gb": 17.75, "ce": 0.4425612008126336, "kl": 0.03311379184162888, "context": 0.4591180934694421} +{"step": 52, "steps": 250, "lr": 4.756134269573832e-05, "grad_norm": 0.5332924723625183, "loss_scale": 4.0, "retries": 0, "seconds": 24.1, "max_memory_gb": 17.75, "ce": 0.021229013276752084, "kl": 0.17538295057602227, "context": 0.10892049293033779} +{"step": 53, "steps": 250, "lr": 4.7422760577197226e-05, "grad_norm": 0.7617713212966919, "loss_scale": 4.0, "retries": 0, "seconds": 24.6, "max_memory_gb": 17.75, "ce": 0.09887583972886205, "kl": 0.4813906028866768, "context": 0.33957114396616817} +{"step": 54, "steps": 250, "lr": 4.728058749488963e-05, "grad_norm": 1.1211837530136108, "loss_scale": 4.0, "retries": 0, "seconds": 25.0, "max_memory_gb": 17.75, "ce": 0.035028248879825696, "kl": 0.28597314207581803, "context": 0.17801482579670846} +{"step": 55, "steps": 250, "lr": 4.713484907470133e-05, "grad_norm": 0.1428728848695755, "loss_scale": 4.0, "retries": 0, "seconds": 25.5, "max_memory_gb": 17.75, "ce": 0.018602756375912577, "kl": 0.014308106041198698, "context": 0.025756810326129198} +{"step": 56, "steps": 250, "lr": 4.6985571585149876e-05, "grad_norm": 0.29583948850631714, "loss_scale": 4.0, "retries": 0, "seconds": 25.9, "max_memory_gb": 17.75, "ce": 0.02185801276937127, "kl": 0.25578948005568236, "context": 0.14975275774486363} +{"step": 57, "steps": 250, "lr": 4.6832781932649884e-05, "grad_norm": 0.12227892130613327, "loss_scale": 4.0, "retries": 0, "seconds": 26.3, "max_memory_gb": 17.75, "ce": 0.010239342853310518, "kl": 0.020706203235022258, "context": 0.020592443586792797} +{"step": 58, "steps": 250, "lr": 4.667650765666325e-05, "grad_norm": 0.035881407558918, "loss_scale": 4.0, "retries": 0, "seconds": 26.7, "max_memory_gb": 17.75, "ce": 0.001304059973335825, "kl": 0.0007606188024169569, "context": 0.0016843693629198242} +{"step": 59, "steps": 250, "lr": 4.651677692473538e-05, "grad_norm": 0.2853132486343384, "loss_scale": 4.0, "retries": 0, "seconds": 27.2, "max_memory_gb": 17.75, "ce": 0.041735532257007435, "kl": 0.06670631683664396, "context": 0.07508869250887074} +{"step": 60, "steps": 250, "lr": 4.635361852741808e-05, "grad_norm": 0.31162741780281067, "loss_scale": 4.0, "retries": 0, "seconds": 27.6, "max_memory_gb": 17.75, "ce": 0.0843187996360939, "kl": 0.06899556837743148, "context": 0.11881658365018666} +{"step": 61, "steps": 250, "lr": 4.6187061873080264e-05, "grad_norm": 0.7945164442062378, "loss_scale": 4.0, "retries": 0, "seconds": 28.0, "max_memory_gb": 17.75, "ce": 0.02396601709187962, "kl": 0.8115472139070334, "context": 0.4297396107867826} +{"step": 62, "steps": 250, "lr": 4.601713698260727e-05, "grad_norm": 0.21081724762916565, "loss_scale": 4.0, "retries": 0, "seconds": 28.5, "max_memory_gb": 17.75, "ce": 0.024591358305769973, "kl": 0.00864206990081584, "context": 0.028912392794154584} +{"step": 63, "steps": 250, "lr": 4.5843874483989746e-05, "grad_norm": 0.2865503132343292, "loss_scale": 4.0, "retries": 0, "seconds": 28.9, "max_memory_gb": 17.75, "ce": 0.04016635240986943, "kl": 0.12653535199933685, "context": 0.10343402717262506} +{"step": 64, "steps": 250, "lr": 4.566730560680316e-05, "grad_norm": 0.12457866966724396, "loss_scale": 4.0, "retries": 0, "seconds": 29.4, "max_memory_gb": 17.75, "ce": 0.01510238234186545, "kl": 0.01126182044390589, "context": 0.020733292447403073} +{"step": 65, "steps": 250, "lr": 4.548746217657878e-05, "grad_norm": 0.7419996857643127, "loss_scale": 4.0, "retries": 0, "seconds": 29.8, "max_memory_gb": 17.75, "ce": 0.14589713071472943, "kl": 0.15358047129120678, "context": 0.2226873654872179} +{"step": 66, "steps": 250, "lr": 4.5304376609067396e-05, "grad_norm": 0.6293042898178101, "loss_scale": 4.0, "retries": 0, "seconds": 30.2, "max_memory_gb": 17.75, "ce": 0.1558044792795954, "kl": 0.06323601368580967, "context": 0.18742248610078605} +{"step": 67, "steps": 250, "lr": 4.5118081904396504e-05, "grad_norm": 0.06744783371686935, "loss_scale": 4.0, "retries": 0, "seconds": 30.7, "max_memory_gb": 17.75, "ce": 0.005310065491357818, "kl": 0.00873762292394531, "context": 0.009678876958787441} +{"step": 68, "steps": 250, "lr": 4.4928611641122225e-05, "grad_norm": 0.16388416290283203, "loss_scale": 4.0, "retries": 0, "seconds": 31.1, "max_memory_gb": 17.75, "ce": 0.005562044607358985, "kl": 0.027530822921107756, "context": 0.01932745586236706} +{"step": 69, "steps": 250, "lr": 4.473599997017701e-05, "grad_norm": 1.026323914527893, "loss_scale": 4.0, "retries": 0, "seconds": 31.6, "max_memory_gb": 17.75, "ce": 0.31745226922794245, "kl": 0.025388464906427544, "context": 0.3301464952819515} +{"step": 70, "steps": 250, "lr": 4.4540281608714053e-05, "grad_norm": 0.03424953296780586, "loss_scale": 4.0, "retries": 0, "seconds": 32.0, "max_memory_gb": 17.75, "ce": 0.0023130252957344055, "kl": 0.0003449123141763266, "context": 0.002485481498297304} +{"step": 71, "steps": 250, "lr": 4.434149183384977e-05, "grad_norm": 0.09006053954362869, "loss_scale": 4.0, "retries": 0, "seconds": 32.4, "max_memory_gb": 17.75, "ce": 0.009005183335830225, "kl": 0.015038112105685286, "context": 0.01652423950145021} +{"step": 72, "steps": 250, "lr": 4.413966647630534e-05, "grad_norm": 0.31693875789642334, "loss_scale": 4.0, "retries": 0, "seconds": 32.9, "max_memory_gb": 17.75, "ce": 0.015285825589671731, "kl": 0.36893973930273205, "context": 0.19975570059614256} +{"step": 73, "steps": 250, "lr": 4.3934841913948323e-05, "grad_norm": 1.4469068050384521, "loss_scale": 4.0, "retries": 0, "seconds": 33.3, "max_memory_gb": 17.75, "ce": 0.28444774417221197, "kl": 0.057368131823750446, "context": 0.3131318128507701} +{"step": 74, "steps": 250, "lr": 4.372705506523583e-05, "grad_norm": 0.2938985824584961, "loss_scale": 4.0, "retries": 0, "seconds": 33.7, "max_memory_gb": 17.75, "ce": 0.06554116206098115, "kl": 0.09210888262896333, "context": 0.11159560332453111} +{"step": 75, "steps": 250, "lr": 4.351634338256017e-05, "grad_norm": 1.0557628870010376, "loss_scale": 4.0, "retries": 0, "seconds": 34.2, "max_memory_gb": 17.75, "ce": 0.44182629125134554, "kl": 0.9357189498841763, "context": 0.9096857588738203} +{"step": 76, "steps": 250, "lr": 4.330274484549825e-05, "grad_norm": 1.3138376474380493, "loss_scale": 4.0, "retries": 0, "seconds": 34.6, "max_memory_gb": 17.75, "ce": 0.018233422364573926, "kl": 1.6337685361504555, "context": 0.8351176885189489} +{"step": 77, "steps": 250, "lr": 4.308629795396599e-05, "grad_norm": 0.0966842994093895, "loss_scale": 4.0, "retries": 0, "seconds": 35.1, "max_memory_gb": 17.75, "ce": 0.012319152854615822, "kl": 0.013002211426282884, "context": 0.01882025843951851} +{"step": 78, "steps": 250, "lr": 4.286704172127894e-05, "grad_norm": 0.18083380162715912, "loss_scale": 4.0, "retries": 0, "seconds": 35.5, "max_memory_gb": 17.75, "ce": 0.002363760519074276, "kl": 0.08512501957011409, "context": 0.0449262717156671} +{"step": 79, "steps": 250, "lr": 4.264501566712033e-05, "grad_norm": 0.29903605580329895, "loss_scale": 4.0, "retries": 0, "seconds": 35.9, "max_memory_gb": 17.75, "ce": 0.015212225578579819, "kl": 0.02127527664470108, "context": 0.0258498629409587} +{"step": 80, "steps": 250, "lr": 4.2420259810417894e-05, "grad_norm": 0.20541560649871826, "loss_scale": 4.0, "retries": 0, "seconds": 36.3, "max_memory_gb": 17.75, "ce": 0.0029040575645922218, "kl": 0.08141858392525592, "context": 0.04361335069552297} +{"step": 81, "steps": 250, "lr": 4.219281466213066e-05, "grad_norm": 0.44740450382232666, "loss_scale": 4.0, "retries": 0, "seconds": 36.8, "max_memory_gb": 17.75, "ce": 0.09062073053792119, "kl": 0.0442758639837848, "context": 0.1127586601069197} +{"step": 82, "steps": 250, "lr": 4.1962721217947135e-05, "grad_norm": 0.06990393996238708, "loss_scale": 4.0, "retries": 0, "seconds": 37.2, "max_memory_gb": 17.75, "ce": 0.006790952058508992, "kl": 0.0019143676818202948, "context": 0.007748135772999376} +{"step": 83, "steps": 250, "lr": 4.173002095089599e-05, "grad_norm": 0.6575981974601746, "loss_scale": 4.0, "retries": 0, "seconds": 37.6, "max_memory_gb": 17.75, "ce": 0.14910627651261166, "kl": 0.702242801176908, "context": 0.5002276778104715} +{"step": 84, "steps": 250, "lr": 4.149475580387084e-05, "grad_norm": 0.3362799286842346, "loss_scale": 4.0, "retries": 0, "seconds": 38.1, "max_memory_gb": 17.75, "ce": 0.0190781129822426, "kl": 0.045475834657736414, "context": 0.0418160305780475} +{"step": 85, "steps": 250, "lr": 4.1256968182070275e-05, "grad_norm": 0.2857571840286255, "loss_scale": 4.0, "retries": 0, "seconds": 38.5, "max_memory_gb": 17.75, "ce": 0.008216898298996966, "kl": 0.00279701614986827, "context": 0.009615406600460119} +{"step": 86, "steps": 250, "lr": 4.101670094535453e-05, "grad_norm": 1.5065709352493286, "loss_scale": 4.0, "retries": 0, "seconds": 38.9, "max_memory_gb": 17.75, "ce": 0.0740868552820757, "kl": 0.29838315036613494, "context": 0.2232784393709153} +{"step": 87, "steps": 250, "lr": 4.077399740052027e-05, "grad_norm": 1.0144890546798706, "loss_scale": 4.0, "retries": 0, "seconds": 39.4, "max_memory_gb": 17.75, "ce": 0.008034655766095966, "kl": 0.644941094564274, "context": 0.3305051983334124} +{"step": 88, "steps": 250, "lr": 4.0528901293494775e-05, "grad_norm": 0.14360055327415466, "loss_scale": 4.0, "retries": 0, "seconds": 39.8, "max_memory_gb": 17.75, "ce": 0.014866362640532316, "kl": 0.017909220701312734, "context": 0.02382097296140273} +{"step": 89, "steps": 250, "lr": 4.0281456801451004e-05, "grad_norm": 0.1699550747871399, "loss_scale": 4.0, "retries": 0, "seconds": 40.2, "max_memory_gb": 17.75, "ce": 0.01868223168185068, "kl": 0.012509164691437036, "context": 0.024936813104432076} +{"step": 90, "steps": 250, "lr": 4.0031708524844884e-05, "grad_norm": 0.09461043030023575, "loss_scale": 4.0, "retries": 0, "seconds": 40.7, "max_memory_gb": 17.75, "ce": 0.009500251355348155, "kl": 0.004519101610640064, "context": 0.011759802233427763} +{"step": 91, "steps": 250, "lr": 3.977970147937635e-05, "grad_norm": 1.9492747783660889, "loss_scale": 4.0, "retries": 0, "seconds": 41.1, "max_memory_gb": 17.75, "ce": 0.41326973207469564, "kl": 0.2824895757221384, "context": 0.5545145311043598} +{"step": 92, "steps": 250, "lr": 3.952548108787553e-05, "grad_norm": 2.584890365600586, "loss_scale": 4.0, "retries": 0, "seconds": 41.6, "max_memory_gb": 17.75, "ce": 0.15758929020375945, "kl": 0.30656878459922154, "context": 0.3108736902795499} +{"step": 93, "steps": 250, "lr": 3.9269093172115496e-05, "grad_norm": 1.7253048419952393, "loss_scale": 4.0, "retries": 0, "seconds": 42.0, "max_memory_gb": 17.75, "ce": 0.0583919171767775, "kl": 0.2885966899630148, "context": 0.20269026415189728} +{"step": 94, "steps": 250, "lr": 3.901058394455323e-05, "grad_norm": 0.8042491674423218, "loss_scale": 4.0, "retries": 0, "seconds": 42.4, "max_memory_gb": 17.75, "ce": 0.0448789456859231, "kl": 0.022187779868545476, "context": 0.05597283644601703} +{"step": 95, "steps": 250, "lr": 3.875e-05, "grad_norm": 2.0449233055114746, "loss_scale": 4.0, "retries": 0, "seconds": 42.9, "max_memory_gb": 17.75, "ce": 0.4036335559503641, "kl": 0.039437796105630696, "context": 0.423352456651628} +{"step": 96, "steps": 250, "lr": 3.8487388307223005e-05, "grad_norm": 1.3956749439239502, "loss_scale": 4.0, "retries": 0, "seconds": 43.3, "max_memory_gb": 17.75, "ce": 0.32938294897030573, "kl": 0.00788552808080567, "context": 0.33332572237122804} +{"step": 97, "steps": 250, "lr": 3.822279620047943e-05, "grad_norm": 0.3960878551006317, "loss_scale": 4.0, "retries": 0, "seconds": 43.8, "max_memory_gb": 17.75, "ce": 0.019558706786483526, "kl": 0.025327188952360302, "context": 0.03222230193205178} +{"step": 98, "steps": 250, "lr": 3.795627137098479e-05, "grad_norm": 0.15226949751377106, "loss_scale": 4.0, "retries": 0, "seconds": 44.2, "max_memory_gb": 17.75, "ce": 0.004288563497539144, "kl": 0.09574983795755543, "context": 0.05216348201793153} +{"step": 99, "steps": 250, "lr": 3.7687861858316844e-05, "grad_norm": 0.09719705581665039, "loss_scale": 4.0, "retries": 0, "seconds": 44.6, "max_memory_gb": 17.75, "ce": 0.012668361494434066, "kl": 0.0065512673354533035, "context": 0.01594399515306577} +{"step": 100, "steps": 250, "lr": 3.741761604175668e-05, "grad_norm": 1.2363700866699219, "loss_scale": 4.0, "retries": 0, "seconds": 45.1, "max_memory_gb": 17.75, "ce": 0.039270549721550196, "kl": 0.4225883500621421, "context": 0.2505647170182783} +{"step": 101, "steps": 250, "lr": 3.714558263156872e-05, "grad_norm": 0.3373444080352783, "loss_scale": 4.0, "retries": 0, "seconds": 45.5, "max_memory_gb": 17.75, "ce": 0.005418326472863555, "kl": 0.06716069096000865, "context": 0.038998671690933406} +{"step": 102, "steps": 250, "lr": 3.6871810660220904e-05, "grad_norm": 0.31649914383888245, "loss_scale": 4.0, "retries": 0, "seconds": 46.0, "max_memory_gb": 17.75, "ce": 0.06671954470220953, "kl": 0.029319228953681886, "context": 0.08137915609404445} +{"step": 103, "steps": 250, "lr": 3.659634947354686e-05, "grad_norm": 0.15366138517856598, "loss_scale": 4.0, "retries": 0, "seconds": 46.4, "max_memory_gb": 17.75, "ce": 0.010742080696218181, "kl": 0.010016119476858876, "context": 0.015750141217722557} +{"step": 104, "steps": 250, "lr": 3.631924872185169e-05, "grad_norm": 0.08618797361850739, "loss_scale": 4.0, "retries": 0, "seconds": 46.8, "max_memory_gb": 17.75, "ce": 0.007370755425654352, "kl": 0.004185648616839899, "context": 0.009463579743169248} +{"step": 105, "steps": 250, "lr": 3.6040558350962674e-05, "grad_norm": 0.6562780141830444, "loss_scale": 4.0, "retries": 0, "seconds": 47.2, "max_memory_gb": 17.75, "ce": 0.032570125767961144, "kl": 0.3463633800856769, "context": 0.20575182256288826} +{"step": 106, "steps": 250, "lr": 3.576032859322687e-05, "grad_norm": 0.05286172777414322, "loss_scale": 4.0, "retries": 0, "seconds": 47.7, "max_memory_gb": 17.75, "ce": 0.004875424288911745, "kl": 0.00257990901809535, "context": 0.0061653788143303245} +{"step": 107, "steps": 250, "lr": 3.5478609958457056e-05, "grad_norm": 1.150240421295166, "loss_scale": 4.0, "retries": 0, "seconds": 48.1, "max_memory_gb": 17.75, "ce": 0.07683309087587986, "kl": 0.13521338615186096, "context": 0.1444397895829752} +{"step": 108, "steps": 250, "lr": 3.519545322482755e-05, "grad_norm": 2.1999425888061523, "loss_scale": 4.0, "retries": 0, "seconds": 48.5, "max_memory_gb": 17.75, "ce": 0.23351499378622975, "kl": 0.016712786240532296, "context": 0.2418713934821426} +{"step": 109, "steps": 250, "lr": 3.491090942972179e-05, "grad_norm": 2.298095464706421, "loss_scale": 4.0, "retries": 0, "seconds": 49.0, "max_memory_gb": 17.75, "ce": 0.1331716906861402, "kl": 0.4183718338608742, "context": 0.34235760662704706} +{"step": 110, "steps": 250, "lr": 3.4625029860533135e-05, "grad_norm": 1.2787846326828003, "loss_scale": 4.0, "retries": 0, "seconds": 49.4, "max_memory_gb": 17.75, "ce": 0.049266109243035316, "kl": 0.15738862217403948, "context": 0.12796042347326875} +{"step": 111, "steps": 250, "lr": 3.433786604542056e-05, "grad_norm": 0.02968861162662506, "loss_scale": 4.0, "retries": 0, "seconds": 49.9, "max_memory_gb": 17.75, "ce": 0.0024493031087331474, "kl": 0.0019338455313118175, "context": 0.0034162257506977767} +{"step": 112, "steps": 250, "lr": 3.404946974402104e-05, "grad_norm": 0.10420659184455872, "loss_scale": 4.0, "retries": 0, "seconds": 50.3, "max_memory_gb": 17.75, "ce": 0.008333651581779122, "kl": 0.0007623517522006296, "context": 0.008714827301446348} +{"step": 113, "steps": 250, "lr": 3.375989293812019e-05, "grad_norm": 0.4173636734485626, "loss_scale": 4.0, "retries": 0, "seconds": 50.7, "max_memory_gb": 17.75, "ce": 0.013523124973289669, "kl": 0.28111930401064456, "context": 0.1540827788412571} +{"step": 114, "steps": 250, "lr": 3.3469187822282766e-05, "grad_norm": 1.690385103225708, "loss_scale": 4.0, "retries": 0, "seconds": 51.2, "max_memory_gb": 17.75, "ce": 0.01768481489853002, "kl": 0.8860530875681434, "context": 0.4607113694655709} +{"step": 115, "steps": 250, "lr": 3.317740679444499e-05, "grad_norm": 0.3118142783641815, "loss_scale": 4.0, "retries": 0, "seconds": 51.6, "max_memory_gb": 17.75, "ce": 0.010880564572289586, "kl": 0.049510176249896176, "context": 0.03563565236981958} +{"step": 116, "steps": 250, "lr": 3.2884602446470056e-05, "grad_norm": 0.6317144632339478, "loss_scale": 4.0, "retries": 0, "seconds": 52.0, "max_memory_gb": 17.75, "ce": 0.06475471268640831, "kl": 0.33452976759735975, "context": 0.23201960395090282} +{"step": 117, "steps": 250, "lr": 3.259082755466876e-05, "grad_norm": 0.24181212484836578, "loss_scale": 4.0, "retries": 0, "seconds": 52.5, "max_memory_gb": 17.76, "ce": 0.008784893609117717, "kl": 0.009290046084061032, "context": 0.013429916551103815} +{"step": 118, "steps": 250, "lr": 3.229613507028686e-05, "grad_norm": 0.8104220032691956, "loss_scale": 4.0, "retries": 0, "seconds": 52.9, "max_memory_gb": 17.76, "ce": 0.2224977606383618, "kl": 0.006784358607546892, "context": 0.2258899469889002} +{"step": 119, "steps": 250, "lr": 3.2000578109961004e-05, "grad_norm": 0.16627328097820282, "loss_scale": 4.0, "retries": 0, "seconds": 53.4, "max_memory_gb": 17.76, "ce": 0.009997440472943708, "kl": 0.020264274721057518, "context": 0.02012957795523107} +{"step": 120, "steps": 250, "lr": 3.1704209946144616e-05, "grad_norm": 2.4856951236724854, "loss_scale": 4.0, "retries": 0, "seconds": 53.8, "max_memory_gb": 17.76, "ce": 0.24491052300436422, "kl": 0.2027280426118523, "context": 0.3462745387223549} +{"step": 121, "steps": 250, "lr": 3.140708399750594e-05, "grad_norm": 0.44285064935684204, "loss_scale": 4.0, "retries": 0, "seconds": 54.3, "max_memory_gb": 17.76, "ce": 0.0668761705965153, "kl": 0.03343300955748418, "context": 0.08359267399646342} +{"step": 122, "steps": 250, "lr": 3.1109253819299606e-05, "grad_norm": 0.93439120054245, "loss_scale": 4.0, "retries": 0, "seconds": 54.7, "max_memory_gb": 17.76, "ce": 0.17231654387433082, "kl": 0.019046180259465473, "context": 0.18183963476622012} +{"step": 123, "steps": 250, "lr": 3.081077309371361e-05, "grad_norm": 0.2629188001155853, "loss_scale": 4.0, "retries": 0, "seconds": 55.1, "max_memory_gb": 17.76, "ce": 0.044872561818920076, "kl": 0.02369675738736987, "context": 0.0567209409782663} +{"step": 124, "steps": 250, "lr": 3.051169562019339e-05, "grad_norm": 1.7695590257644653, "loss_scale": 4.0, "retries": 0, "seconds": 55.5, "max_memory_gb": 17.76, "ce": 0.6627518907771446, "kl": 0.43967319471994415, "context": 0.8825884890393354} +{"step": 125, "steps": 250, "lr": 3.021207530574477e-05, "grad_norm": 0.6798504590988159, "loss_scale": 4.0, "retries": 0, "seconds": 56.0, "max_memory_gb": 17.76, "ce": 0.15063567302422598, "kl": 0.07837077902877354, "context": 0.18982107000192627} +{"step": 126, "steps": 250, "lr": 2.991196615521758e-05, "grad_norm": 2.279172420501709, "loss_scale": 4.0, "retries": 0, "seconds": 56.4, "max_memory_gb": 17.76, "ce": 0.7265258114784956, "kl": 0.09790552221238613, "context": 0.7754785735160112} +{"step": 127, "steps": 250, "lr": 2.9611422261571536e-05, "grad_norm": 0.14486528933048248, "loss_scale": 4.0, "retries": 0, "seconds": 56.8, "max_memory_gb": 17.76, "ce": 0.017357919074129313, "kl": 0.0008631257733213715, "context": 0.017789481658837758} +{"step": 128, "steps": 250, "lr": 2.931049779612634e-05, "grad_norm": 1.505720615386963, "loss_scale": 4.0, "retries": 0, "seconds": 57.3, "max_memory_gb": 17.76, "ce": 0.05704627172963228, "kl": 0.9300430353614502, "context": 0.5220677991164848} +{"step": 129, "steps": 250, "lr": 2.9009246998797557e-05, "grad_norm": 0.3332342803478241, "loss_scale": 4.0, "retries": 0, "seconds": 57.7, "max_memory_gb": 17.76, "ce": 0.08252488600555807, "kl": 0.030831664582365192, "context": 0.09794071922078729} +{"step": 130, "steps": 250, "lr": 2.8707724168320238e-05, "grad_norm": 0.2114253044128418, "loss_scale": 4.0, "retries": 0, "seconds": 58.2, "max_memory_gb": 17.76, "ce": 0.04848963691620156, "kl": 0.2571890303515829, "context": 0.17708415701054037} +{"step": 131, "steps": 250, "lr": 2.840598365246184e-05, "grad_norm": 3.050330400466919, "loss_scale": 4.0, "retries": 0, "seconds": 58.6, "max_memory_gb": 17.76, "ce": 0.40865440876223147, "kl": 0.3746871383918915, "context": 0.5959979774197564} +{"step": 132, "steps": 250, "lr": 2.8104079838226383e-05, "grad_norm": 0.45698508620262146, "loss_scale": 4.0, "retries": 0, "seconds": 59.0, "max_memory_gb": 17.76, "ce": 0.17543839052086696, "kl": 0.031302967116062064, "context": 0.19108986665378325} +{"step": 133, "steps": 250, "lr": 2.7802067142051446e-05, "grad_norm": 1.0265750885009766, "loss_scale": 4.0, "retries": 0, "seconds": 59.5, "max_memory_gb": 17.76, "ce": 0.35228395817102864, "kl": 0.1692846491932869, "context": 0.4369262820109725} +{"step": 134, "steps": 250, "lr": 2.7500000000000004e-05, "grad_norm": 0.5506671071052551, "loss_scale": 4.0, "retries": 0, "seconds": 59.9, "max_memory_gb": 17.76, "ce": 0.12850329758657608, "kl": 0.004555227591936273, "context": 0.13078091185889207} +{"step": 135, "steps": 250, "lr": 2.7197932857948566e-05, "grad_norm": 0.15374143421649933, "loss_scale": 4.0, "retries": 0, "seconds": 60.3, "max_memory_gb": 17.76, "ce": 0.023046563263051212, "kl": 0.09747541211254429, "context": 0.07178426731843501} +{"step": 136, "steps": 250, "lr": 2.689592016177363e-05, "grad_norm": 1.761008381843567, "loss_scale": 4.0, "retries": 0, "seconds": 60.8, "max_memory_gb": 17.76, "ce": 0.04231582389911637, "kl": 0.172406757628778, "context": 0.12851920002140105} +{"step": 137, "steps": 250, "lr": 2.659401634753817e-05, "grad_norm": 0.5376363396644592, "loss_scale": 4.0, "retries": 0, "seconds": 61.2, "max_memory_gb": 17.76, "ce": 0.02902974814060144, "kl": 0.11299000553117367, "context": 0.0855247510044137} +{"step": 138, "steps": 250, "lr": 2.6292275831679764e-05, "grad_norm": 0.5066392421722412, "loss_scale": 4.0, "retries": 0, "seconds": 61.6, "max_memory_gb": 17.76, "ce": 0.06548799620941281, "kl": 0.09427362587302923, "context": 0.11262480914592743} +{"step": 139, "steps": 250, "lr": 2.5990753001202445e-05, "grad_norm": 0.677640974521637, "loss_scale": 4.0, "retries": 0, "seconds": 62.1, "max_memory_gb": 17.76, "ce": 0.007446596850058995, "kl": 0.0803003905020887, "context": 0.047596792377589736} +{"step": 140, "steps": 250, "lr": 2.5689502203873672e-05, "grad_norm": 0.07066713273525238, "loss_scale": 4.0, "retries": 0, "seconds": 62.5, "max_memory_gb": 17.76, "ce": 0.0068557552876882255, "kl": 0.0023523191011918243, "context": 0.008031914883758873} +{"step": 141, "steps": 250, "lr": 2.538857773842847e-05, "grad_norm": 0.13126251101493835, "loss_scale": 4.0, "retries": 0, "seconds": 63.0, "max_memory_gb": 17.76, "ce": 0.011931957735214382, "kl": 0.00779799569863826, "context": 0.015830955671845004} +{"step": 142, "steps": 250, "lr": 2.5088033844782426e-05, "grad_norm": 0.2721940279006958, "loss_scale": 4.0, "retries": 0, "seconds": 63.4, "max_memory_gb": 17.76, "ce": 0.02861450330237858, "kl": 0.0096104888652917, "context": 0.03341974824434146} +{"step": 143, "steps": 250, "lr": 2.4787924694255237e-05, "grad_norm": 2.370868444442749, "loss_scale": 4.0, "retries": 0, "seconds": 63.9, "max_memory_gb": 17.76, "ce": 0.34497529693180695, "kl": 0.03027610812932835, "context": 0.36011335672810674} +{"step": 144, "steps": 250, "lr": 2.4488304379806627e-05, "grad_norm": 0.9684215784072876, "loss_scale": 4.0, "retries": 0, "seconds": 64.3, "max_memory_gb": 17.76, "ce": 0.02707296295557171, "kl": 0.1512584846932441, "context": 0.10270220716483891} +{"step": 145, "steps": 250, "lr": 2.4189226906286388e-05, "grad_norm": 0.05022377893328667, "loss_scale": 4.0, "retries": 0, "seconds": 64.8, "max_memory_gb": 17.76, "ce": 0.002516647888114676, "kl": 0.007422116159887082, "context": 0.006227706035133451} +{"step": 146, "steps": 250, "lr": 2.3890746180700395e-05, "grad_norm": 0.15000703930854797, "loss_scale": 4.0, "retries": 0, "seconds": 65.2, "max_memory_gb": 17.76, "ce": 0.014887655503116548, "kl": 0.014826383147010347, "context": 0.022300846758298576} +{"step": 147, "steps": 250, "lr": 2.359291600249407e-05, "grad_norm": 1.2746680974960327, "loss_scale": 4.0, "retries": 0, "seconds": 65.7, "max_memory_gb": 17.76, "ce": 0.45395328871381935, "kl": 0.47205172292888165, "context": 0.6899791344039841} +{"step": 148, "steps": 250, "lr": 2.3295790053855392e-05, "grad_norm": 0.09196636825799942, "loss_scale": 4.0, "retries": 0, "seconds": 66.2, "max_memory_gb": 17.76, "ce": 0.007128110515623121, "kl": 0.01252257045416627, "context": 0.013389395826379769} +{"step": 149, "steps": 250, "lr": 2.299942189003901e-05, "grad_norm": 0.18821239471435547, "loss_scale": 4.0, "retries": 0, "seconds": 66.7, "max_memory_gb": 17.76, "ce": 0.014128931827144697, "kl": 0.004563076081467443, "context": 0.01641047015436925} +{"step": 150, "steps": 250, "lr": 2.2703864929713144e-05, "grad_norm": 0.27690714597702026, "loss_scale": 4.0, "retries": 0, "seconds": 67.1, "max_memory_gb": 17.76, "ce": 0.06207523384364322, "kl": 0.027053053840063512, "context": 0.07560176216065884} +{"step": 151, "steps": 250, "lr": 2.2409172445331257e-05, "grad_norm": 0.2454889565706253, "loss_scale": 4.0, "retries": 0, "seconds": 67.6, "max_memory_gb": 17.76, "ce": 0.02301519001775887, "kl": 0.026929805702934573, "context": 0.03648009285097942} +{"step": 152, "steps": 250, "lr": 2.211539755352996e-05, "grad_norm": 0.5818483829498291, "loss_scale": 4.0, "retries": 0, "seconds": 68.1, "max_memory_gb": 17.76, "ce": 0.008375917066587135, "kl": 0.5135838402247828, "context": 0.2651678509428166} +{"step": 153, "steps": 250, "lr": 2.1822593205555016e-05, "grad_norm": 0.46188628673553467, "loss_scale": 4.0, "retries": 0, "seconds": 68.6, "max_memory_gb": 17.76, "ce": 0.014320038084406406, "kl": 0.3514323125264127, "context": 0.1900361889929627} +{"step": 154, "steps": 250, "lr": 2.1530812177717236e-05, "grad_norm": 0.23226375877857208, "loss_scale": 4.0, "retries": 0, "seconds": 69.1, "max_memory_gb": 17.76, "ce": 0.04551958345109597, "kl": 0.06921016308479011, "context": 0.08012466120999306} +{"step": 155, "steps": 250, "lr": 2.1240107061879815e-05, "grad_norm": 0.4126468002796173, "loss_scale": 4.0, "retries": 0, "seconds": 69.6, "max_memory_gb": 17.76, "ce": 0.17292694015486632, "kl": 0.009368867811645032, "context": 0.17761137077468447} +{"step": 156, "steps": 250, "lr": 2.095053025597896e-05, "grad_norm": 0.4827142059803009, "loss_scale": 4.0, "retries": 0, "seconds": 70.0, "max_memory_gb": 17.76, "ce": 0.021438008378027007, "kl": 0.009318906705402696, "context": 0.02609746097732568} +{"step": 157, "steps": 250, "lr": 2.066213395457945e-05, "grad_norm": 0.1993989795446396, "loss_scale": 4.0, "retries": 0, "seconds": 70.5, "max_memory_gb": 17.76, "ce": 0.01849867159035057, "kl": 0.025982336490415037, "context": 0.03148983954451978} +{"step": 158, "steps": 250, "lr": 2.0374970139466877e-05, "grad_norm": 0.9820252060890198, "loss_scale": 4.0, "retries": 0, "seconds": 70.9, "max_memory_gb": 17.76, "ce": 0.2360683292208705, "kl": 0.08208657670184039, "context": 0.27711160827311687} +{"step": 159, "steps": 250, "lr": 2.0089090570278213e-05, "grad_norm": 0.544734001159668, "loss_scale": 4.0, "retries": 0, "seconds": 71.3, "max_memory_gb": 17.76, "ce": 0.008403486746829003, "kl": 0.05591864851885475, "context": 0.03636281075887382} +{"step": 160, "steps": 250, "lr": 1.9804546775172457e-05, "grad_norm": 0.5151125192642212, "loss_scale": 4.0, "retries": 0, "seconds": 71.8, "max_memory_gb": 17.76, "ce": 0.06346541142556816, "kl": 0.26437708266894333, "context": 0.19565396028338} +{"step": 161, "steps": 250, "lr": 1.9521390041542946e-05, "grad_norm": 0.48266929388046265, "loss_scale": 4.0, "retries": 0, "seconds": 72.2, "max_memory_gb": 17.76, "ce": 0.10369655440445058, "kl": 0.06143395111166683, "context": 0.1344135290128179} +{"step": 162, "steps": 250, "lr": 1.923967140677313e-05, "grad_norm": 1.219137191772461, "loss_scale": 4.0, "retries": 0, "seconds": 72.6, "max_memory_gb": 17.76, "ce": 0.07102174827014096, "kl": 0.10186438303207979, "context": 0.12195393606089056} +{"step": 163, "steps": 250, "lr": 1.8959441649037335e-05, "grad_norm": 2.255819320678711, "loss_scale": 4.0, "retries": 0, "seconds": 73.0, "max_memory_gb": 17.76, "ce": 0.22634620714234188, "kl": 0.6635856115608476, "context": 0.5581390237202868} +{"step": 164, "steps": 250, "lr": 1.868075127814832e-05, "grad_norm": 1.003279209136963, "loss_scale": 4.0, "retries": 0, "seconds": 73.5, "max_memory_gb": 17.76, "ce": 0.10201258618326392, "kl": 0.0651606731917127, "context": 0.1345929267408792} +{"step": 165, "steps": 250, "lr": 1.8403650526453143e-05, "grad_norm": 0.30607208609580994, "loss_scale": 4.0, "retries": 0, "seconds": 73.9, "max_memory_gb": 17.76, "ce": 0.013783033529762179, "kl": 0.02986034243076574, "context": 0.028713204839732498} +{"step": 166, "steps": 250, "lr": 1.812818933977912e-05, "grad_norm": 1.862757921218872, "loss_scale": 4.0, "retries": 0, "seconds": 74.4, "max_memory_gb": 17.76, "ce": 0.43067175813484937, "kl": 0.09053373607457615, "context": 0.47593861317727715} +{"step": 167, "steps": 250, "lr": 1.785441736843128e-05, "grad_norm": 0.188022643327713, "loss_scale": 4.0, "retries": 0, "seconds": 74.8, "max_memory_gb": 17.76, "ce": 0.02232055552303791, "kl": 0.004482645803363994, "context": 0.0245618789922446} +{"step": 168, "steps": 250, "lr": 1.7582383958243314e-05, "grad_norm": 0.1706693470478058, "loss_scale": 4.0, "retries": 0, "seconds": 75.2, "max_memory_gb": 17.76, "ce": 0.038073682924732566, "kl": 0.0021801362236146815, "context": 0.03916375065455213} +{"step": 169, "steps": 250, "lr": 1.7312138141683164e-05, "grad_norm": 1.279062032699585, "loss_scale": 4.0, "retries": 0, "seconds": 75.7, "max_memory_gb": 17.76, "ce": 0.0854973090463318, "kl": 1.215721234679222, "context": 0.6933579221367836} +{"step": 170, "steps": 250, "lr": 1.704372862901521e-05, "grad_norm": 0.057717856019735336, "loss_scale": 4.0, "retries": 0, "seconds": 76.1, "max_memory_gb": 17.76, "ce": 0.003225672146072611, "kl": 0.0004076594668731559, "context": 0.0034295019577257335} +{"step": 171, "steps": 250, "lr": 1.6777203799520573e-05, "grad_norm": 0.6785839796066284, "loss_scale": 4.0, "retries": 0, "seconds": 76.5, "max_memory_gb": 17.76, "ce": 0.03116613377642352, "kl": 0.24527402082458138, "context": 0.15380314632784575} +{"step": 172, "steps": 250, "lr": 1.6512611692777008e-05, "grad_norm": 0.185169517993927, "loss_scale": 4.0, "retries": 0, "seconds": 77.0, "max_memory_gb": 17.76, "ce": 0.01815204188460484, "kl": 0.015405456826556474, "context": 0.02585477032698691} +{"step": 173, "steps": 250, "lr": 1.6250000000000005e-05, "grad_norm": 0.2321605384349823, "loss_scale": 4.0, "retries": 0, "seconds": 77.4, "max_memory_gb": 17.76, "ce": 0.05330396210774779, "kl": 0.024868192806025036, "context": 0.06573806027881801} +{"step": 174, "steps": 250, "lr": 1.598941605544678e-05, "grad_norm": 0.2701021730899811, "loss_scale": 4.0, "retries": 0, "seconds": 77.8, "max_memory_gb": 17.76, "ce": 0.040360778861213475, "kl": 0.02893872861704949, "context": 0.054830143417348154} +{"step": 175, "steps": 250, "lr": 1.5730906827884506e-05, "grad_norm": 1.0399155616760254, "loss_scale": 4.0, "retries": 0, "seconds": 78.3, "max_memory_gb": 17.76, "ce": 0.096397599671036, "kl": 0.07648251252248883, "context": 0.13463885709643364} +{"step": 176, "steps": 250, "lr": 1.5474518912124476e-05, "grad_norm": 0.08719189465045929, "loss_scale": 4.0, "retries": 0, "seconds": 78.7, "max_memory_gb": 17.76, "ce": 0.010111571697052568, "kl": 0.0022357728848874103, "context": 0.011229458206798881} +{"step": 177, "steps": 250, "lr": 1.5220298520623646e-05, "grad_norm": 0.7653400897979736, "loss_scale": 4.0, "retries": 0, "seconds": 79.1, "max_memory_gb": 17.76, "ce": 0.03325559926452115, "kl": 0.5823945860538515, "context": 0.32445288950111717} +{"step": 178, "steps": 250, "lr": 1.4968291475155121e-05, "grad_norm": 0.7823708653450012, "loss_scale": 4.0, "retries": 0, "seconds": 79.6, "max_memory_gb": 17.76, "ce": 0.08243867487180978, "kl": 0.049154727603308856, "context": 0.1070160404779017} +{"step": 179, "steps": 250, "lr": 1.4718543198548998e-05, "grad_norm": 0.9520983695983887, "loss_scale": 4.0, "retries": 0, "seconds": 80.0, "max_memory_gb": 17.76, "ce": 0.414304635807639, "kl": 0.014149077753245365, "context": 0.4213791764341295} +{"step": 180, "steps": 250, "lr": 1.4471098706505227e-05, "grad_norm": 0.4440711736679077, "loss_scale": 4.0, "retries": 0, "seconds": 80.4, "max_memory_gb": 17.76, "ce": 0.054416080994997174, "kl": 0.03628408996883081, "context": 0.07255812527728267} +{"step": 181, "steps": 250, "lr": 1.4226002599479743e-05, "grad_norm": 0.21577617526054382, "loss_scale": 4.0, "retries": 0, "seconds": 80.9, "max_memory_gb": 17.76, "ce": 0.010128644935321063, "kl": 0.11076293955557048, "context": 0.06551011558622122} +{"step": 182, "steps": 250, "lr": 1.3983299054645471e-05, "grad_norm": 0.5942585468292236, "loss_scale": 4.0, "retries": 0, "seconds": 81.4, "max_memory_gb": 17.76, "ce": 0.032653366622980684, "kl": 0.2576258731423877, "context": 0.16146630130242556} +{"step": 183, "steps": 250, "lr": 1.3743031817929724e-05, "grad_norm": 0.049579519778490067, "loss_scale": 4.0, "retries": 0, "seconds": 81.9, "max_memory_gb": 17.76, "ce": 0.004110555681108963, "kl": 0.0008922657289076596, "context": 0.004556688567390665} +{"step": 184, "steps": 250, "lr": 1.3505244196129164e-05, "grad_norm": 0.11227907240390778, "loss_scale": 4.0, "retries": 0, "seconds": 82.4, "max_memory_gb": 17.76, "ce": 0.005762052373029292, "kl": 0.009009262670588214, "context": 0.010266683471854776} +{"step": 185, "steps": 250, "lr": 1.3269979049104014e-05, "grad_norm": 0.20153909921646118, "loss_scale": 4.0, "retries": 0, "seconds": 82.8, "max_memory_gb": 17.76, "ce": 0.015191715181572363, "kl": 0.07720465350575978, "context": 0.05379404002451338} +{"step": 186, "steps": 250, "lr": 1.3037278782052863e-05, "grad_norm": 1.695488452911377, "loss_scale": 4.0, "retries": 0, "seconds": 83.3, "max_memory_gb": 17.76, "ce": 0.06836436028243043, "kl": 0.1499459714104887, "context": 0.1433373422914883} +{"step": 187, "steps": 250, "lr": 1.2807185337869343e-05, "grad_norm": 0.33492693305015564, "loss_scale": 4.0, "retries": 0, "seconds": 83.8, "max_memory_gb": 17.76, "ce": 0.07645231852075085, "kl": 0.47911122589721344, "context": 0.3160079236840829} +{"step": 188, "steps": 250, "lr": 1.2579740189582115e-05, "grad_norm": 1.1667629480361938, "loss_scale": 4.0, "retries": 0, "seconds": 84.3, "max_memory_gb": 17.76, "ce": 0.20447683627571678, "kl": 0.013956251454146695, "context": 0.21145495827659033} +{"step": 189, "steps": 250, "lr": 1.2354984332879668e-05, "grad_norm": 2.363664150238037, "loss_scale": 4.0, "retries": 0, "seconds": 84.8, "max_memory_gb": 17.76, "ce": 0.3141199405072257, "kl": 0.27507177158258855, "context": 0.4516558228060603} +{"step": 190, "steps": 250, "lr": 1.2132958278721064e-05, "grad_norm": 0.1232505738735199, "loss_scale": 4.0, "retries": 0, "seconds": 85.2, "max_memory_gb": 17.76, "ce": 0.005291048553772271, "kl": 0.03841685570660047, "context": 0.02449947636341676} +{"step": 191, "steps": 250, "lr": 1.1913702046034016e-05, "grad_norm": 2.4564568996429443, "loss_scale": 4.0, "retries": 0, "seconds": 85.7, "max_memory_gb": 17.76, "ce": 0.29739360918756574, "kl": 0.2412491838040296, "context": 0.4180182014242746} +{"step": 192, "steps": 250, "lr": 1.1697255154501756e-05, "grad_norm": 0.9528209567070007, "loss_scale": 4.0, "retries": 0, "seconds": 86.2, "max_memory_gb": 17.76, "ce": 0.4400935329613276, "kl": 0.050634556195291225, "context": 0.465410809207242} +{"step": 193, "steps": 250, "lr": 1.1483656617439842e-05, "grad_norm": 0.07525520026683807, "loss_scale": 4.0, "retries": 0, "seconds": 86.6, "max_memory_gb": 17.76, "ce": 0.005918960276176222, "kl": 0.0036723114644701127, "context": 0.007755116181215271} +{"step": 194, "steps": 250, "lr": 1.1272944934764177e-05, "grad_norm": 0.4375089406967163, "loss_scale": 4.0, "retries": 0, "seconds": 87.1, "max_memory_gb": 17.76, "ce": 0.05792311945697293, "kl": 0.05917314013640862, "context": 0.08750969130778685} +{"step": 195, "steps": 250, "lr": 1.1065158086051684e-05, "grad_norm": 0.19115978479385376, "loss_scale": 4.0, "retries": 0, "seconds": 87.6, "max_memory_gb": 17.76, "ce": 0.018867230173782445, "kl": 0.0055556675397383515, "context": 0.021645064000040293} +{"step": 196, "steps": 250, "lr": 1.086033352369467e-05, "grad_norm": 0.6303121447563171, "loss_scale": 4.0, "retries": 0, "seconds": 88.1, "max_memory_gb": 17.76, "ce": 0.0023015912738628685, "kl": 0.43005563490442, "context": 0.2173294046369847} +{"step": 197, "steps": 250, "lr": 1.0658508166150224e-05, "grad_norm": 1.3285831212997437, "loss_scale": 4.0, "retries": 0, "seconds": 88.6, "max_memory_gb": 17.76, "ce": 0.02578806277597323, "kl": 0.656653359666052, "context": 0.35411475636647083} +{"step": 198, "steps": 250, "lr": 1.0459718391285954e-05, "grad_norm": 0.3062499463558197, "loss_scale": 4.0, "retries": 0, "seconds": 89.0, "max_memory_gb": 17.76, "ce": 0.07232000818476081, "kl": 0.008487863869959256, "context": 0.07656393792422023} +{"step": 199, "steps": 250, "lr": 1.0264000029822999e-05, "grad_norm": 0.6746798753738403, "loss_scale": 4.0, "retries": 0, "seconds": 89.5, "max_memory_gb": 17.76, "ce": 0.14806572685483843, "kl": 0.13468724189442582, "context": 0.21540934860240668} +{"step": 200, "steps": 250, "lr": 1.0071388358877775e-05, "grad_norm": 0.6194244027137756, "loss_scale": 4.0, "retries": 0, "seconds": 90.0, "max_memory_gb": 17.76, "ce": 0.25776460813358426, "kl": 0.03239419839519542, "context": 0.27396170247811824} +{"step": 201, "steps": 250, "lr": 9.881918095603501e-06, "grad_norm": 0.6395638585090637, "loss_scale": 4.0, "retries": 0, "seconds": 90.5, "max_memory_gb": 17.76, "ce": 0.06946144855464809, "kl": 0.08588351105208858, "context": 0.11240320405340753} +{"step": 202, "steps": 250, "lr": 9.695623390932612e-06, "grad_norm": 0.03859371319413185, "loss_scale": 4.0, "retries": 0, "seconds": 90.9, "max_memory_gb": 17.76, "ce": 0.002267458323331084, "kl": 0.003628897128692188, "context": 0.004081906903593335} +{"step": 203, "steps": 250, "lr": 9.51253782342123e-06, "grad_norm": 0.3595825731754303, "loss_scale": 4.0, "retries": 0, "seconds": 91.4, "max_memory_gb": 17.76, "ce": 0.01660524183535017, "kl": 0.01973398239351809, "context": 0.026472233119420707} +{"step": 204, "steps": 250, "lr": 9.332694393196846e-06, "grad_norm": 0.24482221901416779, "loss_scale": 4.0, "retries": 0, "seconds": 91.8, "max_memory_gb": 17.76, "ce": 0.03999587646103464, "kl": 0.011560261191334575, "context": 0.0457760066492483} +{"step": 205, "steps": 250, "lr": 9.156125516010256e-06, "grad_norm": 0.5030831694602966, "loss_scale": 4.0, "retries": 0, "seconds": 92.3, "max_memory_gb": 17.76, "ce": 0.005080799892311916, "kl": 0.5101894842155161, "context": 0.2601755372597836} +{"step": 206, "steps": 250, "lr": 8.982863017392733e-06, "grad_norm": 0.10396161675453186, "loss_scale": 4.0, "retries": 0, "seconds": 92.8, "max_memory_gb": 17.76, "ce": 0.00861313434143085, "kl": 0.0033342337101203157, "context": 0.010280251197400503} +{"step": 207, "steps": 250, "lr": 8.81293812691974e-06, "grad_norm": 0.21354961395263672, "loss_scale": 4.0, "retries": 0, "seconds": 93.3, "max_memory_gb": 17.76, "ce": 0.07337126712081954, "kl": 0.04029238864313811, "context": 0.09351746318861842} +{"step": 208, "steps": 250, "lr": 8.646381472581927e-06, "grad_norm": 0.05763465538620949, "loss_scale": 4.0, "retries": 0, "seconds": 93.8, "max_memory_gb": 17.76, "ce": 0.00567964423680678, "kl": 0.0020556797717290465, "context": 0.006707484048092738} +{"step": 209, "steps": 250, "lr": 8.483223075264622e-06, "grad_norm": 1.3940664529800415, "loss_scale": 4.0, "retries": 0, "seconds": 94.2, "max_memory_gb": 17.76, "ce": 0.213137477170676, "kl": 0.09504482227930566, "context": 0.26065989857306704} +{"step": 210, "steps": 250, "lr": 8.323492343336747e-06, "grad_norm": 0.9328556060791016, "loss_scale": 4.0, "retries": 0, "seconds": 94.7, "max_memory_gb": 17.76, "ce": 0.47157431405503303, "kl": 0.15727548771246802, "context": 0.5502120542078046} +{"step": 211, "steps": 250, "lr": 8.167218067350122e-06, "grad_norm": 0.5555849075317383, "loss_scale": 4.0, "retries": 0, "seconds": 95.2, "max_memory_gb": 17.76, "ce": 0.16292811569292098, "kl": 0.08606490204692818, "context": 0.20596056012436748} +{"step": 212, "steps": 250, "lr": 8.01442841485013e-06, "grad_norm": 2.368835926055908, "loss_scale": 4.0, "retries": 0, "seconds": 95.7, "max_memory_gb": 17.76, "ce": 0.11255127791082487, "kl": 1.0514400774482056, "context": 0.6382713139755651} +{"step": 213, "steps": 250, "lr": 7.865150925298678e-06, "grad_norm": 0.6445244550704956, "loss_scale": 4.0, "retries": 0, "seconds": 96.2, "max_memory_gb": 17.76, "ce": 0.2245038754772395, "kl": 0.03650524090335239, "context": 0.24275649985065684} +{"step": 214, "steps": 250, "lr": 7.719412505110375e-06, "grad_norm": 0.9562677145004272, "loss_scale": 4.0, "retries": 0, "seconds": 96.7, "max_memory_gb": 17.76, "ce": 0.41877738456241786, "kl": 0.1549378631170839, "context": 0.4962463018600829} +{"step": 215, "steps": 250, "lr": 7.57723942280278e-06, "grad_norm": 1.497400164604187, "loss_scale": 4.0, "retries": 0, "seconds": 97.1, "max_memory_gb": 17.76, "ce": 0.5395244858809747, "kl": 0.017734659800908048, "context": 0.5483918237005128} +{"step": 216, "steps": 250, "lr": 7.438657304261687e-06, "grad_norm": 0.2367219775915146, "loss_scale": 4.0, "retries": 0, "seconds": 97.6, "max_memory_gb": 17.76, "ce": 0.017115890368586406, "kl": 0.09497011081839446, "context": 0.06460094579961151} +{"step": 217, "steps": 250, "lr": 7.30369112812219e-06, "grad_norm": 0.05291162431240082, "loss_scale": 4.0, "retries": 0, "seconds": 98.1, "max_memory_gb": 17.76, "ce": 0.003122669630101882, "kl": 0.0032222686568275094, "context": 0.0047338038857560605} +{"step": 218, "steps": 250, "lr": 7.172365221266403e-06, "grad_norm": 0.8192787766456604, "loss_scale": 4.0, "retries": 0, "seconds": 98.6, "max_memory_gb": 17.76, "ce": 0.18519493122585118, "kl": 0.29542140290141106, "context": 0.3329056315124035} +{"step": 219, "steps": 250, "lr": 7.0447032544387154e-06, "grad_norm": 0.6824762225151062, "loss_scale": 4.0, "retries": 0, "seconds": 99.1, "max_memory_gb": 17.76, "ce": 0.011830246963654645, "kl": 0.14615881032659672, "context": 0.08490965407690965} +{"step": 220, "steps": 250, "lr": 6.920728237979274e-06, "grad_norm": 0.2487679421901703, "loss_scale": 4.0, "retries": 0, "seconds": 99.5, "max_memory_gb": 17.76, "ce": 0.06539929447171744, "kl": 0.0054385398270824226, "context": 0.0681185648572864} +{"step": 221, "steps": 250, "lr": 6.800462517676456e-06, "grad_norm": 0.45599135756492615, "loss_scale": 4.0, "retries": 0, "seconds": 100.0, "max_memory_gb": 17.76, "ce": 0.006582457775948569, "kl": 0.48474223726952914, "context": 0.24895357474451885} +{"step": 222, "steps": 250, "lr": 6.683927770739222e-06, "grad_norm": 0.3493168354034424, "loss_scale": 4.0, "retries": 0, "seconds": 100.5, "max_memory_gb": 17.76, "ce": 0.040285558643518016, "kl": 0.0729700718805475, "context": 0.07677059504203498} +{"step": 223, "steps": 250, "lr": 6.571145001889892e-06, "grad_norm": 2.20438551902771, "loss_scale": 4.0, "retries": 0, "seconds": 101.0, "max_memory_gb": 17.76, "ce": 0.16373723090509884, "kl": 0.0106396822957322, "context": 0.1690570694336202} +{"step": 224, "steps": 250, "lr": 6.46213453957817e-06, "grad_norm": 1.3995792865753174, "loss_scale": 4.0, "retries": 0, "seconds": 101.4, "max_memory_gb": 17.76, "ce": 0.08147621026728302, "kl": 0.20465619358037657, "context": 0.18380430416436866} +{"step": 225, "steps": 250, "lr": 6.356916032317063e-06, "grad_norm": 0.3016379177570343, "loss_scale": 4.0, "retries": 0, "seconds": 101.9, "max_memory_gb": 17.76, "ce": 0.007447967771440744, "kl": 0.1809534824351431, "context": 0.09792470608954318} +{"step": 226, "steps": 250, "lr": 6.255508445141349e-06, "grad_norm": 0.31985875964164734, "loss_scale": 4.0, "retries": 0, "seconds": 102.4, "max_memory_gb": 17.76, "ce": 0.02221910160733387, "kl": 0.02354936175106559, "context": 0.03399378265021369} +{"step": 227, "steps": 250, "lr": 6.157930056189228e-06, "grad_norm": 0.22699925303459167, "loss_scale": 4.0, "retries": 0, "seconds": 102.9, "max_memory_gb": 17.76, "ce": 0.01878916163695976, "kl": 0.007332117802434368, "context": 0.022455220110714436} +{"step": 228, "steps": 250, "lr": 6.064198453407796e-06, "grad_norm": 0.45922765135765076, "loss_scale": 4.0, "retries": 0, "seconds": 103.3, "max_memory_gb": 17.76, "ce": 0.00767939371871762, "kl": 0.25652008334873244, "context": 0.13593943428713828} +{"step": 229, "steps": 250, "lr": 5.974330531382935e-06, "grad_norm": 3.201777935028076, "loss_scale": 4.0, "retries": 0, "seconds": 103.8, "max_memory_gb": 17.76, "ce": 0.1228431032068329, "kl": 0.09660664762395754, "context": 0.17114642885280773} +{"step": 230, "steps": 250, "lr": 5.888342488294125e-06, "grad_norm": 1.9272056818008423, "loss_scale": 4.0, "retries": 0, "seconds": 104.3, "max_memory_gb": 17.76, "ce": 0.5027903365553357, "kl": 0.3017790992744267, "context": 0.6536798847373575} +{"step": 231, "steps": 250, "lr": 5.806249822994849e-06, "grad_norm": 0.7826371788978577, "loss_scale": 4.0, "retries": 0, "seconds": 104.7, "max_memory_gb": 17.76, "ce": 0.05612515781831462, "kl": 0.15589997966890223, "context": 0.13407514625578187} +{"step": 232, "steps": 250, "lr": 5.728067332218994e-06, "grad_norm": 0.23935401439666748, "loss_scale": 4.0, "retries": 0, "seconds": 105.2, "max_memory_gb": 17.76, "ce": 0.033443997643189505, "kl": 0.010089904797496274, "context": 0.03848895087139681} +{"step": 233, "steps": 250, "lr": 5.65380910791383e-06, "grad_norm": 0.29737481474876404, "loss_scale": 4.0, "retries": 0, "seconds": 105.7, "max_memory_gb": 17.76, "ce": 0.012174335424788296, "kl": 0.08809996000491083, "context": 0.056224314495921135} +{"step": 234, "steps": 250, "lr": 5.583488534700022e-06, "grad_norm": 0.30905747413635254, "loss_scale": 4.0, "retries": 0, "seconds": 106.2, "max_memory_gb": 17.76, "ce": 0.01578939234605059, "kl": 0.026997083812602796, "context": 0.02928793430328369} +{"step": 235, "steps": 250, "lr": 5.517118287459118e-06, "grad_norm": 1.6592460870742798, "loss_scale": 4.0, "retries": 0, "seconds": 106.6, "max_memory_gb": 17.76, "ce": 0.389337262880872, "kl": 0.009750907549459953, "context": 0.39421270963794086} +{"step": 236, "steps": 250, "lr": 5.454710329048978e-06, "grad_norm": 0.140055850148201, "loss_scale": 4.0, "retries": 0, "seconds": 107.1, "max_memory_gb": 17.76, "ce": 0.013254599703941494, "kl": 0.02081293225637637, "context": 0.023661065351916477} +{"step": 237, "steps": 250, "lr": 5.3962759081475436e-06, "grad_norm": 0.3994230628013611, "loss_scale": 4.0, "retries": 0, "seconds": 107.6, "max_memory_gb": 17.76, "ce": 0.07285418314859271, "kl": 0.01772310498392926, "context": 0.08171573482104577} +{"step": 238, "steps": 250, "lr": 5.341825557225321e-06, "grad_norm": 0.316872775554657, "loss_scale": 4.0, "retries": 0, "seconds": 108.0, "max_memory_gb": 17.76, "ce": 0.03361470889649354, "kl": 0.006567332431586692, "context": 0.036898375779855996} +{"step": 239, "steps": 250, "lr": 5.291369090646961e-06, "grad_norm": 0.4017370045185089, "loss_scale": 4.0, "retries": 0, "seconds": 108.5, "max_memory_gb": 17.76, "ce": 0.13112111610826105, "kl": 0.019825031180516817, "context": 0.14103362511377782} +{"step": 240, "steps": 250, "lr": 5.244915602902289e-06, "grad_norm": 0.4132571518421173, "loss_scale": 4.0, "retries": 0, "seconds": 109.0, "max_memory_gb": 17.76, "ce": 0.11260443681385368, "kl": 0.03632304694474442, "context": 0.1307659560116008} +{"step": 241, "steps": 250, "lr": 5.20247346696706e-06, "grad_norm": 1.3360940217971802, "loss_scale": 4.0, "retries": 0, "seconds": 109.5, "max_memory_gb": 17.76, "ce": 0.19482493663963396, "kl": 0.026850794376514386, "context": 0.20825032808352262} +{"step": 242, "steps": 250, "lr": 5.164050332793787e-06, "grad_norm": 0.16322046518325806, "loss_scale": 4.0, "retries": 0, "seconds": 109.9, "max_memory_gb": 17.76, "ce": 0.011096136586274952, "kl": 0.0530784435059104, "context": 0.037635357410181314} +{"step": 243, "steps": 250, "lr": 5.129653125932873e-06, "grad_norm": 0.08828333020210266, "loss_scale": 4.0, "retries": 0, "seconds": 110.4, "max_memory_gb": 17.76, "ce": 0.007170670200139284, "kl": 0.004954112198902294, "context": 0.009647726430557668} +{"step": 244, "steps": 250, "lr": 5.099288046284329e-06, "grad_norm": 1.620833158493042, "loss_scale": 4.0, "retries": 0, "seconds": 110.9, "max_memory_gb": 17.76, "ce": 0.04731651008478366, "kl": 0.23164171431380964, "context": 0.16313736491429154} +{"step": 245, "steps": 250, "lr": 5.072960566980277e-06, "grad_norm": 1.1102392673492432, "loss_scale": 4.0, "retries": 0, "seconds": 111.4, "max_memory_gb": 17.76, "ce": 0.40667667239904404, "kl": 0.06009319017175585, "context": 0.43672326672822237} +{"step": 246, "steps": 250, "lr": 5.050675433398435e-06, "grad_norm": 0.03187653049826622, "loss_scale": 4.0, "retries": 0, "seconds": 111.9, "max_memory_gb": 17.76, "ce": 0.0021903177694184706, "kl": 0.0010327360305382172, "context": 0.0027066858165198937} +{"step": 247, "steps": 250, "lr": 5.032436662306811e-06, "grad_norm": 0.4834247827529907, "loss_scale": 4.0, "retries": 0, "seconds": 112.3, "max_memory_gb": 17.76, "ce": 0.009634551432100125, "kl": 0.47410824719190714, "context": 0.2466886822367087} +{"step": 248, "steps": 250, "lr": 5.018247541139684e-06, "grad_norm": 0.2816905677318573, "loss_scale": 4.0, "retries": 0, "seconds": 112.8, "max_memory_gb": 17.76, "ce": 0.036656665964983404, "kl": 0.009864516323432326, "context": 0.041588925290852785} +{"step": 249, "steps": 250, "lr": 5.00811062740508e-06, "grad_norm": 0.37335219979286194, "loss_scale": 4.0, "retries": 0, "seconds": 113.3, "max_memory_gb": 17.76, "ce": 0.04693033150397241, "kl": 0.050812262401450425, "context": 0.07233646255917847} +{"step": 250, "steps": 250, "lr": 5.002027748223778e-06, "grad_norm": 0.08638470619916916, "loss_scale": 4.0, "retries": 0, "seconds": 113.8, "max_memory_gb": 17.76, "ce": 0.008923978894017637, "kl": 0.004037753889861051, "context": 0.010942856024485081} diff --git a/runs/single_internvl/context/L20_s13/adapter/README.md b/runs/single_internvl/context/L20_s13/adapter/README.md new file mode 100644 index 0000000000000000000000000000000000000000..4f9f856cf0e1531b47e25d266833b2977a19d3ac --- /dev/null +++ b/runs/single_internvl/context/L20_s13/adapter/README.md @@ -0,0 +1,207 @@ +--- +base_model: /data/OPSD/assets/models/InternVL3_5-8B-HF +library_name: peft +pipeline_tag: text-generation +tags: +- base_model:adapter:/data/OPSD/assets/models/InternVL3_5-8B-HF +- lora +- transformers +--- + +# Model Card for Model ID + + + + + +## Model Details + +### Model Description + + + + + +- **Developed by:** [More Information Needed] +- **Funded by [optional]:** [More Information Needed] +- **Shared by [optional]:** [More Information Needed] +- **Model type:** [More Information Needed] +- **Language(s) (NLP):** [More Information Needed] +- **License:** [More Information Needed] +- **Finetuned from model [optional]:** [More Information Needed] + +### Model Sources [optional] + + + +- **Repository:** [More Information Needed] +- **Paper [optional]:** [More Information Needed] +- **Demo [optional]:** [More Information Needed] + +## Uses + + + +### Direct Use + + + +[More Information Needed] + +### Downstream Use [optional] + + + +[More Information Needed] + +### Out-of-Scope Use + + + +[More Information Needed] + +## Bias, Risks, and Limitations + + + +[More Information Needed] + +### Recommendations + + + +Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. More information needed for further recommendations. + +## How to Get Started with the Model + +Use the code below to get started with the model. + +[More Information Needed] + +## Training Details + +### Training Data + + + +[More Information Needed] + +### Training Procedure + + + +#### Preprocessing [optional] + +[More Information Needed] + + +#### Training Hyperparameters + +- **Training regime:** [More Information Needed] + +#### Speeds, Sizes, Times [optional] + + + +[More Information Needed] + +## Evaluation + + + +### Testing Data, Factors & Metrics + +#### Testing Data + + + +[More Information Needed] + +#### Factors + + + +[More Information Needed] + +#### Metrics + + + +[More Information Needed] + +### Results + +[More Information Needed] + +#### Summary + + + +## Model Examination [optional] + + + +[More Information Needed] + +## Environmental Impact + + + +Carbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700). + +- **Hardware Type:** [More Information Needed] +- **Hours used:** [More Information Needed] +- **Cloud Provider:** [More Information Needed] +- **Compute Region:** [More Information Needed] +- **Carbon Emitted:** [More Information Needed] + +## Technical Specifications [optional] + +### Model Architecture and Objective + +[More Information Needed] + +### Compute Infrastructure + +[More Information Needed] + +#### Hardware + +[More Information Needed] + +#### Software + +[More Information Needed] + +## Citation [optional] + + + +**BibTeX:** + +[More Information Needed] + +**APA:** + +[More Information Needed] + +## Glossary [optional] + + + +[More Information Needed] + +## More Information [optional] + +[More Information Needed] + +## Model Card Authors [optional] + +[More Information Needed] + +## Model Card Contact + +[More Information Needed] +### Framework versions + +- PEFT 0.20.0 \ No newline at end of file diff --git a/runs/single_internvl/context/L20_s13/adapter/adapter_config.json b/runs/single_internvl/context/L20_s13/adapter/adapter_config.json new file mode 100644 index 0000000000000000000000000000000000000000..2355ebc78cf2692622ff7a42db930e33e0e90366 --- /dev/null +++ b/runs/single_internvl/context/L20_s13/adapter/adapter_config.json @@ -0,0 +1,50 @@ +{ + "alora_invocation_tokens": null, + "alpha_pattern": {}, + "arrow_config": null, + "auto_mapping": null, + "base_model_name_or_path": "/data/OPSD/assets/models/InternVL3_5-8B-HF", + "bias": "none", + "corda_config": null, + "ensure_weight_tying": false, + "eva_config": null, + "exclude_modules": null, + "fan_in_fan_out": false, + "inference_mode": true, + "init_lora_weights": true, + "layer_replication": null, + "layers_pattern": null, + "layers_to_transform": null, + "loftq_config": {}, + "lora_alpha": 32, + "lora_bias": false, + "lora_dropout": 0.05, + "lora_ga_config": null, + "megatron_config": null, + "megatron_core": "megatron.core", + "modules_to_save": null, + "monteclora_config": null, + "peft_type": "LORA", + "peft_version": "0.20.0", + "qalora_group_size": 16, + "r": 16, + "rank_pattern": {}, + "revision": null, + "target_modules": [ + "model.language_model.layers.20.mlp.down_proj", + "model.language_model.layers.20.self_attn.o_proj", + "model.language_model.layers.20.mlp.gate_proj", + "model.language_model.layers.20.self_attn.q_proj", + "model.language_model.layers.20.self_attn.k_proj", + "model.language_model.layers.20.self_attn.v_proj", + "model.language_model.layers.20.mlp.up_proj" + ], + "target_parameters": null, + "task_type": "CAUSAL_LM", + "trainable_token_indices": null, + "use_bdlora": null, + "use_dora": false, + "use_qalora": false, + "use_rslora": false, + "velora_config": null +} \ No newline at end of file diff --git a/runs/single_internvl/context/L20_s13/config.json b/runs/single_internvl/context/L20_s13/config.json new file mode 100644 index 0000000000000000000000000000000000000000..220dbe7c4336832b156fdea5dd7545af54c1d69c --- /dev/null +++ b/runs/single_internvl/context/L20_s13/config.json @@ -0,0 +1,24 @@ +{ + "behavior": "context", + "seed": 13, + "model": "/data/OPSD/assets/models/InternVL3_5-8B-HF", + "backbone": "internvl", + "train": "/data/OPSD/assets/v3/context.jsonl", + "items": 1000, + "steps": 250, + "epochs": 1, + "steps_per_epoch": 250, + "groups_per_step": 4, + "lr": 5e-05, + "scheduler": "cosine", + "warmup_steps": 16, + "final_fraction": 0.1, + "weight_decay": 0.0, + "clip": 1.0, + "initial_loss_scale": 4.0, + "adapter": { + "layer": 20, + "rank": 16, + "trainable_parameters": 1212416 + } +} \ No newline at end of file diff --git a/runs/single_internvl/context/L20_s13/steps.jsonl b/runs/single_internvl/context/L20_s13/steps.jsonl new file mode 100644 index 0000000000000000000000000000000000000000..e7854716681f1ae2b796c07be94f4eb1706e8e6f --- /dev/null +++ b/runs/single_internvl/context/L20_s13/steps.jsonl @@ -0,0 +1,250 @@ +{"step": 1, "steps": 250, "lr": 3.125e-06, "grad_norm": 1.4590295553207397, "loss_scale": 4.0, "retries": 0, "seconds": 1.2, "max_memory_gb": 17.41, "ce": 0.5227766726748087, "kl": 0.08654405153356493, "context": 0.5660487064160407} +{"step": 2, "steps": 250, "lr": 6.25e-06, "grad_norm": 1.78007173538208, "loss_scale": 4.0, "retries": 0, "seconds": 1.5, "max_memory_gb": 17.43, "ce": 0.6366294079707586, "kl": 0.4660195497110635, "context": 0.8696392080310034} +{"step": 3, "steps": 250, "lr": 9.375000000000001e-06, "grad_norm": 5.842503070831299, "loss_scale": 4.0, "retries": 0, "seconds": 1.9, "max_memory_gb": 17.43, "ce": 1.0934037640690804, "kl": 0.06840741634368896, "context": 1.1276074573397636} +{"step": 4, "steps": 250, "lr": 1.25e-05, "grad_norm": 3.0623605251312256, "loss_scale": 4.0, "retries": 0, "seconds": 2.4, "max_memory_gb": 17.43, "ce": 0.4437175701750675, "kl": 0.4299729057972854, "context": 0.6587040300000808} +{"step": 5, "steps": 250, "lr": 1.5625e-05, "grad_norm": 0.03569492697715759, "loss_scale": 4.0, "retries": 0, "seconds": 2.8, "max_memory_gb": 17.43, "ce": 0.00918852480026544, "kl": 0.008196011491236277, "context": 0.013286530618643155} +{"step": 6, "steps": 250, "lr": 1.8750000000000002e-05, "grad_norm": 1.599719762802124, "loss_scale": 4.0, "retries": 0, "seconds": 3.2, "max_memory_gb": 17.43, "ce": 0.27817397758190054, "kl": 0.07494046157444245, "context": 0.315644205926219} +{"step": 7, "steps": 250, "lr": 2.1875e-05, "grad_norm": 5.15983247756958, "loss_scale": 4.0, "retries": 0, "seconds": 3.6, "max_memory_gb": 17.43, "ce": 0.7515319935773732, "kl": 0.12480065323666167, "context": 0.813932306924471} +{"step": 8, "steps": 250, "lr": 2.5e-05, "grad_norm": 5.100747108459473, "loss_scale": 4.0, "retries": 0, "seconds": 4.1, "max_memory_gb": 17.43, "ce": 0.6904933545738459, "kl": 0.283041100949049, "context": 0.8320138938724995} +{"step": 9, "steps": 250, "lr": 2.8125000000000003e-05, "grad_norm": 3.38368558883667, "loss_scale": 4.0, "retries": 0, "seconds": 4.5, "max_memory_gb": 17.43, "ce": 0.4274580981582403, "kl": 0.13683928083628416, "context": 0.4958777353167534} +{"step": 10, "steps": 250, "lr": 3.125e-05, "grad_norm": 0.5007935762405396, "loss_scale": 4.0, "retries": 0, "seconds": 5.0, "max_memory_gb": 17.43, "ce": 0.03491228540951852, "kl": 0.04464597976766527, "context": 0.05723527551162988} +{"step": 11, "steps": 250, "lr": 3.4375e-05, "grad_norm": 4.241944313049316, "loss_scale": 4.0, "retries": 0, "seconds": 5.5, "max_memory_gb": 17.43, "ce": 0.49639689180185087, "kl": 0.004847807378610014, "context": 0.4988207812712062} +{"step": 12, "steps": 250, "lr": 3.7500000000000003e-05, "grad_norm": 2.6355199813842773, "loss_scale": 4.0, "retries": 0, "seconds": 5.9, "max_memory_gb": 17.43, "ce": 0.3132792016094754, "kl": 0.003863185357204202, "context": 0.31521079959202325} +{"step": 13, "steps": 250, "lr": 4.0625000000000005e-05, "grad_norm": 2.2507987022399902, "loss_scale": 4.0, "retries": 0, "seconds": 6.3, "max_memory_gb": 17.43, "ce": 0.26802006128491485, "kl": 0.014076146499064635, "context": 0.27505814053074573} +{"step": 14, "steps": 250, "lr": 4.375e-05, "grad_norm": 3.808793067932129, "loss_scale": 4.0, "retries": 0, "seconds": 6.8, "max_memory_gb": 17.43, "ce": 0.3734566271305084, "kl": 0.9031106205657125, "context": 0.8250119350850582} +{"step": 15, "steps": 250, "lr": 4.6875e-05, "grad_norm": 2.638715982437134, "loss_scale": 4.0, "retries": 0, "seconds": 7.2, "max_memory_gb": 17.43, "ce": 0.17491526532876378, "kl": 0.4327442004456543, "context": 0.39128736555881005} +{"step": 16, "steps": 250, "lr": 5e-05, "grad_norm": 0.8238896727561951, "loss_scale": 4.0, "retries": 0, "seconds": 7.7, "max_memory_gb": 17.43, "ce": 0.059422593367344234, "kl": 0.004056053744307064, "context": 0.06145062118594069} +{"step": 17, "steps": 250, "lr": 5e-05, "grad_norm": 1.0013560056686401, "loss_scale": 4.0, "retries": 0, "seconds": 8.1, "max_memory_gb": 17.43, "ce": 0.1371016499469988, "kl": 0.8286110732587986, "context": 0.5514071777288336} +{"step": 18, "steps": 250, "lr": 4.999797225177623e-05, "grad_norm": 0.5703598856925964, "loss_scale": 4.0, "retries": 0, "seconds": 8.5, "max_memory_gb": 17.43, "ce": 0.032141720250365324, "kl": 0.24249948128999677, "context": 0.15339146347832866} +{"step": 19, "steps": 250, "lr": 4.999188937259492e-05, "grad_norm": 0.6333978772163391, "loss_scale": 4.0, "retries": 0, "seconds": 9.0, "max_memory_gb": 17.43, "ce": 0.028871652204543352, "kl": 0.17019960378820542, "context": 0.1139714540913701} +{"step": 20, "steps": 250, "lr": 4.998175245886032e-05, "grad_norm": 0.08504945039749146, "loss_scale": 4.0, "retries": 0, "seconds": 9.4, "max_memory_gb": 17.43, "ce": 0.0023108667473934474, "kl": 0.019928770980186528, "context": 0.012275251910978113} +{"step": 21, "steps": 250, "lr": 4.996756333769319e-05, "grad_norm": 0.08364845812320709, "loss_scale": 4.0, "retries": 0, "seconds": 9.9, "max_memory_gb": 17.43, "ce": 0.0027906876184715657, "kl": 0.01677944063249015, "context": 0.011180408371728845} +{"step": 22, "steps": 250, "lr": 4.994932456660157e-05, "grad_norm": 0.16146539151668549, "loss_scale": 4.0, "retries": 0, "seconds": 10.3, "max_memory_gb": 17.43, "ce": 0.02810489048715681, "kl": 0.013898886591050541, "context": 0.03505433407553937} +{"step": 23, "steps": 250, "lr": 4.992703943301973e-05, "grad_norm": 2.0699622631073, "loss_scale": 4.0, "retries": 0, "seconds": 10.8, "max_memory_gb": 17.43, "ce": 0.2059809923857756, "kl": 0.9915683351068765, "context": 0.7017651593569099} +{"step": 24, "steps": 250, "lr": 4.9900711953715675e-05, "grad_norm": 0.20677560567855835, "loss_scale": 4.0, "retries": 0, "seconds": 11.2, "max_memory_gb": 17.43, "ce": 0.007989924168214202, "kl": 0.0649815768119879, "context": 0.04048071219585836} +{"step": 25, "steps": 250, "lr": 4.987034687406713e-05, "grad_norm": 1.0563938617706299, "loss_scale": 4.0, "retries": 0, "seconds": 11.6, "max_memory_gb": 17.43, "ce": 0.3158848427283374, "kl": 0.11903471150435507, "context": 0.3754021916538477} +{"step": 26, "steps": 250, "lr": 4.9835949667206215e-05, "grad_norm": 0.2925988733768463, "loss_scale": 4.0, "retries": 0, "seconds": 12.1, "max_memory_gb": 17.43, "ce": 0.017313783027930185, "kl": 0.6509802414802834, "context": 0.34280390210915357} +{"step": 27, "steps": 250, "lr": 4.9797526533032945e-05, "grad_norm": 0.3114781677722931, "loss_scale": 4.0, "retries": 0, "seconds": 12.5, "max_memory_gb": 17.43, "ce": 0.08420245898378198, "kl": 0.3152887004639524, "context": 0.24184680593316443} +{"step": 28, "steps": 250, "lr": 4.9755084397097716e-05, "grad_norm": 1.1801263093948364, "loss_scale": 4.0, "retries": 0, "seconds": 12.9, "max_memory_gb": 17.43, "ce": 0.44300341978669167, "kl": 0.987000689143315, "context": 0.9365037470124662} +{"step": 29, "steps": 250, "lr": 4.970863090935304e-05, "grad_norm": 0.4496458172798157, "loss_scale": 4.0, "retries": 0, "seconds": 13.4, "max_memory_gb": 17.43, "ce": 0.05082357977516949, "kl": 0.3629166174941929, "context": 0.23228189459769055} +{"step": 30, "steps": 250, "lr": 4.965817444277468e-05, "grad_norm": 0.23124264180660248, "loss_scale": 4.0, "retries": 0, "seconds": 13.8, "max_memory_gb": 17.43, "ce": 0.012387108705297578, "kl": 0.0008308256960845029, "context": 0.01280252151627792} +{"step": 31, "steps": 250, "lr": 4.9603724091852456e-05, "grad_norm": 0.7295592427253723, "loss_scale": 4.0, "retries": 0, "seconds": 14.3, "max_memory_gb": 17.43, "ce": 0.176368516869843, "kl": 0.11039229782181792, "context": 0.2315646675415337} +{"step": 32, "steps": 250, "lr": 4.9545289670951025e-05, "grad_norm": 0.10880402475595474, "loss_scale": 4.0, "retries": 0, "seconds": 14.7, "max_memory_gb": 17.43, "ce": 0.006317058519925922, "kl": 0.005484528897795826, "context": 0.009059322823304683} +{"step": 33, "steps": 250, "lr": 4.948288171254089e-05, "grad_norm": 0.5952839851379395, "loss_scale": 4.0, "retries": 0, "seconds": 15.2, "max_memory_gb": 17.43, "ce": 0.05794963100925088, "kl": 0.017276032827794552, "context": 0.06658764788880944} +{"step": 34, "steps": 250, "lr": 4.941651146529998e-05, "grad_norm": 0.6218534111976624, "loss_scale": 4.0, "retries": 0, "seconds": 15.6, "max_memory_gb": 17.43, "ce": 0.003538993016263703, "kl": 0.769402709465794, "context": 0.3882403482857626} +{"step": 35, "steps": 250, "lr": 4.9346190892086174e-05, "grad_norm": 0.0334814190864563, "loss_scale": 4.0, "retries": 0, "seconds": 16.0, "max_memory_gb": 17.43, "ce": 0.0021921198786003515, "kl": 0.0008407424511460704, "context": 0.0026124911819351837} +{"step": 36, "steps": 250, "lr": 4.927193266778101e-05, "grad_norm": 0.3715822994709015, "loss_scale": 4.0, "retries": 0, "seconds": 16.5, "max_memory_gb": 17.43, "ce": 0.0866326438554097, "kl": 0.05331891038804315, "context": 0.1132921010866994} +{"step": 37, "steps": 250, "lr": 4.919375017700515e-05, "grad_norm": 0.6451432704925537, "loss_scale": 4.0, "retries": 0, "seconds": 16.9, "max_memory_gb": 17.43, "ce": 0.11707513089640997, "kl": 0.775784456927795, "context": 0.5049673655303195} +{"step": 38, "steps": 250, "lr": 4.911165751170588e-05, "grad_norm": 0.3308546245098114, "loss_scale": 4.0, "retries": 0, "seconds": 17.4, "max_memory_gb": 17.43, "ce": 0.021094653522595763, "kl": 0.46581189129210543, "context": 0.25400060089305043} +{"step": 39, "steps": 250, "lr": 4.902566946861708e-05, "grad_norm": 0.5844442844390869, "loss_scale": 4.0, "retries": 0, "seconds": 17.8, "max_memory_gb": 17.43, "ce": 0.10489695554133505, "kl": 0.009716941283841152, "context": 0.10975542708183639} +{"step": 40, "steps": 250, "lr": 4.89358015465922e-05, "grad_norm": 0.08519139140844345, "loss_scale": 4.0, "retries": 0, "seconds": 18.2, "max_memory_gb": 17.43, "ce": 0.007418603723635897, "kl": 0.006961180624784902, "context": 0.010899193934164941} +{"step": 41, "steps": 250, "lr": 4.884206994381078e-05, "grad_norm": 1.0804206132888794, "loss_scale": 4.0, "retries": 0, "seconds": 18.7, "max_memory_gb": 17.43, "ce": 0.3684101404505782, "kl": 0.11210921686142683, "context": 0.4244647556915879} +{"step": 42, "steps": 250, "lr": 4.874449155485866e-05, "grad_norm": 0.020837802439928055, "loss_scale": 4.0, "retries": 0, "seconds": 19.1, "max_memory_gb": 17.43, "ce": 0.0031220944656524807, "kl": 0.000338321398885455, "context": 0.003291255154181272} +{"step": 43, "steps": 250, "lr": 4.864308396768294e-05, "grad_norm": 0.03592628240585327, "loss_scale": 4.0, "retries": 0, "seconds": 19.5, "max_memory_gb": 17.43, "ce": 0.002527493066736497, "kl": 0.002524338138755411, "context": 0.003789662121562287} +{"step": 44, "steps": 250, "lr": 4.853786546042184e-05, "grad_norm": 5.256229400634766, "loss_scale": 4.0, "retries": 0, "seconds": 20.0, "max_memory_gb": 17.43, "ce": 0.3888126241945429, "kl": 1.2045047816936858, "context": 0.9910650504025398} +{"step": 45, "steps": 250, "lr": 4.842885499811012e-05, "grad_norm": 1.9830766916275024, "loss_scale": 4.0, "retries": 0, "seconds": 20.4, "max_memory_gb": 17.43, "ce": 0.7345142573467456, "kl": 0.3110611511656316, "context": 0.8900448159256484} +{"step": 46, "steps": 250, "lr": 4.831607222926078e-05, "grad_norm": 0.6168208718299866, "loss_scale": 4.0, "retries": 0, "seconds": 20.8, "max_memory_gb": 17.43, "ce": 0.06050255085574463, "kl": 0.24570858286460862, "context": 0.1833568379515782} +{"step": 47, "steps": 250, "lr": 4.8199537482323545e-05, "grad_norm": 0.16110387444496155, "loss_scale": 4.0, "retries": 0, "seconds": 21.3, "max_memory_gb": 17.43, "ce": 0.007095586137438659, "kl": 0.018647240998689085, "context": 0.01641920703696087} +{"step": 48, "steps": 250, "lr": 4.807927176202073e-05, "grad_norm": 0.06652951240539551, "loss_scale": 4.0, "retries": 0, "seconds": 21.7, "max_memory_gb": 17.43, "ce": 0.002800227332045324, "kl": 0.0032193428924074396, "context": 0.004409898887388408} +{"step": 49, "steps": 250, "lr": 4.795529674556129e-05, "grad_norm": 0.43266505002975464, "loss_scale": 4.0, "retries": 0, "seconds": 22.1, "max_memory_gb": 17.43, "ce": 0.012503619909693953, "kl": 0.3061961956577761, "context": 0.16560171075252583} +{"step": 50, "steps": 250, "lr": 4.782763477873361e-05, "grad_norm": 0.44683152437210083, "loss_scale": 4.0, "retries": 0, "seconds": 22.6, "max_memory_gb": 17.43, "ce": 0.07518033167434623, "kl": 0.008170026543666609, "context": 0.07926534212310798} +{"step": 51, "steps": 250, "lr": 4.769630887187782e-05, "grad_norm": 3.214966058731079, "loss_scale": 4.0, "retries": 0, "seconds": 23.0, "max_memory_gb": 17.43, "ce": 0.3396517083856452, "kl": 0.044177739826864126, "context": 0.36174058947653975} +{"step": 52, "steps": 250, "lr": 4.756134269573832e-05, "grad_norm": 0.3487529754638672, "loss_scale": 4.0, "retries": 0, "seconds": 23.4, "max_memory_gb": 17.43, "ce": 0.022971620288444683, "kl": 0.143459607148543, "context": 0.09470142144709826} +{"step": 53, "steps": 250, "lr": 4.7422760577197226e-05, "grad_norm": 0.6475163698196411, "loss_scale": 4.0, "retries": 0, "seconds": 23.9, "max_memory_gb": 17.43, "ce": 0.09023297717794776, "kl": 0.46499473229050636, "context": 0.32273034006357193} +{"step": 54, "steps": 250, "lr": 4.728058749488963e-05, "grad_norm": 1.9940190315246582, "loss_scale": 4.0, "retries": 0, "seconds": 24.3, "max_memory_gb": 17.43, "ce": 0.041798733931500465, "kl": 0.2107516900432529, "context": 0.14717457591905259} +{"step": 55, "steps": 250, "lr": 4.713484907470133e-05, "grad_norm": 0.1597791612148285, "loss_scale": 4.0, "retries": 0, "seconds": 24.7, "max_memory_gb": 17.43, "ce": 0.020175372570520267, "kl": 0.014988486822403502, "context": 0.02766961621819064} +{"step": 56, "steps": 250, "lr": 4.6985571585149876e-05, "grad_norm": 0.30648520588874817, "loss_scale": 4.0, "retries": 0, "seconds": 25.2, "max_memory_gb": 17.43, "ce": 0.02265405119396746, "kl": 0.24025317071937025, "context": 0.14278063410893083} +{"step": 57, "steps": 250, "lr": 4.6832781932649884e-05, "grad_norm": 0.13155564665794373, "loss_scale": 4.0, "retries": 0, "seconds": 25.6, "max_memory_gb": 17.43, "ce": 0.01005603585508652, "kl": 0.019770676117332187, "context": 0.019941373931942508} +{"step": 58, "steps": 250, "lr": 4.667650765666325e-05, "grad_norm": 0.04355127736926079, "loss_scale": 4.0, "retries": 0, "seconds": 26.0, "max_memory_gb": 17.43, "ce": 0.0019352707095094956, "kl": 0.0010422195705359627, "context": 0.0024563805709476583} +{"step": 59, "steps": 250, "lr": 4.651677692473538e-05, "grad_norm": 0.3471390902996063, "loss_scale": 4.0, "retries": 0, "seconds": 26.5, "max_memory_gb": 17.43, "ce": 0.04413880512584001, "kl": 0.06929239176679403, "context": 0.07878499914659187} +{"step": 60, "steps": 250, "lr": 4.635361852741808e-05, "grad_norm": 0.35079559683799744, "loss_scale": 4.0, "retries": 0, "seconds": 26.9, "max_memory_gb": 17.43, "ce": 0.08115507813636214, "kl": 0.06525251944549382, "context": 0.1137813376262784} +{"step": 61, "steps": 250, "lr": 4.6187061873080264e-05, "grad_norm": 0.8055300712585449, "loss_scale": 4.0, "retries": 0, "seconds": 27.3, "max_memory_gb": 17.43, "ce": 0.03022584517020732, "kl": 0.7561344536443357, "context": 0.40829307385138236} +{"step": 62, "steps": 250, "lr": 4.601713698260727e-05, "grad_norm": 0.14643117785453796, "loss_scale": 4.0, "retries": 0, "seconds": 27.8, "max_memory_gb": 17.43, "ce": 0.0203113044553902, "kl": 0.009705642631161027, "context": 0.025164124788716435} +{"step": 63, "steps": 250, "lr": 4.5843874483989746e-05, "grad_norm": 0.3106098473072052, "loss_scale": 4.0, "retries": 0, "seconds": 28.2, "max_memory_gb": 17.43, "ce": 0.04272790369577706, "kl": 0.1165416735020699, "context": 0.10099874250590801} +{"step": 64, "steps": 250, "lr": 4.566730560680316e-05, "grad_norm": 0.12850423157215118, "loss_scale": 4.0, "retries": 0, "seconds": 28.6, "max_memory_gb": 17.43, "ce": 0.015987171849701554, "kl": 0.01221005863044411, "context": 0.022092201048508286} +{"step": 65, "steps": 250, "lr": 4.548746217657878e-05, "grad_norm": 0.6521686911582947, "loss_scale": 4.0, "retries": 0, "seconds": 29.1, "max_memory_gb": 17.43, "ce": 0.12016367982141674, "kl": 0.1478828187682666, "context": 0.19410509569570422} +{"step": 66, "steps": 250, "lr": 4.5304376609067396e-05, "grad_norm": 0.6648927330970764, "loss_scale": 4.0, "retries": 0, "seconds": 29.5, "max_memory_gb": 17.43, "ce": 0.14172255479547857, "kl": 0.06009675246923507, "context": 0.17177093478449024} +{"step": 67, "steps": 250, "lr": 4.5118081904396504e-05, "grad_norm": 0.06657734513282776, "loss_scale": 4.0, "retries": 0, "seconds": 30.0, "max_memory_gb": 17.43, "ce": 0.00612570348312147, "kl": 0.008521442159690196, "context": 0.010386424837633967} +{"step": 68, "steps": 250, "lr": 4.4928611641122225e-05, "grad_norm": 0.1079975962638855, "loss_scale": 4.0, "retries": 0, "seconds": 30.4, "max_memory_gb": 17.43, "ce": 0.006359036844514776, "kl": 0.02101546252379194, "context": 0.016866768193722237} +{"step": 69, "steps": 250, "lr": 4.473599997017701e-05, "grad_norm": 1.5151896476745605, "loss_scale": 4.0, "retries": 0, "seconds": 30.8, "max_memory_gb": 17.43, "ce": 0.3813709410314914, "kl": 0.02287343882198911, "context": 0.3928076521260664} +{"step": 70, "steps": 250, "lr": 4.4540281608714053e-05, "grad_norm": 0.07775301486253738, "loss_scale": 4.0, "retries": 0, "seconds": 31.3, "max_memory_gb": 17.43, "ce": 0.004152380104642361, "kl": 0.0004707925008915481, "context": 0.004387776396470144} +{"step": 71, "steps": 250, "lr": 4.434149183384977e-05, "grad_norm": 0.10408937931060791, "loss_scale": 4.0, "retries": 0, "seconds": 31.7, "max_memory_gb": 17.43, "ce": 0.009601881603884976, "kl": 0.01500730845873477, "context": 0.017105535604059696} +{"step": 72, "steps": 250, "lr": 4.413966647630534e-05, "grad_norm": 0.33736470341682434, "loss_scale": 4.0, "retries": 0, "seconds": 32.2, "max_memory_gb": 17.43, "ce": 0.015269813331542537, "kl": 0.38211002841126174, "context": 0.20632483263034374} +{"step": 73, "steps": 250, "lr": 4.3934841913948323e-05, "grad_norm": 1.3399064540863037, "loss_scale": 4.0, "retries": 0, "seconds": 32.6, "max_memory_gb": 17.43, "ce": 0.21682346364832483, "kl": 0.044994297555604135, "context": 0.2393206086999271} +{"step": 74, "steps": 250, "lr": 4.372705506523583e-05, "grad_norm": 0.2504763603210449, "loss_scale": 4.0, "retries": 0, "seconds": 33.0, "max_memory_gb": 17.43, "ce": 0.0751807359229133, "kl": 0.08913200673123356, "context": 0.11974673937584157} +{"step": 75, "steps": 250, "lr": 4.351634338256017e-05, "grad_norm": 1.0622806549072266, "loss_scale": 4.0, "retries": 0, "seconds": 33.5, "max_memory_gb": 17.43, "ce": 0.47386760004155803, "kl": 0.893038097769022, "context": 0.9203866412863135} +{"step": 76, "steps": 250, "lr": 4.330274484549825e-05, "grad_norm": 1.2058604955673218, "loss_scale": 4.0, "retries": 0, "seconds": 33.9, "max_memory_gb": 17.43, "ce": 0.01687932331697084, "kl": 1.623361369012855, "context": 0.8285599981900305} +{"step": 77, "steps": 250, "lr": 4.308629795396599e-05, "grad_norm": 0.13544629514217377, "loss_scale": 4.0, "retries": 0, "seconds": 34.3, "max_memory_gb": 17.43, "ce": 0.014501469791866839, "kl": 0.014645274786744267, "context": 0.021824107156135142} +{"step": 78, "steps": 250, "lr": 4.286704172127894e-05, "grad_norm": 0.2065483182668686, "loss_scale": 4.0, "retries": 0, "seconds": 34.7, "max_memory_gb": 17.43, "ce": 0.0023367464600596577, "kl": 0.08321637190238107, "context": 0.043944933742750436} +{"step": 79, "steps": 250, "lr": 4.264501566712033e-05, "grad_norm": 0.21659167110919952, "loss_scale": 4.0, "retries": 0, "seconds": 35.2, "max_memory_gb": 17.43, "ce": 0.013792799261864275, "kl": 0.015937344956910238, "context": 0.02176147077989299} +{"step": 80, "steps": 250, "lr": 4.2420259810417894e-05, "grad_norm": 0.1891402304172516, "loss_scale": 4.0, "retries": 0, "seconds": 35.6, "max_memory_gb": 17.43, "ce": 0.003346259831232601, "kl": 0.06757338315844663, "context": 0.037132950468730996} +{"step": 81, "steps": 250, "lr": 4.219281466213066e-05, "grad_norm": 0.4086894989013672, "loss_scale": 4.0, "retries": 0, "seconds": 36.0, "max_memory_gb": 17.43, "ce": 0.08717048610560596, "kl": 0.028728247198159806, "context": 0.10153460968285799} +{"step": 82, "steps": 250, "lr": 4.1962721217947135e-05, "grad_norm": 0.07796826958656311, "loss_scale": 4.0, "retries": 0, "seconds": 36.5, "max_memory_gb": 17.43, "ce": 0.008054286998230964, "kl": 0.0022424659109674394, "context": 0.009175520099233836} +{"step": 83, "steps": 250, "lr": 4.173002095089599e-05, "grad_norm": 0.8515313267707825, "loss_scale": 4.0, "retries": 0, "seconds": 36.9, "max_memory_gb": 17.43, "ce": 0.1511872747214511, "kl": 0.6323062885130639, "context": 0.4673404161585495} +{"step": 84, "steps": 250, "lr": 4.149475580387084e-05, "grad_norm": 0.3333689272403717, "loss_scale": 4.0, "retries": 0, "seconds": 37.3, "max_memory_gb": 17.43, "ce": 0.028020928022669978, "kl": 0.04483404051825346, "context": 0.0504379472185974} +{"step": 85, "steps": 250, "lr": 4.1256968182070275e-05, "grad_norm": 0.06441715359687805, "loss_scale": 4.0, "retries": 0, "seconds": 37.8, "max_memory_gb": 17.43, "ce": 0.002285135213242029, "kl": 0.000849529958472317, "context": 0.0027099002199975075} +{"step": 86, "steps": 250, "lr": 4.101670094535453e-05, "grad_norm": 1.0212243795394897, "loss_scale": 4.0, "retries": 0, "seconds": 38.2, "max_memory_gb": 17.43, "ce": 0.05146839760709554, "kl": 0.23303778923582286, "context": 0.167987288441509} +{"step": 87, "steps": 250, "lr": 4.077399740052027e-05, "grad_norm": 0.919080913066864, "loss_scale": 4.0, "retries": 0, "seconds": 38.6, "max_memory_gb": 17.43, "ce": 0.008744585502427071, "kl": 0.6363969976082444, "context": 0.32694309717044234} +{"step": 88, "steps": 250, "lr": 4.0528901293494775e-05, "grad_norm": 0.2870903015136719, "loss_scale": 4.0, "retries": 0, "seconds": 39.1, "max_memory_gb": 17.43, "ce": 0.0223031496097974, "kl": 0.013695430367576478, "context": 0.029150864778785035} +{"step": 89, "steps": 250, "lr": 4.0281456801451004e-05, "grad_norm": 0.3578581213951111, "loss_scale": 4.0, "retries": 0, "seconds": 39.5, "max_memory_gb": 17.43, "ce": 0.025923942286681267, "kl": 0.021779876267828513, "context": 0.03681387849064777} +{"step": 90, "steps": 250, "lr": 4.0031708524844884e-05, "grad_norm": 0.15714870393276215, "loss_scale": 4.0, "retries": 0, "seconds": 39.9, "max_memory_gb": 17.43, "ce": 0.011107431433629245, "kl": 0.005350622421246953, "context": 0.013782742549665272} +{"step": 91, "steps": 250, "lr": 3.977970147937635e-05, "grad_norm": 2.752161741256714, "loss_scale": 4.0, "retries": 0, "seconds": 40.4, "max_memory_gb": 17.43, "ce": 0.5620770691311918, "kl": 0.22485357311961707, "context": 0.674503842194099} +{"step": 92, "steps": 250, "lr": 3.952548108787553e-05, "grad_norm": 1.9731318950653076, "loss_scale": 4.0, "retries": 0, "seconds": 40.8, "max_memory_gb": 17.43, "ce": 0.06884440993599128, "kl": 0.2411640100481236, "context": 0.18942641423200257} +{"step": 93, "steps": 250, "lr": 3.9269093172115496e-05, "grad_norm": 1.1871824264526367, "loss_scale": 4.0, "retries": 0, "seconds": 41.3, "max_memory_gb": 17.43, "ce": 0.050266940001165494, "kl": 0.1967766959569417, "context": 0.14865528437076136} +{"step": 94, "steps": 250, "lr": 3.901058394455323e-05, "grad_norm": 0.48708540201187134, "loss_scale": 4.0, "retries": 0, "seconds": 41.7, "max_memory_gb": 17.43, "ce": 0.02731466552359052, "kl": 0.022778558763093315, "context": 0.038703944912413135} +{"step": 95, "steps": 250, "lr": 3.875e-05, "grad_norm": 1.8533868789672852, "loss_scale": 4.0, "retries": 0, "seconds": 42.1, "max_memory_gb": 17.43, "ce": 0.32720155267452355, "kl": 0.06440478900913149, "context": 0.3594039441086352} +{"step": 96, "steps": 250, "lr": 3.8487388307223005e-05, "grad_norm": 1.7158684730529785, "loss_scale": 4.0, "retries": 0, "seconds": 42.6, "max_memory_gb": 17.43, "ce": 0.26567180597339757, "kl": 0.012080118562153075, "context": 0.2717118702130392} +{"step": 97, "steps": 250, "lr": 3.822279620047943e-05, "grad_norm": 0.3808671534061432, "loss_scale": 4.0, "retries": 0, "seconds": 43.0, "max_memory_gb": 17.43, "ce": 0.016993138153338805, "kl": 0.0325813292292878, "context": 0.033283802680671215} +{"step": 98, "steps": 250, "lr": 3.795627137098479e-05, "grad_norm": 0.16177786886692047, "loss_scale": 4.0, "retries": 0, "seconds": 43.4, "max_memory_gb": 17.43, "ce": 0.004596342914737761, "kl": 0.08229222648151335, "context": 0.04574245730327675} +{"step": 99, "steps": 250, "lr": 3.7687861858316844e-05, "grad_norm": 0.10119668394327164, "loss_scale": 4.0, "retries": 0, "seconds": 43.9, "max_memory_gb": 17.43, "ce": 0.013608573670353508, "kl": 0.0037594231216644403, "context": 0.015488285003812052} +{"step": 100, "steps": 250, "lr": 3.741761604175668e-05, "grad_norm": 0.7325934767723083, "loss_scale": 4.0, "retries": 0, "seconds": 44.3, "max_memory_gb": 17.43, "ce": 0.04031593873514794, "kl": 0.37348852603827254, "context": 0.22706019831821322} +{"step": 101, "steps": 250, "lr": 3.714558263156872e-05, "grad_norm": 0.39551404118537903, "loss_scale": 4.0, "retries": 0, "seconds": 44.7, "max_memory_gb": 17.43, "ce": 0.005332737433491275, "kl": 0.09055385447572917, "context": 0.05060966464225203} +{"step": 102, "steps": 250, "lr": 3.6871810660220904e-05, "grad_norm": 0.30780673027038574, "loss_scale": 4.0, "retries": 0, "seconds": 45.2, "max_memory_gb": 17.43, "ce": 0.06024105520918965, "kl": 0.02938046772032976, "context": 0.07493129116483033} +{"step": 103, "steps": 250, "lr": 3.659634947354686e-05, "grad_norm": 0.10728354007005692, "loss_scale": 4.0, "retries": 0, "seconds": 45.6, "max_memory_gb": 17.43, "ce": 0.009640236996347085, "kl": 0.008846853670547716, "context": 0.014063664071727544} +{"step": 104, "steps": 250, "lr": 3.631924872185169e-05, "grad_norm": 0.20399275422096252, "loss_scale": 4.0, "retries": 0, "seconds": 46.0, "max_memory_gb": 17.43, "ce": 0.011648617415630724, "kl": 0.005115276864671614, "context": 0.014206255524186417} +{"step": 105, "steps": 250, "lr": 3.6040558350962674e-05, "grad_norm": 0.46492430567741394, "loss_scale": 4.0, "retries": 0, "seconds": 46.4, "max_memory_gb": 17.43, "ce": 0.03758461441611871, "kl": 0.3149016574025154, "context": 0.19503544154576957} +{"step": 106, "steps": 250, "lr": 3.576032859322687e-05, "grad_norm": 0.04618300497531891, "loss_scale": 4.0, "retries": 0, "seconds": 46.9, "max_memory_gb": 17.43, "ce": 0.004708804422989488, "kl": 0.002523512150219176, "context": 0.005970560596324503} +{"step": 107, "steps": 250, "lr": 3.5478609958457056e-05, "grad_norm": 0.4975419044494629, "loss_scale": 4.0, "retries": 0, "seconds": 47.3, "max_memory_gb": 17.43, "ce": 0.018177331949118525, "kl": 0.17779076874853672, "context": 0.10707271473074798} +{"step": 108, "steps": 250, "lr": 3.519545322482755e-05, "grad_norm": 1.897992730140686, "loss_scale": 4.0, "retries": 0, "seconds": 47.7, "max_memory_gb": 17.43, "ce": 0.3628122572990833, "kl": 0.013431661176355192, "context": 0.36952810172806494} +{"step": 109, "steps": 250, "lr": 3.491090942972179e-05, "grad_norm": 1.8871122598648071, "loss_scale": 4.0, "retries": 0, "seconds": 48.2, "max_memory_gb": 17.43, "ce": 0.14720601204317063, "kl": 0.3518263138830662, "context": 0.32311917189508677} +{"step": 110, "steps": 250, "lr": 3.4625029860533135e-05, "grad_norm": 0.9294307231903076, "loss_scale": 4.0, "retries": 0, "seconds": 48.6, "max_memory_gb": 17.43, "ce": 0.03929390059784055, "kl": 0.036022286163643, "context": 0.05730504426173866} +{"step": 111, "steps": 250, "lr": 3.433786604542056e-05, "grad_norm": 0.022702375426888466, "loss_scale": 4.0, "retries": 0, "seconds": 49.0, "max_memory_gb": 17.43, "ce": 0.0018701931403484195, "kl": 0.0017409043575753458, "context": 0.0027406452572904527} +{"step": 112, "steps": 250, "lr": 3.404946974402104e-05, "grad_norm": 0.09127989411354065, "loss_scale": 4.0, "retries": 0, "seconds": 49.4, "max_memory_gb": 17.43, "ce": 0.00872225931379944, "kl": 0.0015916419815766858, "context": 0.009518080158159137} +{"step": 113, "steps": 250, "lr": 3.375989293812019e-05, "grad_norm": 0.7182163000106812, "loss_scale": 4.0, "retries": 0, "seconds": 49.9, "max_memory_gb": 17.43, "ce": 0.016893642488867044, "kl": 0.2631824993295595, "context": 0.14848489011637866} +{"step": 114, "steps": 250, "lr": 3.3469187822282766e-05, "grad_norm": 1.339119791984558, "loss_scale": 4.0, "retries": 0, "seconds": 50.3, "max_memory_gb": 17.43, "ce": 0.016464120970340446, "kl": 0.8116697008663323, "context": 0.42229895264608786} +{"step": 115, "steps": 250, "lr": 3.317740679444499e-05, "grad_norm": 0.7177527546882629, "loss_scale": 4.0, "retries": 0, "seconds": 50.7, "max_memory_gb": 17.43, "ce": 0.017375864612404257, "kl": 0.05089364654122619, "context": 0.04282268678070977} +{"step": 116, "steps": 250, "lr": 3.2884602446470056e-05, "grad_norm": 0.5892314314842224, "loss_scale": 4.0, "retries": 0, "seconds": 51.2, "max_memory_gb": 17.43, "ce": 0.061770426109433174, "kl": 0.36993687456606494, "context": 0.24673887081735302} +{"step": 117, "steps": 250, "lr": 3.259082755466876e-05, "grad_norm": 0.12474849075078964, "loss_scale": 4.0, "retries": 0, "seconds": 51.6, "max_memory_gb": 17.44, "ce": 0.005949444457655773, "kl": 0.005534401621844154, "context": 0.008716645301319659} +{"step": 118, "steps": 250, "lr": 3.229613507028686e-05, "grad_norm": 0.9120950102806091, "loss_scale": 4.0, "retries": 0, "seconds": 52.0, "max_memory_gb": 17.44, "ce": 0.24902319205284584, "kl": 0.004609224737578188, "context": 0.25132780810236} +{"step": 119, "steps": 250, "lr": 3.2000578109961004e-05, "grad_norm": 0.17609155178070068, "loss_scale": 4.0, "retries": 0, "seconds": 52.5, "max_memory_gb": 17.44, "ce": 0.011862875500810333, "kl": 0.019088812479026274, "context": 0.021407281528809108} +{"step": 120, "steps": 250, "lr": 3.1704209946144616e-05, "grad_norm": 1.4973896741867065, "loss_scale": 4.0, "retries": 0, "seconds": 52.9, "max_memory_gb": 17.44, "ce": 0.21465037914458662, "kl": 0.2005518880323507, "context": 0.3149263139348477} +{"step": 121, "steps": 250, "lr": 3.140708399750594e-05, "grad_norm": 0.7288948893547058, "loss_scale": 4.0, "retries": 0, "seconds": 53.4, "max_memory_gb": 17.44, "ce": 0.08601042210648302, "kl": 0.04481054850111832, "context": 0.10841569821059238} +{"step": 122, "steps": 250, "lr": 3.1109253819299606e-05, "grad_norm": 1.4408925771713257, "loss_scale": 4.0, "retries": 0, "seconds": 53.8, "max_memory_gb": 17.44, "ce": 0.2027026844516513, "kl": 0.018982908452017, "context": 0.21219413296785206} +{"step": 123, "steps": 250, "lr": 3.081077309371361e-05, "grad_norm": 0.3471967279911041, "loss_scale": 4.0, "retries": 0, "seconds": 54.2, "max_memory_gb": 17.44, "ce": 0.04475720995105803, "kl": 0.0299777791951783, "context": 0.059746101149357855} +{"step": 124, "steps": 250, "lr": 3.051169562019339e-05, "grad_norm": 1.6607682704925537, "loss_scale": 4.0, "retries": 0, "seconds": 54.7, "max_memory_gb": 17.44, "ce": 0.6766616315289866, "kl": 0.3915491316729458, "context": 0.8724361815548036} +{"step": 125, "steps": 250, "lr": 3.021207530574477e-05, "grad_norm": 0.5751215219497681, "loss_scale": 4.0, "retries": 0, "seconds": 55.1, "max_memory_gb": 17.44, "ce": 0.16993627676856704, "kl": 0.05813453498558374, "context": 0.19900353875709698} +{"step": 126, "steps": 250, "lr": 2.991196615521758e-05, "grad_norm": 1.9519726037979126, "loss_scale": 4.0, "retries": 0, "seconds": 55.5, "max_memory_gb": 17.44, "ce": 0.7117552161216736, "kl": 0.1060451939702034, "context": 0.7647778242826462} +{"step": 127, "steps": 250, "lr": 2.9611422261571536e-05, "grad_norm": 0.1534527987241745, "loss_scale": 4.0, "retries": 0, "seconds": 56.0, "max_memory_gb": 17.44, "ce": 0.017438133945688605, "kl": 0.0008098506586975418, "context": 0.01784305917681195} +{"step": 128, "steps": 250, "lr": 2.931049779612634e-05, "grad_norm": 1.4248194694519043, "loss_scale": 4.0, "retries": 0, "seconds": 56.4, "max_memory_gb": 17.44, "ce": 0.04205541124974843, "kl": 0.8243524878635071, "context": 0.4542316566221416} +{"step": 129, "steps": 250, "lr": 2.9009246998797557e-05, "grad_norm": 0.3368087410926819, "loss_scale": 4.0, "retries": 0, "seconds": 56.9, "max_memory_gb": 17.44, "ce": 0.0843405710766092, "kl": 0.028234510340553243, "context": 0.09845782537013292} +{"step": 130, "steps": 250, "lr": 2.8707724168320238e-05, "grad_norm": 0.25774237513542175, "loss_scale": 4.0, "retries": 0, "seconds": 57.3, "max_memory_gb": 17.44, "ce": 0.04619685828220099, "kl": 0.2835838664250332, "context": 0.18798879568930715} +{"step": 131, "steps": 250, "lr": 2.840598365246184e-05, "grad_norm": 3.063838481903076, "loss_scale": 4.0, "retries": 0, "seconds": 57.7, "max_memory_gb": 17.44, "ce": 0.37667796231107786, "kl": 0.3032837637292687, "context": 0.5283198292599991} +{"step": 132, "steps": 250, "lr": 2.8104079838226383e-05, "grad_norm": 0.5085240602493286, "loss_scale": 4.0, "retries": 0, "seconds": 58.2, "max_memory_gb": 17.44, "ce": 0.1707285970041994, "kl": 0.03403631550463615, "context": 0.18774675985332578} +{"step": 133, "steps": 250, "lr": 2.7802067142051446e-05, "grad_norm": 1.2576524019241333, "loss_scale": 4.0, "retries": 0, "seconds": 58.6, "max_memory_gb": 17.44, "ce": 0.3815022477065213, "kl": 0.15178912552073598, "context": 0.4573968155309558} +{"step": 134, "steps": 250, "lr": 2.7500000000000004e-05, "grad_norm": 0.5528141260147095, "loss_scale": 4.0, "retries": 0, "seconds": 59.0, "max_memory_gb": 17.44, "ce": 0.13555632450152189, "kl": 0.0024999873628530622, "context": 0.1368063256086316} +{"step": 135, "steps": 250, "lr": 2.7197932857948566e-05, "grad_norm": 0.21274833381175995, "loss_scale": 4.0, "retries": 0, "seconds": 59.5, "max_memory_gb": 17.44, "ce": 0.016190930851735175, "kl": 0.11563970544375479, "context": 0.07401078712427989} +{"step": 136, "steps": 250, "lr": 2.689592016177363e-05, "grad_norm": 1.5158169269561768, "loss_scale": 4.0, "retries": 0, "seconds": 59.9, "max_memory_gb": 17.44, "ce": 0.0353808244690299, "kl": 0.18906222764053382, "context": 0.12991193949710578} +{"step": 137, "steps": 250, "lr": 2.659401634753817e-05, "grad_norm": 0.49042657017707825, "loss_scale": 4.0, "retries": 0, "seconds": 60.3, "max_memory_gb": 17.44, "ce": 0.03656268924532924, "kl": 0.0768446993897669, "context": 0.07498503886745311} +{"step": 138, "steps": 250, "lr": 2.6292275831679764e-05, "grad_norm": 0.4863108992576599, "loss_scale": 4.0, "retries": 0, "seconds": 60.7, "max_memory_gb": 17.44, "ce": 0.06822736607864499, "kl": 0.09333021892234683, "context": 0.11489247251302004} +{"step": 139, "steps": 250, "lr": 2.5990753001202445e-05, "grad_norm": 0.9081736207008362, "loss_scale": 4.0, "retries": 0, "seconds": 61.2, "max_memory_gb": 17.44, "ce": 0.008133156254189089, "kl": 0.09661234832128685, "context": 0.05643933133978862} +{"step": 140, "steps": 250, "lr": 2.5689502203873672e-05, "grad_norm": 0.0850965678691864, "loss_scale": 4.0, "retries": 0, "seconds": 61.6, "max_memory_gb": 17.44, "ce": 0.009195544669637457, "kl": 0.0038366234584827907, "context": 0.011113856104202569} +{"step": 141, "steps": 250, "lr": 2.538857773842847e-05, "grad_norm": 0.1403789222240448, "loss_scale": 4.0, "retries": 0, "seconds": 62.1, "max_memory_gb": 17.44, "ce": 0.011927231680601835, "kl": 0.005820182210300118, "context": 0.014837322407402098} +{"step": 142, "steps": 250, "lr": 2.5088033844782426e-05, "grad_norm": 0.26932889223098755, "loss_scale": 4.0, "retries": 0, "seconds": 62.5, "max_memory_gb": 17.44, "ce": 0.03195401199627668, "kl": 0.012189406290417537, "context": 0.03804871469037607} +{"step": 143, "steps": 250, "lr": 2.4787924694255237e-05, "grad_norm": 1.8961385488510132, "loss_scale": 4.0, "retries": 0, "seconds": 62.9, "max_memory_gb": 17.44, "ce": 0.3258642994769616, "kl": 0.023536186057754094, "context": 0.3376323847915046} +{"step": 144, "steps": 250, "lr": 2.4488304379806627e-05, "grad_norm": 0.7084206938743591, "loss_scale": 4.0, "retries": 0, "seconds": 63.4, "max_memory_gb": 17.44, "ce": 0.04122256813570857, "kl": 0.189561280538328, "context": 0.13600320648401976} +{"step": 145, "steps": 250, "lr": 2.4189226906286388e-05, "grad_norm": 0.042604122310876846, "loss_scale": 4.0, "retries": 0, "seconds": 63.8, "max_memory_gb": 17.44, "ce": 0.002653244649991393, "kl": 0.006617161083340761, "context": 0.005961825110716745} +{"step": 146, "steps": 250, "lr": 2.3890746180700395e-05, "grad_norm": 0.15715667605400085, "loss_scale": 4.0, "retries": 0, "seconds": 64.2, "max_memory_gb": 17.44, "ce": 0.01736276736482978, "kl": 0.007787728405673988, "context": 0.02125663112383336} +{"step": 147, "steps": 250, "lr": 2.359291600249407e-05, "grad_norm": 1.3081551790237427, "loss_scale": 4.0, "retries": 0, "seconds": 64.7, "max_memory_gb": 17.44, "ce": 0.43453878158470616, "kl": 0.4365038828109391, "context": 0.6527907423733268} +{"step": 148, "steps": 250, "lr": 2.3295790053855392e-05, "grad_norm": 0.14775905013084412, "loss_scale": 4.0, "retries": 0, "seconds": 65.1, "max_memory_gb": 17.44, "ce": 0.009060713062353898, "kl": 0.019024658380658366, "context": 0.01857304225268308} +{"step": 149, "steps": 250, "lr": 2.299942189003901e-05, "grad_norm": 0.12054844945669174, "loss_scale": 4.0, "retries": 0, "seconds": 65.5, "max_memory_gb": 17.44, "ce": 0.014305992342997342, "kl": 0.0043464515038067475, "context": 0.016479217680171132} +{"step": 150, "steps": 250, "lr": 2.2703864929713144e-05, "grad_norm": 0.26041179895401, "loss_scale": 4.0, "retries": 0, "seconds": 65.9, "max_memory_gb": 17.44, "ce": 0.05841323744971305, "kl": 0.02325980761088431, "context": 0.07004314288496971} +{"step": 151, "steps": 250, "lr": 2.2409172445331257e-05, "grad_norm": 0.36780160665512085, "loss_scale": 4.0, "retries": 0, "seconds": 66.4, "max_memory_gb": 17.44, "ce": 0.029550086765084416, "kl": 0.03179397237181547, "context": 0.04544707291643135} +{"step": 152, "steps": 250, "lr": 2.211539755352996e-05, "grad_norm": 0.6610993146896362, "loss_scale": 4.0, "retries": 0, "seconds": 66.8, "max_memory_gb": 17.44, "ce": 0.009157744701951742, "kl": 0.4712741695384466, "context": 0.2447948312619701} +{"step": 153, "steps": 250, "lr": 2.1822593205555016e-05, "grad_norm": 0.49421021342277527, "loss_scale": 4.0, "retries": 0, "seconds": 67.3, "max_memory_gb": 17.44, "ce": 0.013039601792115718, "kl": 0.36418798775230243, "context": 0.19513359008124098} +{"step": 154, "steps": 250, "lr": 2.1530812177717236e-05, "grad_norm": 0.32275745272636414, "loss_scale": 4.0, "retries": 0, "seconds": 67.7, "max_memory_gb": 17.44, "ce": 0.05691598710836843, "kl": 0.06833901716163382, "context": 0.09108549583470449} +{"step": 155, "steps": 250, "lr": 2.1240107061879815e-05, "grad_norm": 0.47943341732025146, "loss_scale": 4.0, "retries": 0, "seconds": 68.1, "max_memory_gb": 17.44, "ce": 0.16189597420452628, "kl": 0.008772267028689384, "context": 0.16628210166527424} +{"step": 156, "steps": 250, "lr": 2.095053025597896e-05, "grad_norm": 0.19050337374210358, "loss_scale": 4.0, "retries": 0, "seconds": 68.6, "max_memory_gb": 17.44, "ce": 0.016567753606068436, "kl": 0.007365813882643124, "context": 0.020250660374586005} +{"step": 157, "steps": 250, "lr": 2.066213395457945e-05, "grad_norm": 0.23434270918369293, "loss_scale": 4.0, "retries": 0, "seconds": 69.0, "max_memory_gb": 17.44, "ce": 0.02033697720617056, "kl": 0.029195878640166484, "context": 0.034934915718622506} +{"step": 158, "steps": 250, "lr": 2.0374970139466877e-05, "grad_norm": 0.8608682155609131, "loss_scale": 4.0, "retries": 0, "seconds": 69.4, "max_memory_gb": 17.44, "ce": 0.21446965765790083, "kl": 0.11723991252074484, "context": 0.27308962476672605} +{"step": 159, "steps": 250, "lr": 2.0089090570278213e-05, "grad_norm": 0.5015860795974731, "loss_scale": 4.0, "retries": 0, "seconds": 69.9, "max_memory_gb": 17.44, "ce": 0.0070190281549002975, "kl": 0.0646919715218246, "context": 0.039365015807561576} +{"step": 160, "steps": 250, "lr": 1.9804546775172457e-05, "grad_norm": 0.4162678122520447, "loss_scale": 4.0, "retries": 0, "seconds": 70.3, "max_memory_gb": 17.44, "ce": 0.05985602951841429, "kl": 0.2420002314975136, "context": 0.1808561374200508} +{"step": 161, "steps": 250, "lr": 1.9521390041542946e-05, "grad_norm": 0.6541949510574341, "loss_scale": 4.0, "retries": 0, "seconds": 70.7, "max_memory_gb": 17.44, "ce": 0.11182200101029593, "kl": 0.06641335651556801, "context": 0.1450286792969564} +{"step": 162, "steps": 250, "lr": 1.923967140677313e-05, "grad_norm": 1.4594573974609375, "loss_scale": 4.0, "retries": 0, "seconds": 71.1, "max_memory_gb": 17.44, "ce": 0.0659897358273156, "kl": 0.0912122153386008, "context": 0.11159584351116791} +{"step": 163, "steps": 250, "lr": 1.8959441649037335e-05, "grad_norm": 2.075258493423462, "loss_scale": 4.0, "retries": 0, "seconds": 71.6, "max_memory_gb": 17.44, "ce": 0.21696927316952497, "kl": 0.5818815374514088, "context": 0.5079100439324975} +{"step": 164, "steps": 250, "lr": 1.868075127814832e-05, "grad_norm": 0.7456574440002441, "loss_scale": 4.0, "retries": 0, "seconds": 72.0, "max_memory_gb": 17.44, "ce": 0.08760684843582567, "kl": 0.0620378679705027, "context": 0.11862577828287613} +{"step": 165, "steps": 250, "lr": 1.8403650526453143e-05, "grad_norm": 0.17111727595329285, "loss_scale": 4.0, "retries": 0, "seconds": 72.4, "max_memory_gb": 17.44, "ce": 0.010323645605240017, "kl": 0.021768846330814995, "context": 0.02120806893799454} +{"step": 166, "steps": 250, "lr": 1.812818933977912e-05, "grad_norm": 1.9159077405929565, "loss_scale": 4.0, "retries": 0, "seconds": 72.9, "max_memory_gb": 17.44, "ce": 0.41746362240519375, "kl": 0.09069868677761406, "context": 0.46281296154484153} +{"step": 167, "steps": 250, "lr": 1.785441736843128e-05, "grad_norm": 0.19487303495407104, "loss_scale": 4.0, "retries": 0, "seconds": 73.3, "max_memory_gb": 17.44, "ce": 0.02524505346082151, "kl": 0.006156618008390069, "context": 0.028323363047093153} +{"step": 168, "steps": 250, "lr": 1.7582383958243314e-05, "grad_norm": 0.180902361869812, "loss_scale": 4.0, "retries": 0, "seconds": 73.7, "max_memory_gb": 17.44, "ce": 0.03457109333248809, "kl": 0.004197972040856257, "context": 0.036670078116003424} +{"step": 169, "steps": 250, "lr": 1.7312138141683164e-05, "grad_norm": 1.248425006866455, "loss_scale": 4.0, "retries": 0, "seconds": 74.2, "max_memory_gb": 17.44, "ce": 0.06932701205369085, "kl": 1.1409137174487114, "context": 0.6397838816046715} +{"step": 170, "steps": 250, "lr": 1.704372862901521e-05, "grad_norm": 0.042603496462106705, "loss_scale": 4.0, "retries": 0, "seconds": 74.6, "max_memory_gb": 17.44, "ce": 0.0030363374098669738, "kl": 0.0006252676330404938, "context": 0.0033489711640868336} +{"step": 171, "steps": 250, "lr": 1.6777203799520573e-05, "grad_norm": 0.7066062092781067, "loss_scale": 4.0, "retries": 0, "seconds": 75.0, "max_memory_gb": 17.44, "ce": 0.03169715098920278, "kl": 0.23517959972377867, "context": 0.14928695454727858} +{"step": 172, "steps": 250, "lr": 1.6512611692777008e-05, "grad_norm": 0.17722022533416748, "loss_scale": 4.0, "retries": 0, "seconds": 75.5, "max_memory_gb": 17.44, "ce": 0.018491230788640678, "kl": 0.01706780877429992, "context": 0.02702513540862128} +{"step": 173, "steps": 250, "lr": 1.6250000000000005e-05, "grad_norm": 0.2589870095252991, "loss_scale": 4.0, "retries": 0, "seconds": 75.9, "max_memory_gb": 17.44, "ce": 0.04960518382722512, "kl": 0.01926867212750949, "context": 0.05923952045850456} +{"step": 174, "steps": 250, "lr": 1.598941605544678e-05, "grad_norm": 0.2125617116689682, "loss_scale": 4.0, "retries": 0, "seconds": 76.3, "max_memory_gb": 17.44, "ce": 0.033515953211463057, "kl": 0.020001492689516454, "context": 0.04351669974857941} +{"step": 175, "steps": 250, "lr": 1.5730906827884506e-05, "grad_norm": 1.4802006483078003, "loss_scale": 4.0, "retries": 0, "seconds": 76.8, "max_memory_gb": 17.44, "ce": 0.12013053707778454, "kl": 0.07493369374424219, "context": 0.15759738441556692} +{"step": 176, "steps": 250, "lr": 1.5474518912124476e-05, "grad_norm": 0.08370137959718704, "loss_scale": 4.0, "retries": 0, "seconds": 77.2, "max_memory_gb": 17.44, "ce": 0.010495648690266535, "kl": 0.00220549815276172, "context": 0.011598397832131013} +{"step": 177, "steps": 250, "lr": 1.5220298520623646e-05, "grad_norm": 0.7317657470703125, "loss_scale": 4.0, "retries": 0, "seconds": 77.6, "max_memory_gb": 17.44, "ce": 0.03860377741511911, "kl": 0.5573228736539022, "context": 0.317265212390339} +{"step": 178, "steps": 250, "lr": 1.4968291475155121e-05, "grad_norm": 0.527668833732605, "loss_scale": 4.0, "retries": 0, "seconds": 78.0, "max_memory_gb": 17.44, "ce": 0.06103824399178848, "kl": 0.04334524666774087, "context": 0.0827108679804951} +{"step": 179, "steps": 250, "lr": 1.4718543198548998e-05, "grad_norm": 1.0000944137573242, "loss_scale": 4.0, "retries": 0, "seconds": 78.4, "max_memory_gb": 17.44, "ce": 0.41815048691933043, "kl": 0.0131310685792414, "context": 0.4247160154336598} +{"step": 180, "steps": 250, "lr": 1.4471098706505227e-05, "grad_norm": 0.5152769684791565, "loss_scale": 4.0, "retries": 0, "seconds": 78.9, "max_memory_gb": 17.44, "ce": 0.052722403954248875, "kl": 0.03363932605134323, "context": 0.06954206674708985} +{"step": 181, "steps": 250, "lr": 1.4226002599479743e-05, "grad_norm": 0.28051501512527466, "loss_scale": 4.0, "retries": 0, "seconds": 79.3, "max_memory_gb": 17.44, "ce": 0.00949212076375261, "kl": 0.12885896637453698, "context": 0.0739216028014198} +{"step": 182, "steps": 250, "lr": 1.3983299054645471e-05, "grad_norm": 0.5531642436981201, "loss_scale": 4.0, "retries": 0, "seconds": 79.7, "max_memory_gb": 17.44, "ce": 0.03319926781114191, "kl": 0.2634169900629786, "context": 0.1649077633046545} +{"step": 183, "steps": 250, "lr": 1.3743031817929724e-05, "grad_norm": 0.05661175027489662, "loss_scale": 4.0, "retries": 0, "seconds": 80.2, "max_memory_gb": 17.44, "ce": 0.005179335435968824, "kl": 0.0011161453257955145, "context": 0.005737408238928765} +{"step": 184, "steps": 250, "lr": 1.3505244196129164e-05, "grad_norm": 0.07351866364479065, "loss_scale": 4.0, "retries": 0, "seconds": 80.6, "max_memory_gb": 17.44, "ce": 0.004764364304719493, "kl": 0.005951668354100548, "context": 0.007740198343526572} +{"step": 185, "steps": 250, "lr": 1.3269979049104014e-05, "grad_norm": 0.18691422045230865, "loss_scale": 4.0, "retries": 0, "seconds": 81.0, "max_memory_gb": 17.44, "ce": 0.015058443532325327, "kl": 0.07531367841147585, "context": 0.052715283760335296} +{"step": 186, "steps": 250, "lr": 1.3037278782052863e-05, "grad_norm": 1.520707130432129, "loss_scale": 4.0, "retries": 0, "seconds": 81.5, "max_memory_gb": 17.44, "ce": 0.07788864079338964, "kl": 0.1503932776322472, "context": 0.15308527827437501} +{"step": 187, "steps": 250, "lr": 1.2807185337869343e-05, "grad_norm": 0.41876477003097534, "loss_scale": 4.0, "retries": 0, "seconds": 81.9, "max_memory_gb": 17.44, "ce": 0.06944575277157128, "kl": 0.4812805590336211, "context": 0.31008602562360466} +{"step": 188, "steps": 250, "lr": 1.2579740189582115e-05, "grad_norm": 1.203551173210144, "loss_scale": 4.0, "retries": 0, "seconds": 82.3, "max_memory_gb": 17.44, "ce": 0.2245698753540637, "kl": 0.01599428779900336, "context": 0.23256702601793222} +{"step": 189, "steps": 250, "lr": 1.2354984332879668e-05, "grad_norm": 1.8283036947250366, "loss_scale": 4.0, "retries": 0, "seconds": 82.8, "max_memory_gb": 17.44, "ce": 0.2637939705164172, "kl": 0.27883846129407175, "context": 0.4032132007414475} +{"step": 190, "steps": 250, "lr": 1.2132958278721064e-05, "grad_norm": 0.11457131057977676, "loss_scale": 4.0, "retries": 0, "seconds": 83.2, "max_memory_gb": 17.44, "ce": 0.005814751726575196, "kl": 0.027241703821346164, "context": 0.01943560387007892} +{"step": 191, "steps": 250, "lr": 1.1913702046034016e-05, "grad_norm": 2.0050671100616455, "loss_scale": 4.0, "retries": 0, "seconds": 83.6, "max_memory_gb": 17.44, "ce": 0.2972669439332094, "kl": 0.21998150852596154, "context": 0.4072576828766614} +{"step": 192, "steps": 250, "lr": 1.1697255154501756e-05, "grad_norm": 0.9663605093955994, "loss_scale": 4.0, "retries": 0, "seconds": 84.1, "max_memory_gb": 17.44, "ce": 0.4491837488603778, "kl": 0.07109053510066587, "context": 0.48472900525666773} +{"step": 193, "steps": 250, "lr": 1.1483656617439842e-05, "grad_norm": 0.0800260603427887, "loss_scale": 4.0, "retries": 0, "seconds": 84.5, "max_memory_gb": 17.44, "ce": 0.0074702304700622335, "kl": 0.0029333267848414835, "context": 0.008936894009821117} +{"step": 194, "steps": 250, "lr": 1.1272944934764177e-05, "grad_norm": 0.5466418266296387, "loss_scale": 4.0, "retries": 0, "seconds": 84.9, "max_memory_gb": 17.44, "ce": 0.0650794118992053, "kl": 0.06081318186079443, "context": 0.0954860026249662} +{"step": 195, "steps": 250, "lr": 1.1065158086051684e-05, "grad_norm": 0.1674991399049759, "loss_scale": 4.0, "retries": 0, "seconds": 85.4, "max_memory_gb": 17.44, "ce": 0.016751740564359352, "kl": 0.002962971462693531, "context": 0.01823322643758729} +{"step": 196, "steps": 250, "lr": 1.086033352369467e-05, "grad_norm": 0.6412990093231201, "loss_scale": 4.0, "retries": 0, "seconds": 85.8, "max_memory_gb": 17.44, "ce": 0.0035091330792056397, "kl": 0.4053885666737642, "context": 0.20620341948233545} +{"step": 197, "steps": 250, "lr": 1.0658508166150224e-05, "grad_norm": 1.885278582572937, "loss_scale": 4.0, "retries": 0, "seconds": 86.2, "max_memory_gb": 17.44, "ce": 0.05178657631040551, "kl": 0.6882797123917044, "context": 0.3959264423174318} +{"step": 198, "steps": 250, "lr": 1.0459718391285954e-05, "grad_norm": 0.2917458117008209, "loss_scale": 4.0, "retries": 0, "seconds": 86.7, "max_memory_gb": 17.44, "ce": 0.06871766920085065, "kl": 0.0058837608758040005, "context": 0.07165955239906907} +{"step": 199, "steps": 250, "lr": 1.0264000029822999e-05, "grad_norm": 0.5847971439361572, "loss_scale": 4.0, "retries": 0, "seconds": 87.1, "max_memory_gb": 17.44, "ce": 0.1288332884432748, "kl": 0.09973152847669553, "context": 0.17869905184488744} +{"step": 200, "steps": 250, "lr": 1.0071388358877775e-05, "grad_norm": 0.7401729226112366, "loss_scale": 4.0, "retries": 0, "seconds": 87.5, "max_memory_gb": 17.44, "ce": 0.25229497451800853, "kl": 0.0337247340939939, "context": 0.26915733818896115} +{"step": 201, "steps": 250, "lr": 9.881918095603501e-06, "grad_norm": 0.7894347906112671, "loss_scale": 4.0, "retries": 0, "seconds": 88.0, "max_memory_gb": 17.44, "ce": 0.07442833195091225, "kl": 0.10747541533783078, "context": 0.1281660395907238} +{"step": 202, "steps": 250, "lr": 9.695623390932612e-06, "grad_norm": 0.04209136217832565, "loss_scale": 4.0, "retries": 0, "seconds": 88.4, "max_memory_gb": 17.44, "ce": 0.003090486214205157, "kl": 0.00444782794738785, "context": 0.005314400375937112} +{"step": 203, "steps": 250, "lr": 9.51253782342123e-06, "grad_norm": 0.2548060715198517, "loss_scale": 4.0, "retries": 0, "seconds": 88.8, "max_memory_gb": 17.44, "ce": 0.01634390295657795, "kl": 0.018208438530564308, "context": 0.025448121887166053} +{"step": 204, "steps": 250, "lr": 9.332694393196846e-06, "grad_norm": 0.2685844302177429, "loss_scale": 4.0, "retries": 0, "seconds": 89.2, "max_memory_gb": 17.44, "ce": 0.044059862673748285, "kl": 0.010874023078940809, "context": 0.04949687194311991} +{"step": 205, "steps": 250, "lr": 9.156125516010256e-06, "grad_norm": 0.5298177599906921, "loss_scale": 4.0, "retries": 0, "seconds": 89.7, "max_memory_gb": 17.44, "ce": 0.005732922232709825, "kl": 0.48565543360018637, "context": 0.24856063607148826} +{"step": 206, "steps": 250, "lr": 8.982863017392733e-06, "grad_norm": 0.14007796347141266, "loss_scale": 4.0, "retries": 0, "seconds": 90.1, "max_memory_gb": 17.44, "ce": 0.012079191968950909, "kl": 0.003837137782738864, "context": 0.01399776036851108} +{"step": 207, "steps": 250, "lr": 8.81293812691974e-06, "grad_norm": 0.27375704050064087, "loss_scale": 4.0, "retries": 0, "seconds": 90.5, "max_memory_gb": 17.44, "ce": 0.076754204579629, "kl": 0.03649671084713191, "context": 0.0950025599449873} +{"step": 208, "steps": 250, "lr": 8.646381472581927e-06, "grad_norm": 0.060229431837797165, "loss_scale": 4.0, "retries": 0, "seconds": 91.0, "max_memory_gb": 17.44, "ce": 0.006746736151399091, "kl": 0.0025243611744372174, "context": 0.008008916571270674} +{"step": 209, "steps": 250, "lr": 8.483223075264622e-06, "grad_norm": 1.1093343496322632, "loss_scale": 4.0, "retries": 0, "seconds": 91.4, "max_memory_gb": 17.44, "ce": 0.20636107865720987, "kl": 0.08581813272758154, "context": 0.24927014153217897} +{"step": 210, "steps": 250, "lr": 8.323492343336747e-06, "grad_norm": 0.9411027431488037, "loss_scale": 4.0, "retries": 0, "seconds": 91.8, "max_memory_gb": 17.44, "ce": 0.4690406000299845, "kl": 0.13642690098640742, "context": 0.5372540621319786} +{"step": 211, "steps": 250, "lr": 8.167218067350122e-06, "grad_norm": 0.5314802527427673, "loss_scale": 4.0, "retries": 0, "seconds": 92.3, "max_memory_gb": 17.44, "ce": 0.173253926448524, "kl": 0.06779424662818201, "context": 0.20715105452109128} +{"step": 212, "steps": 250, "lr": 8.01442841485013e-06, "grad_norm": 1.8546481132507324, "loss_scale": 4.0, "retries": 0, "seconds": 92.7, "max_memory_gb": 17.44, "ce": 0.09362812910694629, "kl": 0.9933095221058466, "context": 0.5902828923426569} +{"step": 213, "steps": 250, "lr": 7.865150925298678e-06, "grad_norm": 0.720806896686554, "loss_scale": 4.0, "retries": 0, "seconds": 93.1, "max_memory_gb": 17.44, "ce": 0.2395758143684361, "kl": 0.043788866147224326, "context": 0.2614702539285645} +{"step": 214, "steps": 250, "lr": 7.719412505110375e-06, "grad_norm": 1.0639702081680298, "loss_scale": 4.0, "retries": 0, "seconds": 93.6, "max_memory_gb": 17.44, "ce": 0.41170463801245205, "kl": 0.12135869672056288, "context": 0.4723839989164844} +{"step": 215, "steps": 250, "lr": 7.57723942280278e-06, "grad_norm": 1.7505913972854614, "loss_scale": 4.0, "retries": 0, "seconds": 94.0, "max_memory_gb": 17.44, "ce": 0.5811073939257767, "kl": 0.025278120606230914, "context": 0.5937464571325108} +{"step": 216, "steps": 250, "lr": 7.438657304261687e-06, "grad_norm": 0.2282540649175644, "loss_scale": 4.0, "retries": 0, "seconds": 94.5, "max_memory_gb": 17.44, "ce": 0.01245240870048292, "kl": 0.08342059279311798, "context": 0.054162703279871494} +{"step": 217, "steps": 250, "lr": 7.30369112812219e-06, "grad_norm": 0.05403276905417442, "loss_scale": 4.0, "retries": 0, "seconds": 94.9, "max_memory_gb": 17.44, "ce": 0.004129835491767153, "kl": 0.003164582871249877, "context": 0.005712127021979541} +{"step": 218, "steps": 250, "lr": 7.172365221266403e-06, "grad_norm": 0.6711013317108154, "loss_scale": 4.0, "retries": 0, "seconds": 95.3, "max_memory_gb": 17.44, "ce": 0.1758000128902495, "kl": 0.2621798515319824, "context": 0.30688993260264397} +{"step": 219, "steps": 250, "lr": 7.0447032544387154e-06, "grad_norm": 0.3883483111858368, "loss_scale": 4.0, "retries": 0, "seconds": 95.8, "max_memory_gb": 17.44, "ce": 0.013117191774654202, "kl": 0.11797631619265303, "context": 0.0721053495653905} +{"step": 220, "steps": 250, "lr": 6.920728237979274e-06, "grad_norm": 0.27148377895355225, "loss_scale": 4.0, "retries": 0, "seconds": 96.2, "max_memory_gb": 17.44, "ce": 0.05005647960933857, "kl": 0.0045545400062110275, "context": 0.0523337482009083} +{"step": 221, "steps": 250, "lr": 6.800462517676456e-06, "grad_norm": 0.5294294953346252, "loss_scale": 4.0, "retries": 0, "seconds": 96.6, "max_memory_gb": 17.44, "ce": 0.006631538533838466, "kl": 0.4163482530420879, "context": 0.21480565797537565} +{"step": 222, "steps": 250, "lr": 6.683927770739222e-06, "grad_norm": 0.5469995141029358, "loss_scale": 4.0, "retries": 0, "seconds": 97.0, "max_memory_gb": 17.44, "ce": 0.05788625404238701, "kl": 0.05812537183828681, "context": 0.08694894163636491} +{"step": 223, "steps": 250, "lr": 6.571145001889892e-06, "grad_norm": 3.687875747680664, "loss_scale": 4.0, "retries": 0, "seconds": 97.5, "max_memory_gb": 17.44, "ce": 0.24216359097044915, "kl": 0.024172105277102673, "context": 0.2542496373062022} +{"step": 224, "steps": 250, "lr": 6.46213453957817e-06, "grad_norm": 1.4247239828109741, "loss_scale": 4.0, "retries": 0, "seconds": 97.9, "max_memory_gb": 17.44, "ce": 0.069984003494028, "kl": 0.16778120887101977, "context": 0.15387460181955248} +{"step": 225, "steps": 250, "lr": 6.356916032317063e-06, "grad_norm": 0.36880454421043396, "loss_scale": 4.0, "retries": 0, "seconds": 98.3, "max_memory_gb": 17.44, "ce": 0.006192745349835604, "kl": 0.14612657608449808, "context": 0.07925603384501301} +{"step": 226, "steps": 250, "lr": 6.255508445141349e-06, "grad_norm": 0.6794047355651855, "loss_scale": 4.0, "retries": 0, "seconds": 98.8, "max_memory_gb": 17.44, "ce": 0.0391638966393657, "kl": 0.028843454289017245, "context": 0.053585624729748815} +{"step": 227, "steps": 250, "lr": 6.157930056189228e-06, "grad_norm": 0.3253277540206909, "loss_scale": 4.0, "retries": 0, "seconds": 99.2, "max_memory_gb": 17.44, "ce": 0.032869173170183785, "kl": 0.013139800947783442, "context": 0.03943907179927919} +{"step": 228, "steps": 250, "lr": 6.064198453407796e-06, "grad_norm": 0.5358420014381409, "loss_scale": 4.0, "retries": 0, "seconds": 99.6, "max_memory_gb": 17.44, "ce": 0.009110593920922838, "kl": 0.21209347067633644, "context": 0.11515733192209154} +{"step": 229, "steps": 250, "lr": 5.974330531382935e-06, "grad_norm": 1.7915294170379639, "loss_scale": 4.0, "retries": 0, "seconds": 100.1, "max_memory_gb": 17.44, "ce": 0.10304591790190898, "kl": 0.06915788346350382, "context": 0.13762486216728576} +{"step": 230, "steps": 250, "lr": 5.888342488294125e-06, "grad_norm": 1.824634075164795, "loss_scale": 4.0, "retries": 0, "seconds": 100.5, "max_memory_gb": 17.44, "ce": 0.5261201687389985, "kl": 0.28143306635320187, "context": 0.6668367111124098} +{"step": 231, "steps": 250, "lr": 5.806249822994849e-06, "grad_norm": 1.1312415599822998, "loss_scale": 4.0, "retries": 0, "seconds": 100.9, "max_memory_gb": 17.44, "ce": 0.0698705087415874, "kl": 0.1662152048374992, "context": 0.15297811396885663} +{"step": 232, "steps": 250, "lr": 5.728067332218994e-06, "grad_norm": 0.2568143308162689, "loss_scale": 4.0, "retries": 0, "seconds": 101.3, "max_memory_gb": 17.44, "ce": 0.03950274342787452, "kl": 0.007589619708596729, "context": 0.04329755186336115} +{"step": 233, "steps": 250, "lr": 5.65380910791383e-06, "grad_norm": 0.3629458248615265, "loss_scale": 4.0, "retries": 0, "seconds": 101.8, "max_memory_gb": 17.44, "ce": 0.01434144377708435, "kl": 0.09552034339867532, "context": 0.062101616989821196} +{"step": 234, "steps": 250, "lr": 5.583488534700022e-06, "grad_norm": 0.22956934571266174, "loss_scale": 4.0, "retries": 0, "seconds": 102.2, "max_memory_gb": 17.44, "ce": 0.01293437328422442, "kl": 0.012563310730911326, "context": 0.01921602856600657} +{"step": 235, "steps": 250, "lr": 5.517118287459118e-06, "grad_norm": 1.6706289052963257, "loss_scale": 4.0, "retries": 0, "seconds": 102.6, "max_memory_gb": 17.44, "ce": 0.3650902077788487, "kl": 0.021604967630992178, "context": 0.3758926876762416} +{"step": 236, "steps": 250, "lr": 5.454710329048978e-06, "grad_norm": 0.13671788573265076, "loss_scale": 4.0, "retries": 0, "seconds": 103.1, "max_memory_gb": 17.44, "ce": 0.013609801622806117, "kl": 0.020432003191672266, "context": 0.02382580313133076} +{"step": 237, "steps": 250, "lr": 5.3962759081475436e-06, "grad_norm": 0.3887299597263336, "loss_scale": 4.0, "retries": 0, "seconds": 103.5, "max_memory_gb": 17.44, "ce": 0.08051875146338716, "kl": 0.024658569807286312, "context": 0.09284803568152711} +{"step": 238, "steps": 250, "lr": 5.341825557225321e-06, "grad_norm": 0.22685550153255463, "loss_scale": 4.0, "retries": 0, "seconds": 103.9, "max_memory_gb": 17.44, "ce": 0.0278759922075551, "kl": 0.006583646181752556, "context": 0.03116781475546304} +{"step": 239, "steps": 250, "lr": 5.291369090646961e-06, "grad_norm": 0.3455101251602173, "loss_scale": 4.0, "retries": 0, "seconds": 104.3, "max_memory_gb": 17.44, "ce": 0.123907262342982, "kl": 0.02159249763644766, "context": 0.1347035076469183} +{"step": 240, "steps": 250, "lr": 5.244915602902289e-06, "grad_norm": 0.4530338943004608, "loss_scale": 4.0, "retries": 0, "seconds": 104.8, "max_memory_gb": 17.44, "ce": 0.10762354370672256, "kl": 0.02973196018137969, "context": 0.12248952657682821} +{"step": 241, "steps": 250, "lr": 5.20247346696706e-06, "grad_norm": 1.423787236213684, "loss_scale": 4.0, "retries": 0, "seconds": 105.2, "max_memory_gb": 17.44, "ce": 0.22502508186153136, "kl": 0.03015421591408085, "context": 0.24010218895273283} +{"step": 242, "steps": 250, "lr": 5.164050332793787e-06, "grad_norm": 0.15139326453208923, "loss_scale": 4.0, "retries": 0, "seconds": 105.6, "max_memory_gb": 17.44, "ce": 0.009298542921897024, "kl": 0.04830929993204336, "context": 0.033453192620072514} +{"step": 243, "steps": 250, "lr": 5.129653125932873e-06, "grad_norm": 0.1091562807559967, "loss_scale": 4.0, "retries": 0, "seconds": 106.1, "max_memory_gb": 17.44, "ce": 0.00955700408667326, "kl": 0.005369872262235731, "context": 0.012241940246894956} +{"step": 244, "steps": 250, "lr": 5.099288046284329e-06, "grad_norm": 0.5413085222244263, "loss_scale": 4.0, "retries": 0, "seconds": 106.5, "max_memory_gb": 17.44, "ce": 0.060302945305011235, "kl": 0.14859933517072932, "context": 0.1346026128885569} +{"step": 245, "steps": 250, "lr": 5.072960566980277e-06, "grad_norm": 1.0055218935012817, "loss_scale": 4.0, "retries": 0, "seconds": 106.9, "max_memory_gb": 17.44, "ce": 0.3816550327464938, "kl": 0.07939321978483349, "context": 0.4213516318704933} +{"step": 246, "steps": 250, "lr": 5.050675433398435e-06, "grad_norm": 0.026745416224002838, "loss_scale": 4.0, "retries": 0, "seconds": 107.4, "max_memory_gb": 17.44, "ce": 0.0022504933731397614, "kl": 0.000935023035708582, "context": 0.002718004834605381} +{"step": 247, "steps": 250, "lr": 5.032436662306811e-06, "grad_norm": 0.5240446925163269, "loss_scale": 4.0, "retries": 0, "seconds": 107.8, "max_memory_gb": 17.44, "ce": 0.009204930916894227, "kl": 0.46004432132758666, "context": 0.23922708712052554} +{"step": 248, "steps": 250, "lr": 5.018247541139684e-06, "grad_norm": 0.2734729051589966, "loss_scale": 4.0, "retries": 0, "seconds": 108.2, "max_memory_gb": 17.44, "ce": 0.03674278501421213, "kl": 0.015257685910910368, "context": 0.04437162773683667} +{"step": 249, "steps": 250, "lr": 5.00811062740508e-06, "grad_norm": 0.5165212154388428, "loss_scale": 4.0, "retries": 0, "seconds": 108.6, "max_memory_gb": 17.44, "ce": 0.061084243236109614, "kl": 0.059503668162506074, "context": 0.09083607792854309} +{"step": 250, "steps": 250, "lr": 5.002027748223778e-06, "grad_norm": 0.09728644043207169, "loss_scale": 4.0, "retries": 0, "seconds": 109.1, "max_memory_gb": 17.44, "ce": 0.010132987168617547, "kl": 0.004580398577672895, "context": 0.012423186388332397} diff --git a/runs/single_internvl/context/L25_s13/adapter/README.md b/runs/single_internvl/context/L25_s13/adapter/README.md new file mode 100644 index 0000000000000000000000000000000000000000..4f9f856cf0e1531b47e25d266833b2977a19d3ac --- /dev/null +++ b/runs/single_internvl/context/L25_s13/adapter/README.md @@ -0,0 +1,207 @@ +--- +base_model: /data/OPSD/assets/models/InternVL3_5-8B-HF +library_name: peft +pipeline_tag: text-generation +tags: +- base_model:adapter:/data/OPSD/assets/models/InternVL3_5-8B-HF +- lora +- transformers +--- + +# Model Card for Model ID + + + + + +## Model Details + +### Model Description + + + + + +- **Developed by:** [More Information Needed] +- **Funded by [optional]:** [More Information Needed] +- **Shared by [optional]:** [More Information Needed] +- **Model type:** [More Information Needed] +- **Language(s) (NLP):** [More Information Needed] +- **License:** [More Information Needed] +- **Finetuned from model [optional]:** [More Information Needed] + +### Model Sources [optional] + + + +- **Repository:** [More Information Needed] +- **Paper [optional]:** [More Information Needed] +- **Demo [optional]:** [More Information Needed] + +## Uses + + + +### Direct Use + + + +[More Information Needed] + +### Downstream Use [optional] + + + +[More Information Needed] + +### Out-of-Scope Use + + + +[More Information Needed] + +## Bias, Risks, and Limitations + + + +[More Information Needed] + +### Recommendations + + + +Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. More information needed for further recommendations. + +## How to Get Started with the Model + +Use the code below to get started with the model. + +[More Information Needed] + +## Training Details + +### Training Data + + + +[More Information Needed] + +### Training Procedure + + + +#### Preprocessing [optional] + +[More Information Needed] + + +#### Training Hyperparameters + +- **Training regime:** [More Information Needed] + +#### Speeds, Sizes, Times [optional] + + + +[More Information Needed] + +## Evaluation + + + +### Testing Data, Factors & Metrics + +#### Testing Data + + + +[More Information Needed] + +#### Factors + + + +[More Information Needed] + +#### Metrics + + + +[More Information Needed] + +### Results + +[More Information Needed] + +#### Summary + + + +## Model Examination [optional] + + + +[More Information Needed] + +## Environmental Impact + + + +Carbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700). + +- **Hardware Type:** [More Information Needed] +- **Hours used:** [More Information Needed] +- **Cloud Provider:** [More Information Needed] +- **Compute Region:** [More Information Needed] +- **Carbon Emitted:** [More Information Needed] + +## Technical Specifications [optional] + +### Model Architecture and Objective + +[More Information Needed] + +### Compute Infrastructure + +[More Information Needed] + +#### Hardware + +[More Information Needed] + +#### Software + +[More Information Needed] + +## Citation [optional] + + + +**BibTeX:** + +[More Information Needed] + +**APA:** + +[More Information Needed] + +## Glossary [optional] + + + +[More Information Needed] + +## More Information [optional] + +[More Information Needed] + +## Model Card Authors [optional] + +[More Information Needed] + +## Model Card Contact + +[More Information Needed] +### Framework versions + +- PEFT 0.20.0 \ No newline at end of file diff --git a/runs/single_internvl/context/L25_s13/adapter/adapter_config.json b/runs/single_internvl/context/L25_s13/adapter/adapter_config.json new file mode 100644 index 0000000000000000000000000000000000000000..f495b3bed9ab3161fa2c1980a8a0013264fa6b34 --- /dev/null +++ b/runs/single_internvl/context/L25_s13/adapter/adapter_config.json @@ -0,0 +1,50 @@ +{ + "alora_invocation_tokens": null, + "alpha_pattern": {}, + "arrow_config": null, + "auto_mapping": null, + "base_model_name_or_path": "/data/OPSD/assets/models/InternVL3_5-8B-HF", + "bias": "none", + "corda_config": null, + "ensure_weight_tying": false, + "eva_config": null, + "exclude_modules": null, + "fan_in_fan_out": false, + "inference_mode": true, + "init_lora_weights": true, + "layer_replication": null, + "layers_pattern": null, + "layers_to_transform": null, + "loftq_config": {}, + "lora_alpha": 32, + "lora_bias": false, + "lora_dropout": 0.05, + "lora_ga_config": null, + "megatron_config": null, + "megatron_core": "megatron.core", + "modules_to_save": null, + "monteclora_config": null, + "peft_type": "LORA", + "peft_version": "0.20.0", + "qalora_group_size": 16, + "r": 16, + "rank_pattern": {}, + "revision": null, + "target_modules": [ + "model.language_model.layers.25.self_attn.k_proj", + "model.language_model.layers.25.self_attn.v_proj", + "model.language_model.layers.25.self_attn.o_proj", + "model.language_model.layers.25.self_attn.q_proj", + "model.language_model.layers.25.mlp.gate_proj", + "model.language_model.layers.25.mlp.up_proj", + "model.language_model.layers.25.mlp.down_proj" + ], + "target_parameters": null, + "task_type": "CAUSAL_LM", + "trainable_token_indices": null, + "use_bdlora": null, + "use_dora": false, + "use_qalora": false, + "use_rslora": false, + "velora_config": null +} \ No newline at end of file diff --git a/runs/single_internvl/context/L25_s13/config.json b/runs/single_internvl/context/L25_s13/config.json new file mode 100644 index 0000000000000000000000000000000000000000..8e8da27f10e3d3efe46003c66c5813224a9b76a2 --- /dev/null +++ b/runs/single_internvl/context/L25_s13/config.json @@ -0,0 +1,24 @@ +{ + "behavior": "context", + "seed": 13, + "model": "/data/OPSD/assets/models/InternVL3_5-8B-HF", + "backbone": "internvl", + "train": "/data/OPSD/assets/v3/context.jsonl", + "items": 1000, + "steps": 250, + "epochs": 1, + "steps_per_epoch": 250, + "groups_per_step": 4, + "lr": 5e-05, + "scheduler": "cosine", + "warmup_steps": 16, + "final_fraction": 0.1, + "weight_decay": 0.0, + "clip": 1.0, + "initial_loss_scale": 4.0, + "adapter": { + "layer": 25, + "rank": 16, + "trainable_parameters": 1212416 + } +} \ No newline at end of file diff --git a/runs/single_internvl/context/L25_s13/steps.jsonl b/runs/single_internvl/context/L25_s13/steps.jsonl new file mode 100644 index 0000000000000000000000000000000000000000..bc7ebc8f9d688514a6e3f402c55b41673c8b2479 --- /dev/null +++ b/runs/single_internvl/context/L25_s13/steps.jsonl @@ -0,0 +1,250 @@ +{"step": 1, "steps": 250, "lr": 3.125e-06, "grad_norm": 0.5883803963661194, "loss_scale": 4.0, "retries": 0, "seconds": 1.4, "max_memory_gb": 17.01, "ce": 0.5227766726748087, "kl": 0.08654405153356493, "context": 0.5660487064160407} +{"step": 2, "steps": 250, "lr": 6.25e-06, "grad_norm": 0.6018528938293457, "loss_scale": 4.0, "retries": 0, "seconds": 1.7, "max_memory_gb": 17.02, "ce": 0.6404333652244532, "kl": 0.4672744294241511, "context": 0.8740705911695841} +{"step": 3, "steps": 250, "lr": 9.375000000000001e-06, "grad_norm": 2.07371187210083, "loss_scale": 4.0, "retries": 0, "seconds": 2.2, "max_memory_gb": 17.02, "ce": 1.0982282981276512, "kl": 0.06708946172147989, "context": 1.1317730098962784} +{"step": 4, "steps": 250, "lr": 1.25e-05, "grad_norm": 1.2216459512710571, "loss_scale": 4.0, "retries": 0, "seconds": 2.6, "max_memory_gb": 17.02, "ce": 0.4485335658682743, "kl": 0.4298224066997136, "context": 0.6634447768738028} +{"step": 5, "steps": 250, "lr": 1.5625e-05, "grad_norm": 0.012583052739501, "loss_scale": 4.0, "retries": 0, "seconds": 3.1, "max_memory_gb": 17.02, "ce": 0.0091884375524387, "kl": 0.008232076364947716, "context": 0.013304475676704897} +{"step": 6, "steps": 250, "lr": 1.8750000000000002e-05, "grad_norm": 0.5756418704986572, "loss_scale": 4.0, "retries": 0, "seconds": 3.5, "max_memory_gb": 17.02, "ce": 0.28596588954678737, "kl": 0.07419521610427182, "context": 0.3230635010986589} +{"step": 7, "steps": 250, "lr": 2.1875e-05, "grad_norm": 1.8725969791412354, "loss_scale": 4.0, "retries": 0, "seconds": 4.0, "max_memory_gb": 17.02, "ce": 0.8147359094618878, "kl": 0.12323955487056537, "context": 0.8763556851517933} +{"step": 8, "steps": 250, "lr": 2.5e-05, "grad_norm": 2.097053289413452, "loss_scale": 4.0, "retries": 0, "seconds": 4.4, "max_memory_gb": 17.03, "ce": 0.7779182773083448, "kl": 0.2881745472550392, "context": 0.9220055602490902} +{"step": 9, "steps": 250, "lr": 2.8125000000000003e-05, "grad_norm": 1.5860474109649658, "loss_scale": 4.0, "retries": 0, "seconds": 4.9, "max_memory_gb": 17.03, "ce": 0.5088652726262808, "kl": 0.13699499261565506, "context": 0.5773627609014511} +{"step": 10, "steps": 250, "lr": 3.125e-05, "grad_norm": 0.27133557200431824, "loss_scale": 4.0, "retries": 0, "seconds": 5.3, "max_memory_gb": 17.03, "ce": 0.049309223009913694, "kl": 0.04537611495470628, "context": 0.07199727860279381} +{"step": 11, "steps": 250, "lr": 3.4375e-05, "grad_norm": 1.7590290307998657, "loss_scale": 4.0, "retries": 0, "seconds": 5.7, "max_memory_gb": 17.03, "ce": 0.6654312550235773, "kl": 0.0050657084905196825, "context": 0.6679641216105665} +{"step": 12, "steps": 250, "lr": 3.7500000000000003e-05, "grad_norm": 1.1741905212402344, "loss_scale": 4.0, "retries": 0, "seconds": 6.1, "max_memory_gb": 17.03, "ce": 0.44441517303948785, "kl": 0.002710095401653234, "context": 0.44577022876546835} +{"step": 13, "steps": 250, "lr": 4.0625000000000005e-05, "grad_norm": 1.145578145980835, "loss_scale": 4.0, "retries": 0, "seconds": 6.5, "max_memory_gb": 17.03, "ce": 0.39925142271385994, "kl": 0.0187787204158667, "context": 0.4086407827453513} +{"step": 14, "steps": 250, "lr": 4.375e-05, "grad_norm": 2.44775652885437, "loss_scale": 4.0, "retries": 0, "seconds": 6.9, "max_memory_gb": 17.03, "ce": 0.6799478977918625, "kl": 0.9486082820221782, "context": 1.1542520076036453} +{"step": 15, "steps": 250, "lr": 4.6875e-05, "grad_norm": 2.014955520629883, "loss_scale": 4.0, "retries": 0, "seconds": 7.3, "max_memory_gb": 17.03, "ce": 0.4455733961704027, "kl": 0.4663476336850181, "context": 0.6787472111502666} +{"step": 16, "steps": 250, "lr": 5e-05, "grad_norm": 0.8571985363960266, "loss_scale": 4.0, "retries": 0, "seconds": 7.8, "max_memory_gb": 17.03, "ce": 0.17199777761561563, "kl": 0.0008986846428342687, "context": 0.17244711599778384} +{"step": 17, "steps": 250, "lr": 5e-05, "grad_norm": 0.9717519283294678, "loss_scale": 4.0, "retries": 0, "seconds": 8.2, "max_memory_gb": 17.03, "ce": 0.2769453158834949, "kl": 0.825930904422421, "context": 0.6899107478384394} +{"step": 18, "steps": 250, "lr": 4.999797225177623e-05, "grad_norm": 2.7226998805999756, "loss_scale": 4.0, "retries": 0, "seconds": 8.6, "max_memory_gb": 17.03, "ce": 0.3590149078518152, "kl": 0.21990333545545582, "context": 0.4689665929181501} +{"step": 19, "steps": 250, "lr": 4.999188937259492e-05, "grad_norm": 4.263156414031982, "loss_scale": 4.0, "retries": 0, "seconds": 9.1, "max_memory_gb": 17.03, "ce": 0.4625355862081051, "kl": 0.17147006560117006, "context": 0.5482706166803837} +{"step": 20, "steps": 250, "lr": 4.998175245886032e-05, "grad_norm": 0.03494376316666603, "loss_scale": 4.0, "retries": 0, "seconds": 9.5, "max_memory_gb": 17.03, "ce": 0.002319384506336064, "kl": 0.021213895099208457, "context": 0.012926331626658794} +{"step": 21, "steps": 250, "lr": 4.996756333769319e-05, "grad_norm": 2.152287483215332, "loss_scale": 4.0, "retries": 0, "seconds": 9.9, "max_memory_gb": 17.03, "ce": 0.18082309379860817, "kl": 0.036486873455032764, "context": 0.1990665305256698} +{"step": 22, "steps": 250, "lr": 4.994932456660157e-05, "grad_norm": 1.8844300508499146, "loss_scale": 4.0, "retries": 0, "seconds": 10.3, "max_memory_gb": 17.03, "ce": 0.18615324969869107, "kl": 0.033913985134859104, "context": 0.20311024208785966} +{"step": 23, "steps": 250, "lr": 4.992703943301973e-05, "grad_norm": 2.0362184047698975, "loss_scale": 4.0, "retries": 0, "seconds": 10.8, "max_memory_gb": 17.03, "ce": 0.35921427206176304, "kl": 0.970214748394028, "context": 0.8443216398700315} +{"step": 24, "steps": 250, "lr": 4.9900711953715675e-05, "grad_norm": 2.0983030796051025, "loss_scale": 4.0, "retries": 0, "seconds": 11.2, "max_memory_gb": 17.03, "ce": 0.09754919586703181, "kl": 0.22379334713332355, "context": 0.20944586396217346} +{"step": 25, "steps": 250, "lr": 4.987034687406713e-05, "grad_norm": 1.9635854959487915, "loss_scale": 4.0, "retries": 0, "seconds": 11.6, "max_memory_gb": 17.03, "ce": 0.42098192686898983, "kl": 0.16148384660482407, "context": 0.50172383710742} +{"step": 26, "steps": 250, "lr": 4.9835949667206215e-05, "grad_norm": 1.2059119939804077, "loss_scale": 4.0, "retries": 0, "seconds": 12.0, "max_memory_gb": 17.03, "ce": 0.0757646500132978, "kl": 0.6657166795339435, "context": 0.4086229810491204} +{"step": 27, "steps": 250, "lr": 4.9797526533032945e-05, "grad_norm": 1.5041186809539795, "loss_scale": 4.0, "retries": 0, "seconds": 12.5, "max_memory_gb": 17.03, "ce": 0.13551380793796852, "kl": 0.36740465975344705, "context": 0.31921613297163276} +{"step": 28, "steps": 250, "lr": 4.9755084397097716e-05, "grad_norm": 0.6702799797058105, "loss_scale": 4.0, "retries": 0, "seconds": 12.9, "max_memory_gb": 17.03, "ce": 0.4857026836834848, "kl": 0.9502642345614731, "context": 0.9608347956091166} +{"step": 29, "steps": 250, "lr": 4.970863090935304e-05, "grad_norm": 0.7582566738128662, "loss_scale": 4.0, "retries": 0, "seconds": 13.3, "max_memory_gb": 17.03, "ce": 0.08213966991752386, "kl": 0.37779853062238544, "context": 0.27103893365710974} +{"step": 30, "steps": 250, "lr": 4.965817444277468e-05, "grad_norm": 0.8398661613464355, "loss_scale": 4.0, "retries": 0, "seconds": 13.7, "max_memory_gb": 17.03, "ce": 0.06107084699215193, "kl": 0.002009998933317547, "context": 0.062075847299638554} +{"step": 31, "steps": 250, "lr": 4.9603724091852456e-05, "grad_norm": 0.9056372046470642, "loss_scale": 4.0, "retries": 0, "seconds": 14.2, "max_memory_gb": 17.03, "ce": 0.19285876769572496, "kl": 0.11618714593350887, "context": 0.2509523360058665} +{"step": 32, "steps": 250, "lr": 4.9545289670951025e-05, "grad_norm": 0.5165380239486694, "loss_scale": 4.0, "retries": 0, "seconds": 14.6, "max_memory_gb": 17.03, "ce": 0.018824590952135623, "kl": 0.016624662865069695, "context": 0.027136922348290682} +{"step": 33, "steps": 250, "lr": 4.948288171254089e-05, "grad_norm": 0.46021685004234314, "loss_scale": 4.0, "retries": 0, "seconds": 15.0, "max_memory_gb": 17.03, "ce": 0.053450383711606264, "kl": 0.027688548434525728, "context": 0.0672946572303772} +{"step": 34, "steps": 250, "lr": 4.941651146529998e-05, "grad_norm": 0.5405283570289612, "loss_scale": 4.0, "retries": 0, "seconds": 15.5, "max_memory_gb": 17.03, "ce": 0.0036093350317969453, "kl": 0.7894694572219123, "context": 0.39834406021509494} +{"step": 35, "steps": 250, "lr": 4.9346190892086174e-05, "grad_norm": 0.09846634417772293, "loss_scale": 4.0, "retries": 0, "seconds": 15.9, "max_memory_gb": 17.03, "ce": 0.003590363048715517, "kl": 0.0029023185270489193, "context": 0.005041522250394337} +{"step": 36, "steps": 250, "lr": 4.927193266778101e-05, "grad_norm": 0.4762556552886963, "loss_scale": 4.0, "retries": 0, "seconds": 16.4, "max_memory_gb": 17.03, "ce": 0.0904133084404748, "kl": 0.053309373288357165, "context": 0.1170679933129577} +{"step": 37, "steps": 250, "lr": 4.919375017700515e-05, "grad_norm": 0.6668826341629028, "loss_scale": 4.0, "retries": 0, "seconds": 16.8, "max_memory_gb": 17.03, "ce": 0.12812985161144752, "kl": 0.7802973578218371, "context": 0.5182785344077274} +{"step": 38, "steps": 250, "lr": 4.911165751170588e-05, "grad_norm": 0.3699702024459839, "loss_scale": 4.0, "retries": 0, "seconds": 17.3, "max_memory_gb": 17.03, "ce": 0.019534793682396412, "kl": 0.46295720801572315, "context": 0.25101339165121317} +{"step": 39, "steps": 250, "lr": 4.902566946861708e-05, "grad_norm": 0.32537129521369934, "loss_scale": 4.0, "retries": 0, "seconds": 17.7, "max_memory_gb": 17.03, "ce": 0.11225691212166566, "kl": 0.009604482431313954, "context": 0.11705915149650536} +{"step": 40, "steps": 250, "lr": 4.89358015465922e-05, "grad_norm": 0.059092652052640915, "loss_scale": 4.0, "retries": 0, "seconds": 18.2, "max_memory_gb": 17.03, "ce": 0.005978546178084798, "kl": 0.00878378565539606, "context": 0.010370439063990489} +{"step": 41, "steps": 250, "lr": 4.884206994381078e-05, "grad_norm": 0.6495232582092285, "loss_scale": 4.0, "retries": 0, "seconds": 18.6, "max_memory_gb": 17.03, "ce": 0.3648601232562214, "kl": 0.11942991754040122, "context": 0.42457508598454297} +{"step": 42, "steps": 250, "lr": 4.874449155485866e-05, "grad_norm": 0.020746495574712753, "loss_scale": 4.0, "retries": 0, "seconds": 19.1, "max_memory_gb": 17.03, "ce": 0.002792502440570388, "kl": 0.0005278027238091454, "context": 0.0030564037733711302} +{"step": 43, "steps": 250, "lr": 4.864308396768294e-05, "grad_norm": 0.022785823792219162, "loss_scale": 4.0, "retries": 0, "seconds": 19.5, "max_memory_gb": 17.03, "ce": 0.0019012618213309906, "kl": 0.0026951846230076626, "context": 0.0032488540455233306} +{"step": 44, "steps": 250, "lr": 4.853786546042184e-05, "grad_norm": 1.8987876176834106, "loss_scale": 4.0, "retries": 0, "seconds": 20.0, "max_memory_gb": 17.03, "ce": 0.5095766170343268, "kl": 1.2970087654393865, "context": 1.1580809855367988} +{"step": 45, "steps": 250, "lr": 4.842885499811012e-05, "grad_norm": 0.9252567291259766, "loss_scale": 4.0, "retries": 0, "seconds": 20.4, "max_memory_gb": 17.03, "ce": 0.8199016902362928, "kl": 0.27302640629932284, "context": 0.9564148568315431} +{"step": 46, "steps": 250, "lr": 4.831607222926078e-05, "grad_norm": 0.3026173710823059, "loss_scale": 4.0, "retries": 0, "seconds": 20.9, "max_memory_gb": 17.03, "ce": 0.07065047482319642, "kl": 0.26014358559041284, "context": 0.20072227058699355} +{"step": 47, "steps": 250, "lr": 4.8199537482323545e-05, "grad_norm": 0.05316576734185219, "loss_scale": 4.0, "retries": 0, "seconds": 21.3, "max_memory_gb": 17.03, "ce": 0.004817881068447605, "kl": 0.014133454591501504, "context": 0.011884608014952391} +{"step": 48, "steps": 250, "lr": 4.807927176202073e-05, "grad_norm": 0.0166016835719347, "loss_scale": 4.0, "retries": 0, "seconds": 21.8, "max_memory_gb": 17.03, "ce": 0.0009886067709885538, "kl": 0.002823416686624114, "context": 0.002400315148406662} +{"step": 49, "steps": 250, "lr": 4.795529674556129e-05, "grad_norm": 0.5973345041275024, "loss_scale": 4.0, "retries": 0, "seconds": 22.2, "max_memory_gb": 17.03, "ce": 0.012892144892248325, "kl": 0.3017775985586013, "context": 0.16378094900210272} +{"step": 50, "steps": 250, "lr": 4.782763477873361e-05, "grad_norm": 0.6920309066772461, "loss_scale": 4.0, "retries": 0, "seconds": 22.7, "max_memory_gb": 17.03, "ce": 0.09822942760001752, "kl": 0.00798040757945273, "context": 0.10221963142976165} +{"step": 51, "steps": 250, "lr": 4.769630887187782e-05, "grad_norm": 1.3096792697906494, "loss_scale": 4.0, "retries": 0, "seconds": 23.2, "max_memory_gb": 17.03, "ce": 0.47753341624411405, "kl": 0.02411270804407195, "context": 0.48958976817084476} +{"step": 52, "steps": 250, "lr": 4.756134269573832e-05, "grad_norm": 0.3982016444206238, "loss_scale": 4.0, "retries": 0, "seconds": 23.6, "max_memory_gb": 17.03, "ce": 0.022759321378543973, "kl": 0.1893149031093344, "context": 0.11741677485406399} +{"step": 53, "steps": 250, "lr": 4.7422760577197226e-05, "grad_norm": 0.7930741906166077, "loss_scale": 4.0, "retries": 0, "seconds": 24.0, "max_memory_gb": 17.03, "ce": 0.12042401020880789, "kl": 0.39120370615273714, "context": 0.31602586060762405} +{"step": 54, "steps": 250, "lr": 4.728058749488963e-05, "grad_norm": 0.5449915528297424, "loss_scale": 4.0, "retries": 0, "seconds": 24.5, "max_memory_gb": 17.03, "ce": 0.04203539360605646, "kl": 0.3255150227123522, "context": 0.20479290471121203} +{"step": 55, "steps": 250, "lr": 4.713484907470133e-05, "grad_norm": 0.19594796001911163, "loss_scale": 4.0, "retries": 0, "seconds": 25.0, "max_memory_gb": 17.03, "ce": 0.021414118236862123, "kl": 0.014417447110190551, "context": 0.02862284248112701} +{"step": 56, "steps": 250, "lr": 4.6985571585149876e-05, "grad_norm": 0.3875943720340729, "loss_scale": 4.0, "retries": 0, "seconds": 25.4, "max_memory_gb": 17.03, "ce": 0.02237821533344686, "kl": 0.21401527954731137, "context": 0.12938585225492716} +{"step": 57, "steps": 250, "lr": 4.6832781932649884e-05, "grad_norm": 0.16711507737636566, "loss_scale": 4.0, "retries": 0, "seconds": 25.9, "max_memory_gb": 17.03, "ce": 0.008422325270657893, "kl": 0.01662750477771624, "context": 0.016736078578105662} +{"step": 58, "steps": 250, "lr": 4.667650765666325e-05, "grad_norm": 0.028492286801338196, "loss_scale": 4.0, "retries": 0, "seconds": 26.3, "max_memory_gb": 17.03, "ce": 0.0016132289383676834, "kl": 0.0008999560209304036, "context": 0.002063206884486135} +{"step": 59, "steps": 250, "lr": 4.651677692473538e-05, "grad_norm": 0.2726878225803375, "loss_scale": 4.0, "retries": 0, "seconds": 26.8, "max_memory_gb": 17.03, "ce": 0.03286314845900051, "kl": 0.0623879442573525, "context": 0.06405711872503161} +{"step": 60, "steps": 250, "lr": 4.635361852741808e-05, "grad_norm": 0.5190157890319824, "loss_scale": 4.0, "retries": 0, "seconds": 27.2, "max_memory_gb": 17.03, "ce": 0.07814182684523985, "kl": 0.0654287159559317, "context": 0.11085618479410186} +{"step": 61, "steps": 250, "lr": 4.6187061873080264e-05, "grad_norm": 0.9531986713409424, "loss_scale": 4.0, "retries": 0, "seconds": 27.7, "max_memory_gb": 17.03, "ce": 0.02158917547785677, "kl": 0.7772277932162979, "context": 0.41020306485006586} +{"step": 62, "steps": 250, "lr": 4.601713698260727e-05, "grad_norm": 0.10969998687505722, "loss_scale": 4.0, "retries": 0, "seconds": 28.1, "max_memory_gb": 17.03, "ce": 0.02016347204335034, "kl": 0.009075642519746907, "context": 0.024701293325051665} +{"step": 63, "steps": 250, "lr": 4.5843874483989746e-05, "grad_norm": 0.5174950361251831, "loss_scale": 4.0, "retries": 0, "seconds": 28.6, "max_memory_gb": 17.03, "ce": 0.05134840466780588, "kl": 0.11910521311801858, "context": 0.11090101022273302} +{"step": 64, "steps": 250, "lr": 4.566730560680316e-05, "grad_norm": 0.18598923087120056, "loss_scale": 4.0, "retries": 0, "seconds": 29.0, "max_memory_gb": 17.03, "ce": 0.01995462510967627, "kl": 0.016171357448911294, "context": 0.028040303324814886} +{"step": 65, "steps": 250, "lr": 4.548746217657878e-05, "grad_norm": 0.5916532874107361, "loss_scale": 4.0, "retries": 0, "seconds": 29.5, "max_memory_gb": 17.03, "ce": 0.15141260949894786, "kl": 0.1517234374769032, "context": 0.22727432753890753} +{"step": 66, "steps": 250, "lr": 4.5304376609067396e-05, "grad_norm": 0.840674638748169, "loss_scale": 4.0, "retries": 0, "seconds": 30.0, "max_memory_gb": 17.03, "ce": 0.15794898173521688, "kl": 0.06900354827098454, "context": 0.1924507596032754} +{"step": 67, "steps": 250, "lr": 4.5118081904396504e-05, "grad_norm": 0.06331270933151245, "loss_scale": 4.0, "retries": 0, "seconds": 30.4, "max_memory_gb": 17.03, "ce": 0.0058982543559977785, "kl": 0.007234948658151552, "context": 0.009515728859696537} +{"step": 68, "steps": 250, "lr": 4.4928611641122225e-05, "grad_norm": 0.09075813740491867, "loss_scale": 4.0, "retries": 0, "seconds": 30.9, "max_memory_gb": 17.03, "ce": 0.005249471274510142, "kl": 0.03186450110842998, "context": 0.021181722499022726} +{"step": 69, "steps": 250, "lr": 4.473599997017701e-05, "grad_norm": 0.6685600280761719, "loss_scale": 4.0, "retries": 0, "seconds": 31.4, "max_memory_gb": 17.03, "ce": 0.3244938444986474, "kl": 0.026773757024784572, "context": 0.33788071089657024} +{"step": 70, "steps": 250, "lr": 4.4540281608714053e-05, "grad_norm": 0.01318266335874796, "loss_scale": 4.0, "retries": 0, "seconds": 31.8, "max_memory_gb": 17.03, "ce": 0.0017261885077459738, "kl": 0.00016616403718217043, "context": 0.00180927048495505} +{"step": 71, "steps": 250, "lr": 4.434149183384977e-05, "grad_norm": 0.11517094820737839, "loss_scale": 4.0, "retries": 0, "seconds": 32.3, "max_memory_gb": 17.03, "ce": 0.010707456231102697, "kl": 0.010895552870351821, "context": 0.016155232908204198} +{"step": 72, "steps": 250, "lr": 4.413966647630534e-05, "grad_norm": 0.42940908670425415, "loss_scale": 4.0, "retries": 0, "seconds": 32.7, "max_memory_gb": 17.03, "ce": 0.01719764742301777, "kl": 0.39265445206547156, "context": 0.21352487371768802} +{"step": 73, "steps": 250, "lr": 4.3934841913948323e-05, "grad_norm": 0.9861831068992615, "loss_scale": 4.0, "retries": 0, "seconds": 33.2, "max_memory_gb": 17.03, "ce": 0.2911085589657887, "kl": 0.0838282144650293, "context": 0.3330226680773194} +{"step": 74, "steps": 250, "lr": 4.372705506523583e-05, "grad_norm": 0.17128324508666992, "loss_scale": 4.0, "retries": 0, "seconds": 33.6, "max_memory_gb": 17.03, "ce": 0.05912718737090472, "kl": 0.11091211012899294, "context": 0.11458324615705351} +{"step": 75, "steps": 250, "lr": 4.351634338256017e-05, "grad_norm": 0.7294386625289917, "loss_scale": 4.0, "retries": 0, "seconds": 34.1, "max_memory_gb": 17.03, "ce": 0.4628721833432792, "kl": 0.9788778331130743, "context": 0.9523110999725759} +{"step": 76, "steps": 250, "lr": 4.330274484549825e-05, "grad_norm": 1.0860651731491089, "loss_scale": 4.0, "retries": 0, "seconds": 34.6, "max_memory_gb": 17.03, "ce": 0.011887717846548185, "kl": 1.6225410388142336, "context": 0.8231582619482651} +{"step": 77, "steps": 250, "lr": 4.308629795396599e-05, "grad_norm": 0.1326388269662857, "loss_scale": 4.0, "retries": 0, "seconds": 35.0, "max_memory_gb": 17.03, "ce": 0.010985344764776528, "kl": 0.011062886529543903, "context": 0.016516787800355814} +{"step": 78, "steps": 250, "lr": 4.286704172127894e-05, "grad_norm": 0.44485628604888916, "loss_scale": 4.0, "retries": 0, "seconds": 35.5, "max_memory_gb": 17.03, "ce": 0.00188755095587112, "kl": 0.11202920452342369, "context": 0.057902154338080436} +{"step": 79, "steps": 250, "lr": 4.264501566712033e-05, "grad_norm": 0.1760922074317932, "loss_scale": 4.0, "retries": 0, "seconds": 35.9, "max_memory_gb": 17.03, "ce": 0.016876722447705106, "kl": 0.02667155336439464, "context": 0.030212498242690344} +{"step": 80, "steps": 250, "lr": 4.2420259810417894e-05, "grad_norm": 0.3171546161174774, "loss_scale": 4.0, "retries": 0, "seconds": 36.3, "max_memory_gb": 17.03, "ce": 0.004161332311923616, "kl": 0.06979448854917791, "context": 0.039058576001480105} +{"step": 81, "steps": 250, "lr": 4.219281466213066e-05, "grad_norm": 0.492328941822052, "loss_scale": 4.0, "retries": 0, "seconds": 36.8, "max_memory_gb": 17.03, "ce": 0.08422820980194956, "kl": 0.0447734919434879, "context": 0.10661495721433312} +{"step": 82, "steps": 250, "lr": 4.1962721217947135e-05, "grad_norm": 0.04819710925221443, "loss_scale": 4.0, "retries": 0, "seconds": 37.2, "max_memory_gb": 17.03, "ce": 0.006598527077585459, "kl": 0.001668263808824122, "context": 0.007432659040205181} +{"step": 83, "steps": 250, "lr": 4.173002095089599e-05, "grad_norm": 1.3910094499588013, "loss_scale": 4.0, "retries": 0, "seconds": 37.7, "max_memory_gb": 17.03, "ce": 0.13659551524324343, "kl": 0.6275358727871208, "context": 0.45036343904212117} +{"step": 84, "steps": 250, "lr": 4.149475580387084e-05, "grad_norm": 0.56479412317276, "loss_scale": 4.0, "retries": 0, "seconds": 38.2, "max_memory_gb": 17.03, "ce": 0.037734911948973604, "kl": 0.04911188094411045, "context": 0.06229085201721318} +{"step": 85, "steps": 250, "lr": 4.1256968182070275e-05, "grad_norm": 0.22084961831569672, "loss_scale": 4.0, "retries": 0, "seconds": 38.6, "max_memory_gb": 17.03, "ce": 0.007968294897636952, "kl": 0.00545471440161549, "context": 0.010695652100366715} +{"step": 86, "steps": 250, "lr": 4.101670094535453e-05, "grad_norm": 1.4515538215637207, "loss_scale": 4.0, "retries": 0, "seconds": 39.1, "max_memory_gb": 17.03, "ce": 0.11365919938543811, "kl": 0.3424575293902308, "context": 0.2848879718221724} +{"step": 87, "steps": 250, "lr": 4.077399740052027e-05, "grad_norm": 0.9042577147483826, "loss_scale": 4.0, "retries": 0, "seconds": 39.5, "max_memory_gb": 17.03, "ce": 0.007699473819229752, "kl": 0.6548030693084002, "context": 0.33510101214051247} +{"step": 88, "steps": 250, "lr": 4.0528901293494775e-05, "grad_norm": 0.1283699870109558, "loss_scale": 4.0, "retries": 0, "seconds": 40.0, "max_memory_gb": 17.03, "ce": 0.010951541540634935, "kl": 0.014890067492530079, "context": 0.018396575766928436} +{"step": 89, "steps": 250, "lr": 4.0281456801451004e-05, "grad_norm": 0.08313681930303574, "loss_scale": 4.0, "retries": 0, "seconds": 40.4, "max_memory_gb": 17.03, "ce": 0.014786536752126267, "kl": 0.011726911579899024, "context": 0.02064999149115465} +{"step": 90, "steps": 250, "lr": 4.0031708524844884e-05, "grad_norm": 0.0570790097117424, "loss_scale": 4.0, "retries": 0, "seconds": 40.9, "max_memory_gb": 17.03, "ce": 0.007018532312940806, "kl": 0.002639731450472027, "context": 0.00833839809638448} +{"step": 91, "steps": 250, "lr": 3.977970147937635e-05, "grad_norm": 1.4531819820404053, "loss_scale": 4.0, "retries": 0, "seconds": 41.3, "max_memory_gb": 17.03, "ce": 0.40020177182304906, "kl": 0.3458884698775364, "context": 0.5731460137467366} +{"step": 92, "steps": 250, "lr": 3.952548108787553e-05, "grad_norm": 1.944313645362854, "loss_scale": 4.0, "retries": 0, "seconds": 41.8, "max_memory_gb": 17.03, "ce": 0.21000844473746838, "kl": 0.38547047879501406, "context": 0.40274368419341045} +{"step": 93, "steps": 250, "lr": 3.9269093172115496e-05, "grad_norm": 0.9694845080375671, "loss_scale": 4.0, "retries": 0, "seconds": 42.2, "max_memory_gb": 17.03, "ce": 0.06704297458054498, "kl": 0.3743771935987752, "context": 0.2542315673199482} +{"step": 94, "steps": 250, "lr": 3.901058394455323e-05, "grad_norm": 1.05013906955719, "loss_scale": 4.0, "retries": 0, "seconds": 42.7, "max_memory_gb": 17.03, "ce": 0.06471366251935251, "kl": 0.03785419254563749, "context": 0.08364076240104623} +{"step": 95, "steps": 250, "lr": 3.875e-05, "grad_norm": 1.5430810451507568, "loss_scale": 4.0, "retries": 0, "seconds": 43.2, "max_memory_gb": 17.03, "ce": 0.40936883155518444, "kl": 0.040847307071089745, "context": 0.4297924926504493} +{"step": 96, "steps": 250, "lr": 3.8487388307223005e-05, "grad_norm": 1.0396592617034912, "loss_scale": 4.0, "retries": 0, "seconds": 43.6, "max_memory_gb": 17.03, "ce": 0.37167296788538806, "kl": 0.006465722606662894, "context": 0.37490584058105014} +{"step": 97, "steps": 250, "lr": 3.822279620047943e-05, "grad_norm": 0.9739329814910889, "loss_scale": 4.0, "retries": 0, "seconds": 44.1, "max_memory_gb": 17.03, "ce": 0.014180020494677592, "kl": 0.1091620335355401, "context": 0.06876103766262531} +{"step": 98, "steps": 250, "lr": 3.795627137098479e-05, "grad_norm": 0.282056987285614, "loss_scale": 4.0, "retries": 0, "seconds": 44.5, "max_memory_gb": 17.03, "ce": 0.0036496852844720706, "kl": 0.0873579473081918, "context": 0.047328658933111} +{"step": 99, "steps": 250, "lr": 3.7687861858316844e-05, "grad_norm": 0.07208988815546036, "loss_scale": 4.0, "retries": 0, "seconds": 45.0, "max_memory_gb": 17.03, "ce": 0.01028372865766869, "kl": 0.006813220421463484, "context": 0.013690338399101165} +{"step": 100, "steps": 250, "lr": 3.741761604175668e-05, "grad_norm": 1.021178126335144, "loss_scale": 4.0, "retries": 0, "seconds": 45.4, "max_memory_gb": 17.03, "ce": 0.028298218559939414, "kl": 0.44205462044192245, "context": 0.24932552941027097} +{"step": 101, "steps": 250, "lr": 3.714558263156872e-05, "grad_norm": 0.25038596987724304, "loss_scale": 4.0, "retries": 0, "seconds": 45.9, "max_memory_gb": 17.03, "ce": 0.003833352995570749, "kl": 0.08867719446425326, "context": 0.04817195073701441} +{"step": 102, "steps": 250, "lr": 3.6871810660220904e-05, "grad_norm": 0.45333877205848694, "loss_scale": 4.0, "retries": 0, "seconds": 46.4, "max_memory_gb": 17.03, "ce": 0.05208042968297377, "kl": 0.02731550403404981, "context": 0.06573818216565996} +{"step": 103, "steps": 250, "lr": 3.659634947354686e-05, "grad_norm": 0.11849364638328552, "loss_scale": 4.0, "retries": 0, "seconds": 46.8, "max_memory_gb": 17.03, "ce": 0.010483227506483672, "kl": 0.013883957879443187, "context": 0.017425206387997605} +{"step": 104, "steps": 250, "lr": 3.631924872185169e-05, "grad_norm": 0.06797442585229874, "loss_scale": 4.0, "retries": 0, "seconds": 47.2, "max_memory_gb": 17.03, "ce": 0.004509986229095375, "kl": 0.007347288940763974, "context": 0.00818363036523806} +{"step": 105, "steps": 250, "lr": 3.6040558350962674e-05, "grad_norm": 0.49751296639442444, "loss_scale": 4.0, "retries": 0, "seconds": 47.7, "max_memory_gb": 17.03, "ce": 0.03095902723725885, "kl": 0.3512296425178647, "context": 0.20657384453807026} +{"step": 106, "steps": 250, "lr": 3.576032859322687e-05, "grad_norm": 0.06552024930715561, "loss_scale": 4.0, "retries": 0, "seconds": 48.1, "max_memory_gb": 17.03, "ce": 0.00574508604768198, "kl": 0.002659450219653081, "context": 0.007074811335769482} +{"step": 107, "steps": 250, "lr": 3.5478609958457056e-05, "grad_norm": 0.42443516850471497, "loss_scale": 4.0, "retries": 0, "seconds": 48.6, "max_memory_gb": 17.03, "ce": 0.03529062820598483, "kl": 0.18507945343844767, "context": 0.12783035865868442} +{"step": 108, "steps": 250, "lr": 3.519545322482755e-05, "grad_norm": 2.0436384677886963, "loss_scale": 4.0, "retries": 0, "seconds": 49.0, "max_memory_gb": 17.03, "ce": 0.24757855100688175, "kl": 0.03148094756750197, "context": 0.2633190097476472} +{"step": 109, "steps": 250, "lr": 3.491090942972179e-05, "grad_norm": 1.1742371320724487, "loss_scale": 4.0, "retries": 0, "seconds": 49.5, "max_memory_gb": 17.03, "ce": 0.12017881235806271, "kl": 0.3695967085659504, "context": 0.3049771748483181} +{"step": 110, "steps": 250, "lr": 3.4625029860533135e-05, "grad_norm": 1.0979045629501343, "loss_scale": 4.0, "retries": 0, "seconds": 50.0, "max_memory_gb": 17.03, "ce": 0.03651611343957484, "kl": 0.10795420443173498, "context": 0.09049321501515806} +{"step": 111, "steps": 250, "lr": 3.433786604542056e-05, "grad_norm": 0.02202354371547699, "loss_scale": 4.0, "retries": 0, "seconds": 50.4, "max_memory_gb": 17.03, "ce": 0.001631044497116818, "kl": 0.0015675078921049135, "context": 0.002414798436802812} +{"step": 112, "steps": 250, "lr": 3.404946974402104e-05, "grad_norm": 0.09950650483369827, "loss_scale": 4.0, "retries": 0, "seconds": 50.9, "max_memory_gb": 17.03, "ce": 0.009248501854017377, "kl": 0.0009517269354546443, "context": 0.009724364877911285} +{"step": 113, "steps": 250, "lr": 3.375989293812019e-05, "grad_norm": 0.547526478767395, "loss_scale": 4.0, "retries": 0, "seconds": 51.3, "max_memory_gb": 17.03, "ce": 0.010414859454613179, "kl": 0.22404723684303463, "context": 0.12243848224170506} +{"step": 114, "steps": 250, "lr": 3.3469187822282766e-05, "grad_norm": 1.2829176187515259, "loss_scale": 4.0, "retries": 0, "seconds": 51.8, "max_memory_gb": 17.03, "ce": 0.010290073769283481, "kl": 0.8492824825807475, "context": 0.43493131629657} +{"step": 115, "steps": 250, "lr": 3.317740679444499e-05, "grad_norm": 0.12803243100643158, "loss_scale": 4.0, "retries": 0, "seconds": 52.3, "max_memory_gb": 17.03, "ce": 0.006330150514259003, "kl": 0.028514359622931806, "context": 0.02058733043668326} +{"step": 116, "steps": 250, "lr": 3.2884602446470056e-05, "grad_norm": 0.7060248255729675, "loss_scale": 4.0, "retries": 0, "seconds": 52.7, "max_memory_gb": 17.03, "ce": 0.05228008536869311, "kl": 0.4870599882636384, "context": 0.29581007576416596} +{"step": 117, "steps": 250, "lr": 3.259082755466876e-05, "grad_norm": 0.23245637118816376, "loss_scale": 4.0, "retries": 0, "seconds": 53.2, "max_memory_gb": 17.03, "ce": 0.007272361792274751, "kl": 0.011291879483906087, "context": 0.0129183019453194} +{"step": 118, "steps": 250, "lr": 3.229613507028686e-05, "grad_norm": 0.7925398349761963, "loss_scale": 4.0, "retries": 0, "seconds": 53.6, "max_memory_gb": 17.03, "ce": 0.2854174791100377, "kl": 0.00544866241307318, "context": 0.2881418038268748} +{"step": 119, "steps": 250, "lr": 3.2000578109961004e-05, "grad_norm": 0.15989096462726593, "loss_scale": 4.0, "retries": 0, "seconds": 54.1, "max_memory_gb": 17.03, "ce": 0.01509233702745405, "kl": 0.023622276677201626, "context": 0.026903475361905294} +{"step": 120, "steps": 250, "lr": 3.1704209946144616e-05, "grad_norm": 1.8607845306396484, "loss_scale": 4.0, "retries": 0, "seconds": 54.6, "max_memory_gb": 17.03, "ce": 0.25362767750630155, "kl": 0.1670017147262115, "context": 0.3371285400353372} +{"step": 121, "steps": 250, "lr": 3.140708399750594e-05, "grad_norm": 0.32582011818885803, "loss_scale": 4.0, "retries": 0, "seconds": 55.1, "max_memory_gb": 17.03, "ce": 0.07047239800886018, "kl": 0.040578877395091695, "context": 0.09076183903380297} +{"step": 122, "steps": 250, "lr": 3.1109253819299606e-05, "grad_norm": 1.1192219257354736, "loss_scale": 4.0, "retries": 0, "seconds": 55.5, "max_memory_gb": 17.03, "ce": 0.19138703450153116, "kl": 0.01662173381919274, "context": 0.1996978996648977} +{"step": 123, "steps": 250, "lr": 3.081077309371361e-05, "grad_norm": 0.3873857259750366, "loss_scale": 4.0, "retries": 0, "seconds": 56.0, "max_memory_gb": 17.03, "ce": 0.039366116165183485, "kl": 0.026533628930337727, "context": 0.05263292888412252} +{"step": 124, "steps": 250, "lr": 3.051169562019339e-05, "grad_norm": 1.5917096138000488, "loss_scale": 4.0, "retries": 0, "seconds": 56.4, "max_memory_gb": 17.03, "ce": 0.7518948805081891, "kl": 0.4963836315146182, "context": 1.0000866627524374} +{"step": 125, "steps": 250, "lr": 3.021207530574477e-05, "grad_norm": 0.3557108938694, "loss_scale": 4.0, "retries": 0, "seconds": 56.9, "max_memory_gb": 17.03, "ce": 0.13643087587843183, "kl": 0.06694549463645671, "context": 0.16990363056538627} +{"step": 126, "steps": 250, "lr": 2.991196615521758e-05, "grad_norm": 1.581441879272461, "loss_scale": 4.0, "retries": 0, "seconds": 57.3, "max_memory_gb": 17.03, "ce": 0.757691165432334, "kl": 0.09041343908756971, "context": 0.802897859364748} +{"step": 127, "steps": 250, "lr": 2.9611422261571536e-05, "grad_norm": 0.10198111087083817, "loss_scale": 4.0, "retries": 0, "seconds": 57.8, "max_memory_gb": 17.03, "ce": 0.009884103994409088, "kl": 0.0015159232170844916, "context": 0.010642065739375539} +{"step": 128, "steps": 250, "lr": 2.931049779612634e-05, "grad_norm": 1.7708078622817993, "loss_scale": 4.0, "retries": 0, "seconds": 58.3, "max_memory_gb": 17.03, "ce": 0.06705376431636978, "kl": 0.8782407676917501, "context": 0.50617413490545} +{"step": 129, "steps": 250, "lr": 2.9009246998797557e-05, "grad_norm": 0.7329121232032776, "loss_scale": 4.0, "retries": 0, "seconds": 58.7, "max_memory_gb": 17.03, "ce": 0.11665922921383753, "kl": 0.02876431388722267, "context": 0.13104138663038611} +{"step": 130, "steps": 250, "lr": 2.8707724168320238e-05, "grad_norm": 0.39186763763427734, "loss_scale": 4.0, "retries": 0, "seconds": 59.2, "max_memory_gb": 17.03, "ce": 0.05048683303175494, "kl": 0.28171021211892366, "context": 0.19134194235084578} +{"step": 131, "steps": 250, "lr": 2.840598365246184e-05, "grad_norm": 2.1831634044647217, "loss_scale": 4.0, "retries": 0, "seconds": 59.6, "max_memory_gb": 17.03, "ce": 0.2979839983745478, "kl": 0.32291715347673744, "context": 0.4594425620161928} +{"step": 132, "steps": 250, "lr": 2.8104079838226383e-05, "grad_norm": 1.01661217212677, "loss_scale": 4.0, "retries": 0, "seconds": 60.1, "max_memory_gb": 17.03, "ce": 0.20938155775365885, "kl": 0.029919255834101932, "context": 0.22434117825469002} +{"step": 133, "steps": 250, "lr": 2.7802067142051446e-05, "grad_norm": 1.0701940059661865, "loss_scale": 4.0, "retries": 0, "seconds": 60.5, "max_memory_gb": 17.03, "ce": 0.4348538512713276, "kl": 0.09802910219877958, "context": 0.4838683931156993} +{"step": 134, "steps": 250, "lr": 2.7500000000000004e-05, "grad_norm": 0.4816232919692993, "loss_scale": 4.0, "retries": 0, "seconds": 61.0, "max_memory_gb": 17.03, "ce": 0.11785231171597843, "kl": 0.0026169502093580377, "context": 0.11916078612557612} +{"step": 135, "steps": 250, "lr": 2.7197932857948566e-05, "grad_norm": 0.38143593072891235, "loss_scale": 4.0, "retries": 0, "seconds": 61.4, "max_memory_gb": 17.03, "ce": 0.017068182001821697, "kl": 0.13301423390112177, "context": 0.08357529900968075} +{"step": 136, "steps": 250, "lr": 2.689592016177363e-05, "grad_norm": 0.3733215928077698, "loss_scale": 4.0, "retries": 0, "seconds": 61.9, "max_memory_gb": 17.03, "ce": 0.027781492477515712, "kl": 0.050107816539821215, "context": 0.052835402137134224} +{"step": 137, "steps": 250, "lr": 2.659401634753817e-05, "grad_norm": 0.43028372526168823, "loss_scale": 4.0, "retries": 0, "seconds": 62.3, "max_memory_gb": 17.03, "ce": 0.038209167192690074, "kl": 0.09418106934754178, "context": 0.08529969834489748} +{"step": 138, "steps": 250, "lr": 2.6292275831679764e-05, "grad_norm": 0.7217217087745667, "loss_scale": 4.0, "retries": 0, "seconds": 62.8, "max_memory_gb": 17.03, "ce": 0.07018623524345458, "kl": 0.10873509990051389, "context": 0.12455378472805023} +{"step": 139, "steps": 250, "lr": 2.5990753001202445e-05, "grad_norm": 0.7847253680229187, "loss_scale": 4.0, "retries": 0, "seconds": 63.3, "max_memory_gb": 17.03, "ce": 0.0034984926824108697, "kl": 0.10119933201758613, "context": 0.054098159900604514} +{"step": 140, "steps": 250, "lr": 2.5689502203873672e-05, "grad_norm": 0.11256065964698792, "loss_scale": 4.0, "retries": 0, "seconds": 63.7, "max_memory_gb": 17.03, "ce": 0.00865374477871228, "kl": 0.0037992392244632356, "context": 0.010553364336374216} +{"step": 141, "steps": 250, "lr": 2.538857773842847e-05, "grad_norm": 0.06278890371322632, "loss_scale": 4.0, "retries": 0, "seconds": 64.2, "max_memory_gb": 17.03, "ce": 0.006333414858090691, "kl": 0.0051343380473554134, "context": 0.008900583910872228} +{"step": 142, "steps": 250, "lr": 2.5088033844782426e-05, "grad_norm": 0.1468425989151001, "loss_scale": 4.0, "retries": 0, "seconds": 64.7, "max_memory_gb": 17.03, "ce": 0.012729324902466033, "kl": 0.008390945615246892, "context": 0.01692479767370969} +{"step": 143, "steps": 250, "lr": 2.4787924694255237e-05, "grad_norm": 1.5532841682434082, "loss_scale": 4.0, "retries": 0, "seconds": 65.2, "max_memory_gb": 17.03, "ce": 0.33014116615231615, "kl": 0.054901811660784006, "context": 0.35759206553484546} +{"step": 144, "steps": 250, "lr": 2.4488304379806627e-05, "grad_norm": 0.6458211541175842, "loss_scale": 4.0, "retries": 0, "seconds": 65.6, "max_memory_gb": 17.03, "ce": 0.027421454549767077, "kl": 0.22410322073847055, "context": 0.13947306433692575} +{"step": 145, "steps": 250, "lr": 2.4189226906286388e-05, "grad_norm": 0.044459860771894455, "loss_scale": 4.0, "retries": 0, "seconds": 66.1, "max_memory_gb": 17.03, "ce": 0.002290274016559124, "kl": 0.006311983289378986, "context": 0.005446265582577325} +{"step": 146, "steps": 250, "lr": 2.3890746180700395e-05, "grad_norm": 0.08699104934930801, "loss_scale": 4.0, "retries": 0, "seconds": 66.6, "max_memory_gb": 17.03, "ce": 0.01069367304444313, "kl": 0.006202428268807125, "context": 0.013794886996038258} +{"step": 147, "steps": 250, "lr": 2.359291600249407e-05, "grad_norm": 1.3830406665802002, "loss_scale": 4.0, "retries": 0, "seconds": 67.0, "max_memory_gb": 17.03, "ce": 0.5247288077953272, "kl": 0.39880173426354304, "context": 0.7241296680876985} +{"step": 148, "steps": 250, "lr": 2.3295790053855392e-05, "grad_norm": 0.07576215267181396, "loss_scale": 4.0, "retries": 0, "seconds": 67.5, "max_memory_gb": 17.03, "ce": 0.004946048153215088, "kl": 0.010359929321566597, "context": 0.010126012726686895} +{"step": 149, "steps": 250, "lr": 2.299942189003901e-05, "grad_norm": 0.06464395672082901, "loss_scale": 4.0, "retries": 0, "seconds": 67.9, "max_memory_gb": 17.03, "ce": 0.007749440526822582, "kl": 0.002852310872185626, "context": 0.009175596118438989} +{"step": 150, "steps": 250, "lr": 2.2703864929713144e-05, "grad_norm": 0.19881440699100494, "loss_scale": 4.0, "retries": 0, "seconds": 68.4, "max_memory_gb": 17.03, "ce": 0.04221794509794563, "kl": 0.030742385773919523, "context": 0.05758913769386709} +{"step": 151, "steps": 250, "lr": 2.2409172445331257e-05, "grad_norm": 0.32710912823677063, "loss_scale": 4.0, "retries": 0, "seconds": 68.9, "max_memory_gb": 17.03, "ce": 0.0159028768830467, "kl": 0.019501950771882548, "context": 0.02565385226625949} +{"step": 152, "steps": 250, "lr": 2.211539755352996e-05, "grad_norm": 0.9660107493400574, "loss_scale": 4.0, "retries": 0, "seconds": 69.3, "max_memory_gb": 17.03, "ce": 0.011791162985900883, "kl": 0.46118513574992903, "context": 0.24238372579566203} +{"step": 153, "steps": 250, "lr": 2.1822593205555016e-05, "grad_norm": 0.6168968081474304, "loss_scale": 4.0, "retries": 0, "seconds": 69.8, "max_memory_gb": 17.03, "ce": 0.010861288395972224, "kl": 0.3862015298309416, "context": 0.20396205785073107} +{"step": 154, "steps": 250, "lr": 2.1530812177717236e-05, "grad_norm": 0.393751323223114, "loss_scale": 4.0, "retries": 0, "seconds": 70.2, "max_memory_gb": 17.03, "ce": 0.04448456986574456, "kl": 0.07473974306776654, "context": 0.08185443776892498} +{"step": 155, "steps": 250, "lr": 2.1240107061879815e-05, "grad_norm": 0.9528965950012207, "loss_scale": 4.0, "retries": 0, "seconds": 70.7, "max_memory_gb": 17.03, "ce": 0.20300727489666315, "kl": 0.008841387428901726, "context": 0.20742796948252362} +{"step": 156, "steps": 250, "lr": 2.095053025597896e-05, "grad_norm": 0.3426593542098999, "loss_scale": 4.0, "retries": 0, "seconds": 71.1, "max_memory_gb": 17.03, "ce": 0.015056069423735607, "kl": 0.008825134327139494, "context": 0.01946863617922645} +{"step": 157, "steps": 250, "lr": 2.066213395457945e-05, "grad_norm": 0.10129007697105408, "loss_scale": 4.0, "retries": 0, "seconds": 71.5, "max_memory_gb": 17.03, "ce": 0.008060934487730265, "kl": 0.01840615033142967, "context": 0.017264009919017553} +{"step": 158, "steps": 250, "lr": 2.0374970139466877e-05, "grad_norm": 0.9789071679115295, "loss_scale": 4.0, "retries": 0, "seconds": 71.9, "max_memory_gb": 17.03, "ce": 0.24024197226390243, "kl": 0.09044754020214896, "context": 0.2854657488060184} +{"step": 159, "steps": 250, "lr": 2.0089090570278213e-05, "grad_norm": 0.47290173172950745, "loss_scale": 4.0, "retries": 0, "seconds": 72.4, "max_memory_gb": 17.03, "ce": 0.005474655183206778, "kl": 0.07463376154191792, "context": 0.042791535961441696} +{"step": 160, "steps": 250, "lr": 1.9804546775172457e-05, "grad_norm": 0.4661954641342163, "loss_scale": 4.0, "retries": 0, "seconds": 72.8, "max_memory_gb": 17.03, "ce": 0.04864725150400773, "kl": 0.33110148830746766, "context": 0.21419799292925745} +{"step": 161, "steps": 250, "lr": 1.9521390041542946e-05, "grad_norm": 0.6289879679679871, "loss_scale": 4.0, "retries": 0, "seconds": 73.2, "max_memory_gb": 17.03, "ce": 0.09312230058276327, "kl": 0.08761238165743634, "context": 0.13692849327344447} +{"step": 162, "steps": 250, "lr": 1.923967140677313e-05, "grad_norm": 1.0681747198104858, "loss_scale": 4.0, "retries": 0, "seconds": 73.6, "max_memory_gb": 17.03, "ce": 0.07321049377787858, "kl": 0.126152463562903, "context": 0.13628672555205412} +{"step": 163, "steps": 250, "lr": 1.8959441649037335e-05, "grad_norm": 2.490954637527466, "loss_scale": 4.0, "retries": 0, "seconds": 74.0, "max_memory_gb": 17.03, "ce": 0.2198253006208688, "kl": 0.6236569188185968, "context": 0.5316537577891722} +{"step": 164, "steps": 250, "lr": 1.868075127814832e-05, "grad_norm": 0.9794778823852539, "loss_scale": 4.0, "retries": 0, "seconds": 74.5, "max_memory_gb": 17.03, "ce": 0.10525978803343605, "kl": 0.07853399952728068, "context": 0.14452678785892203} +{"step": 165, "steps": 250, "lr": 1.8403650526453143e-05, "grad_norm": 0.10609472543001175, "loss_scale": 4.0, "retries": 0, "seconds": 74.9, "max_memory_gb": 17.03, "ce": 0.005300105374772102, "kl": 0.014486485774796165, "context": 0.012543348435428925} +{"step": 166, "steps": 250, "lr": 1.812818933977912e-05, "grad_norm": 1.5901341438293457, "loss_scale": 4.0, "retries": 0, "seconds": 75.3, "max_memory_gb": 17.03, "ce": 0.5688748210377526, "kl": 0.11415106684580678, "context": 0.6259503744659014} +{"step": 167, "steps": 250, "lr": 1.785441736843128e-05, "grad_norm": 0.2725522816181183, "loss_scale": 4.0, "retries": 0, "seconds": 75.8, "max_memory_gb": 17.03, "ce": 0.024177291430532932, "kl": 0.006739090389601188, "context": 0.027546836179681122} +{"step": 168, "steps": 250, "lr": 1.7582383958243314e-05, "grad_norm": 0.23770394921302795, "loss_scale": 4.0, "retries": 0, "seconds": 76.2, "max_memory_gb": 17.03, "ce": 0.026351861393777654, "kl": 0.005469133306178264, "context": 0.029086427646689117} +{"step": 169, "steps": 250, "lr": 1.7312138141683164e-05, "grad_norm": 1.319625735282898, "loss_scale": 4.0, "retries": 0, "seconds": 76.6, "max_memory_gb": 17.03, "ce": 0.06356582866283134, "kl": 1.1812528371810913, "context": 0.6541922390460968} +{"step": 170, "steps": 250, "lr": 1.704372862901521e-05, "grad_norm": 0.06064365431666374, "loss_scale": 4.0, "retries": 0, "seconds": 77.0, "max_memory_gb": 17.03, "ce": 0.003419166663661599, "kl": 0.0005057900798419723, "context": 0.0036720616626553237} +{"step": 171, "steps": 250, "lr": 1.6777203799520573e-05, "grad_norm": 0.6917200088500977, "loss_scale": 4.0, "retries": 0, "seconds": 77.4, "max_memory_gb": 17.03, "ce": 0.0356911159997253, "kl": 0.19726267957594246, "context": 0.13432245544390753} +{"step": 172, "steps": 250, "lr": 1.6512611692777008e-05, "grad_norm": 0.19312314689159393, "loss_scale": 4.0, "retries": 0, "seconds": 77.9, "max_memory_gb": 17.03, "ce": 0.01872201805235818, "kl": 0.020736611360916868, "context": 0.029090323485434055} +{"step": 173, "steps": 250, "lr": 1.6250000000000005e-05, "grad_norm": 0.2933748960494995, "loss_scale": 4.0, "retries": 0, "seconds": 78.3, "max_memory_gb": 17.03, "ce": 0.031157539633568376, "kl": 0.017510380159365013, "context": 0.03991272987332195} +{"step": 174, "steps": 250, "lr": 1.598941605544678e-05, "grad_norm": 0.33721041679382324, "loss_scale": 4.0, "retries": 0, "seconds": 78.7, "max_memory_gb": 17.03, "ce": 0.03263972975764773, "kl": 0.02438539386548655, "context": 0.044832428564404836} +{"step": 175, "steps": 250, "lr": 1.5730906827884506e-05, "grad_norm": 1.0812634229660034, "loss_scale": 4.0, "retries": 0, "seconds": 79.1, "max_memory_gb": 17.03, "ce": 0.1329997160937637, "kl": 0.0756831404287368, "context": 0.1708412878215313} +{"step": 176, "steps": 250, "lr": 1.5474518912124476e-05, "grad_norm": 0.11225103586912155, "loss_scale": 4.0, "retries": 0, "seconds": 79.5, "max_memory_gb": 17.03, "ce": 0.01110675124800764, "kl": 0.002949470261228271, "context": 0.012581486138515174} +{"step": 177, "steps": 250, "lr": 1.5220298520623646e-05, "grad_norm": 0.7594441175460815, "loss_scale": 4.0, "retries": 0, "seconds": 79.9, "max_memory_gb": 17.03, "ce": 0.030622158985352144, "kl": 0.6516925141768297, "context": 0.35646841234120075} +{"step": 178, "steps": 250, "lr": 1.4968291475155121e-05, "grad_norm": 0.27180856466293335, "loss_scale": 4.0, "retries": 0, "seconds": 80.3, "max_memory_gb": 17.03, "ce": 0.0382986111799255, "kl": 0.04731054467265494, "context": 0.06195388198830187} +{"step": 179, "steps": 250, "lr": 1.4718543198548998e-05, "grad_norm": 0.8271092176437378, "loss_scale": 4.0, "retries": 0, "seconds": 80.7, "max_memory_gb": 17.03, "ce": 0.42855861806310713, "kl": 0.016734159144107252, "context": 0.43692570440180134} +{"step": 180, "steps": 250, "lr": 1.4471098706505227e-05, "grad_norm": 0.36870071291923523, "loss_scale": 4.0, "retries": 0, "seconds": 81.1, "max_memory_gb": 17.03, "ce": 0.04123671640991233, "kl": 0.024486517522746, "context": 0.05347997319768183} +{"step": 181, "steps": 250, "lr": 1.4226002599479743e-05, "grad_norm": 0.39366614818573, "loss_scale": 4.0, "retries": 0, "seconds": 81.6, "max_memory_gb": 17.03, "ce": 0.00895782490260899, "kl": 0.12747281495830975, "context": 0.0726942319306545} +{"step": 182, "steps": 250, "lr": 1.3983299054645471e-05, "grad_norm": 0.43555688858032227, "loss_scale": 4.0, "retries": 0, "seconds": 82.0, "max_memory_gb": 17.03, "ce": 0.01700750837335363, "kl": 0.35860173347646196, "context": 0.19630837347358465} +{"step": 183, "steps": 250, "lr": 1.3743031817929724e-05, "grad_norm": 0.06283027678728104, "loss_scale": 4.0, "retries": 0, "seconds": 82.4, "max_memory_gb": 17.03, "ce": 0.004917282309179427, "kl": 0.0008670056849950925, "context": 0.005350785228074528} +{"step": 184, "steps": 250, "lr": 1.3505244196129164e-05, "grad_norm": 0.05623235926032066, "loss_scale": 4.0, "retries": 0, "seconds": 82.8, "max_memory_gb": 17.03, "ce": 0.0030248176408349536, "kl": 0.0039012392808217555, "context": 0.004975437303073704} +{"step": 185, "steps": 250, "lr": 1.3269979049104014e-05, "grad_norm": 0.2644278407096863, "loss_scale": 4.0, "retries": 0, "seconds": 83.2, "max_memory_gb": 17.03, "ce": 0.018095926148816943, "kl": 0.10234395134466467, "context": 0.06926790380384773} +{"step": 186, "steps": 250, "lr": 1.3037278782052863e-05, "grad_norm": 1.4592841863632202, "loss_scale": 4.0, "retries": 0, "seconds": 83.7, "max_memory_gb": 17.03, "ce": 0.119017002536566, "kl": 0.23699821200352744, "context": 0.2375161081145052} +{"step": 187, "steps": 250, "lr": 1.2807185337869343e-05, "grad_norm": 0.5812423825263977, "loss_scale": 4.0, "retries": 0, "seconds": 84.1, "max_memory_gb": 17.03, "ce": 0.05227982407086529, "kl": 0.4761710772436345, "context": 0.29036535893101245} +{"step": 188, "steps": 250, "lr": 1.2579740189582115e-05, "grad_norm": 1.0277632474899292, "loss_scale": 4.0, "retries": 0, "seconds": 84.5, "max_memory_gb": 17.03, "ce": 0.23303340493839642, "kl": 0.017826313420982842, "context": 0.24194655861901992} +{"step": 189, "steps": 250, "lr": 1.2354984332879668e-05, "grad_norm": 1.1544181108474731, "loss_scale": 4.0, "retries": 0, "seconds": 84.9, "max_memory_gb": 17.03, "ce": 0.2938855163520202, "kl": 0.36039742588764057, "context": 0.4740842153551057} +{"step": 190, "steps": 250, "lr": 1.2132958278721064e-05, "grad_norm": 0.11300656199455261, "loss_scale": 4.0, "retries": 0, "seconds": 85.3, "max_memory_gb": 17.03, "ce": 0.004160032054642215, "kl": 0.01837022410472855, "context": 0.013345143757760525} +{"step": 191, "steps": 250, "lr": 1.1913702046034016e-05, "grad_norm": 1.307254672050476, "loss_scale": 4.0, "retries": 0, "seconds": 85.7, "max_memory_gb": 17.03, "ce": 0.35139350923418533, "kl": 0.22044350085889164, "context": 0.46161525140632875} +{"step": 192, "steps": 250, "lr": 1.1697255154501756e-05, "grad_norm": 1.0238513946533203, "loss_scale": 4.0, "retries": 0, "seconds": 86.1, "max_memory_gb": 17.03, "ce": 0.5667436913354322, "kl": 0.10011550260242075, "context": 0.6168014630675316} +{"step": 193, "steps": 250, "lr": 1.1483656617439842e-05, "grad_norm": 0.051361437886953354, "loss_scale": 4.0, "retries": 0, "seconds": 86.6, "max_memory_gb": 17.03, "ce": 0.004702752106823027, "kl": 0.0023796807172402623, "context": 0.005892592558666365} +{"step": 194, "steps": 250, "lr": 1.1272944934764177e-05, "grad_norm": 0.33124712109565735, "loss_scale": 4.0, "retries": 0, "seconds": 87.0, "max_memory_gb": 17.03, "ce": 0.0437122427392751, "kl": 0.10149333086883416, "context": 0.09445891188806854} +{"step": 195, "steps": 250, "lr": 1.1065158086051684e-05, "grad_norm": 0.08974657207727432, "loss_scale": 4.0, "retries": 0, "seconds": 87.4, "max_memory_gb": 17.03, "ce": 0.01138039561919868, "kl": 0.0042718109471024945, "context": 0.0135163011145778} +{"step": 196, "steps": 250, "lr": 1.086033352369467e-05, "grad_norm": 1.0476049184799194, "loss_scale": 4.0, "retries": 0, "seconds": 87.8, "max_memory_gb": 17.03, "ce": 0.001800829795683967, "kl": 0.4329603006092668, "context": 0.21828098478363245} +{"step": 197, "steps": 250, "lr": 1.0658508166150224e-05, "grad_norm": 2.3913276195526123, "loss_scale": 4.0, "retries": 0, "seconds": 88.2, "max_memory_gb": 17.03, "ce": 0.10487320221727714, "kl": 0.7637474794382797, "context": 0.48674695240333676} +{"step": 198, "steps": 250, "lr": 1.0459718391285954e-05, "grad_norm": 0.19971363246440887, "loss_scale": 4.0, "retries": 0, "seconds": 88.7, "max_memory_gb": 17.03, "ce": 0.06271432648645714, "kl": 0.010338675603634329, "context": 0.06788366244290955} +{"step": 199, "steps": 250, "lr": 1.0264000029822999e-05, "grad_norm": 1.0951231718063354, "loss_scale": 4.0, "retries": 0, "seconds": 89.1, "max_memory_gb": 17.03, "ce": 0.1811512049753219, "kl": 0.07933440090255317, "context": 0.2208183978800662} +{"step": 200, "steps": 250, "lr": 1.0071388358877775e-05, "grad_norm": 0.4814234972000122, "loss_scale": 4.0, "retries": 0, "seconds": 89.5, "max_memory_gb": 17.03, "ce": 0.2591826236748602, "kl": 0.043484976566105615, "context": 0.28092511917930096} +{"step": 201, "steps": 250, "lr": 9.881918095603501e-06, "grad_norm": 0.4331781566143036, "loss_scale": 4.0, "retries": 0, "seconds": 89.9, "max_memory_gb": 17.03, "ce": 0.0689631151908543, "kl": 0.1178324810207414, "context": 0.1278793631354347} +{"step": 202, "steps": 250, "lr": 9.695623390932612e-06, "grad_norm": 0.030282583087682724, "loss_scale": 4.0, "retries": 0, "seconds": 90.3, "max_memory_gb": 17.03, "ce": 0.0018578129893285222, "kl": 0.002896241433063551, "context": 0.003305933681986062} +{"step": 203, "steps": 250, "lr": 9.51253782342123e-06, "grad_norm": 0.40700089931488037, "loss_scale": 4.0, "retries": 0, "seconds": 90.7, "max_memory_gb": 17.03, "ce": 0.012786234830855392, "kl": 0.02110551981604658, "context": 0.023338994767982513} +{"step": 204, "steps": 250, "lr": 9.332694393196846e-06, "grad_norm": 0.33653727173805237, "loss_scale": 4.0, "retries": 0, "seconds": 91.1, "max_memory_gb": 17.03, "ce": 0.044432983675505966, "kl": 0.010142287355847657, "context": 0.04950412758626044} +{"step": 205, "steps": 250, "lr": 9.156125516010256e-06, "grad_norm": 0.7399264574050903, "loss_scale": 4.0, "retries": 0, "seconds": 91.6, "max_memory_gb": 17.03, "ce": 0.0023914349294500425, "kl": 0.42553503411181737, "context": 0.21515894995536655} +{"step": 206, "steps": 250, "lr": 8.982863017392733e-06, "grad_norm": 0.1276760697364807, "loss_scale": 4.0, "retries": 0, "seconds": 92.0, "max_memory_gb": 17.03, "ce": 0.009654847923229681, "kl": 0.003691989180993005, "context": 0.011500842687382828} +{"step": 207, "steps": 250, "lr": 8.81293812691974e-06, "grad_norm": 0.15303780138492584, "loss_scale": 4.0, "retries": 0, "seconds": 92.4, "max_memory_gb": 17.03, "ce": 0.0644440292235231, "kl": 0.04007509260554798, "context": 0.08448157738894224} +{"step": 208, "steps": 250, "lr": 8.646381472581927e-06, "grad_norm": 0.07133025676012039, "loss_scale": 4.0, "retries": 0, "seconds": 92.8, "max_memory_gb": 17.03, "ce": 0.005688674500561319, "kl": 0.0023967228462424828, "context": 0.006887035706313327} +{"step": 209, "steps": 250, "lr": 8.483223075264622e-06, "grad_norm": 0.7325005531311035, "loss_scale": 4.0, "retries": 0, "seconds": 93.2, "max_memory_gb": 17.03, "ce": 0.17692685109796003, "kl": 0.09298358572414145, "context": 0.22341864299960434} +{"step": 210, "steps": 250, "lr": 8.323492343336747e-06, "grad_norm": 0.8628749251365662, "loss_scale": 4.0, "retries": 0, "seconds": 93.7, "max_memory_gb": 17.03, "ce": 0.5233511367041501, "kl": 0.13913243213573878, "context": 0.5929173672229808} +{"step": 211, "steps": 250, "lr": 8.167218067350122e-06, "grad_norm": 0.30474504828453064, "loss_scale": 4.0, "retries": 0, "seconds": 94.1, "max_memory_gb": 17.03, "ce": 0.11878074309788644, "kl": 0.12145576096372679, "context": 0.1795086159836501} +{"step": 212, "steps": 250, "lr": 8.01442841485013e-06, "grad_norm": 1.28254234790802, "loss_scale": 4.0, "retries": 0, "seconds": 94.5, "max_memory_gb": 17.03, "ce": 0.0645886771089863, "kl": 0.9029845556542568, "context": 0.5160809705266729} +{"step": 213, "steps": 250, "lr": 7.865150925298678e-06, "grad_norm": 0.6484095454216003, "loss_scale": 4.0, "retries": 0, "seconds": 94.9, "max_memory_gb": 17.03, "ce": 0.26835493374528596, "kl": 0.04778774274745956, "context": 0.2922488045951468} +{"step": 214, "steps": 250, "lr": 7.719412505110375e-06, "grad_norm": 1.629351258277893, "loss_scale": 4.0, "retries": 0, "seconds": 95.4, "max_memory_gb": 17.03, "ce": 0.4034676400551689, "kl": 0.1292928293987643, "context": 0.4681140653556213} +{"step": 215, "steps": 250, "lr": 7.57723942280278e-06, "grad_norm": 1.2515625953674316, "loss_scale": 4.0, "retries": 0, "seconds": 95.8, "max_memory_gb": 17.03, "ce": 0.6144464304015855, "kl": 0.03008011630083729, "context": 0.629486465237278} +{"step": 216, "steps": 250, "lr": 7.438657304261687e-06, "grad_norm": 0.32815074920654297, "loss_scale": 4.0, "retries": 0, "seconds": 96.2, "max_memory_gb": 17.03, "ce": 0.006636434191023, "kl": 0.06980620152171468, "context": 0.041539535915944725} +{"step": 217, "steps": 250, "lr": 7.30369112812219e-06, "grad_norm": 0.056760117411613464, "loss_scale": 4.0, "retries": 0, "seconds": 96.6, "max_memory_gb": 17.03, "ce": 0.002676919859368354, "kl": 0.004019138428702718, "context": 0.004686489060986787} +{"step": 218, "steps": 250, "lr": 7.172365221266403e-06, "grad_norm": 0.6022874712944031, "loss_scale": 4.0, "retries": 0, "seconds": 97.1, "max_memory_gb": 17.03, "ce": 0.16158835915848613, "kl": 0.24128401651978493, "context": 0.28223036974668503} +{"step": 219, "steps": 250, "lr": 7.0447032544387154e-06, "grad_norm": 0.32800039649009705, "loss_scale": 4.0, "retries": 0, "seconds": 97.5, "max_memory_gb": 17.03, "ce": 0.008013668200874235, "kl": 0.12004988442640752, "context": 0.0680386119929608} +{"step": 220, "steps": 250, "lr": 6.920728237979274e-06, "grad_norm": 0.3862667679786682, "loss_scale": 4.0, "retries": 0, "seconds": 97.9, "max_memory_gb": 17.03, "ce": 0.05245736882352503, "kl": 0.007584188860164431, "context": 0.05624946278112475} +{"step": 221, "steps": 250, "lr": 6.800462517676456e-06, "grad_norm": 0.685825765132904, "loss_scale": 4.0, "retries": 0, "seconds": 98.3, "max_memory_gb": 17.03, "ce": 0.004895298188785091, "kl": 0.35147889505606145, "context": 0.18063473975053057} +{"step": 222, "steps": 250, "lr": 6.683927770739222e-06, "grad_norm": 0.7714501023292542, "loss_scale": 4.0, "retries": 0, "seconds": 98.7, "max_memory_gb": 17.03, "ce": 0.050158682686742395, "kl": 0.06593817769135057, "context": 0.08312777307583019} +{"step": 223, "steps": 250, "lr": 6.571145001889892e-06, "grad_norm": 2.066169261932373, "loss_scale": 4.0, "retries": 0, "seconds": 99.2, "max_memory_gb": 17.03, "ce": 0.14881653164411546, "kl": 0.012514362124420586, "context": 0.15507371576313744} +{"step": 224, "steps": 250, "lr": 6.46213453957817e-06, "grad_norm": 1.075538992881775, "loss_scale": 4.0, "retries": 0, "seconds": 99.6, "max_memory_gb": 17.03, "ce": 0.0503439761523623, "kl": 0.12940231251377554, "context": 0.1150451367138885} +{"step": 225, "steps": 250, "lr": 6.356916032317063e-06, "grad_norm": 0.4573914408683777, "loss_scale": 4.0, "retries": 0, "seconds": 100.0, "max_memory_gb": 17.03, "ce": 0.0034533535363152623, "kl": 0.16257655280060135, "context": 0.08474162861239165} +{"step": 226, "steps": 250, "lr": 6.255508445141349e-06, "grad_norm": 0.5417963266372681, "loss_scale": 4.0, "retries": 0, "seconds": 100.4, "max_memory_gb": 17.03, "ce": 0.029309203804587014, "kl": 0.02707476970681455, "context": 0.04284658876713365} +{"step": 227, "steps": 250, "lr": 6.157930056189228e-06, "grad_norm": 0.24052166938781738, "loss_scale": 4.0, "retries": 0, "seconds": 100.8, "max_memory_gb": 17.03, "ce": 0.02697902946238173, "kl": 0.012231526166715412, "context": 0.03309479253948666} +{"step": 228, "steps": 250, "lr": 6.064198453407796e-06, "grad_norm": 0.6167959570884705, "loss_scale": 4.0, "retries": 0, "seconds": 101.2, "max_memory_gb": 17.03, "ce": 0.003786434550420381, "kl": 0.12579308165004477, "context": 0.06668297463329509} +{"step": 229, "steps": 250, "lr": 5.974330531382935e-06, "grad_norm": 0.5709551572799683, "loss_scale": 4.0, "retries": 0, "seconds": 101.7, "max_memory_gb": 17.03, "ce": 0.066492917285359, "kl": 0.040216479904984226, "context": 0.08660115630482323} +{"step": 230, "steps": 250, "lr": 5.888342488294125e-06, "grad_norm": 1.7019537687301636, "loss_scale": 4.0, "retries": 0, "seconds": 102.1, "max_memory_gb": 17.03, "ce": 0.6184547471930273, "kl": 0.35084958863444626, "context": 0.7938795284717344} +{"step": 231, "steps": 250, "lr": 5.806249822994849e-06, "grad_norm": 0.49168601632118225, "loss_scale": 4.0, "retries": 0, "seconds": 102.5, "max_memory_gb": 17.03, "ce": 0.038199090224225074, "kl": 0.13003486705565592, "context": 0.10321652445418295} +{"step": 232, "steps": 250, "lr": 5.728067332218994e-06, "grad_norm": 0.30870383977890015, "loss_scale": 4.0, "retries": 0, "seconds": 102.9, "max_memory_gb": 17.03, "ce": 0.028533693781355396, "kl": 0.008247519610449672, "context": 0.03265745419776067} +{"step": 233, "steps": 250, "lr": 5.65380910791383e-06, "grad_norm": 0.42206403613090515, "loss_scale": 4.0, "retries": 0, "seconds": 103.3, "max_memory_gb": 17.03, "ce": 0.01063293102197349, "kl": 0.10418614011723548, "context": 0.06272600218653679} +{"step": 234, "steps": 250, "lr": 5.583488534700022e-06, "grad_norm": 0.22887767851352692, "loss_scale": 4.0, "retries": 0, "seconds": 103.7, "max_memory_gb": 17.03, "ce": 0.008867055323207751, "kl": 0.021643368614604697, "context": 0.019688739368575625} +{"step": 235, "steps": 250, "lr": 5.517118287459118e-06, "grad_norm": 1.38666832447052, "loss_scale": 4.0, "retries": 0, "seconds": 104.1, "max_memory_gb": 17.03, "ce": 0.49075317502138205, "kl": 0.01695840112370206, "context": 0.499232382979244} +{"step": 236, "steps": 250, "lr": 5.454710329048978e-06, "grad_norm": 0.11291448771953583, "loss_scale": 4.0, "retries": 0, "seconds": 104.6, "max_memory_gb": 17.03, "ce": 0.009931797343597282, "kl": 0.013937149218691047, "context": 0.01690037218213547} +{"step": 237, "steps": 250, "lr": 5.3962759081475436e-06, "grad_norm": 0.23813332617282867, "loss_scale": 4.0, "retries": 0, "seconds": 105.0, "max_memory_gb": 17.03, "ce": 0.06159646951709874, "kl": 0.029016946004503552, "context": 0.07610493966785725} +{"step": 238, "steps": 250, "lr": 5.341825557225321e-06, "grad_norm": 0.20570287108421326, "loss_scale": 4.0, "retries": 0, "seconds": 105.4, "max_memory_gb": 17.03, "ce": 0.021645031025400385, "kl": 0.0060105731718067545, "context": 0.024650317413033918} +{"step": 239, "steps": 250, "lr": 5.291369090646961e-06, "grad_norm": 0.266798734664917, "loss_scale": 4.0, "retries": 0, "seconds": 105.8, "max_memory_gb": 17.03, "ce": 0.10413038835395128, "kl": 0.026976854613167234, "context": 0.11761881277197972} +{"step": 240, "steps": 250, "lr": 5.244915602902289e-06, "grad_norm": 0.6823749542236328, "loss_scale": 4.0, "retries": 0, "seconds": 106.2, "max_memory_gb": 17.03, "ce": 0.09551352561538806, "kl": 0.03468780497496482, "context": 0.11285742369364016} +{"step": 241, "steps": 250, "lr": 5.20247346696706e-06, "grad_norm": 1.0921671390533447, "loss_scale": 4.0, "retries": 0, "seconds": 106.6, "max_memory_gb": 17.03, "ce": 0.2379776086017955, "kl": 0.03334883469506167, "context": 0.2546520227115252} +{"step": 242, "steps": 250, "lr": 5.164050332793787e-06, "grad_norm": 0.16821356117725372, "loss_scale": 4.0, "retries": 0, "seconds": 107.0, "max_memory_gb": 17.03, "ce": 0.004588107512972783, "kl": 0.03732858348962509, "context": 0.02325239867786877} +{"step": 243, "steps": 250, "lr": 5.129653125932873e-06, "grad_norm": 0.08887011557817459, "loss_scale": 4.0, "retries": 0, "seconds": 107.5, "max_memory_gb": 17.03, "ce": 0.006061189633328468, "kl": 0.005029521475080401, "context": 0.008575950341764838} +{"step": 244, "steps": 250, "lr": 5.099288046284329e-06, "grad_norm": 0.35974183678627014, "loss_scale": 4.0, "retries": 0, "seconds": 107.9, "max_memory_gb": 17.03, "ce": 0.028948739913175814, "kl": 0.1389826595041086, "context": 0.09844007140782196} +{"step": 245, "steps": 250, "lr": 5.072960566980277e-06, "grad_norm": 0.9534862041473389, "loss_scale": 4.0, "retries": 0, "seconds": 108.3, "max_memory_gb": 17.03, "ce": 0.39290873729623854, "kl": 0.09546407300513238, "context": 0.4406407750211656} +{"step": 246, "steps": 250, "lr": 5.050675433398435e-06, "grad_norm": 0.021206054836511612, "loss_scale": 4.0, "retries": 0, "seconds": 108.7, "max_memory_gb": 17.03, "ce": 0.001529951608972624, "kl": 0.0007448694477716344, "context": 0.0019023862987523898} +{"step": 247, "steps": 250, "lr": 5.032436662306811e-06, "grad_norm": 0.6905167698860168, "loss_scale": 4.0, "retries": 0, "seconds": 109.1, "max_memory_gb": 17.03, "ce": 0.002328258167835884, "kl": 0.44038506851211423, "context": 0.2225207869487349} +{"step": 248, "steps": 250, "lr": 5.018247541139684e-06, "grad_norm": 0.2342296689748764, "loss_scale": 4.0, "retries": 0, "seconds": 109.6, "max_memory_gb": 17.03, "ce": 0.02894393193128053, "kl": 0.008247015066444874, "context": 0.03306744003202766} +{"step": 249, "steps": 250, "lr": 5.00811062740508e-06, "grad_norm": 0.6701333522796631, "loss_scale": 4.0, "retries": 0, "seconds": 110.0, "max_memory_gb": 17.03, "ce": 0.046999166370369494, "kl": 0.04473653715103865, "context": 0.06936743739061058} +{"step": 250, "steps": 250, "lr": 5.002027748223778e-06, "grad_norm": 0.09018982946872711, "loss_scale": 4.0, "retries": 0, "seconds": 110.4, "max_memory_gb": 17.03, "ce": 0.009507904745987616, "kl": 0.005324674028088339, "context": 0.012170241825515404} diff --git a/runs/single_internvl/context/L29_s13/adapter/README.md b/runs/single_internvl/context/L29_s13/adapter/README.md new file mode 100644 index 0000000000000000000000000000000000000000..4f9f856cf0e1531b47e25d266833b2977a19d3ac --- /dev/null +++ b/runs/single_internvl/context/L29_s13/adapter/README.md @@ -0,0 +1,207 @@ +--- +base_model: /data/OPSD/assets/models/InternVL3_5-8B-HF +library_name: peft +pipeline_tag: text-generation +tags: +- base_model:adapter:/data/OPSD/assets/models/InternVL3_5-8B-HF +- lora +- transformers +--- + +# Model Card for Model ID + + + + + +## Model Details + +### Model Description + + + + + +- **Developed by:** [More Information Needed] +- **Funded by [optional]:** [More Information Needed] +- **Shared by [optional]:** [More Information Needed] +- **Model type:** [More Information Needed] +- **Language(s) (NLP):** [More Information Needed] +- **License:** [More Information Needed] +- **Finetuned from model [optional]:** [More Information Needed] + +### Model Sources [optional] + + + +- **Repository:** [More Information Needed] +- **Paper [optional]:** [More Information Needed] +- **Demo [optional]:** [More Information Needed] + +## Uses + + + +### Direct Use + + + +[More Information Needed] + +### Downstream Use [optional] + + + +[More Information Needed] + +### Out-of-Scope Use + + + +[More Information Needed] + +## Bias, Risks, and Limitations + + + +[More Information Needed] + +### Recommendations + + + +Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. More information needed for further recommendations. + +## How to Get Started with the Model + +Use the code below to get started with the model. + +[More Information Needed] + +## Training Details + +### Training Data + + + +[More Information Needed] + +### Training Procedure + + + +#### Preprocessing [optional] + +[More Information Needed] + + +#### Training Hyperparameters + +- **Training regime:** [More Information Needed] + +#### Speeds, Sizes, Times [optional] + + + +[More Information Needed] + +## Evaluation + + + +### Testing Data, Factors & Metrics + +#### Testing Data + + + +[More Information Needed] + +#### Factors + + + +[More Information Needed] + +#### Metrics + + + +[More Information Needed] + +### Results + +[More Information Needed] + +#### Summary + + + +## Model Examination [optional] + + + +[More Information Needed] + +## Environmental Impact + + + +Carbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700). + +- **Hardware Type:** [More Information Needed] +- **Hours used:** [More Information Needed] +- **Cloud Provider:** [More Information Needed] +- **Compute Region:** [More Information Needed] +- **Carbon Emitted:** [More Information Needed] + +## Technical Specifications [optional] + +### Model Architecture and Objective + +[More Information Needed] + +### Compute Infrastructure + +[More Information Needed] + +#### Hardware + +[More Information Needed] + +#### Software + +[More Information Needed] + +## Citation [optional] + + + +**BibTeX:** + +[More Information Needed] + +**APA:** + +[More Information Needed] + +## Glossary [optional] + + + +[More Information Needed] + +## More Information [optional] + +[More Information Needed] + +## Model Card Authors [optional] + +[More Information Needed] + +## Model Card Contact + +[More Information Needed] +### Framework versions + +- PEFT 0.20.0 \ No newline at end of file diff --git a/runs/single_internvl/context/L29_s13/adapter/adapter_config.json b/runs/single_internvl/context/L29_s13/adapter/adapter_config.json new file mode 100644 index 0000000000000000000000000000000000000000..bc0708cdf52e1b3f333ccfa1b15116aa54e7dbcc --- /dev/null +++ b/runs/single_internvl/context/L29_s13/adapter/adapter_config.json @@ -0,0 +1,50 @@ +{ + "alora_invocation_tokens": null, + "alpha_pattern": {}, + "arrow_config": null, + "auto_mapping": null, + "base_model_name_or_path": "/data/OPSD/assets/models/InternVL3_5-8B-HF", + "bias": "none", + "corda_config": null, + "ensure_weight_tying": false, + "eva_config": null, + "exclude_modules": null, + "fan_in_fan_out": false, + "inference_mode": true, + "init_lora_weights": true, + "layer_replication": null, + "layers_pattern": null, + "layers_to_transform": null, + "loftq_config": {}, + "lora_alpha": 32, + "lora_bias": false, + "lora_dropout": 0.05, + "lora_ga_config": null, + "megatron_config": null, + "megatron_core": "megatron.core", + "modules_to_save": null, + "monteclora_config": null, + "peft_type": "LORA", + "peft_version": "0.20.0", + "qalora_group_size": 16, + "r": 16, + "rank_pattern": {}, + "revision": null, + "target_modules": [ + "model.language_model.layers.29.self_attn.q_proj", + "model.language_model.layers.29.self_attn.v_proj", + "model.language_model.layers.29.mlp.gate_proj", + "model.language_model.layers.29.self_attn.o_proj", + "model.language_model.layers.29.mlp.down_proj", + "model.language_model.layers.29.self_attn.k_proj", + "model.language_model.layers.29.mlp.up_proj" + ], + "target_parameters": null, + "task_type": "CAUSAL_LM", + "trainable_token_indices": null, + "use_bdlora": null, + "use_dora": false, + "use_qalora": false, + "use_rslora": false, + "velora_config": null +} \ No newline at end of file diff --git a/runs/single_internvl/context/L29_s13/config.json b/runs/single_internvl/context/L29_s13/config.json new file mode 100644 index 0000000000000000000000000000000000000000..209ba994998d062cc752b2d966329f2f829a15ff --- /dev/null +++ b/runs/single_internvl/context/L29_s13/config.json @@ -0,0 +1,24 @@ +{ + "behavior": "context", + "seed": 13, + "model": "/data/OPSD/assets/models/InternVL3_5-8B-HF", + "backbone": "internvl", + "train": "/data/OPSD/assets/v3/context.jsonl", + "items": 1000, + "steps": 250, + "epochs": 1, + "steps_per_epoch": 250, + "groups_per_step": 4, + "lr": 5e-05, + "scheduler": "cosine", + "warmup_steps": 16, + "final_fraction": 0.1, + "weight_decay": 0.0, + "clip": 1.0, + "initial_loss_scale": 4.0, + "adapter": { + "layer": 29, + "rank": 16, + "trainable_parameters": 1212416 + } +} \ No newline at end of file diff --git a/runs/single_internvl/context/L29_s13/steps.jsonl b/runs/single_internvl/context/L29_s13/steps.jsonl new file mode 100644 index 0000000000000000000000000000000000000000..6b5029f0f3fb5fd5732966a0986f75b8925623ac --- /dev/null +++ b/runs/single_internvl/context/L29_s13/steps.jsonl @@ -0,0 +1,250 @@ +{"step": 1, "steps": 250, "lr": 3.125e-06, "grad_norm": 0.2368045300245285, "loss_scale": 4.0, "retries": 0, "seconds": 1.4, "max_memory_gb": 16.69, "ce": 0.5227766726748087, "kl": 0.08654405153356493, "context": 0.5660487064160407} +{"step": 2, "steps": 250, "lr": 6.25e-06, "grad_norm": 0.327043354511261, "loss_scale": 4.0, "retries": 0, "seconds": 1.8, "max_memory_gb": 16.7, "ce": 0.6404375040219747, "kl": 0.46727469330153326, "context": 0.874074881321576} +{"step": 3, "steps": 250, "lr": 9.375000000000001e-06, "grad_norm": 0.769045889377594, "loss_scale": 4.0, "retries": 0, "seconds": 2.2, "max_memory_gb": 16.7, "ce": 1.0932224094867706, "kl": 0.06784487329423428, "context": 1.1271448582410812} +{"step": 4, "steps": 250, "lr": 1.25e-05, "grad_norm": 0.3907756507396698, "loss_scale": 4.0, "retries": 0, "seconds": 2.7, "max_memory_gb": 16.7, "ce": 0.451626935711829, "kl": 0.43012601514419657, "context": 0.6666899229676346} +{"step": 5, "steps": 250, "lr": 1.5625e-05, "grad_norm": 0.008808257058262825, "loss_scale": 4.0, "retries": 0, "seconds": 3.1, "max_memory_gb": 16.7, "ce": 0.009193224415867007, "kl": 0.008233361233578762, "context": 0.01330990514907171} +{"step": 6, "steps": 250, "lr": 1.8750000000000002e-05, "grad_norm": 0.25807544589042664, "loss_scale": 4.0, "retries": 0, "seconds": 3.6, "max_memory_gb": 16.7, "ce": 0.29110770285478793, "kl": 0.0742342584162543, "context": 0.3282248232717393} +{"step": 7, "steps": 250, "lr": 2.1875e-05, "grad_norm": 0.6202875971794128, "loss_scale": 4.0, "retries": 0, "seconds": 4.0, "max_memory_gb": 16.7, "ce": 0.8280366486087587, "kl": 0.1227430608563509, "context": 0.8894081706548604} +{"step": 8, "steps": 250, "lr": 2.5e-05, "grad_norm": 0.6656560301780701, "loss_scale": 4.0, "retries": 0, "seconds": 4.5, "max_memory_gb": 16.71, "ce": 0.795151861384511, "kl": 0.2897674401756376, "context": 0.9400355815887451} +{"step": 9, "steps": 250, "lr": 2.8125000000000003e-05, "grad_norm": 0.48557519912719727, "loss_scale": 4.0, "retries": 0, "seconds": 4.9, "max_memory_gb": 16.71, "ce": 0.531519902870059, "kl": 0.13788561860565096, "context": 0.6004627235233784} +{"step": 10, "steps": 250, "lr": 3.125e-05, "grad_norm": 0.11990529298782349, "loss_scale": 4.0, "retries": 0, "seconds": 5.4, "max_memory_gb": 16.71, "ce": 0.05340433833771385, "kl": 0.045304756436962634, "context": 0.0760567169636488} +{"step": 11, "steps": 250, "lr": 3.4375e-05, "grad_norm": 0.6163947582244873, "loss_scale": 4.0, "retries": 0, "seconds": 5.9, "max_memory_gb": 16.71, "ce": 0.7089882403961383, "kl": 0.004855293829677976, "context": 0.7114158784897882} +{"step": 12, "steps": 250, "lr": 3.7500000000000003e-05, "grad_norm": 0.4634014666080475, "loss_scale": 4.0, "retries": 0, "seconds": 6.3, "max_memory_gb": 16.71, "ce": 0.4747669590760779, "kl": 0.002514318157409434, "context": 0.47602413246568176} +{"step": 13, "steps": 250, "lr": 4.0625000000000005e-05, "grad_norm": 0.4194768965244293, "loss_scale": 4.0, "retries": 0, "seconds": 6.7, "max_memory_gb": 16.71, "ce": 0.42903357234899886, "kl": 0.019577729583033943, "context": 0.4388224463800725} +{"step": 14, "steps": 250, "lr": 4.375e-05, "grad_norm": 0.8380067944526672, "loss_scale": 4.0, "retries": 0, "seconds": 7.2, "max_memory_gb": 16.71, "ce": 0.7802784293889999, "kl": 0.9456784292124212, "context": 1.2531176656484604} +{"step": 15, "steps": 250, "lr": 4.6875e-05, "grad_norm": 0.7394029498100281, "loss_scale": 4.0, "retries": 0, "seconds": 7.6, "max_memory_gb": 16.71, "ce": 0.5328527746364671, "kl": 0.4689808783755325, "context": 0.7673432250001042} +{"step": 16, "steps": 250, "lr": 5e-05, "grad_norm": 0.32328182458877563, "loss_scale": 4.0, "retries": 0, "seconds": 8.1, "max_memory_gb": 16.71, "ce": 0.20833582297200337, "kl": 0.0009543582834794506, "context": 0.2088130015254137} +{"step": 17, "steps": 250, "lr": 5e-05, "grad_norm": 0.39328280091285706, "loss_scale": 4.0, "retries": 0, "seconds": 8.5, "max_memory_gb": 16.71, "ce": 0.33080662053544074, "kl": 0.8246882430976257, "context": 0.7431507483124733} +{"step": 18, "steps": 250, "lr": 4.999797225177623e-05, "grad_norm": 0.8903608918190002, "loss_scale": 4.0, "retries": 0, "seconds": 8.9, "max_memory_gb": 16.71, "ce": 0.5575144251051825, "kl": 0.20974945105990628, "context": 0.6623891585477395} +{"step": 19, "steps": 250, "lr": 4.999188937259492e-05, "grad_norm": 1.8549394607543945, "loss_scale": 4.0, "retries": 0, "seconds": 9.4, "max_memory_gb": 16.71, "ce": 0.8175516128540039, "kl": 0.1628367763478309, "context": 0.8989700078964233} +{"step": 20, "steps": 250, "lr": 4.998175245886032e-05, "grad_norm": 0.019542271271348, "loss_scale": 4.0, "retries": 0, "seconds": 9.8, "max_memory_gb": 16.71, "ce": 0.0023010416589386296, "kl": 0.02123097529693041, "context": 0.012916529067297233} +{"step": 21, "steps": 250, "lr": 4.996756333769319e-05, "grad_norm": 1.0111737251281738, "loss_scale": 4.0, "retries": 0, "seconds": 10.3, "max_memory_gb": 16.71, "ce": 0.38248012181884405, "kl": 0.048789063383082976, "context": 0.4068746656321309} +{"step": 22, "steps": 250, "lr": 4.994932456660157e-05, "grad_norm": 0.9598909020423889, "loss_scale": 4.0, "retries": 0, "seconds": 10.7, "max_memory_gb": 16.71, "ce": 0.423254440058372, "kl": 0.026696948742028326, "context": 0.43660291978449095} +{"step": 23, "steps": 250, "lr": 4.992703943301973e-05, "grad_norm": 1.3687055110931396, "loss_scale": 4.0, "retries": 0, "seconds": 11.2, "max_memory_gb": 16.71, "ce": 0.5924570706556551, "kl": 0.9890876810237614, "context": 1.0870009209465934} +{"step": 24, "steps": 250, "lr": 4.9900711953715675e-05, "grad_norm": 1.0363179445266724, "loss_scale": 4.0, "retries": 0, "seconds": 11.6, "max_memory_gb": 16.71, "ce": 0.3033435009419918, "kl": 0.3332179016433656, "context": 0.46995246410369873} +{"step": 25, "steps": 250, "lr": 4.987034687406713e-05, "grad_norm": 1.4582756757736206, "loss_scale": 4.0, "retries": 0, "seconds": 12.0, "max_memory_gb": 16.71, "ce": 0.7006395137796062, "kl": 0.21715917019173503, "context": 0.8092190958559513} +{"step": 26, "steps": 250, "lr": 4.9835949667206215e-05, "grad_norm": 1.3045154809951782, "loss_scale": 4.0, "retries": 0, "seconds": 12.5, "max_memory_gb": 16.71, "ce": 0.35067669302225113, "kl": 0.7001139237545431, "context": 0.7007336467504501} +{"step": 27, "steps": 250, "lr": 4.9797526533032945e-05, "grad_norm": 1.391527771949768, "loss_scale": 4.0, "retries": 0, "seconds": 12.9, "max_memory_gb": 16.71, "ce": 0.4600100731040584, "kl": 0.4283303649935988, "context": 0.6741752581001492} +{"step": 28, "steps": 250, "lr": 4.9755084397097716e-05, "grad_norm": 0.8539785742759705, "loss_scale": 4.0, "retries": 0, "seconds": 13.4, "max_memory_gb": 16.71, "ce": 0.6140342019498348, "kl": 0.9417435270734131, "context": 1.0849059522151947} +{"step": 29, "steps": 250, "lr": 4.970863090935304e-05, "grad_norm": 1.3099579811096191, "loss_scale": 4.0, "retries": 0, "seconds": 13.8, "max_memory_gb": 16.71, "ce": 0.3630714048631489, "kl": 0.3595145691651851, "context": 0.5428286790847778} +{"step": 30, "steps": 250, "lr": 4.965817444277468e-05, "grad_norm": 1.3632243871688843, "loss_scale": 4.0, "retries": 0, "seconds": 14.2, "max_memory_gb": 16.71, "ce": 0.3477202722351649, "kl": 0.007463099724645872, "context": 0.35145183234271826} +{"step": 31, "steps": 250, "lr": 4.9603724091852456e-05, "grad_norm": 1.9161518812179565, "loss_scale": 4.0, "retries": 0, "seconds": 14.7, "max_memory_gb": 16.71, "ce": 0.6607154347002506, "kl": 0.1619139164686203, "context": 0.7416723966598511} +{"step": 32, "steps": 250, "lr": 4.9545289670951025e-05, "grad_norm": 2.2006258964538574, "loss_scale": 4.0, "retries": 0, "seconds": 15.1, "max_memory_gb": 16.71, "ce": 0.3962809219956398, "kl": 0.015312236821046099, "context": 0.40393703803420067} +{"step": 33, "steps": 250, "lr": 4.948288171254089e-05, "grad_norm": 0.9485607147216797, "loss_scale": 4.0, "retries": 0, "seconds": 15.6, "max_memory_gb": 16.71, "ce": 0.21800466580316424, "kl": 0.07039269781671464, "context": 0.25320101249963045} +{"step": 34, "steps": 250, "lr": 4.941651146529998e-05, "grad_norm": 1.100024700164795, "loss_scale": 4.0, "retries": 0, "seconds": 16.0, "max_memory_gb": 16.71, "ce": 0.10886622611360508, "kl": 0.7500241809002546, "context": 0.48387830905630835} +{"step": 35, "steps": 250, "lr": 4.9346190892086174e-05, "grad_norm": 1.0855989456176758, "loss_scale": 4.0, "retries": 0, "seconds": 16.5, "max_memory_gb": 16.71, "ce": 0.1670604961545905, "kl": 0.004122340652429557, "context": 0.1691216647886904} +{"step": 36, "steps": 250, "lr": 4.927193266778101e-05, "grad_norm": 1.0965862274169922, "loss_scale": 4.0, "retries": 0, "seconds": 17.0, "max_memory_gb": 16.71, "ce": 0.25313852942053927, "kl": 0.05021007238246966, "context": 0.27824357282952406} +{"step": 37, "steps": 250, "lr": 4.919375017700515e-05, "grad_norm": 1.0173462629318237, "loss_scale": 4.0, "retries": 0, "seconds": 17.4, "max_memory_gb": 16.71, "ce": 0.2645794189738808, "kl": 0.7572606479516253, "context": 0.643209726200439} +{"step": 38, "steps": 250, "lr": 4.911165751170588e-05, "grad_norm": 1.0166475772857666, "loss_scale": 4.0, "retries": 0, "seconds": 17.8, "max_memory_gb": 16.71, "ce": 0.14396110828965902, "kl": 0.5090394093422219, "context": 0.3984808037057519} +{"step": 39, "steps": 250, "lr": 4.902566946861708e-05, "grad_norm": 0.3153465688228607, "loss_scale": 4.0, "retries": 0, "seconds": 18.2, "max_memory_gb": 16.71, "ce": 0.13841690152912633, "kl": 0.009311500383773819, "context": 0.1430726507678628} +{"step": 40, "steps": 250, "lr": 4.89358015465922e-05, "grad_norm": 0.4071977734565735, "loss_scale": 4.0, "retries": 0, "seconds": 18.7, "max_memory_gb": 16.71, "ce": 0.0360455556656234, "kl": 0.01092245674226433, "context": 0.04150678447331302} +{"step": 41, "steps": 250, "lr": 4.884206994381078e-05, "grad_norm": 0.5732959508895874, "loss_scale": 4.0, "retries": 0, "seconds": 19.1, "max_memory_gb": 16.71, "ce": 0.3898785043274984, "kl": 0.12497198674827814, "context": 0.4523644857108593} +{"step": 42, "steps": 250, "lr": 4.874449155485866e-05, "grad_norm": 0.7198561429977417, "loss_scale": 4.0, "retries": 0, "seconds": 19.5, "max_memory_gb": 16.71, "ce": 0.0629174248824711, "kl": 0.006988826884480659, "context": 0.06641183742613066} +{"step": 43, "steps": 250, "lr": 4.864308396768294e-05, "grad_norm": 0.7262969613075256, "loss_scale": 4.0, "retries": 0, "seconds": 20.0, "max_memory_gb": 16.71, "ce": 0.05828394429408945, "kl": 0.09589021779538598, "context": 0.1062290531990584} +{"step": 44, "steps": 250, "lr": 4.853786546042184e-05, "grad_norm": 1.2888867855072021, "loss_scale": 4.0, "retries": 0, "seconds": 20.4, "max_memory_gb": 16.71, "ce": 0.5736977792184916, "kl": 1.343023726662068, "context": 1.2452096462802729} +{"step": 45, "steps": 250, "lr": 4.842885499811012e-05, "grad_norm": 0.8340447545051575, "loss_scale": 4.0, "retries": 0, "seconds": 20.9, "max_memory_gb": 16.71, "ce": 0.8888028300134465, "kl": 0.24021743292541942, "context": 1.0089115366863552} +{"step": 46, "steps": 250, "lr": 4.831607222926078e-05, "grad_norm": 0.27493610978126526, "loss_scale": 4.0, "retries": 0, "seconds": 21.3, "max_memory_gb": 16.71, "ce": 0.08701110724359751, "kl": 0.27264758537057787, "context": 0.22333490825258195} +{"step": 47, "steps": 250, "lr": 4.8199537482323545e-05, "grad_norm": 0.37824442982673645, "loss_scale": 4.0, "retries": 0, "seconds": 21.8, "max_memory_gb": 16.71, "ce": 0.01760654376266757, "kl": 0.04345504107186571, "context": 0.03933406452415511} +{"step": 48, "steps": 250, "lr": 4.807927176202073e-05, "grad_norm": 0.5810438394546509, "loss_scale": 4.0, "retries": 0, "seconds": 22.2, "max_memory_gb": 16.71, "ce": 0.03545440763991792, "kl": 0.00795467289572116, "context": 0.03943174460437149} +{"step": 49, "steps": 250, "lr": 4.795529674556129e-05, "grad_norm": 0.747589111328125, "loss_scale": 4.0, "retries": 0, "seconds": 22.7, "max_memory_gb": 16.71, "ce": 0.0445215639701928, "kl": 0.31313182958592733, "context": 0.2010874750376388} +{"step": 50, "steps": 250, "lr": 4.782763477873361e-05, "grad_norm": 1.2603031396865845, "loss_scale": 4.0, "retries": 0, "seconds": 23.1, "max_memory_gb": 16.71, "ce": 0.13851953029734432, "kl": 0.011725601812941022, "context": 0.14438233120017685} +{"step": 51, "steps": 250, "lr": 4.769630887187782e-05, "grad_norm": 0.8463196754455566, "loss_scale": 4.0, "retries": 0, "seconds": 23.6, "max_memory_gb": 16.71, "ce": 0.5454443816597632, "kl": 0.020497909824030103, "context": 0.5556933388979814} +{"step": 52, "steps": 250, "lr": 4.756134269573832e-05, "grad_norm": 0.5768530964851379, "loss_scale": 4.0, "retries": 0, "seconds": 24.1, "max_memory_gb": 16.71, "ce": 0.0444487122585997, "kl": 0.21368631371296942, "context": 0.15129186701960862} +{"step": 53, "steps": 250, "lr": 4.7422760577197226e-05, "grad_norm": 0.9573751091957092, "loss_scale": 4.0, "retries": 0, "seconds": 24.5, "max_memory_gb": 16.71, "ce": 0.15706212702207267, "kl": 0.36911876406520605, "context": 0.34162150509655476} +{"step": 54, "steps": 250, "lr": 4.728058749488963e-05, "grad_norm": 0.5050751566886902, "loss_scale": 4.0, "retries": 0, "seconds": 25.0, "max_memory_gb": 16.71, "ce": 0.03886442378279753, "kl": 0.33992804048466496, "context": 0.20882845192681998} +{"step": 55, "steps": 250, "lr": 4.713484907470133e-05, "grad_norm": 0.1312277764081955, "loss_scale": 4.0, "retries": 0, "seconds": 25.4, "max_memory_gb": 16.71, "ce": 0.015492717400775291, "kl": 0.011091876755017438, "context": 0.02103865600656718} +{"step": 56, "steps": 250, "lr": 4.6985571585149876e-05, "grad_norm": 0.5078900456428528, "loss_scale": 4.0, "retries": 0, "seconds": 25.8, "max_memory_gb": 16.71, "ce": 0.025040312204509974, "kl": 0.27726728352718055, "context": 0.16367395734414458} +{"step": 57, "steps": 250, "lr": 4.6832781932649884e-05, "grad_norm": 0.29348644614219666, "loss_scale": 4.0, "retries": 0, "seconds": 26.3, "max_memory_gb": 16.71, "ce": 0.014825327656581067, "kl": 0.018851741904654773, "context": 0.024251197661214974} +{"step": 58, "steps": 250, "lr": 4.667650765666325e-05, "grad_norm": 0.10412448644638062, "loss_scale": 4.0, "retries": 0, "seconds": 26.7, "max_memory_gb": 16.71, "ce": 0.007554385181720136, "kl": 0.0011070561163109005, "context": 0.008107913443382131} +{"step": 59, "steps": 250, "lr": 4.651677692473538e-05, "grad_norm": 0.2867010831832886, "loss_scale": 4.0, "retries": 0, "seconds": 27.2, "max_memory_gb": 16.71, "ce": 0.02975150541169569, "kl": 0.05942369518743362, "context": 0.05946335208136588} +{"step": 60, "steps": 250, "lr": 4.635361852741808e-05, "grad_norm": 0.4637559950351715, "loss_scale": 4.0, "retries": 0, "seconds": 27.6, "max_memory_gb": 16.71, "ce": 0.05299242938053794, "kl": 0.0498827209812589, "context": 0.07793378975475207} +{"step": 61, "steps": 250, "lr": 4.6187061873080264e-05, "grad_norm": 1.3525638580322266, "loss_scale": 4.0, "retries": 0, "seconds": 28.1, "max_memory_gb": 16.71, "ce": 0.0235990472137928, "kl": 0.862341269152239, "context": 0.45476967887952924} +{"step": 62, "steps": 250, "lr": 4.601713698260727e-05, "grad_norm": 0.12171317636966705, "loss_scale": 4.0, "retries": 0, "seconds": 28.5, "max_memory_gb": 16.71, "ce": 0.028137307715951465, "kl": 0.008328624655405292, "context": 0.03230162005638704} +{"step": 63, "steps": 250, "lr": 4.5843874483989746e-05, "grad_norm": 0.5662246942520142, "loss_scale": 4.0, "retries": 0, "seconds": 29.0, "max_memory_gb": 16.71, "ce": 0.04043019510572776, "kl": 0.12189843076339457, "context": 0.1013794126920402} +{"step": 64, "steps": 250, "lr": 4.566730560680316e-05, "grad_norm": 0.16281436383724213, "loss_scale": 4.0, "retries": 0, "seconds": 29.4, "max_memory_gb": 16.71, "ce": 0.016869472339749336, "kl": 0.016335611027898267, "context": 0.025037277373485267} +{"step": 65, "steps": 250, "lr": 4.548746217657878e-05, "grad_norm": 0.6464684009552002, "loss_scale": 4.0, "retries": 0, "seconds": 29.9, "max_memory_gb": 16.71, "ce": 0.20452961651608348, "kl": 0.15183181827887893, "context": 0.2804455179721117} +{"step": 66, "steps": 250, "lr": 4.5304376609067396e-05, "grad_norm": 0.9418254494667053, "loss_scale": 4.0, "retries": 0, "seconds": 30.3, "max_memory_gb": 16.71, "ce": 0.1586373824961811, "kl": 0.06898525123745003, "context": 0.19313000070428643} +{"step": 67, "steps": 250, "lr": 4.5118081904396504e-05, "grad_norm": 0.09022855758666992, "loss_scale": 4.0, "retries": 0, "seconds": 30.8, "max_memory_gb": 16.71, "ce": 0.005944273114437237, "kl": 0.008874883002135903, "context": 0.010381714528193697} +{"step": 68, "steps": 250, "lr": 4.4928611641122225e-05, "grad_norm": 0.1474025696516037, "loss_scale": 4.0, "retries": 0, "seconds": 31.2, "max_memory_gb": 16.71, "ce": 0.01047287659093854, "kl": 0.043806473400763934, "context": 0.03237611320219003} +{"step": 69, "steps": 250, "lr": 4.473599997017701e-05, "grad_norm": 0.5302525162696838, "loss_scale": 4.0, "retries": 0, "seconds": 31.7, "max_memory_gb": 16.71, "ce": 0.2958965701982379, "kl": 0.028403541538864374, "context": 0.31009834236465394} +{"step": 70, "steps": 250, "lr": 4.4540281608714053e-05, "grad_norm": 0.0688260868191719, "loss_scale": 4.0, "retries": 0, "seconds": 32.1, "max_memory_gb": 16.71, "ce": 0.003104300209088251, "kl": 0.0008234810875364929, "context": 0.0035160408151568845} +{"step": 71, "steps": 250, "lr": 4.434149183384977e-05, "grad_norm": 0.12043743580579758, "loss_scale": 4.0, "retries": 0, "seconds": 32.6, "max_memory_gb": 16.71, "ce": 0.00810248024208704, "kl": 0.010576749264146201, "context": 0.013390854757744819} +{"step": 72, "steps": 250, "lr": 4.413966647630534e-05, "grad_norm": 0.47359955310821533, "loss_scale": 4.0, "retries": 0, "seconds": 33.0, "max_memory_gb": 16.71, "ce": 0.018363282666541636, "kl": 0.40850812022108585, "context": 0.22261733701452613} +{"step": 73, "steps": 250, "lr": 4.3934841913948323e-05, "grad_norm": 0.753589391708374, "loss_scale": 4.0, "retries": 0, "seconds": 33.4, "max_memory_gb": 16.71, "ce": 0.3673154679345316, "kl": 0.10377357929974096, "context": 0.4192022445749899} +{"step": 74, "steps": 250, "lr": 4.372705506523583e-05, "grad_norm": 0.15636923909187317, "loss_scale": 4.0, "retries": 0, "seconds": 33.9, "max_memory_gb": 16.71, "ce": 0.061261469214514364, "kl": 0.12201028881463571, "context": 0.12226661739805422} +{"step": 75, "steps": 250, "lr": 4.351634338256017e-05, "grad_norm": 0.5778779983520508, "loss_scale": 4.0, "retries": 0, "seconds": 34.3, "max_memory_gb": 16.71, "ce": 0.45779934019083157, "kl": 1.000898611266166, "context": 0.9582486720755696} +{"step": 76, "steps": 250, "lr": 4.330274484549825e-05, "grad_norm": 1.1487363576889038, "loss_scale": 4.0, "retries": 0, "seconds": 34.8, "max_memory_gb": 16.71, "ce": 0.011480047221994027, "kl": 1.7031282427487895, "context": 0.8630441925488412} +{"step": 77, "steps": 250, "lr": 4.308629795396599e-05, "grad_norm": 0.10541275143623352, "loss_scale": 4.0, "retries": 0, "seconds": 35.2, "max_memory_gb": 16.71, "ce": 0.006289156423008535, "kl": 0.010978091879223939, "context": 0.011778202417190187} +{"step": 78, "steps": 250, "lr": 4.286704172127894e-05, "grad_norm": 0.39423996210098267, "loss_scale": 4.0, "retries": 0, "seconds": 35.6, "max_memory_gb": 16.71, "ce": 0.0013913031871197745, "kl": 0.07212100806646049, "context": 0.03745180694386363} +{"step": 79, "steps": 250, "lr": 4.264501566712033e-05, "grad_norm": 0.15400977432727814, "loss_scale": 4.0, "retries": 0, "seconds": 36.1, "max_memory_gb": 16.71, "ce": 0.018754970208647137, "kl": 0.03447577588440254, "context": 0.03599285818017961} +{"step": 80, "steps": 250, "lr": 4.2420259810417894e-05, "grad_norm": 0.4575210213661194, "loss_scale": 4.0, "retries": 0, "seconds": 36.5, "max_memory_gb": 16.71, "ce": 0.003171216400915, "kl": 0.08595484515484486, "context": 0.04614864039740496} +{"step": 81, "steps": 250, "lr": 4.219281466213066e-05, "grad_norm": 0.514958918094635, "loss_scale": 4.0, "retries": 0, "seconds": 36.9, "max_memory_gb": 16.71, "ce": 0.0736928636324592, "kl": 0.06157279672333971, "context": 0.10447926499182358} +{"step": 82, "steps": 250, "lr": 4.1962721217947135e-05, "grad_norm": 0.07922427356243134, "loss_scale": 4.0, "retries": 0, "seconds": 37.4, "max_memory_gb": 16.71, "ce": 0.00797699237591587, "kl": 0.002039927025180077, "context": 0.008996955846669152} +{"step": 83, "steps": 250, "lr": 4.173002095089599e-05, "grad_norm": 1.4993987083435059, "loss_scale": 4.0, "retries": 0, "seconds": 37.8, "max_memory_gb": 16.71, "ce": 0.1598056461662054, "kl": 0.7504231357888784, "context": 0.5350172028993256} +{"step": 84, "steps": 250, "lr": 4.149475580387084e-05, "grad_norm": 0.4829157292842865, "loss_scale": 4.0, "retries": 0, "seconds": 38.3, "max_memory_gb": 16.71, "ce": 0.02589954252289317, "kl": 0.0441833840104664, "context": 0.04799123632164992} +{"step": 85, "steps": 250, "lr": 4.1256968182070275e-05, "grad_norm": 0.7096356153488159, "loss_scale": 4.0, "retries": 0, "seconds": 38.8, "max_memory_gb": 16.71, "ce": 0.03429349738962628, "kl": 0.02974953286744153, "context": 0.04916826568705801} +{"step": 86, "steps": 250, "lr": 4.101670094535453e-05, "grad_norm": 1.2602087259292603, "loss_scale": 4.0, "retries": 0, "seconds": 39.2, "max_memory_gb": 16.71, "ce": 0.217115031555295, "kl": 0.45480364421382546, "context": 0.4445168530801311} +{"step": 87, "steps": 250, "lr": 4.077399740052027e-05, "grad_norm": 0.7413747310638428, "loss_scale": 4.0, "retries": 0, "seconds": 39.6, "max_memory_gb": 16.71, "ce": 0.006965069842408411, "kl": 0.7334089117357507, "context": 0.3736695144325495} +{"step": 88, "steps": 250, "lr": 4.0528901293494775e-05, "grad_norm": 0.17610470950603485, "loss_scale": 4.0, "retries": 0, "seconds": 40.1, "max_memory_gb": 16.71, "ce": 0.008511748852470191, "kl": 0.020101023950588548, "context": 0.01856226056588639} +{"step": 89, "steps": 250, "lr": 4.0281456801451004e-05, "grad_norm": 0.0433182418346405, "loss_scale": 4.0, "retries": 0, "seconds": 40.5, "max_memory_gb": 16.71, "ce": 0.012362602326447814, "kl": 0.010057802206574706, "context": 0.017391503433827893} +{"step": 90, "steps": 250, "lr": 4.0031708524844884e-05, "grad_norm": 0.056681931018829346, "loss_scale": 4.0, "retries": 0, "seconds": 40.9, "max_memory_gb": 16.71, "ce": 0.006081424071453512, "kl": 0.004243359493557364, "context": 0.00820310364360921} +{"step": 91, "steps": 250, "lr": 3.977970147937635e-05, "grad_norm": 0.6189141273498535, "loss_scale": 4.0, "retries": 0, "seconds": 41.4, "max_memory_gb": 16.71, "ce": 0.23181540091900388, "kl": 0.5263204759685323, "context": 0.4949756344794878} +{"step": 92, "steps": 250, "lr": 3.952548108787553e-05, "grad_norm": 1.274174690246582, "loss_scale": 4.0, "retries": 0, "seconds": 41.8, "max_memory_gb": 16.71, "ce": 0.3520932191313477, "kl": 0.48381724672526616, "context": 0.5940018416185922} +{"step": 93, "steps": 250, "lr": 3.9269093172115496e-05, "grad_norm": 0.7131308913230896, "loss_scale": 4.0, "retries": 0, "seconds": 42.3, "max_memory_gb": 16.71, "ce": 0.07104845946014393, "kl": 0.48102535076759523, "context": 0.3115611294924747} +{"step": 94, "steps": 250, "lr": 3.901058394455323e-05, "grad_norm": 1.2233120203018188, "loss_scale": 4.0, "retries": 0, "seconds": 42.7, "max_memory_gb": 16.71, "ce": 0.18158194301940966, "kl": 0.10346697343629785, "context": 0.23331542624509893} +{"step": 95, "steps": 250, "lr": 3.875e-05, "grad_norm": 1.4160939455032349, "loss_scale": 4.0, "retries": 0, "seconds": 43.2, "max_memory_gb": 16.71, "ce": 0.500761864852393, "kl": 0.030716669163666666, "context": 0.5161202121526003} +{"step": 96, "steps": 250, "lr": 3.8487388307223005e-05, "grad_norm": 0.7726480960845947, "loss_scale": 4.0, "retries": 0, "seconds": 43.6, "max_memory_gb": 16.71, "ce": 0.46687257062876597, "kl": 0.006340614743749029, "context": 0.47004288247262593} +{"step": 97, "steps": 250, "lr": 3.822279620047943e-05, "grad_norm": 1.1073535680770874, "loss_scale": 4.0, "retries": 0, "seconds": 44.1, "max_memory_gb": 16.71, "ce": 0.015857250669796485, "kl": 0.395851100445725, "context": 0.21378279710188508} +{"step": 98, "steps": 250, "lr": 3.795627137098479e-05, "grad_norm": 0.3725271224975586, "loss_scale": 4.0, "retries": 0, "seconds": 44.5, "max_memory_gb": 16.71, "ce": 0.0031024569761939347, "kl": 0.08320402721801656, "context": 0.04470447197672911} +{"step": 99, "steps": 250, "lr": 3.7687861858316844e-05, "grad_norm": 0.11662140488624573, "loss_scale": 4.0, "retries": 0, "seconds": 45.0, "max_memory_gb": 16.71, "ce": 0.014580606004528818, "kl": 0.009454738430576981, "context": 0.019307975802803412} +{"step": 100, "steps": 250, "lr": 3.741761604175668e-05, "grad_norm": 0.7715573310852051, "loss_scale": 4.0, "retries": 0, "seconds": 45.4, "max_memory_gb": 16.71, "ce": 0.02947282303648535, "kl": 0.6977854511351325, "context": 0.378365551281604} +{"step": 101, "steps": 250, "lr": 3.714558263156872e-05, "grad_norm": 0.17554689943790436, "loss_scale": 4.0, "retries": 0, "seconds": 45.9, "max_memory_gb": 16.71, "ce": 0.004645833309041336, "kl": 0.09139214095193893, "context": 0.05034190486185253} +{"step": 102, "steps": 250, "lr": 3.6871810660220904e-05, "grad_norm": 0.5939478874206543, "loss_scale": 4.0, "retries": 0, "seconds": 46.3, "max_memory_gb": 16.71, "ce": 0.04481267894152552, "kl": 0.027034853002987802, "context": 0.05833010550122708} +{"step": 103, "steps": 250, "lr": 3.659634947354686e-05, "grad_norm": 0.1252141147851944, "loss_scale": 4.0, "retries": 0, "seconds": 46.7, "max_memory_gb": 16.71, "ce": 0.010280071532179136, "kl": 0.020277320763852913, "context": 0.02041873225243762} +{"step": 104, "steps": 250, "lr": 3.631924872185169e-05, "grad_norm": 0.0947798565030098, "loss_scale": 4.0, "retries": 0, "seconds": 47.2, "max_memory_gb": 16.71, "ce": 0.0047238194674719125, "kl": 0.008344574938291771, "context": 0.00889610669673857} +{"step": 105, "steps": 250, "lr": 3.6040558350962674e-05, "grad_norm": 0.634733259677887, "loss_scale": 4.0, "retries": 0, "seconds": 47.6, "max_memory_gb": 16.71, "ce": 0.05101200641365722, "kl": 0.40641220181714743, "context": 0.2542181061580777} +{"step": 106, "steps": 250, "lr": 3.576032859322687e-05, "grad_norm": 0.03221597149968147, "loss_scale": 4.0, "retries": 0, "seconds": 48.1, "max_memory_gb": 16.71, "ce": 0.002580111045972444, "kl": 0.0014188248405844206, "context": 0.003289523534476757} +{"step": 107, "steps": 250, "lr": 3.5478609958457056e-05, "grad_norm": 2.7471413612365723, "loss_scale": 4.0, "retries": 0, "seconds": 48.5, "max_memory_gb": 16.71, "ce": 0.34437734522362007, "kl": 0.14398403624477396, "context": 0.41636937403382035} +{"step": 108, "steps": 250, "lr": 3.519545322482755e-05, "grad_norm": 0.7013013362884521, "loss_scale": 4.0, "retries": 0, "seconds": 49.0, "max_memory_gb": 16.71, "ce": 0.04860791179089574, "kl": 0.013260778294224451, "context": 0.05523830228776205} +{"step": 109, "steps": 250, "lr": 3.491090942972179e-05, "grad_norm": 0.9138124585151672, "loss_scale": 4.0, "retries": 0, "seconds": 49.4, "max_memory_gb": 16.71, "ce": 0.05922757138614543, "kl": 0.29234934551641345, "context": 0.20540223829448223} +{"step": 110, "steps": 250, "lr": 3.4625029860533135e-05, "grad_norm": 1.3075295686721802, "loss_scale": 4.0, "retries": 0, "seconds": 49.8, "max_memory_gb": 16.71, "ce": 0.08226949721574783, "kl": 0.27057445084210485, "context": 0.21755672874860466} +{"step": 111, "steps": 250, "lr": 3.433786604542056e-05, "grad_norm": 0.015218937769532204, "loss_scale": 4.0, "retries": 0, "seconds": 50.3, "max_memory_gb": 16.71, "ce": 0.0011360972375769052, "kl": 0.0010239974653813988, "context": 0.0016480959166074172} +{"step": 112, "steps": 250, "lr": 3.404946974402104e-05, "grad_norm": 0.05767499655485153, "loss_scale": 4.0, "retries": 0, "seconds": 50.7, "max_memory_gb": 16.71, "ce": 0.004154552392719779, "kl": 0.0006125797162894742, "context": 0.004460842224943917} +{"step": 113, "steps": 250, "lr": 3.375989293812019e-05, "grad_norm": 0.9578631520271301, "loss_scale": 4.0, "retries": 0, "seconds": 51.2, "max_memory_gb": 16.71, "ce": 0.006171071727294475, "kl": 0.2911490532569587, "context": 0.1517456010915339} +{"step": 114, "steps": 250, "lr": 3.3469187822282766e-05, "grad_norm": 1.383124589920044, "loss_scale": 4.0, "retries": 0, "seconds": 51.6, "max_memory_gb": 16.71, "ce": 0.010361607099184766, "kl": 0.9193729919497855, "context": 0.4700480934116058} +{"step": 115, "steps": 250, "lr": 3.317740679444499e-05, "grad_norm": 0.18402741849422455, "loss_scale": 4.0, "retries": 0, "seconds": 52.1, "max_memory_gb": 16.71, "ce": 0.005406147425674135, "kl": 0.032201576887018746, "context": 0.021506935721845366} +{"step": 116, "steps": 250, "lr": 3.2884602446470056e-05, "grad_norm": 1.069904088973999, "loss_scale": 4.0, "retries": 0, "seconds": 52.5, "max_memory_gb": 16.71, "ce": 0.10144173370099452, "kl": 0.5243120627246185, "context": 0.36359776506469643} +{"step": 117, "steps": 250, "lr": 3.259082755466876e-05, "grad_norm": 1.5753618478775024, "loss_scale": 4.0, "retries": 0, "seconds": 52.9, "max_memory_gb": 16.71, "ce": 0.05465306757832877, "kl": 0.10171535922836483, "context": 0.10551074721297482} +{"step": 118, "steps": 250, "lr": 3.229613507028686e-05, "grad_norm": 0.46985381841659546, "loss_scale": 4.0, "retries": 0, "seconds": 53.4, "max_memory_gb": 16.71, "ce": 0.21198910542443627, "kl": 0.008683587510176949, "context": 0.21633089733950328} +{"step": 119, "steps": 250, "lr": 3.2000578109961004e-05, "grad_norm": 0.13115957379341125, "loss_scale": 4.0, "retries": 0, "seconds": 53.8, "max_memory_gb": 16.71, "ce": 0.01580141131853452, "kl": 0.021349271857843632, "context": 0.026476046739844605} +{"step": 120, "steps": 250, "lr": 3.1704209946144616e-05, "grad_norm": 2.2988944053649902, "loss_scale": 4.0, "retries": 0, "seconds": 54.3, "max_memory_gb": 16.71, "ce": 0.42899152600148227, "kl": 0.16052397682506125, "context": 0.5092534966242965} +{"step": 121, "steps": 250, "lr": 3.140708399750594e-05, "grad_norm": 0.19642528891563416, "loss_scale": 4.0, "retries": 0, "seconds": 54.8, "max_memory_gb": 16.71, "ce": 0.060893563448189525, "kl": 0.038475211476907134, "context": 0.08013116963775246} +{"step": 122, "steps": 250, "lr": 3.1109253819299606e-05, "grad_norm": 0.5918744206428528, "loss_scale": 4.0, "retries": 0, "seconds": 55.2, "max_memory_gb": 16.71, "ce": 0.09097812068648636, "kl": 0.019989633190562017, "context": 0.10097293716535205} +{"step": 123, "steps": 250, "lr": 3.081077309371361e-05, "grad_norm": 0.3699270188808441, "loss_scale": 4.0, "retries": 0, "seconds": 55.6, "max_memory_gb": 16.71, "ce": 0.030098459334112704, "kl": 0.019091302761808038, "context": 0.03964411071501672} +{"step": 124, "steps": 250, "lr": 3.051169562019339e-05, "grad_norm": 1.3693232536315918, "loss_scale": 4.0, "retries": 0, "seconds": 56.1, "max_memory_gb": 16.71, "ce": 0.734637500892859, "kl": 0.6113216503581498, "context": 1.040298349398654} +{"step": 125, "steps": 250, "lr": 3.021207530574477e-05, "grad_norm": 0.15954823791980743, "loss_scale": 4.0, "retries": 0, "seconds": 56.5, "max_memory_gb": 16.71, "ce": 0.11157838342478499, "kl": 0.07891165509499842, "context": 0.15103420364903286} +{"step": 126, "steps": 250, "lr": 2.991196615521758e-05, "grad_norm": 1.4899884462356567, "loss_scale": 4.0, "retries": 0, "seconds": 57.0, "max_memory_gb": 16.71, "ce": 0.6607798151671886, "kl": 0.12692392198368907, "context": 0.7242417596280575} +{"step": 127, "steps": 250, "lr": 2.9611422261571536e-05, "grad_norm": 0.05602286756038666, "loss_scale": 4.0, "retries": 0, "seconds": 57.4, "max_memory_gb": 16.71, "ce": 0.004240061403834261, "kl": 0.0006616969058086397, "context": 0.00457090996496845} +{"step": 128, "steps": 250, "lr": 2.931049779612634e-05, "grad_norm": 2.9427216053009033, "loss_scale": 4.0, "retries": 0, "seconds": 57.8, "max_memory_gb": 16.71, "ce": 0.16898957910598256, "kl": 1.0649576223513577, "context": 0.7014683860470541} +{"step": 129, "steps": 250, "lr": 2.9009246998797557e-05, "grad_norm": 0.7505726218223572, "loss_scale": 4.0, "retries": 0, "seconds": 58.3, "max_memory_gb": 16.71, "ce": 0.07602168811718002, "kl": 0.03306893883200246, "context": 0.09255616122391075} +{"step": 130, "steps": 250, "lr": 2.8707724168320238e-05, "grad_norm": 0.4982530474662781, "loss_scale": 4.0, "retries": 0, "seconds": 58.7, "max_memory_gb": 16.71, "ce": 0.04235470632556826, "kl": 0.2709699947772606, "context": 0.17783969570882618} +{"step": 131, "steps": 250, "lr": 2.840598365246184e-05, "grad_norm": 1.537609577178955, "loss_scale": 4.0, "retries": 0, "seconds": 59.2, "max_memory_gb": 16.71, "ce": 0.21041193984274287, "kl": 0.30502206430537626, "context": 0.36292297317413613} +{"step": 132, "steps": 250, "lr": 2.8104079838226383e-05, "grad_norm": 1.28470778465271, "loss_scale": 4.0, "retries": 0, "seconds": 59.6, "max_memory_gb": 16.71, "ce": 0.19866602565161884, "kl": 0.02989489381434396, "context": 0.21361347768106498} +{"step": 133, "steps": 250, "lr": 2.7802067142051446e-05, "grad_norm": 0.8809653520584106, "loss_scale": 4.0, "retries": 0, "seconds": 60.0, "max_memory_gb": 16.71, "ce": 0.38239821937168017, "kl": 0.05607438343577087, "context": 0.41043541952967644} +{"step": 134, "steps": 250, "lr": 2.7500000000000004e-05, "grad_norm": 0.5466724634170532, "loss_scale": 4.0, "retries": 0, "seconds": 60.5, "max_memory_gb": 16.71, "ce": 0.16299361077471985, "kl": 0.002304727406198026, "context": 0.16414596912363777} +{"step": 135, "steps": 250, "lr": 2.7197932857948566e-05, "grad_norm": 0.553953230381012, "loss_scale": 4.0, "retries": 0, "seconds": 60.9, "max_memory_gb": 16.71, "ce": 0.009703548159450293, "kl": 0.10276824159518583, "context": 0.06108766995021142} +{"step": 136, "steps": 250, "lr": 2.689592016177363e-05, "grad_norm": 1.234946846961975, "loss_scale": 4.0, "retries": 0, "seconds": 61.3, "max_memory_gb": 16.71, "ce": 0.05667556058324408, "kl": 0.07236247430773801, "context": 0.0928568017843645} +{"step": 137, "steps": 250, "lr": 2.659401634753817e-05, "grad_norm": 0.3837984800338745, "loss_scale": 4.0, "retries": 0, "seconds": 61.8, "max_memory_gb": 16.71, "ce": 0.014168832887662575, "kl": 0.05299561898573302, "context": 0.04066664236597717} +{"step": 138, "steps": 250, "lr": 2.6292275831679764e-05, "grad_norm": 0.8844560980796814, "loss_scale": 4.0, "retries": 0, "seconds": 62.3, "max_memory_gb": 16.71, "ce": 0.06040463445242494, "kl": 0.11135801766067743, "context": 0.11608364339917898} +{"step": 139, "steps": 250, "lr": 2.5990753001202445e-05, "grad_norm": 1.0091288089752197, "loss_scale": 4.0, "retries": 0, "seconds": 62.7, "max_memory_gb": 16.71, "ce": 0.003378009765583556, "kl": 0.22437468711268593, "context": 0.11556535107956734} +{"step": 140, "steps": 250, "lr": 2.5689502203873672e-05, "grad_norm": 0.08126889169216156, "loss_scale": 4.0, "retries": 0, "seconds": 63.2, "max_memory_gb": 16.71, "ce": 0.0047739853253006, "kl": 0.0028628174532059347, "context": 0.006205394180142321} +{"step": 141, "steps": 250, "lr": 2.538857773842847e-05, "grad_norm": 0.04643251374363899, "loss_scale": 4.0, "retries": 0, "seconds": 63.6, "max_memory_gb": 16.71, "ce": 0.003603604760428425, "kl": 0.0034481337534089107, "context": 0.005327671664417721} +{"step": 142, "steps": 250, "lr": 2.5088033844782426e-05, "grad_norm": 0.21796993911266327, "loss_scale": 4.0, "retries": 0, "seconds": 64.1, "max_memory_gb": 16.71, "ce": 0.014615579864766914, "kl": 0.008864728617481887, "context": 0.019047943991608918} +{"step": 143, "steps": 250, "lr": 2.4787924694255237e-05, "grad_norm": 1.8241688013076782, "loss_scale": 4.0, "retries": 0, "seconds": 64.5, "max_memory_gb": 16.71, "ce": 0.339815295432345, "kl": 0.08667577264941428, "context": 0.38315318897366524} +{"step": 144, "steps": 250, "lr": 2.4488304379806627e-05, "grad_norm": 0.7050882577896118, "loss_scale": 4.0, "retries": 0, "seconds": 65.0, "max_memory_gb": 16.71, "ce": 0.035232391936006024, "kl": 0.2145392156962771, "context": 0.14250200189417228} +{"step": 145, "steps": 250, "lr": 2.4189226906286388e-05, "grad_norm": 0.07749767601490021, "loss_scale": 4.0, "retries": 0, "seconds": 65.4, "max_memory_gb": 16.71, "ce": 0.001700727385468781, "kl": 0.008759223919696524, "context": 0.0060803393280366436} +{"step": 146, "steps": 250, "lr": 2.3890746180700395e-05, "grad_norm": 0.18957076966762543, "loss_scale": 4.0, "retries": 0, "seconds": 65.9, "max_memory_gb": 16.71, "ce": 0.008986407774500549, "kl": 0.013926351761256228, "context": 0.015949583961628377} +{"step": 147, "steps": 250, "lr": 2.359291600249407e-05, "grad_norm": 1.4763610363006592, "loss_scale": 4.0, "retries": 0, "seconds": 66.3, "max_memory_gb": 16.71, "ce": 0.5092421722511062, "kl": 0.484220547194127, "context": 0.7513524332462111} +{"step": 148, "steps": 250, "lr": 2.3295790053855392e-05, "grad_norm": 0.08148222416639328, "loss_scale": 4.0, "retries": 0, "seconds": 66.7, "max_memory_gb": 16.71, "ce": 0.004460379013835336, "kl": 0.011859442092827521, "context": 0.010390099756477866} +{"step": 149, "steps": 250, "lr": 2.299942189003901e-05, "grad_norm": 0.07100266963243484, "loss_scale": 4.0, "retries": 0, "seconds": 67.2, "max_memory_gb": 16.71, "ce": 0.0078302816618816, "kl": 0.0020040814706590027, "context": 0.008832322404487059} +{"step": 150, "steps": 250, "lr": 2.2703864929713144e-05, "grad_norm": 0.20121361315250397, "loss_scale": 4.0, "retries": 0, "seconds": 67.6, "max_memory_gb": 16.71, "ce": 0.03698073371197097, "kl": 0.025577153894118965, "context": 0.04976930859265849} +{"step": 151, "steps": 250, "lr": 2.2409172445331257e-05, "grad_norm": 0.31113266944885254, "loss_scale": 4.0, "retries": 0, "seconds": 68.1, "max_memory_gb": 16.71, "ce": 0.011391480704332935, "kl": 0.016186903050765977, "context": 0.01948493222880643} +{"step": 152, "steps": 250, "lr": 2.211539755352996e-05, "grad_norm": 1.2242207527160645, "loss_scale": 4.0, "retries": 0, "seconds": 68.5, "max_memory_gb": 16.71, "ce": 0.010872962331632152, "kl": 0.47305833307200373, "context": 0.24740212771575898} +{"step": 153, "steps": 250, "lr": 2.1822593205555016e-05, "grad_norm": 0.7952424883842468, "loss_scale": 4.0, "retries": 0, "seconds": 68.9, "max_memory_gb": 16.71, "ce": 0.0072291401847905945, "kl": 0.4160452163305308, "context": 0.2152517509966856} +{"step": 154, "steps": 250, "lr": 2.1530812177717236e-05, "grad_norm": 0.8161841034889221, "loss_scale": 4.0, "retries": 0, "seconds": 69.4, "max_memory_gb": 16.71, "ce": 0.06713232496986166, "kl": 0.07820802432252094, "context": 0.10623634164221585} +{"step": 155, "steps": 250, "lr": 2.1240107061879815e-05, "grad_norm": 1.5322774648666382, "loss_scale": 4.0, "retries": 0, "seconds": 69.8, "max_memory_gb": 16.71, "ce": 0.22141521272351383, "kl": 0.011130610574582533, "context": 0.22698051802581176} +{"step": 156, "steps": 250, "lr": 2.095053025597896e-05, "grad_norm": 0.18527641892433167, "loss_scale": 4.0, "retries": 0, "seconds": 70.3, "max_memory_gb": 16.71, "ce": 0.010600751425954513, "kl": 0.005980090337061483, "context": 0.013590796377684455} +{"step": 157, "steps": 250, "lr": 2.066213395457945e-05, "grad_norm": 0.24941034615039825, "loss_scale": 4.0, "retries": 0, "seconds": 70.7, "max_memory_gb": 16.71, "ce": 0.011947727238293737, "kl": 0.026220194915367756, "context": 0.02505782520165667} +{"step": 158, "steps": 250, "lr": 2.0374970139466877e-05, "grad_norm": 0.8233588337898254, "loss_scale": 4.0, "retries": 0, "seconds": 71.2, "max_memory_gb": 16.71, "ce": 0.18439486376882996, "kl": 0.0673708093672758, "context": 0.2180802694929298} +{"step": 159, "steps": 250, "lr": 2.0089090570278213e-05, "grad_norm": 0.3812306523323059, "loss_scale": 4.0, "retries": 0, "seconds": 71.6, "max_memory_gb": 16.71, "ce": 0.005530665173864691, "kl": 0.08463068748824298, "context": 0.04784600998391397} +{"step": 160, "steps": 250, "lr": 1.9804546775172457e-05, "grad_norm": 0.850827157497406, "loss_scale": 4.0, "retries": 0, "seconds": 72.1, "max_memory_gb": 16.71, "ce": 0.03811356643564068, "kl": 0.3835885763692204, "context": 0.22990785795263946} +{"step": 161, "steps": 250, "lr": 1.9521390041542946e-05, "grad_norm": 0.6899948120117188, "loss_scale": 4.0, "retries": 0, "seconds": 72.5, "max_memory_gb": 16.71, "ce": 0.0813293045212049, "kl": 0.08082721672963089, "context": 0.12174291288101813} +{"step": 162, "steps": 250, "lr": 1.923967140677313e-05, "grad_norm": 1.1962462663650513, "loss_scale": 4.0, "retries": 0, "seconds": 73.0, "max_memory_gb": 16.71, "ce": 0.0927574787492631, "kl": 0.1785185935877962, "context": 0.18201677556498908} +{"step": 163, "steps": 250, "lr": 1.8959441649037335e-05, "grad_norm": 2.5697832107543945, "loss_scale": 4.0, "retries": 0, "seconds": 73.3, "max_memory_gb": 16.71, "ce": 0.2330172373331152, "kl": 0.6724347355775535, "context": 0.5692346016294323} +{"step": 164, "steps": 250, "lr": 1.868075127814832e-05, "grad_norm": 0.7112966179847717, "loss_scale": 4.0, "retries": 0, "seconds": 73.7, "max_memory_gb": 16.71, "ce": 0.08604453012230806, "kl": 0.07414408546173945, "context": 0.12311657308600843} +{"step": 165, "steps": 250, "lr": 1.8403650526453143e-05, "grad_norm": 0.10644535720348358, "loss_scale": 4.0, "retries": 0, "seconds": 74.1, "max_memory_gb": 16.71, "ce": 0.0051848420553142205, "kl": 0.012808190885152726, "context": 0.011588937064516358} +{"step": 166, "steps": 250, "lr": 1.812818933977912e-05, "grad_norm": 1.696427583694458, "loss_scale": 4.0, "retries": 0, "seconds": 74.5, "max_memory_gb": 16.71, "ce": 0.4945782715512905, "kl": 0.10635858309979085, "context": 0.5477575621625874} +{"step": 167, "steps": 250, "lr": 1.785441736843128e-05, "grad_norm": 0.4064151346683502, "loss_scale": 4.0, "retries": 0, "seconds": 74.9, "max_memory_gb": 16.71, "ce": 0.02520511334296316, "kl": 0.00878405045659747, "context": 0.029597138112876564} +{"step": 168, "steps": 250, "lr": 1.7582383958243314e-05, "grad_norm": 0.41470953822135925, "loss_scale": 4.0, "retries": 0, "seconds": 75.4, "max_memory_gb": 16.71, "ce": 0.028323719416221138, "kl": 0.004052875010529533, "context": 0.030350156150234398} +{"step": 169, "steps": 250, "lr": 1.7312138141683164e-05, "grad_norm": 1.5999475717544556, "loss_scale": 4.0, "retries": 0, "seconds": 75.7, "max_memory_gb": 16.71, "ce": 0.06426927947904915, "kl": 1.153262972831726, "context": 0.6409007497131824} +{"step": 170, "steps": 250, "lr": 1.704372862901521e-05, "grad_norm": 0.02439856342971325, "loss_scale": 4.0, "retries": 0, "seconds": 76.1, "max_memory_gb": 16.71, "ce": 0.0013683197248610668, "kl": 0.0002985957644341397, "context": 0.0015176176020759158} +{"step": 171, "steps": 250, "lr": 1.6777203799520573e-05, "grad_norm": 0.9757713079452515, "loss_scale": 4.0, "retries": 0, "seconds": 76.5, "max_memory_gb": 16.71, "ce": 0.02459436873323284, "kl": 0.21867510967422277, "context": 0.1339319198159501} +{"step": 172, "steps": 250, "lr": 1.6512611692777008e-05, "grad_norm": 0.2097717970609665, "loss_scale": 4.0, "retries": 0, "seconds": 76.9, "max_memory_gb": 16.71, "ce": 0.012605406431248412, "kl": 0.01918991914135404, "context": 0.02220036496873945} +{"step": 173, "steps": 250, "lr": 1.6250000000000005e-05, "grad_norm": 0.4217269718647003, "loss_scale": 4.0, "retries": 0, "seconds": 77.3, "max_memory_gb": 16.71, "ce": 0.02773283055284992, "kl": 0.01804577963775955, "context": 0.036755720793735236} +{"step": 174, "steps": 250, "lr": 1.598941605544678e-05, "grad_norm": 0.4552392363548279, "loss_scale": 4.0, "retries": 0, "seconds": 77.7, "max_memory_gb": 16.71, "ce": 0.03040705765306484, "kl": 0.02159059582595546, "context": 0.04120235451409826} +{"step": 175, "steps": 250, "lr": 1.5730906827884506e-05, "grad_norm": 1.860949993133545, "loss_scale": 4.0, "retries": 0, "seconds": 78.2, "max_memory_gb": 16.71, "ce": 0.19531967234797776, "kl": 0.06897702277638018, "context": 0.22980818897485733} +{"step": 176, "steps": 250, "lr": 1.5474518912124476e-05, "grad_norm": 0.1081150472164154, "loss_scale": 4.0, "retries": 0, "seconds": 78.6, "max_memory_gb": 16.71, "ce": 0.0071619321970501915, "kl": 0.003324082614199142, "context": 0.008823973592370749} +{"step": 177, "steps": 250, "lr": 1.5220298520623646e-05, "grad_norm": 0.8179239630699158, "loss_scale": 4.0, "retries": 0, "seconds": 79.1, "max_memory_gb": 16.71, "ce": 0.02780293249816168, "kl": 0.7513666799204657, "context": 0.4034862500993768} +{"step": 178, "steps": 250, "lr": 1.4968291475155121e-05, "grad_norm": 0.7400793433189392, "loss_scale": 4.0, "retries": 0, "seconds": 79.5, "max_memory_gb": 16.71, "ce": 0.044825607270468026, "kl": 0.08244425214070361, "context": 0.0860477332607843} +{"step": 179, "steps": 250, "lr": 1.4718543198548998e-05, "grad_norm": 0.8545299172401428, "loss_scale": 4.0, "retries": 0, "seconds": 79.9, "max_memory_gb": 16.71, "ce": 0.4289486824563937, "kl": 0.01345342428976437, "context": 0.43567540602816734} +{"step": 180, "steps": 250, "lr": 1.4471098706505227e-05, "grad_norm": 0.47611570358276367, "loss_scale": 4.0, "retries": 0, "seconds": 80.4, "max_memory_gb": 16.71, "ce": 0.03903971931140404, "kl": 0.023571549403641257, "context": 0.05082549389771884} +{"step": 181, "steps": 250, "lr": 1.4226002599479743e-05, "grad_norm": 0.5193780064582825, "loss_scale": 4.0, "retries": 0, "seconds": 80.8, "max_memory_gb": 16.71, "ce": 0.00861426406481769, "kl": 0.1349772744870279, "context": 0.07610290152661037} +{"step": 182, "steps": 250, "lr": 1.3983299054645471e-05, "grad_norm": 0.4446464776992798, "loss_scale": 4.0, "retries": 0, "seconds": 81.3, "max_memory_gb": 16.71, "ce": 0.01883307858952321, "kl": 0.2979753613435605, "context": 0.16782076298841275} +{"step": 183, "steps": 250, "lr": 1.3743031817929724e-05, "grad_norm": 0.044058188796043396, "loss_scale": 4.0, "retries": 0, "seconds": 81.8, "max_memory_gb": 16.71, "ce": 0.003362491022926406, "kl": 0.0007088639176799916, "context": 0.0037169229181017727} +{"step": 184, "steps": 250, "lr": 1.3505244196129164e-05, "grad_norm": 0.04901159927248955, "loss_scale": 4.0, "retries": 0, "seconds": 82.3, "max_memory_gb": 16.71, "ce": 0.0021696796393371187, "kl": 0.0027329839367666864, "context": 0.0035361715854378417} +{"step": 185, "steps": 250, "lr": 1.3269979049104014e-05, "grad_norm": 0.47745996713638306, "loss_scale": 4.0, "retries": 0, "seconds": 82.8, "max_memory_gb": 16.71, "ce": 0.009616556228138506, "kl": 0.12086866940080654, "context": 0.07005089096492156} +{"step": 186, "steps": 250, "lr": 1.3037278782052863e-05, "grad_norm": 1.643538236618042, "loss_scale": 4.0, "retries": 0, "seconds": 83.3, "max_memory_gb": 16.71, "ce": 0.12910731192096137, "kl": 0.26206237154838163, "context": 0.2601385050002136} +{"step": 187, "steps": 250, "lr": 1.2807185337869343e-05, "grad_norm": 0.7913607954978943, "loss_scale": 4.0, "retries": 0, "seconds": 83.7, "max_memory_gb": 16.71, "ce": 0.049782474408857524, "kl": 0.47572728982777335, "context": 0.28764611075166613} +{"step": 188, "steps": 250, "lr": 1.2579740189582115e-05, "grad_norm": 1.180463433265686, "loss_scale": 4.0, "retries": 0, "seconds": 84.2, "max_memory_gb": 16.71, "ce": 0.25717698784865206, "kl": 0.018517542281642818, "context": 0.2664357487410598} +{"step": 189, "steps": 250, "lr": 1.2354984332879668e-05, "grad_norm": 1.0000441074371338, "loss_scale": 4.0, "retries": 0, "seconds": 84.6, "max_memory_gb": 16.71, "ce": 0.2927595311193727, "kl": 0.2647052349639125, "context": 0.42511215931153856} +{"step": 190, "steps": 250, "lr": 1.2132958278721064e-05, "grad_norm": 0.09843477606773376, "loss_scale": 4.0, "retries": 0, "seconds": 85.0, "max_memory_gb": 16.71, "ce": 0.0018258185591548681, "kl": 0.010077834594994783, "context": 0.006864735827548429} +{"step": 191, "steps": 250, "lr": 1.1913702046034016e-05, "grad_norm": 1.4671674966812134, "loss_scale": 4.0, "retries": 0, "seconds": 85.5, "max_memory_gb": 16.71, "ce": 0.3119582537910901, "kl": 0.1817436372384691, "context": 0.4028300869103987} +{"step": 192, "steps": 250, "lr": 1.1697255154501756e-05, "grad_norm": 1.269222617149353, "loss_scale": 4.0, "retries": 0, "seconds": 85.9, "max_memory_gb": 16.71, "ce": 0.5636607748747338, "kl": 0.09788094550458482, "context": 0.6126012234017253} +{"step": 193, "steps": 250, "lr": 1.1483656617439842e-05, "grad_norm": 0.07623818516731262, "loss_scale": 4.0, "retries": 0, "seconds": 86.3, "max_memory_gb": 16.71, "ce": 0.0049179950437974185, "kl": 0.0031530402739008423, "context": 0.006494515320810024} +{"step": 194, "steps": 250, "lr": 1.1272944934764177e-05, "grad_norm": 0.4359196126461029, "loss_scale": 4.0, "retries": 0, "seconds": 86.8, "max_memory_gb": 16.71, "ce": 0.049083784222602844, "kl": 0.07084876646695193, "context": 0.08450817118864506} +{"step": 195, "steps": 250, "lr": 1.1065158086051684e-05, "grad_norm": 0.09845951944589615, "loss_scale": 4.0, "retries": 0, "seconds": 87.2, "max_memory_gb": 16.71, "ce": 0.008838967332849279, "kl": 0.0037996953178662807, "context": 0.01073881532647647} +{"step": 196, "steps": 250, "lr": 1.086033352369467e-05, "grad_norm": 1.4749562740325928, "loss_scale": 4.0, "retries": 0, "seconds": 87.7, "max_memory_gb": 16.71, "ce": 0.001243407568836119, "kl": 0.4389821417932467, "context": 0.22073447624279652} +{"step": 197, "steps": 250, "lr": 1.0658508166150224e-05, "grad_norm": 1.3682529926300049, "loss_scale": 4.0, "retries": 0, "seconds": 88.1, "max_memory_gb": 16.71, "ce": 0.060160838984302245, "kl": 0.7639652977741207, "context": 0.44214347773231566} +{"step": 198, "steps": 250, "lr": 1.0459718391285954e-05, "grad_norm": 0.26359426975250244, "loss_scale": 4.0, "retries": 0, "seconds": 88.6, "max_memory_gb": 16.71, "ce": 0.07070977652620059, "kl": 0.009663644486863632, "context": 0.07554159691790119} +{"step": 199, "steps": 250, "lr": 1.0264000029822999e-05, "grad_norm": 1.761228084564209, "loss_scale": 4.0, "retries": 0, "seconds": 89.0, "max_memory_gb": 16.71, "ce": 0.1747294524757308, "kl": 0.04913011817552615, "context": 0.19929451344069093} +{"step": 200, "steps": 250, "lr": 1.0071388358877775e-05, "grad_norm": 0.7693033218383789, "loss_scale": 4.0, "retries": 0, "seconds": 89.5, "max_memory_gb": 16.71, "ce": 0.296380577608943, "kl": 0.03837903249950614, "context": 0.315570087346714} +{"step": 201, "steps": 250, "lr": 9.881918095603501e-06, "grad_norm": 0.38673102855682373, "loss_scale": 4.0, "retries": 0, "seconds": 89.9, "max_memory_gb": 16.71, "ce": 0.07155851217976306, "kl": 0.11776564522369881, "context": 0.13044133846415207} +{"step": 202, "steps": 250, "lr": 9.695623390932612e-06, "grad_norm": 0.05055180937051773, "loss_scale": 4.0, "retries": 0, "seconds": 90.3, "max_memory_gb": 16.71, "ce": 0.0019501004753692541, "kl": 0.0033988879138178163, "context": 0.003649544552899897} +{"step": 203, "steps": 250, "lr": 9.51253782342123e-06, "grad_norm": 0.416126549243927, "loss_scale": 4.0, "retries": 0, "seconds": 90.8, "max_memory_gb": 16.71, "ce": 0.011354575064615346, "kl": 0.017855685873655602, "context": 0.020282418001443148} +{"step": 204, "steps": 250, "lr": 9.332694393196846e-06, "grad_norm": 0.4761333465576172, "loss_scale": 4.0, "retries": 0, "seconds": 91.2, "max_memory_gb": 16.71, "ce": 0.03969378164038062, "kl": 0.009160347166471183, "context": 0.044273955281823874} +{"step": 205, "steps": 250, "lr": 9.156125516010256e-06, "grad_norm": 1.0375559329986572, "loss_scale": 4.0, "retries": 0, "seconds": 91.7, "max_memory_gb": 16.71, "ce": 0.002855285791156348, "kl": 0.42509326541767223, "context": 0.21540192337124608} +{"step": 206, "steps": 250, "lr": 8.982863017392733e-06, "grad_norm": 0.08228770643472672, "loss_scale": 4.0, "retries": 0, "seconds": 92.2, "max_memory_gb": 16.71, "ce": 0.004621352389222011, "kl": 0.002827683219948085, "context": 0.0060351940264808945} +{"step": 207, "steps": 250, "lr": 8.81293812691974e-06, "grad_norm": 0.16019822657108307, "loss_scale": 4.0, "retries": 0, "seconds": 92.6, "max_memory_gb": 16.71, "ce": 0.06152741461119149, "kl": 0.04397523097577505, "context": 0.08351503382436931} +{"step": 208, "steps": 250, "lr": 8.646381472581927e-06, "grad_norm": 0.07645486295223236, "loss_scale": 4.0, "retries": 0, "seconds": 93.1, "max_memory_gb": 16.71, "ce": 0.004391890070110094, "kl": 0.0022215715689526405, "context": 0.005502675703610294} +{"step": 209, "steps": 250, "lr": 8.483223075264622e-06, "grad_norm": 0.7476220726966858, "loss_scale": 4.0, "retries": 0, "seconds": 93.5, "max_memory_gb": 16.71, "ce": 0.178247246745741, "kl": 0.07938163804647047, "context": 0.2179380698944442} +{"step": 210, "steps": 250, "lr": 8.323492343336747e-06, "grad_norm": 1.0427380800247192, "loss_scale": 4.0, "retries": 0, "seconds": 94.0, "max_memory_gb": 16.71, "ce": 0.5398981235193787, "kl": 0.1377046208435786, "context": 0.608750427069026} +{"step": 211, "steps": 250, "lr": 8.167218067350122e-06, "grad_norm": 0.43511027097702026, "loss_scale": 4.0, "retries": 0, "seconds": 94.4, "max_memory_gb": 16.71, "ce": 0.10635353380348533, "kl": 0.13491400322527625, "context": 0.17381052812561393} +{"step": 212, "steps": 250, "lr": 8.01442841485013e-06, "grad_norm": 1.3100106716156006, "loss_scale": 4.0, "retries": 0, "seconds": 94.9, "max_memory_gb": 16.71, "ce": 0.05497748948982917, "kl": 0.9837258050720266, "context": 0.546840381633956} +{"step": 213, "steps": 250, "lr": 7.865150925298678e-06, "grad_norm": 0.751321017742157, "loss_scale": 4.0, "retries": 0, "seconds": 95.3, "max_memory_gb": 16.71, "ce": 0.25982198168640025, "kl": 0.04257186244285549, "context": 0.28110791107610567} +{"step": 214, "steps": 250, "lr": 7.719412505110375e-06, "grad_norm": 2.266218662261963, "loss_scale": 4.0, "retries": 0, "seconds": 95.8, "max_memory_gb": 16.71, "ce": 0.40552738566475455, "kl": 0.14177078567445278, "context": 0.4764127812231891} +{"step": 215, "steps": 250, "lr": 7.57723942280278e-06, "grad_norm": 1.3241349458694458, "loss_scale": 4.0, "retries": 0, "seconds": 96.2, "max_memory_gb": 16.71, "ce": 0.6486403101444012, "kl": 0.029734225485526622, "context": 0.6635074503865326} +{"step": 216, "steps": 250, "lr": 7.438657304261687e-06, "grad_norm": 0.7546284198760986, "loss_scale": 4.0, "retries": 0, "seconds": 96.7, "max_memory_gb": 16.71, "ce": 0.008609694014012348, "kl": 0.14874649154080544, "context": 0.08298294222913682} +{"step": 217, "steps": 250, "lr": 7.30369112812219e-06, "grad_norm": 0.03741403669118881, "loss_scale": 4.0, "retries": 0, "seconds": 97.1, "max_memory_gb": 16.71, "ce": 0.001924399723066017, "kl": 0.0017690397426122217, "context": 0.002808919583912939} +{"step": 218, "steps": 250, "lr": 7.172365221266403e-06, "grad_norm": 0.7428407073020935, "loss_scale": 4.0, "retries": 0, "seconds": 97.6, "max_memory_gb": 16.71, "ce": 0.15308568067848682, "kl": 0.23455457389354706, "context": 0.2703629657626152} +{"step": 219, "steps": 250, "lr": 7.0447032544387154e-06, "grad_norm": 0.296287477016449, "loss_scale": 4.0, "retries": 0, "seconds": 98.0, "max_memory_gb": 16.71, "ce": 0.008535549881344195, "kl": 0.11415630863120896, "context": 0.06561370413692202} +{"step": 220, "steps": 250, "lr": 6.920728237979274e-06, "grad_norm": 0.4003655016422272, "loss_scale": 4.0, "retries": 0, "seconds": 98.5, "max_memory_gb": 16.71, "ce": 0.03625480303890072, "kl": 0.005281375062622828, "context": 0.03889549011364579} +{"step": 221, "steps": 250, "lr": 6.800462517676456e-06, "grad_norm": 1.352281093597412, "loss_scale": 4.0, "retries": 0, "seconds": 98.9, "max_memory_gb": 16.71, "ce": 0.0049794182705227286, "kl": 0.40758624994850834, "context": 0.20877254553488456} +{"step": 222, "steps": 250, "lr": 6.683927770739222e-06, "grad_norm": 0.8316126465797424, "loss_scale": 4.0, "retries": 0, "seconds": 99.4, "max_memory_gb": 16.71, "ce": 0.03155179225723259, "kl": 0.0663077851113485, "context": 0.06470568527583964} +{"step": 223, "steps": 250, "lr": 6.571145001889892e-06, "grad_norm": 1.6795313358306885, "loss_scale": 4.0, "retries": 0, "seconds": 99.8, "max_memory_gb": 16.71, "ce": 0.09228127076858073, "kl": 0.011393118536943803, "context": 0.09797783142857952} +{"step": 224, "steps": 250, "lr": 6.46213453957817e-06, "grad_norm": 0.922089159488678, "loss_scale": 4.0, "retries": 0, "seconds": 100.3, "max_memory_gb": 16.71, "ce": 0.04135878733359277, "kl": 0.12818968301326095, "context": 0.10545363259734586} +{"step": 225, "steps": 250, "lr": 6.356916032317063e-06, "grad_norm": 0.7243639826774597, "loss_scale": 4.0, "retries": 0, "seconds": 100.7, "max_memory_gb": 16.71, "ce": 0.002670426416443661, "kl": 0.1342218959944148, "context": 0.06978137584519573} +{"step": 226, "steps": 250, "lr": 6.255508445141349e-06, "grad_norm": 0.7595767974853516, "loss_scale": 4.0, "retries": 0, "seconds": 101.2, "max_memory_gb": 16.71, "ce": 0.039773064461769536, "kl": 0.031141035186010413, "context": 0.05534358086879365} +{"step": 227, "steps": 250, "lr": 6.157930056189228e-06, "grad_norm": 0.2559490203857422, "loss_scale": 4.0, "retries": 0, "seconds": 101.6, "max_memory_gb": 16.71, "ce": 0.03021840141445864, "kl": 0.013403967200702027, "context": 0.03692038501321804} +{"step": 228, "steps": 250, "lr": 6.064198453407796e-06, "grad_norm": 1.0401334762573242, "loss_scale": 4.0, "retries": 0, "seconds": 102.0, "max_memory_gb": 16.71, "ce": 0.0034088095162587706, "kl": 0.16077539324760437, "context": 0.08379650418646634} +{"step": 229, "steps": 250, "lr": 5.974330531382935e-06, "grad_norm": 0.40622901916503906, "loss_scale": 4.0, "retries": 0, "seconds": 102.5, "max_memory_gb": 16.71, "ce": 0.0639987960603321, "kl": 0.025303671811116146, "context": 0.07665063382592052} +{"step": 230, "steps": 250, "lr": 5.888342488294125e-06, "grad_norm": 1.8644514083862305, "loss_scale": 4.0, "retries": 0, "seconds": 102.9, "max_memory_gb": 16.71, "ce": 0.5671948347880971, "kl": 0.3379666559631005, "context": 0.7361781593644992} +{"step": 231, "steps": 250, "lr": 5.806249822994849e-06, "grad_norm": 0.554814875125885, "loss_scale": 4.0, "retries": 0, "seconds": 103.3, "max_memory_gb": 16.71, "ce": 0.03927257026953157, "kl": 0.15143142243323382, "context": 0.1149882786849048} +{"step": 232, "steps": 250, "lr": 5.728067332218994e-06, "grad_norm": 0.4297504425048828, "loss_scale": 4.0, "retries": 0, "seconds": 103.7, "max_memory_gb": 16.71, "ce": 0.023474101879401132, "kl": 0.01196859165065689, "context": 0.029458398115821183} +{"step": 233, "steps": 250, "lr": 5.65380910791383e-06, "grad_norm": 0.5879163146018982, "loss_scale": 4.0, "retries": 0, "seconds": 104.2, "max_memory_gb": 16.71, "ce": 0.006988727313000709, "kl": 0.11971918802009895, "context": 0.06684832088649273} +{"step": 234, "steps": 250, "lr": 5.583488534700022e-06, "grad_norm": 0.3307815492153168, "loss_scale": 4.0, "retries": 0, "seconds": 104.6, "max_memory_gb": 16.71, "ce": 0.00797227508155629, "kl": 0.03647392056154786, "context": 0.026209234725683928} +{"step": 235, "steps": 250, "lr": 5.517118287459118e-06, "grad_norm": 1.5439237356185913, "loss_scale": 4.0, "retries": 0, "seconds": 105.1, "max_memory_gb": 16.71, "ce": 0.46142754128959496, "kl": 0.02313649763163994, "context": 0.47299579583341256} +{"step": 236, "steps": 250, "lr": 5.454710329048978e-06, "grad_norm": 0.17799057066440582, "loss_scale": 4.0, "retries": 0, "seconds": 105.5, "max_memory_gb": 16.71, "ce": 0.009183914822642691, "kl": 0.01849630217839149, "context": 0.01843206601915881} +{"step": 237, "steps": 250, "lr": 5.3962759081475436e-06, "grad_norm": 0.2599031329154968, "loss_scale": 4.0, "retries": 0, "seconds": 106.0, "max_memory_gb": 16.71, "ce": 0.06663757175556384, "kl": 0.029207479503781997, "context": 0.08124130958458409} +{"step": 238, "steps": 250, "lr": 5.341825557225321e-06, "grad_norm": 0.25375667214393616, "loss_scale": 4.0, "retries": 0, "seconds": 106.4, "max_memory_gb": 16.71, "ce": 0.021733979774580803, "kl": 0.008173513910151087, "context": 0.025820736918831244} +{"step": 239, "steps": 250, "lr": 5.291369090646961e-06, "grad_norm": 0.2965341806411743, "loss_scale": 4.0, "retries": 0, "seconds": 106.8, "max_memory_gb": 16.71, "ce": 0.097240423085168, "kl": 0.02391043478564825, "context": 0.10919563670177013} +{"step": 240, "steps": 250, "lr": 5.244915602902289e-06, "grad_norm": 1.0109524726867676, "loss_scale": 4.0, "retries": 0, "seconds": 107.2, "max_memory_gb": 16.71, "ce": 0.1069060717272805, "kl": 0.046794740854238626, "context": 0.13030344530125149} +{"step": 241, "steps": 250, "lr": 5.20247346696706e-06, "grad_norm": 1.1630357503890991, "loss_scale": 4.0, "retries": 0, "seconds": 107.7, "max_memory_gb": 16.71, "ce": 0.2056777434045216, "kl": 0.028419492031389382, "context": 0.21988749648153316} +{"step": 242, "steps": 250, "lr": 5.164050332793787e-06, "grad_norm": 0.19477276504039764, "loss_scale": 4.0, "retries": 0, "seconds": 108.1, "max_memory_gb": 16.71, "ce": 0.003290535882115364, "kl": 0.026884311171102127, "context": 0.016732691816287115} +{"step": 243, "steps": 250, "lr": 5.129653125932873e-06, "grad_norm": 0.06715390086174011, "loss_scale": 4.0, "retries": 0, "seconds": 108.6, "max_memory_gb": 16.71, "ce": 0.003166153939673677, "kl": 0.0041159137326758355, "context": 0.005224110733252019} +{"step": 244, "steps": 250, "lr": 5.099288046284329e-06, "grad_norm": 0.5538956522941589, "loss_scale": 4.0, "retries": 0, "seconds": 109.0, "max_memory_gb": 16.71, "ce": 0.03460235166858183, "kl": 0.13033297130459687, "context": 0.09976883883064147} +{"step": 245, "steps": 250, "lr": 5.072960566980277e-06, "grad_norm": 0.9753162860870361, "loss_scale": 4.0, "retries": 0, "seconds": 109.5, "max_memory_gb": 16.71, "ce": 0.3950878292089328, "kl": 0.062057274975813925, "context": 0.4261164660565555} +{"step": 246, "steps": 250, "lr": 5.050675433398435e-06, "grad_norm": 0.016602231189608574, "loss_scale": 4.0, "retries": 0, "seconds": 109.9, "max_memory_gb": 16.71, "ce": 0.001144015957834199, "kl": 0.0004790911129930464, "context": 0.0013835614954587072} +{"step": 247, "steps": 250, "lr": 5.032436662306811e-06, "grad_norm": 0.8274972438812256, "loss_scale": 4.0, "retries": 0, "seconds": 110.4, "max_memory_gb": 16.71, "ce": 0.003601079515647143, "kl": 0.46074639431390096, "context": 0.233974274204229} +{"step": 248, "steps": 250, "lr": 5.018247541139684e-06, "grad_norm": 0.2989153563976288, "loss_scale": 4.0, "retries": 0, "seconds": 110.8, "max_memory_gb": 16.71, "ce": 0.0247425195411779, "kl": 0.008599508670158684, "context": 0.029042274341918528} +{"step": 249, "steps": 250, "lr": 5.00811062740508e-06, "grad_norm": 1.4432365894317627, "loss_scale": 4.0, "retries": 0, "seconds": 111.3, "max_memory_gb": 16.71, "ce": 0.05559809447731823, "kl": 0.0712228994234465, "context": 0.09120954782702029} +{"step": 250, "steps": 250, "lr": 5.002027748223778e-06, "grad_norm": 0.0649186298251152, "loss_scale": 4.0, "retries": 0, "seconds": 111.7, "max_memory_gb": 16.71, "ce": 0.0064979665039572865, "kl": 0.0031367175397463143, "context": 0.0080663250409998} diff --git a/runs/single_internvl/context/L34_s13/adapter/README.md b/runs/single_internvl/context/L34_s13/adapter/README.md new file mode 100644 index 0000000000000000000000000000000000000000..4f9f856cf0e1531b47e25d266833b2977a19d3ac --- /dev/null +++ b/runs/single_internvl/context/L34_s13/adapter/README.md @@ -0,0 +1,207 @@ +--- +base_model: /data/OPSD/assets/models/InternVL3_5-8B-HF +library_name: peft +pipeline_tag: text-generation +tags: +- base_model:adapter:/data/OPSD/assets/models/InternVL3_5-8B-HF +- lora +- transformers +--- + +# Model Card for Model ID + + + + + +## Model Details + +### Model Description + + + + + +- **Developed by:** [More Information Needed] +- **Funded by [optional]:** [More Information Needed] +- **Shared by [optional]:** [More Information Needed] +- **Model type:** [More Information Needed] +- **Language(s) (NLP):** [More Information Needed] +- **License:** [More Information Needed] +- **Finetuned from model [optional]:** [More Information Needed] + +### Model Sources [optional] + + + +- **Repository:** [More Information Needed] +- **Paper [optional]:** [More Information Needed] +- **Demo [optional]:** [More Information Needed] + +## Uses + + + +### Direct Use + + + +[More Information Needed] + +### Downstream Use [optional] + + + +[More Information Needed] + +### Out-of-Scope Use + + + +[More Information Needed] + +## Bias, Risks, and Limitations + + + +[More Information Needed] + +### Recommendations + + + +Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. More information needed for further recommendations. + +## How to Get Started with the Model + +Use the code below to get started with the model. + +[More Information Needed] + +## Training Details + +### Training Data + + + +[More Information Needed] + +### Training Procedure + + + +#### Preprocessing [optional] + +[More Information Needed] + + +#### Training Hyperparameters + +- **Training regime:** [More Information Needed] + +#### Speeds, Sizes, Times [optional] + + + +[More Information Needed] + +## Evaluation + + + +### Testing Data, Factors & Metrics + +#### Testing Data + + + +[More Information Needed] + +#### Factors + + + +[More Information Needed] + +#### Metrics + + + +[More Information Needed] + +### Results + +[More Information Needed] + +#### Summary + + + +## Model Examination [optional] + + + +[More Information Needed] + +## Environmental Impact + + + +Carbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700). + +- **Hardware Type:** [More Information Needed] +- **Hours used:** [More Information Needed] +- **Cloud Provider:** [More Information Needed] +- **Compute Region:** [More Information Needed] +- **Carbon Emitted:** [More Information Needed] + +## Technical Specifications [optional] + +### Model Architecture and Objective + +[More Information Needed] + +### Compute Infrastructure + +[More Information Needed] + +#### Hardware + +[More Information Needed] + +#### Software + +[More Information Needed] + +## Citation [optional] + + + +**BibTeX:** + +[More Information Needed] + +**APA:** + +[More Information Needed] + +## Glossary [optional] + + + +[More Information Needed] + +## More Information [optional] + +[More Information Needed] + +## Model Card Authors [optional] + +[More Information Needed] + +## Model Card Contact + +[More Information Needed] +### Framework versions + +- PEFT 0.20.0 \ No newline at end of file diff --git a/runs/single_internvl/context/L34_s13/adapter/adapter_config.json b/runs/single_internvl/context/L34_s13/adapter/adapter_config.json new file mode 100644 index 0000000000000000000000000000000000000000..bf37324039cfb8c6db94986da06db8431c123d89 --- /dev/null +++ b/runs/single_internvl/context/L34_s13/adapter/adapter_config.json @@ -0,0 +1,50 @@ +{ + "alora_invocation_tokens": null, + "alpha_pattern": {}, + "arrow_config": null, + "auto_mapping": null, + "base_model_name_or_path": "/data/OPSD/assets/models/InternVL3_5-8B-HF", + "bias": "none", + "corda_config": null, + "ensure_weight_tying": false, + "eva_config": null, + "exclude_modules": null, + "fan_in_fan_out": false, + "inference_mode": true, + "init_lora_weights": true, + "layer_replication": null, + "layers_pattern": null, + "layers_to_transform": null, + "loftq_config": {}, + "lora_alpha": 32, + "lora_bias": false, + "lora_dropout": 0.05, + "lora_ga_config": null, + "megatron_config": null, + "megatron_core": "megatron.core", + "modules_to_save": null, + "monteclora_config": null, + "peft_type": "LORA", + "peft_version": "0.20.0", + "qalora_group_size": 16, + "r": 16, + "rank_pattern": {}, + "revision": null, + "target_modules": [ + "model.language_model.layers.34.mlp.down_proj", + "model.language_model.layers.34.self_attn.q_proj", + "model.language_model.layers.34.mlp.gate_proj", + "model.language_model.layers.34.self_attn.o_proj", + "model.language_model.layers.34.mlp.up_proj", + "model.language_model.layers.34.self_attn.k_proj", + "model.language_model.layers.34.self_attn.v_proj" + ], + "target_parameters": null, + "task_type": "CAUSAL_LM", + "trainable_token_indices": null, + "use_bdlora": null, + "use_dora": false, + "use_qalora": false, + "use_rslora": false, + "velora_config": null +} \ No newline at end of file diff --git a/runs/single_internvl/context/L34_s13/config.json b/runs/single_internvl/context/L34_s13/config.json new file mode 100644 index 0000000000000000000000000000000000000000..3cc33e9dd86c88687bbad75bf76b5ae4f2353921 --- /dev/null +++ b/runs/single_internvl/context/L34_s13/config.json @@ -0,0 +1,24 @@ +{ + "behavior": "context", + "seed": 13, + "model": "/data/OPSD/assets/models/InternVL3_5-8B-HF", + "backbone": "internvl", + "train": "/data/OPSD/assets/v3/context.jsonl", + "items": 1000, + "steps": 250, + "epochs": 1, + "steps_per_epoch": 250, + "groups_per_step": 4, + "lr": 5e-05, + "scheduler": "cosine", + "warmup_steps": 16, + "final_fraction": 0.1, + "weight_decay": 0.0, + "clip": 1.0, + "initial_loss_scale": 4.0, + "adapter": { + "layer": 34, + "rank": 16, + "trainable_parameters": 1212416 + } +} \ No newline at end of file diff --git a/runs/single_internvl/context/L34_s13/steps.jsonl b/runs/single_internvl/context/L34_s13/steps.jsonl new file mode 100644 index 0000000000000000000000000000000000000000..8bccba9bdca4b4dd6749389f152a50572e1603bf --- /dev/null +++ b/runs/single_internvl/context/L34_s13/steps.jsonl @@ -0,0 +1,250 @@ +{"step": 1, "steps": 250, "lr": 3.125e-06, "grad_norm": 0.37761035561561584, "loss_scale": 4.0, "retries": 0, "seconds": 1.3, "max_memory_gb": 16.47, "ce": 0.5227766726748087, "kl": 0.08654405153356493, "context": 0.5660487064160407} +{"step": 2, "steps": 250, "lr": 6.25e-06, "grad_norm": 0.5186718106269836, "loss_scale": 4.0, "retries": 0, "seconds": 1.7, "max_memory_gb": 16.47, "ce": 0.6404395870849839, "kl": 0.4683591094252506, "context": 0.874619144036842} +{"step": 3, "steps": 250, "lr": 9.375000000000001e-06, "grad_norm": 1.0828431844711304, "loss_scale": 4.0, "retries": 0, "seconds": 2.0, "max_memory_gb": 16.47, "ce": 1.0941549465060234, "kl": 0.06738869473338127, "context": 1.127849280834198} +{"step": 4, "steps": 250, "lr": 1.25e-05, "grad_norm": 0.6129411458969116, "loss_scale": 4.0, "retries": 0, "seconds": 2.4, "max_memory_gb": 16.47, "ce": 0.45207827698322944, "kl": 0.42941547930308843, "context": 0.666786012268858} +{"step": 5, "steps": 250, "lr": 1.5625e-05, "grad_norm": 0.013608083128929138, "loss_scale": 4.0, "retries": 0, "seconds": 2.8, "max_memory_gb": 16.47, "ce": 0.009195431375701446, "kl": 0.00819971573582734, "context": 0.013295289332745597} +{"step": 6, "steps": 250, "lr": 1.8750000000000002e-05, "grad_norm": 0.3128042221069336, "loss_scale": 4.0, "retries": 0, "seconds": 3.1, "max_memory_gb": 16.47, "ce": 0.29097037747851573, "kl": 0.0746842822409235, "context": 0.3283125229063444} +{"step": 7, "steps": 250, "lr": 2.1875e-05, "grad_norm": 0.8752612471580505, "loss_scale": 4.0, "retries": 0, "seconds": 3.5, "max_memory_gb": 16.47, "ce": 0.8299291558323603, "kl": 0.12237092932673477, "context": 0.8911146149075648} +{"step": 8, "steps": 250, "lr": 2.5e-05, "grad_norm": 1.103765606880188, "loss_scale": 4.0, "retries": 0, "seconds": 3.9, "max_memory_gb": 16.47, "ce": 0.7970096506178379, "kl": 0.2904033618979156, "context": 0.9422113485634327} +{"step": 9, "steps": 250, "lr": 2.8125000000000003e-05, "grad_norm": 0.7307001948356628, "loss_scale": 4.0, "retries": 0, "seconds": 4.3, "max_memory_gb": 16.47, "ce": 0.529319217428565, "kl": 0.13603398314444348, "context": 0.5973362177610397} +{"step": 10, "steps": 250, "lr": 3.125e-05, "grad_norm": 0.14946189522743225, "loss_scale": 4.0, "retries": 0, "seconds": 4.6, "max_memory_gb": 16.47, "ce": 0.05406907516589854, "kl": 0.04538091190624982, "context": 0.07675953186117113} +{"step": 11, "steps": 250, "lr": 3.4375e-05, "grad_norm": 0.9589400291442871, "loss_scale": 4.0, "retries": 0, "seconds": 5.0, "max_memory_gb": 16.47, "ce": 0.707253715780098, "kl": 0.004943941062265367, "context": 0.7097257009372697} +{"step": 12, "steps": 250, "lr": 3.7500000000000003e-05, "grad_norm": 0.6590546369552612, "loss_scale": 4.0, "retries": 0, "seconds": 5.4, "max_memory_gb": 16.47, "ce": 0.47689084079775057, "kl": 0.002505653024854837, "context": 0.47814365851445473} +{"step": 13, "steps": 250, "lr": 4.0625000000000005e-05, "grad_norm": 0.6652080416679382, "loss_scale": 4.0, "retries": 0, "seconds": 5.7, "max_memory_gb": 16.47, "ce": 0.4309131596091902, "kl": 0.01905656784583698, "context": 0.44044144966755994} +{"step": 14, "steps": 250, "lr": 4.375e-05, "grad_norm": 1.219215989112854, "loss_scale": 4.0, "retries": 0, "seconds": 6.1, "max_memory_gb": 16.47, "ce": 0.7773014008998871, "kl": 0.9423743607476354, "context": 1.2484886199235916} +{"step": 15, "steps": 250, "lr": 4.6875e-05, "grad_norm": 1.0410902500152588, "loss_scale": 4.0, "retries": 0, "seconds": 6.4, "max_memory_gb": 16.47, "ce": 0.5317166905338127, "kl": 0.46841133894531595, "context": 0.7659223339294385} +{"step": 16, "steps": 250, "lr": 5e-05, "grad_norm": 0.5452935099601746, "loss_scale": 4.0, "retries": 0, "seconds": 6.8, "max_memory_gb": 16.47, "ce": 0.20521015085250838, "kl": 0.0009570344000167097, "context": 0.20568866158282617} +{"step": 17, "steps": 250, "lr": 5e-05, "grad_norm": 0.7989943623542786, "loss_scale": 4.0, "retries": 0, "seconds": 7.2, "max_memory_gb": 16.47, "ce": 0.3286481923132669, "kl": 0.8176099721167702, "context": 0.7374531725654379} +{"step": 18, "steps": 250, "lr": 4.999797225177623e-05, "grad_norm": 1.458905577659607, "loss_scale": 4.0, "retries": 0, "seconds": 7.5, "max_memory_gb": 16.47, "ce": 0.5464778242021566, "kl": 0.2091637388075469, "context": 0.6510596894077025} +{"step": 19, "steps": 250, "lr": 4.999188937259492e-05, "grad_norm": 2.792428970336914, "loss_scale": 4.0, "retries": 0, "seconds": 7.9, "max_memory_gb": 16.47, "ce": 0.8029928058385849, "kl": 0.16543459240347147, "context": 0.8857101127505302} +{"step": 20, "steps": 250, "lr": 4.998175245886032e-05, "grad_norm": 0.021210303530097008, "loss_scale": 4.0, "retries": 0, "seconds": 8.2, "max_memory_gb": 16.47, "ce": 0.0023261074329639086, "kl": 0.021195027917201514, "context": 0.012923620975925587} +{"step": 21, "steps": 250, "lr": 4.996756333769319e-05, "grad_norm": 1.5256315469741821, "loss_scale": 4.0, "retries": 0, "seconds": 8.6, "max_memory_gb": 16.47, "ce": 0.3759386041592734, "kl": 0.05062838021240168, "context": 0.40125279519634205} +{"step": 22, "steps": 250, "lr": 4.994932456660157e-05, "grad_norm": 1.6551647186279297, "loss_scale": 4.0, "retries": 0, "seconds": 9.0, "max_memory_gb": 16.47, "ce": 0.4098642658937024, "kl": 0.027115583594422787, "context": 0.4234220684011234} +{"step": 23, "steps": 250, "lr": 4.992703943301973e-05, "grad_norm": 2.134916305541992, "loss_scale": 4.0, "retries": 0, "seconds": 9.3, "max_memory_gb": 16.47, "ce": 0.5732468810547289, "kl": 0.980443716443574, "context": 1.0634687118026704} +{"step": 24, "steps": 250, "lr": 4.9900711953715675e-05, "grad_norm": 1.6743794679641724, "loss_scale": 4.0, "retries": 0, "seconds": 9.6, "max_memory_gb": 16.47, "ce": 0.2790686786174774, "kl": 0.3385501215234399, "context": 0.4483437482267618} +{"step": 25, "steps": 250, "lr": 4.987034687406713e-05, "grad_norm": 2.6279423236846924, "loss_scale": 4.0, "retries": 0, "seconds": 10.0, "max_memory_gb": 16.47, "ce": 0.6429021182066208, "kl": 0.21956343296915293, "context": 0.7526838295161724} +{"step": 26, "steps": 250, "lr": 4.9835949667206215e-05, "grad_norm": 2.2995517253875732, "loss_scale": 4.0, "retries": 0, "seconds": 10.4, "max_memory_gb": 16.47, "ce": 0.31581367179751396, "kl": 0.6854472709819674, "context": 0.6585372984409332} +{"step": 27, "steps": 250, "lr": 4.9797526533032945e-05, "grad_norm": 2.2470264434814453, "loss_scale": 4.0, "retries": 0, "seconds": 10.7, "max_memory_gb": 16.47, "ce": 0.42172479729924817, "kl": 0.41442567374542705, "context": 0.6289376515196636} +{"step": 28, "steps": 250, "lr": 4.9755084397097716e-05, "grad_norm": 1.3578767776489258, "loss_scale": 4.0, "retries": 0, "seconds": 11.1, "max_memory_gb": 16.47, "ce": 0.586861427873373, "kl": 0.9242429820587859, "context": 1.0489829033613205} +{"step": 29, "steps": 250, "lr": 4.970863090935304e-05, "grad_norm": 1.9864777326583862, "loss_scale": 4.0, "retries": 0, "seconds": 11.5, "max_memory_gb": 16.47, "ce": 0.35618973802775145, "kl": 0.3606715979985893, "context": 0.5365255456417799} +{"step": 30, "steps": 250, "lr": 4.965817444277468e-05, "grad_norm": 2.278364896774292, "loss_scale": 4.0, "retries": 0, "seconds": 11.8, "max_memory_gb": 16.47, "ce": 0.30045791371412633, "kl": 0.009465701120461745, "context": 0.30519075582742516} +{"step": 31, "steps": 250, "lr": 4.9603724091852456e-05, "grad_norm": 3.0495870113372803, "loss_scale": 4.0, "retries": 0, "seconds": 12.2, "max_memory_gb": 16.47, "ce": 0.6068007834255695, "kl": 0.15922486782073975, "context": 0.6864132359623909} +{"step": 32, "steps": 250, "lr": 4.9545289670951025e-05, "grad_norm": 3.303603172302246, "loss_scale": 4.0, "retries": 0, "seconds": 12.6, "max_memory_gb": 16.47, "ce": 0.32879478484392166, "kl": 0.01751886901911348, "context": 0.3375542163848877} +{"step": 33, "steps": 250, "lr": 4.948288171254089e-05, "grad_norm": 1.644532322883606, "loss_scale": 4.0, "retries": 0, "seconds": 13.0, "max_memory_gb": 16.47, "ce": 0.1716457987204194, "kl": 0.06106753833591938, "context": 0.2021795753389597} +{"step": 34, "steps": 250, "lr": 4.941651146529998e-05, "grad_norm": 1.4247065782546997, "loss_scale": 4.0, "retries": 0, "seconds": 13.4, "max_memory_gb": 16.47, "ce": 0.09061342109998805, "kl": 0.7398322010240008, "context": 0.4605295068286068} +{"step": 35, "steps": 250, "lr": 4.9346190892086174e-05, "grad_norm": 1.5644445419311523, "loss_scale": 4.0, "retries": 0, "seconds": 13.8, "max_memory_gb": 16.47, "ce": 0.11415828792087268, "kl": 0.004089853685400158, "context": 0.11620321893133223} +{"step": 36, "steps": 250, "lr": 4.927193266778101e-05, "grad_norm": 1.680168628692627, "loss_scale": 4.0, "retries": 0, "seconds": 14.2, "max_memory_gb": 16.47, "ce": 0.24066647718427703, "kl": 0.04151513797114603, "context": 0.26142404349229764} +{"step": 37, "steps": 250, "lr": 4.919375017700515e-05, "grad_norm": 1.2988855838775635, "loss_scale": 4.0, "retries": 0, "seconds": 14.6, "max_memory_gb": 16.47, "ce": 0.20160232647322118, "kl": 0.7391269218642265, "context": 0.5711657957872376} +{"step": 38, "steps": 250, "lr": 4.911165751170588e-05, "grad_norm": 1.5227571725845337, "loss_scale": 4.0, "retries": 0, "seconds": 15.0, "max_memory_gb": 16.47, "ce": 0.13600982073694468, "kl": 0.5118064093985595, "context": 0.39191302191466093} +{"step": 39, "steps": 250, "lr": 4.902566946861708e-05, "grad_norm": 0.27973541617393494, "loss_scale": 4.0, "retries": 0, "seconds": 15.4, "max_memory_gb": 16.47, "ce": 0.1229011340692523, "kl": 0.009806103291339241, "context": 0.1278041836048942} +{"step": 40, "steps": 250, "lr": 4.89358015465922e-05, "grad_norm": 0.38713106513023376, "loss_scale": 4.0, "retries": 0, "seconds": 15.8, "max_memory_gb": 16.47, "ce": 0.02480542630655691, "kl": 0.007932591193821281, "context": 0.028771722689270973} +{"step": 41, "steps": 250, "lr": 4.884206994381078e-05, "grad_norm": 0.6594567894935608, "loss_scale": 4.0, "retries": 0, "seconds": 16.1, "max_memory_gb": 16.47, "ce": 0.3783110890071839, "kl": 0.12381945899687707, "context": 0.44022082770243287} +{"step": 42, "steps": 250, "lr": 4.874449155485866e-05, "grad_norm": 1.0558871030807495, "loss_scale": 4.0, "retries": 0, "seconds": 16.5, "max_memory_gb": 16.47, "ce": 0.06884965513745556, "kl": 0.010017899410740938, "context": 0.07385860378417419} +{"step": 43, "steps": 250, "lr": 4.864308396768294e-05, "grad_norm": 0.9899312257766724, "loss_scale": 4.0, "retries": 0, "seconds": 16.9, "max_memory_gb": 16.47, "ce": 0.03732533962465823, "kl": 0.04601217378512956, "context": 0.06033142654632684} +{"step": 44, "steps": 250, "lr": 4.853786546042184e-05, "grad_norm": 1.5511966943740845, "loss_scale": 4.0, "retries": 0, "seconds": 17.3, "max_memory_gb": 16.47, "ce": 0.5580377080113976, "kl": 1.3266719271414331, "context": 1.2213736734556733} +{"step": 45, "steps": 250, "lr": 4.842885499811012e-05, "grad_norm": 1.4160007238388062, "loss_scale": 4.0, "retries": 0, "seconds": 17.7, "max_memory_gb": 16.47, "ce": 0.8767894050979521, "kl": 0.21264849188446533, "context": 0.983113668160513} +{"step": 46, "steps": 250, "lr": 4.831607222926078e-05, "grad_norm": 0.3288963735103607, "loss_scale": 4.0, "retries": 0, "seconds": 18.0, "max_memory_gb": 16.47, "ce": 0.08710252097807825, "kl": 0.25664880813565105, "context": 0.21542692324146628} +{"step": 47, "steps": 250, "lr": 4.8199537482323545e-05, "grad_norm": 0.3190377354621887, "loss_scale": 4.0, "retries": 0, "seconds": 18.4, "max_memory_gb": 16.47, "ce": 0.012811800683266483, "kl": 0.026383135351352394, "context": 0.0260033673257567} +{"step": 48, "steps": 250, "lr": 4.807927176202073e-05, "grad_norm": 0.7356830835342407, "loss_scale": 4.0, "retries": 0, "seconds": 18.8, "max_memory_gb": 16.47, "ce": 0.027977079793345183, "kl": 0.008884651942935307, "context": 0.03241940477164462} +{"step": 49, "steps": 250, "lr": 4.795529674556129e-05, "grad_norm": 1.2481225728988647, "loss_scale": 4.0, "retries": 0, "seconds": 19.2, "max_memory_gb": 16.47, "ce": 0.03372259564639535, "kl": 0.25314849342180423, "context": 0.16029684049499338} +{"step": 50, "steps": 250, "lr": 4.782763477873361e-05, "grad_norm": 1.3849934339523315, "loss_scale": 4.0, "retries": 0, "seconds": 19.6, "max_memory_gb": 16.47, "ce": 0.09054279038537061, "kl": 0.01462170206650626, "context": 0.09785364205890801} +{"step": 51, "steps": 250, "lr": 4.769630887187782e-05, "grad_norm": 1.1050041913986206, "loss_scale": 4.0, "retries": 0, "seconds": 20.0, "max_memory_gb": 16.47, "ce": 0.5346348782659334, "kl": 0.021364169223176077, "context": 0.5453169430802518} +{"step": 52, "steps": 250, "lr": 4.756134269573832e-05, "grad_norm": 0.6214925050735474, "loss_scale": 4.0, "retries": 0, "seconds": 20.3, "max_memory_gb": 16.47, "ce": 0.024725373863475397, "kl": 0.22056901163887233, "context": 0.1350098834373057} +{"step": 53, "steps": 250, "lr": 4.7422760577197226e-05, "grad_norm": 1.3009779453277588, "loss_scale": 4.0, "retries": 0, "seconds": 20.7, "max_memory_gb": 16.47, "ce": 0.18340773973613977, "kl": 0.32028638990595937, "context": 0.3435509353876114} +{"step": 54, "steps": 250, "lr": 4.728058749488963e-05, "grad_norm": 0.7596674561500549, "loss_scale": 4.0, "retries": 0, "seconds": 21.1, "max_memory_gb": 16.47, "ce": 0.04197191438288428, "kl": 0.3070386714389315, "context": 0.1954912464134395} +{"step": 55, "steps": 250, "lr": 4.713484907470133e-05, "grad_norm": 0.2958359122276306, "loss_scale": 4.0, "retries": 0, "seconds": 21.4, "max_memory_gb": 16.47, "ce": 0.01915532586281188, "kl": 0.013050557648057293, "context": 0.025680605875095353} +{"step": 56, "steps": 250, "lr": 4.6985571585149876e-05, "grad_norm": 0.6505310535430908, "loss_scale": 4.0, "retries": 0, "seconds": 21.8, "max_memory_gb": 16.47, "ce": 0.026160964975133538, "kl": 0.22079393197782338, "context": 0.13655793317593634} +{"step": 57, "steps": 250, "lr": 4.6832781932649884e-05, "grad_norm": 1.0424697399139404, "loss_scale": 4.0, "retries": 0, "seconds": 22.2, "max_memory_gb": 16.47, "ce": 0.03314849469461478, "kl": 0.03183406507378095, "context": 0.049065525410696864} +{"step": 58, "steps": 250, "lr": 4.667650765666325e-05, "grad_norm": 0.05739349126815796, "loss_scale": 4.0, "retries": 0, "seconds": 22.6, "max_memory_gb": 16.47, "ce": 0.003399612905923277, "kl": 0.0016180769915834503, "context": 0.004208651473163627} +{"step": 59, "steps": 250, "lr": 4.651677692473538e-05, "grad_norm": 1.015641689300537, "loss_scale": 4.0, "retries": 0, "seconds": 23.0, "max_memory_gb": 16.47, "ce": 0.03994853119365871, "kl": 0.05865753337275237, "context": 0.0692772978800349} +{"step": 60, "steps": 250, "lr": 4.635361852741808e-05, "grad_norm": 1.6636821031570435, "loss_scale": 4.0, "retries": 0, "seconds": 23.3, "max_memory_gb": 16.47, "ce": 0.06407979904906824, "kl": 0.05871231440687552, "context": 0.09343595802783966} +{"step": 61, "steps": 250, "lr": 4.6187061873080264e-05, "grad_norm": 1.6113951206207275, "loss_scale": 4.0, "retries": 0, "seconds": 23.7, "max_memory_gb": 16.47, "ce": 0.023819448659196496, "kl": 0.8361918111331761, "context": 0.4419153684284538} +{"step": 62, "steps": 250, "lr": 4.601713698260727e-05, "grad_norm": 0.18082670867443085, "loss_scale": 4.0, "retries": 0, "seconds": 24.1, "max_memory_gb": 16.47, "ce": 0.03501736468751915, "kl": 0.009694977670733351, "context": 0.039864853548351675} +{"step": 63, "steps": 250, "lr": 4.5843874483989746e-05, "grad_norm": 1.0709245204925537, "loss_scale": 4.0, "retries": 0, "seconds": 24.5, "max_memory_gb": 16.47, "ce": 0.04108122136676684, "kl": 0.07051847489492502, "context": 0.0763404592871666} +{"step": 64, "steps": 250, "lr": 4.566730560680316e-05, "grad_norm": 0.2672809362411499, "loss_scale": 4.0, "retries": 0, "seconds": 24.8, "max_memory_gb": 16.47, "ce": 0.016198195808101445, "kl": 0.012789112370228395, "context": 0.022592751774936914} +{"step": 65, "steps": 250, "lr": 4.548746217657878e-05, "grad_norm": 1.5865362882614136, "loss_scale": 4.0, "retries": 0, "seconds": 25.2, "max_memory_gb": 16.47, "ce": 0.2636011531576514, "kl": 0.11128308088518679, "context": 0.3192427009344101} +{"step": 66, "steps": 250, "lr": 4.5304376609067396e-05, "grad_norm": 2.257746696472168, "loss_scale": 4.0, "retries": 0, "seconds": 25.6, "max_memory_gb": 16.47, "ce": 0.16411478790632827, "kl": 0.07415785459784274, "context": 0.20119370774000345} +{"step": 67, "steps": 250, "lr": 4.5118081904396504e-05, "grad_norm": 0.16051939129829407, "loss_scale": 4.0, "retries": 0, "seconds": 26.0, "max_memory_gb": 16.47, "ce": 0.00829287149826996, "kl": 0.010847485915292054, "context": 0.013716614805161953} +{"step": 68, "steps": 250, "lr": 4.4928611641122225e-05, "grad_norm": 0.1115679070353508, "loss_scale": 4.0, "retries": 0, "seconds": 26.4, "max_memory_gb": 16.47, "ce": 0.008297624244733015, "kl": 0.04493565726079396, "context": 0.030765453542699106} +{"step": 69, "steps": 250, "lr": 4.473599997017701e-05, "grad_norm": 0.7930982708930969, "loss_scale": 4.0, "retries": 0, "seconds": 26.8, "max_memory_gb": 16.47, "ce": 0.25218116329051554, "kl": 0.029988321155542508, "context": 0.26717532600741833} +{"step": 70, "steps": 250, "lr": 4.4540281608714053e-05, "grad_norm": 0.055742401629686356, "loss_scale": 4.0, "retries": 0, "seconds": 27.2, "max_memory_gb": 16.47, "ce": 0.00292969866131898, "kl": 0.0006226273399079219, "context": 0.003241012367652729} +{"step": 71, "steps": 250, "lr": 4.434149183384977e-05, "grad_norm": 0.42156440019607544, "loss_scale": 4.0, "retries": 0, "seconds": 27.5, "max_memory_gb": 16.47, "ce": 0.02006226755111129, "kl": 0.012907995289424434, "context": 0.026516265017562546} +{"step": 72, "steps": 250, "lr": 4.413966647630534e-05, "grad_norm": 0.8080861568450928, "loss_scale": 4.0, "retries": 0, "seconds": 27.9, "max_memory_gb": 16.47, "ce": 0.019341501174494624, "kl": 0.4133800126146525, "context": 0.22603151015937328} +{"step": 73, "steps": 250, "lr": 4.3934841913948323e-05, "grad_norm": 0.9471169710159302, "loss_scale": 4.0, "retries": 0, "seconds": 28.3, "max_memory_gb": 16.47, "ce": 0.3684332705233828, "kl": 0.11612568562850356, "context": 0.4264961263979785} +{"step": 74, "steps": 250, "lr": 4.372705506523583e-05, "grad_norm": 0.270221471786499, "loss_scale": 4.0, "retries": 0, "seconds": 28.7, "max_memory_gb": 16.47, "ce": 0.07602135361776163, "kl": 0.10834721320134122, "context": 0.13019496022025123} +{"step": 75, "steps": 250, "lr": 4.351634338256017e-05, "grad_norm": 1.0820882320404053, "loss_scale": 4.0, "retries": 0, "seconds": 29.1, "max_memory_gb": 16.47, "ce": 0.4637548547470942, "kl": 0.9706895519047976, "context": 0.9490996268577874} +{"step": 76, "steps": 250, "lr": 4.330274484549825e-05, "grad_norm": 1.7432626485824585, "loss_scale": 4.0, "retries": 0, "seconds": 29.4, "max_memory_gb": 16.47, "ce": 0.005576135183218867, "kl": 1.531206208979711, "context": 0.7711792411282659} +{"step": 77, "steps": 250, "lr": 4.308629795396599e-05, "grad_norm": 0.3452589809894562, "loss_scale": 4.0, "retries": 0, "seconds": 29.8, "max_memory_gb": 16.47, "ce": 0.01312896772287786, "kl": 0.014786228392040357, "context": 0.020522082384559326} +{"step": 78, "steps": 250, "lr": 4.286704172127894e-05, "grad_norm": 0.9154940843582153, "loss_scale": 4.0, "retries": 0, "seconds": 30.2, "max_memory_gb": 16.47, "ce": 0.0019845255301333964, "kl": 0.08114306336210575, "context": 0.04255605611251667} +{"step": 79, "steps": 250, "lr": 4.264501566712033e-05, "grad_norm": 0.26931676268577576, "loss_scale": 4.0, "retries": 0, "seconds": 30.5, "max_memory_gb": 16.47, "ce": 0.025921085327354376, "kl": 0.04202663330488576, "context": 0.046934401980252005} +{"step": 80, "steps": 250, "lr": 4.2420259810417894e-05, "grad_norm": 1.0699714422225952, "loss_scale": 4.0, "retries": 0, "seconds": 30.9, "max_memory_gb": 16.47, "ce": 0.005854386288774549, "kl": 0.11774213060067495, "context": 0.06472545211363467} +{"step": 81, "steps": 250, "lr": 4.219281466213066e-05, "grad_norm": 0.8156082630157471, "loss_scale": 4.0, "retries": 0, "seconds": 31.2, "max_memory_gb": 16.47, "ce": 0.08651831519091502, "kl": 0.06888969092688058, "context": 0.12096315552480519} +{"step": 82, "steps": 250, "lr": 4.1962721217947135e-05, "grad_norm": 0.11501165479421616, "loss_scale": 4.0, "retries": 0, "seconds": 31.6, "max_memory_gb": 16.47, "ce": 0.011510388692840934, "kl": 0.003772030917389202, "context": 0.013396404567174613} +{"step": 83, "steps": 250, "lr": 4.173002095089599e-05, "grad_norm": 2.092132091522217, "loss_scale": 4.0, "retries": 0, "seconds": 32.0, "max_memory_gb": 16.47, "ce": 0.12405738295638002, "kl": 0.634978319700167, "context": 0.44154655071906745} +{"step": 84, "steps": 250, "lr": 4.149475580387084e-05, "grad_norm": 1.3233458995819092, "loss_scale": 4.0, "retries": 0, "seconds": 32.4, "max_memory_gb": 16.47, "ce": 0.049353984278241114, "kl": 0.04187541298870201, "context": 0.07029169155066484} +{"step": 85, "steps": 250, "lr": 4.1256968182070275e-05, "grad_norm": 0.8251069784164429, "loss_scale": 4.0, "retries": 0, "seconds": 32.8, "max_memory_gb": 16.47, "ce": 0.07316299463764153, "kl": 0.046257248103245274, "context": 0.09629161868906522} +{"step": 86, "steps": 250, "lr": 4.101670094535453e-05, "grad_norm": 1.5796597003936768, "loss_scale": 4.0, "retries": 0, "seconds": 33.1, "max_memory_gb": 16.47, "ce": 0.22184788656886667, "kl": 0.40838580249692313, "context": 0.42604078841395676} +{"step": 87, "steps": 250, "lr": 4.077399740052027e-05, "grad_norm": 0.9872198700904846, "loss_scale": 4.0, "retries": 0, "seconds": 33.5, "max_memory_gb": 16.47, "ce": 0.009165915922494605, "kl": 0.7055542243178934, "context": 0.3619430218823254} +{"step": 88, "steps": 250, "lr": 4.0528901293494775e-05, "grad_norm": 0.25431588292121887, "loss_scale": 4.0, "retries": 0, "seconds": 33.9, "max_memory_gb": 16.47, "ce": 0.012707001244052663, "kl": 0.01959943644510531, "context": 0.022506719917146256} +{"step": 89, "steps": 250, "lr": 4.0281456801451004e-05, "grad_norm": 0.07689652591943741, "loss_scale": 4.0, "retries": 0, "seconds": 34.2, "max_memory_gb": 16.47, "ce": 0.014672392970169312, "kl": 0.011188653214048827, "context": 0.020266718522179872} +{"step": 90, "steps": 250, "lr": 4.0031708524844884e-05, "grad_norm": 0.1032879427075386, "loss_scale": 4.0, "retries": 0, "seconds": 34.6, "max_memory_gb": 16.47, "ce": 0.010993881296599284, "kl": 0.008115267126413528, "context": 0.015051514375954866} +{"step": 91, "steps": 250, "lr": 3.977970147937635e-05, "grad_norm": 0.6612429618835449, "loss_scale": 4.0, "retries": 0, "seconds": 35.0, "max_memory_gb": 16.47, "ce": 0.1581683697659173, "kl": 0.5502810619436787, "context": 0.4333089053980075} +{"step": 92, "steps": 250, "lr": 3.952548108787553e-05, "grad_norm": 1.5353977680206299, "loss_scale": 4.0, "retries": 0, "seconds": 35.4, "max_memory_gb": 16.47, "ce": 0.37572451653977623, "kl": 0.42926444474142045, "context": 0.5903567541463417} +{"step": 93, "steps": 250, "lr": 3.9269093172115496e-05, "grad_norm": 0.8907526731491089, "loss_scale": 4.0, "retries": 0, "seconds": 35.8, "max_memory_gb": 16.47, "ce": 0.07501881229109131, "kl": 0.45712568430099054, "context": 0.30358166160294786} +{"step": 94, "steps": 250, "lr": 3.901058394455323e-05, "grad_norm": 1.1790088415145874, "loss_scale": 4.0, "retries": 0, "seconds": 36.2, "max_memory_gb": 16.47, "ce": 0.2124678185209632, "kl": 0.09898994903778657, "context": 0.2619627930398565} +{"step": 95, "steps": 250, "lr": 3.875e-05, "grad_norm": 1.687898874282837, "loss_scale": 4.0, "retries": 0, "seconds": 36.5, "max_memory_gb": 16.47, "ce": 0.5213842554658186, "kl": 0.026118438225239515, "context": 0.5344434762373567} +{"step": 96, "steps": 250, "lr": 3.8487388307223005e-05, "grad_norm": 1.1689647436141968, "loss_scale": 4.0, "retries": 0, "seconds": 36.9, "max_memory_gb": 16.47, "ce": 0.41177474657888524, "kl": 0.004433104957570322, "context": 0.4139912980608642} +{"step": 97, "steps": 250, "lr": 3.822279620047943e-05, "grad_norm": 0.9632822275161743, "loss_scale": 4.0, "retries": 0, "seconds": 37.3, "max_memory_gb": 16.47, "ce": 0.04259736306266859, "kl": 0.4333916368195787, "context": 0.2592931818217039} +{"step": 98, "steps": 250, "lr": 3.795627137098479e-05, "grad_norm": 0.6146625876426697, "loss_scale": 4.0, "retries": 0, "seconds": 37.7, "max_memory_gb": 16.47, "ce": 0.00371652607282158, "kl": 0.08380205888897763, "context": 0.04561755574832205} +{"step": 99, "steps": 250, "lr": 3.7687861858316844e-05, "grad_norm": 0.39156803488731384, "loss_scale": 4.0, "retries": 0, "seconds": 38.0, "max_memory_gb": 16.47, "ce": 0.04217839246848598, "kl": 0.013318693287146743, "context": 0.04883774027985055} +{"step": 100, "steps": 250, "lr": 3.741761604175668e-05, "grad_norm": 1.103805422782898, "loss_scale": 4.0, "retries": 0, "seconds": 38.4, "max_memory_gb": 16.47, "ce": 0.058493990916758776, "kl": 0.5362830762969679, "context": 0.3266355280065909} +{"step": 101, "steps": 250, "lr": 3.714558263156872e-05, "grad_norm": 0.28458812832832336, "loss_scale": 4.0, "retries": 0, "seconds": 38.8, "max_memory_gb": 16.47, "ce": 0.009554541087709367, "kl": 0.09514158536330797, "context": 0.05712533416226506} +{"step": 102, "steps": 250, "lr": 3.6871810660220904e-05, "grad_norm": 0.7398524284362793, "loss_scale": 4.0, "retries": 0, "seconds": 39.2, "max_memory_gb": 16.47, "ce": 0.03467882494442165, "kl": 0.015646414132788777, "context": 0.04250203096307814} +{"step": 103, "steps": 250, "lr": 3.659634947354686e-05, "grad_norm": 0.3240455389022827, "loss_scale": 4.0, "retries": 0, "seconds": 39.6, "max_memory_gb": 16.47, "ce": 0.020816633943468332, "kl": 0.03899145418108674, "context": 0.04031236085575074} +{"step": 104, "steps": 250, "lr": 3.631924872185169e-05, "grad_norm": 0.23146109282970428, "loss_scale": 4.0, "retries": 0, "seconds": 39.9, "max_memory_gb": 16.47, "ce": 0.008133341238135472, "kl": 0.0204123499988782, "context": 0.018339516020205338} +{"step": 105, "steps": 250, "lr": 3.6040558350962674e-05, "grad_norm": 1.2316596508026123, "loss_scale": 4.0, "retries": 0, "seconds": 40.3, "max_memory_gb": 16.47, "ce": 0.08627842932764906, "kl": 0.28396609362971503, "context": 0.22826147021260113} +{"step": 106, "steps": 250, "lr": 3.576032859322687e-05, "grad_norm": 0.07407756894826889, "loss_scale": 4.0, "retries": 0, "seconds": 40.7, "max_memory_gb": 16.47, "ce": 0.004106615291675553, "kl": 0.0011316654217807809, "context": 0.004672448048950173} +{"step": 107, "steps": 250, "lr": 3.5478609958457056e-05, "grad_norm": 2.3115715980529785, "loss_scale": 4.0, "retries": 0, "seconds": 41.1, "max_memory_gb": 16.47, "ce": 0.46876133265323006, "kl": 0.08135742007084446, "context": 0.5094400303642033} +{"step": 108, "steps": 250, "lr": 3.519545322482755e-05, "grad_norm": 0.3444644808769226, "loss_scale": 4.0, "retries": 0, "seconds": 41.4, "max_memory_gb": 16.47, "ce": 0.03545179047068814, "kl": 0.002393407071394904, "context": 0.03664849404594861} +{"step": 109, "steps": 250, "lr": 3.491090942972179e-05, "grad_norm": 2.1307523250579834, "loss_scale": 4.0, "retries": 0, "seconds": 41.8, "max_memory_gb": 16.47, "ce": 0.09097798162838444, "kl": 0.26249488443136215, "context": 0.2222254192456603} +{"step": 110, "steps": 250, "lr": 3.4625029860533135e-05, "grad_norm": 1.41319739818573, "loss_scale": 4.0, "retries": 0, "seconds": 42.2, "max_memory_gb": 16.47, "ce": 0.17683868203312159, "kl": 0.2137802231591195, "context": 0.2837287988513708} +{"step": 111, "steps": 250, "lr": 3.433786604542056e-05, "grad_norm": 0.026664255186915398, "loss_scale": 4.0, "retries": 0, "seconds": 42.5, "max_memory_gb": 16.47, "ce": 0.0015646039028069936, "kl": 0.001495811087806942, "context": 0.0023125095176510513} +{"step": 112, "steps": 250, "lr": 3.404946974402104e-05, "grad_norm": 0.2606724500656128, "loss_scale": 4.0, "retries": 0, "seconds": 42.9, "max_memory_gb": 16.47, "ce": 0.013972120432299562, "kl": 0.0007974577747518197, "context": 0.014370849370607175} +{"step": 113, "steps": 250, "lr": 3.375989293812019e-05, "grad_norm": 1.003149390220642, "loss_scale": 4.0, "retries": 0, "seconds": 43.3, "max_memory_gb": 16.47, "ce": 0.01589862850960344, "kl": 0.21297428314574063, "context": 0.12238577241078019} +{"step": 114, "steps": 250, "lr": 3.3469187822282766e-05, "grad_norm": 2.2295455932617188, "loss_scale": 4.0, "retries": 0, "seconds": 43.6, "max_memory_gb": 16.47, "ce": 0.020181935295113362, "kl": 0.7746704818273429, "context": 0.40751716651720926} +{"step": 115, "steps": 250, "lr": 3.317740679444499e-05, "grad_norm": 0.3809361755847931, "loss_scale": 4.0, "retries": 0, "seconds": 44.0, "max_memory_gb": 16.47, "ce": 0.018725941015873104, "kl": 0.05096399039030075, "context": 0.044207935337908566} +{"step": 116, "steps": 250, "lr": 3.2884602446470056e-05, "grad_norm": 1.2033956050872803, "loss_scale": 4.0, "retries": 0, "seconds": 44.4, "max_memory_gb": 16.47, "ce": 0.14461008729995228, "kl": 0.3844216663172233, "context": 0.33682090560614597} +{"step": 117, "steps": 250, "lr": 3.259082755466876e-05, "grad_norm": 1.3950989246368408, "loss_scale": 4.0, "retries": 0, "seconds": 44.8, "max_memory_gb": 16.47, "ce": 0.10672100364172366, "kl": 0.10078440966026392, "context": 0.15711320110131055} +{"step": 118, "steps": 250, "lr": 3.229613507028686e-05, "grad_norm": 0.4858960509300232, "loss_scale": 4.0, "retries": 0, "seconds": 45.1, "max_memory_gb": 16.47, "ce": 0.1820290530740749, "kl": 0.008557275924431451, "context": 0.18630768642469775} +{"step": 119, "steps": 250, "lr": 3.2000578109961004e-05, "grad_norm": 0.21822938323020935, "loss_scale": 4.0, "retries": 0, "seconds": 45.5, "max_memory_gb": 16.47, "ce": 0.033931256337382365, "kl": 0.01667159140026797, "context": 0.042267051481758244} +{"step": 120, "steps": 250, "lr": 3.1704209946144616e-05, "grad_norm": 2.6122398376464844, "loss_scale": 4.0, "retries": 0, "seconds": 45.9, "max_memory_gb": 16.47, "ce": 0.426505598181393, "kl": 0.15100115039967932, "context": 0.5020061662653461} +{"step": 121, "steps": 250, "lr": 3.140708399750594e-05, "grad_norm": 0.3272062838077545, "loss_scale": 4.0, "retries": 0, "seconds": 46.3, "max_memory_gb": 16.47, "ce": 0.07083234805759275, "kl": 0.044951880950975465, "context": 0.09330829084501602} +{"step": 122, "steps": 250, "lr": 3.1109253819299606e-05, "grad_norm": 0.6482672095298767, "loss_scale": 4.0, "retries": 0, "seconds": 46.7, "max_memory_gb": 16.47, "ce": 0.10292622481210856, "kl": 0.016833016234159004, "context": 0.11134273106290493} +{"step": 123, "steps": 250, "lr": 3.081077309371361e-05, "grad_norm": 0.9160653948783875, "loss_scale": 4.0, "retries": 0, "seconds": 47.0, "max_memory_gb": 16.47, "ce": 0.07066653715446591, "kl": 0.029950531956274062, "context": 0.08564180415123701} +{"step": 124, "steps": 250, "lr": 3.051169562019339e-05, "grad_norm": 1.7597171068191528, "loss_scale": 4.0, "retries": 0, "seconds": 47.4, "max_memory_gb": 16.47, "ce": 0.6099894177459646, "kl": 0.4873403614219569, "context": 0.8536596052581444} +{"step": 125, "steps": 250, "lr": 3.021207530574477e-05, "grad_norm": 0.44371581077575684, "loss_scale": 4.0, "retries": 0, "seconds": 47.8, "max_memory_gb": 16.47, "ce": 0.1635266748489812, "kl": 0.07616794607019983, "context": 0.20161064231069759} +{"step": 126, "steps": 250, "lr": 2.991196615521758e-05, "grad_norm": 1.5422512292861938, "loss_scale": 4.0, "retries": 0, "seconds": 48.1, "max_memory_gb": 16.47, "ce": 0.5987408012151718, "kl": 0.11619376996532083, "context": 0.6568376813083887} +{"step": 127, "steps": 250, "lr": 2.9611422261571536e-05, "grad_norm": 0.06374993175268173, "loss_scale": 4.0, "retries": 0, "seconds": 48.5, "max_memory_gb": 16.47, "ce": 0.0037477003497770056, "kl": 0.0011063418023695704, "context": 0.0043008712091250345} +{"step": 128, "steps": 250, "lr": 2.931049779612634e-05, "grad_norm": 2.90609073638916, "loss_scale": 4.0, "retries": 0, "seconds": 48.9, "max_memory_gb": 16.47, "ce": 0.23128023685421795, "kl": 0.7317641217669006, "context": 0.5971623013028875} +{"step": 129, "steps": 250, "lr": 2.9009246998797557e-05, "grad_norm": 1.3704829216003418, "loss_scale": 4.0, "retries": 0, "seconds": 49.3, "max_memory_gb": 16.47, "ce": 0.0839061014703475, "kl": 0.03113722954003606, "context": 0.09947471623308957} +{"step": 130, "steps": 250, "lr": 2.8707724168320238e-05, "grad_norm": 0.7709740996360779, "loss_scale": 4.0, "retries": 0, "seconds": 49.6, "max_memory_gb": 16.47, "ce": 0.05090293777175248, "kl": 0.31602115221176064, "context": 0.208913512295112} +{"step": 131, "steps": 250, "lr": 2.840598365246184e-05, "grad_norm": 1.1484224796295166, "loss_scale": 4.0, "retries": 0, "seconds": 50.0, "max_memory_gb": 16.47, "ce": 0.13161687966203317, "kl": 0.17231077142059803, "context": 0.21777226449921727} +{"step": 132, "steps": 250, "lr": 2.8104079838226383e-05, "grad_norm": 2.306229829788208, "loss_scale": 4.0, "retries": 0, "seconds": 50.4, "max_memory_gb": 16.47, "ce": 0.23444913735147566, "kl": 0.03915017869803705, "context": 0.25402422205661424} +{"step": 133, "steps": 250, "lr": 2.7802067142051446e-05, "grad_norm": 0.9727643132209778, "loss_scale": 4.0, "retries": 0, "seconds": 50.8, "max_memory_gb": 16.47, "ce": 0.30812741311092395, "kl": 0.03786984644830227, "context": 0.32706233114004135} +{"step": 134, "steps": 250, "lr": 2.7500000000000004e-05, "grad_norm": 0.6157997846603394, "loss_scale": 4.0, "retries": 0, "seconds": 51.1, "max_memory_gb": 16.47, "ce": 0.17459103028522804, "kl": 0.002068694026434059, "context": 0.17562537307094317} +{"step": 135, "steps": 250, "lr": 2.7197932857948566e-05, "grad_norm": 0.42000654339790344, "loss_scale": 4.0, "retries": 0, "seconds": 51.5, "max_memory_gb": 16.47, "ce": 0.0187924865167588, "kl": 0.03616737360425759, "context": 0.03687617293326184} +{"step": 136, "steps": 250, "lr": 2.689592016177363e-05, "grad_norm": 1.5617010593414307, "loss_scale": 4.0, "retries": 0, "seconds": 51.9, "max_memory_gb": 16.47, "ce": 0.16347522870637476, "kl": 0.17225511044671293, "context": 0.24960278428625315} +{"step": 137, "steps": 250, "lr": 2.659401634753817e-05, "grad_norm": 0.23609937727451324, "loss_scale": 4.0, "retries": 0, "seconds": 52.2, "max_memory_gb": 16.47, "ce": 0.018513087939936668, "kl": 0.016684538015397266, "context": 0.026855356874875724} +{"step": 138, "steps": 250, "lr": 2.6292275831679764e-05, "grad_norm": 2.640674352645874, "loss_scale": 4.0, "retries": 0, "seconds": 52.6, "max_memory_gb": 16.47, "ce": 0.1533635735977441, "kl": 0.19895915407687426, "context": 0.2528431573882699} +{"step": 139, "steps": 250, "lr": 2.5990753001202445e-05, "grad_norm": 0.5753344893455505, "loss_scale": 4.0, "retries": 0, "seconds": 53.0, "max_memory_gb": 16.47, "ce": 0.01417516719084233, "kl": 0.17562367348728003, "context": 0.101987000904046} +{"step": 140, "steps": 250, "lr": 2.5689502203873672e-05, "grad_norm": 0.1508786380290985, "loss_scale": 4.0, "retries": 0, "seconds": 53.4, "max_memory_gb": 16.47, "ce": 0.009626108178053983, "kl": 0.004571168523398228, "context": 0.01191169218509458} +{"step": 141, "steps": 250, "lr": 2.538857773842847e-05, "grad_norm": 0.09940017759799957, "loss_scale": 4.0, "retries": 0, "seconds": 53.8, "max_memory_gb": 16.47, "ce": 0.007711589008977171, "kl": 0.002533894316002261, "context": 0.008978535974165425} +{"step": 142, "steps": 250, "lr": 2.5088033844782426e-05, "grad_norm": 0.4540635645389557, "loss_scale": 4.0, "retries": 0, "seconds": 54.2, "max_memory_gb": 16.47, "ce": 0.05035934518673457, "kl": 0.022265802705078386, "context": 0.061492245411500335} +{"step": 143, "steps": 250, "lr": 2.4787924694255237e-05, "grad_norm": 1.64764404296875, "loss_scale": 4.0, "retries": 0, "seconds": 54.5, "max_memory_gb": 16.47, "ce": 0.3434288035496138, "kl": 0.037122779143828666, "context": 0.36199019983178005} +{"step": 144, "steps": 250, "lr": 2.4488304379806627e-05, "grad_norm": 0.8765448927879333, "loss_scale": 4.0, "retries": 0, "seconds": 54.9, "max_memory_gb": 16.47, "ce": 0.08237503044074401, "kl": 0.15535728106624447, "context": 0.16005367378238589} +{"step": 145, "steps": 250, "lr": 2.4189226906286388e-05, "grad_norm": 0.10428954660892487, "loss_scale": 4.0, "retries": 0, "seconds": 55.3, "max_memory_gb": 16.47, "ce": 0.003535341063980013, "kl": 0.007715902484051185, "context": 0.0073932922969106585} +{"step": 146, "steps": 250, "lr": 2.3890746180700395e-05, "grad_norm": 0.35915717482566833, "loss_scale": 4.0, "retries": 0, "seconds": 55.7, "max_memory_gb": 16.47, "ce": 0.03469140542438254, "kl": 0.007582697919133352, "context": 0.038482753559947014} +{"step": 147, "steps": 250, "lr": 2.359291600249407e-05, "grad_norm": 1.6535667181015015, "loss_scale": 4.0, "retries": 0, "seconds": 56.1, "max_memory_gb": 16.47, "ce": 0.373264098278014, "kl": 0.4573690151446499, "context": 0.6019486117875203} +{"step": 148, "steps": 250, "lr": 2.3295790053855392e-05, "grad_norm": 0.2816430926322937, "loss_scale": 4.0, "retries": 0, "seconds": 56.5, "max_memory_gb": 16.47, "ce": 0.013541343956603669, "kl": 0.03600516918231733, "context": 0.03154392752912827} +{"step": 149, "steps": 250, "lr": 2.299942189003901e-05, "grad_norm": 0.32513654232025146, "loss_scale": 4.0, "retries": 0, "seconds": 56.8, "max_memory_gb": 16.47, "ce": 0.0397790651768446, "kl": 0.003909030961949611, "context": 0.04173358163097873} +{"step": 150, "steps": 250, "lr": 2.2703864929713144e-05, "grad_norm": 0.423282265663147, "loss_scale": 4.0, "retries": 0, "seconds": 57.2, "max_memory_gb": 16.47, "ce": 0.061545148069853894, "kl": 0.01929120472050272, "context": 0.0711907485674601} +{"step": 151, "steps": 250, "lr": 2.2409172445331257e-05, "grad_norm": 0.3939381539821625, "loss_scale": 4.0, "retries": 0, "seconds": 57.6, "max_memory_gb": 16.47, "ce": 0.012647942290641367, "kl": 0.01364965246830252, "context": 0.019472768006380647} +{"step": 152, "steps": 250, "lr": 2.211539755352996e-05, "grad_norm": 1.7087198495864868, "loss_scale": 4.0, "retries": 0, "seconds": 58.0, "max_memory_gb": 16.47, "ce": 0.014376570441527292, "kl": 0.5900594571476176, "context": 0.3094062876771204} +{"step": 153, "steps": 250, "lr": 2.1822593205555016e-05, "grad_norm": 1.2607839107513428, "loss_scale": 4.0, "retries": 0, "seconds": 58.4, "max_memory_gb": 16.47, "ce": 0.014673112069431227, "kl": 0.41883540638445993, "context": 0.22409081817022525} +{"step": 154, "steps": 250, "lr": 2.1530812177717236e-05, "grad_norm": 1.3185656070709229, "loss_scale": 4.0, "retries": 0, "seconds": 58.8, "max_memory_gb": 16.47, "ce": 0.1410541163641028, "kl": 0.08757490083371522, "context": 0.18484156604972668} +{"step": 155, "steps": 250, "lr": 2.1240107061879815e-05, "grad_norm": 3.389451265335083, "loss_scale": 4.0, "retries": 0, "seconds": 59.1, "max_memory_gb": 16.47, "ce": 0.25146054945798824, "kl": 0.009440482960599184, "context": 0.2561807904203306} +{"step": 156, "steps": 250, "lr": 2.095053025597896e-05, "grad_norm": 0.35409489274024963, "loss_scale": 4.0, "retries": 0, "seconds": 59.5, "max_memory_gb": 16.47, "ce": 0.011382283992134035, "kl": 0.04879733971574751, "context": 0.035780953359790146} +{"step": 157, "steps": 250, "lr": 2.066213395457945e-05, "grad_norm": 0.7731274962425232, "loss_scale": 4.0, "retries": 0, "seconds": 59.9, "max_memory_gb": 16.47, "ce": 0.08323618758004159, "kl": 0.0508612533158157, "context": 0.10866681789048016} +{"step": 158, "steps": 250, "lr": 2.0374970139466877e-05, "grad_norm": 0.9155433773994446, "loss_scale": 4.0, "retries": 0, "seconds": 60.3, "max_memory_gb": 16.47, "ce": 0.16574142553145066, "kl": 0.07312176068080589, "context": 0.20230230828747153} +{"step": 159, "steps": 250, "lr": 2.0089090570278213e-05, "grad_norm": 0.6482465267181396, "loss_scale": 4.0, "retries": 0, "seconds": 60.6, "max_memory_gb": 16.47, "ce": 0.017321132792858407, "kl": 0.12422748806420714, "context": 0.07943487667944282} +{"step": 160, "steps": 250, "lr": 1.9804546775172457e-05, "grad_norm": 1.3648407459259033, "loss_scale": 4.0, "retries": 0, "seconds": 61.0, "max_memory_gb": 16.47, "ce": 0.05285628873389214, "kl": 0.28020469000330195, "context": 0.1929586308542639} +{"step": 161, "steps": 250, "lr": 1.9521390041542946e-05, "grad_norm": 1.6535708904266357, "loss_scale": 4.0, "retries": 0, "seconds": 61.4, "max_memory_gb": 16.47, "ce": 0.14873241024906747, "kl": 0.10015133182787395, "context": 0.19880807242589071} +{"step": 162, "steps": 250, "lr": 1.923967140677313e-05, "grad_norm": 1.228124976158142, "loss_scale": 4.0, "retries": 0, "seconds": 61.8, "max_memory_gb": 16.47, "ce": 0.14658278910792433, "kl": 0.20012966719514225, "context": 0.24664762275642715} +{"step": 163, "steps": 250, "lr": 1.8959441649037335e-05, "grad_norm": 3.0756523609161377, "loss_scale": 4.0, "retries": 0, "seconds": 62.1, "max_memory_gb": 16.47, "ce": 0.2736607240221929, "kl": 0.5433752180251759, "context": 0.5453483503661118} +{"step": 164, "steps": 250, "lr": 1.868075127814832e-05, "grad_norm": 0.6024351119995117, "loss_scale": 4.0, "retries": 0, "seconds": 62.5, "max_memory_gb": 16.47, "ce": 0.09910765997483395, "kl": 0.07888251366603072, "context": 0.13854891833034344} +{"step": 165, "steps": 250, "lr": 1.8403650526453143e-05, "grad_norm": 0.23399659991264343, "loss_scale": 4.0, "retries": 0, "seconds": 62.9, "max_memory_gb": 16.47, "ce": 0.013955814509245101, "kl": 0.01891580155643169, "context": 0.023413715534843504} +{"step": 166, "steps": 250, "lr": 1.812818933977912e-05, "grad_norm": 1.534469485282898, "loss_scale": 4.0, "retries": 0, "seconds": 63.3, "max_memory_gb": 16.47, "ce": 0.25932274910155684, "kl": 0.14050840091658756, "context": 0.3295769449323416} +{"step": 167, "steps": 250, "lr": 1.785441736843128e-05, "grad_norm": 0.7162711024284363, "loss_scale": 4.0, "retries": 0, "seconds": 63.7, "max_memory_gb": 16.47, "ce": 0.06563666253350675, "kl": 0.010640355059877038, "context": 0.07095683924853802} +{"step": 168, "steps": 250, "lr": 1.7582383958243314e-05, "grad_norm": 0.23449170589447021, "loss_scale": 4.0, "retries": 0, "seconds": 64.1, "max_memory_gb": 16.47, "ce": 0.015161020601226483, "kl": 0.003211345392628573, "context": 0.016766693588579074} +{"step": 169, "steps": 250, "lr": 1.7312138141683164e-05, "grad_norm": 1.7217001914978027, "loss_scale": 4.0, "retries": 0, "seconds": 64.4, "max_memory_gb": 16.47, "ce": 0.08413251704769209, "kl": 0.952951405197382, "context": 0.5606082323938608} +{"step": 170, "steps": 250, "lr": 1.704372862901521e-05, "grad_norm": 0.04280557483434677, "loss_scale": 4.0, "retries": 0, "seconds": 64.8, "max_memory_gb": 16.47, "ce": 0.002306578680872917, "kl": 0.00031897708504402544, "context": 0.0024660671770107} +{"step": 171, "steps": 250, "lr": 1.6777203799520573e-05, "grad_norm": 1.6938555240631104, "loss_scale": 4.0, "retries": 0, "seconds": 65.2, "max_memory_gb": 16.47, "ce": 0.04754379038058687, "kl": 0.17426306777633727, "context": 0.13467532116919756} +{"step": 172, "steps": 250, "lr": 1.6512611692777008e-05, "grad_norm": 0.5032947063446045, "loss_scale": 4.0, "retries": 0, "seconds": 65.6, "max_memory_gb": 16.47, "ce": 0.018851318513043225, "kl": 0.025597029700293206, "context": 0.03164983360329643} +{"step": 173, "steps": 250, "lr": 1.6250000000000005e-05, "grad_norm": 0.33910253643989563, "loss_scale": 4.0, "retries": 0, "seconds": 65.9, "max_memory_gb": 16.47, "ce": 0.018326543096918613, "kl": 0.007650433326489292, "context": 0.0221517599420622} +{"step": 174, "steps": 250, "lr": 1.598941605544678e-05, "grad_norm": 0.9617010354995728, "loss_scale": 4.0, "retries": 0, "seconds": 66.3, "max_memory_gb": 16.47, "ce": 0.04340904932178091, "kl": 0.042734970625588176, "context": 0.06477653277397621} +{"step": 175, "steps": 250, "lr": 1.5730906827884506e-05, "grad_norm": 1.3058862686157227, "loss_scale": 4.0, "retries": 0, "seconds": 66.7, "max_memory_gb": 16.47, "ce": 0.26696532836649567, "kl": 0.0374913098057732, "context": 0.28571098297834396} +{"step": 176, "steps": 250, "lr": 1.5474518912124476e-05, "grad_norm": 0.1587214469909668, "loss_scale": 4.0, "retries": 0, "seconds": 67.1, "max_memory_gb": 16.47, "ce": 0.007426342985127121, "kl": 0.0027485323807923123, "context": 0.008800609211903065} +{"step": 177, "steps": 250, "lr": 1.5220298520623646e-05, "grad_norm": 1.0197340250015259, "loss_scale": 4.0, "retries": 0, "seconds": 67.4, "max_memory_gb": 16.47, "ce": 0.02566056548675988, "kl": 0.6996946984472743, "context": 0.3755079165712232} +{"step": 178, "steps": 250, "lr": 1.4968291475155121e-05, "grad_norm": 0.6800634264945984, "loss_scale": 4.0, "retries": 0, "seconds": 67.8, "max_memory_gb": 16.47, "ce": 0.15252121147932485, "kl": 0.03588447594665922, "context": 0.17046344478148967} +{"step": 179, "steps": 250, "lr": 1.4718543198548998e-05, "grad_norm": 0.8461829423904419, "loss_scale": 4.0, "retries": 0, "seconds": 68.2, "max_memory_gb": 16.47, "ce": 0.41354010306531563, "kl": 0.009129666072112741, "context": 0.4181049325852655} +{"step": 180, "steps": 250, "lr": 1.4471098706505227e-05, "grad_norm": 1.2973538637161255, "loss_scale": 4.0, "retries": 0, "seconds": 68.5, "max_memory_gb": 16.47, "ce": 0.055293579905992374, "kl": 0.03419249832586502, "context": 0.07238983023853507} +{"step": 181, "steps": 250, "lr": 1.4226002599479743e-05, "grad_norm": 0.6040036678314209, "loss_scale": 4.0, "retries": 0, "seconds": 68.9, "max_memory_gb": 16.47, "ce": 0.01714630408969242, "kl": 0.15360285257338546, "context": 0.09394773273379542} +{"step": 182, "steps": 250, "lr": 1.3983299054645471e-05, "grad_norm": 0.7336249351501465, "loss_scale": 4.0, "retries": 0, "seconds": 69.3, "max_memory_gb": 16.47, "ce": 0.11308077222201973, "kl": 0.11812185827147914, "context": 0.17214170203078538} +{"step": 183, "steps": 250, "lr": 1.3743031817929724e-05, "grad_norm": 0.06218017637729645, "loss_scale": 4.0, "retries": 0, "seconds": 69.7, "max_memory_gb": 16.47, "ce": 0.0037850177905056626, "kl": 0.0006555576401297003, "context": 0.004112796625122428} +{"step": 184, "steps": 250, "lr": 1.3505244196129164e-05, "grad_norm": 0.07674992084503174, "loss_scale": 4.0, "retries": 0, "seconds": 70.1, "max_memory_gb": 16.47, "ce": 0.0031217253126669675, "kl": 0.002835927480191458, "context": 0.004539689020020887} +{"step": 185, "steps": 250, "lr": 1.3269979049104014e-05, "grad_norm": 0.8030687570571899, "loss_scale": 4.0, "retries": 0, "seconds": 70.5, "max_memory_gb": 16.47, "ce": 0.017595787707250565, "kl": 0.1121953700640006, "context": 0.07369347184430808} +{"step": 186, "steps": 250, "lr": 1.3037278782052863e-05, "grad_norm": 1.466934323310852, "loss_scale": 4.0, "retries": 0, "seconds": 70.9, "max_memory_gb": 16.47, "ce": 0.17902536226029042, "kl": 0.2646285150767653, "context": 0.31133960490114987} +{"step": 187, "steps": 250, "lr": 1.2807185337869343e-05, "grad_norm": 1.2443082332611084, "loss_scale": 4.0, "retries": 0, "seconds": 71.2, "max_memory_gb": 16.47, "ce": 0.03670887931366451, "kl": 0.43585354535025544, "context": 0.2546356492675841} +{"step": 188, "steps": 250, "lr": 1.2579740189582115e-05, "grad_norm": 1.0394351482391357, "loss_scale": 4.0, "retries": 0, "seconds": 71.6, "max_memory_gb": 16.47, "ce": 0.24790627509355545, "kl": 0.019005087700861623, "context": 0.2574088138353545} +{"step": 189, "steps": 250, "lr": 1.2354984332879668e-05, "grad_norm": 1.3132929801940918, "loss_scale": 4.0, "retries": 0, "seconds": 72.0, "max_memory_gb": 16.47, "ce": 0.3354244108486455, "kl": 0.092861259370693, "context": 0.38185504809371196} +{"step": 190, "steps": 250, "lr": 1.2132958278721064e-05, "grad_norm": 0.1094052717089653, "loss_scale": 4.0, "retries": 0, "seconds": 72.4, "max_memory_gb": 16.47, "ce": 0.0019801703965640627, "kl": 0.010180479439441115, "context": 0.007070409948937595} +{"step": 191, "steps": 250, "lr": 1.1913702046034016e-05, "grad_norm": 0.9418196082115173, "loss_scale": 4.0, "retries": 0, "seconds": 72.8, "max_memory_gb": 16.47, "ce": 0.1574037903919816, "kl": 0.08402810191182652, "context": 0.19941784534603357} +{"step": 192, "steps": 250, "lr": 1.1697255154501756e-05, "grad_norm": 1.490408182144165, "loss_scale": 4.0, "retries": 0, "seconds": 73.1, "max_memory_gb": 16.47, "ce": 0.485149459913373, "kl": 0.14138677113078302, "context": 0.5558428450021893} +{"step": 193, "steps": 250, "lr": 1.1483656617439842e-05, "grad_norm": 0.10789324343204498, "loss_scale": 4.0, "retries": 0, "seconds": 73.6, "max_memory_gb": 16.47, "ce": 0.005615483969449997, "kl": 0.0023205966881505447, "context": 0.00677578235627152} +{"step": 194, "steps": 250, "lr": 1.1272944934764177e-05, "grad_norm": 0.5004081130027771, "loss_scale": 4.0, "retries": 0, "seconds": 74.0, "max_memory_gb": 16.47, "ce": 0.0645717044826597, "kl": 0.02822284633293748, "context": 0.0786831303848885} +{"step": 195, "steps": 250, "lr": 1.1065158086051684e-05, "grad_norm": 0.18108510971069336, "loss_scale": 4.0, "retries": 0, "seconds": 74.4, "max_memory_gb": 16.47, "ce": 0.01270779746118933, "kl": 0.004684885861934163, "context": 0.015050240617711097} +{"step": 196, "steps": 250, "lr": 1.086033352369467e-05, "grad_norm": 2.215398073196411, "loss_scale": 4.0, "retries": 0, "seconds": 74.7, "max_memory_gb": 16.47, "ce": 0.0014248650841182098, "kl": 0.33015964144988175, "context": 0.1665046868729405} +{"step": 197, "steps": 250, "lr": 1.0658508166150224e-05, "grad_norm": 0.7424083948135376, "loss_scale": 4.0, "retries": 0, "seconds": 75.1, "max_memory_gb": 16.47, "ce": 0.03646673623006791, "kl": 0.6661994516962295, "context": 0.36956646654289216} +{"step": 198, "steps": 250, "lr": 1.0459718391285954e-05, "grad_norm": 0.5255900621414185, "loss_scale": 4.0, "retries": 0, "seconds": 75.5, "max_memory_gb": 16.47, "ce": 0.11584945549839176, "kl": 0.008037711209908593, "context": 0.11986831336980686} +{"step": 199, "steps": 250, "lr": 1.0264000029822999e-05, "grad_norm": 1.8391447067260742, "loss_scale": 4.0, "retries": 0, "seconds": 75.9, "max_memory_gb": 16.47, "ce": 0.12454358024115209, "kl": 0.03778899661392643, "context": 0.1434380780265201} +{"step": 200, "steps": 250, "lr": 1.0071388358877775e-05, "grad_norm": 0.7818291187286377, "loss_scale": 4.0, "retries": 0, "seconds": 76.2, "max_memory_gb": 16.47, "ce": 0.34851751499809325, "kl": 0.020121883702813648, "context": 0.35857845784630626} +{"step": 201, "steps": 250, "lr": 9.881918095603501e-06, "grad_norm": 0.5142010450363159, "loss_scale": 4.0, "retries": 0, "seconds": 76.6, "max_memory_gb": 16.47, "ce": 0.0739532053557923, "kl": 0.09347408126632217, "context": 0.12069024598167744} +{"step": 202, "steps": 250, "lr": 9.695623390932612e-06, "grad_norm": 0.060144100338220596, "loss_scale": 4.0, "retries": 0, "seconds": 77.0, "max_memory_gb": 16.47, "ce": 0.0020258083677617833, "kl": 0.003010268395428284, "context": 0.0035309425875311717} +{"step": 203, "steps": 250, "lr": 9.51253782342123e-06, "grad_norm": 0.8705191612243652, "loss_scale": 4.0, "retries": 0, "seconds": 77.3, "max_memory_gb": 16.47, "ce": 0.06307820872461889, "kl": 0.07465846079867333, "context": 0.10040744283469394} +{"step": 204, "steps": 250, "lr": 9.332694393196846e-06, "grad_norm": 0.6567627191543579, "loss_scale": 4.0, "retries": 0, "seconds": 77.7, "max_memory_gb": 16.47, "ce": 0.03844300241325982, "kl": 0.00752836384344846, "context": 0.04220718389842659} +{"step": 205, "steps": 250, "lr": 9.156125516010256e-06, "grad_norm": 1.5584237575531006, "loss_scale": 4.0, "retries": 0, "seconds": 78.1, "max_memory_gb": 16.47, "ce": 0.009527946298476309, "kl": 0.3570271915086778, "context": 0.18804154571262188} +{"step": 206, "steps": 250, "lr": 8.982863017392733e-06, "grad_norm": 0.11572178453207016, "loss_scale": 4.0, "retries": 0, "seconds": 78.4, "max_memory_gb": 16.47, "ce": 0.005019896474550478, "kl": 0.0020912714326186688, "context": 0.0060655323177343234} +{"step": 207, "steps": 250, "lr": 8.81293812691974e-06, "grad_norm": 0.25836363434791565, "loss_scale": 4.0, "retries": 0, "seconds": 78.8, "max_memory_gb": 16.47, "ce": 0.07002195087261498, "kl": 0.0471033600915689, "context": 0.0935736270621419} +{"step": 208, "steps": 250, "lr": 8.646381472581927e-06, "grad_norm": 0.08919896930456161, "loss_scale": 4.0, "retries": 0, "seconds": 79.2, "max_memory_gb": 16.47, "ce": 0.0038312537071760744, "kl": 0.001783654992323136, "context": 0.004723081292468123} +{"step": 209, "steps": 250, "lr": 8.483223075264622e-06, "grad_norm": 0.5937061309814453, "loss_scale": 4.0, "retries": 0, "seconds": 79.6, "max_memory_gb": 16.47, "ce": 0.16860436991555616, "kl": 0.04636991332517937, "context": 0.19178932750946842} +{"step": 210, "steps": 250, "lr": 8.323492343336747e-06, "grad_norm": 1.0133178234100342, "loss_scale": 4.0, "retries": 0, "seconds": 80.0, "max_memory_gb": 16.47, "ce": 0.46624342462746426, "kl": 0.11074703005579067, "context": 0.521616913494654} +{"step": 211, "steps": 250, "lr": 8.167218067350122e-06, "grad_norm": 0.6531597375869751, "loss_scale": 4.0, "retries": 0, "seconds": 80.4, "max_memory_gb": 16.47, "ce": 0.11403090530075133, "kl": 0.11175189632922411, "context": 0.16990684706252068} +{"step": 212, "steps": 250, "lr": 8.01442841485013e-06, "grad_norm": 1.4864791631698608, "loss_scale": 4.0, "retries": 0, "seconds": 80.7, "max_memory_gb": 16.47, "ce": 0.04915522175724618, "kl": 0.8859698821033817, "context": 0.49214015633333474} +{"step": 213, "steps": 250, "lr": 7.865150925298678e-06, "grad_norm": 0.849037230014801, "loss_scale": 4.0, "retries": 0, "seconds": 81.1, "max_memory_gb": 16.47, "ce": 0.2380294131289702, "kl": 0.030955231388361426, "context": 0.2535070267331321} +{"step": 214, "steps": 250, "lr": 7.719412505110375e-06, "grad_norm": 3.607100009918213, "loss_scale": 4.0, "retries": 0, "seconds": 81.4, "max_memory_gb": 16.47, "ce": 0.4058263941842597, "kl": 0.12988696433603764, "context": 0.47076985659077764} +{"step": 215, "steps": 250, "lr": 7.57723942280278e-06, "grad_norm": 1.3108117580413818, "loss_scale": 4.0, "retries": 0, "seconds": 81.8, "max_memory_gb": 16.47, "ce": 0.5340528840606567, "kl": 0.02390852714893299, "context": 0.5460071610868908} +{"step": 216, "steps": 250, "lr": 7.438657304261687e-06, "grad_norm": 1.063254952430725, "loss_scale": 4.0, "retries": 0, "seconds": 82.2, "max_memory_gb": 16.47, "ce": 0.040259367786347866, "kl": 0.13043024099897593, "context": 0.10547448880970478} +{"step": 217, "steps": 250, "lr": 7.30369112812219e-06, "grad_norm": 0.08914871513843536, "loss_scale": 4.0, "retries": 0, "seconds": 82.6, "max_memory_gb": 16.47, "ce": 0.0038588158786296844, "kl": 0.0045954680767863465, "context": 0.0061565499054268} +{"step": 218, "steps": 250, "lr": 7.172365221266403e-06, "grad_norm": 0.7415753602981567, "loss_scale": 4.0, "retries": 0, "seconds": 83.0, "max_memory_gb": 16.47, "ce": 0.19583761971443892, "kl": 0.1829968011006713, "context": 0.2873360179364681} +{"step": 219, "steps": 250, "lr": 7.0447032544387154e-06, "grad_norm": 0.22523267567157745, "loss_scale": 4.0, "retries": 0, "seconds": 83.3, "max_memory_gb": 16.47, "ce": 0.018487931505660526, "kl": 0.06847485355683602, "context": 0.05272535956464708} +{"step": 220, "steps": 250, "lr": 6.920728237979274e-06, "grad_norm": 0.666539192199707, "loss_scale": 4.0, "retries": 0, "seconds": 83.7, "max_memory_gb": 16.47, "ce": 0.04661873294389807, "kl": 0.007183560617704643, "context": 0.05021051276708022} +{"step": 221, "steps": 250, "lr": 6.800462517676456e-06, "grad_norm": 1.8918818235397339, "loss_scale": 4.0, "retries": 0, "seconds": 84.1, "max_memory_gb": 16.47, "ce": 0.00478581314382609, "kl": 0.3446688579933834, "context": 0.17712024180218577} +{"step": 222, "steps": 250, "lr": 6.683927770739222e-06, "grad_norm": 1.2580595016479492, "loss_scale": 4.0, "retries": 0, "seconds": 84.5, "max_memory_gb": 16.47, "ce": 0.02413000285741873, "kl": 0.056485004231490166, "context": 0.052372505713719875} +{"step": 223, "steps": 250, "lr": 6.571145001889892e-06, "grad_norm": 0.6261196136474609, "loss_scale": 4.0, "retries": 0, "seconds": 84.9, "max_memory_gb": 16.47, "ce": 0.0556555257353466, "kl": 0.0044502636446850374, "context": 0.05788065874367021} +{"step": 224, "steps": 250, "lr": 6.46213453957817e-06, "grad_norm": 0.6913971304893494, "loss_scale": 4.0, "retries": 0, "seconds": 85.2, "max_memory_gb": 16.47, "ce": 0.0538810086727608, "kl": 0.08219753256707918, "context": 0.0949797754874453} +{"step": 225, "steps": 250, "lr": 6.356916032317063e-06, "grad_norm": 1.2869007587432861, "loss_scale": 4.0, "retries": 0, "seconds": 85.6, "max_memory_gb": 16.47, "ce": 0.007214898447273299, "kl": 0.11727036199954455, "context": 0.06585007975809276} +{"step": 226, "steps": 250, "lr": 6.255508445141349e-06, "grad_norm": 0.7540601491928101, "loss_scale": 4.0, "retries": 0, "seconds": 86.0, "max_memory_gb": 16.47, "ce": 0.10156975369318388, "kl": 0.019935157732106745, "context": 0.11153733351966366} +{"step": 227, "steps": 250, "lr": 6.157930056189228e-06, "grad_norm": 0.4910772740840912, "loss_scale": 4.0, "retries": 0, "seconds": 86.3, "max_memory_gb": 16.47, "ce": 0.06049316137796268, "kl": 0.006443053913244512, "context": 0.06371468855650164} +{"step": 228, "steps": 250, "lr": 6.064198453407796e-06, "grad_norm": 1.4194769859313965, "loss_scale": 4.0, "retries": 0, "seconds": 86.7, "max_memory_gb": 16.47, "ce": 0.014695379315526225, "kl": 0.17346279649063945, "context": 0.1014267758000642} +{"step": 229, "steps": 250, "lr": 5.974330531382935e-06, "grad_norm": 0.4714776873588562, "loss_scale": 4.0, "retries": 0, "seconds": 87.1, "max_memory_gb": 16.47, "ce": 0.0855556153546786, "kl": 0.016333155815118516, "context": 0.09372219443321228} +{"step": 230, "steps": 250, "lr": 5.888342488294125e-06, "grad_norm": 1.5537902116775513, "loss_scale": 4.0, "retries": 0, "seconds": 87.5, "max_memory_gb": 16.47, "ce": 0.3353754512500018, "kl": 0.22757658828049898, "context": 0.4491637524915859} +{"step": 231, "steps": 250, "lr": 5.806249822994849e-06, "grad_norm": 0.8339364528656006, "loss_scale": 4.0, "retries": 0, "seconds": 87.8, "max_memory_gb": 16.47, "ce": 0.055992231995332986, "kl": 0.11839321695151739, "context": 0.1151888442109339} +{"step": 232, "steps": 250, "lr": 5.728067332218994e-06, "grad_norm": 0.5641066431999207, "loss_scale": 4.0, "retries": 0, "seconds": 88.2, "max_memory_gb": 16.47, "ce": 0.02124386018840596, "kl": 0.00885069889773149, "context": 0.025669209222542122} +{"step": 233, "steps": 250, "lr": 5.65380910791383e-06, "grad_norm": 0.7081896066665649, "loss_scale": 4.0, "retries": 0, "seconds": 88.6, "max_memory_gb": 16.47, "ce": 0.017782746348530054, "kl": 0.1475083266850561, "context": 0.09153691027313471} +{"step": 234, "steps": 250, "lr": 5.583488534700022e-06, "grad_norm": 0.3769172728061676, "loss_scale": 4.0, "retries": 0, "seconds": 89.0, "max_memory_gb": 16.47, "ce": 0.04113022590172477, "kl": 0.03511330056062434, "context": 0.05868687623296864} +{"step": 235, "steps": 250, "lr": 5.517118287459118e-06, "grad_norm": 1.0042308568954468, "loss_scale": 4.0, "retries": 0, "seconds": 89.4, "max_memory_gb": 16.47, "ce": 0.22526582670980133, "kl": 0.026682250914745964, "context": 0.23860695547773503} +{"step": 236, "steps": 250, "lr": 5.454710329048978e-06, "grad_norm": 0.37550005316734314, "loss_scale": 4.0, "retries": 0, "seconds": 89.8, "max_memory_gb": 16.47, "ce": 0.014762225182494149, "kl": 0.02047308747205534, "context": 0.0249987686984241} +{"step": 237, "steps": 250, "lr": 5.3962759081475436e-06, "grad_norm": 0.44837456941604614, "loss_scale": 4.0, "retries": 0, "seconds": 90.1, "max_memory_gb": 16.47, "ce": 0.10213250410743058, "kl": 0.0246649350028747, "context": 0.11446497173164971} +{"step": 238, "steps": 250, "lr": 5.341825557225321e-06, "grad_norm": 0.3780568540096283, "loss_scale": 4.0, "retries": 0, "seconds": 90.5, "max_memory_gb": 16.47, "ce": 0.03611299971817061, "kl": 0.013946773751740693, "context": 0.043086387173389085} +{"step": 239, "steps": 250, "lr": 5.291369090646961e-06, "grad_norm": 0.5128703117370605, "loss_scale": 4.0, "retries": 0, "seconds": 90.8, "max_memory_gb": 16.47, "ce": 0.12954806687776, "kl": 0.02059185587859247, "context": 0.13984399114269763} +{"step": 240, "steps": 250, "lr": 5.244915602902289e-06, "grad_norm": 1.7161273956298828, "loss_scale": 4.0, "retries": 0, "seconds": 91.2, "max_memory_gb": 16.47, "ce": 0.13532035559182987, "kl": 0.027717215809389018, "context": 0.1491789638530463} +{"step": 241, "steps": 250, "lr": 5.20247346696706e-06, "grad_norm": 0.6830705404281616, "loss_scale": 4.0, "retries": 0, "seconds": 91.6, "max_memory_gb": 16.47, "ce": 0.19186534800974187, "kl": 0.020527619912172668, "context": 0.20212915621232241} +{"step": 242, "steps": 250, "lr": 5.164050332793787e-06, "grad_norm": 0.24240662157535553, "loss_scale": 4.0, "retries": 0, "seconds": 92.0, "max_memory_gb": 16.47, "ce": 0.008872782433172688, "kl": 0.01967269080614642, "context": 0.01870912784943357} +{"step": 243, "steps": 250, "lr": 5.129653125932873e-06, "grad_norm": 0.08302410691976547, "loss_scale": 4.0, "retries": 0, "seconds": 92.3, "max_memory_gb": 16.47, "ce": 0.003994766273535788, "kl": 0.003012652858160436, "context": 0.005501092702616006} +{"step": 244, "steps": 250, "lr": 5.099288046284329e-06, "grad_norm": 1.138001561164856, "loss_scale": 4.0, "retries": 0, "seconds": 92.7, "max_memory_gb": 16.47, "ce": 0.07862470201507676, "kl": 0.1288427395575127, "context": 0.14304606667428743} +{"step": 245, "steps": 250, "lr": 5.072960566980277e-06, "grad_norm": 0.9014365673065186, "loss_scale": 4.0, "retries": 0, "seconds": 93.1, "max_memory_gb": 16.47, "ce": 0.3761249631352257, "kl": 0.034601066072355025, "context": 0.3934254952473566} +{"step": 246, "steps": 250, "lr": 5.050675433398435e-06, "grad_norm": 0.0575554184615612, "loss_scale": 4.0, "retries": 0, "seconds": 93.5, "max_memory_gb": 16.47, "ce": 0.0029902705718996003, "kl": 0.0010475618037162349, "context": 0.0035140514955855906} +{"step": 247, "steps": 250, "lr": 5.032436662306811e-06, "grad_norm": 1.3495700359344482, "loss_scale": 4.0, "retries": 0, "seconds": 93.9, "max_memory_gb": 16.47, "ce": 0.045711888902587816, "kl": 0.43700347534468165, "context": 0.26421362462861} +{"step": 248, "steps": 250, "lr": 5.018247541139684e-06, "grad_norm": 0.7220816612243652, "loss_scale": 4.0, "retries": 0, "seconds": 94.3, "max_memory_gb": 16.47, "ce": 0.03527180524542928, "kl": 0.032312741968780756, "context": 0.05142817588057369} +{"step": 249, "steps": 250, "lr": 5.00811062740508e-06, "grad_norm": 1.5349928140640259, "loss_scale": 4.0, "retries": 0, "seconds": 94.6, "max_memory_gb": 16.47, "ce": 0.04839531797915697, "kl": 0.05487796082161367, "context": 0.07583429664373398} +{"step": 250, "steps": 250, "lr": 5.002027748223778e-06, "grad_norm": 0.21956956386566162, "loss_scale": 4.0, "retries": 0, "seconds": 95.0, "max_memory_gb": 16.47, "ce": 0.01584162237122655, "kl": 0.007120237700291909, "context": 0.019401741272304207} diff --git a/runs/single_internvl/replacement/L01_s13/adapter/README.md b/runs/single_internvl/replacement/L01_s13/adapter/README.md new file mode 100644 index 0000000000000000000000000000000000000000..4f9f856cf0e1531b47e25d266833b2977a19d3ac --- /dev/null +++ b/runs/single_internvl/replacement/L01_s13/adapter/README.md @@ -0,0 +1,207 @@ +--- +base_model: /data/OPSD/assets/models/InternVL3_5-8B-HF +library_name: peft +pipeline_tag: text-generation +tags: +- base_model:adapter:/data/OPSD/assets/models/InternVL3_5-8B-HF +- lora +- transformers +--- + +# Model Card for Model ID + + + + + +## Model Details + +### Model Description + + + + + +- **Developed by:** [More Information Needed] +- **Funded by [optional]:** [More Information Needed] +- **Shared by [optional]:** [More Information Needed] +- **Model type:** [More Information Needed] +- **Language(s) (NLP):** [More Information Needed] +- **License:** [More Information Needed] +- **Finetuned from model [optional]:** [More Information Needed] + +### Model Sources [optional] + + + +- **Repository:** [More Information Needed] +- **Paper [optional]:** [More Information Needed] +- **Demo [optional]:** [More Information Needed] + +## Uses + + + +### Direct Use + + + +[More Information Needed] + +### Downstream Use [optional] + + + +[More Information Needed] + +### Out-of-Scope Use + + + +[More Information Needed] + +## Bias, Risks, and Limitations + + + +[More Information Needed] + +### Recommendations + + + +Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. More information needed for further recommendations. + +## How to Get Started with the Model + +Use the code below to get started with the model. + +[More Information Needed] + +## Training Details + +### Training Data + + + +[More Information Needed] + +### Training Procedure + + + +#### Preprocessing [optional] + +[More Information Needed] + + +#### Training Hyperparameters + +- **Training regime:** [More Information Needed] + +#### Speeds, Sizes, Times [optional] + + + +[More Information Needed] + +## Evaluation + + + +### Testing Data, Factors & Metrics + +#### Testing Data + + + +[More Information Needed] + +#### Factors + + + +[More Information Needed] + +#### Metrics + + + +[More Information Needed] + +### Results + +[More Information Needed] + +#### Summary + + + +## Model Examination [optional] + + + +[More Information Needed] + +## Environmental Impact + + + +Carbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700). + +- **Hardware Type:** [More Information Needed] +- **Hours used:** [More Information Needed] +- **Cloud Provider:** [More Information Needed] +- **Compute Region:** [More Information Needed] +- **Carbon Emitted:** [More Information Needed] + +## Technical Specifications [optional] + +### Model Architecture and Objective + +[More Information Needed] + +### Compute Infrastructure + +[More Information Needed] + +#### Hardware + +[More Information Needed] + +#### Software + +[More Information Needed] + +## Citation [optional] + + + +**BibTeX:** + +[More Information Needed] + +**APA:** + +[More Information Needed] + +## Glossary [optional] + + + +[More Information Needed] + +## More Information [optional] + +[More Information Needed] + +## Model Card Authors [optional] + +[More Information Needed] + +## Model Card Contact + +[More Information Needed] +### Framework versions + +- PEFT 0.20.0 \ No newline at end of file diff --git a/runs/single_internvl/replacement/L01_s13/adapter/adapter_config.json b/runs/single_internvl/replacement/L01_s13/adapter/adapter_config.json new file mode 100644 index 0000000000000000000000000000000000000000..232af3a1e06e46afcd4c1ad0dd6e7cf3bb96296c --- /dev/null +++ b/runs/single_internvl/replacement/L01_s13/adapter/adapter_config.json @@ -0,0 +1,50 @@ +{ + "alora_invocation_tokens": null, + "alpha_pattern": {}, + "arrow_config": null, + "auto_mapping": null, + "base_model_name_or_path": "/data/OPSD/assets/models/InternVL3_5-8B-HF", + "bias": "none", + "corda_config": null, + "ensure_weight_tying": false, + "eva_config": null, + "exclude_modules": null, + "fan_in_fan_out": false, + "inference_mode": true, + "init_lora_weights": true, + "layer_replication": null, + "layers_pattern": null, + "layers_to_transform": null, + "loftq_config": {}, + "lora_alpha": 32, + "lora_bias": false, + "lora_dropout": 0.05, + "lora_ga_config": null, + "megatron_config": null, + "megatron_core": "megatron.core", + "modules_to_save": null, + "monteclora_config": null, + "peft_type": "LORA", + "peft_version": "0.20.0", + "qalora_group_size": 16, + "r": 16, + "rank_pattern": {}, + "revision": null, + "target_modules": [ + "model.language_model.layers.1.mlp.gate_proj", + "model.language_model.layers.1.mlp.down_proj", + "model.language_model.layers.1.self_attn.q_proj", + "model.language_model.layers.1.self_attn.v_proj", + "model.language_model.layers.1.self_attn.o_proj", + "model.language_model.layers.1.self_attn.k_proj", + "model.language_model.layers.1.mlp.up_proj" + ], + "target_parameters": null, + "task_type": "CAUSAL_LM", + "trainable_token_indices": null, + "use_bdlora": null, + "use_dora": false, + "use_qalora": false, + "use_rslora": false, + "velora_config": null +} \ No newline at end of file diff --git a/runs/single_internvl/replacement/L01_s13/config.json b/runs/single_internvl/replacement/L01_s13/config.json new file mode 100644 index 0000000000000000000000000000000000000000..5276da981e4207ad5fcdbaa293f43c4243019a11 --- /dev/null +++ b/runs/single_internvl/replacement/L01_s13/config.json @@ -0,0 +1,24 @@ +{ + "behavior": "replacement", + "seed": 13, + "model": "/data/OPSD/assets/models/InternVL3_5-8B-HF", + "backbone": "internvl", + "train": "/data/OPSD/assets/v3/replacement.jsonl", + "items": 1000, + "steps": 250, + "epochs": 1, + "steps_per_epoch": 250, + "groups_per_step": 4, + "lr": 5e-05, + "scheduler": "cosine", + "warmup_steps": 16, + "final_fraction": 0.1, + "weight_decay": 0.0, + "clip": 1.0, + "initial_loss_scale": 4.0, + "adapter": { + "layer": 1, + "rank": 16, + "trainable_parameters": 1212416 + } +} \ No newline at end of file diff --git a/runs/single_internvl/replacement/L01_s13/steps.jsonl b/runs/single_internvl/replacement/L01_s13/steps.jsonl new file mode 100644 index 0000000000000000000000000000000000000000..bcfe85d0b96da658d122f9ca4b631cb9a85d9fc3 --- /dev/null +++ b/runs/single_internvl/replacement/L01_s13/steps.jsonl @@ -0,0 +1,250 @@ +{"step": 1, "steps": 250, "lr": 3.125e-06, "grad_norm": 0.9593239426612854, "loss_scale": 4.0, "retries": 0, "seconds": 1.3, "max_memory_gb": 24.84, "pair": 1.0986123085021973, "sign": 0.7328193015418947, "keep": 1.137254401575774, "leak": 0.0, "replacement": 2.0336491465568542} +{"step": 2, "steps": 250, "lr": 6.25e-06, "grad_norm": 0.8836205005645752, "loss_scale": 4.0, "retries": 0, "seconds": 2.4, "max_memory_gb": 24.84, "pair": 1.100912630558014, "sign": 0.18113453849218786, "keep": 1.3659147868456785, "leak": 4.569975430968043e-05, "replacement": 1.8744487464427948} +{"step": 3, "steps": 250, "lr": 9.375000000000001e-06, "grad_norm": 2.0255348682403564, "loss_scale": 4.0, "retries": 0, "seconds": 3.5, "max_memory_gb": 24.84, "pair": 1.096558153629303, "sign": 0.1154714091680944, "keep": 2.063725635409355, "leak": 0.0001009879802040814, "replacement": 2.186181962490082} +{"step": 4, "steps": 250, "lr": 1.25e-05, "grad_norm": 1.3091331720352173, "loss_scale": 4.0, "retries": 0, "seconds": 4.6, "max_memory_gb": 24.84, "pair": 1.0949964225292206, "sign": 0.03801619075238705, "keep": 0.8714177075889893, "leak": 5.002560396860645e-05, "replacement": 1.549725890159607} +{"step": 5, "steps": 250, "lr": 1.5625e-05, "grad_norm": 0.5621684193611145, "loss_scale": 4.0, "retries": 0, "seconds": 5.7, "max_memory_gb": 24.84, "pair": 1.1011099517345428, "sign": 0.11726495902985334, "keep": 0.02337648676621029, "leak": 6.079615431886509e-05, "replacement": 1.1714458763599396} +{"step": 6, "steps": 250, "lr": 1.8750000000000002e-05, "grad_norm": 0.7710880041122437, "loss_scale": 4.0, "retries": 0, "seconds": 6.7, "max_memory_gb": 24.84, "pair": 1.1010949909687042, "sign": 0.20000055711716413, "keep": 0.6361354511464015, "leak": 3.651940033932988e-05, "replacement": 1.5191721022129059} +{"step": 7, "steps": 250, "lr": 2.1875e-05, "grad_norm": 1.5538718700408936, "loss_scale": 4.0, "retries": 0, "seconds": 7.8, "max_memory_gb": 24.84, "pair": 1.0963671505451202, "sign": 0.08985515031963587, "keep": 1.5254834839870455, "leak": 7.902192828623811e-05, "replacement": 1.9040562808513641} +{"step": 8, "steps": 250, "lr": 2.5e-05, "grad_norm": 1.6165627241134644, "loss_scale": 4.0, "retries": 0, "seconds": 8.9, "max_memory_gb": 24.88, "pair": 1.090606689453125, "sign": 0.20960095198825002, "keep": 1.4586929976940155, "leak": 0.00017665238647168735, "replacement": 1.9247978031635284} +{"step": 9, "steps": 250, "lr": 2.8125000000000003e-05, "grad_norm": 4.158328056335449, "loss_scale": 4.0, "retries": 0, "seconds": 10.0, "max_memory_gb": 24.88, "pair": 1.0889300405979156, "sign": 0.5744876121170819, "keep": 1.380729965865612, "leak": 2.7532164441623763e-05, "replacement": 2.0665457248687744} +{"step": 10, "steps": 250, "lr": 3.125e-05, "grad_norm": 0.6829278469085693, "loss_scale": 4.0, "retries": 0, "seconds": 11.1, "max_memory_gb": 24.88, "pair": 1.094979465007782, "sign": 0.0802498459815979, "keep": 0.09905106155201793, "leak": 9.233138098352356e-05, "replacement": 1.1846530139446259} +{"step": 11, "steps": 250, "lr": 3.4375e-05, "grad_norm": 1.3712434768676758, "loss_scale": 4.0, "retries": 0, "seconds": 12.2, "max_memory_gb": 24.88, "pair": 1.0926148891448975, "sign": 0.20463008247315884, "keep": 1.3226584108633688, "leak": 8.867381075106096e-05, "replacement": 1.8562813103199005} +{"step": 12, "steps": 250, "lr": 3.7500000000000003e-05, "grad_norm": 1.155125617980957, "loss_scale": 4.0, "retries": 0, "seconds": 13.3, "max_memory_gb": 24.88, "pair": 1.086828202009201, "sign": 0.07588598993606865, "keep": 0.9347169754619244, "leak": 0.00013547539482772208, "replacement": 1.592163622379303} +{"step": 13, "steps": 250, "lr": 4.0625000000000005e-05, "grad_norm": 0.9987849593162537, "loss_scale": 4.0, "retries": 0, "seconds": 14.3, "max_memory_gb": 24.88, "pair": 1.0809769034385681, "sign": 0.05775862652808428, "keep": 0.8571424642868806, "leak": 0.00036681828265727745, "replacement": 1.53851917386055} +{"step": 14, "steps": 250, "lr": 4.375e-05, "grad_norm": 1.6895848512649536, "loss_scale": 4.0, "retries": 0, "seconds": 15.4, "max_memory_gb": 24.88, "pair": 1.0866911113262177, "sign": 0.26730811037123203, "keep": 1.6208590865135193, "leak": 0.0003212282181266346, "replacement": 2.030855029821396} +{"step": 15, "steps": 250, "lr": 4.6875e-05, "grad_norm": 1.566087245941162, "loss_scale": 4.0, "retries": 0, "seconds": 16.5, "max_memory_gb": 24.88, "pair": 1.0758999586105347, "sign": 0.19864547508768737, "keep": 1.008882045336577, "leak": 0.0009738420776557177, "replacement": 1.679907202720642} +{"step": 16, "steps": 250, "lr": 5e-05, "grad_norm": 1.0744260549545288, "loss_scale": 4.0, "retries": 0, "seconds": 17.5, "max_memory_gb": 24.88, "pair": 1.0773667693138123, "sign": 0.30122254928573966, "keep": 0.37775913847144693, "leak": 0.0003974992396251764, "replacement": 1.4169570207595825} +{"step": 17, "steps": 250, "lr": 5e-05, "grad_norm": 1.2805174589157104, "loss_scale": 4.0, "retries": 0, "seconds": 18.6, "max_memory_gb": 24.88, "pair": 1.0755679309368134, "sign": 0.7401372450403869, "keep": 0.8349960019113496, "leak": 0.0009162000933429226, "replacement": 1.8633635640144348} +{"step": 18, "steps": 250, "lr": 4.999797225177623e-05, "grad_norm": 1.6800304651260376, "loss_scale": 4.0, "retries": 0, "seconds": 19.7, "max_memory_gb": 24.91, "pair": 1.0667639374732971, "sign": 0.1887393263168633, "keep": 1.0721882901561912, "leak": 0.0012781343248207122, "replacement": 1.6975472569465637} +{"step": 19, "steps": 250, "lr": 4.999188937259492e-05, "grad_norm": 2.6647350788116455, "loss_scale": 4.0, "retries": 0, "seconds": 20.8, "max_memory_gb": 24.91, "pair": 1.034154251217842, "sign": 0.036760090151801705, "keep": 1.3786970674991608, "leak": 0.008728663691726979, "replacement": 1.74406498670578} +{"step": 20, "steps": 250, "lr": 4.998175245886032e-05, "grad_norm": 0.9031949639320374, "loss_scale": 4.0, "retries": 0, "seconds": 21.9, "max_memory_gb": 24.91, "pair": 1.0687974989414215, "sign": 0.03151711728423834, "keep": 0.0038333598313329276, "leak": 0.0008942735148593783, "replacement": 1.086696296930313} +{"step": 21, "steps": 250, "lr": 4.996756333769319e-05, "grad_norm": 1.7205910682678223, "loss_scale": 4.0, "retries": 0, "seconds": 23.0, "max_memory_gb": 24.91, "pair": 1.038803517818451, "sign": 0.04534841806162149, "keep": 0.6147950096492423, "leak": 0.006071922078490388, "replacement": 1.3703932166099548} +{"step": 22, "steps": 250, "lr": 4.994932456660157e-05, "grad_norm": 1.7196253538131714, "loss_scale": 4.0, "retries": 0, "seconds": 24.1, "max_memory_gb": 24.91, "pair": 1.0368816256523132, "sign": 0.0672085927799344, "keep": 0.7375379111617804, "leak": 0.0032519787055207416, "replacement": 1.440067857503891} +{"step": 23, "steps": 250, "lr": 4.992703943301973e-05, "grad_norm": 2.0640406608581543, "loss_scale": 4.0, "retries": 0, "seconds": 25.3, "max_memory_gb": 24.91, "pair": 1.027148813009262, "sign": 0.03629625914618373, "keep": 1.2526440611909493, "leak": 0.0027307907585054636, "replacement": 1.6723016798496246} +{"step": 24, "steps": 250, "lr": 4.9900711953715675e-05, "grad_norm": 2.1688501834869385, "loss_scale": 4.0, "retries": 0, "seconds": 26.3, "max_memory_gb": 24.91, "pair": 0.9735378921031952, "sign": 0.016917557921260595, "keep": 0.6271993592381477, "leak": 0.008676064317114651, "replacement": 1.2977654039859772} +{"step": 25, "steps": 250, "lr": 4.987034687406713e-05, "grad_norm": 2.0932400226593018, "loss_scale": 4.0, "retries": 0, "seconds": 27.4, "max_memory_gb": 24.91, "pair": 0.9956439137458801, "sign": 0.11262324918061495, "keep": 1.3639576763453078, "leak": 0.009263821542845108, "replacement": 1.7362503111362457} +{"step": 26, "steps": 250, "lr": 4.9835949667206215e-05, "grad_norm": 2.2887678146362305, "loss_scale": 4.0, "retries": 0, "seconds": 28.5, "max_memory_gb": 24.91, "pair": 0.9629136174917221, "sign": 0.06361226190347224, "keep": 0.6764476895332336, "leak": 0.015040741243865341, "replacement": 1.33670374751091} +{"step": 27, "steps": 250, "lr": 4.9797526533032945e-05, "grad_norm": 2.127100944519043, "loss_scale": 4.0, "retries": 0, "seconds": 29.5, "max_memory_gb": 24.91, "pair": 0.9679032266139984, "sign": 0.14422611729241908, "keep": 0.8377881149644963, "leak": 0.015461106318980455, "replacement": 1.4627755731344223} +{"step": 28, "steps": 250, "lr": 4.9755084397097716e-05, "grad_norm": 2.200202703475952, "loss_scale": 4.0, "retries": 0, "seconds": 30.6, "max_memory_gb": 24.91, "pair": 1.0048754215240479, "sign": 0.4199048785958439, "keep": 1.4303466156125069, "leak": 0.019662330771097913, "replacement": 1.9349167346954346} +{"step": 29, "steps": 250, "lr": 4.970863090935304e-05, "grad_norm": 2.0573318004608154, "loss_scale": 4.0, "retries": 0, "seconds": 31.6, "max_memory_gb": 24.91, "pair": 1.0216076076030731, "sign": 0.7323151044547558, "keep": 0.5860957559198141, "leak": 0.005672101583513722, "replacement": 1.6822310388088226} +{"step": 30, "steps": 250, "lr": 4.965817444277468e-05, "grad_norm": 2.2713782787323, "loss_scale": 4.0, "retries": 0, "seconds": 32.6, "max_memory_gb": 24.91, "pair": 0.9004113376140594, "sign": 0.011818741215392947, "keep": 0.6846532040217426, "leak": 0.020499725826084614, "replacement": 1.2537721991539001} +{"step": 31, "steps": 250, "lr": 4.9603724091852456e-05, "grad_norm": 3.529432535171509, "loss_scale": 4.0, "retries": 0, "seconds": 33.6, "max_memory_gb": 24.91, "pair": 0.9632610827684402, "sign": 0.21891301800496876, "keep": 1.3696066588163376, "leak": 0.026491371827432886, "replacement": 1.7641438245773315} +{"step": 32, "steps": 250, "lr": 4.9545289670951025e-05, "grad_norm": 3.1211459636688232, "loss_scale": 4.0, "retries": 0, "seconds": 34.7, "max_memory_gb": 24.91, "pair": 0.8854369521141052, "sign": 0.1204133415594697, "keep": 1.0019936636090279, "leak": 0.07520221918821335, "replacement": 1.4654409885406494} +{"step": 33, "steps": 250, "lr": 4.948288171254089e-05, "grad_norm": 1.7438175678253174, "loss_scale": 4.0, "retries": 0, "seconds": 35.8, "max_memory_gb": 24.91, "pair": 0.9792991578578949, "sign": 0.2037269426509738, "keep": 0.5346342120319605, "leak": 0.019816807907773182, "replacement": 1.353433907032013} +{"step": 34, "steps": 250, "lr": 4.941651146529998e-05, "grad_norm": 1.8584084510803223, "loss_scale": 4.0, "retries": 0, "seconds": 36.9, "max_memory_gb": 24.91, "pair": 0.9075607806444168, "sign": 0.0335038504563272, "keep": 0.3647868746847962, "leak": 0.05572750326246023, "replacement": 1.1206380128860474} +{"step": 35, "steps": 250, "lr": 4.9346190892086174e-05, "grad_norm": 2.1501827239990234, "loss_scale": 4.0, "retries": 0, "seconds": 38.0, "max_memory_gb": 24.91, "pair": 0.9112263023853302, "sign": 0.06441305577754974, "keep": 0.5002845347626135, "leak": 0.031040637753903866, "replacement": 1.2013352662324905} +{"step": 36, "steps": 250, "lr": 4.927193266778101e-05, "grad_norm": 1.9061715602874756, "loss_scale": 4.0, "retries": 0, "seconds": 39.1, "max_memory_gb": 24.91, "pair": 0.8837364464998245, "sign": 0.04627762152813375, "keep": 0.47674247436225414, "leak": 0.07598071033135056, "replacement": 1.1642416715621948} +{"step": 37, "steps": 250, "lr": 4.919375017700515e-05, "grad_norm": 1.7615516185760498, "loss_scale": 4.0, "retries": 0, "seconds": 40.2, "max_memory_gb": 24.91, "pair": 0.9151197075843811, "sign": 0.21506958827376366, "keep": 0.9008815992856398, "leak": 0.05712219490123971, "replacement": 1.4873758554458618} +{"step": 38, "steps": 250, "lr": 4.911165751170588e-05, "grad_norm": 2.0702669620513916, "loss_scale": 4.0, "retries": 0, "seconds": 41.3, "max_memory_gb": 24.91, "pair": 0.8231582939624786, "sign": 0.09208180801942945, "keep": 0.3433177024126053, "leak": 0.11076125130057335, "replacement": 1.0685483813285828} +{"step": 39, "steps": 250, "lr": 4.902566946861708e-05, "grad_norm": 1.4943896532058716, "loss_scale": 4.0, "retries": 0, "seconds": 42.3, "max_memory_gb": 24.91, "pair": 0.9385769069194794, "sign": 1.3502098609460518, "keep": 0.3742325920102303, "leak": 0.038869730415171944, "replacement": 1.8105155676603317} +{"step": 40, "steps": 250, "lr": 4.89358015465922e-05, "grad_norm": 1.884460687637329, "loss_scale": 4.0, "retries": 0, "seconds": 43.4, "max_memory_gb": 24.91, "pair": 0.8147080540657043, "sign": 0.021244398318231106, "keep": 0.28902948620088864, "leak": 0.10730892606079578, "replacement": 0.9966722130775452} +{"step": 41, "steps": 250, "lr": 4.884206994381078e-05, "grad_norm": 1.861406922340393, "loss_scale": 4.0, "retries": 0, "seconds": 44.5, "max_memory_gb": 24.91, "pair": 0.9391740560531616, "sign": 0.862089367583394, "keep": 0.9869738142006099, "leak": 0.02878342568874359, "replacement": 1.8709015101194382} +{"step": 42, "steps": 250, "lr": 4.874449155485866e-05, "grad_norm": 1.8960652351379395, "loss_scale": 4.0, "retries": 0, "seconds": 45.6, "max_memory_gb": 24.91, "pair": 0.7361322194337845, "sign": 0.056262134574353695, "keep": 0.13470058426901232, "leak": 0.11910193227231503, "replacement": 0.8613890558481216} +{"step": 43, "steps": 250, "lr": 4.864308396768294e-05, "grad_norm": 2.140575885772705, "loss_scale": 4.0, "retries": 0, "seconds": 46.6, "max_memory_gb": 24.91, "pair": 0.728646382689476, "sign": 0.15320531511679292, "keep": 0.38966699813317973, "leak": 0.410024621989578, "replacement": 1.102588728070259} +{"step": 44, "steps": 250, "lr": 4.853786546042184e-05, "grad_norm": 2.8155951499938965, "loss_scale": 4.0, "retries": 0, "seconds": 47.7, "max_memory_gb": 24.91, "pair": 0.7938971817493439, "sign": 0.1722112197894603, "keep": 1.0035638708795886, "leak": 0.12069942988455296, "replacement": 1.4119596034288406} +{"step": 45, "steps": 250, "lr": 4.842885499811012e-05, "grad_norm": 2.78945255279541, "loss_scale": 4.0, "retries": 0, "seconds": 48.8, "max_memory_gb": 24.91, "pair": 0.9258239567279816, "sign": 0.6827853759750724, "keep": 1.9325696126325056, "leak": 0.03271326544222575, "replacement": 2.241679832339287} +{"step": 46, "steps": 250, "lr": 4.831607222926078e-05, "grad_norm": 2.1906940937042236, "loss_scale": 4.0, "retries": 0, "seconds": 49.9, "max_memory_gb": 24.91, "pair": 0.772977277636528, "sign": 0.2694443682557903, "keep": 0.5847670249640942, "leak": 0.10058769630268216, "replacement": 1.2252298891544342} +{"step": 47, "steps": 250, "lr": 4.8199537482323545e-05, "grad_norm": 1.8550363779067993, "loss_scale": 4.0, "retries": 0, "seconds": 51.0, "max_memory_gb": 24.91, "pair": 0.7494458630681038, "sign": 0.10781487240456045, "keep": 0.1399290138942888, "leak": 0.1479154495173134, "replacement": 0.9102966487407684} +{"step": 48, "steps": 250, "lr": 4.807927176202073e-05, "grad_norm": 2.113006114959717, "loss_scale": 4.0, "retries": 0, "seconds": 52.1, "max_memory_gb": 24.91, "pair": 0.6056891828775406, "sign": 0.02908642269903794, "keep": 0.3788295137928799, "leak": 0.3252095356583595, "replacement": 0.890949547290802} +{"step": 49, "steps": 250, "lr": 4.795529674556129e-05, "grad_norm": 2.433511257171631, "loss_scale": 4.0, "retries": 0, "seconds": 53.2, "max_memory_gb": 24.91, "pair": 0.6061426177620888, "sign": 0.05482121976092458, "keep": 0.2475166850927053, "leak": 0.1772209107875824, "replacement": 0.8016167879104614} +{"step": 50, "steps": 250, "lr": 4.782763477873361e-05, "grad_norm": 1.9321521520614624, "loss_scale": 4.0, "retries": 0, "seconds": 54.2, "max_memory_gb": 24.91, "pair": 0.7025566250085831, "sign": 0.11669068038463593, "keep": 0.5151523369349889, "leak": 0.31896121241152287, "replacement": 1.0982184410095215} +{"step": 51, "steps": 250, "lr": 4.769630887187782e-05, "grad_norm": 2.8775508403778076, "loss_scale": 4.0, "retries": 0, "seconds": 55.2, "max_memory_gb": 24.91, "pair": 0.6460344195365906, "sign": 0.034270408912561834, "keep": 1.1264112496719463, "leak": 0.12568371527595446, "replacement": 1.257796198129654} +{"step": 52, "steps": 250, "lr": 4.756134269573832e-05, "grad_norm": 1.928645133972168, "loss_scale": 4.0, "retries": 0, "seconds": 56.3, "max_memory_gb": 24.91, "pair": 0.6637492924928665, "sign": 0.2721002446487546, "keep": 0.4724055406986736, "leak": 0.21040969155728817, "replacement": 1.0886046141386032} +{"step": 53, "steps": 250, "lr": 4.7422760577197226e-05, "grad_norm": 2.3556747436523438, "loss_scale": 4.0, "retries": 0, "seconds": 57.3, "max_memory_gb": 24.91, "pair": 0.5731634199619293, "sign": 0.09998900070786476, "keep": 0.4027402396313846, "leak": 0.4756558556109667, "replacement": 0.9434420168399811} +{"step": 54, "steps": 250, "lr": 4.728058749488963e-05, "grad_norm": 2.57547664642334, "loss_scale": 4.0, "retries": 0, "seconds": 58.4, "max_memory_gb": 24.91, "pair": 0.6269023194909096, "sign": 0.16912469873204827, "keep": 0.35750041739083827, "leak": 0.41954235220327973, "replacement": 0.9951004981994629} +{"step": 55, "steps": 250, "lr": 4.713484907470133e-05, "grad_norm": 1.2889399528503418, "loss_scale": 4.0, "retries": 0, "seconds": 59.4, "max_memory_gb": 24.91, "pair": 0.5369869694113731, "sign": 0.03741335088852793, "keep": 0.01811919348256197, "leak": 0.2997436113655567, "replacement": 0.6396891474723816} +{"step": 56, "steps": 250, "lr": 4.6985571585149876e-05, "grad_norm": 1.4326708316802979, "loss_scale": 4.0, "retries": 0, "seconds": 60.4, "max_memory_gb": 24.91, "pair": 0.4692816957831383, "sign": 0.01882627629674971, "keep": 0.11444142973050475, "leak": 0.3823040798306465, "replacement": 0.6314915716648102} +{"step": 57, "steps": 250, "lr": 4.6832781932649884e-05, "grad_norm": 1.4226309061050415, "loss_scale": 4.0, "retries": 0, "seconds": 61.5, "max_memory_gb": 24.91, "pair": 0.6088783219456673, "sign": 0.3788689957000315, "keep": 0.04120112683449406, "leak": 0.3841096750693396, "replacement": 0.914940819144249} +{"step": 58, "steps": 250, "lr": 4.667650765666325e-05, "grad_norm": 1.3140082359313965, "loss_scale": 4.0, "retries": 0, "seconds": 62.6, "max_memory_gb": 24.91, "pair": 0.4019218981266022, "sign": 0.01698891445994377, "keep": 0.10793496236146893, "leak": 0.49546221643686295, "replacement": 0.5882494077086449} +{"step": 59, "steps": 250, "lr": 4.651677692473538e-05, "grad_norm": 1.532514214515686, "loss_scale": 4.0, "retries": 0, "seconds": 63.6, "max_memory_gb": 24.91, "pair": 0.527865506708622, "sign": 0.044620986678637564, "keep": 0.030759991248487495, "leak": 0.28746331646107137, "replacement": 0.6374218240380287} +{"step": 60, "steps": 250, "lr": 4.635361852741808e-05, "grad_norm": 1.5194724798202515, "loss_scale": 4.0, "retries": 0, "seconds": 64.7, "max_memory_gb": 24.91, "pair": 0.4796089790761471, "sign": 0.0640085821505636, "keep": 0.10910004400648177, "leak": 0.7658028483911039, "replacement": 0.757614016532898} +{"step": 61, "steps": 250, "lr": 4.6187061873080264e-05, "grad_norm": 1.5239684581756592, "loss_scale": 4.0, "retries": 0, "seconds": 65.8, "max_memory_gb": 24.91, "pair": 0.504695039242506, "sign": 0.24969042581506073, "keep": 0.1104187211021781, "leak": 0.9446919541805983, "replacement": 0.9209225922822952} +{"step": 62, "steps": 250, "lr": 4.601713698260727e-05, "grad_norm": 1.0000324249267578, "loss_scale": 4.0, "retries": 0, "seconds": 66.9, "max_memory_gb": 24.91, "pair": 0.49943047016859055, "sign": 0.033917239401489496, "keep": 0.09513175894971937, "leak": 0.6745178150013089, "replacement": 0.7325844317674637} +{"step": 63, "steps": 250, "lr": 4.5843874483989746e-05, "grad_norm": 1.1779537200927734, "loss_scale": 4.0, "retries": 0, "seconds": 68.0, "max_memory_gb": 24.91, "pair": 0.4339828584343195, "sign": 0.14946638653054833, "keep": 0.10187394195236266, "leak": 0.1898023618850857, "replacement": 0.6071036159992218} +{"step": 64, "steps": 250, "lr": 4.566730560680316e-05, "grad_norm": 0.860977053642273, "loss_scale": 4.0, "retries": 0, "seconds": 69.1, "max_memory_gb": 24.91, "pair": 0.3329704962670803, "sign": 0.018753668060526252, "keep": 0.04903025319799781, "leak": 0.5910621248185635, "replacement": 0.5146280005574226} +{"step": 65, "steps": 250, "lr": 4.548746217657878e-05, "grad_norm": 3.2860939502716064, "loss_scale": 4.0, "retries": 0, "seconds": 70.2, "max_memory_gb": 24.91, "pair": 0.41611362993717194, "sign": 0.358366958796978, "keep": 0.3174682538956404, "leak": 2.0285084173083305, "replacement": 1.2611583769321442} +{"step": 66, "steps": 250, "lr": 4.5304376609067396e-05, "grad_norm": 1.528635859489441, "loss_scale": 4.0, "retries": 0, "seconds": 71.3, "max_memory_gb": 24.91, "pair": 0.4837736636400223, "sign": 0.13629374478477985, "keep": 0.35867646333639414, "leak": 0.24402592983096838, "replacement": 0.7922652587294579} +{"step": 67, "steps": 250, "lr": 4.5118081904396504e-05, "grad_norm": 0.919424831867218, "loss_scale": 4.0, "retries": 0, "seconds": 72.4, "max_memory_gb": 24.91, "pair": 0.24085511453449726, "sign": 0.021444848272949457, "keep": 0.1414259089215193, "leak": 1.071868747472763, "replacement": 0.5902576744556427} +{"step": 68, "steps": 250, "lr": 4.4928611641122225e-05, "grad_norm": 1.3649795055389404, "loss_scale": 4.0, "retries": 0, "seconds": 73.5, "max_memory_gb": 24.91, "pair": 0.5065605565905571, "sign": 0.054017911199480295, "keep": 0.030905174504823663, "leak": 0.4684041514992714, "replacement": 0.6661231368780136} +{"step": 69, "steps": 250, "lr": 4.473599997017701e-05, "grad_norm": 2.295872211456299, "loss_scale": 4.0, "retries": 0, "seconds": 74.6, "max_memory_gb": 24.91, "pair": 0.6220027878880501, "sign": 0.8636098445858806, "keep": 0.649433043319732, "leak": 0.8501729383133352, "replacement": 1.5910674929618835} +{"step": 70, "steps": 250, "lr": 4.4540281608714053e-05, "grad_norm": 2.2771055698394775, "loss_scale": 4.0, "retries": 0, "seconds": 75.7, "max_memory_gb": 24.91, "pair": 0.42841141670942307, "sign": 0.03860668194829486, "keep": 0.04521861298417207, "leak": 1.866131566464901, "replacement": 0.9368569552898407} +{"step": 71, "steps": 250, "lr": 4.434149183384977e-05, "grad_norm": 1.298382043838501, "loss_scale": 4.0, "retries": 0, "seconds": 76.8, "max_memory_gb": 24.91, "pair": 0.199148740619421, "sign": 0.00781807373277843, "keep": 0.20065851504841703, "leak": 0.877310000360012, "replacement": 0.5227145403623581} +{"step": 72, "steps": 250, "lr": 4.413966647630534e-05, "grad_norm": 1.6637517213821411, "loss_scale": 4.0, "retries": 0, "seconds": 77.9, "max_memory_gb": 24.91, "pair": 0.28785306960344315, "sign": 0.19592745508998632, "keep": 0.1429770067334175, "leak": 1.3263638615608215, "replacement": 0.7888962626457214} +{"step": 73, "steps": 250, "lr": 4.3934841913948323e-05, "grad_norm": 1.6548287868499756, "loss_scale": 4.0, "retries": 0, "seconds": 78.9, "max_memory_gb": 24.91, "pair": 0.49113987386226654, "sign": 0.12668264005333185, "keep": 0.5229754740921635, "leak": 0.4790707267820835, "replacement": 0.9357366114854813} +{"step": 74, "steps": 250, "lr": 4.372705506523583e-05, "grad_norm": 1.998927116394043, "loss_scale": 4.0, "retries": 0, "seconds": 80.0, "max_memory_gb": 24.91, "pair": 0.36229724809527397, "sign": 0.2446999920066446, "keep": 0.23086390643766208, "leak": 1.5239223489770666, "replacement": 0.981059804558754} +{"step": 75, "steps": 250, "lr": 4.351634338256017e-05, "grad_norm": 1.6441998481750488, "loss_scale": 4.0, "retries": 0, "seconds": 81.1, "max_memory_gb": 24.91, "pair": 0.7146413065493107, "sign": 0.5235644610365853, "keep": 0.8656133045442402, "leak": 0.2414648192934692, "replacement": 1.4695963859558105} +{"step": 76, "steps": 250, "lr": 4.330274484549825e-05, "grad_norm": 1.0517312288284302, "loss_scale": 4.0, "retries": 0, "seconds": 82.2, "max_memory_gb": 24.91, "pair": 0.3612573556602001, "sign": 0.10788466990925372, "keep": 0.03121067295433022, "leak": 0.76962910592556, "replacement": 0.6232122778892517} +{"step": 77, "steps": 250, "lr": 4.308629795396599e-05, "grad_norm": 2.7209420204162598, "loss_scale": 4.0, "retries": 0, "seconds": 83.3, "max_memory_gb": 24.91, "pair": 0.18452558107674122, "sign": 0.008171118330210447, "keep": 0.010215282119133917, "leak": 2.032524287700653, "replacement": 0.7018498554825783} +{"step": 78, "steps": 250, "lr": 4.286704172127894e-05, "grad_norm": 2.094033718109131, "loss_scale": 4.0, "retries": 0, "seconds": 84.3, "max_memory_gb": 24.91, "pair": 0.4612255357205868, "sign": 0.6581724279094487, "keep": 0.002558058753493242, "leak": 0.4953908510506153, "replacement": 0.9154385477304459} +{"step": 79, "steps": 250, "lr": 4.264501566712033e-05, "grad_norm": 1.4861345291137695, "loss_scale": 4.0, "retries": 0, "seconds": 85.4, "max_memory_gb": 24.91, "pair": 0.46879126876592636, "sign": 0.031184740364551544, "keep": 0.1327907797131047, "leak": 0.5021803677082062, "replacement": 0.6763241291046143} +{"step": 80, "steps": 250, "lr": 4.2420259810417894e-05, "grad_norm": 1.0156691074371338, "loss_scale": 4.0, "retries": 0, "seconds": 86.4, "max_memory_gb": 24.91, "pair": 0.5084546804428101, "sign": 0.10036893608048558, "keep": 0.06729683261710306, "leak": 0.6280559562146664, "replacement": 0.749301552772522} +{"step": 81, "steps": 250, "lr": 4.219281466213066e-05, "grad_norm": 1.0555386543273926, "loss_scale": 4.0, "retries": 0, "seconds": 87.5, "max_memory_gb": 24.91, "pair": 0.2903808318078518, "sign": 0.06096653948770836, "keep": 0.16856108838692307, "leak": 1.2339978758245707, "replacement": 0.7136441022157669} +{"step": 82, "steps": 250, "lr": 4.1962721217947135e-05, "grad_norm": 3.1040585041046143, "loss_scale": 4.0, "retries": 0, "seconds": 88.6, "max_memory_gb": 24.91, "pair": 0.2593694142997265, "sign": 1.0351668803486973, "keep": 0.11068129900377244, "leak": 1.988919958472252, "replacement": 1.3295235186815262} +{"step": 83, "steps": 250, "lr": 4.173002095089599e-05, "grad_norm": 1.183480978012085, "loss_scale": 4.0, "retries": 0, "seconds": 89.7, "max_memory_gb": 24.91, "pair": 0.29837838746607304, "sign": 0.03216676786541939, "keep": 0.3582564471871592, "leak": 1.085783015936613, "replacement": 0.7650357559323311} +{"step": 84, "steps": 250, "lr": 4.149475580387084e-05, "grad_norm": 1.478989601135254, "loss_scale": 4.0, "retries": 0, "seconds": 90.8, "max_memory_gb": 24.91, "pair": 0.25139553658664227, "sign": 0.029099896550178528, "keep": 0.05490097764345592, "leak": 1.1824527978897095, "replacement": 0.589009165763855} +{"step": 85, "steps": 250, "lr": 4.1256968182070275e-05, "grad_norm": 3.204007863998413, "loss_scale": 4.0, "retries": 0, "seconds": 91.9, "max_memory_gb": 24.91, "pair": 0.29698001593351364, "sign": 0.051450155093334615, "keep": 0.11955138284892541, "leak": 0.7830079197883606, "replacement": 0.5782327800989151} +{"step": 86, "steps": 250, "lr": 4.101670094535453e-05, "grad_norm": 2.198681592941284, "loss_scale": 4.0, "retries": 0, "seconds": 93.0, "max_memory_gb": 24.91, "pair": 0.31244171783328056, "sign": 0.2843579100444913, "keep": 0.3249598501715809, "leak": 0.3660430945456028, "replacement": 0.7086113691329956} +{"step": 87, "steps": 250, "lr": 4.077399740052027e-05, "grad_norm": 1.8662866353988647, "loss_scale": 4.0, "retries": 0, "seconds": 94.0, "max_memory_gb": 24.91, "pair": 0.4569992944598198, "sign": 0.09859325061552227, "keep": 0.019736574176931754, "leak": 0.5683194734156132, "replacement": 0.6582440808415413} +{"step": 88, "steps": 250, "lr": 4.0528901293494775e-05, "grad_norm": 2.1579151153564453, "loss_scale": 4.0, "retries": 0, "seconds": 95.1, "max_memory_gb": 24.91, "pair": 0.42954133450984955, "sign": 0.2302305530756712, "keep": 0.03827836864979872, "leak": 0.653175838291645, "replacement": 0.7270897626876831} +{"step": 89, "steps": 250, "lr": 4.0281456801451004e-05, "grad_norm": 1.3979181051254272, "loss_scale": 4.0, "retries": 0, "seconds": 96.2, "max_memory_gb": 24.91, "pair": 0.5724548101425171, "sign": 0.18656899756751955, "keep": 0.016759189938568397, "leak": 0.18955759145319462, "replacement": 0.7215082794427872} +{"step": 90, "steps": 250, "lr": 4.0031708524844884e-05, "grad_norm": 2.065509557723999, "loss_scale": 4.0, "retries": 0, "seconds": 97.2, "max_memory_gb": 24.91, "pair": 0.4241482503712177, "sign": 0.1060818734113127, "keep": 0.06110549397999421, "leak": 1.17609441280365, "replacement": 0.8017655164003372} +{"step": 91, "steps": 250, "lr": 3.977970147937635e-05, "grad_norm": 4.381953239440918, "loss_scale": 4.0, "retries": 0, "seconds": 98.4, "max_memory_gb": 24.91, "pair": 0.576403483748436, "sign": 0.24230129946954548, "keep": 0.2939742944945465, "leak": 0.5023393221199512, "replacement": 0.9701261520385742} +{"step": 92, "steps": 250, "lr": 3.952548108787553e-05, "grad_norm": 4.595095157623291, "loss_scale": 4.0, "retries": 0, "seconds": 99.5, "max_memory_gb": 24.91, "pair": 0.4428306110203266, "sign": 0.7479272527853027, "keep": 1.2177159899756589, "leak": 0.6894774107495323, "replacement": 1.5980216711759567} +{"step": 93, "steps": 250, "lr": 3.9269093172115496e-05, "grad_norm": 1.6285349130630493, "loss_scale": 4.0, "retries": 0, "seconds": 100.6, "max_memory_gb": 24.91, "pair": 0.3534163273870945, "sign": 0.0387694031232968, "keep": 0.18897054356057197, "leak": 0.5697673810645938, "replacement": 0.609728142619133} +{"step": 94, "steps": 250, "lr": 3.901058394455323e-05, "grad_norm": 3.102586030960083, "loss_scale": 4.0, "retries": 0, "seconds": 101.7, "max_memory_gb": 24.91, "pair": 0.27160472236573696, "sign": 0.03475168254226446, "keep": 0.7331659314222634, "leak": 1.3338413685560226, "replacement": 0.9890238493680954} +{"step": 95, "steps": 250, "lr": 3.875e-05, "grad_norm": 2.6145823001861572, "loss_scale": 4.0, "retries": 0, "seconds": 102.7, "max_memory_gb": 24.91, "pair": 0.4677175208926201, "sign": 0.30721723777242005, "keep": 1.4163293838500977, "leak": 0.6765773685765453, "replacement": 1.4986352026462555} +{"step": 96, "steps": 250, "lr": 3.8487388307223005e-05, "grad_norm": 1.8379371166229248, "loss_scale": 4.0, "retries": 0, "seconds": 103.8, "max_memory_gb": 24.91, "pair": 0.39782610535621643, "sign": 0.11552224715705961, "keep": 0.9671072978526354, "leak": 0.5707640768960118, "replacement": 1.0818318873643875} +{"step": 97, "steps": 250, "lr": 3.822279620047943e-05, "grad_norm": 1.2721278667449951, "loss_scale": 4.0, "retries": 0, "seconds": 104.9, "max_memory_gb": 24.91, "pair": 0.302354346960783, "sign": 0.017706738435663283, "keep": 0.03737335402001918, "leak": 0.9607368595898151, "replacement": 0.5700786113739014} +{"step": 98, "steps": 250, "lr": 3.795627137098479e-05, "grad_norm": 1.0255733728408813, "loss_scale": 4.0, "retries": 0, "seconds": 106.0, "max_memory_gb": 24.91, "pair": 0.43252531811594963, "sign": 0.15593825647374615, "keep": 0.030070498430859516, "leak": 0.7909931689500809, "replacement": 0.7232779860496521} +{"step": 99, "steps": 250, "lr": 3.7687861858316844e-05, "grad_norm": 2.613483428955078, "loss_scale": 4.0, "retries": 0, "seconds": 107.1, "max_memory_gb": 24.91, "pair": 0.6475401818752289, "sign": 0.6941652958630584, "keep": 0.04238144040073166, "leak": 0.24268474825657904, "replacement": 1.0764847919344902} +{"step": 100, "steps": 250, "lr": 3.741761604175668e-05, "grad_norm": 3.5690970420837402, "loss_scale": 4.0, "retries": 0, "seconds": 108.1, "max_memory_gb": 24.91, "pair": 0.542979009449482, "sign": 0.9291125233285129, "keep": 0.37865057843737304, "leak": 0.940795011818409, "replacement": 1.432059332728386} +{"step": 101, "steps": 250, "lr": 3.714558263156872e-05, "grad_norm": 3.6809451580047607, "loss_scale": 4.0, "retries": 0, "seconds": 109.2, "max_memory_gb": 24.91, "pair": 0.1731370585039258, "sign": 0.010188606946030632, "keep": 0.03546982655825559, "leak": 3.00321476906538, "replacement": 0.9467699304223061} +{"step": 102, "steps": 250, "lr": 3.6871810660220904e-05, "grad_norm": 1.1764495372772217, "loss_scale": 4.0, "retries": 0, "seconds": 110.3, "max_memory_gb": 24.91, "pair": 0.2798560597002506, "sign": 0.024191282922402024, "keep": 0.0656227616709657, "leak": 0.3883320251479745, "replacement": 0.42184609547257423} +{"step": 103, "steps": 250, "lr": 3.659634947354686e-05, "grad_norm": 0.8349543213844299, "loss_scale": 4.0, "retries": 0, "seconds": 111.4, "max_memory_gb": 24.91, "pair": 0.29803847521543503, "sign": 0.03369372908491641, "keep": 0.02818828178988042, "leak": 0.7264191210269928, "replacement": 0.5105842500925064} +{"step": 104, "steps": 250, "lr": 3.631924872185169e-05, "grad_norm": 2.9172067642211914, "loss_scale": 4.0, "retries": 0, "seconds": 112.4, "max_memory_gb": 24.91, "pair": 0.4421111047267914, "sign": 0.39588362781796604, "keep": 0.0068165764496370684, "leak": 0.7245193719863892, "replacement": 0.8245910257101059} +{"step": 105, "steps": 250, "lr": 3.6040558350962674e-05, "grad_norm": 1.1476163864135742, "loss_scale": 4.0, "retries": 0, "seconds": 113.4, "max_memory_gb": 24.91, "pair": 0.25864996016025543, "sign": 0.022716065286658704, "keep": 0.10362828988581896, "leak": 1.3286185413599014, "replacement": 0.653976783156395} +{"step": 106, "steps": 250, "lr": 3.576032859322687e-05, "grad_norm": 1.2531541585922241, "loss_scale": 4.0, "retries": 0, "seconds": 114.4, "max_memory_gb": 24.91, "pair": 0.33804773911833763, "sign": 0.03968671336770058, "keep": 0.06845818219153443, "leak": 0.7167413919232786, "replacement": 0.5713055282831192} +{"step": 107, "steps": 250, "lr": 3.5478609958457056e-05, "grad_norm": 2.9077298641204834, "loss_scale": 4.0, "retries": 0, "seconds": 115.5, "max_memory_gb": 24.91, "pair": 0.37590742111206055, "sign": 0.3811760365497321, "keep": 2.2467839462606207, "leak": 0.4252070225775242, "replacement": 1.7961892634630203} +{"step": 108, "steps": 250, "lr": 3.519545322482755e-05, "grad_norm": 1.6107070446014404, "loss_scale": 4.0, "retries": 0, "seconds": 116.6, "max_memory_gb": 24.91, "pair": 0.4323650151491165, "sign": 0.26637348672375083, "keep": 0.038253881124092004, "leak": 0.5542995585128665, "replacement": 0.7232535630464554} +{"step": 109, "steps": 250, "lr": 3.491090942972179e-05, "grad_norm": 1.858938455581665, "loss_scale": 4.0, "retries": 0, "seconds": 117.7, "max_memory_gb": 24.91, "pair": 0.4138353392481804, "sign": 0.0919773387722671, "keep": 0.08554625511169434, "leak": 0.5896672746166587, "replacement": 0.6500139534473419} +{"step": 110, "steps": 250, "lr": 3.4625029860533135e-05, "grad_norm": 4.200831890106201, "loss_scale": 4.0, "retries": 0, "seconds": 118.8, "max_memory_gb": 24.91, "pair": 0.3625268414616585, "sign": 0.21520765614695847, "keep": 0.47288758493959904, "leak": 0.7359690815210342, "replacement": 0.8905667439103127} +{"step": 111, "steps": 250, "lr": 3.433786604542056e-05, "grad_norm": 3.321861982345581, "loss_scale": 4.0, "retries": 0, "seconds": 119.9, "max_memory_gb": 24.91, "pair": 0.4086840348318219, "sign": 1.5195088152540848, "keep": 0.0030406703992866824, "leak": 1.8026190474629402, "replacement": 1.6206136047840118} +{"step": 112, "steps": 250, "lr": 3.404946974402104e-05, "grad_norm": 4.054687976837158, "loss_scale": 4.0, "retries": 0, "seconds": 121.0, "max_memory_gb": 24.91, "pair": 0.2070870567113161, "sign": 0.010054697515442967, "keep": 0.005094756510516163, "leak": 2.1414314284920692, "replacement": 0.7500196397304535} +{"step": 113, "steps": 250, "lr": 3.375989293812019e-05, "grad_norm": 1.0585137605667114, "loss_scale": 4.0, "retries": 0, "seconds": 122.1, "max_memory_gb": 24.91, "pair": 0.34578029811382294, "sign": 0.04156517377123237, "keep": 0.01277232525171712, "leak": 0.33465573098510504, "replacement": 0.4566129967570305} +{"step": 114, "steps": 250, "lr": 3.3469187822282766e-05, "grad_norm": 1.172142505645752, "loss_scale": 4.0, "retries": 0, "seconds": 123.1, "max_memory_gb": 24.91, "pair": 0.2706954590976238, "sign": 0.05027043307200074, "keep": 0.03894834592938423, "leak": 1.2099412754178047, "replacement": 0.6177901774644852} +{"step": 115, "steps": 250, "lr": 3.317740679444499e-05, "grad_norm": 1.1659260988235474, "loss_scale": 4.0, "retries": 0, "seconds": 124.2, "max_memory_gb": 24.91, "pair": 0.24824467860162258, "sign": 0.012026338605210185, "keep": 0.02842355681650588, "leak": 0.8268828559666872, "replacement": 0.47519034892320633} +{"step": 116, "steps": 250, "lr": 3.2884602446470056e-05, "grad_norm": 4.187062740325928, "loss_scale": 4.0, "retries": 0, "seconds": 125.3, "max_memory_gb": 24.91, "pair": 0.44520123675465584, "sign": 0.7278933066409081, "keep": 0.8925373759229842, "leak": 0.6573405861854553, "replacement": 1.4197516217827797} +{"step": 117, "steps": 250, "lr": 3.259082755466876e-05, "grad_norm": 6.019753932952881, "loss_scale": 4.0, "retries": 0, "seconds": 126.4, "max_memory_gb": 24.93, "pair": 0.30928342789411545, "sign": 0.38798998261336237, "keep": 1.0958164720836976, "leak": 1.5612314194440842, "replacement": 1.4414945244789124} +{"step": 118, "steps": 250, "lr": 3.229613507028686e-05, "grad_norm": 1.532555341720581, "loss_scale": 4.0, "retries": 0, "seconds": 127.5, "max_memory_gb": 24.93, "pair": 0.47221649810671806, "sign": 0.35717565880622715, "keep": 0.48929161829255463, "leak": 0.7214965559542179, "replacement": 1.0758242756128311} +{"step": 119, "steps": 250, "lr": 3.2000578109961004e-05, "grad_norm": 1.544098138809204, "loss_scale": 4.0, "retries": 0, "seconds": 128.6, "max_memory_gb": 24.93, "pair": 0.44692740589380264, "sign": 0.2955343248322606, "keep": 0.017713204107622005, "leak": 0.426730003207922, "replacement": 0.7102336883544922} +{"step": 120, "steps": 250, "lr": 3.1704209946144616e-05, "grad_norm": 3.6731114387512207, "loss_scale": 4.0, "retries": 0, "seconds": 129.7, "max_memory_gb": 24.93, "pair": 0.47094255685806274, "sign": 0.8820948172360659, "keep": 1.7356908440124243, "leak": 0.47101191990259395, "replacement": 1.8975883275270462} +{"step": 121, "steps": 250, "lr": 3.140708399750594e-05, "grad_norm": 1.8598378896713257, "loss_scale": 4.0, "retries": 0, "seconds": 130.8, "max_memory_gb": 24.93, "pair": 0.6006423756480217, "sign": 0.4185390619095415, "keep": 0.10404306649047612, "leak": 0.31466369703412056, "replacement": 0.940599337220192} +{"step": 122, "steps": 250, "lr": 3.1109253819299606e-05, "grad_norm": 1.21583092212677, "loss_scale": 4.0, "retries": 0, "seconds": 131.9, "max_memory_gb": 24.93, "pair": 0.41824057325720787, "sign": 0.10898036090657115, "keep": 0.04729924815569575, "leak": 0.6686311303637922, "replacement": 0.6635381430387497} +{"step": 123, "steps": 250, "lr": 3.081077309371361e-05, "grad_norm": 3.7164225578308105, "loss_scale": 4.0, "retries": 0, "seconds": 133.0, "max_memory_gb": 24.93, "pair": 0.43956473004072905, "sign": 1.1730663797352463, "keep": 0.3180158889153972, "leak": 0.9120279811322689, "replacement": 1.4131128564476967} +{"step": 124, "steps": 250, "lr": 3.051169562019339e-05, "grad_norm": 1.905483603477478, "loss_scale": 4.0, "retries": 0, "seconds": 134.0, "max_memory_gb": 24.93, "pair": 0.6137754581868649, "sign": 0.5760065759532154, "keep": 1.7199488950136583, "leak": 0.37657239101827145, "replacement": 1.8558962941169739} +{"step": 125, "steps": 250, "lr": 3.021207530574477e-05, "grad_norm": 1.9457882642745972, "loss_scale": 4.0, "retries": 0, "seconds": 135.1, "max_memory_gb": 24.93, "pair": 0.4287236388772726, "sign": 0.25548437098041177, "keep": 0.3269997055940621, "leak": 1.0851105019100942, "replacement": 0.9912433177232742} +{"step": 126, "steps": 250, "lr": 2.991196615521758e-05, "grad_norm": 3.3027384281158447, "loss_scale": 4.0, "retries": 0, "seconds": 136.2, "max_memory_gb": 24.93, "pair": 0.6699004136025906, "sign": 0.5140186174539849, "keep": 1.0309038301929832, "leak": 1.0580214275760227, "replacement": 1.7068670094013214} +{"step": 127, "steps": 250, "lr": 2.9611422261571536e-05, "grad_norm": 1.0694621801376343, "loss_scale": 4.0, "retries": 0, "seconds": 137.2, "max_memory_gb": 24.93, "pair": 0.2785356044769287, "sign": 0.010739287245087326, "keep": 0.024149430349325485, "leak": 0.7059322223067284, "replacement": 0.47246304154396057} +{"step": 128, "steps": 250, "lr": 2.931049779612634e-05, "grad_norm": 3.8562347888946533, "loss_scale": 4.0, "retries": 0, "seconds": 138.3, "max_memory_gb": 24.93, "pair": 0.528052244335413, "sign": 0.9396564872004092, "keep": 1.265245403687004, "leak": 0.4326715711504221, "replacement": 1.7386710867285728} +{"step": 129, "steps": 250, "lr": 2.9009246998797557e-05, "grad_norm": 1.3259423971176147, "loss_scale": 4.0, "retries": 0, "seconds": 139.4, "max_memory_gb": 24.93, "pair": 0.40960243344306946, "sign": 0.09500663005746901, "keep": 0.1646482888609171, "leak": 0.4405415372457355, "replacement": 0.6495652869343758} +{"step": 130, "steps": 250, "lr": 2.8707724168320238e-05, "grad_norm": 1.9972788095474243, "loss_scale": 4.0, "retries": 0, "seconds": 140.5, "max_memory_gb": 24.93, "pair": 0.5518851242959499, "sign": 1.7759046335704625, "keep": 0.07064478460233659, "leak": 0.2774111714679748, "replacement": 1.5445126444101334} +{"step": 131, "steps": 250, "lr": 2.840598365246184e-05, "grad_norm": 5.230543613433838, "loss_scale": 4.0, "retries": 0, "seconds": 141.5, "max_memory_gb": 24.93, "pair": 0.7281762957572937, "sign": 0.341538829728961, "keep": 0.3835207244119374, "leak": 0.5325461020693183, "replacement": 1.2238426059484482} +{"step": 132, "steps": 250, "lr": 2.8104079838226383e-05, "grad_norm": 1.2322065830230713, "loss_scale": 4.0, "retries": 0, "seconds": 142.6, "max_memory_gb": 24.93, "pair": 0.2905115634202957, "sign": 0.030296982440631837, "keep": 0.4390941962665238, "leak": 0.7045866325497627, "replacement": 0.7013538181781769} +{"step": 133, "steps": 250, "lr": 2.7802067142051446e-05, "grad_norm": 1.6812834739685059, "loss_scale": 4.0, "retries": 0, "seconds": 143.7, "max_memory_gb": 24.93, "pair": 0.5613155215978622, "sign": 0.49056821479462087, "keep": 0.8230503501254134, "leak": 0.2715425053611398, "replacement": 1.2860103994607925} +{"step": 134, "steps": 250, "lr": 2.7500000000000004e-05, "grad_norm": 2.724658250808716, "loss_scale": 4.0, "retries": 0, "seconds": 144.8, "max_memory_gb": 24.93, "pair": 0.37417713552713394, "sign": 0.3060669509577565, "keep": 0.5371280969261534, "leak": 0.17757286317646503, "replacement": 0.8401678279042244} +{"step": 135, "steps": 250, "lr": 2.7197932857948566e-05, "grad_norm": 1.26642906665802, "loss_scale": 4.0, "retries": 0, "seconds": 145.9, "max_memory_gb": 24.93, "pair": 0.26085035502910614, "sign": 0.03837483818642795, "keep": 0.1579753588885069, "leak": 0.7956230118870735, "replacement": 0.5579312145709991} +{"step": 136, "steps": 250, "lr": 2.689592016177363e-05, "grad_norm": 4.883063316345215, "loss_scale": 4.0, "retries": 0, "seconds": 146.9, "max_memory_gb": 24.93, "pair": 0.25460266694426537, "sign": 0.22575442446395755, "keep": 0.37516088658594526, "leak": 1.6637781858444214, "replacement": 0.971004918217659} +{"step": 137, "steps": 250, "lr": 2.659401634753817e-05, "grad_norm": 3.0485987663269043, "loss_scale": 4.0, "retries": 0, "seconds": 148.0, "max_memory_gb": 24.93, "pair": 0.2361427452415228, "sign": 0.019267166790086776, "keep": 0.03660562734876294, "leak": 1.8488797275349498, "replacement": 0.7262990698218346} +{"step": 138, "steps": 250, "lr": 2.6292275831679764e-05, "grad_norm": 2.8134753704071045, "loss_scale": 4.0, "retries": 0, "seconds": 149.2, "max_memory_gb": 24.93, "pair": 0.37436958961188793, "sign": 0.036338417092338204, "keep": 0.3120726686902344, "leak": 0.46869363193400204, "replacement": 0.6657485365867615} +{"step": 139, "steps": 250, "lr": 2.5990753001202445e-05, "grad_norm": 2.8296473026275635, "loss_scale": 4.0, "retries": 0, "seconds": 150.3, "max_memory_gb": 24.93, "pair": 0.35645992308855057, "sign": 0.7758504537050612, "keep": 0.027349790470907465, "leak": 1.3997674733400345, "replacement": 1.1080019026994705} +{"step": 140, "steps": 250, "lr": 2.5689502203873672e-05, "grad_norm": 1.3065638542175293, "loss_scale": 4.0, "retries": 0, "seconds": 151.4, "max_memory_gb": 24.93, "pair": 0.20971788465976715, "sign": 0.03199217893416062, "keep": 0.018716996773946448, "leak": 1.1309322640299797, "replacement": 0.5178055465221405} +{"step": 141, "steps": 250, "lr": 2.538857773842847e-05, "grad_norm": 1.0335960388183594, "loss_scale": 4.0, "retries": 0, "seconds": 152.6, "max_memory_gb": 24.93, "pair": 0.33864186704158783, "sign": 0.2296614813967608, "keep": 0.01346135109270108, "leak": 0.8704527732916176, "replacement": 0.6778164803981781} +{"step": 142, "steps": 250, "lr": 2.5088033844782426e-05, "grad_norm": 2.8041460514068604, "loss_scale": 4.0, "retries": 0, "seconds": 153.6, "max_memory_gb": 24.93, "pair": 0.2906288616359234, "sign": 0.43369505659211427, "keep": 0.36443934338967665, "leak": 0.8499573171138763, "replacement": 0.902185395359993} +{"step": 143, "steps": 250, "lr": 2.4787924694255237e-05, "grad_norm": 4.285581111907959, "loss_scale": 4.0, "retries": 0, "seconds": 154.8, "max_memory_gb": 24.93, "pair": 0.4347127266228199, "sign": 0.14871118683367968, "keep": 1.4863541302574959, "leak": 0.21535199508070946, "replacement": 1.3060833737254143} +{"step": 144, "steps": 250, "lr": 2.4488304379806627e-05, "grad_norm": 3.385026454925537, "loss_scale": 4.0, "retries": 0, "seconds": 155.9, "max_memory_gb": 24.93, "pair": 0.4185349605977535, "sign": 0.6860082827042788, "keep": 0.10067321872338653, "leak": 0.6282849349081516, "replacement": 0.9689469635486603} +{"step": 145, "steps": 250, "lr": 2.4189226906286388e-05, "grad_norm": 0.9929678440093994, "loss_scale": 4.0, "retries": 0, "seconds": 157.0, "max_memory_gb": 24.93, "pair": 0.32683079689741135, "sign": 0.076187803177163, "keep": 0.02257899972209998, "leak": 0.7278052717447281, "replacement": 0.55816550552845} +{"step": 146, "steps": 250, "lr": 2.3890746180700395e-05, "grad_norm": 3.6484501361846924, "loss_scale": 4.0, "retries": 0, "seconds": 158.1, "max_memory_gb": 24.93, "pair": 0.45163427386432886, "sign": 0.17421144200488925, "keep": 0.021891324780881405, "leak": 1.554620198905468, "replacement": 0.9383406937122345} +{"step": 147, "steps": 250, "lr": 2.359291600249407e-05, "grad_norm": 2.310969829559326, "loss_scale": 4.0, "retries": 0, "seconds": 159.2, "max_memory_gb": 24.93, "pair": 0.6591605395078659, "sign": 0.8994787838310003, "keep": 1.1398554370607599, "leak": 0.5147867724299431, "replacement": 1.8075242191553116} +{"step": 148, "steps": 250, "lr": 2.3295790053855392e-05, "grad_norm": 4.07835054397583, "loss_scale": 4.0, "retries": 0, "seconds": 160.3, "max_memory_gb": 24.93, "pair": 0.35081420093774796, "sign": 0.1657156510045752, "keep": 0.005783470040029215, "leak": 0.8096556700766087, "replacement": 0.638977661728859} +{"step": 149, "steps": 250, "lr": 2.299942189003901e-05, "grad_norm": 1.3288289308547974, "loss_scale": 4.0, "retries": 0, "seconds": 161.4, "max_memory_gb": 24.93, "pair": 0.2891420852392912, "sign": 0.08687075390480459, "keep": 0.049861780455103144, "leak": 0.9175247028470039, "replacement": 0.5868895277380943} +{"step": 150, "steps": 250, "lr": 2.2703864929713144e-05, "grad_norm": 1.193445086479187, "loss_scale": 4.0, "retries": 0, "seconds": 162.4, "max_memory_gb": 24.93, "pair": 0.33917665481567383, "sign": 0.13270351674873382, "keep": 0.17403953755274415, "leak": 0.5453286473639309, "replacement": 0.6288803443312645} +{"step": 151, "steps": 250, "lr": 2.2409172445331257e-05, "grad_norm": 1.1364271640777588, "loss_scale": 4.0, "retries": 0, "seconds": 163.6, "max_memory_gb": 24.93, "pair": 0.34255581721663475, "sign": 0.02782122523058206, "keep": 0.017507838494566386, "leak": 0.6540457159280777, "replacement": 0.5287317782640457} +{"step": 152, "steps": 250, "lr": 2.211539755352996e-05, "grad_norm": 1.0524426698684692, "loss_scale": 4.0, "retries": 0, "seconds": 164.7, "max_memory_gb": 24.93, "pair": 0.22804156690835953, "sign": 0.008357671438716352, "keep": 0.04537333872212912, "leak": 0.9813272580504417, "replacement": 0.5002388656139374} +{"step": 153, "steps": 250, "lr": 2.1822593205555016e-05, "grad_norm": 2.34171199798584, "loss_scale": 4.0, "retries": 0, "seconds": 165.8, "max_memory_gb": 24.93, "pair": 0.24884680286049843, "sign": 0.011378215276636183, "keep": 0.020689811548891157, "leak": 1.236374944448471, "replacement": 0.5739745423197746} +{"step": 154, "steps": 250, "lr": 2.1530812177717236e-05, "grad_norm": 1.9993990659713745, "loss_scale": 4.0, "retries": 0, "seconds": 166.8, "max_memory_gb": 24.93, "pair": 0.13861635141074657, "sign": 0.008888072101399302, "keep": 0.3424466662108898, "leak": 1.7980482876300812, "replacement": 0.763795793056488} +{"step": 155, "steps": 250, "lr": 2.1240107061879815e-05, "grad_norm": 1.184498906135559, "loss_scale": 4.0, "retries": 0, "seconds": 168.0, "max_memory_gb": 24.93, "pair": 0.21212404407560825, "sign": 0.00961054116487503, "keep": 0.45747234754401234, "leak": 0.7844105139374733, "replacement": 0.6417680978775024} +{"step": 156, "steps": 250, "lr": 2.095053025597896e-05, "grad_norm": 1.985670804977417, "loss_scale": 4.0, "retries": 0, "seconds": 169.0, "max_memory_gb": 24.93, "pair": 0.30387186631560326, "sign": 0.006324557005427778, "keep": 0.03155219216580463, "leak": 0.48527682200074196, "replacement": 0.44412943720817566} +{"step": 157, "steps": 250, "lr": 2.066213395457945e-05, "grad_norm": 3.8058700561523438, "loss_scale": 4.0, "retries": 0, "seconds": 170.1, "max_memory_gb": 24.93, "pair": 0.3134430325590074, "sign": 1.0892468402162194, "keep": 0.09617990866536275, "leak": 3.1508861798793077, "replacement": 1.69387786090374} +{"step": 158, "steps": 250, "lr": 2.0374970139466877e-05, "grad_norm": 2.5355281829833984, "loss_scale": 4.0, "retries": 0, "seconds": 171.2, "max_memory_gb": 24.93, "pair": 0.4485981948673725, "sign": 0.3766232007765211, "keep": 0.4648508028994911, "leak": 0.42905305279418826, "replacement": 0.976598471403122} +{"step": 159, "steps": 250, "lr": 2.0089090570278213e-05, "grad_norm": 1.3076965808868408, "loss_scale": 4.0, "retries": 0, "seconds": 172.2, "max_memory_gb": 24.99, "pair": 0.2940502427518368, "sign": 0.016818295931443572, "keep": 0.003081015922361985, "leak": 0.8783110864460468, "replacement": 0.5235776752233505} +{"step": 160, "steps": 250, "lr": 1.9804546775172457e-05, "grad_norm": 3.7664358615875244, "loss_scale": 4.0, "retries": 0, "seconds": 173.3, "max_memory_gb": 24.99, "pair": 0.22144816257059574, "sign": 0.1467078490240965, "keep": 0.5245042748283595, "leak": 1.0973334144800901, "replacement": 0.8313875608146191} +{"step": 161, "steps": 250, "lr": 1.9521390041542946e-05, "grad_norm": 1.729967474937439, "loss_scale": 4.0, "retries": 0, "seconds": 174.4, "max_memory_gb": 24.99, "pair": 0.41211695596575737, "sign": 0.15593015006743371, "keep": 0.2260102191180522, "leak": 0.3094064365141094, "replacement": 0.6804387643933296} +{"step": 162, "steps": 250, "lr": 1.923967140677313e-05, "grad_norm": 1.83511221408844, "loss_scale": 4.0, "retries": 0, "seconds": 175.5, "max_memory_gb": 24.99, "pair": 0.2401308063417673, "sign": 0.037572961940895766, "keep": 0.2242560943054741, "leak": 1.0927541702985764, "replacement": 0.6442338675260544} +{"step": 163, "steps": 250, "lr": 1.8959441649037335e-05, "grad_norm": 12.363283157348633, "loss_scale": 4.0, "retries": 0, "seconds": 176.5, "max_memory_gb": 24.99, "pair": 0.9788151532411575, "sign": 0.6186982106883079, "keep": 0.9413886107504368, "leak": 1.1784784680930898, "replacement": 2.0534781143069267} +{"step": 164, "steps": 250, "lr": 1.868075127814832e-05, "grad_norm": 5.020566940307617, "loss_scale": 4.0, "retries": 0, "seconds": 177.7, "max_memory_gb": 24.99, "pair": 0.5817694962024689, "sign": 0.3549208453623578, "keep": 0.2679030189756304, "leak": 0.39315043756505474, "replacement": 0.9914690032601357} +{"step": 165, "steps": 250, "lr": 1.8403650526453143e-05, "grad_norm": 1.1555222272872925, "loss_scale": 4.0, "retries": 0, "seconds": 178.8, "max_memory_gb": 24.99, "pair": 0.27640434354543686, "sign": 0.02555088623194024, "keep": 0.004771091381826409, "leak": 0.47965491749346256, "replacement": 0.41147907078266144} +{"step": 166, "steps": 250, "lr": 1.812818933977912e-05, "grad_norm": 7.497646808624268, "loss_scale": 4.0, "retries": 0, "seconds": 179.8, "max_memory_gb": 24.99, "pair": 0.7848343215882778, "sign": 0.7235962064005435, "keep": 1.1378323092940263, "leak": 0.8868104907451198, "replacement": 1.9372511208057404} +{"step": 167, "steps": 250, "lr": 1.785441736843128e-05, "grad_norm": 2.0417230129241943, "loss_scale": 4.0, "retries": 0, "seconds": 180.9, "max_memory_gb": 24.99, "pair": 0.3921002969145775, "sign": 0.07392324972897768, "keep": 0.08019361412152648, "leak": 0.40396696142852306, "replacement": 0.5701504498720169} +{"step": 168, "steps": 250, "lr": 1.7582383958243314e-05, "grad_norm": 0.8404585123062134, "loss_scale": 4.0, "retries": 0, "seconds": 182.0, "max_memory_gb": 24.99, "pair": 0.2766610737890005, "sign": 0.017286977090407163, "keep": 0.08653535446865135, "leak": 0.8852465599775314, "replacement": 0.5498838871717453} +{"step": 169, "steps": 250, "lr": 1.7312138141683164e-05, "grad_norm": 1.7634210586547852, "loss_scale": 4.0, "retries": 0, "seconds": 183.1, "max_memory_gb": 24.99, "pair": 0.46562105044722557, "sign": 0.16707195434719324, "keep": 0.21189073706045747, "leak": 0.3617795594036579, "replacement": 0.7455473095178604} +{"step": 170, "steps": 250, "lr": 1.704372862901521e-05, "grad_norm": 1.6742719411849976, "loss_scale": 4.0, "retries": 0, "seconds": 184.2, "max_memory_gb": 24.99, "pair": 0.2743111327290535, "sign": 0.0064439522684551775, "keep": 0.1397808197998529, "leak": 1.0640602111816406, "replacement": 0.6134385764598846} +{"step": 171, "steps": 250, "lr": 1.6777203799520573e-05, "grad_norm": 1.982043743133545, "loss_scale": 4.0, "retries": 0, "seconds": 185.3, "max_memory_gb": 24.99, "pair": 0.3241423089057207, "sign": 0.14557083987165242, "keep": 0.09257607080314756, "leak": 0.254929818212986, "replacement": 0.5069482289254665} +{"step": 172, "steps": 250, "lr": 1.6512611692777008e-05, "grad_norm": 1.1445151567459106, "loss_scale": 4.0, "retries": 0, "seconds": 186.4, "max_memory_gb": 24.99, "pair": 0.15897092781960964, "sign": 0.00411990872817114, "keep": 0.03367975656874478, "leak": 0.741442084312439, "replacement": 0.36323128640651703} +{"step": 173, "steps": 250, "lr": 1.6250000000000005e-05, "grad_norm": 0.906279444694519, "loss_scale": 4.0, "retries": 0, "seconds": 187.4, "max_memory_gb": 24.99, "pair": 0.23009561188519, "sign": 0.09614683655672707, "keep": 0.12514976819511503, "leak": 0.7172860510181636, "replacement": 0.5200654119253159} +{"step": 174, "steps": 250, "lr": 1.598941605544678e-05, "grad_norm": 1.2286524772644043, "loss_scale": 4.0, "retries": 0, "seconds": 188.5, "max_memory_gb": 24.99, "pair": 0.2366097755730152, "sign": 0.01990867091808468, "keep": 0.0803719137495591, "leak": 0.37574900686740875, "replacement": 0.38068731874227524} +{"step": 175, "steps": 250, "lr": 1.5730906827884506e-05, "grad_norm": 4.2647576332092285, "loss_scale": 4.0, "retries": 0, "seconds": 189.8, "max_memory_gb": 24.99, "pair": 0.3731683362275362, "sign": 0.2722278961446136, "keep": 1.306439084932208, "leak": 0.6584702329710126, "replacement": 1.3271193578839302} +{"step": 176, "steps": 250, "lr": 1.5474518912124476e-05, "grad_norm": 1.0375542640686035, "loss_scale": 4.0, "retries": 0, "seconds": 190.9, "max_memory_gb": 24.99, "pair": 0.23104071244597435, "sign": 0.03964825233560987, "keep": 0.020476743578910828, "leak": 0.45497722338768654, "replacement": 0.37484750896692276} +{"step": 177, "steps": 250, "lr": 1.5220298520623646e-05, "grad_norm": 1.6760914325714111, "loss_scale": 4.0, "retries": 0, "seconds": 192.0, "max_memory_gb": 24.99, "pair": 0.43331005051732063, "sign": 0.4676265332964249, "keep": 0.00913747777667595, "leak": 0.3693727634381503, "replacement": 0.7640352845191956} +{"step": 178, "steps": 250, "lr": 1.4968291475155121e-05, "grad_norm": 3.7500810623168945, "loss_scale": 4.0, "retries": 0, "seconds": 193.1, "max_memory_gb": 24.99, "pair": 0.2140225013718009, "sign": 0.09733506082557142, "keep": 0.2732184684209642, "leak": 1.5806738324463367, "replacement": 0.7944677174091339} +{"step": 179, "steps": 250, "lr": 1.4718543198548998e-05, "grad_norm": 1.8318873643875122, "loss_scale": 4.0, "retries": 0, "seconds": 194.1, "max_memory_gb": 24.99, "pair": 0.6923937350511551, "sign": 0.5702927092206664, "keep": 0.6247464066691464, "leak": 0.20215080212801695, "replacement": 1.3404510393738747} +{"step": 180, "steps": 250, "lr": 1.4471098706505227e-05, "grad_norm": 0.9772827625274658, "loss_scale": 4.0, "retries": 0, "seconds": 195.2, "max_memory_gb": 24.99, "pair": 0.42059263959527016, "sign": 0.04382743185851723, "keep": 0.07488440105225891, "leak": 0.481336077908054, "replacement": 0.6002825796604156} +{"step": 181, "steps": 250, "lr": 1.4226002599479743e-05, "grad_norm": 2.469465970993042, "loss_scale": 4.0, "retries": 0, "seconds": 196.2, "max_memory_gb": 24.99, "pair": 0.18559584766626358, "sign": 0.015292140538804233, "keep": 0.06957277865149081, "leak": 2.0368454456329346, "replacement": 0.7372396886348724} +{"step": 182, "steps": 250, "lr": 1.3983299054645471e-05, "grad_norm": 4.148369789123535, "loss_scale": 4.0, "retries": 0, "seconds": 197.3, "max_memory_gb": 24.99, "pair": 0.4791129529476166, "sign": 0.31713391258381307, "keep": 0.024831723087118007, "leak": 1.418541144579649, "replacement": 1.00473103672266} +{"step": 183, "steps": 250, "lr": 1.3743031817929724e-05, "grad_norm": 0.8816785216331482, "loss_scale": 4.0, "retries": 0, "seconds": 198.4, "max_memory_gb": 24.99, "pair": 0.32513774931430817, "sign": 0.08334699302213266, "keep": 0.010504899315037619, "leak": 0.7418855167925358, "replacement": 0.5575350672006607} +{"step": 184, "steps": 250, "lr": 1.3505244196129164e-05, "grad_norm": 1.3782780170440674, "loss_scale": 4.0, "retries": 0, "seconds": 199.6, "max_memory_gb": 24.99, "pair": 0.28254637494683266, "sign": 0.008981412567663938, "keep": 0.020976212013920303, "leak": 0.7028708010911942, "replacement": 0.4732428789138794} +{"step": 185, "steps": 250, "lr": 1.3269979049104014e-05, "grad_norm": 1.3064557313919067, "loss_scale": 4.0, "retries": 0, "seconds": 200.7, "max_memory_gb": 24.99, "pair": 0.3010259550064802, "sign": 0.03992174629820511, "keep": 0.14196350472047925, "leak": 0.8695204071700573, "replacement": 0.6093486919999123} +{"step": 186, "steps": 250, "lr": 1.3037278782052863e-05, "grad_norm": 2.22420334815979, "loss_scale": 4.0, "retries": 0, "seconds": 201.8, "max_memory_gb": 24.99, "pair": 0.43661223724484444, "sign": 0.02706250798655674, "keep": 0.8571585751910789, "leak": 0.4008036106824875, "replacement": 0.9789236634969711} +{"step": 187, "steps": 250, "lr": 1.2807185337869343e-05, "grad_norm": 1.1716539859771729, "loss_scale": 4.0, "retries": 0, "seconds": 202.9, "max_memory_gb": 24.99, "pair": 0.29500249586999416, "sign": 0.009634711313992739, "keep": 0.46586935652885586, "leak": 0.779317220672965, "replacement": 0.7275838479399681} +{"step": 188, "steps": 250, "lr": 1.2579740189582115e-05, "grad_norm": 2.3495354652404785, "loss_scale": 4.0, "retries": 0, "seconds": 203.9, "max_memory_gb": 24.99, "pair": 0.5926392152905464, "sign": 0.8545813377131708, "keep": 0.7122217144424212, "leak": 0.2811112813651562, "replacement": 1.4463184960186481} +{"step": 189, "steps": 250, "lr": 1.2354984332879668e-05, "grad_norm": 3.7598488330841064, "loss_scale": 4.0, "retries": 0, "seconds": 205.0, "max_memory_gb": 24.99, "pair": 0.33589346520602703, "sign": 0.3942911250051111, "keep": 0.8193914893781766, "leak": 1.607711043174504, "replacement": 1.3446625173091888} +{"step": 190, "steps": 250, "lr": 1.2132958278721064e-05, "grad_norm": 1.4050366878509521, "loss_scale": 4.0, "retries": 0, "seconds": 206.1, "max_memory_gb": 24.99, "pair": 0.19082124903798103, "sign": 0.008124373620375991, "keep": 0.01612045877845958, "leak": 1.1482974141836166, "replacement": 0.49001801759004593} +{"step": 191, "steps": 250, "lr": 1.1913702046034016e-05, "grad_norm": 8.395153045654297, "loss_scale": 4.0, "retries": 0, "seconds": 207.1, "max_memory_gb": 24.99, "pair": 0.5257293079048395, "sign": 0.4240441604051739, "keep": 0.5857828713913023, "leak": 0.8973417282104492, "replacement": 1.2549782916903496} +{"step": 192, "steps": 250, "lr": 1.1697255154501756e-05, "grad_norm": 1.2581760883331299, "loss_scale": 4.0, "retries": 0, "seconds": 208.1, "max_memory_gb": 24.99, "pair": 0.5972996801137924, "sign": 0.488755667116493, "keep": 1.1125326175242662, "leak": 0.5088561633601785, "replacement": 1.5251579135656357} +{"step": 193, "steps": 250, "lr": 1.1483656617439842e-05, "grad_norm": 1.3761824369430542, "loss_scale": 4.0, "retries": 0, "seconds": 209.1, "max_memory_gb": 24.99, "pair": 0.31209372356534004, "sign": 0.032937017269432545, "keep": 0.03414852600872109, "leak": 0.6709902733564377, "replacement": 0.5133840590715408} +{"step": 194, "steps": 250, "lr": 1.1272944934764177e-05, "grad_norm": 2.062492847442627, "loss_scale": 4.0, "retries": 0, "seconds": 210.2, "max_memory_gb": 24.99, "pair": 0.5228137522935867, "sign": 0.3200417053885758, "keep": 0.024076367411907995, "leak": 0.41804844699800014, "replacement": 0.7993848994374275} +{"step": 195, "steps": 250, "lr": 1.1065158086051684e-05, "grad_norm": 2.010516405105591, "loss_scale": 4.0, "retries": 0, "seconds": 211.2, "max_memory_gb": 24.99, "pair": 0.39275528863072395, "sign": 0.07077908486826345, "keep": 0.010677798869437538, "leak": 1.005004070699215, "replacement": 0.6847347468137741} +{"step": 196, "steps": 250, "lr": 1.086033352369467e-05, "grad_norm": 1.2690080404281616, "loss_scale": 4.0, "retries": 0, "seconds": 212.4, "max_memory_gb": 24.99, "pair": 0.14749355614185333, "sign": 0.00850968511076644, "keep": 0.024042396398954224, "leak": 1.1161699146032333, "replacement": 0.44281207397580147} +{"step": 197, "steps": 250, "lr": 1.0658508166150224e-05, "grad_norm": 1.2366952896118164, "loss_scale": 4.0, "retries": 0, "seconds": 213.5, "max_memory_gb": 24.99, "pair": 0.2860703580081463, "sign": 0.03664178459439427, "keep": 0.009499028892605565, "leak": 0.8494780138134956, "replacement": 0.5215102657675743} +{"step": 198, "steps": 250, "lr": 1.0459718391285954e-05, "grad_norm": 2.5155980587005615, "loss_scale": 4.0, "retries": 0, "seconds": 214.6, "max_memory_gb": 24.99, "pair": 0.23835106939077377, "sign": 0.08098439080640674, "keep": 0.0960849372568191, "leak": 0.9375838525593281, "replacement": 0.5612816959619522} +{"step": 199, "steps": 250, "lr": 1.0264000029822999e-05, "grad_norm": 3.3235890865325928, "loss_scale": 4.0, "retries": 0, "seconds": 215.7, "max_memory_gb": 24.99, "pair": 0.17599445302039385, "sign": 0.034568350500194356, "keep": 0.5687340884032892, "leak": 1.573245532810688, "replacement": 0.870957039296627} +{"step": 200, "steps": 250, "lr": 1.0071388358877775e-05, "grad_norm": 2.677016019821167, "loss_scale": 4.0, "retries": 0, "seconds": 216.8, "max_memory_gb": 24.99, "pair": 0.44903623778373003, "sign": 0.3678488750301767, "keep": 1.2037163924542256, "leak": 1.039696453139186, "replacement": 1.494742974638939} +{"step": 201, "steps": 250, "lr": 9.881918095603501e-06, "grad_norm": 2.8060665130615234, "loss_scale": 4.0, "retries": 0, "seconds": 217.9, "max_memory_gb": 24.99, "pair": 0.42836498096585274, "sign": 0.2014997557271272, "keep": 0.19323230828376836, "leak": 0.42390753142535686, "replacement": 0.7317078933119774} +{"step": 202, "steps": 250, "lr": 9.695623390932612e-06, "grad_norm": 1.2002732753753662, "loss_scale": 4.0, "retries": 0, "seconds": 219.0, "max_memory_gb": 24.99, "pair": 0.3927311599254608, "sign": 0.14990553091047332, "keep": 0.007730083510068653, "leak": 0.7046972066164017, "replacement": 0.6477232575416565} +{"step": 203, "steps": 250, "lr": 9.51253782342123e-06, "grad_norm": 1.0474112033843994, "loss_scale": 4.0, "retries": 0, "seconds": 220.1, "max_memory_gb": 24.99, "pair": 0.15404482372105122, "sign": 0.04516937187872827, "keep": 0.01009875995805487, "leak": 0.8938828110694885, "replacement": 0.40514960139989853} +{"step": 204, "steps": 250, "lr": 9.332694393196846e-06, "grad_norm": 0.8149201273918152, "loss_scale": 4.0, "retries": 0, "seconds": 221.2, "max_memory_gb": 24.99, "pair": 0.3718704879283905, "sign": 0.028529081493616104, "keep": 0.053709254949353635, "leak": 0.8863686682889238, "replacement": 0.6345818117260933} +{"step": 205, "steps": 250, "lr": 9.156125516010256e-06, "grad_norm": 0.8895628452301025, "loss_scale": 4.0, "retries": 0, "seconds": 222.2, "max_memory_gb": 24.99, "pair": 0.245581628754735, "sign": 0.008125437656417489, "keep": 0.0030012432289368007, "leak": 0.5750560015439987, "replacement": 0.39490897208452225} +{"step": 206, "steps": 250, "lr": 8.982863017392733e-06, "grad_norm": 1.2300633192062378, "loss_scale": 4.0, "retries": 0, "seconds": 223.3, "max_memory_gb": 24.99, "pair": 0.26778467185795307, "sign": 0.03904356784187257, "keep": 0.04004437103003511, "leak": 0.4835614189505577, "replacement": 0.4282190129160881} +{"step": 207, "steps": 250, "lr": 8.81293812691974e-06, "grad_norm": 1.1300227642059326, "loss_scale": 4.0, "retries": 0, "seconds": 224.4, "max_memory_gb": 24.99, "pair": 0.48043570667505264, "sign": 0.1702093841158785, "keep": 0.08346565888086843, "leak": 0.3058642678661272, "replacement": 0.6837393417954445} +{"step": 208, "steps": 250, "lr": 8.646381472581927e-06, "grad_norm": 0.9903798699378967, "loss_scale": 4.0, "retries": 0, "seconds": 225.5, "max_memory_gb": 24.99, "pair": 0.16263315454125404, "sign": 0.004163047473412007, "keep": 0.030736499807517248, "leak": 0.657559422776103, "replacement": 0.3444727808237076} +{"step": 209, "steps": 250, "lr": 8.483223075264622e-06, "grad_norm": 2.0491628646850586, "loss_scale": 4.0, "retries": 0, "seconds": 226.6, "max_memory_gb": 24.99, "pair": 0.49101484939455986, "sign": 0.3926476505585015, "keep": 0.3188402026571566, "leak": 0.9482863899320364, "replacement": 1.0838303864002228} +{"step": 210, "steps": 250, "lr": 8.323492343336747e-06, "grad_norm": 2.835543632507324, "loss_scale": 4.0, "retries": 0, "seconds": 227.7, "max_memory_gb": 24.99, "pair": 0.5375588834285736, "sign": 0.49868948257062584, "keep": 1.3273123729718463, "leak": 0.45124703738838434, "replacement": 1.5633715391159058} +{"step": 211, "steps": 250, "lr": 8.167218067350122e-06, "grad_norm": 0.9926357269287109, "loss_scale": 4.0, "retries": 0, "seconds": 228.8, "max_memory_gb": 24.99, "pair": 0.49684442579746246, "sign": 0.2651340625016019, "keep": 0.09437710233032703, "leak": 0.6011285916902125, "replacement": 0.8268821537494659} +{"step": 212, "steps": 250, "lr": 8.01442841485013e-06, "grad_norm": 6.977097988128662, "loss_scale": 4.0, "retries": 0, "seconds": 230.0, "max_memory_gb": 24.99, "pair": 0.7231140099465847, "sign": 0.24852213641861454, "keep": 0.1210326119326055, "leak": 1.378215602133423, "replacement": 1.252445325255394} +{"step": 213, "steps": 250, "lr": 7.865150925298678e-06, "grad_norm": 2.227473497390747, "loss_scale": 4.0, "retries": 0, "seconds": 231.1, "max_memory_gb": 24.99, "pair": 0.4637271389365196, "sign": 0.3504747621482238, "keep": 0.541645475444966, "leak": 0.5293718059547246, "replacement": 1.042130209505558} +{"step": 214, "steps": 250, "lr": 7.719412505110375e-06, "grad_norm": 1.433350920677185, "loss_scale": 4.0, "retries": 0, "seconds": 232.2, "max_memory_gb": 24.99, "pair": 0.4955456294119358, "sign": 0.3860320364474319, "keep": 1.2428011884794614, "leak": 0.0678311123047024, "replacement": 1.326919972896576} +{"step": 215, "steps": 250, "lr": 7.57723942280278e-06, "grad_norm": 2.8206026554107666, "loss_scale": 4.0, "retries": 0, "seconds": 233.3, "max_memory_gb": 24.99, "pair": 0.7728674449026585, "sign": 0.8214359041303396, "keep": 1.3611158429516763, "leak": 0.07518915338636134, "replacement": 1.8829406052827835} +{"step": 216, "steps": 250, "lr": 7.438657304261687e-06, "grad_norm": 0.9174391627311707, "loss_scale": 4.0, "retries": 0, "seconds": 234.4, "max_memory_gb": 24.99, "pair": 0.17493557929992676, "sign": 0.030091025168076158, "keep": 0.01958649525840883, "leak": 0.8337089195847511, "replacement": 0.40820156782865524} +{"step": 217, "steps": 250, "lr": 7.30369112812219e-06, "grad_norm": 1.2611151933670044, "loss_scale": 4.0, "retries": 0, "seconds": 235.5, "max_memory_gb": 24.99, "pair": 0.21983375772833824, "sign": 0.022175160353071988, "keep": 0.0006954239488550229, "leak": 0.9773836508393288, "replacement": 0.4756149649620056} +{"step": 218, "steps": 250, "lr": 7.172365221266403e-06, "grad_norm": 8.273880004882812, "loss_scale": 4.0, "retries": 0, "seconds": 236.6, "max_memory_gb": 24.99, "pair": 1.1362624764442444, "sign": 1.8333762288093567, "keep": 0.4450632850639522, "leak": 0.5772938574664295, "replacement": 2.4198057055473328} +{"step": 219, "steps": 250, "lr": 7.0447032544387154e-06, "grad_norm": 0.973691463470459, "loss_scale": 4.0, "retries": 0, "seconds": 237.7, "max_memory_gb": 24.99, "pair": 0.3146728128194809, "sign": 0.0845340231899172, "keep": 0.0047265142275136895, "leak": 0.512618824839592, "replacement": 0.4874577820301056} +{"step": 220, "steps": 250, "lr": 6.920728237979274e-06, "grad_norm": 1.4642994403839111, "loss_scale": 4.0, "retries": 0, "seconds": 238.8, "max_memory_gb": 24.99, "pair": 0.3964524269104004, "sign": 0.055544751056004316, "keep": 0.06559567411204625, "leak": 0.23486266657710075, "replacement": 0.5157383158802986} +{"step": 221, "steps": 250, "lr": 6.800462517676456e-06, "grad_norm": 0.8963322639465332, "loss_scale": 4.0, "retries": 0, "seconds": 239.9, "max_memory_gb": 24.99, "pair": 0.3693454675376415, "sign": 0.17790049209725112, "keep": 0.012877494329586625, "leak": 0.7536231745034456, "replacement": 0.6531402617692947} +{"step": 222, "steps": 250, "lr": 6.683927770739222e-06, "grad_norm": 1.9718118906021118, "loss_scale": 4.0, "retries": 0, "seconds": 241.0, "max_memory_gb": 24.99, "pair": 0.2758192718029022, "sign": 0.029670532152522355, "keep": 0.056898416793046636, "leak": 0.5292431870475411, "replacement": 0.4514145404100418} +{"step": 223, "steps": 250, "lr": 6.571145001889892e-06, "grad_norm": 5.309587478637695, "loss_scale": 4.0, "retries": 0, "seconds": 242.1, "max_memory_gb": 24.99, "pair": 0.5834416970610619, "sign": 0.09785410697804764, "keep": 0.018058152605590294, "leak": 0.7232361994683743, "replacement": 0.8222068697214127} +{"step": 224, "steps": 250, "lr": 6.46213453957817e-06, "grad_norm": 1.5152078866958618, "loss_scale": 4.0, "retries": 0, "seconds": 243.2, "max_memory_gb": 24.99, "pair": 0.39440203458070755, "sign": 0.04787377826869488, "keep": 0.008997296172310598, "leak": 0.3068964993581176, "replacement": 0.49956168234348297} +{"step": 225, "steps": 250, "lr": 6.356916032317063e-06, "grad_norm": 1.34420645236969, "loss_scale": 4.0, "retries": 0, "seconds": 244.3, "max_memory_gb": 24.99, "pair": 0.2567629776895046, "sign": 0.03627667436376214, "keep": 0.003438861433096463, "leak": 0.8082209415733814, "replacement": 0.478675976395607} +{"step": 226, "steps": 250, "lr": 6.255508445141349e-06, "grad_norm": 6.642570972442627, "loss_scale": 4.0, "retries": 0, "seconds": 245.4, "max_memory_gb": 24.99, "pair": 0.2684350134804845, "sign": 0.036483674542978406, "keep": 0.8457033586746547, "leak": 0.9200051221996546, "replacement": 0.9395297914743423} +{"step": 227, "steps": 250, "lr": 6.157930056189228e-06, "grad_norm": 1.3736262321472168, "loss_scale": 4.0, "retries": 0, "seconds": 246.4, "max_memory_gb": 24.99, "pair": 0.23158452659845352, "sign": 0.016892935382202268, "keep": 0.05601030691605047, "leak": 0.566243963316083, "replacement": 0.40959715098142624} +{"step": 228, "steps": 250, "lr": 6.064198453407796e-06, "grad_norm": 1.0417975187301636, "loss_scale": 4.0, "retries": 0, "seconds": 247.5, "max_memory_gb": 24.99, "pair": 0.23069331422448158, "sign": 0.05388856143690646, "keep": 0.0017416144314665871, "leak": 1.0159521028399467, "replacement": 0.5124964267015457} +{"step": 229, "steps": 250, "lr": 5.974330531382935e-06, "grad_norm": 1.7731822729110718, "loss_scale": 4.0, "retries": 0, "seconds": 248.6, "max_memory_gb": 24.99, "pair": 0.3830191530287266, "sign": 0.1205246530007571, "keep": 0.1103319058883585, "leak": 0.9129511440114584, "replacement": 0.7266852110624313} +{"step": 230, "steps": 250, "lr": 5.888342488294125e-06, "grad_norm": 39.756385803222656, "loss_scale": 4.0, "retries": 0, "seconds": 249.6, "max_memory_gb": 24.99, "pair": 1.0833229720592499, "sign": 0.649001732468605, "keep": 0.9139517678577249, "leak": 0.510363171575591, "replacement": 1.992390513420105} +{"step": 231, "steps": 250, "lr": 5.806249822994849e-06, "grad_norm": 1.6890212297439575, "loss_scale": 4.0, "retries": 0, "seconds": 250.7, "max_memory_gb": 24.99, "pair": 0.5734516717493534, "sign": 0.7052554795518517, "keep": 0.03385873494198677, "leak": 0.44706688076257706, "replacement": 1.054775483906269} +{"step": 232, "steps": 250, "lr": 5.728067332218994e-06, "grad_norm": 2.685453176498413, "loss_scale": 4.0, "retries": 0, "seconds": 251.8, "max_memory_gb": 24.99, "pair": 0.20572122558951378, "sign": 0.02415109591674991, "keep": 0.05330219061579555, "leak": 1.7820310518145561, "replacement": 0.6899556368589401} +{"step": 233, "steps": 250, "lr": 5.65380910791383e-06, "grad_norm": 1.0131847858428955, "loss_scale": 4.0, "retries": 0, "seconds": 252.9, "max_memory_gb": 24.99, "pair": 0.27292584627866745, "sign": 0.00877245661104098, "keep": 0.012286557350307703, "leak": 0.9153244532644749, "replacement": 0.5122864693403244} +{"step": 234, "steps": 250, "lr": 5.583488534700022e-06, "grad_norm": 2.6371700763702393, "loss_scale": 4.0, "retries": 0, "seconds": 254.0, "max_memory_gb": 24.99, "pair": 0.22101532109081745, "sign": 0.0248621697537601, "keep": 0.05236437737039523, "leak": 0.7801490761339664, "replacement": 0.454665869474411} +{"step": 235, "steps": 250, "lr": 5.517118287459118e-06, "grad_norm": 9.997612953186035, "loss_scale": 4.0, "retries": 0, "seconds": 255.1, "max_memory_gb": 24.99, "pair": 1.0767176058143377, "sign": 0.6876704229507595, "keep": 1.3044942498563614, "leak": 2.022232748568058, "replacement": 2.578358009457588} +{"step": 236, "steps": 250, "lr": 5.454710329048978e-06, "grad_norm": 1.1501895189285278, "loss_scale": 4.0, "retries": 0, "seconds": 256.2, "max_memory_gb": 24.99, "pair": 0.34980154410004616, "sign": 0.10414957979810424, "keep": 0.005787847787132705, "leak": 0.19838217552751303, "replacement": 0.4543657973408699} +{"step": 237, "steps": 250, "lr": 5.3962759081475436e-06, "grad_norm": 5.138010501861572, "loss_scale": 4.0, "retries": 0, "seconds": 257.2, "max_memory_gb": 24.99, "pair": 0.4382684752345085, "sign": 0.42636554641649127, "keep": 0.14265052443442983, "leak": 0.3259022943675518, "replacement": 0.8042520508170128} +{"step": 238, "steps": 250, "lr": 5.341825557225321e-06, "grad_norm": 1.977783203125, "loss_scale": 4.0, "retries": 0, "seconds": 258.3, "max_memory_gb": 24.99, "pair": 0.24123477563261986, "sign": 0.01903329137712717, "keep": 0.034335087053477764, "leak": 0.7060666233301163, "replacement": 0.44443561881780624} +{"step": 239, "steps": 250, "lr": 5.291369090646961e-06, "grad_norm": 2.5061237812042236, "loss_scale": 4.0, "retries": 0, "seconds": 259.4, "max_memory_gb": 24.99, "pair": 0.6328091993927956, "sign": 1.4336691470816731, "keep": 0.16068989378982224, "leak": 0.44574308009350716, "replacement": 1.5414244532585144} +{"step": 240, "steps": 250, "lr": 5.244915602902289e-06, "grad_norm": 1.1575303077697754, "loss_scale": 4.0, "retries": 0, "seconds": 260.4, "max_memory_gb": 24.99, "pair": 0.40914881229400635, "sign": 0.33875303337117657, "keep": 0.1996708688857325, "leak": 0.5939733497798443, "replacement": 0.8268541246652603} +{"step": 241, "steps": 250, "lr": 5.20247346696706e-06, "grad_norm": 6.019176959991455, "loss_scale": 4.0, "retries": 0, "seconds": 261.6, "max_memory_gb": 24.99, "pair": 0.9024174474179745, "sign": 0.6317698795464821, "keep": 0.3639980854893565, "leak": 0.500407219864428, "replacement": 1.5254032388329506} +{"step": 242, "steps": 250, "lr": 5.164050332793787e-06, "grad_norm": 1.487187385559082, "loss_scale": 4.0, "retries": 0, "seconds": 262.6, "max_memory_gb": 24.99, "pair": 0.42722487077116966, "sign": 0.050640727393329144, "keep": 0.004231036638657315, "leak": 0.6532794162631035, "replacement": 0.617980606853962} +{"step": 243, "steps": 250, "lr": 5.129653125932873e-06, "grad_norm": 1.2078847885131836, "loss_scale": 4.0, "retries": 0, "seconds": 263.7, "max_memory_gb": 24.99, "pair": 0.24766818061470985, "sign": 0.021910291514359415, "keep": 0.059002846508519724, "leak": 0.9007081463932991, "replacement": 0.513301782310009} +{"step": 244, "steps": 250, "lr": 5.099288046284329e-06, "grad_norm": 2.4074032306671143, "loss_scale": 4.0, "retries": 0, "seconds": 264.8, "max_memory_gb": 24.99, "pair": 0.3866242989897728, "sign": 0.18010614393278956, "keep": 0.026819340884685516, "leak": 1.1639584824442863, "replacement": 0.7810766249895096} +{"step": 245, "steps": 250, "lr": 5.072960566980277e-06, "grad_norm": 4.175290584564209, "loss_scale": 4.0, "retries": 0, "seconds": 265.9, "max_memory_gb": 24.99, "pair": 0.6567663475871086, "sign": 0.40703338012099266, "keep": 0.7492658756673336, "leak": 0.3459197459742427, "replacement": 1.3213958963751793} +{"step": 246, "steps": 250, "lr": 5.050675433398435e-06, "grad_norm": 0.8467075824737549, "loss_scale": 4.0, "retries": 0, "seconds": 267.1, "max_memory_gb": 24.99, "pair": 0.22003832645714283, "sign": 0.006694875308312476, "keep": 0.006287017838985776, "leak": 0.5727590546011925, "replacement": 0.369719035923481} +{"step": 247, "steps": 250, "lr": 5.032436662306811e-06, "grad_norm": 3.2035775184631348, "loss_scale": 4.0, "retries": 0, "seconds": 268.2, "max_memory_gb": 24.99, "pair": 0.26987424492836, "sign": 0.02557332639116794, "keep": 0.29556808683423696, "leak": 0.8317966461181641, "replacement": 0.6383941173553467} +{"step": 248, "steps": 250, "lr": 5.018247541139684e-06, "grad_norm": 0.8486275672912598, "loss_scale": 4.0, "retries": 0, "seconds": 269.2, "max_memory_gb": 24.99, "pair": 0.24245628714561462, "sign": 0.036714589688926935, "keep": 0.04881329067211482, "leak": 0.6852042805403471, "replacement": 0.45652129501104355} +{"step": 249, "steps": 250, "lr": 5.00811062740508e-06, "grad_norm": 1.6075814962387085, "loss_scale": 4.0, "retries": 0, "seconds": 270.4, "max_memory_gb": 24.99, "pair": 0.2608849871903658, "sign": 0.08294130954891443, "keep": 0.05676294007571414, "leak": 0.8112590275704861, "replacement": 0.5335518717765808} +{"step": 250, "steps": 250, "lr": 5.002027748223778e-06, "grad_norm": 1.9267163276672363, "loss_scale": 4.0, "retries": 0, "seconds": 271.4, "max_memory_gb": 24.99, "pair": 0.4951863959431648, "sign": 0.04435571865178645, "keep": 0.11805228857338079, "leak": 0.2158895842730999, "replacement": 0.6303628012537956} diff --git a/runs/single_internvl/replacement/L02_s13/adapter/README.md b/runs/single_internvl/replacement/L02_s13/adapter/README.md new file mode 100644 index 0000000000000000000000000000000000000000..4f9f856cf0e1531b47e25d266833b2977a19d3ac --- /dev/null +++ b/runs/single_internvl/replacement/L02_s13/adapter/README.md @@ -0,0 +1,207 @@ +--- +base_model: /data/OPSD/assets/models/InternVL3_5-8B-HF +library_name: peft +pipeline_tag: text-generation +tags: +- base_model:adapter:/data/OPSD/assets/models/InternVL3_5-8B-HF +- lora +- transformers +--- + +# Model Card for Model ID + + + + + +## Model Details + +### Model Description + + + + + +- **Developed by:** [More Information Needed] +- **Funded by [optional]:** [More Information Needed] +- **Shared by [optional]:** [More Information Needed] +- **Model type:** [More Information Needed] +- **Language(s) (NLP):** [More Information Needed] +- **License:** [More Information Needed] +- **Finetuned from model [optional]:** [More Information Needed] + +### Model Sources [optional] + + + +- **Repository:** [More Information Needed] +- **Paper [optional]:** [More Information Needed] +- **Demo [optional]:** [More Information Needed] + +## Uses + + + +### Direct Use + + + +[More Information Needed] + +### Downstream Use [optional] + + + +[More Information Needed] + +### Out-of-Scope Use + + + +[More Information Needed] + +## Bias, Risks, and Limitations + + + +[More Information Needed] + +### Recommendations + + + +Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. More information needed for further recommendations. + +## How to Get Started with the Model + +Use the code below to get started with the model. + +[More Information Needed] + +## Training Details + +### Training Data + + + +[More Information Needed] + +### Training Procedure + + + +#### Preprocessing [optional] + +[More Information Needed] + + +#### Training Hyperparameters + +- **Training regime:** [More Information Needed] + +#### Speeds, Sizes, Times [optional] + + + +[More Information Needed] + +## Evaluation + + + +### Testing Data, Factors & Metrics + +#### Testing Data + + + +[More Information Needed] + +#### Factors + + + +[More Information Needed] + +#### Metrics + + + +[More Information Needed] + +### Results + +[More Information Needed] + +#### Summary + + + +## Model Examination [optional] + + + +[More Information Needed] + +## Environmental Impact + + + +Carbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700). + +- **Hardware Type:** [More Information Needed] +- **Hours used:** [More Information Needed] +- **Cloud Provider:** [More Information Needed] +- **Compute Region:** [More Information Needed] +- **Carbon Emitted:** [More Information Needed] + +## Technical Specifications [optional] + +### Model Architecture and Objective + +[More Information Needed] + +### Compute Infrastructure + +[More Information Needed] + +#### Hardware + +[More Information Needed] + +#### Software + +[More Information Needed] + +## Citation [optional] + + + +**BibTeX:** + +[More Information Needed] + +**APA:** + +[More Information Needed] + +## Glossary [optional] + + + +[More Information Needed] + +## More Information [optional] + +[More Information Needed] + +## Model Card Authors [optional] + +[More Information Needed] + +## Model Card Contact + +[More Information Needed] +### Framework versions + +- PEFT 0.20.0 \ No newline at end of file diff --git a/runs/single_internvl/replacement/L02_s13/adapter/adapter_config.json b/runs/single_internvl/replacement/L02_s13/adapter/adapter_config.json new file mode 100644 index 0000000000000000000000000000000000000000..c814dcc9fa1a10490d9d55d1f165cd3476888925 --- /dev/null +++ b/runs/single_internvl/replacement/L02_s13/adapter/adapter_config.json @@ -0,0 +1,50 @@ +{ + "alora_invocation_tokens": null, + "alpha_pattern": {}, + "arrow_config": null, + "auto_mapping": null, + "base_model_name_or_path": "/data/OPSD/assets/models/InternVL3_5-8B-HF", + "bias": "none", + "corda_config": null, + "ensure_weight_tying": false, + "eva_config": null, + "exclude_modules": null, + "fan_in_fan_out": false, + "inference_mode": true, + "init_lora_weights": true, + "layer_replication": null, + "layers_pattern": null, + "layers_to_transform": null, + "loftq_config": {}, + "lora_alpha": 32, + "lora_bias": false, + "lora_dropout": 0.05, + "lora_ga_config": null, + "megatron_config": null, + "megatron_core": "megatron.core", + "modules_to_save": null, + "monteclora_config": null, + "peft_type": "LORA", + "peft_version": "0.20.0", + "qalora_group_size": 16, + "r": 16, + "rank_pattern": {}, + "revision": null, + "target_modules": [ + "model.language_model.layers.2.mlp.up_proj", + "model.language_model.layers.2.mlp.down_proj", + "model.language_model.layers.2.self_attn.o_proj", + "model.language_model.layers.2.mlp.gate_proj", + "model.language_model.layers.2.self_attn.k_proj", + "model.language_model.layers.2.self_attn.v_proj", + "model.language_model.layers.2.self_attn.q_proj" + ], + "target_parameters": null, + "task_type": "CAUSAL_LM", + "trainable_token_indices": null, + "use_bdlora": null, + "use_dora": false, + "use_qalora": false, + "use_rslora": false, + "velora_config": null +} \ No newline at end of file diff --git a/runs/single_internvl/replacement/L02_s13/config.json b/runs/single_internvl/replacement/L02_s13/config.json new file mode 100644 index 0000000000000000000000000000000000000000..1736755114411b93ff48b5928501feb572fa1d3e --- /dev/null +++ b/runs/single_internvl/replacement/L02_s13/config.json @@ -0,0 +1,24 @@ +{ + "behavior": "replacement", + "seed": 13, + "model": "/data/OPSD/assets/models/InternVL3_5-8B-HF", + "backbone": "internvl", + "train": "/data/OPSD/assets/v3/replacement.jsonl", + "items": 1000, + "steps": 250, + "epochs": 1, + "steps_per_epoch": 250, + "groups_per_step": 4, + "lr": 5e-05, + "scheduler": "cosine", + "warmup_steps": 16, + "final_fraction": 0.1, + "weight_decay": 0.0, + "clip": 1.0, + "initial_loss_scale": 4.0, + "adapter": { + "layer": 2, + "rank": 16, + "trainable_parameters": 1212416 + } +} \ No newline at end of file diff --git a/runs/single_internvl/replacement/L02_s13/steps.jsonl b/runs/single_internvl/replacement/L02_s13/steps.jsonl new file mode 100644 index 0000000000000000000000000000000000000000..b2496c58bf6994fff87c9e301a24d8b38f4f924a --- /dev/null +++ b/runs/single_internvl/replacement/L02_s13/steps.jsonl @@ -0,0 +1,250 @@ +{"step": 1, "steps": 250, "lr": 3.125e-06, "grad_norm": 0.9583612084388733, "loss_scale": 4.0, "retries": 0, "seconds": 1.1, "max_memory_gb": 24.6, "pair": 1.0986123085021973, "sign": 0.7328193015418947, "keep": 1.137254401575774, "leak": 0.0, "replacement": 2.0336491465568542} +{"step": 2, "steps": 250, "lr": 6.25e-06, "grad_norm": 0.9831684231758118, "loss_scale": 4.0, "retries": 0, "seconds": 2.1, "max_memory_gb": 24.6, "pair": 1.096964567899704, "sign": 0.1813456095987931, "keep": 1.368505568461842, "leak": 2.480725449061083e-05, "replacement": 1.8718963265419006} +{"step": 3, "steps": 250, "lr": 9.375000000000001e-06, "grad_norm": 2.420527696609497, "loss_scale": 4.0, "retries": 0, "seconds": 3.1, "max_memory_gb": 24.6, "pair": 1.0907756984233856, "sign": 0.11447693407535553, "keep": 2.064627856016159, "leak": 2.4098124242755148e-05, "replacement": 2.1803340315818787} +{"step": 4, "steps": 250, "lr": 1.25e-05, "grad_norm": 1.5746196508407593, "loss_scale": 4.0, "retries": 0, "seconds": 4.1, "max_memory_gb": 24.6, "pair": 1.0951945185661316, "sign": 0.03809042857028544, "keep": 0.8673102994653163, "leak": 6.120599465475607e-05, "replacement": 1.5479102432727814} +{"step": 5, "steps": 250, "lr": 1.5625e-05, "grad_norm": 0.6794475317001343, "loss_scale": 4.0, "retries": 0, "seconds": 5.1, "max_memory_gb": 24.6, "pair": 1.097258299589157, "sign": 0.11683737323619425, "keep": 0.023240134498337284, "leak": 1.527200950590668e-05, "replacement": 1.1673009097576141} +{"step": 6, "steps": 250, "lr": 1.8750000000000002e-05, "grad_norm": 0.9000606536865234, "loss_scale": 4.0, "retries": 0, "seconds": 6.1, "max_memory_gb": 24.6, "pair": 1.0976634919643402, "sign": 0.19926619064062834, "keep": 0.6332892416394316, "leak": 4.5745800861141106e-05, "replacement": 1.513952612876892} +{"step": 7, "steps": 250, "lr": 2.1875e-05, "grad_norm": 1.8879647254943848, "loss_scale": 4.0, "retries": 0, "seconds": 7.0, "max_memory_gb": 24.6, "pair": 1.0944688618183136, "sign": 0.08967955759726465, "keep": 1.524530688933737, "leak": 5.184311477535175e-05, "replacement": 1.9015869200229645} +{"step": 8, "steps": 250, "lr": 2.5e-05, "grad_norm": 2.045976161956787, "loss_scale": 4.0, "retries": 0, "seconds": 8.0, "max_memory_gb": 24.64, "pair": 1.0944971144199371, "sign": 0.20976255694404244, "keep": 1.4593652412295341, "leak": 0.00013179336474422598, "replacement": 1.929093986749649} +{"step": 9, "steps": 250, "lr": 2.8125000000000003e-05, "grad_norm": 3.6866044998168945, "loss_scale": 4.0, "retries": 0, "seconds": 9.0, "max_memory_gb": 24.64, "pair": 1.100559264421463, "sign": 0.5766326216980815, "keep": 1.3853222876787186, "leak": 7.79694858010771e-07, "replacement": 2.0815370678901672} +{"step": 10, "steps": 250, "lr": 3.125e-05, "grad_norm": 0.8040004968643188, "loss_scale": 4.0, "retries": 0, "seconds": 10.0, "max_memory_gb": 24.64, "pair": 1.092244178056717, "sign": 0.08064881293103099, "keep": 0.09780377201968804, "leak": 0.00015258977146004327, "replacement": 1.1815086007118225} +{"step": 11, "steps": 250, "lr": 3.4375e-05, "grad_norm": 1.7379133701324463, "loss_scale": 4.0, "retries": 0, "seconds": 11.0, "max_memory_gb": 24.64, "pair": 1.085358202457428, "sign": 0.20401584031060338, "keep": 1.3154912420432083, "leak": 9.137156212091213e-05, "replacement": 1.8451346158981323} +{"step": 12, "steps": 250, "lr": 3.7500000000000003e-05, "grad_norm": 1.3809114694595337, "loss_scale": 4.0, "retries": 0, "seconds": 12.0, "max_memory_gb": 24.64, "pair": 1.0844092965126038, "sign": 0.07545977435074747, "keep": 0.9349518977105618, "leak": 0.0001359367652327137, "replacement": 1.5896491706371307} +{"step": 13, "steps": 250, "lr": 4.0625000000000005e-05, "grad_norm": 1.308713674545288, "loss_scale": 4.0, "retries": 0, "seconds": 13.0, "max_memory_gb": 24.64, "pair": 1.0782186090946198, "sign": 0.057775964960455894, "keep": 0.8533943975780858, "leak": 0.0004578279385896167, "replacement": 1.5339182913303375} +{"step": 14, "steps": 250, "lr": 4.375e-05, "grad_norm": 1.943528413772583, "loss_scale": 4.0, "retries": 0, "seconds": 13.9, "max_memory_gb": 24.64, "pair": 1.0801407396793365, "sign": 0.26642381865531206, "keep": 1.6105240881443024, "leak": 0.0007258682817337103, "replacement": 2.0187961757183075} +{"step": 15, "steps": 250, "lr": 4.6875e-05, "grad_norm": 1.9395681619644165, "loss_scale": 4.0, "retries": 0, "seconds": 14.9, "max_memory_gb": 24.64, "pair": 1.069372683763504, "sign": 0.19765539490617812, "keep": 0.993158270164713, "leak": 0.002023095468757674, "replacement": 1.6652851998806} +{"step": 16, "steps": 250, "lr": 5e-05, "grad_norm": 1.1997666358947754, "loss_scale": 4.0, "retries": 0, "seconds": 15.9, "max_memory_gb": 24.64, "pair": 1.0717414915561676, "sign": 0.3019342829938978, "keep": 0.3747153547301423, "leak": 0.0006249765574466437, "replacement": 1.4102225601673126} +{"step": 17, "steps": 250, "lr": 5e-05, "grad_norm": 2.976576566696167, "loss_scale": 4.0, "retries": 0, "seconds": 16.9, "max_memory_gb": 24.64, "pair": 1.0833787322044373, "sign": 0.7484222706407309, "keep": 0.831612793088425, "leak": 0.0014533174035022967, "replacement": 1.8737595081329346} +{"step": 18, "steps": 250, "lr": 4.999797225177623e-05, "grad_norm": 1.921366572380066, "loss_scale": 4.0, "retries": 0, "seconds": 17.9, "max_memory_gb": 24.66, "pair": 1.0613307058811188, "sign": 0.18896084697917104, "keep": 1.056324827921344, "leak": 0.0017677610530881793, "replacement": 1.6844155490398407} +{"step": 19, "steps": 250, "lr": 4.999188937259492e-05, "grad_norm": 3.1113827228546143, "loss_scale": 4.0, "retries": 0, "seconds": 18.9, "max_memory_gb": 24.66, "pair": 1.0263586938381195, "sign": 0.03659565723501146, "keep": 1.3573853373527527, "leak": 0.01062368307611905, "replacement": 1.7260051369667053} +{"step": 20, "steps": 250, "lr": 4.998175245886032e-05, "grad_norm": 1.1257145404815674, "loss_scale": 4.0, "retries": 0, "seconds": 19.9, "max_memory_gb": 24.66, "pair": 1.0584298968315125, "sign": 0.03150372998788953, "keep": 0.003740171483514132, "leak": 0.0019348568766872631, "replacement": 1.0765355229377747} +{"step": 21, "steps": 250, "lr": 4.996756333769319e-05, "grad_norm": 2.07155442237854, "loss_scale": 4.0, "retries": 0, "seconds": 20.9, "max_memory_gb": 24.66, "pair": 1.0229685604572296, "sign": 0.04505528323352337, "keep": 0.5889426099347475, "leak": 0.008522697258740664, "replacement": 1.3420981466770172} +{"step": 22, "steps": 250, "lr": 4.994932456660157e-05, "grad_norm": 2.077216625213623, "loss_scale": 4.0, "retries": 0, "seconds": 21.9, "max_memory_gb": 24.66, "pair": 1.0253704190254211, "sign": 0.0668171988800168, "keep": 0.7205783311801497, "leak": 0.004258246503013652, "replacement": 1.420132726430893} +{"step": 23, "steps": 250, "lr": 4.992703943301973e-05, "grad_norm": 2.3394546508789062, "loss_scale": 4.0, "retries": 0, "seconds": 22.9, "max_memory_gb": 24.66, "pair": 1.0084541141986847, "sign": 0.0357923770789057, "keep": 1.2176667890962563, "leak": 0.004282996931578964, "replacement": 1.6362545192241669} +{"step": 24, "steps": 250, "lr": 4.9900711953715675e-05, "grad_norm": 2.5804288387298584, "loss_scale": 4.0, "retries": 0, "seconds": 23.8, "max_memory_gb": 24.66, "pair": 0.9505388140678406, "sign": 0.016616527223959565, "keep": 0.5954832471907139, "leak": 0.012381731648929417, "replacement": 1.2596841752529144} +{"step": 25, "steps": 250, "lr": 4.987034687406713e-05, "grad_norm": 2.387911796569824, "loss_scale": 4.0, "retries": 0, "seconds": 24.8, "max_memory_gb": 24.66, "pair": 0.9822705686092377, "sign": 0.11136306449770927, "keep": 1.347866703494219, "leak": 0.013590652961283922, "replacement": 1.7152830958366394} +{"step": 26, "steps": 250, "lr": 4.9835949667206215e-05, "grad_norm": 2.6690642833709717, "loss_scale": 4.0, "retries": 0, "seconds": 25.8, "max_memory_gb": 24.66, "pair": 0.9420401453971863, "sign": 0.06373409216757864, "keep": 0.6415662467479706, "leak": 0.02197742486896459, "replacement": 1.3001846671104431} +{"step": 27, "steps": 250, "lr": 4.9797526533032945e-05, "grad_norm": 2.5473341941833496, "loss_scale": 4.0, "retries": 0, "seconds": 26.8, "max_memory_gb": 24.66, "pair": 0.9418675452470779, "sign": 0.14459697343409061, "keep": 0.791927744721761, "leak": 0.022170746000483632, "replacement": 1.4156726151704788} +{"step": 28, "steps": 250, "lr": 4.9755084397097716e-05, "grad_norm": 1.9068999290466309, "loss_scale": 4.0, "retries": 0, "seconds": 27.8, "max_memory_gb": 24.66, "pair": 0.9878570735454559, "sign": 0.42101318994536996, "keep": 1.409144975244999, "leak": 0.027624098816886544, "replacement": 1.9098421037197113} +{"step": 29, "steps": 250, "lr": 4.970863090935304e-05, "grad_norm": 2.3723690509796143, "loss_scale": 4.0, "retries": 0, "seconds": 28.8, "max_memory_gb": 24.66, "pair": 1.0086399167776108, "sign": 0.735750287771225, "keep": 0.5285235373303294, "leak": 0.008001487716683187, "replacement": 1.6427772343158722} +{"step": 30, "steps": 250, "lr": 4.965817444277468e-05, "grad_norm": 2.865506887435913, "loss_scale": 4.0, "retries": 0, "seconds": 29.8, "max_memory_gb": 24.66, "pair": 0.8498270213603973, "sign": 0.011454866966232657, "keep": 0.6413502853247337, "leak": 0.04144103452563286, "replacement": 1.1865897923707962} +{"step": 31, "steps": 250, "lr": 4.9603724091852456e-05, "grad_norm": 3.16275954246521, "loss_scale": 4.0, "retries": 0, "seconds": 30.8, "max_memory_gb": 24.66, "pair": 0.9178417921066284, "sign": 0.21522133611142635, "keep": 1.3149241246283054, "leak": 0.040764375356957316, "replacement": 1.6931056380271912} +{"step": 32, "steps": 250, "lr": 4.9545289670951025e-05, "grad_norm": 3.521577835083008, "loss_scale": 4.0, "retries": 0, "seconds": 31.8, "max_memory_gb": 24.66, "pair": 0.8373127579689026, "sign": 0.11897367518395185, "keep": 0.920768640935421, "leak": 0.10850454494357109, "replacement": 1.3843100368976593} +{"step": 33, "steps": 250, "lr": 4.948288171254089e-05, "grad_norm": 1.975196361541748, "loss_scale": 4.0, "retries": 0, "seconds": 32.8, "max_memory_gb": 24.66, "pair": 0.9441412687301636, "sign": 0.1999318622983992, "keep": 0.5188692957162857, "leak": 0.03205039264867082, "replacement": 1.311554491519928} +{"step": 34, "steps": 250, "lr": 4.941651146529998e-05, "grad_norm": 2.288963794708252, "loss_scale": 4.0, "retries": 0, "seconds": 33.8, "max_memory_gb": 24.66, "pair": 0.858988881111145, "sign": 0.0324381769169122, "keep": 0.30815399949642597, "leak": 0.09200108703225851, "replacement": 1.0522852540016174} +{"step": 35, "steps": 250, "lr": 4.9346190892086174e-05, "grad_norm": 2.4940593242645264, "loss_scale": 4.0, "retries": 0, "seconds": 34.8, "max_memory_gb": 24.66, "pair": 0.8584348857402802, "sign": 0.06355391722172499, "keep": 0.4384362563723698, "leak": 0.05551717570051551, "replacement": 1.1233092546463013} +{"step": 36, "steps": 250, "lr": 4.927193266778101e-05, "grad_norm": 2.2380595207214355, "loss_scale": 4.0, "retries": 0, "seconds": 35.8, "max_memory_gb": 24.66, "pair": 0.7981594502925873, "sign": 0.0437262614723295, "keep": 0.4316729459242197, "leak": 0.14465833315625787, "replacement": 1.0720236748456955} +{"step": 37, "steps": 250, "lr": 4.919375017700515e-05, "grad_norm": 2.0182392597198486, "loss_scale": 4.0, "retries": 0, "seconds": 36.7, "max_memory_gb": 24.66, "pair": 0.8672910183668137, "sign": 0.21615763986483216, "keep": 0.8887733938754536, "leak": 0.10831619708187645, "replacement": 1.4468356370925903} +{"step": 38, "steps": 250, "lr": 4.911165751170588e-05, "grad_norm": 2.4127581119537354, "loss_scale": 4.0, "retries": 0, "seconds": 37.7, "max_memory_gb": 24.66, "pair": 0.7364848554134369, "sign": 0.09165746159851551, "keep": 0.26575225219130516, "leak": 0.19899964146316051, "replacement": 0.9649396389722824} +{"step": 39, "steps": 250, "lr": 4.902566946861708e-05, "grad_norm": 1.644340991973877, "loss_scale": 4.0, "retries": 0, "seconds": 38.7, "max_memory_gb": 24.66, "pair": 0.9048540443181992, "sign": 1.3901672773063183, "keep": 0.3717589436346316, "leak": 0.08477264270186424, "replacement": 1.8070103526115417} +{"step": 40, "steps": 250, "lr": 4.89358015465922e-05, "grad_norm": 2.481245756149292, "loss_scale": 4.0, "retries": 0, "seconds": 39.7, "max_memory_gb": 24.66, "pair": 0.7080312073230743, "sign": 0.0190753978677094, "keep": 0.2544433439907152, "leak": 0.1743155950680375, "replacement": 0.8883694708347321} +{"step": 41, "steps": 250, "lr": 4.884206994381078e-05, "grad_norm": 2.1692371368408203, "loss_scale": 4.0, "retries": 0, "seconds": 40.7, "max_memory_gb": 24.66, "pair": 0.9046269804239273, "sign": 0.8865694310516119, "keep": 1.0307181058451533, "leak": 0.044736487325280905, "replacement": 1.874454841017723} +{"step": 42, "steps": 250, "lr": 4.874449155485866e-05, "grad_norm": 1.7763490676879883, "loss_scale": 4.0, "retries": 0, "seconds": 41.6, "max_memory_gb": 24.66, "pair": 0.6533639580011368, "sign": 0.05498817120678723, "keep": 0.0974711838789517, "leak": 0.24752528965473175, "replacement": 0.7914749383926392} +{"step": 43, "steps": 250, "lr": 4.864308396768294e-05, "grad_norm": 2.024191379547119, "loss_scale": 4.0, "retries": 0, "seconds": 42.6, "max_memory_gb": 24.66, "pair": 0.6400236021727324, "sign": 0.15295484708622098, "keep": 0.31538343837019056, "leak": 0.698502330109477, "replacement": 1.0488183051347733} +{"step": 44, "steps": 250, "lr": 4.853786546042184e-05, "grad_norm": 2.752363681793213, "loss_scale": 4.0, "retries": 0, "seconds": 43.6, "max_memory_gb": 24.66, "pair": 0.6713006719946861, "sign": 0.17985882493667305, "keep": 0.9986480736843077, "leak": 0.2586264666169882, "replacement": 1.3252107203006744} +{"step": 45, "steps": 250, "lr": 4.842885499811012e-05, "grad_norm": 2.703026056289673, "loss_scale": 4.0, "retries": 0, "seconds": 44.6, "max_memory_gb": 24.66, "pair": 0.9554942101240158, "sign": 0.7460218775086105, "keep": 1.9431093638995662, "leak": 0.06456891939160414, "replacement": 2.3162019848823547} +{"step": 46, "steps": 250, "lr": 4.831607222926078e-05, "grad_norm": 2.1827712059020996, "loss_scale": 4.0, "retries": 0, "seconds": 45.6, "max_memory_gb": 24.66, "pair": 0.6985536515712738, "sign": 0.2693364628357813, "keep": 0.5921103209257126, "leak": 0.15887169539928436, "replacement": 1.168994978070259} +{"step": 47, "steps": 250, "lr": 4.8199537482323545e-05, "grad_norm": 1.7091879844665527, "loss_scale": 4.0, "retries": 0, "seconds": 46.7, "max_memory_gb": 24.66, "pair": 0.6582672856748104, "sign": 0.10962767276214436, "keep": 0.10021453895024024, "leak": 0.2652213447727263, "replacement": 0.8294937014579773} +{"step": 48, "steps": 250, "lr": 4.807927176202073e-05, "grad_norm": 1.8036415576934814, "loss_scale": 4.0, "retries": 0, "seconds": 47.7, "max_memory_gb": 24.66, "pair": 0.45369047671556473, "sign": 0.02889237756608054, "keep": 0.32379862444940954, "leak": 0.6704546511173248, "replacement": 0.7976496517658234} +{"step": 49, "steps": 250, "lr": 4.795529674556129e-05, "grad_norm": 1.8968255519866943, "loss_scale": 4.0, "retries": 0, "seconds": 48.8, "max_memory_gb": 24.66, "pair": 0.4546547085046768, "sign": 0.04574205307289958, "keep": 0.1768071377009619, "leak": 0.38017382472753525, "replacement": 0.6609727591276169} +{"step": 50, "steps": 250, "lr": 4.782763477873361e-05, "grad_norm": 1.9136351346969604, "loss_scale": 4.0, "retries": 0, "seconds": 49.8, "max_memory_gb": 24.66, "pair": 0.5578595697879791, "sign": 0.10361083783209324, "keep": 0.43794060830623494, "leak": 0.5914833173155785, "replacement": 0.9765061289072037} +{"step": 51, "steps": 250, "lr": 4.769630887187782e-05, "grad_norm": 2.5538735389709473, "loss_scale": 4.0, "retries": 0, "seconds": 50.9, "max_memory_gb": 24.66, "pair": 0.5149668827652931, "sign": 0.03198935999535024, "keep": 1.1336502969243156, "leak": 0.2812798556078633, "replacement": 1.1681066900491714} +{"step": 52, "steps": 250, "lr": 4.756134269573832e-05, "grad_norm": 1.756796956062317, "loss_scale": 4.0, "retries": 0, "seconds": 51.9, "max_memory_gb": 24.66, "pair": 0.5423272550106049, "sign": 0.26681355526670814, "keep": 0.4081090726540424, "leak": 0.42363209649920464, "replacement": 0.9856965988874435} +{"step": 53, "steps": 250, "lr": 4.7422760577197226e-05, "grad_norm": 3.481926441192627, "loss_scale": 4.0, "retries": 0, "seconds": 52.8, "max_memory_gb": 24.66, "pair": 0.5006474480032921, "sign": 0.0958057168754749, "keep": 0.4421991645358503, "leak": 0.6694219466298819, "replacement": 0.9370054006576538} +{"step": 54, "steps": 250, "lr": 4.728058749488963e-05, "grad_norm": 2.298646926879883, "loss_scale": 4.0, "retries": 0, "seconds": 53.8, "max_memory_gb": 24.66, "pair": 0.5048634707927704, "sign": 0.16976024548057467, "keep": 0.3467544511659071, "leak": 0.7618602174334228, "replacement": 0.9535858482122421} +{"step": 55, "steps": 250, "lr": 4.713484907470133e-05, "grad_norm": 1.108834147453308, "loss_scale": 4.0, "retries": 0, "seconds": 54.8, "max_memory_gb": 24.66, "pair": 0.41196709126234055, "sign": 0.03490909386891872, "keep": 0.0133626511451439, "leak": 0.5329079367220402, "replacement": 0.5693299397826195} +{"step": 56, "steps": 250, "lr": 4.6985571585149876e-05, "grad_norm": 0.8909606337547302, "loss_scale": 4.0, "retries": 0, "seconds": 55.7, "max_memory_gb": 24.66, "pair": 0.3465124163776636, "sign": 0.01532230875454843, "keep": 0.103188322391361, "leak": 0.7490449994802475, "replacement": 0.5930289775133133} +{"step": 57, "steps": 250, "lr": 4.6832781932649884e-05, "grad_norm": 1.3335822820663452, "loss_scale": 4.0, "retries": 0, "seconds": 56.7, "max_memory_gb": 24.66, "pair": 0.4817569926381111, "sign": 0.39958301186561584, "keep": 0.02182653116506117, "leak": 0.8005374024888994, "replacement": 0.8925961330533028} +{"step": 58, "steps": 250, "lr": 4.667650765666325e-05, "grad_norm": 1.002448320388794, "loss_scale": 4.0, "retries": 0, "seconds": 57.7, "max_memory_gb": 24.66, "pair": 0.24914909340441227, "sign": 0.01223599398508668, "keep": 0.07959034198211157, "leak": 1.0670633763074875, "replacement": 0.5618281066417694} +{"step": 59, "steps": 250, "lr": 4.651677692473538e-05, "grad_norm": 1.2724955081939697, "loss_scale": 4.0, "retries": 0, "seconds": 58.7, "max_memory_gb": 24.66, "pair": 0.42641347646713257, "sign": 0.03751878038747236, "keep": 0.03095673782081576, "leak": 0.48933885409496725, "replacement": 0.5829859524965286} +{"step": 60, "steps": 250, "lr": 4.635361852741808e-05, "grad_norm": 2.2295453548431396, "loss_scale": 4.0, "retries": 0, "seconds": 59.7, "max_memory_gb": 24.66, "pair": 0.4791686711832881, "sign": 0.0631140394252725, "keep": 0.10639960586559027, "leak": 1.1441137964029622, "replacement": 0.8499539345502853} +{"step": 61, "steps": 250, "lr": 4.6187061873080264e-05, "grad_norm": 3.6485376358032227, "loss_scale": 4.0, "retries": 0, "seconds": 60.7, "max_memory_gb": 24.66, "pair": 0.437805213034153, "sign": 0.2419138951227069, "keep": 0.08167383540421724, "leak": 1.8879885794594884, "replacement": 1.0715962201356888} +{"step": 62, "steps": 250, "lr": 4.601713698260727e-05, "grad_norm": 1.4375462532043457, "loss_scale": 4.0, "retries": 0, "seconds": 61.6, "max_memory_gb": 24.66, "pair": 0.4585406444966793, "sign": 0.03353908518329263, "keep": 0.07708484054455766, "leak": 1.0326422676444054, "replacement": 0.7720131576061249} +{"step": 63, "steps": 250, "lr": 4.5843874483989746e-05, "grad_norm": 1.1096991300582886, "loss_scale": 4.0, "retries": 0, "seconds": 62.6, "max_memory_gb": 24.66, "pair": 0.3737536892294884, "sign": 0.1423791740089655, "keep": 0.1175845805555582, "leak": 0.23508719308301806, "replacement": 0.5625073462724686} +{"step": 64, "steps": 250, "lr": 4.566730560680316e-05, "grad_norm": 0.6844059824943542, "loss_scale": 4.0, "retries": 0, "seconds": 63.6, "max_memory_gb": 24.66, "pair": 0.2666590139269829, "sign": 0.01634377404116094, "keep": 0.03513277415186167, "leak": 0.7809305414557457, "replacement": 0.4876299351453781} +{"step": 65, "steps": 250, "lr": 4.548746217657878e-05, "grad_norm": 3.0959537029266357, "loss_scale": 4.0, "retries": 0, "seconds": 64.6, "max_memory_gb": 24.66, "pair": 0.40972965210676193, "sign": 0.3660194796975702, "keep": 0.3073175959289074, "leak": 2.0613107439130545, "replacement": 1.261725902557373} +{"step": 66, "steps": 250, "lr": 4.5304376609067396e-05, "grad_norm": 1.402345061302185, "loss_scale": 4.0, "retries": 0, "seconds": 65.6, "max_memory_gb": 24.66, "pair": 0.4641815945506096, "sign": 0.13505766831804067, "keep": 0.3417984559782212, "leak": 0.28400288335978985, "replacement": 0.7736103758215904} +{"step": 67, "steps": 250, "lr": 4.5118081904396504e-05, "grad_norm": 6.1691670417785645, "loss_scale": 4.0, "retries": 0, "seconds": 66.6, "max_memory_gb": 24.66, "pair": 0.21390890888869762, "sign": 0.020173318102024496, "keep": 0.1177414103076444, "leak": 1.1401393413543701, "replacement": 0.5679011195898056} +{"step": 68, "steps": 250, "lr": 4.4928611641122225e-05, "grad_norm": 1.4553927183151245, "loss_scale": 4.0, "retries": 0, "seconds": 67.6, "max_memory_gb": 24.66, "pair": 0.5638898760080338, "sign": 0.07239087345078588, "keep": 0.01944893612380838, "leak": 0.4855845794081688, "replacement": 0.731205940246582} +{"step": 69, "steps": 250, "lr": 4.473599997017701e-05, "grad_norm": 2.3669536113739014, "loss_scale": 4.0, "retries": 0, "seconds": 68.6, "max_memory_gb": 24.66, "pair": 0.5674436669796705, "sign": 0.8514063064940274, "keep": 0.5781269040890038, "leak": 1.006084989523515, "replacement": 1.5337315499782562} +{"step": 70, "steps": 250, "lr": 4.4540281608714053e-05, "grad_norm": 2.122206687927246, "loss_scale": 4.0, "retries": 0, "seconds": 69.6, "max_memory_gb": 24.66, "pair": 0.4220619569532573, "sign": 0.036985050624934956, "keep": 0.03786211875558365, "leak": 1.8019311428070068, "replacement": 0.9099683314561844} +{"step": 71, "steps": 250, "lr": 4.434149183384977e-05, "grad_norm": 1.2520971298217773, "loss_scale": 4.0, "retries": 0, "seconds": 70.5, "max_memory_gb": 24.66, "pair": 0.18568865582346916, "sign": 0.007623986341059208, "keep": 0.16326946272647547, "leak": 0.8296209499239922, "replacement": 0.4785406142473221} +{"step": 72, "steps": 250, "lr": 4.413966647630534e-05, "grad_norm": 1.9414350986480713, "loss_scale": 4.0, "retries": 0, "seconds": 71.5, "max_memory_gb": 24.66, "pair": 0.28387574199587107, "sign": 0.19949716818518937, "keep": 0.10217262338846922, "leak": 1.4600576683878899, "replacement": 0.7997250258922577} +{"step": 73, "steps": 250, "lr": 4.3934841913948323e-05, "grad_norm": 1.6074433326721191, "loss_scale": 4.0, "retries": 0, "seconds": 72.5, "max_memory_gb": 24.66, "pair": 0.5129876472055912, "sign": 0.13282848987728357, "keep": 0.5463389169781294, "leak": 0.3680935800075531, "replacement": 0.9445948228240013} +{"step": 74, "steps": 250, "lr": 4.372705506523583e-05, "grad_norm": 1.9558775424957275, "loss_scale": 4.0, "retries": 0, "seconds": 73.5, "max_memory_gb": 24.66, "pair": 0.3874411843717098, "sign": 0.25732381641864777, "keep": 0.220347783520765, "leak": 1.4530226937495172, "replacement": 0.9895326644182205} +{"step": 75, "steps": 250, "lr": 4.351634338256017e-05, "grad_norm": 1.7386842966079712, "loss_scale": 4.0, "retries": 0, "seconds": 74.5, "max_memory_gb": 24.66, "pair": 0.7576713971793652, "sign": 0.524545130552724, "keep": 0.8487801133305766, "leak": 0.24184099296689965, "replacement": 1.504794254899025} +{"step": 76, "steps": 250, "lr": 4.330274484549825e-05, "grad_norm": 1.0755178928375244, "loss_scale": 4.0, "retries": 0, "seconds": 75.5, "max_memory_gb": 24.66, "pair": 0.34665507078170776, "sign": 0.10477638116572052, "keep": 0.025347669317852706, "leak": 0.7690584026277065, "replacement": 0.6039817184209824} +{"step": 77, "steps": 250, "lr": 4.308629795396599e-05, "grad_norm": 2.1013667583465576, "loss_scale": 4.0, "retries": 0, "seconds": 76.4, "max_memory_gb": 24.66, "pair": 0.21090046875178814, "sign": 0.00877962401136756, "keep": 0.008565520895899681, "leak": 1.6523834317922592, "replacement": 0.6326688826084137} +{"step": 78, "steps": 250, "lr": 4.286704172127894e-05, "grad_norm": 2.1413016319274902, "loss_scale": 4.0, "retries": 0, "seconds": 77.4, "max_memory_gb": 24.66, "pair": 0.4950392171740532, "sign": 0.7009562803432345, "keep": 0.0021382997510954738, "leak": 0.4661720935255289, "replacement": 0.9631294906139374} +{"step": 79, "steps": 250, "lr": 4.264501566712033e-05, "grad_norm": 1.385023593902588, "loss_scale": 4.0, "retries": 0, "seconds": 78.4, "max_memory_gb": 24.66, "pair": 0.5097723230719566, "sign": 0.03510914300568402, "keep": 0.1283182966021741, "leak": 0.45784442126750946, "replacement": 0.7059471607208252} +{"step": 80, "steps": 250, "lr": 4.2420259810417894e-05, "grad_norm": 1.0412802696228027, "loss_scale": 4.0, "retries": 0, "seconds": 79.3, "max_memory_gb": 24.66, "pair": 0.5504274368286133, "sign": 0.10364799248054624, "keep": 0.047161212577293554, "leak": 0.6270403228700161, "replacement": 0.7825921177864075} +{"step": 81, "steps": 250, "lr": 4.219281466213066e-05, "grad_norm": 0.9949823021888733, "loss_scale": 4.0, "retries": 0, "seconds": 80.3, "max_memory_gb": 24.66, "pair": 0.31056103855371475, "sign": 0.06323168851668015, "keep": 0.18131924141198397, "leak": 1.0818393509835005, "replacement": 0.703296348452568} +{"step": 82, "steps": 250, "lr": 4.1962721217947135e-05, "grad_norm": 2.794419050216675, "loss_scale": 4.0, "retries": 0, "seconds": 81.3, "max_memory_gb": 24.66, "pair": 0.2753683142364025, "sign": 1.0118075896752998, "keep": 0.09059798822272569, "leak": 1.768195554614067, "replacement": 1.268620029091835} +{"step": 83, "steps": 250, "lr": 4.173002095089599e-05, "grad_norm": 0.9002965688705444, "loss_scale": 4.0, "retries": 0, "seconds": 82.3, "max_memory_gb": 24.66, "pair": 0.33017314970493317, "sign": 0.03234747762326151, "keep": 0.3640815284452401, "leak": 0.8088610209524632, "replacement": 0.7306028828024864} +{"step": 84, "steps": 250, "lr": 4.149475580387084e-05, "grad_norm": 1.1176789999008179, "loss_scale": 4.0, "retries": 0, "seconds": 83.3, "max_memory_gb": 24.66, "pair": 0.2998217474669218, "sign": 0.030811138567514718, "keep": 0.048419292419453086, "leak": 0.9583147317171097, "replacement": 0.5790156498551369} +{"step": 85, "steps": 250, "lr": 4.1256968182070275e-05, "grad_norm": 4.599983215332031, "loss_scale": 4.0, "retries": 0, "seconds": 84.3, "max_memory_gb": 24.66, "pair": 0.3382658585906029, "sign": 0.06270264345221221, "keep": 0.0768634229204963, "leak": 0.7978003211319447, "replacement": 0.6074989587068558} +{"step": 86, "steps": 250, "lr": 4.101670094535453e-05, "grad_norm": 2.0579488277435303, "loss_scale": 4.0, "retries": 0, "seconds": 85.2, "max_memory_gb": 24.66, "pair": 0.3559749610722065, "sign": 0.3135653231292963, "keep": 0.39646017155610025, "leak": 0.2673886101692915, "replacement": 0.7778348848223686} +{"step": 87, "steps": 250, "lr": 4.077399740052027e-05, "grad_norm": 1.648735761642456, "loss_scale": 4.0, "retries": 0, "seconds": 86.2, "max_memory_gb": 24.66, "pair": 0.4878396540880203, "sign": 0.09880043694283813, "keep": 0.015030964277684689, "leak": 0.4194616936329112, "replacement": 0.649620771408081} +{"step": 88, "steps": 250, "lr": 4.0528901293494775e-05, "grad_norm": 3.226145029067993, "loss_scale": 4.0, "retries": 0, "seconds": 87.2, "max_memory_gb": 24.66, "pair": 0.4518614485859871, "sign": 0.2189642246812582, "keep": 0.028454314916530166, "leak": 0.5987624377012253, "replacement": 0.7252613455057144} +{"step": 89, "steps": 250, "lr": 4.0281456801451004e-05, "grad_norm": 1.6483436822891235, "loss_scale": 4.0, "retries": 0, "seconds": 88.1, "max_memory_gb": 24.66, "pair": 0.5994590073823929, "sign": 0.1955470060929656, "keep": 0.018005205628810472, "leak": 0.11220102908555418, "replacement": 0.7342853993177414} +{"step": 90, "steps": 250, "lr": 4.0031708524844884e-05, "grad_norm": 1.990323781967163, "loss_scale": 4.0, "retries": 0, "seconds": 89.1, "max_memory_gb": 24.66, "pair": 0.5011741667985916, "sign": 0.133425181847997, "keep": 0.06518608616897836, "leak": 1.0415531545877457, "replacement": 0.8608680963516235} +{"step": 91, "steps": 250, "lr": 3.977970147937635e-05, "grad_norm": 2.691159963607788, "loss_scale": 4.0, "retries": 0, "seconds": 90.1, "max_memory_gb": 24.66, "pair": 0.5535736158490181, "sign": 0.22438134881667793, "keep": 0.19863681607239414, "leak": 0.2605085727991536, "replacement": 0.8302098512649536} +{"step": 92, "steps": 250, "lr": 3.952548108787553e-05, "grad_norm": 3.736649513244629, "loss_scale": 4.0, "retries": 0, "seconds": 91.1, "max_memory_gb": 24.66, "pair": 0.5214594006538391, "sign": 0.7962845609290525, "keep": 1.3222555919705883, "leak": 0.6176423830911517, "replacement": 1.7351400628685951} +{"step": 93, "steps": 250, "lr": 3.9269093172115496e-05, "grad_norm": 1.5101782083511353, "loss_scale": 4.0, "retries": 0, "seconds": 92.1, "max_memory_gb": 24.66, "pair": 0.38712257519364357, "sign": 0.04169383563566953, "keep": 0.17010751937050372, "leak": 0.4959554527304135, "replacement": 0.6170121282339096} +{"step": 94, "steps": 250, "lr": 3.901058394455323e-05, "grad_norm": 2.7410333156585693, "loss_scale": 4.0, "retries": 0, "seconds": 93.1, "max_memory_gb": 24.66, "pair": 0.3056482672691345, "sign": 0.03770825994433835, "keep": 0.811612605350092, "leak": 1.1731159687042236, "replacement": 1.0235877260565758} +{"step": 95, "steps": 250, "lr": 3.875e-05, "grad_norm": 2.2016639709472656, "loss_scale": 4.0, "retries": 0, "seconds": 94.0, "max_memory_gb": 24.66, "pair": 0.4971481002867222, "sign": 0.313059814972803, "keep": 1.5178969078697264, "leak": 0.6209772012662143, "replacement": 1.5678707957267761} +{"step": 96, "steps": 250, "lr": 3.8487388307223005e-05, "grad_norm": 1.5700916051864624, "loss_scale": 4.0, "retries": 0, "seconds": 95.0, "max_memory_gb": 24.66, "pair": 0.41828399524092674, "sign": 0.12330188206396997, "keep": 1.0207501631230116, "leak": 0.5149951588828117, "replacement": 1.1190588399767876} +{"step": 97, "steps": 250, "lr": 3.822279620047943e-05, "grad_norm": 1.4360700845718384, "loss_scale": 4.0, "retries": 0, "seconds": 96.0, "max_memory_gb": 24.66, "pair": 0.3381325751543045, "sign": 0.01881279144436121, "keep": 0.06589419210649794, "leak": 0.8009067960083485, "replacement": 0.5807127729058266} +{"step": 98, "steps": 250, "lr": 3.795627137098479e-05, "grad_norm": 1.1904908418655396, "loss_scale": 4.0, "retries": 0, "seconds": 97.0, "max_memory_gb": 24.66, "pair": 0.4598841778934002, "sign": 0.17253416345920414, "keep": 0.02691756257399902, "leak": 0.7387669384479523, "replacement": 0.7443018034100533} +{"step": 99, "steps": 250, "lr": 3.7687861858316844e-05, "grad_norm": 2.367474317550659, "loss_scale": 4.0, "retries": 0, "seconds": 98.0, "max_memory_gb": 24.66, "pair": 0.6849384009838104, "sign": 0.7379139505792409, "keep": 0.039800777860818926, "leak": 0.21708634495735168, "replacement": 1.128067396581173} +{"step": 100, "steps": 250, "lr": 3.741761604175668e-05, "grad_norm": 3.215153217315674, "loss_scale": 4.0, "retries": 0, "seconds": 99.0, "max_memory_gb": 24.66, "pair": 0.5360939688980579, "sign": 0.9248852665768936, "keep": 0.5062148390570655, "leak": 0.8768063690513372, "replacement": 1.4708456173539162} +{"step": 101, "steps": 250, "lr": 3.714558263156872e-05, "grad_norm": 4.146965980529785, "loss_scale": 4.0, "retries": 0, "seconds": 100.0, "max_memory_gb": 24.66, "pair": 0.18346290895715356, "sign": 0.010901302623096853, "keep": 0.04563795323338127, "leak": 3.3531845584511757, "replacement": 1.0500286892056465} +{"step": 102, "steps": 250, "lr": 3.6871810660220904e-05, "grad_norm": 1.3754444122314453, "loss_scale": 4.0, "retries": 0, "seconds": 100.9, "max_memory_gb": 24.66, "pair": 0.2928492836654186, "sign": 0.0260408038739115, "keep": 0.07221900322474539, "leak": 0.34172313765157014, "replacement": 0.42740996554493904} +{"step": 103, "steps": 250, "lr": 3.659634947354686e-05, "grad_norm": 0.8876057863235474, "loss_scale": 4.0, "retries": 0, "seconds": 101.9, "max_memory_gb": 24.66, "pair": 0.3244105502963066, "sign": 0.03569386212620884, "keep": 0.02410866920126864, "leak": 0.6111603602766991, "replacement": 0.5071019157767296} +{"step": 104, "steps": 250, "lr": 3.631924872185169e-05, "grad_norm": 2.6864821910858154, "loss_scale": 4.0, "retries": 0, "seconds": 102.8, "max_memory_gb": 24.66, "pair": 0.5094882622361183, "sign": 0.4579635856789537, "keep": 0.004726395145553397, "leak": 0.5714577063918114, "replacement": 0.8836976885795593} +{"step": 105, "steps": 250, "lr": 3.6040558350962674e-05, "grad_norm": 1.3168597221374512, "loss_scale": 4.0, "retries": 0, "seconds": 103.8, "max_memory_gb": 24.66, "pair": 0.2621566243469715, "sign": 0.02317306841723621, "keep": 0.14227793784812093, "leak": 1.2781480550765991, "replacement": 0.6644191294908524} +{"step": 106, "steps": 250, "lr": 3.576032859322687e-05, "grad_norm": 1.2692667245864868, "loss_scale": 4.0, "retries": 0, "seconds": 104.8, "max_memory_gb": 24.66, "pair": 0.33026277646422386, "sign": 0.039815488969907165, "keep": 0.04928145610028878, "leak": 0.7930160553660244, "replacement": 0.573065273463726} +{"step": 107, "steps": 250, "lr": 3.5478609958457056e-05, "grad_norm": 2.390712022781372, "loss_scale": 4.0, "retries": 0, "seconds": 105.8, "max_memory_gb": 24.66, "pair": 0.39411743730306625, "sign": 0.36402428639121354, "keep": 2.362560058766576, "leak": 0.37404681742191315, "replacement": 1.8509212732315063} +{"step": 108, "steps": 250, "lr": 3.519545322482755e-05, "grad_norm": 1.5443437099456787, "loss_scale": 4.0, "retries": 0, "seconds": 106.8, "max_memory_gb": 24.66, "pair": 0.4689839705824852, "sign": 0.28420941391959786, "keep": 0.03314842368126847, "leak": 0.5026003571692854, "replacement": 0.7533130049705505} +{"step": 109, "steps": 250, "lr": 3.491090942972179e-05, "grad_norm": 1.5381715297698975, "loss_scale": 4.0, "retries": 0, "seconds": 107.7, "max_memory_gb": 24.66, "pair": 0.37417547032237053, "sign": 0.08339214499574155, "keep": 0.07928919931873679, "leak": 0.5927829146385193, "replacement": 0.6037118583917618} +{"step": 110, "steps": 250, "lr": 3.4625029860533135e-05, "grad_norm": 7.4922308921813965, "loss_scale": 4.0, "retries": 0, "seconds": 108.7, "max_memory_gb": 24.66, "pair": 0.39331139624118805, "sign": 0.24781543284188956, "keep": 0.587587489746511, "leak": 0.6681324290111661, "replacement": 0.9780459553003311} +{"step": 111, "steps": 250, "lr": 3.433786604542056e-05, "grad_norm": 3.350646495819092, "loss_scale": 4.0, "retries": 0, "seconds": 109.7, "max_memory_gb": 24.66, "pair": 0.42238283809274435, "sign": 1.5694062771799508, "keep": 0.0015500471621407996, "leak": 1.6441380381584167, "replacement": 1.6188954263925552} +{"step": 112, "steps": 250, "lr": 3.404946974402104e-05, "grad_norm": 3.925124406814575, "loss_scale": 4.0, "retries": 0, "seconds": 110.7, "max_memory_gb": 24.66, "pair": 0.21175910625606775, "sign": 0.010140179190784693, "keep": 0.002822799331624992, "leak": 2.0168115496635437, "replacement": 0.7224435061216354} +{"step": 113, "steps": 250, "lr": 3.375989293812019e-05, "grad_norm": 1.180708408355713, "loss_scale": 4.0, "retries": 0, "seconds": 111.7, "max_memory_gb": 24.66, "pair": 0.34355632215738297, "sign": 0.043174868798814714, "keep": 0.011776210449170321, "leak": 0.31978112272918224, "replacement": 0.4509771540760994} +{"step": 114, "steps": 250, "lr": 3.3469187822282766e-05, "grad_norm": 1.2147942781448364, "loss_scale": 4.0, "retries": 0, "seconds": 112.7, "max_memory_gb": 24.66, "pair": 0.26552725210785866, "sign": 0.05251597240567207, "keep": 0.03622193168848753, "leak": 1.2170369625091553, "replacement": 0.6141554415225983} +{"step": 115, "steps": 250, "lr": 3.317740679444499e-05, "grad_norm": 1.351759672164917, "loss_scale": 4.0, "retries": 0, "seconds": 113.6, "max_memory_gb": 24.66, "pair": 0.242323974147439, "sign": 0.011544771492481232, "keep": 0.035530563028260076, "leak": 0.7979355305433273, "replacement": 0.46534550935029984} +{"step": 116, "steps": 250, "lr": 3.2884602446470056e-05, "grad_norm": 3.791081428527832, "loss_scale": 4.0, "retries": 0, "seconds": 114.6, "max_memory_gb": 24.66, "pair": 0.5313710719347, "sign": 0.7918064665864222, "keep": 0.996148562396229, "leak": 0.5595980286598206, "replacement": 1.5652481019496918} +{"step": 117, "steps": 250, "lr": 3.259082755466876e-05, "grad_norm": 6.322779655456543, "loss_scale": 4.0, "retries": 0, "seconds": 115.6, "max_memory_gb": 24.68, "pair": 0.3275335542857647, "sign": 0.43122464464977384, "keep": 1.134960435287212, "leak": 1.8033497482538223, "replacement": 1.5614635348320007} +{"step": 118, "steps": 250, "lr": 3.229613507028686e-05, "grad_norm": 1.2590782642364502, "loss_scale": 4.0, "retries": 0, "seconds": 116.6, "max_memory_gb": 24.68, "pair": 0.45944229513406754, "sign": 0.33394566958304495, "keep": 0.4274619113996323, "leak": 0.6744770845398307, "replacement": 1.0087653696537018} +{"step": 119, "steps": 250, "lr": 3.2000578109961004e-05, "grad_norm": 1.2258919477462769, "loss_scale": 4.0, "retries": 0, "seconds": 117.6, "max_memory_gb": 24.68, "pair": 0.45102881640195847, "sign": 0.3040920081548393, "keep": 0.0182412240847043, "leak": 0.4504522867500782, "replacement": 0.7248084992170334} +{"step": 120, "steps": 250, "lr": 3.1704209946144616e-05, "grad_norm": 2.987356662750244, "loss_scale": 4.0, "retries": 0, "seconds": 118.6, "max_memory_gb": 24.68, "pair": 0.5235935673117638, "sign": 0.9000303689390421, "keep": 1.8368452274880838, "leak": 0.37226338870823383, "replacement": 1.9850972592830658} +{"step": 121, "steps": 250, "lr": 3.140708399750594e-05, "grad_norm": 1.6891783475875854, "loss_scale": 4.0, "retries": 0, "seconds": 119.6, "max_memory_gb": 24.68, "pair": 0.5993961170315742, "sign": 0.4473472014069557, "keep": 0.09137027709857648, "leak": 0.34318349370732903, "replacement": 0.9545507207512856} +{"step": 122, "steps": 250, "lr": 3.1109253819299606e-05, "grad_norm": 1.0295237302780151, "loss_scale": 4.0, "retries": 0, "seconds": 120.6, "max_memory_gb": 24.68, "pair": 0.4124405672773719, "sign": 0.10593914374476299, "keep": 0.039784901764505776, "leak": 0.6791188009083271, "replacement": 0.6550822854042053} +{"step": 123, "steps": 250, "lr": 3.081077309371361e-05, "grad_norm": 6.203326225280762, "loss_scale": 4.0, "retries": 0, "seconds": 121.6, "max_memory_gb": 24.68, "pair": 0.4529771786183119, "sign": 1.3413779526017606, "keep": 0.42224278289359063, "leak": 0.7961041256785393, "replacement": 1.5338135808706284} +{"step": 124, "steps": 250, "lr": 3.051169562019339e-05, "grad_norm": 1.7249661684036255, "loss_scale": 4.0, "retries": 0, "seconds": 122.5, "max_memory_gb": 24.68, "pair": 0.6190766207873821, "sign": 0.586926179472357, "keep": 1.6236588235769887, "leak": 0.27785826226387655, "replacement": 1.7938337177038193} +{"step": 125, "steps": 250, "lr": 3.021207530574477e-05, "grad_norm": 2.7465410232543945, "loss_scale": 4.0, "retries": 0, "seconds": 123.5, "max_memory_gb": 24.68, "pair": 0.4314076229929924, "sign": 0.25371776800602674, "keep": 0.5426819508502376, "leak": 0.7791642630490978, "replacement": 1.0243985652923584} +{"step": 126, "steps": 250, "lr": 2.991196615521758e-05, "grad_norm": 2.471228837966919, "loss_scale": 4.0, "retries": 0, "seconds": 124.5, "max_memory_gb": 24.68, "pair": 0.679855614900589, "sign": 0.5176421091891825, "keep": 0.958994921296835, "leak": 0.9730621133930981, "replacement": 1.6614396572113037} +{"step": 127, "steps": 250, "lr": 2.9611422261571536e-05, "grad_norm": 1.1127760410308838, "loss_scale": 4.0, "retries": 0, "seconds": 125.4, "max_memory_gb": 24.68, "pair": 0.30372118204832077, "sign": 0.011344788043061271, "keep": 0.02300305589915297, "leak": 0.5256867595016956, "replacement": 0.4523167908191681} +{"step": 128, "steps": 250, "lr": 2.931049779612634e-05, "grad_norm": 3.1657896041870117, "loss_scale": 4.0, "retries": 0, "seconds": 126.4, "max_memory_gb": 24.68, "pair": 0.5646733194589615, "sign": 0.9768932670122012, "keep": 1.4034409972082358, "leak": 0.339621027931571, "replacement": 1.839745618402958} +{"step": 129, "steps": 250, "lr": 2.9009246998797557e-05, "grad_norm": 1.6516714096069336, "loss_scale": 4.0, "retries": 0, "seconds": 127.4, "max_memory_gb": 24.68, "pair": 0.4304255470633507, "sign": 0.09408192103728652, "keep": 0.15641345363110304, "leak": 0.38970046024769545, "replacement": 0.6530983299016953} +{"step": 130, "steps": 250, "lr": 2.8707724168320238e-05, "grad_norm": 2.420579671859741, "loss_scale": 4.0, "retries": 0, "seconds": 128.3, "max_memory_gb": 24.68, "pair": 0.6045572534203529, "sign": 1.8358284700661898, "keep": 0.06546512347995304, "leak": 0.2914478639140725, "replacement": 1.6280660033226013} +{"step": 131, "steps": 250, "lr": 2.840598365246184e-05, "grad_norm": 3.4105288982391357, "loss_scale": 4.0, "retries": 0, "seconds": 129.3, "max_memory_gb": 24.68, "pair": 0.6524209529161453, "sign": 0.2768017053604126, "keep": 0.2665095748088788, "leak": 0.4002143112011254, "replacement": 1.024130180478096} +{"step": 132, "steps": 250, "lr": 2.8104079838226383e-05, "grad_norm": 1.5018386840820312, "loss_scale": 4.0, "retries": 0, "seconds": 130.3, "max_memory_gb": 24.68, "pair": 0.27325224317610264, "sign": 0.03154206706676632, "keep": 0.4081881655129109, "leak": 0.8655259497463703, "replacement": 0.7094988524913788} +{"step": 133, "steps": 250, "lr": 2.7802067142051446e-05, "grad_norm": 1.5678316354751587, "loss_scale": 4.0, "retries": 0, "seconds": 131.3, "max_memory_gb": 24.68, "pair": 0.5532828718423843, "sign": 0.4709280231036246, "keep": 0.7809847589232959, "leak": 0.24742108257487416, "replacement": 1.2410945035517216} +{"step": 134, "steps": 250, "lr": 2.7500000000000004e-05, "grad_norm": 2.126420497894287, "loss_scale": 4.0, "retries": 0, "seconds": 132.2, "max_memory_gb": 24.68, "pair": 0.40391374565660954, "sign": 0.3420941570948344, "keep": 0.6075877867606323, "leak": 0.18452776968479156, "replacement": 0.9248866587877274} +{"step": 135, "steps": 250, "lr": 2.7197932857948566e-05, "grad_norm": 2.141313314437866, "loss_scale": 4.0, "retries": 0, "seconds": 133.2, "max_memory_gb": 24.68, "pair": 0.2745865248143673, "sign": 0.04515840986277908, "keep": 0.1493565067066811, "leak": 0.8897400200366974, "replacement": 0.5942789614200592} +{"step": 136, "steps": 250, "lr": 2.689592016177363e-05, "grad_norm": 5.485594749450684, "loss_scale": 4.0, "retries": 0, "seconds": 134.1, "max_memory_gb": 24.68, "pair": 0.3970496505498886, "sign": 0.528050844091922, "keep": 0.8786217436099832, "leak": 1.012985222041607, "replacement": 1.3536321893334389} +{"step": 137, "steps": 250, "lr": 2.659401634753817e-05, "grad_norm": 3.6370160579681396, "loss_scale": 4.0, "retries": 0, "seconds": 135.2, "max_memory_gb": 24.68, "pair": 0.23684372566640377, "sign": 0.019436235481407493, "keep": 0.035330946106114425, "leak": 1.8963804300874472, "replacement": 0.738322414457798} +{"step": 138, "steps": 250, "lr": 2.6292275831679764e-05, "grad_norm": 3.2138564586639404, "loss_scale": 4.0, "retries": 0, "seconds": 136.1, "max_memory_gb": 24.68, "pair": 0.35114831291139126, "sign": 0.036172071122564375, "keep": 0.4469276072923094, "leak": 0.4187017120420933, "replacement": 0.6973735764622688} +{"step": 139, "steps": 250, "lr": 2.5990753001202445e-05, "grad_norm": 2.92756986618042, "loss_scale": 4.0, "retries": 0, "seconds": 137.1, "max_memory_gb": 24.68, "pair": 0.40564409643411636, "sign": 0.8267145385907497, "keep": 0.02258693582552951, "leak": 1.3506988361477852, "replacement": 1.167969599366188} +{"step": 140, "steps": 250, "lr": 2.5689502203873672e-05, "grad_norm": 1.2596755027770996, "loss_scale": 4.0, "retries": 0, "seconds": 138.1, "max_memory_gb": 24.68, "pair": 0.22285116836428642, "sign": 0.033618945570196956, "keep": 0.024103640324028675, "leak": 0.9406469240784645, "replacement": 0.48687418550252914} +{"step": 141, "steps": 250, "lr": 2.538857773842847e-05, "grad_norm": 0.9429095983505249, "loss_scale": 4.0, "retries": 0, "seconds": 139.2, "max_memory_gb": 24.68, "pair": 0.3478705547749996, "sign": 0.21518962061963975, "keep": 0.009758755291841226, "leak": 0.8158927707772818, "replacement": 0.6643179208040237} +{"step": 142, "steps": 250, "lr": 2.5088033844782426e-05, "grad_norm": 2.900627851486206, "loss_scale": 4.0, "retries": 0, "seconds": 140.1, "max_memory_gb": 24.68, "pair": 0.3412773013114929, "sign": 0.5030464074807242, "keep": 0.4116782279616018, "leak": 0.7105276361107826, "replacement": 0.9762715101242065} +{"step": 143, "steps": 250, "lr": 2.4787924694255237e-05, "grad_norm": 3.2071032524108887, "loss_scale": 4.0, "retries": 0, "seconds": 141.1, "max_memory_gb": 24.68, "pair": 0.48483793437480927, "sign": 0.16131535754539073, "keep": 1.6388544037181418, "leak": 0.21299783140420914, "replacement": 1.4381722509860992} +{"step": 144, "steps": 250, "lr": 2.4488304379806627e-05, "grad_norm": 3.4348576068878174, "loss_scale": 4.0, "retries": 0, "seconds": 142.1, "max_memory_gb": 24.68, "pair": 0.5065350495278835, "sign": 0.8060922706499696, "keep": 0.1742171193473041, "leak": 0.5699788816273212, "replacement": 1.1391844302415848} +{"step": 145, "steps": 250, "lr": 2.4189226906286388e-05, "grad_norm": 1.0660526752471924, "loss_scale": 4.0, "retries": 0, "seconds": 143.1, "max_memory_gb": 24.68, "pair": 0.31241314485669136, "sign": 0.0787529197987169, "keep": 0.021484358804627846, "leak": 0.7568880170583725, "replacement": 0.5517537891864777} +{"step": 146, "steps": 250, "lr": 2.3890746180700395e-05, "grad_norm": 2.3744094371795654, "loss_scale": 4.0, "retries": 0, "seconds": 144.1, "max_memory_gb": 24.68, "pair": 0.41263275034725666, "sign": 0.14270207902882248, "keep": 0.01848960356437601, "leak": 1.5713757574558258, "replacement": 0.8860725313425064} +{"step": 147, "steps": 250, "lr": 2.359291600249407e-05, "grad_norm": 2.6210482120513916, "loss_scale": 4.0, "retries": 0, "seconds": 145.1, "max_memory_gb": 24.68, "pair": 0.6525727733969688, "sign": 0.8666579904966056, "keep": 1.049173059524037, "leak": 0.6075557470321655, "replacement": 1.7623772099614143} +{"step": 148, "steps": 250, "lr": 2.3295790053855392e-05, "grad_norm": 4.772956848144531, "loss_scale": 4.0, "retries": 0, "seconds": 146.1, "max_memory_gb": 24.68, "pair": 0.43165119737386703, "sign": 0.2871141789946705, "keep": 0.007964545569734582, "leak": 0.778653210029006, "replacement": 0.7738538905978203} +{"step": 149, "steps": 250, "lr": 2.299942189003901e-05, "grad_norm": 1.5993372201919556, "loss_scale": 4.0, "retries": 0, "seconds": 147.0, "max_memory_gb": 24.68, "pair": 0.3110211957246065, "sign": 0.10131788207218051, "keep": 0.04651652264328732, "leak": 0.9160260334610939, "replacement": 0.613944910466671} +{"step": 150, "steps": 250, "lr": 2.2703864929713144e-05, "grad_norm": 4.407200813293457, "loss_scale": 4.0, "retries": 0, "seconds": 148.0, "max_memory_gb": 24.68, "pair": 0.3587878979742527, "sign": 0.14255786139983684, "keep": 0.19251474435441196, "leak": 0.5277333895210177, "replacement": 0.6582575440406799} +{"step": 151, "steps": 250, "lr": 2.2409172445331257e-05, "grad_norm": 1.1575767993927002, "loss_scale": 4.0, "retries": 0, "seconds": 149.0, "max_memory_gb": 24.68, "pair": 0.3693622350692749, "sign": 0.0308405130635947, "keep": 0.017062029532098677, "leak": 0.5678703337907791, "replacement": 0.5352811068296432} +{"step": 152, "steps": 250, "lr": 2.211539755352996e-05, "grad_norm": 1.0181719064712524, "loss_scale": 4.0, "retries": 0, "seconds": 150.0, "max_memory_gb": 24.68, "pair": 0.24353107437491417, "sign": 0.008667381131090224, "keep": 0.04199328434651761, "leak": 0.8922502733767033, "replacement": 0.4919239580631256} +{"step": 153, "steps": 250, "lr": 2.1822593205555016e-05, "grad_norm": 2.3358850479125977, "loss_scale": 4.0, "retries": 0, "seconds": 151.0, "max_memory_gb": 24.68, "pair": 0.2439179103821516, "sign": 0.010678358259610832, "keep": 0.01679060560763901, "leak": 1.1095673143863678, "replacement": 0.5350442230701447} +{"step": 154, "steps": 250, "lr": 2.1530812177717236e-05, "grad_norm": 2.964175224304199, "loss_scale": 4.0, "retries": 0, "seconds": 152.0, "max_memory_gb": 24.68, "pair": 0.18215937726199627, "sign": 0.01367277029203251, "keep": 0.7884106142446399, "leak": 1.1250727325677872, "replacement": 0.8644692525267601} +{"step": 155, "steps": 250, "lr": 2.1240107061879815e-05, "grad_norm": 1.2480281591415405, "loss_scale": 4.0, "retries": 0, "seconds": 153.0, "max_memory_gb": 24.68, "pair": 0.21841161511838436, "sign": 0.010050420998595655, "keep": 0.4437708201029409, "leak": 0.7753098905086517, "replacement": 0.6391497254371643} +{"step": 156, "steps": 250, "lr": 2.095053025597896e-05, "grad_norm": 1.8952786922454834, "loss_scale": 4.0, "retries": 0, "seconds": 154.0, "max_memory_gb": 24.68, "pair": 0.29259344562888145, "sign": 0.006144313374534249, "keep": 0.029084746095350056, "leak": 0.42083728313446045, "replacement": 0.41541730612516403} +{"step": 157, "steps": 250, "lr": 2.066213395457945e-05, "grad_norm": 4.378449440002441, "loss_scale": 4.0, "retries": 0, "seconds": 154.9, "max_memory_gb": 24.68, "pair": 0.33126164227724075, "sign": 1.153975584777072, "keep": 0.13409005863650236, "leak": 3.069933071732521, "replacement": 1.742777705192566} +{"step": 158, "steps": 250, "lr": 2.0374970139466877e-05, "grad_norm": 2.2127127647399902, "loss_scale": 4.0, "retries": 0, "seconds": 155.9, "max_memory_gb": 24.68, "pair": 0.5135011933743954, "sign": 0.41975810602889396, "keep": 0.5128259602856815, "leak": 0.28320625447668135, "replacement": 1.0505947545170784} +{"step": 159, "steps": 250, "lr": 2.0089090570278213e-05, "grad_norm": 1.195952296257019, "loss_scale": 4.0, "retries": 0, "seconds": 156.9, "max_memory_gb": 24.74, "pair": 0.30005433037877083, "sign": 0.017511889338493347, "keep": 0.0034001142885244917, "leak": 0.799357570707798, "replacement": 0.5103497058153152} +{"step": 160, "steps": 250, "lr": 1.9804546775172457e-05, "grad_norm": 3.452138662338257, "loss_scale": 4.0, "retries": 0, "seconds": 157.9, "max_memory_gb": 24.74, "pair": 0.2846377184614539, "sign": 0.20410702704975847, "keep": 0.6169904968701303, "leak": 1.1810213346034288, "replacement": 0.9904417842626572} +{"step": 161, "steps": 250, "lr": 1.9521390041542946e-05, "grad_norm": 1.8808056116104126, "loss_scale": 4.0, "retries": 0, "seconds": 158.8, "max_memory_gb": 24.74, "pair": 0.4421350695192814, "sign": 0.17133392090909183, "keep": 0.24806637929850694, "leak": 0.25223573739640415, "replacement": 0.7148941606283188} +{"step": 162, "steps": 250, "lr": 1.923967140677313e-05, "grad_norm": 1.7440357208251953, "loss_scale": 4.0, "retries": 0, "seconds": 159.8, "max_memory_gb": 24.74, "pair": 0.2578346636146307, "sign": 0.03973015281371772, "keep": 0.23291424014587392, "leak": 0.8154453560709953, "replacement": 0.598018191754818} +{"step": 163, "steps": 250, "lr": 1.8959441649037335e-05, "grad_norm": 5.919051647186279, "loss_scale": 4.0, "retries": 0, "seconds": 160.8, "max_memory_gb": 24.74, "pair": 0.6913284696638584, "sign": 0.3652959226164967, "keep": 0.5121570862829685, "leak": 0.30195926735177636, "replacement": 1.205544799566269} +{"step": 164, "steps": 250, "lr": 1.868075127814832e-05, "grad_norm": 3.13069486618042, "loss_scale": 4.0, "retries": 0, "seconds": 161.8, "max_memory_gb": 24.74, "pair": 0.5458705686032772, "sign": 0.3196977456100285, "keep": 0.25766932958504185, "leak": 0.38526895456016064, "replacement": 0.9308713227510452} +{"step": 165, "steps": 250, "lr": 1.8403650526453143e-05, "grad_norm": 1.197651982307434, "loss_scale": 4.0, "retries": 0, "seconds": 162.8, "max_memory_gb": 24.74, "pair": 0.2818778473883867, "sign": 0.026033568079583347, "keep": 0.0037100082299730275, "leak": 0.42449150048196316, "replacement": 0.4028725177049637} +{"step": 166, "steps": 250, "lr": 1.812818933977912e-05, "grad_norm": 7.111949443817139, "loss_scale": 4.0, "retries": 0, "seconds": 163.8, "max_memory_gb": 24.74, "pair": 0.7320393733680248, "sign": 0.6860975371673703, "keep": 1.051590766030131, "leak": 0.7363628547172993, "replacement": 1.7849742472171783} +{"step": 167, "steps": 250, "lr": 1.785441736843128e-05, "grad_norm": 2.0594003200531006, "loss_scale": 4.0, "retries": 0, "seconds": 164.8, "max_memory_gb": 24.74, "pair": 0.4206007868051529, "sign": 0.08581647975370288, "keep": 0.08115256065502763, "leak": 0.36355146393179893, "replacement": 0.5949731767177582} +{"step": 168, "steps": 250, "lr": 1.7582383958243314e-05, "grad_norm": 0.8969761729240417, "loss_scale": 4.0, "retries": 0, "seconds": 165.7, "max_memory_gb": 24.74, "pair": 0.29294706881046295, "sign": 0.01769259970751591, "keep": 0.08450165075919358, "leak": 0.7336046695709229, "replacement": 0.5274453461170197} +{"step": 169, "steps": 250, "lr": 1.7312138141683164e-05, "grad_norm": 1.9390813112258911, "loss_scale": 4.0, "retries": 0, "seconds": 166.8, "max_memory_gb": 24.74, "pair": 0.44401348009705544, "sign": 0.16360813425853848, "keep": 0.2113099144771695, "leak": 0.3709780741482973, "replacement": 0.7242170348763466} +{"step": 170, "steps": 250, "lr": 1.704372862901521e-05, "grad_norm": 1.8429088592529297, "loss_scale": 4.0, "retries": 0, "seconds": 167.8, "max_memory_gb": 24.74, "pair": 0.3056911528110504, "sign": 0.006597108149435371, "keep": 0.10903658702954999, "leak": 0.8796000331640244, "replacement": 0.583407998085022} +{"step": 171, "steps": 250, "lr": 1.6777203799520573e-05, "grad_norm": 2.103221893310547, "loss_scale": 4.0, "retries": 0, "seconds": 168.8, "max_memory_gb": 24.74, "pair": 0.335673613473773, "sign": 0.18224081593507435, "keep": 0.09951791533808318, "leak": 0.2915319986641407, "replacement": 0.5494359731674194} +{"step": 172, "steps": 250, "lr": 1.6512611692777008e-05, "grad_norm": 0.8600014448165894, "loss_scale": 4.0, "retries": 0, "seconds": 169.8, "max_memory_gb": 24.74, "pair": 0.14692682772874832, "sign": 0.004003676702268422, "keep": 0.028461400303058326, "leak": 0.7182058077305555, "replacement": 0.34271081164479256} +{"step": 173, "steps": 250, "lr": 1.6250000000000005e-05, "grad_norm": 0.925463080406189, "loss_scale": 4.0, "retries": 0, "seconds": 170.7, "max_memory_gb": 24.74, "pair": 0.24233432998880744, "sign": 0.08960155304521322, "keep": 0.12419226591009647, "leak": 0.6210681625525467, "replacement": 0.5044982954859734} +{"step": 174, "steps": 250, "lr": 1.598941605544678e-05, "grad_norm": 1.3093831539154053, "loss_scale": 4.0, "retries": 0, "seconds": 171.7, "max_memory_gb": 24.74, "pair": 0.2097362782806158, "sign": 0.01917744503589347, "keep": 0.0707409664169063, "leak": 0.3770945370197296, "replacement": 0.34896911308169365} +{"step": 175, "steps": 250, "lr": 1.5730906827884506e-05, "grad_norm": 4.3768391609191895, "loss_scale": 4.0, "retries": 0, "seconds": 172.7, "max_memory_gb": 24.74, "pair": 0.38066601008176804, "sign": 0.29151175590232015, "keep": 1.5519672585651278, "leak": 0.362284058064688, "replacement": 1.3929765224456787} +{"step": 176, "steps": 250, "lr": 1.5474518912124476e-05, "grad_norm": 1.190477728843689, "loss_scale": 4.0, "retries": 0, "seconds": 173.7, "max_memory_gb": 24.74, "pair": 0.2009126003831625, "sign": 0.033548048842931166, "keep": 0.018340546783292666, "leak": 0.3932546186260879, "replacement": 0.3251705542206764} +{"step": 177, "steps": 250, "lr": 1.5220298520623646e-05, "grad_norm": 1.9467300176620483, "loss_scale": 4.0, "retries": 0, "seconds": 174.6, "max_memory_gb": 24.74, "pair": 0.3811505474150181, "sign": 0.4797499429550953, "keep": 0.008828286909192684, "leak": 0.3197923987172544, "replacement": 0.7053877487778664} +{"step": 178, "steps": 250, "lr": 1.4968291475155121e-05, "grad_norm": 6.643339157104492, "loss_scale": 4.0, "retries": 0, "seconds": 175.6, "max_memory_gb": 24.74, "pair": 0.323183988686651, "sign": 0.1982637852197513, "keep": 0.7369928188272752, "leak": 1.2258214393659728, "replacement": 1.0972676277160645} +{"step": 179, "steps": 250, "lr": 1.4718543198548998e-05, "grad_norm": 1.8378536701202393, "loss_scale": 4.0, "retries": 0, "seconds": 176.6, "max_memory_gb": 24.74, "pair": 0.6455861814320087, "sign": 0.5480197243159637, "keep": 0.5969122966125724, "leak": 0.17126962868496776, "replacement": 1.2608695775270462} +{"step": 180, "steps": 250, "lr": 1.4471098706505227e-05, "grad_norm": 1.1774799823760986, "loss_scale": 4.0, "retries": 0, "seconds": 177.5, "max_memory_gb": 24.74, "pair": 0.42974692955613136, "sign": 0.04521738493349403, "keep": 0.07555981492623687, "leak": 0.46087250020354986, "replacement": 0.6053536534309387} +{"step": 181, "steps": 250, "lr": 1.4226002599479743e-05, "grad_norm": 3.349534273147583, "loss_scale": 4.0, "retries": 0, "seconds": 178.5, "max_memory_gb": 24.74, "pair": 0.1607363298535347, "sign": 0.014400601154193282, "keep": 0.06311604578513652, "leak": 2.156872808933258, "replacement": 0.7387128621339798} +{"step": 182, "steps": 250, "lr": 1.3983299054645471e-05, "grad_norm": 2.1869895458221436, "loss_scale": 4.0, "retries": 0, "seconds": 179.5, "max_memory_gb": 24.74, "pair": 0.48240770399570465, "sign": 0.30007244797889143, "keep": 0.04563671709911432, "leak": 0.9386080056428909, "replacement": 0.8899142816662788} +{"step": 183, "steps": 250, "lr": 1.3743031817929724e-05, "grad_norm": 1.005779504776001, "loss_scale": 4.0, "retries": 0, "seconds": 180.5, "max_memory_gb": 24.74, "pair": 0.31044916063547134, "sign": 0.08587539324071258, "keep": 0.006576593546924414, "leak": 0.7186725661158562, "replacement": 0.5363432988524437} +{"step": 184, "steps": 250, "lr": 1.3505244196129164e-05, "grad_norm": 1.51499342918396, "loss_scale": 4.0, "retries": 0, "seconds": 181.5, "max_memory_gb": 24.74, "pair": 0.3093450926244259, "sign": 0.009474072372540832, "keep": 0.014866826171783032, "leak": 0.5942841041833162, "replacement": 0.47008658945560455} +{"step": 185, "steps": 250, "lr": 1.3269979049104014e-05, "grad_norm": 1.5799096822738647, "loss_scale": 4.0, "retries": 0, "seconds": 182.5, "max_memory_gb": 24.74, "pair": 0.2782104145735502, "sign": 0.04021154245128855, "keep": 0.10650027316296473, "leak": 0.806831508903997, "replacement": 0.5532741993665695} +{"step": 186, "steps": 250, "lr": 1.3037278782052863e-05, "grad_norm": 2.3505427837371826, "loss_scale": 4.0, "retries": 0, "seconds": 183.5, "max_memory_gb": 24.74, "pair": 0.3946960009634495, "sign": 0.02587278315331787, "keep": 0.9038320262850448, "leak": 0.4328845962882042, "replacement": 0.9677695631980896} +{"step": 187, "steps": 250, "lr": 1.2807185337869343e-05, "grad_norm": 1.2628318071365356, "loss_scale": 4.0, "retries": 0, "seconds": 184.5, "max_memory_gb": 24.74, "pair": 0.27524733915925026, "sign": 0.009416723623871803, "keep": 0.43667835829546675, "leak": 0.7982598748058081, "replacement": 0.6978598535060883} +{"step": 188, "steps": 250, "lr": 1.2579740189582115e-05, "grad_norm": 2.129549503326416, "loss_scale": 4.0, "retries": 0, "seconds": 185.4, "max_memory_gb": 24.74, "pair": 0.6088897809386253, "sign": 0.851779750577407, "keep": 0.8403266891627936, "leak": 0.2092035785317421, "replacement": 1.5072439312934875} +{"step": 189, "steps": 250, "lr": 1.2354984332879668e-05, "grad_norm": 2.501600980758667, "loss_scale": 4.0, "retries": 0, "seconds": 186.4, "max_memory_gb": 24.74, "pair": 0.2646272834390402, "sign": 0.3826061834115535, "keep": 0.9298709113791119, "leak": 1.0335824297235376, "replacement": 1.1792614609003067} +{"step": 190, "steps": 250, "lr": 1.2132958278721064e-05, "grad_norm": 2.083996295928955, "loss_scale": 4.0, "retries": 0, "seconds": 187.4, "max_memory_gb": 24.74, "pair": 0.14123068377375603, "sign": 0.006985537242144346, "keep": 0.010601666159345768, "leak": 1.134407415986061, "replacement": 0.43362613767385483} +{"step": 191, "steps": 250, "lr": 1.1913702046034016e-05, "grad_norm": 11.4720458984375, "loss_scale": 4.0, "retries": 0, "seconds": 188.3, "max_memory_gb": 24.74, "pair": 0.4142060950398445, "sign": 0.3206815130542964, "keep": 0.4568268754044311, "leak": 0.8636348620057106, "replacement": 1.0188689678907394} +{"step": 192, "steps": 250, "lr": 1.1697255154501756e-05, "grad_norm": 1.3229846954345703, "loss_scale": 4.0, "retries": 0, "seconds": 189.3, "max_memory_gb": 24.74, "pair": 0.5774802044034004, "sign": 0.4806948087643832, "keep": 1.0591058093123138, "leak": 0.5161787131801248, "replacement": 1.4764251708984375} +{"step": 193, "steps": 250, "lr": 1.1483656617439842e-05, "grad_norm": 1.1804660558700562, "loss_scale": 4.0, "retries": 0, "seconds": 190.3, "max_memory_gb": 24.74, "pair": 0.2750133126974106, "sign": 0.03187513304874301, "keep": 0.02539148191544882, "leak": 0.7256048172712326, "replacement": 0.4850478321313858} +{"step": 194, "steps": 250, "lr": 1.1272944934764177e-05, "grad_norm": 1.7860713005065918, "loss_scale": 4.0, "retries": 0, "seconds": 191.3, "max_memory_gb": 24.74, "pair": 0.48775939643383026, "sign": 0.3252329397946596, "keep": 0.022208700433111517, "leak": 0.45614401675993577, "replacement": 0.7755162119865417} +{"step": 195, "steps": 250, "lr": 1.1065158086051684e-05, "grad_norm": 3.5462958812713623, "loss_scale": 4.0, "retries": 0, "seconds": 192.3, "max_memory_gb": 24.74, "pair": 0.34055744111537933, "sign": 0.06319637046544813, "keep": 0.009815837467613164, "leak": 1.089040668681264, "replacement": 0.6493237316608429} +{"step": 196, "steps": 250, "lr": 1.086033352369467e-05, "grad_norm": 1.5107567310333252, "loss_scale": 4.0, "retries": 0, "seconds": 193.3, "max_memory_gb": 24.74, "pair": 0.1270954292267561, "sign": 0.007908565632533282, "keep": 0.02212886786890067, "leak": 1.138126090168953, "replacement": 0.4266456700861454} +{"step": 197, "steps": 250, "lr": 1.0658508166150224e-05, "grad_norm": 1.7415881156921387, "loss_scale": 4.0, "retries": 0, "seconds": 194.3, "max_memory_gb": 24.74, "pair": 0.2336271032691002, "sign": 0.03188267751829699, "keep": 0.006007691714330576, "leak": 1.043013647198677, "replacement": 0.5133256986737251} +{"step": 198, "steps": 250, "lr": 1.0459718391285954e-05, "grad_norm": 3.5624310970306396, "loss_scale": 4.0, "retries": 0, "seconds": 195.3, "max_memory_gb": 24.74, "pair": 0.23703095503151417, "sign": 0.08420685806777328, "keep": 0.1411397090087121, "leak": 1.037549201399088, "replacement": 0.6090915352106094} +{"step": 199, "steps": 250, "lr": 1.0264000029822999e-05, "grad_norm": 4.891290664672852, "loss_scale": 4.0, "retries": 0, "seconds": 196.2, "max_memory_gb": 24.74, "pair": 0.17690121615305543, "sign": 0.03408937591302674, "keep": 0.5517067907203455, "leak": 1.89957071444951, "replacement": 0.9446919932961464} +{"step": 200, "steps": 250, "lr": 1.0071388358877775e-05, "grad_norm": 2.9174036979675293, "loss_scale": 4.0, "retries": 0, "seconds": 197.2, "max_memory_gb": 24.74, "pair": 0.4843905568122864, "sign": 0.38059647026238963, "keep": 1.5242545399378287, "leak": 0.8041630471125245, "replacement": 1.6378567814826965} +{"step": 201, "steps": 250, "lr": 9.881918095603501e-06, "grad_norm": 2.371588945388794, "loss_scale": 4.0, "retries": 0, "seconds": 198.2, "max_memory_gb": 24.74, "pair": 0.4327060580253601, "sign": 0.18774729873985052, "keep": 0.16135402317740954, "leak": 0.374745465349406, "replacement": 0.7009431049227715} +{"step": 202, "steps": 250, "lr": 9.695623390932612e-06, "grad_norm": 2.046698570251465, "loss_scale": 4.0, "retries": 0, "seconds": 199.2, "max_memory_gb": 24.74, "pair": 0.3701499179005623, "sign": 0.15056091744918376, "keep": 0.006144828402284475, "leak": 0.7333631739020348, "replacement": 0.631843589246273} +{"step": 203, "steps": 250, "lr": 9.51253782342123e-06, "grad_norm": 1.3998466730117798, "loss_scale": 4.0, "retries": 0, "seconds": 200.1, "max_memory_gb": 24.74, "pair": 0.16567211411893368, "sign": 0.0700190979114268, "keep": 0.0185822104831459, "leak": 0.7682474255561829, "replacement": 0.40203462541103363} +{"step": 204, "steps": 250, "lr": 9.332694393196846e-06, "grad_norm": 1.1246370077133179, "loss_scale": 4.0, "retries": 0, "seconds": 201.1, "max_memory_gb": 24.74, "pair": 0.36870966851711273, "sign": 0.02887213509529829, "keep": 0.04513326473534107, "leak": 0.8907244328584056, "replacement": 0.6283934786915779} +{"step": 205, "steps": 250, "lr": 9.156125516010256e-06, "grad_norm": 1.0618175268173218, "loss_scale": 4.0, "retries": 0, "seconds": 202.1, "max_memory_gb": 24.74, "pair": 0.2240634672343731, "sign": 0.007742583460640162, "keep": 0.002498421774362214, "leak": 0.5581760108470917, "replacement": 0.36872797831892967} +{"step": 206, "steps": 250, "lr": 8.982863017392733e-06, "grad_norm": 1.7629729509353638, "loss_scale": 4.0, "retries": 0, "seconds": 203.1, "max_memory_gb": 24.74, "pair": 0.26132189668715, "sign": 0.04295516636921093, "keep": 0.023978435427125078, "leak": 0.5144080258905888, "replacement": 0.423390693962574} +{"step": 207, "steps": 250, "lr": 8.81293812691974e-06, "grad_norm": 1.0684878826141357, "loss_scale": 4.0, "retries": 0, "seconds": 204.1, "max_memory_gb": 24.74, "pair": 0.4828185476362705, "sign": 0.17439320450648665, "keep": 0.08673075627211801, "leak": 0.2956607041414827, "replacement": 0.6872957237064838} +{"step": 208, "steps": 250, "lr": 8.646381472581927e-06, "grad_norm": 0.9857134222984314, "loss_scale": 4.0, "retries": 0, "seconds": 205.1, "max_memory_gb": 24.74, "pair": 0.14013178646564484, "sign": 0.004056560166645795, "keep": 0.028732059182402736, "leak": 0.6279943622648716, "replacement": 0.31352468580007553} +{"step": 209, "steps": 250, "lr": 8.483223075264622e-06, "grad_norm": 2.091301918029785, "loss_scale": 4.0, "retries": 0, "seconds": 206.1, "max_memory_gb": 24.74, "pair": 0.48070348612964153, "sign": 0.3862815306056291, "keep": 0.2951915383746382, "leak": 0.8960195276886225, "replacement": 1.0454448387026787} +{"step": 210, "steps": 250, "lr": 8.323492343336747e-06, "grad_norm": 1.6871013641357422, "loss_scale": 4.0, "retries": 0, "seconds": 207.0, "max_memory_gb": 24.74, "pair": 0.5287983790040016, "sign": 0.49231750273611397, "keep": 1.2032946277159908, "leak": 0.425346240401268, "replacement": 1.4829409569501877} +{"step": 211, "steps": 250, "lr": 8.167218067350122e-06, "grad_norm": 1.0033769607543945, "loss_scale": 4.0, "retries": 0, "seconds": 208.0, "max_memory_gb": 24.74, "pair": 0.506950818002224, "sign": 0.2647515068529174, "keep": 0.09285984933376312, "leak": 0.5255058240145445, "replacement": 0.8171329498291016} +{"step": 212, "steps": 250, "lr": 8.01442841485013e-06, "grad_norm": 3.4949138164520264, "loss_scale": 4.0, "retries": 0, "seconds": 209.0, "max_memory_gb": 24.74, "pair": 0.5569616360589862, "sign": 0.16003151988843456, "keep": 0.040282426634803414, "leak": 1.1970049431547523, "replacement": 0.9563698470592499} +{"step": 213, "steps": 250, "lr": 7.865150925298678e-06, "grad_norm": 2.9296116828918457, "loss_scale": 4.0, "retries": 0, "seconds": 210.0, "max_memory_gb": 24.74, "pair": 0.4737689904868603, "sign": 0.33381319395266473, "keep": 0.47312572484042903, "leak": 0.49495464749634266, "replacement": 1.000977136194706} +{"step": 214, "steps": 250, "lr": 7.719412505110375e-06, "grad_norm": 1.617417573928833, "loss_scale": 4.0, "retries": 0, "seconds": 211.1, "max_memory_gb": 24.74, "pair": 0.46893009543418884, "sign": 0.38097140251193196, "keep": 1.190083509997521, "leak": 0.07369629386812449, "replacement": 1.2728816121816635} +{"step": 215, "steps": 250, "lr": 7.57723942280278e-06, "grad_norm": 2.5690321922302246, "loss_scale": 4.0, "retries": 0, "seconds": 212.1, "max_memory_gb": 24.74, "pair": 0.7520340923219919, "sign": 0.8124584667384624, "keep": 1.335634930688002, "leak": 0.06989107791201832, "replacement": 1.8435536064207554} +{"step": 216, "steps": 250, "lr": 7.438657304261687e-06, "grad_norm": 1.4995585680007935, "loss_scale": 4.0, "retries": 0, "seconds": 213.1, "max_memory_gb": 24.74, "pair": 0.18960301019251347, "sign": 0.03982136072590947, "keep": 0.021589665633655386, "leak": 0.8255067840218544, "replacement": 0.42668522894382477} +{"step": 217, "steps": 250, "lr": 7.30369112812219e-06, "grad_norm": 1.5815818309783936, "loss_scale": 4.0, "retries": 0, "seconds": 214.1, "max_memory_gb": 24.74, "pair": 0.21437561139464378, "sign": 0.022580822667805478, "keep": 0.000701454293448478, "leak": 0.9977532885968685, "replacement": 0.47545506805181503} +{"step": 218, "steps": 250, "lr": 7.172365221266403e-06, "grad_norm": 4.834105968475342, "loss_scale": 4.0, "retries": 0, "seconds": 215.1, "max_memory_gb": 24.74, "pair": 0.9167595282196999, "sign": 1.685480386018753, "keep": 0.27066416945308447, "leak": 0.20418442549271276, "replacement": 1.9458779394626617} +{"step": 219, "steps": 250, "lr": 7.0447032544387154e-06, "grad_norm": 1.0190497636795044, "loss_scale": 4.0, "retries": 0, "seconds": 216.1, "max_memory_gb": 24.74, "pair": 0.2967536374926567, "sign": 0.08112943809828721, "keep": 0.0055534764906042255, "leak": 0.42936646938323975, "replacement": 0.44743672758340836} +{"step": 220, "steps": 250, "lr": 6.920728237979274e-06, "grad_norm": 1.7751972675323486, "loss_scale": 4.0, "retries": 0, "seconds": 217.2, "max_memory_gb": 24.74, "pair": 0.38288741558790207, "sign": 0.05504863930400461, "keep": 0.06588912999541208, "leak": 0.17596310563385487, "replacement": 0.4873470738530159} +{"step": 221, "steps": 250, "lr": 6.800462517676456e-06, "grad_norm": 0.9419345855712891, "loss_scale": 4.0, "retries": 0, "seconds": 218.1, "max_memory_gb": 24.74, "pair": 0.35291279293596745, "sign": 0.18416035926202312, "keep": 0.008232416468672454, "leak": 0.7376070022583008, "replacement": 0.6335109248757362} +{"step": 222, "steps": 250, "lr": 6.683927770739222e-06, "grad_norm": 1.2969467639923096, "loss_scale": 4.0, "retries": 0, "seconds": 219.1, "max_memory_gb": 24.74, "pair": 0.2707657655701041, "sign": 0.0322465953649953, "keep": 0.05516422617256467, "leak": 0.47926873131655157, "replacement": 0.4342883676290512} +{"step": 223, "steps": 250, "lr": 6.571145001889892e-06, "grad_norm": 4.133554935455322, "loss_scale": 4.0, "retries": 0, "seconds": 220.1, "max_memory_gb": 24.74, "pair": 0.48890840262174606, "sign": 0.06695785454940051, "keep": 0.007497886304008716, "leak": 0.5132066905498505, "replacement": 0.6544379591941833} +{"step": 224, "steps": 250, "lr": 6.46213453957817e-06, "grad_norm": 1.5987920761108398, "loss_scale": 4.0, "retries": 0, "seconds": 221.1, "max_memory_gb": 24.74, "pair": 0.39486249536275864, "sign": 0.04846268240362406, "keep": 0.008707286287972238, "leak": 0.30649596940202173, "replacement": 0.5000714883208275} +{"step": 225, "steps": 250, "lr": 6.356916032317063e-06, "grad_norm": 2.07977032661438, "loss_scale": 4.0, "retries": 0, "seconds": 222.1, "max_memory_gb": 24.74, "pair": 0.24058892577886581, "sign": 0.03514283709228039, "keep": 0.0025526075423840666, "leak": 0.865639753639698, "replacement": 0.4758465774357319} +{"step": 226, "steps": 250, "lr": 6.255508445141349e-06, "grad_norm": 2.5339818000793457, "loss_scale": 4.0, "retries": 0, "seconds": 223.1, "max_memory_gb": 24.74, "pair": 0.2159843109548092, "sign": 0.04118296655360609, "keep": 1.090549261556589, "leak": 0.7499679075554013, "replacement": 0.969342440366745} +{"step": 227, "steps": 250, "lr": 6.157930056189228e-06, "grad_norm": 1.8237663507461548, "loss_scale": 4.0, "retries": 0, "seconds": 224.0, "max_memory_gb": 24.74, "pair": 0.22300690785050392, "sign": 0.016372972051613033, "keep": 0.08424234494077609, "leak": 0.5470219943672419, "replacement": 0.4100700579583645} +{"step": 228, "steps": 250, "lr": 6.064198453407796e-06, "grad_norm": 1.1488839387893677, "loss_scale": 4.0, "retries": 0, "seconds": 225.0, "max_memory_gb": 24.74, "pair": 0.2196543626487255, "sign": 0.05304894014261663, "keep": 0.0013933365426055389, "leak": 1.0136514455080032, "replacement": 0.5002883672714233} +{"step": 229, "steps": 250, "lr": 5.974330531382935e-06, "grad_norm": 1.9709510803222656, "loss_scale": 4.0, "retries": 0, "seconds": 226.0, "max_memory_gb": 24.74, "pair": 0.3890041373670101, "sign": 0.12932838289998472, "keep": 0.12085858813910022, "leak": 1.0145957876520697, "replacement": 0.7677465677261353} +{"step": 230, "steps": 250, "lr": 5.888342488294125e-06, "grad_norm": 18.176668167114258, "loss_scale": 4.0, "retries": 0, "seconds": 227.0, "max_memory_gb": 24.74, "pair": 0.8744138032197952, "sign": 0.5127525217831135, "keep": 0.6344066113888402, "leak": 0.27520023472607136, "replacement": 1.5167934149503708} +{"step": 231, "steps": 250, "lr": 5.806249822994849e-06, "grad_norm": 2.004164695739746, "loss_scale": 4.0, "retries": 0, "seconds": 227.9, "max_memory_gb": 24.74, "pair": 0.6058543808758259, "sign": 0.7472716895863414, "keep": 0.03558024297899465, "leak": 0.5152346715331078, "replacement": 1.12608902156353} +{"step": 232, "steps": 250, "lr": 5.728067332218994e-06, "grad_norm": 3.038922071456909, "loss_scale": 4.0, "retries": 0, "seconds": 228.9, "max_memory_gb": 24.74, "pair": 0.1928755436092615, "sign": 0.023541983246104792, "keep": 0.04706697358051315, "leak": 1.8473247811198235, "replacement": 0.6900112107396126} +{"step": 233, "steps": 250, "lr": 5.65380910791383e-06, "grad_norm": 1.165693998336792, "loss_scale": 4.0, "retries": 0, "seconds": 229.9, "max_memory_gb": 24.74, "pair": 0.24169407598674297, "sign": 0.00825476658064872, "keep": 0.011380698531866074, "leak": 0.8753743432462215, "replacement": 0.4703553840517998} +{"step": 234, "steps": 250, "lr": 5.583488534700022e-06, "grad_norm": 1.892722725868225, "loss_scale": 4.0, "retries": 0, "seconds": 230.8, "max_memory_gb": 24.74, "pair": 0.23441801965236664, "sign": 0.028082188218832016, "keep": 0.14596780325427972, "leak": 0.6646902710199356, "replacement": 0.48761559277772903} +{"step": 235, "steps": 250, "lr": 5.517118287459118e-06, "grad_norm": 8.428054809570312, "loss_scale": 4.0, "retries": 0, "seconds": 231.8, "max_memory_gb": 24.74, "pair": 0.7598515395075083, "sign": 0.47231031709816307, "keep": 0.9348353873810993, "leak": 1.784026812762022, "replacement": 1.9094310104846954} +{"step": 236, "steps": 250, "lr": 5.454710329048978e-06, "grad_norm": 1.3471378087997437, "loss_scale": 4.0, "retries": 0, "seconds": 232.8, "max_memory_gb": 24.74, "pair": 0.33639834355562925, "sign": 0.11192536179441959, "keep": 0.006353531843842575, "leak": 0.151479653082788, "replacement": 0.43340769596397877} +{"step": 237, "steps": 250, "lr": 5.3962759081475436e-06, "grad_norm": 3.346125602722168, "loss_scale": 4.0, "retries": 0, "seconds": 233.8, "max_memory_gb": 24.74, "pair": 0.450339263305068, "sign": 0.46552169462665915, "keep": 0.19974413250020007, "leak": 0.4377942308783531, "replacement": 0.8924207240343094} +{"step": 238, "steps": 250, "lr": 5.341825557225321e-06, "grad_norm": 0.9563127160072327, "loss_scale": 4.0, "retries": 0, "seconds": 234.7, "max_memory_gb": 24.74, "pair": 0.22309228964149952, "sign": 0.01841454254463315, "keep": 0.03445546329021454, "leak": 0.6894334144890308, "replacement": 0.4218856580555439} +{"step": 239, "steps": 250, "lr": 5.291369090646961e-06, "grad_norm": 2.5931529998779297, "loss_scale": 4.0, "retries": 0, "seconds": 235.7, "max_memory_gb": 24.74, "pair": 0.6473049819469452, "sign": 1.4842074606567621, "keep": 0.16969206732755993, "leak": 0.5069979461841285, "replacement": 1.6010043025016785} +{"step": 240, "steps": 250, "lr": 5.244915602902289e-06, "grad_norm": 1.3975422382354736, "loss_scale": 4.0, "retries": 0, "seconds": 236.7, "max_memory_gb": 24.74, "pair": 0.4015575051307678, "sign": 0.3671450054971501, "keep": 0.1927434224053286, "leak": 0.5408448986709118, "replacement": 0.8167129382491112} +{"step": 241, "steps": 250, "lr": 5.20247346696706e-06, "grad_norm": 4.938520908355713, "loss_scale": 4.0, "retries": 0, "seconds": 237.8, "max_memory_gb": 24.74, "pair": 0.8988118954002857, "sign": 0.6350676498841494, "keep": 0.3276834362145564, "leak": 0.35816960463125724, "replacement": 1.4697298556566238} +{"step": 242, "steps": 250, "lr": 5.164050332793787e-06, "grad_norm": 1.6443690061569214, "loss_scale": 4.0, "retries": 0, "seconds": 238.7, "max_memory_gb": 24.74, "pair": 0.44303503073751926, "sign": 0.048998785903677344, "keep": 0.004161012405802467, "leak": 0.615154180675745, "replacement": 0.6234034523367882} +{"step": 243, "steps": 250, "lr": 5.129653125932873e-06, "grad_norm": 0.9997146725654602, "loss_scale": 4.0, "retries": 0, "seconds": 239.8, "max_memory_gb": 24.74, "pair": 0.23803298361599445, "sign": 0.02163693925831467, "keep": 0.049687670019920915, "leak": 0.8157923296093941, "replacement": 0.47764337062835693} +{"step": 244, "steps": 250, "lr": 5.099288046284329e-06, "grad_norm": 2.7651679515838623, "loss_scale": 4.0, "retries": 0, "seconds": 240.7, "max_memory_gb": 24.74, "pair": 0.40322357043623924, "sign": 0.22297428781166673, "keep": 0.052469597198069096, "leak": 1.2197195496410131, "replacement": 0.8458753824234009} +{"step": 245, "steps": 250, "lr": 5.072960566980277e-06, "grad_norm": 3.420501470565796, "loss_scale": 4.0, "retries": 0, "seconds": 241.7, "max_memory_gb": 24.74, "pair": 0.6719750314950943, "sign": 0.3994084168225527, "keep": 0.7294289264827967, "leak": 0.21866723615676165, "replacement": 1.2910604923963547} +{"step": 246, "steps": 250, "lr": 5.050675433398435e-06, "grad_norm": 0.9370740056037903, "loss_scale": 4.0, "retries": 0, "seconds": 242.7, "max_memory_gb": 24.74, "pair": 0.2067292369902134, "sign": 0.006454047746956348, "keep": 0.004825383817660622, "leak": 0.4621390737593174, "replacement": 0.32790372148156166} +{"step": 247, "steps": 250, "lr": 5.032436662306811e-06, "grad_norm": 3.4916539192199707, "loss_scale": 4.0, "retries": 0, "seconds": 243.7, "max_memory_gb": 24.74, "pair": 0.27882567793130875, "sign": 0.026062700781039894, "keep": 0.40578716462414377, "leak": 0.588326059281826, "replacement": 0.6418321058154106} +{"step": 248, "steps": 250, "lr": 5.018247541139684e-06, "grad_norm": 1.0092524290084839, "loss_scale": 4.0, "retries": 0, "seconds": 244.7, "max_memory_gb": 24.74, "pair": 0.2234644591808319, "sign": 0.03656419966137037, "keep": 0.04224802361932234, "leak": 0.6100469517987221, "replacement": 0.4153823032975197} +{"step": 249, "steps": 250, "lr": 5.00811062740508e-06, "grad_norm": 1.568901538848877, "loss_scale": 4.0, "retries": 0, "seconds": 245.7, "max_memory_gb": 24.74, "pair": 0.25128657929599285, "sign": 0.08098955918103456, "keep": 0.04905144526128424, "leak": 0.8649901300668716, "replacement": 0.5325546264648438} +{"step": 250, "steps": 250, "lr": 5.002027748223778e-06, "grad_norm": 2.3615927696228027, "loss_scale": 4.0, "retries": 0, "seconds": 246.7, "max_memory_gb": 24.74, "pair": 0.48619531095027924, "sign": 0.042292216792702675, "keep": 0.12492530727104167, "leak": 0.1866348385810852, "replacement": 0.6164627745747566} diff --git a/runs/single_internvl/replacement/L07_s13/adapter/README.md b/runs/single_internvl/replacement/L07_s13/adapter/README.md new file mode 100644 index 0000000000000000000000000000000000000000..4f9f856cf0e1531b47e25d266833b2977a19d3ac --- /dev/null +++ b/runs/single_internvl/replacement/L07_s13/adapter/README.md @@ -0,0 +1,207 @@ +--- +base_model: /data/OPSD/assets/models/InternVL3_5-8B-HF +library_name: peft +pipeline_tag: text-generation +tags: +- base_model:adapter:/data/OPSD/assets/models/InternVL3_5-8B-HF +- lora +- transformers +--- + +# Model Card for Model ID + + + + + +## Model Details + +### Model Description + + + + + +- **Developed by:** [More Information Needed] +- **Funded by [optional]:** [More Information Needed] +- **Shared by [optional]:** [More Information Needed] +- **Model type:** [More Information Needed] +- **Language(s) (NLP):** [More Information Needed] +- **License:** [More Information Needed] +- **Finetuned from model [optional]:** [More Information Needed] + +### Model Sources [optional] + + + +- **Repository:** [More Information Needed] +- **Paper [optional]:** [More Information Needed] +- **Demo [optional]:** [More Information Needed] + +## Uses + + + +### Direct Use + + + +[More Information Needed] + +### Downstream Use [optional] + + + +[More Information Needed] + +### Out-of-Scope Use + + + +[More Information Needed] + +## Bias, Risks, and Limitations + + + +[More Information Needed] + +### Recommendations + + + +Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. More information needed for further recommendations. + +## How to Get Started with the Model + +Use the code below to get started with the model. + +[More Information Needed] + +## Training Details + +### Training Data + + + +[More Information Needed] + +### Training Procedure + + + +#### Preprocessing [optional] + +[More Information Needed] + + +#### Training Hyperparameters + +- **Training regime:** [More Information Needed] + +#### Speeds, Sizes, Times [optional] + + + +[More Information Needed] + +## Evaluation + + + +### Testing Data, Factors & Metrics + +#### Testing Data + + + +[More Information Needed] + +#### Factors + + + +[More Information Needed] + +#### Metrics + + + +[More Information Needed] + +### Results + +[More Information Needed] + +#### Summary + + + +## Model Examination [optional] + + + +[More Information Needed] + +## Environmental Impact + + + +Carbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700). + +- **Hardware Type:** [More Information Needed] +- **Hours used:** [More Information Needed] +- **Cloud Provider:** [More Information Needed] +- **Compute Region:** [More Information Needed] +- **Carbon Emitted:** [More Information Needed] + +## Technical Specifications [optional] + +### Model Architecture and Objective + +[More Information Needed] + +### Compute Infrastructure + +[More Information Needed] + +#### Hardware + +[More Information Needed] + +#### Software + +[More Information Needed] + +## Citation [optional] + + + +**BibTeX:** + +[More Information Needed] + +**APA:** + +[More Information Needed] + +## Glossary [optional] + + + +[More Information Needed] + +## More Information [optional] + +[More Information Needed] + +## Model Card Authors [optional] + +[More Information Needed] + +## Model Card Contact + +[More Information Needed] +### Framework versions + +- PEFT 0.20.0 \ No newline at end of file diff --git a/runs/single_internvl/replacement/L07_s13/adapter/adapter_config.json b/runs/single_internvl/replacement/L07_s13/adapter/adapter_config.json new file mode 100644 index 0000000000000000000000000000000000000000..0c5d322e71237c11a36a5d37c84ee0f8b5851e4d --- /dev/null +++ b/runs/single_internvl/replacement/L07_s13/adapter/adapter_config.json @@ -0,0 +1,50 @@ +{ + "alora_invocation_tokens": null, + "alpha_pattern": {}, + "arrow_config": null, + "auto_mapping": null, + "base_model_name_or_path": "/data/OPSD/assets/models/InternVL3_5-8B-HF", + "bias": "none", + "corda_config": null, + "ensure_weight_tying": false, + "eva_config": null, + "exclude_modules": null, + "fan_in_fan_out": false, + "inference_mode": true, + "init_lora_weights": true, + "layer_replication": null, + "layers_pattern": null, + "layers_to_transform": null, + "loftq_config": {}, + "lora_alpha": 32, + "lora_bias": false, + "lora_dropout": 0.05, + "lora_ga_config": null, + "megatron_config": null, + "megatron_core": "megatron.core", + "modules_to_save": null, + "monteclora_config": null, + "peft_type": "LORA", + "peft_version": "0.20.0", + "qalora_group_size": 16, + "r": 16, + "rank_pattern": {}, + "revision": null, + "target_modules": [ + "model.language_model.layers.7.mlp.down_proj", + "model.language_model.layers.7.mlp.gate_proj", + "model.language_model.layers.7.self_attn.v_proj", + "model.language_model.layers.7.self_attn.k_proj", + "model.language_model.layers.7.self_attn.o_proj", + "model.language_model.layers.7.self_attn.q_proj", + "model.language_model.layers.7.mlp.up_proj" + ], + "target_parameters": null, + "task_type": "CAUSAL_LM", + "trainable_token_indices": null, + "use_bdlora": null, + "use_dora": false, + "use_qalora": false, + "use_rslora": false, + "velora_config": null +} \ No newline at end of file diff --git a/runs/single_internvl/replacement/L07_s13/config.json b/runs/single_internvl/replacement/L07_s13/config.json new file mode 100644 index 0000000000000000000000000000000000000000..741f91d2782638bccb79d7b03e65a5119b0da06f --- /dev/null +++ b/runs/single_internvl/replacement/L07_s13/config.json @@ -0,0 +1,24 @@ +{ + "behavior": "replacement", + "seed": 13, + "model": "/data/OPSD/assets/models/InternVL3_5-8B-HF", + "backbone": "internvl", + "train": "/data/OPSD/assets/v3/replacement.jsonl", + "items": 1000, + "steps": 250, + "epochs": 1, + "steps_per_epoch": 250, + "groups_per_step": 4, + "lr": 5e-05, + "scheduler": "cosine", + "warmup_steps": 16, + "final_fraction": 0.1, + "weight_decay": 0.0, + "clip": 1.0, + "initial_loss_scale": 4.0, + "adapter": { + "layer": 7, + "rank": 16, + "trainable_parameters": 1212416 + } +} \ No newline at end of file diff --git a/runs/single_internvl/replacement/L07_s13/steps.jsonl b/runs/single_internvl/replacement/L07_s13/steps.jsonl new file mode 100644 index 0000000000000000000000000000000000000000..41e4573b4a752c3a973afbe9f1aa03723785db98 --- /dev/null +++ b/runs/single_internvl/replacement/L07_s13/steps.jsonl @@ -0,0 +1,250 @@ +{"step": 1, "steps": 250, "lr": 3.125e-06, "grad_norm": 2.162787914276123, "loss_scale": 4.0, "retries": 0, "seconds": 1.1, "max_memory_gb": 23.36, "pair": 1.0986123085021973, "sign": 0.7328193015418947, "keep": 1.137254401575774, "leak": 0.0, "replacement": 2.0336491465568542} +{"step": 2, "steps": 250, "lr": 6.25e-06, "grad_norm": 1.7790180444717407, "loss_scale": 4.0, "retries": 0, "seconds": 2.0, "max_memory_gb": 23.36, "pair": 1.0922463238239288, "sign": 0.18002297915518284, "keep": 1.3658381718123564, "leak": 4.985610416952113e-05, "replacement": 1.865189403295517} +{"step": 3, "steps": 250, "lr": 9.375000000000001e-06, "grad_norm": 3.5915815830230713, "loss_scale": 4.0, "retries": 0, "seconds": 3.0, "max_memory_gb": 23.36, "pair": 1.094288557767868, "sign": 0.11583439353853464, "keep": 2.0617590844631195, "leak": 0.00012573591334330558, "replacement": 2.18311670422554} +{"step": 4, "steps": 250, "lr": 1.25e-05, "grad_norm": 2.3662922382354736, "loss_scale": 4.0, "retries": 0, "seconds": 3.9, "max_memory_gb": 23.36, "pair": 1.0974768698215485, "sign": 0.038223347160965204, "keep": 0.8675388980191201, "leak": 4.792941314235577e-05, "replacement": 1.5503699779510498} +{"step": 5, "steps": 250, "lr": 1.5625e-05, "grad_norm": 1.103846788406372, "loss_scale": 4.0, "retries": 0, "seconds": 4.8, "max_memory_gb": 23.36, "pair": 1.0984489619731903, "sign": 0.11669256491586566, "keep": 0.02300260469201021, "leak": 2.9599455274365027e-05, "replacement": 1.1683039367198944} +{"step": 6, "steps": 250, "lr": 1.8750000000000002e-05, "grad_norm": 1.587615966796875, "loss_scale": 4.0, "retries": 0, "seconds": 5.8, "max_memory_gb": 23.36, "pair": 1.0972223281860352, "sign": 0.19902580697089434, "keep": 0.6277192853740416, "leak": 1.729096752001169e-05, "replacement": 1.5105991959571838} +{"step": 7, "steps": 250, "lr": 2.1875e-05, "grad_norm": 2.873260736465454, "loss_scale": 4.0, "retries": 0, "seconds": 6.7, "max_memory_gb": 23.36, "pair": 1.0878406167030334, "sign": 0.08976889122277498, "keep": 1.4890896927136055, "leak": 0.00011780801884242464, "replacement": 1.877299427986145} +{"step": 8, "steps": 250, "lr": 2.5e-05, "grad_norm": 3.2157115936279297, "loss_scale": 4.0, "retries": 0, "seconds": 7.7, "max_memory_gb": 23.41, "pair": 1.0788930356502533, "sign": 0.20915243052877486, "keep": 1.4006801769137383, "leak": 0.00036843359339400195, "replacement": 1.8839014172554016} +{"step": 9, "steps": 250, "lr": 2.8125000000000003e-05, "grad_norm": 6.647846698760986, "loss_scale": 4.0, "retries": 0, "seconds": 8.6, "max_memory_gb": 23.41, "pair": 1.0913479030132294, "sign": 0.574765267316252, "keep": 1.3252605572342873, "leak": 0.0002503874716239807, "replacement": 2.0414233803749084} +{"step": 10, "steps": 250, "lr": 3.125e-05, "grad_norm": 1.4734400510787964, "loss_scale": 4.0, "retries": 0, "seconds": 9.6, "max_memory_gb": 23.41, "pair": 1.0741369724273682, "sign": 0.07994978968054056, "keep": 0.0883315607206896, "leak": 0.0012800896511180326, "replacement": 1.1585976779460907} +{"step": 11, "steps": 250, "lr": 3.4375e-05, "grad_norm": 2.748549222946167, "loss_scale": 4.0, "retries": 0, "seconds": 10.5, "max_memory_gb": 23.41, "pair": 1.0670102834701538, "sign": 0.20375767070800066, "keep": 1.2029244758014102, "leak": 0.0010171913636440877, "replacement": 1.7706057131290436} +{"step": 12, "steps": 250, "lr": 3.7500000000000003e-05, "grad_norm": 2.3269784450531006, "loss_scale": 4.0, "retries": 0, "seconds": 11.5, "max_memory_gb": 23.41, "pair": 1.0501243472099304, "sign": 0.0743319783359766, "keep": 0.8398110465786885, "leak": 0.0016837403527460992, "replacement": 1.50761678814888} +{"step": 13, "steps": 250, "lr": 4.0625000000000005e-05, "grad_norm": 2.2397382259368896, "loss_scale": 4.0, "retries": 0, "seconds": 12.4, "max_memory_gb": 23.41, "pair": 1.0301498472690582, "sign": 0.05607802979648113, "keep": 0.7652640016676742, "leak": 0.004554707178613171, "replacement": 1.4419595897197723} +{"step": 14, "steps": 250, "lr": 4.375e-05, "grad_norm": 3.2864768505096436, "loss_scale": 4.0, "retries": 0, "seconds": 13.3, "max_memory_gb": 23.41, "pair": 1.0286542177200317, "sign": 0.2659809719771147, "keep": 1.37636037170887, "leak": 0.004777612077305093, "replacement": 1.8510193228721619} +{"step": 15, "steps": 250, "lr": 4.6875e-05, "grad_norm": 3.0243775844573975, "loss_scale": 4.0, "retries": 0, "seconds": 14.2, "max_memory_gb": 23.41, "pair": 1.007834941148758, "sign": 0.20131894294172525, "keep": 0.7532704338154872, "leak": 0.015331546775996685, "replacement": 1.488962560892105} +{"step": 16, "steps": 250, "lr": 5e-05, "grad_norm": 2.2034733295440674, "loss_scale": 4.0, "retries": 0, "seconds": 15.2, "max_memory_gb": 23.41, "pair": 1.0088015347719193, "sign": 0.29987702751532197, "keep": 0.28924457012908533, "leak": 0.007505566696636379, "replacement": 1.3052387237548828} +{"step": 17, "steps": 250, "lr": 5e-05, "grad_norm": 1.9154609441757202, "loss_scale": 4.0, "retries": 0, "seconds": 16.1, "max_memory_gb": 23.41, "pair": 1.0093559473752975, "sign": 0.7477152827195823, "keep": 0.7024710410041735, "leak": 0.014680952648632228, "replacement": 1.7381193339824677} +{"step": 18, "steps": 250, "lr": 4.999797225177623e-05, "grad_norm": 3.3781306743621826, "loss_scale": 4.0, "retries": 0, "seconds": 17.1, "max_memory_gb": 23.43, "pair": 0.9422355592250824, "sign": 0.18015331332571805, "keep": 0.668152688012924, "leak": 0.019542515809007455, "replacement": 1.3712742030620575} +{"step": 19, "steps": 250, "lr": 4.999188937259492e-05, "grad_norm": 4.004026412963867, "loss_scale": 4.0, "retries": 0, "seconds": 18.1, "max_memory_gb": 23.43, "pair": 0.8702391088008881, "sign": 0.033793956972658634, "keep": 0.6788946986198425, "leak": 0.15616068337112665, "replacement": 1.265623539686203} +{"step": 20, "steps": 250, "lr": 4.998175245886032e-05, "grad_norm": 2.3777356147766113, "loss_scale": 4.0, "retries": 0, "seconds": 19.0, "max_memory_gb": 23.43, "pair": 0.9306115359067917, "sign": 0.028510298114269972, "keep": 0.003744349884073017, "leak": 0.017706566955894232, "replacement": 0.9511654675006866} +{"step": 21, "steps": 250, "lr": 4.996756333769319e-05, "grad_norm": 3.0729868412017822, "loss_scale": 4.0, "retries": 0, "seconds": 20.0, "max_memory_gb": 23.43, "pair": 0.8329271376132965, "sign": 0.03871762100607157, "keep": 0.19975387393242272, "leak": 0.10052814520895481, "replacement": 0.9772948771715164} +{"step": 22, "steps": 250, "lr": 4.994932456660157e-05, "grad_norm": 2.751854419708252, "loss_scale": 4.0, "retries": 0, "seconds": 20.9, "max_memory_gb": 23.43, "pair": 0.8361071944236755, "sign": 0.06039868900552392, "keep": 0.293519810991711, "leak": 0.06251543760299683, "replacement": 1.0286953300237656} +{"step": 23, "steps": 250, "lr": 4.992703943301973e-05, "grad_norm": 3.700265645980835, "loss_scale": 4.0, "retries": 0, "seconds": 21.9, "max_memory_gb": 23.43, "pair": 0.7676955312490463, "sign": 0.02995861368253827, "keep": 0.8276876430154516, "leak": 0.04329622583463788, "replacement": 1.2073427438735962} +{"step": 24, "steps": 250, "lr": 4.9900711953715675e-05, "grad_norm": 2.9563546180725098, "loss_scale": 4.0, "retries": 0, "seconds": 22.8, "max_memory_gb": 23.43, "pair": 0.5956771075725555, "sign": 0.01229036832228303, "keep": 0.16337388008832932, "leak": 0.20886478945612907, "replacement": 0.7357254475355148} +{"step": 25, "steps": 250, "lr": 4.987034687406713e-05, "grad_norm": 2.662015676498413, "loss_scale": 4.0, "retries": 0, "seconds": 23.8, "max_memory_gb": 23.43, "pair": 0.6511845588684082, "sign": 0.09782146150246263, "keep": 0.8259060814416443, "leak": 0.24672531429678202, "replacement": 1.1747296154499054} +{"step": 26, "steps": 250, "lr": 4.9835949667206215e-05, "grad_norm": 2.1123037338256836, "loss_scale": 4.0, "retries": 0, "seconds": 24.7, "max_memory_gb": 23.43, "pair": 0.5371898710727692, "sign": 0.058424976421520114, "keep": 0.12901876214891672, "leak": 0.3969436576589942, "replacement": 0.7301476746797562} +{"step": 27, "steps": 250, "lr": 4.9797526533032945e-05, "grad_norm": 2.106210947036743, "loss_scale": 4.0, "retries": 0, "seconds": 25.7, "max_memory_gb": 23.43, "pair": 0.5498655512928963, "sign": 0.1407992064487189, "keep": 0.22047155600739643, "leak": 0.416517979465425, "replacement": 0.8346303999423981} +{"step": 28, "steps": 250, "lr": 4.9755084397097716e-05, "grad_norm": 2.9662258625030518, "loss_scale": 4.0, "retries": 0, "seconds": 26.6, "max_memory_gb": 23.43, "pair": 0.7326758652925491, "sign": 0.42340195551514626, "keep": 1.2397014861926436, "leak": 0.3751884298399091, "replacement": 1.658024713397026} +{"step": 29, "steps": 250, "lr": 4.970863090935304e-05, "grad_norm": 2.1697750091552734, "loss_scale": 4.0, "retries": 0, "seconds": 27.5, "max_memory_gb": 23.43, "pair": 0.7785688042640686, "sign": 0.7512690434232354, "keep": 0.08761908952146769, "leak": 0.21835474437102675, "replacement": 1.2526015639305115} +{"step": 30, "steps": 250, "lr": 4.965817444277468e-05, "grad_norm": 1.1618014574050903, "loss_scale": 4.0, "retries": 0, "seconds": 28.5, "max_memory_gb": 23.43, "pair": 0.28561738319694996, "sign": 0.005311981542035937, "keep": 0.15638737786503043, "leak": 0.755240686237812, "replacement": 0.5552772507071495} +{"step": 31, "steps": 250, "lr": 4.9603724091852456e-05, "grad_norm": 1.6228219270706177, "loss_scale": 4.0, "retries": 0, "seconds": 29.4, "max_memory_gb": 23.43, "pair": 0.4778448287397623, "sign": 0.17748612334253266, "keep": 0.5141878556460142, "leak": 0.8905695956200361, "replacement": 1.0463242083787918} +{"step": 32, "steps": 250, "lr": 4.9545289670951025e-05, "grad_norm": 2.9331774711608887, "loss_scale": 4.0, "retries": 0, "seconds": 30.4, "max_memory_gb": 23.43, "pair": 0.31499629095196724, "sign": 0.09403842486790381, "keep": 0.08998937020078301, "leak": 1.8618148267269135, "replacement": 0.8724639117717743} +{"step": 33, "steps": 250, "lr": 4.948288171254089e-05, "grad_norm": 1.8926849365234375, "loss_scale": 4.0, "retries": 0, "seconds": 31.4, "max_memory_gb": 23.43, "pair": 0.6170758083462715, "sign": 0.1742902183905244, "keep": 0.1190455574542284, "leak": 0.2929313015192747, "replacement": 0.8369765281677246} +{"step": 34, "steps": 250, "lr": 4.941651146529998e-05, "grad_norm": 3.1656455993652344, "loss_scale": 4.0, "retries": 0, "seconds": 32.3, "max_memory_gb": 23.43, "pair": 0.30107822455465794, "sign": 0.0176654199603945, "keep": 0.013612460545118665, "leak": 1.4115097373723984, "replacement": 0.6695945858955383} +{"step": 35, "steps": 250, "lr": 4.9346190892086174e-05, "grad_norm": 1.2496098279953003, "loss_scale": 4.0, "retries": 0, "seconds": 33.2, "max_memory_gb": 23.43, "pair": 0.3291480168700218, "sign": 0.040371623588725924, "keep": 0.022725200658896938, "leak": 0.9002835899591446, "replacement": 0.5857673287391663} +{"step": 36, "steps": 250, "lr": 4.927193266778101e-05, "grad_norm": 5.676942348480225, "loss_scale": 4.0, "retries": 0, "seconds": 34.2, "max_memory_gb": 23.43, "pair": 0.2763976417481899, "sign": 0.025842588220257312, "keep": 0.1576052556047216, "leak": 2.4095836244523525, "replacement": 0.9705174565315247} +{"step": 37, "steps": 250, "lr": 4.919375017700515e-05, "grad_norm": 2.6021389961242676, "loss_scale": 4.0, "retries": 0, "seconds": 35.1, "max_memory_gb": 23.43, "pair": 0.4670400246977806, "sign": 0.18383209267631173, "keep": 0.46993492741603404, "leak": 1.3716134848073125, "replacement": 1.1368268877267838} +{"step": 38, "steps": 250, "lr": 4.911165751170588e-05, "grad_norm": 6.188747882843018, "loss_scale": 4.0, "retries": 0, "seconds": 36.0, "max_memory_gb": 23.43, "pair": 0.15302111580967903, "sign": 0.07528647151775658, "keep": 0.031000610906630754, "leak": 2.5751233100891113, "replacement": 0.8499454855918884} +{"step": 39, "steps": 250, "lr": 4.902566946861708e-05, "grad_norm": 2.479248523712158, "loss_scale": 4.0, "retries": 0, "seconds": 37.0, "max_memory_gb": 23.43, "pair": 0.5293263271450996, "sign": 1.3897623284137808, "keep": 0.23529119155500666, "leak": 0.7227742932736874, "replacement": 1.522546611726284} +{"step": 40, "steps": 250, "lr": 4.89358015465922e-05, "grad_norm": 2.251873254776001, "loss_scale": 4.0, "retries": 0, "seconds": 38.0, "max_memory_gb": 23.43, "pair": 0.16977032646536827, "sign": 0.008626544731669128, "keep": 0.014494477523840033, "leak": 1.3805446773767471, "replacement": 0.5264669954776764} +{"step": 41, "steps": 250, "lr": 4.884206994381078e-05, "grad_norm": 3.5696322917938232, "loss_scale": 4.0, "retries": 0, "seconds": 38.9, "max_memory_gb": 23.43, "pair": 0.6543965488672256, "sign": 0.9261564631015062, "keep": 1.0109675847925246, "leak": 0.3158297794871032, "replacement": 1.7019160315394402} +{"step": 42, "steps": 250, "lr": 4.874449155485866e-05, "grad_norm": 3.610121011734009, "loss_scale": 4.0, "retries": 0, "seconds": 39.9, "max_memory_gb": 23.43, "pair": 0.2907885638996959, "sign": 0.035891797160729766, "keep": 0.002432378034427529, "leak": 1.6292379051446915, "replacement": 0.7172601372003555} +{"step": 43, "steps": 250, "lr": 4.864308396768294e-05, "grad_norm": 3.49001407623291, "loss_scale": 4.0, "retries": 0, "seconds": 40.8, "max_memory_gb": 23.43, "pair": 0.3554923189803958, "sign": 0.14650890621123835, "keep": 0.00993974591983715, "leak": 2.0739843398332596, "replacement": 0.9522127285599709} +{"step": 44, "steps": 250, "lr": 4.853786546042184e-05, "grad_norm": 3.0384275913238525, "loss_scale": 4.0, "retries": 0, "seconds": 41.7, "max_memory_gb": 23.43, "pair": 0.2963879741728306, "sign": 0.15193393791560084, "keep": 0.8769943062688981, "leak": 1.1290882974863052, "replacement": 1.0931241810321808} +{"step": 45, "steps": 250, "lr": 4.842885499811012e-05, "grad_norm": 3.707704782485962, "loss_scale": 4.0, "retries": 0, "seconds": 42.7, "max_memory_gb": 23.43, "pair": 0.68573447316885, "sign": 0.7520620529539883, "keep": 1.970680497965077, "leak": 0.3260110808796526, "replacement": 2.1286085695028305} +{"step": 46, "steps": 250, "lr": 4.831607222926078e-05, "grad_norm": 2.617318630218506, "loss_scale": 4.0, "retries": 0, "seconds": 43.6, "max_memory_gb": 23.43, "pair": 0.4986509196460247, "sign": 0.2652284739888273, "keep": 0.5140901901759207, "leak": 0.5712809301912785, "replacement": 1.031130462884903} +{"step": 47, "steps": 250, "lr": 4.8199537482323545e-05, "grad_norm": 1.2739577293395996, "loss_scale": 4.0, "retries": 0, "seconds": 44.5, "max_memory_gb": 23.43, "pair": 0.4119605477899313, "sign": 0.08986009494401515, "keep": 0.015027218016257393, "leak": 0.7356176492758095, "replacement": 0.6483086124062538} +{"step": 48, "steps": 250, "lr": 4.807927176202073e-05, "grad_norm": 2.315520763397217, "loss_scale": 4.0, "retries": 0, "seconds": 45.5, "max_memory_gb": 23.43, "pair": 0.2487061694264412, "sign": 0.02511381806107238, "keep": 0.03875698856427334, "leak": 1.3053460717201233, "replacement": 0.606978066265583} +{"step": 49, "steps": 250, "lr": 4.795529674556129e-05, "grad_norm": 0.8420862555503845, "loss_scale": 4.0, "retries": 0, "seconds": 46.4, "max_memory_gb": 23.43, "pair": 0.2815859131515026, "sign": 0.03262929292395711, "keep": 0.03148282686379389, "leak": 0.7058496326208115, "replacement": 0.49010439217090607} +{"step": 50, "steps": 250, "lr": 4.782763477873361e-05, "grad_norm": 1.4225752353668213, "loss_scale": 4.0, "retries": 0, "seconds": 47.3, "max_memory_gb": 23.43, "pair": 0.3815109059214592, "sign": 0.07673921715468168, "keep": 0.2710065934297745, "leak": 0.7071739435195923, "replacement": 0.7321773022413254} +{"step": 51, "steps": 250, "lr": 4.769630887187782e-05, "grad_norm": 3.346916437149048, "loss_scale": 4.0, "retries": 0, "seconds": 48.3, "max_memory_gb": 23.43, "pair": 0.409505270421505, "sign": 0.028503723617177457, "keep": 1.1723317365176626, "leak": 0.28582810576881457, "replacement": 1.0813799947500229} +{"step": 52, "steps": 250, "lr": 4.756134269573832e-05, "grad_norm": 2.0578203201293945, "loss_scale": 4.0, "retries": 0, "seconds": 49.2, "max_memory_gb": 23.43, "pair": 0.48970944434404373, "sign": 0.26324764895252883, "keep": 0.14227695381850936, "leak": 0.30736272409558296, "replacement": 0.7693124264478683} +{"step": 53, "steps": 250, "lr": 4.7422760577197226e-05, "grad_norm": 2.3647754192352295, "loss_scale": 4.0, "retries": 0, "seconds": 50.1, "max_memory_gb": 23.43, "pair": 0.448251448571682, "sign": 0.08253262069774792, "keep": 0.3392390649532899, "leak": 0.7549437884008512, "replacement": 0.8478732258081436} +{"step": 54, "steps": 250, "lr": 4.728058749488963e-05, "grad_norm": 2.7028205394744873, "loss_scale": 4.0, "retries": 0, "seconds": 51.1, "max_memory_gb": 23.43, "pair": 0.5015356093645096, "sign": 0.14322137634735554, "keep": 0.3269465902412776, "leak": 0.48928632447496057, "replacement": 0.8589412346482277} +{"step": 55, "steps": 250, "lr": 4.713484907470133e-05, "grad_norm": 1.8221986293792725, "loss_scale": 4.0, "retries": 0, "seconds": 52.0, "max_memory_gb": 23.43, "pair": 0.4119367301464081, "sign": 0.03084509412292391, "keep": 0.014848115271888673, "leak": 0.3368956111371517, "replacement": 0.5190072432160378} +{"step": 56, "steps": 250, "lr": 4.6985571585149876e-05, "grad_norm": 1.8124983310699463, "loss_scale": 4.0, "retries": 0, "seconds": 52.9, "max_memory_gb": 23.43, "pair": 0.37251538783311844, "sign": 0.016331946477293968, "keep": 0.02943078288808465, "leak": 0.42488790675997734, "replacement": 0.5016187354922295} +{"step": 57, "steps": 250, "lr": 4.6832781932649884e-05, "grad_norm": 1.9899630546569824, "loss_scale": 4.0, "retries": 0, "seconds": 53.8, "max_memory_gb": 23.43, "pair": 0.46517321467399597, "sign": 0.34611019981093705, "keep": 0.009257287227228517, "leak": 0.4777869489043951, "replacement": 0.7623036950826645} +{"step": 58, "steps": 250, "lr": 4.667650765666325e-05, "grad_norm": 1.1417855024337769, "loss_scale": 4.0, "retries": 0, "seconds": 54.8, "max_memory_gb": 23.43, "pair": 0.18619763106107712, "sign": 0.010609558667056262, "keep": 0.02953772874752758, "leak": 0.8285586461424828, "replacement": 0.4134109392762184} +{"step": 59, "steps": 250, "lr": 4.651677692473538e-05, "grad_norm": 1.8903694152832031, "loss_scale": 4.0, "retries": 0, "seconds": 55.7, "max_memory_gb": 23.43, "pair": 0.4290323406457901, "sign": 0.0412773551652208, "keep": 0.02902667547459714, "leak": 0.46357482990424614, "replacement": 0.5800780802965164} +{"step": 60, "steps": 250, "lr": 4.635361852741808e-05, "grad_norm": 2.6582086086273193, "loss_scale": 4.0, "retries": 0, "seconds": 56.6, "max_memory_gb": 23.43, "pair": 0.45812173560261726, "sign": 0.05972059548366815, "keep": 0.11312351352535188, "leak": 0.4762666691094637, "replacement": 0.6636104583740234} +{"step": 61, "steps": 250, "lr": 4.6187061873080264e-05, "grad_norm": 3.6834070682525635, "loss_scale": 4.0, "retries": 0, "seconds": 57.6, "max_memory_gb": 23.43, "pair": 0.4521036893129349, "sign": 0.25436142343096435, "keep": 0.033708996255882084, "leak": 1.5678822537884116, "replacement": 0.9881094694137573} +{"step": 62, "steps": 250, "lr": 4.601713698260727e-05, "grad_norm": 1.2968741655349731, "loss_scale": 4.0, "retries": 0, "seconds": 58.5, "max_memory_gb": 23.43, "pair": 0.36067812144756317, "sign": 0.02820330735994503, "keep": 0.05889030109392479, "leak": 0.5954962112009525, "replacement": 0.5530989766120911} +{"step": 63, "steps": 250, "lr": 4.5843874483989746e-05, "grad_norm": 2.1204965114593506, "loss_scale": 4.0, "retries": 0, "seconds": 59.4, "max_memory_gb": 23.43, "pair": 0.3926043752580881, "sign": 0.13295923289842904, "keep": 0.052859046030789614, "leak": 0.1303510912002821, "replacement": 0.5181012824177742} +{"step": 64, "steps": 250, "lr": 4.566730560680316e-05, "grad_norm": 1.4228435754776, "loss_scale": 4.0, "retries": 0, "seconds": 60.4, "max_memory_gb": 23.43, "pair": 0.3028241656720638, "sign": 0.01765251182951033, "keep": 0.025521793169900775, "leak": 0.45080004539340734, "replacement": 0.43711134046316147} +{"step": 65, "steps": 250, "lr": 4.548746217657878e-05, "grad_norm": 5.241293430328369, "loss_scale": 4.0, "retries": 0, "seconds": 61.3, "max_memory_gb": 23.43, "pair": 0.39559594728052616, "sign": 0.3629033360630274, "keep": 0.2646809252910316, "leak": 1.8734366185963154, "replacement": 1.1777472794055939} +{"step": 66, "steps": 250, "lr": 4.5304376609067396e-05, "grad_norm": 1.7848939895629883, "loss_scale": 4.0, "retries": 0, "seconds": 62.3, "max_memory_gb": 23.43, "pair": 0.29973768815398216, "sign": 0.12467419283348136, "keep": 0.31803884531835536, "leak": 0.6136035695672035, "replacement": 0.6744951047003269} +{"step": 67, "steps": 250, "lr": 4.5118081904396504e-05, "grad_norm": 2.151949167251587, "loss_scale": 4.0, "retries": 0, "seconds": 63.2, "max_memory_gb": 23.43, "pair": 0.12907516211271286, "sign": 0.016712591401301324, "keep": 0.0192680098261917, "leak": 1.2136421650648117, "replacement": 0.4504760056734085} +{"step": 68, "steps": 250, "lr": 4.4928611641122225e-05, "grad_norm": 2.2074363231658936, "loss_scale": 4.0, "retries": 0, "seconds": 64.2, "max_memory_gb": 23.43, "pair": 0.43772396817803383, "sign": 0.052335010725073516, "keep": 0.01286561247343343, "leak": 0.535466693341732, "replacement": 0.6041909456253052} +{"step": 69, "steps": 250, "lr": 4.473599997017701e-05, "grad_norm": 2.7785136699676514, "loss_scale": 4.0, "retries": 0, "seconds": 65.1, "max_memory_gb": 23.43, "pair": 0.4885138273239136, "sign": 0.8048906794283539, "keep": 0.6379037096630782, "leak": 0.7081329114735126, "replacement": 1.3869442641735077} +{"step": 70, "steps": 250, "lr": 4.4540281608714053e-05, "grad_norm": 3.694175958633423, "loss_scale": 4.0, "retries": 0, "seconds": 66.1, "max_memory_gb": 23.43, "pair": 0.32655699411407113, "sign": 0.03187456013984047, "keep": 0.005331222142558545, "leak": 1.8631578162312508, "replacement": 0.810949333012104} +{"step": 71, "steps": 250, "lr": 4.434149183384977e-05, "grad_norm": 1.3625158071517944, "loss_scale": 4.0, "retries": 0, "seconds": 67.0, "max_memory_gb": 23.43, "pair": 0.14642864279448986, "sign": 0.0073734180186875165, "keep": 0.027461686929655116, "leak": 0.7588595375418663, "replacement": 0.35356108844280243} +{"step": 72, "steps": 250, "lr": 4.413966647630534e-05, "grad_norm": 3.413479804992676, "loss_scale": 4.0, "retries": 0, "seconds": 68.0, "max_memory_gb": 23.43, "pair": 0.25316796265542507, "sign": 0.18421640439191833, "keep": 0.035754885349888355, "leak": 1.474262535572052, "replacement": 0.7317192554473877} +{"step": 73, "steps": 250, "lr": 4.3934841913948323e-05, "grad_norm": 2.161421060562134, "loss_scale": 4.0, "retries": 0, "seconds": 68.9, "max_memory_gb": 23.43, "pair": 0.35940189100801945, "sign": 0.1178611553623341, "keep": 0.5728825208170747, "leak": 0.4275151900947094, "replacement": 0.8116525113582611} +{"step": 74, "steps": 250, "lr": 4.372705506523583e-05, "grad_norm": 3.4018497467041016, "loss_scale": 4.0, "retries": 0, "seconds": 69.8, "max_memory_gb": 23.43, "pair": 0.3509333524852991, "sign": 0.26195360836572945, "keep": 0.14099002725095033, "leak": 1.191484001930803, "replacement": 0.8502761423587799} +{"step": 75, "steps": 250, "lr": 4.351634338256017e-05, "grad_norm": 2.2237906455993652, "loss_scale": 4.0, "retries": 0, "seconds": 70.8, "max_memory_gb": 23.43, "pair": 0.7172423824667931, "sign": 0.513746329001151, "keep": 0.868597692809999, "leak": 0.11448593728709966, "replacement": 1.4370359107851982} +{"step": 76, "steps": 250, "lr": 4.330274484549825e-05, "grad_norm": 1.5104966163635254, "loss_scale": 4.0, "retries": 0, "seconds": 71.8, "max_memory_gb": 23.43, "pair": 0.30708964727818966, "sign": 0.09726376854814589, "keep": 0.014937254760297947, "leak": 0.5388349145650864, "replacement": 0.4978988766670227} +{"step": 77, "steps": 250, "lr": 4.308629795396599e-05, "grad_norm": 2.0999300479888916, "loss_scale": 4.0, "retries": 0, "seconds": 72.7, "max_memory_gb": 23.43, "pair": 0.20639179460704327, "sign": 0.008796520764008164, "keep": 0.004669753419875633, "leak": 1.3043484836816788, "replacement": 0.5392120480537415} +{"step": 78, "steps": 250, "lr": 4.286704172127894e-05, "grad_norm": 2.358626365661621, "loss_scale": 4.0, "retries": 0, "seconds": 73.7, "max_memory_gb": 23.43, "pair": 0.32533910125494003, "sign": 0.520401035901159, "keep": 0.0012965413625352085, "leak": 0.4376417165622115, "replacement": 0.695598304271698} +{"step": 79, "steps": 250, "lr": 4.264501566712033e-05, "grad_norm": 1.6346110105514526, "loss_scale": 4.0, "retries": 0, "seconds": 74.7, "max_memory_gb": 23.43, "pair": 0.34796734899282455, "sign": 0.02356525929644704, "keep": 0.042136901019148354, "leak": 0.6618729513138533, "replacement": 0.5462866723537445} +{"step": 80, "steps": 250, "lr": 4.2420259810417894e-05, "grad_norm": 1.782342553138733, "loss_scale": 4.0, "retries": 0, "seconds": 75.6, "max_memory_gb": 23.43, "pair": 0.4468158558011055, "sign": 0.0988583117723465, "keep": 0.017739387504207116, "leak": 0.7202931568026543, "replacement": 0.6851879879832268} +{"step": 81, "steps": 250, "lr": 4.219281466213066e-05, "grad_norm": 1.3017668724060059, "loss_scale": 4.0, "retries": 0, "seconds": 76.6, "max_memory_gb": 23.43, "pair": 0.2502060066908598, "sign": 0.054528905980987474, "keep": 0.12223856803029776, "leak": 0.716611921787262, "replacement": 0.5177427232265472} +{"step": 82, "steps": 250, "lr": 4.1962721217947135e-05, "grad_norm": 2.902204990386963, "loss_scale": 4.0, "retries": 0, "seconds": 77.5, "max_memory_gb": 23.43, "pair": 0.33269573748111725, "sign": 1.0292077487101778, "keep": 0.031043578288517892, "leak": 1.0742954388260841, "replacement": 1.1313952058553696} +{"step": 83, "steps": 250, "lr": 4.173002095089599e-05, "grad_norm": 2.903637647628784, "loss_scale": 4.0, "retries": 0, "seconds": 78.4, "max_memory_gb": 23.43, "pair": 0.3516880078241229, "sign": 0.03264042836963199, "keep": 0.35726223257370293, "leak": 1.089578090235591, "replacement": 0.8190338760614395} +{"step": 84, "steps": 250, "lr": 4.149475580387084e-05, "grad_norm": 1.5673177242279053, "loss_scale": 4.0, "retries": 0, "seconds": 79.4, "max_memory_gb": 23.43, "pair": 0.27006617560982704, "sign": 0.03097688767593354, "keep": 0.03921565282945494, "leak": 0.5029965043067932, "replacement": 0.4309115707874298} +{"step": 85, "steps": 250, "lr": 4.1256968182070275e-05, "grad_norm": 4.530928611755371, "loss_scale": 4.0, "retries": 0, "seconds": 80.4, "max_memory_gb": 23.43, "pair": 0.17588939517736435, "sign": 0.04075464283232577, "keep": 0.0774557589935796, "leak": 0.8661367371678352, "replacement": 0.45152876526117325} +{"step": 86, "steps": 250, "lr": 4.101670094535453e-05, "grad_norm": 3.05298113822937, "loss_scale": 4.0, "retries": 0, "seconds": 81.3, "max_memory_gb": 23.43, "pair": 0.3456668518483639, "sign": 0.2917229626327753, "keep": 0.3205030568642542, "leak": 0.1716942461207509, "replacement": 0.6947034299373627} +{"step": 87, "steps": 250, "lr": 4.077399740052027e-05, "grad_norm": 1.2957053184509277, "loss_scale": 4.0, "retries": 0, "seconds": 82.2, "max_memory_gb": 23.43, "pair": 0.3130181021988392, "sign": 0.09939421620219946, "keep": 0.009092205567867495, "leak": 0.7789954245090485, "replacement": 0.5620101764798164} +{"step": 88, "steps": 250, "lr": 4.0528901293494775e-05, "grad_norm": 3.497528314590454, "loss_scale": 4.0, "retries": 0, "seconds": 83.2, "max_memory_gb": 23.43, "pair": 0.35885140486061573, "sign": 0.21167799131944776, "keep": 0.04824439866877128, "leak": 0.7624498400837183, "replacement": 0.6794250681996346} +{"step": 89, "steps": 250, "lr": 4.0281456801451004e-05, "grad_norm": 2.654996156692505, "loss_scale": 4.0, "retries": 0, "seconds": 84.1, "max_memory_gb": 23.43, "pair": 0.4575229287147522, "sign": 0.16653813468292356, "keep": 0.013339064675903956, "leak": 0.10041731513047125, "replacement": 0.5725658386945724} +{"step": 90, "steps": 250, "lr": 4.0031708524844884e-05, "grad_norm": 2.147432804107666, "loss_scale": 4.0, "retries": 0, "seconds": 85.0, "max_memory_gb": 23.43, "pair": 0.37607138231396675, "sign": 0.09454014262882993, "keep": 0.03928432319662534, "leak": 0.6235717665404081, "replacement": 0.598876565694809} +{"step": 91, "steps": 250, "lr": 3.977970147937635e-05, "grad_norm": 5.095761775970459, "loss_scale": 4.0, "retries": 0, "seconds": 86.0, "max_memory_gb": 23.43, "pair": 0.5734001323580742, "sign": 0.2656703954562545, "keep": 0.31970192719018087, "leak": 0.2765026115812361, "replacement": 0.9352119863033295} +{"step": 92, "steps": 250, "lr": 3.952548108787553e-05, "grad_norm": 7.415094375610352, "loss_scale": 4.0, "retries": 0, "seconds": 86.9, "max_memory_gb": 23.43, "pair": 0.39771395549178123, "sign": 0.7348297265707515, "keep": 1.1931013155740402, "leak": 0.43798617215361446, "replacement": 1.4711759611964226} +{"step": 93, "steps": 250, "lr": 3.9269093172115496e-05, "grad_norm": 1.9920302629470825, "loss_scale": 4.0, "retries": 0, "seconds": 87.8, "max_memory_gb": 23.43, "pair": 0.2735743448138237, "sign": 0.037121063098311424, "keep": 0.12915157340466976, "leak": 0.4413526132630068, "replacement": 0.4670488312840462} +{"step": 94, "steps": 250, "lr": 3.901058394455323e-05, "grad_norm": 2.4006333351135254, "loss_scale": 4.0, "retries": 0, "seconds": 88.8, "max_memory_gb": 23.43, "pair": 0.23625306226313114, "sign": 0.02991907554678619, "keep": 0.6168351897504181, "leak": 0.7982058972120285, "replacement": 0.7591816633939743} +{"step": 95, "steps": 250, "lr": 3.875e-05, "grad_norm": 3.512723207473755, "loss_scale": 4.0, "retries": 0, "seconds": 89.7, "max_memory_gb": 23.43, "pair": 0.4274404477328062, "sign": 0.3125529931858182, "keep": 1.4341261929366738, "leak": 0.671357563642232, "replacement": 1.4686193689703941} +{"step": 96, "steps": 250, "lr": 3.8487388307223005e-05, "grad_norm": 2.4550607204437256, "loss_scale": 4.0, "retries": 0, "seconds": 90.6, "max_memory_gb": 23.43, "pair": 0.3326185904443264, "sign": 0.10821547533851117, "keep": 1.086013991269283, "leak": 0.6889756117016077, "replacement": 1.1019772440195084} +{"step": 97, "steps": 250, "lr": 3.822279620047943e-05, "grad_norm": 2.5896618366241455, "loss_scale": 4.0, "retries": 0, "seconds": 91.6, "max_memory_gb": 23.43, "pair": 0.22968185041099787, "sign": 0.013142607582267374, "keep": 0.02427556571956302, "leak": 1.0959296207875013, "replacement": 0.5223733335733414} +{"step": 98, "steps": 250, "lr": 3.795627137098479e-05, "grad_norm": 1.6697300672531128, "loss_scale": 4.0, "retries": 0, "seconds": 92.5, "max_memory_gb": 23.43, "pair": 0.37407128885388374, "sign": 0.1285845577949658, "keep": 0.02528164410136924, "leak": 0.6796749122440815, "replacement": 0.6209231317043304} +{"step": 99, "steps": 250, "lr": 3.7687861858316844e-05, "grad_norm": 2.5964155197143555, "loss_scale": 4.0, "retries": 0, "seconds": 93.4, "max_memory_gb": 23.43, "pair": 0.5354307480156422, "sign": 0.6935827269917354, "keep": 0.027473722621152774, "leak": 0.2894129669293761, "replacement": 0.9683122038841248} +{"step": 100, "steps": 250, "lr": 3.741761604175668e-05, "grad_norm": 4.668414115905762, "loss_scale": 4.0, "retries": 0, "seconds": 94.4, "max_memory_gb": 23.43, "pair": 0.6014309339225292, "sign": 0.9477564936969429, "keep": 0.40026007808046415, "leak": 0.8841850294847973, "replacement": 1.4964854270219803} +{"step": 101, "steps": 250, "lr": 3.714558263156872e-05, "grad_norm": 7.433661937713623, "loss_scale": 4.0, "retries": 0, "seconds": 95.3, "max_memory_gb": 23.43, "pair": 0.13615011633373797, "sign": 0.008239906572271138, "keep": 0.02098070773354266, "leak": 3.2854518741369247, "replacement": 0.9721233732998371} +{"step": 102, "steps": 250, "lr": 3.6871810660220904e-05, "grad_norm": 2.2297847270965576, "loss_scale": 4.0, "retries": 0, "seconds": 96.3, "max_memory_gb": 23.43, "pair": 0.22786992136389017, "sign": 0.018042932031676173, "keep": 0.07904569106176496, "leak": 0.3005859726690687, "replacement": 0.35156071931123734} +{"step": 103, "steps": 250, "lr": 3.659634947354686e-05, "grad_norm": 1.1055598258972168, "loss_scale": 4.0, "retries": 0, "seconds": 97.2, "max_memory_gb": 23.43, "pair": 0.18019857071340084, "sign": 0.027340923610609025, "keep": 0.01578252613398945, "leak": 0.7509824056178331, "replacement": 0.38950589299201965} +{"step": 104, "steps": 250, "lr": 3.631924872185169e-05, "grad_norm": 3.2208974361419678, "loss_scale": 4.0, "retries": 0, "seconds": 98.1, "max_memory_gb": 23.43, "pair": 0.2506478186696768, "sign": 0.3167006073636003, "keep": 0.003997353749582544, "leak": 0.3626389019191265, "replacement": 0.5016565173864365} +{"step": 105, "steps": 250, "lr": 3.6040558350962674e-05, "grad_norm": 2.3126211166381836, "loss_scale": 4.0, "retries": 0, "seconds": 99.0, "max_memory_gb": 23.43, "pair": 0.30668284744024277, "sign": 0.0201218135189265, "keep": 0.11390526057220995, "leak": 0.7542151901870966, "replacement": 0.5622501820325851} +{"step": 106, "steps": 250, "lr": 3.576032859322687e-05, "grad_norm": 1.7859752178192139, "loss_scale": 4.0, "retries": 0, "seconds": 99.9, "max_memory_gb": 23.43, "pair": 0.28096496500074863, "sign": 0.0293224579654634, "keep": 0.021498018519196194, "leak": 0.3849252685904503, "replacement": 0.4026065170764923} +{"step": 107, "steps": 250, "lr": 3.5478609958457056e-05, "grad_norm": 4.734681606292725, "loss_scale": 4.0, "retries": 0, "seconds": 100.9, "max_memory_gb": 23.43, "pair": 0.22652654349803925, "sign": 0.2935387345496565, "keep": 1.8784450628500053, "leak": 0.5101410896386369, "replacement": 1.440053716301918} +{"step": 108, "steps": 250, "lr": 3.519545322482755e-05, "grad_norm": 3.18163800239563, "loss_scale": 4.0, "retries": 0, "seconds": 101.8, "max_memory_gb": 23.43, "pair": 0.4376695454120636, "sign": 0.21549892681650817, "keep": 0.027321450122144597, "leak": 0.36452317982912064, "replacement": 0.6502105370163918} +{"step": 109, "steps": 250, "lr": 3.491090942972179e-05, "grad_norm": 3.1277101039886475, "loss_scale": 4.0, "retries": 0, "seconds": 102.7, "max_memory_gb": 23.43, "pair": 0.40523529425263405, "sign": 0.1012512412853539, "keep": 0.07174067513551563, "leak": 0.7364335947204381, "replacement": 0.6758396476507187} +{"step": 110, "steps": 250, "lr": 3.4625029860533135e-05, "grad_norm": 3.2522408962249756, "loss_scale": 4.0, "retries": 0, "seconds": 103.7, "max_memory_gb": 23.43, "pair": 0.32106732577085495, "sign": 0.16699032858014107, "keep": 0.31447117682546377, "leak": 0.4650222547352314, "replacement": 0.6780536398291588} +{"step": 111, "steps": 250, "lr": 3.433786604542056e-05, "grad_norm": 5.52815055847168, "loss_scale": 4.0, "retries": 0, "seconds": 104.6, "max_memory_gb": 23.43, "pair": 0.27034147176891565, "sign": 1.3621001445135335, "keep": 0.0011954406018048758, "leak": 1.106247752904892, "replacement": 1.2285511568188667} +{"step": 112, "steps": 250, "lr": 3.404946974402104e-05, "grad_norm": 7.3009185791015625, "loss_scale": 4.0, "retries": 0, "seconds": 105.5, "max_memory_gb": 23.43, "pair": 0.07970986515283585, "sign": 0.007122495910152793, "keep": 0.0056398618398816325, "leak": 1.9660929664969444, "replacement": 0.5776142999529839} +{"step": 113, "steps": 250, "lr": 3.375989293812019e-05, "grad_norm": 1.5980778932571411, "loss_scale": 4.0, "retries": 0, "seconds": 106.5, "max_memory_gb": 23.43, "pair": 0.22213077545166016, "sign": 0.031544362544082105, "keep": 0.011494553822558373, "leak": 0.39974418841302395, "replacement": 0.3435862921178341} +{"step": 114, "steps": 250, "lr": 3.3469187822282766e-05, "grad_norm": 1.9814022779464722, "loss_scale": 4.0, "retries": 0, "seconds": 107.4, "max_memory_gb": 23.43, "pair": 0.25501564145088196, "sign": 0.04421134071890265, "keep": 0.0260289516299963, "leak": 0.9940837807953358, "replacement": 0.5386567190289497} +{"step": 115, "steps": 250, "lr": 3.317740679444499e-05, "grad_norm": 1.6697343587875366, "loss_scale": 4.0, "retries": 0, "seconds": 108.4, "max_memory_gb": 23.43, "pair": 0.21741814631968737, "sign": 0.010878703207708895, "keep": 0.0111518594349036, "leak": 0.5675025465898216, "replacement": 0.3703090585768223} +{"step": 116, "steps": 250, "lr": 3.2884602446470056e-05, "grad_norm": 7.350651741027832, "loss_scale": 4.0, "retries": 0, "seconds": 109.3, "max_memory_gb": 23.43, "pair": 0.30259391386061907, "sign": 0.7082603675080463, "keep": 0.9193541486056347, "leak": 0.45972348749637604, "replacement": 1.2313321121037006} +{"step": 117, "steps": 250, "lr": 3.259082755466876e-05, "grad_norm": 16.60085678100586, "loss_scale": 4.0, "retries": 0, "seconds": 110.2, "max_memory_gb": 23.45, "pair": 0.1503546901512891, "sign": 0.3555125201528426, "keep": 0.8720492994852975, "leak": 3.3646678179502487, "replacement": 1.6053025126457214} +{"step": 118, "steps": 250, "lr": 3.229613507028686e-05, "grad_norm": 2.134652853012085, "loss_scale": 4.0, "retries": 0, "seconds": 111.2, "max_memory_gb": 23.45, "pair": 0.32402639649808407, "sign": 0.3354846637812443, "keep": 0.4520779922913789, "leak": 0.42334791272878647, "replacement": 0.8236447311937809} +{"step": 119, "steps": 250, "lr": 3.2000578109961004e-05, "grad_norm": 2.811208724975586, "loss_scale": 4.0, "retries": 0, "seconds": 112.1, "max_memory_gb": 23.45, "pair": 0.3613111162558198, "sign": 0.2751817898824811, "keep": 0.019668732022864788, "leak": 0.282186602242291, "replacement": 0.5792830139398575} +{"step": 120, "steps": 250, "lr": 3.1704209946144616e-05, "grad_norm": 4.318571090698242, "loss_scale": 4.0, "retries": 0, "seconds": 113.1, "max_memory_gb": 23.45, "pair": 0.42520109936594963, "sign": 0.855667961994186, "keep": 1.7410708963871002, "leak": 0.18387312203412876, "replacement": 1.7695388048887253} +{"step": 121, "steps": 250, "lr": 3.140708399750594e-05, "grad_norm": 2.505143880844116, "loss_scale": 4.0, "retries": 0, "seconds": 114.0, "max_memory_gb": 23.45, "pair": 0.5442040339112282, "sign": 0.42790249292738736, "keep": 0.0822288778051643, "leak": 0.49186233524233103, "replacement": 0.9222352840006351} +{"step": 122, "steps": 250, "lr": 3.1109253819299606e-05, "grad_norm": 2.0329246520996094, "loss_scale": 4.0, "retries": 0, "seconds": 115.0, "max_memory_gb": 23.45, "pair": 0.33864055993035436, "sign": 0.10072763479547575, "keep": 0.04858303267712927, "leak": 0.6005778685212135, "replacement": 0.5634403377771378} +{"step": 123, "steps": 250, "lr": 3.081077309371361e-05, "grad_norm": 6.014634609222412, "loss_scale": 4.0, "retries": 0, "seconds": 115.9, "max_memory_gb": 23.45, "pair": 0.4239013008773327, "sign": 1.4220025076065212, "keep": 0.19582758843898773, "leak": 0.7380575747811235, "replacement": 1.4173307195305824} +{"step": 124, "steps": 250, "lr": 3.051169562019339e-05, "grad_norm": 2.8137025833129883, "loss_scale": 4.0, "retries": 0, "seconds": 116.9, "max_memory_gb": 23.45, "pair": 0.548218822106719, "sign": 0.5942223013844341, "keep": 1.7841247567994287, "leak": 0.21368369745323434, "replacement": 1.7908133640885353} +{"step": 125, "steps": 250, "lr": 3.021207530574477e-05, "grad_norm": 5.90842342376709, "loss_scale": 4.0, "retries": 0, "seconds": 117.8, "max_memory_gb": 23.45, "pair": 0.3322723659221083, "sign": 0.23151406436227262, "keep": 0.21896703887250624, "leak": 1.68733901809901, "replacement": 0.9793476909399033} +{"step": 126, "steps": 250, "lr": 2.991196615521758e-05, "grad_norm": 6.477382183074951, "loss_scale": 4.0, "retries": 0, "seconds": 118.7, "max_memory_gb": 23.45, "pair": 0.5670872484333813, "sign": 0.5007757761632092, "keep": 0.9433432752266526, "leak": 1.5703907627612352, "replacement": 1.6817444413900375} +{"step": 127, "steps": 250, "lr": 2.9611422261571536e-05, "grad_norm": 2.765192985534668, "loss_scale": 4.0, "retries": 0, "seconds": 119.6, "max_memory_gb": 23.45, "pair": 0.17367257410660386, "sign": 0.00956070596294012, "keep": 0.03973570266725801, "leak": 0.6401233747601509, "replacement": 0.3583516199141741} +{"step": 128, "steps": 250, "lr": 2.931049779612634e-05, "grad_norm": 4.677792072296143, "loss_scale": 4.0, "retries": 0, "seconds": 120.6, "max_memory_gb": 23.45, "pair": 0.3075571581721306, "sign": 0.9384273437899537, "keep": 1.0858632007148117, "leak": 0.38328404165804386, "replacement": 1.4155234899371862} +{"step": 129, "steps": 250, "lr": 2.9009246998797557e-05, "grad_norm": 1.8730922937393188, "loss_scale": 4.0, "retries": 0, "seconds": 121.5, "max_memory_gb": 23.45, "pair": 0.25931581016629934, "sign": 0.07681559619959444, "keep": 0.12511501088738441, "leak": 0.48416105285286903, "replacement": 0.48132139444351196} +{"step": 130, "steps": 250, "lr": 2.8707724168320238e-05, "grad_norm": 3.427035331726074, "loss_scale": 4.0, "retries": 0, "seconds": 122.4, "max_memory_gb": 23.45, "pair": 0.5119465216994286, "sign": 1.6612053886055946, "keep": 0.05248401322751306, "leak": 0.16990245878696442, "replacement": 1.4112668186426163} +{"step": 131, "steps": 250, "lr": 2.840598365246184e-05, "grad_norm": 6.158157825469971, "loss_scale": 4.0, "retries": 0, "seconds": 123.3, "max_memory_gb": 23.45, "pair": 0.689109917730093, "sign": 0.36360359471291304, "keep": 0.48814747351570986, "leak": 0.6117621038574725, "replacement": 1.2679259404540062} +{"step": 132, "steps": 250, "lr": 2.8104079838226383e-05, "grad_norm": 6.903400421142578, "loss_scale": 4.0, "retries": 0, "seconds": 124.3, "max_memory_gb": 23.45, "pair": 0.13943704986013472, "sign": 0.01566219999222085, "keep": 0.40556767904672597, "leak": 1.9597996212542057, "replacement": 0.8400018699467182} +{"step": 133, "steps": 250, "lr": 2.7802067142051446e-05, "grad_norm": 2.5720767974853516, "loss_scale": 4.0, "retries": 0, "seconds": 125.2, "max_memory_gb": 23.45, "pair": 0.515967383980751, "sign": 0.5056824109051377, "keep": 0.8479683444020338, "leak": 0.30907997569465806, "replacement": 1.2700627632439137} +{"step": 134, "steps": 250, "lr": 2.7500000000000004e-05, "grad_norm": 4.986662864685059, "loss_scale": 4.0, "retries": 0, "seconds": 126.1, "max_memory_gb": 23.45, "pair": 0.2928565554320812, "sign": 0.3747846372425556, "keep": 0.681976918869168, "leak": 0.06871272064745426, "replacement": 0.8384155072271824} +{"step": 135, "steps": 250, "lr": 2.7197932857948566e-05, "grad_norm": 7.134878635406494, "loss_scale": 4.0, "retries": 0, "seconds": 127.0, "max_memory_gb": 23.45, "pair": 0.18219815148040652, "sign": 0.024692174440133385, "keep": 0.1272329969506245, "leak": 1.4690100867301226, "replacement": 0.6254132613539696} +{"step": 136, "steps": 250, "lr": 2.689592016177363e-05, "grad_norm": 6.44675350189209, "loss_scale": 4.0, "retries": 0, "seconds": 127.9, "max_memory_gb": 23.45, "pair": 0.25495646707713604, "sign": 0.44931553956121206, "keep": 0.5303056661687151, "leak": 1.098879899829626, "replacement": 1.0194871053099632} +{"step": 137, "steps": 250, "lr": 2.659401634753817e-05, "grad_norm": 5.539999961853027, "loss_scale": 4.0, "retries": 0, "seconds": 128.9, "max_memory_gb": 23.45, "pair": 0.15980367129668593, "sign": 0.016485403262777254, "keep": 0.038304081026581116, "leak": 1.3554214092437178, "replacement": 0.5260537639260292} +{"step": 138, "steps": 250, "lr": 2.6292275831679764e-05, "grad_norm": 3.6907119750976562, "loss_scale": 4.0, "retries": 0, "seconds": 129.8, "max_memory_gb": 23.45, "pair": 0.18131259130313993, "sign": 0.02692589204525575, "keep": 0.1842979658395052, "leak": 0.7901282366365194, "replacement": 0.4844565838575363} +{"step": 139, "steps": 250, "lr": 2.5990753001202445e-05, "grad_norm": 3.255094528198242, "loss_scale": 4.0, "retries": 0, "seconds": 130.8, "max_memory_gb": 23.45, "pair": 0.352696992456913, "sign": 0.686155160976341, "keep": 0.029829924227669835, "leak": 0.8396575525403023, "replacement": 0.9206039160490036} +{"step": 140, "steps": 250, "lr": 2.5689502203873672e-05, "grad_norm": 2.5284955501556396, "loss_scale": 4.0, "retries": 0, "seconds": 131.7, "max_memory_gb": 23.45, "pair": 0.2301942277699709, "sign": 0.03544451810012106, "keep": 0.03009344747988507, "leak": 0.21320971194654703, "replacement": 0.31626565009355545} +{"step": 141, "steps": 250, "lr": 2.538857773842847e-05, "grad_norm": 2.8801279067993164, "loss_scale": 4.0, "retries": 0, "seconds": 132.7, "max_memory_gb": 23.45, "pair": 0.3259907625615597, "sign": 0.21077248570509255, "keep": 0.007722865964751691, "leak": 0.7564506442286074, "replacement": 0.6243511363863945} +{"step": 142, "steps": 250, "lr": 2.5088033844782426e-05, "grad_norm": 5.373021602630615, "loss_scale": 4.0, "retries": 0, "seconds": 133.6, "max_memory_gb": 23.45, "pair": 0.3777002254500985, "sign": 0.5046544776414521, "keep": 0.37911992328008637, "leak": 0.2653914578258991, "replacement": 0.8859352730214596} +{"step": 143, "steps": 250, "lr": 2.4787924694255237e-05, "grad_norm": 5.009029865264893, "loss_scale": 4.0, "retries": 0, "seconds": 134.6, "max_memory_gb": 23.45, "pair": 0.5101301949471235, "sign": 0.19248697522561997, "keep": 1.3461784391147376, "leak": 0.2535308450460434, "replacement": 1.34284558147192} +{"step": 144, "steps": 250, "lr": 2.4488304379806627e-05, "grad_norm": 4.727694034576416, "loss_scale": 4.0, "retries": 0, "seconds": 135.5, "max_memory_gb": 23.45, "pair": 0.5290768072009087, "sign": 0.7444486068561673, "keep": 0.12815045076422393, "leak": 0.2915655798278749, "replacement": 1.0382676869630814} +{"step": 145, "steps": 250, "lr": 2.4189226906286388e-05, "grad_norm": 2.164614677429199, "loss_scale": 4.0, "retries": 0, "seconds": 136.5, "max_memory_gb": 23.45, "pair": 0.2585819251835346, "sign": 0.049616823438555, "keep": 0.011390460560960491, "leak": 0.38557537086308, "replacement": 0.38547941483557224} +{"step": 146, "steps": 250, "lr": 2.3890746180700395e-05, "grad_norm": 3.3982772827148438, "loss_scale": 4.0, "retries": 0, "seconds": 137.4, "max_memory_gb": 23.45, "pair": 0.43507399782538414, "sign": 0.10604742111172527, "keep": 0.012557574431411922, "leak": 1.0404286496341228, "replacement": 0.754483662545681} +{"step": 147, "steps": 250, "lr": 2.359291600249407e-05, "grad_norm": 5.811394214630127, "loss_scale": 4.0, "retries": 0, "seconds": 138.4, "max_memory_gb": 23.45, "pair": 0.607807919383049, "sign": 0.8993321405723691, "keep": 1.0915160043514334, "leak": 0.778187770396471, "replacement": 1.7977788895368576} +{"step": 148, "steps": 250, "lr": 2.3295790053855392e-05, "grad_norm": 4.943589210510254, "loss_scale": 4.0, "retries": 0, "seconds": 139.3, "max_memory_gb": 23.45, "pair": 0.24474860727787018, "sign": 0.1068154705571942, "keep": 0.005117615741937698, "leak": 0.35829640948213637, "replacement": 0.3902892582118511} +{"step": 149, "steps": 250, "lr": 2.299942189003901e-05, "grad_norm": 4.800914287567139, "loss_scale": 4.0, "retries": 0, "seconds": 140.2, "max_memory_gb": 23.45, "pair": 0.2946464754641056, "sign": 0.0814661526819691, "keep": 0.022828185639809817, "leak": 0.8256185640348122, "replacement": 0.5531982854008675} +{"step": 150, "steps": 250, "lr": 2.2703864929713144e-05, "grad_norm": 4.116060256958008, "loss_scale": 4.0, "retries": 0, "seconds": 141.2, "max_memory_gb": 23.45, "pair": 0.3836465310305357, "sign": 0.1338792215101421, "keep": 0.21651139063760638, "leak": 0.15304641160400934, "replacement": 0.59710343927145} +{"step": 151, "steps": 250, "lr": 2.2409172445331257e-05, "grad_norm": 2.8845226764678955, "loss_scale": 4.0, "retries": 0, "seconds": 142.1, "max_memory_gb": 23.45, "pair": 0.3048166558146477, "sign": 0.02391186961904168, "keep": 0.019786005606874824, "leak": 0.12981912726536393, "replacement": 0.3591203838586807} +{"step": 152, "steps": 250, "lr": 2.211539755352996e-05, "grad_norm": 2.581256628036499, "loss_scale": 4.0, "retries": 0, "seconds": 143.1, "max_memory_gb": 23.45, "pair": 0.2988779917359352, "sign": 0.009038657764904201, "keep": 0.03276753648970043, "leak": 0.37232253327965736, "replacement": 0.4128617197275162} +{"step": 153, "steps": 250, "lr": 2.1822593205555016e-05, "grad_norm": 5.1495513916015625, "loss_scale": 4.0, "retries": 0, "seconds": 144.0, "max_memory_gb": 23.45, "pair": 0.2155266753397882, "sign": 0.0068855645658914, "keep": 0.022332034714054316, "leak": 1.3184960344806314, "replacement": 0.5597594901919365} +{"step": 154, "steps": 250, "lr": 2.1530812177717236e-05, "grad_norm": 3.726895570755005, "loss_scale": 4.0, "retries": 0, "seconds": 145.0, "max_memory_gb": 23.45, "pair": 0.19890210358425975, "sign": 0.011393359396606684, "keep": 0.4803458289243281, "leak": 0.7290552444756031, "replacement": 0.6270355135202408} +{"step": 155, "steps": 250, "lr": 2.1240107061879815e-05, "grad_norm": 2.0304386615753174, "loss_scale": 4.0, "retries": 0, "seconds": 145.9, "max_memory_gb": 23.45, "pair": 0.16618232056498528, "sign": 0.008314794162288308, "keep": 0.5403599166306776, "leak": 0.7858584970235825, "replacement": 0.6369842682033777} +{"step": 156, "steps": 250, "lr": 2.095053025597896e-05, "grad_norm": 3.192566156387329, "loss_scale": 4.0, "retries": 0, "seconds": 146.9, "max_memory_gb": 23.45, "pair": 0.14222054556012154, "sign": 0.0036413881462067366, "keep": 0.05075257745784256, "leak": 0.05569462256971747, "replacement": 0.18334118090569973} +{"step": 157, "steps": 250, "lr": 2.066213395457945e-05, "grad_norm": 9.960979461669922, "loss_scale": 4.0, "retries": 0, "seconds": 147.8, "max_memory_gb": 23.45, "pair": 0.18772575724869967, "sign": 1.2314082382945344, "keep": 0.05083644625847228, "leak": 4.6256523029878736, "replacement": 1.9852612353861332} +{"step": 158, "steps": 250, "lr": 2.0374970139466877e-05, "grad_norm": 3.695852518081665, "loss_scale": 4.0, "retries": 0, "seconds": 148.7, "max_memory_gb": 23.45, "pair": 0.42417849972844124, "sign": 0.4407242441084236, "keep": 0.5417294786821003, "leak": 0.030904082872439176, "replacement": 0.9231313951313496} +{"step": 159, "steps": 250, "lr": 2.0089090570278213e-05, "grad_norm": 4.652343273162842, "loss_scale": 4.0, "retries": 0, "seconds": 149.7, "max_memory_gb": 23.5, "pair": 0.3028228133916855, "sign": 0.013092241249978542, "keep": 0.005301560246152803, "leak": 0.5078412909060717, "replacement": 0.4389800392091274} +{"step": 160, "steps": 250, "lr": 1.9804546775172457e-05, "grad_norm": 4.239564418792725, "loss_scale": 4.0, "retries": 0, "seconds": 150.6, "max_memory_gb": 23.5, "pair": 0.29546225629746914, "sign": 0.18406057794345543, "keep": 0.5708176319021732, "leak": 0.6063869152467305, "replacement": 0.8244981020689011} +{"step": 161, "steps": 250, "lr": 1.9521390041542946e-05, "grad_norm": 2.7788279056549072, "loss_scale": 4.0, "retries": 0, "seconds": 151.5, "max_memory_gb": 23.5, "pair": 0.34456988982856274, "sign": 0.1310394894098863, "keep": 0.16409564523473819, "leak": 0.1655544089153409, "replacement": 0.5335260629653931} +{"step": 162, "steps": 250, "lr": 1.923967140677313e-05, "grad_norm": 1.9491991996765137, "loss_scale": 4.0, "retries": 0, "seconds": 152.4, "max_memory_gb": 23.5, "pair": 0.12131927162408829, "sign": 0.020973764010705054, "keep": 0.15211636992899003, "leak": 0.4997031898237765, "replacement": 0.33279013261199} +{"step": 163, "steps": 250, "lr": 1.8959441649037335e-05, "grad_norm": 5.053961753845215, "loss_scale": 4.0, "retries": 0, "seconds": 153.4, "max_memory_gb": 23.5, "pair": 0.35671701468527317, "sign": 0.28039544471539557, "keep": 0.40105443773791194, "leak": 0.1786356121301651, "replacement": 0.7421008553355932} +{"step": 164, "steps": 250, "lr": 1.868075127814832e-05, "grad_norm": 6.869218349456787, "loss_scale": 4.0, "retries": 0, "seconds": 154.3, "max_memory_gb": 23.5, "pair": 0.3374380450695753, "sign": 0.29140359757002443, "keep": 0.11610537413798738, "leak": 0.40399906784296036, "replacement": 0.6421923041343689} +{"step": 165, "steps": 250, "lr": 1.8403650526453143e-05, "grad_norm": 1.2993327379226685, "loss_scale": 4.0, "retries": 0, "seconds": 155.3, "max_memory_gb": 23.5, "pair": 0.1313467165455222, "sign": 0.01596185751259327, "keep": 0.0028870401838503312, "leak": 0.31404657289385796, "replacement": 0.2192828133702278} +{"step": 166, "steps": 250, "lr": 1.812818933977912e-05, "grad_norm": 8.856720924377441, "loss_scale": 4.0, "retries": 0, "seconds": 156.3, "max_memory_gb": 23.5, "pair": 0.5777293182909489, "sign": 0.7502329214476049, "keep": 1.1821127836592495, "leak": 0.9124044113559648, "replacement": 1.772003322839737} +{"step": 167, "steps": 250, "lr": 1.785441736843128e-05, "grad_norm": 3.505953311920166, "loss_scale": 4.0, "retries": 0, "seconds": 157.2, "max_memory_gb": 23.5, "pair": 0.32562143355607986, "sign": 0.06494474341161549, "keep": 0.051596821867860854, "leak": 0.5488553643226624, "replacement": 0.52110605686903} +{"step": 168, "steps": 250, "lr": 1.7582383958243314e-05, "grad_norm": 2.003533124923706, "loss_scale": 4.0, "retries": 0, "seconds": 158.2, "max_memory_gb": 23.5, "pair": 0.22223750315606594, "sign": 0.014279566006734967, "keep": 0.10612880456028506, "leak": 0.20760933216661215, "replacement": 0.3343440294265747} +{"step": 169, "steps": 250, "lr": 1.7312138141683164e-05, "grad_norm": 2.364988327026367, "loss_scale": 4.0, "retries": 0, "seconds": 159.1, "max_memory_gb": 23.5, "pair": 0.36305018700659275, "sign": 0.138581091305241, "keep": 0.16027683415450156, "leak": 0.5091082379221916, "replacement": 0.6397561877965927} +{"step": 170, "steps": 250, "lr": 1.704372862901521e-05, "grad_norm": 7.51207160949707, "loss_scale": 4.0, "retries": 0, "seconds": 160.0, "max_memory_gb": 23.5, "pair": 0.19210129254497588, "sign": 0.004783135271281935, "keep": 0.032256672842777334, "leak": 1.8213392570614815, "replacement": 0.6659560017287731} +{"step": 171, "steps": 250, "lr": 1.6777203799520573e-05, "grad_norm": 2.758462429046631, "loss_scale": 4.0, "retries": 0, "seconds": 161.0, "max_memory_gb": 23.5, "pair": 0.16825165087357163, "sign": 0.10084015520260436, "keep": 0.07658776346215745, "leak": 0.19105762615799904, "replacement": 0.30473001953214407} +{"step": 172, "steps": 250, "lr": 1.6512611692777008e-05, "grad_norm": 1.556313395500183, "loss_scale": 4.0, "retries": 0, "seconds": 161.9, "max_memory_gb": 23.5, "pair": 0.09103768621571362, "sign": 0.0022401553578674793, "keep": 0.016317733796313405, "leak": 0.5696980878710747, "replacement": 0.24274115078151226} +{"step": 173, "steps": 250, "lr": 1.6250000000000005e-05, "grad_norm": 6.084509372711182, "loss_scale": 4.0, "retries": 0, "seconds": 162.8, "max_memory_gb": 23.5, "pair": 0.111186517868191, "sign": 0.08506812651467044, "keep": 0.10527784255100414, "leak": 1.445675351365935, "replacement": 0.5677783414721489} +{"step": 174, "steps": 250, "lr": 1.598941605544678e-05, "grad_norm": 2.8956844806671143, "loss_scale": 4.0, "retries": 0, "seconds": 163.8, "max_memory_gb": 23.5, "pair": 0.11431170313153416, "sign": 0.008417917168117128, "keep": 0.025749208843990345, "leak": 0.7757948376238346, "replacement": 0.3253439739346504} +{"step": 175, "steps": 250, "lr": 1.5730906827884506e-05, "grad_norm": 4.828185081481934, "loss_scale": 4.0, "retries": 0, "seconds": 164.7, "max_memory_gb": 23.5, "pair": 0.28954800497740507, "sign": 0.2864628175739199, "keep": 0.9768359586596489, "leak": 0.5471714052837342, "replacement": 1.0579902194440365} +{"step": 176, "steps": 250, "lr": 1.5474518912124476e-05, "grad_norm": 1.3822336196899414, "loss_scale": 4.0, "retries": 0, "seconds": 165.6, "max_memory_gb": 23.5, "pair": 0.06629678793251514, "sign": 0.02587925162515603, "keep": 0.01321652575279586, "leak": 0.42115557938814163, "replacement": 0.19113357365131378} +{"step": 177, "steps": 250, "lr": 1.5220298520623646e-05, "grad_norm": 4.861684322357178, "loss_scale": 4.0, "retries": 0, "seconds": 166.5, "max_memory_gb": 23.5, "pair": 0.22792528849095106, "sign": 0.39088111044839025, "keep": 0.0070441175612359075, "leak": 0.5434999351600709, "replacement": 0.5627628825604916} +{"step": 178, "steps": 250, "lr": 1.4968291475155121e-05, "grad_norm": 7.560845375061035, "loss_scale": 4.0, "retries": 0, "seconds": 167.5, "max_memory_gb": 23.5, "pair": 0.0988771726260893, "sign": 0.09048616212385241, "keep": 0.6633108395035379, "leak": 1.1360304933041334, "replacement": 0.759783286601305} +{"step": 179, "steps": 250, "lr": 1.4718543198548998e-05, "grad_norm": 2.9308128356933594, "loss_scale": 4.0, "retries": 0, "seconds": 168.4, "max_memory_gb": 23.5, "pair": 0.5523253865540028, "sign": 0.5354432988679036, "keep": 0.634722882805363, "leak": 0.11393846555438358, "replacement": 1.1658931244164705} +{"step": 180, "steps": 250, "lr": 1.4471098706505227e-05, "grad_norm": 2.441545009613037, "loss_scale": 4.0, "retries": 0, "seconds": 169.3, "max_memory_gb": 23.5, "pair": 0.30946889892220497, "sign": 0.034743663389235735, "keep": 0.06286283489316702, "leak": 0.4144070746297075, "replacement": 0.4618739187717438} +{"step": 181, "steps": 250, "lr": 1.4226002599479743e-05, "grad_norm": 12.38836669921875, "loss_scale": 4.0, "retries": 0, "seconds": 170.2, "max_memory_gb": 23.5, "pair": 0.07020519202342257, "sign": 0.005059695104137063, "keep": 0.03909879759885371, "leak": 2.7801708579063416, "replacement": 0.7873271629214287} +{"step": 182, "steps": 250, "lr": 1.3983299054645471e-05, "grad_norm": 5.9453277587890625, "loss_scale": 4.0, "retries": 0, "seconds": 171.2, "max_memory_gb": 23.5, "pair": 0.36403010599315166, "sign": 0.29692805965896696, "keep": 0.01344447990777553, "leak": 1.8604828175157309, "replacement": 0.9843371007591486} +{"step": 183, "steps": 250, "lr": 1.3743031817929724e-05, "grad_norm": 2.262068271636963, "loss_scale": 4.0, "retries": 0, "seconds": 172.1, "max_memory_gb": 23.5, "pair": 0.17344152741134167, "sign": 0.05796470007044263, "keep": 0.0038769960260651715, "leak": 0.5655492469668388, "replacement": 0.3457496725022793} +{"step": 184, "steps": 250, "lr": 1.3505244196129164e-05, "grad_norm": 3.608475923538208, "loss_scale": 4.0, "retries": 0, "seconds": 173.1, "max_memory_gb": 23.5, "pair": 0.2343648737296462, "sign": 0.005119273962918669, "keep": 0.008632730001409072, "leak": 0.9310399717796827, "replacement": 0.4740008618682623} +{"step": 185, "steps": 250, "lr": 1.3269979049104014e-05, "grad_norm": 2.3191120624542236, "loss_scale": 4.0, "retries": 0, "seconds": 174.0, "max_memory_gb": 23.5, "pair": 0.20315502723678946, "sign": 0.028640498363529332, "keep": 0.04970302729634568, "leak": 0.47746759292203933, "replacement": 0.36169368773698807} +{"step": 186, "steps": 250, "lr": 1.3037278782052863e-05, "grad_norm": 3.304003953933716, "loss_scale": 4.0, "retries": 0, "seconds": 175.0, "max_memory_gb": 23.5, "pair": 0.25364899018313736, "sign": 0.023034011261188425, "keep": 0.7729915093714226, "leak": 0.5611933246254921, "replacement": 0.7919600605964661} +{"step": 187, "steps": 250, "lr": 1.2807185337869343e-05, "grad_norm": 4.955257415771484, "loss_scale": 4.0, "retries": 0, "seconds": 175.9, "max_memory_gb": 23.5, "pair": 0.32169679505750537, "sign": 0.010771787056000903, "keep": 0.2526653312961571, "leak": 1.1454856533091515, "replacement": 0.7397867813706398} +{"step": 188, "steps": 250, "lr": 1.2579740189582115e-05, "grad_norm": 3.646132707595825, "loss_scale": 4.0, "retries": 0, "seconds": 176.9, "max_memory_gb": 23.5, "pair": 0.6362797524780035, "sign": 0.8894701217650436, "keep": 0.7905791374255386, "leak": 0.09116349792111578, "replacement": 1.4990952592343092} +{"step": 189, "steps": 250, "lr": 1.2354984332879668e-05, "grad_norm": 6.091403484344482, "loss_scale": 4.0, "retries": 0, "seconds": 177.8, "max_memory_gb": 23.5, "pair": 0.07544150529429317, "sign": 0.4906339992885478, "keep": 1.0743883424438536, "leak": 1.352046095766127, "replacement": 1.1959642246365547} +{"step": 190, "steps": 250, "lr": 1.2132958278721064e-05, "grad_norm": 2.966768264770508, "loss_scale": 4.0, "retries": 0, "seconds": 178.7, "max_memory_gb": 23.5, "pair": 0.06865718634799123, "sign": 0.005012298468500376, "keep": 0.010531702922889963, "leak": 0.813278041779995, "replacement": 0.27974870800971985} +{"step": 191, "steps": 250, "lr": 1.1913702046034016e-05, "grad_norm": 11.949451446533203, "loss_scale": 4.0, "retries": 0, "seconds": 179.6, "max_memory_gb": 23.5, "pair": 0.3130302201025188, "sign": 0.4621710896026343, "keep": 0.7327231789413418, "leak": 1.532055839896202, "replacement": 1.293491292744875} +{"step": 192, "steps": 250, "lr": 1.1697255154501756e-05, "grad_norm": 2.318822145462036, "loss_scale": 4.0, "retries": 0, "seconds": 180.6, "max_memory_gb": 23.5, "pair": 0.4618484191596508, "sign": 0.465757581172511, "keep": 1.0760598997585475, "leak": 0.40415441850200295, "replacement": 1.333795703947544} +{"step": 193, "steps": 250, "lr": 1.1483656617439842e-05, "grad_norm": 2.410166025161743, "loss_scale": 4.0, "retries": 0, "seconds": 181.5, "max_memory_gb": 23.5, "pair": 0.15708715235814452, "sign": 0.02031170978443697, "keep": 0.01573695935167052, "leak": 0.659122884273529, "replacement": 0.33989221416413784} +{"step": 194, "steps": 250, "lr": 1.1272944934764177e-05, "grad_norm": 2.403402805328369, "loss_scale": 4.0, "retries": 0, "seconds": 182.4, "max_memory_gb": 23.5, "pair": 0.2741525564342737, "sign": 0.21942797512747347, "keep": 0.022075949162172037, "leak": 0.4694997303013224, "replacement": 0.5122794508934021} +{"step": 195, "steps": 250, "lr": 1.1065158086051684e-05, "grad_norm": 6.218924045562744, "loss_scale": 4.0, "retries": 0, "seconds": 183.4, "max_memory_gb": 23.5, "pair": 0.2650020846631378, "sign": 0.06594177742954344, "keep": 0.007129984456696548, "leak": 1.202555313706398, "replacement": 0.6021767929196358} +{"step": 196, "steps": 250, "lr": 1.086033352369467e-05, "grad_norm": 4.2800750732421875, "loss_scale": 4.0, "retries": 0, "seconds": 184.4, "max_memory_gb": 23.5, "pair": 0.1256793518550694, "sign": 0.008021734043722972, "keep": 0.009896546906020376, "leak": 1.5555541003122926, "replacement": 0.5235270131379366} +{"step": 197, "steps": 250, "lr": 1.0658508166150224e-05, "grad_norm": 6.688468933105469, "loss_scale": 4.0, "retries": 0, "seconds": 185.3, "max_memory_gb": 23.5, "pair": 0.08653516846243292, "sign": 0.03011489796335809, "keep": 0.0028768260963261127, "leak": 2.1113070901483297, "replacement": 0.6308577861636877} +{"step": 198, "steps": 250, "lr": 1.0459718391285954e-05, "grad_norm": 4.8920512199401855, "loss_scale": 4.0, "retries": 0, "seconds": 186.2, "max_memory_gb": 23.5, "pair": 0.16838349448516965, "sign": 0.07136286440072581, "keep": 0.12267011529183947, "leak": 0.6170565914362669, "replacement": 0.4196641370654106} +{"step": 199, "steps": 250, "lr": 1.0264000029822999e-05, "grad_norm": 6.541852951049805, "loss_scale": 4.0, "retries": 0, "seconds": 187.2, "max_memory_gb": 23.5, "pair": 0.16451295441947877, "sign": 0.03159110051637981, "keep": 0.47052400052780285, "leak": 1.139233217574656, "replacement": 0.7003787942230701} +{"step": 200, "steps": 250, "lr": 1.0071388358877775e-05, "grad_norm": 4.9275803565979, "loss_scale": 4.0, "retries": 0, "seconds": 188.1, "max_memory_gb": 23.5, "pair": 0.4078225011471659, "sign": 0.35005124093731865, "keep": 1.188329095573863, "leak": 0.8437770139425993, "replacement": 1.3879569321870804} +{"step": 201, "steps": 250, "lr": 9.881918095603501e-06, "grad_norm": 5.298229694366455, "loss_scale": 4.0, "retries": 0, "seconds": 189.1, "max_memory_gb": 23.5, "pair": 0.46836917847394943, "sign": 0.229027031455189, "keep": 0.2353604190866463, "leak": 0.31115353481436614, "replacement": 0.7783513069152832} +{"step": 202, "steps": 250, "lr": 9.695623390932612e-06, "grad_norm": 9.523954391479492, "loss_scale": 4.0, "retries": 0, "seconds": 190.0, "max_memory_gb": 23.5, "pair": 0.22056032274849713, "sign": 0.10296744031074923, "keep": 0.012148418130635719, "leak": 2.4662885181605816, "replacement": 0.8946903944015503} +{"step": 203, "steps": 250, "lr": 9.51253782342123e-06, "grad_norm": 3.2422142028808594, "loss_scale": 4.0, "retries": 0, "seconds": 191.0, "max_memory_gb": 23.5, "pair": 0.1623641476035118, "sign": 0.07214483275311068, "keep": 0.08572021371219307, "leak": 0.3416653759777546, "replacement": 0.3267130143940449} +{"step": 204, "steps": 250, "lr": 9.332694393196846e-06, "grad_norm": 4.308422088623047, "loss_scale": 4.0, "retries": 0, "seconds": 191.9, "max_memory_gb": 23.5, "pair": 0.37937968224287033, "sign": 0.025122873950749636, "keep": 0.06094139412743971, "leak": 0.6358598303049803, "replacement": 0.5813767835497856} +{"step": 205, "steps": 250, "lr": 9.156125516010256e-06, "grad_norm": 2.2761218547821045, "loss_scale": 4.0, "retries": 0, "seconds": 192.8, "max_memory_gb": 23.5, "pair": 0.18865554872900248, "sign": 0.006700021360302344, "keep": 0.011965400743065402, "leak": 0.15350190683966503, "replacement": 0.23636373691260815} +{"step": 206, "steps": 250, "lr": 8.982863017392733e-06, "grad_norm": 3.3790132999420166, "loss_scale": 4.0, "retries": 0, "seconds": 193.7, "max_memory_gb": 23.5, "pair": 0.20199072733521461, "sign": 0.02698435290949419, "keep": 0.01684729917792538, "leak": 0.2340179979801178, "replacement": 0.2824110575020313} +{"step": 207, "steps": 250, "lr": 8.81293812691974e-06, "grad_norm": 2.574406147003174, "loss_scale": 4.0, "retries": 0, "seconds": 194.7, "max_memory_gb": 23.5, "pair": 0.3759353458881378, "sign": 0.1640704208984971, "keep": 0.09033419587012759, "leak": 0.15787844301667064, "replacement": 0.5426072627305984} +{"step": 208, "steps": 250, "lr": 8.646381472581927e-06, "grad_norm": 2.134964942932129, "loss_scale": 4.0, "retries": 0, "seconds": 195.7, "max_memory_gb": 23.5, "pair": 0.11132004065439105, "sign": 0.00383609754499048, "keep": 0.008553459284030396, "leak": 0.619347333908081, "replacement": 0.2723516598343849} +{"step": 209, "steps": 250, "lr": 8.483223075264622e-06, "grad_norm": 5.268187999725342, "loss_scale": 4.0, "retries": 0, "seconds": 196.6, "max_memory_gb": 23.5, "pair": 0.645466759800911, "sign": 0.4058011608431116, "keep": 0.3675132947973907, "leak": 0.5958432592451572, "replacement": 1.1810847967863083} +{"step": 210, "steps": 250, "lr": 8.323492343336747e-06, "grad_norm": 5.530877113342285, "loss_scale": 4.0, "retries": 0, "seconds": 197.6, "max_memory_gb": 23.5, "pair": 0.5340253189206123, "sign": 0.5147991530247964, "keep": 1.2652276446715405, "leak": 0.554513081908226, "replacement": 1.5626671090722084} +{"step": 211, "steps": 250, "lr": 8.167218067350122e-06, "grad_norm": 2.7444634437561035, "loss_scale": 4.0, "retries": 0, "seconds": 198.5, "max_memory_gb": 23.5, "pair": 0.5570114646106958, "sign": 0.27988557994831353, "keep": 0.11875587818212807, "leak": 0.16519427113234997, "replacement": 0.7976307831704617} +{"step": 212, "steps": 250, "lr": 8.01442841485013e-06, "grad_norm": 4.716338634490967, "loss_scale": 4.0, "retries": 0, "seconds": 199.5, "max_memory_gb": 23.5, "pair": 0.5809070188552141, "sign": 0.1625072363531217, "keep": 0.04424880584701896, "leak": 1.022642394527793, "replacement": 0.9399456083774567} +{"step": 213, "steps": 250, "lr": 7.865150925298678e-06, "grad_norm": 5.467340469360352, "loss_scale": 4.0, "retries": 0, "seconds": 200.4, "max_memory_gb": 23.5, "pair": 0.5115186534821987, "sign": 0.33201438665855676, "keep": 0.4640562967979349, "leak": 0.3946792894275859, "replacement": 1.008223870769143} +{"step": 214, "steps": 250, "lr": 7.719412505110375e-06, "grad_norm": 2.25899076461792, "loss_scale": 4.0, "retries": 0, "seconds": 201.4, "max_memory_gb": 23.5, "pair": 0.4383850609883666, "sign": 0.3764524938160321, "keep": 1.1568467101733404, "leak": 0.016567298636800842, "replacement": 1.2091764444485307} +{"step": 215, "steps": 250, "lr": 7.57723942280278e-06, "grad_norm": 4.080733776092529, "loss_scale": 4.0, "retries": 0, "seconds": 202.3, "max_memory_gb": 23.5, "pair": 0.6811352516524494, "sign": 0.8051796716172248, "keep": 1.4441743241150107, "leak": 0.06449871882796288, "replacement": 1.8219367978163064} +{"step": 216, "steps": 250, "lr": 7.438657304261687e-06, "grad_norm": 3.2836837768554688, "loss_scale": 4.0, "retries": 0, "seconds": 203.3, "max_memory_gb": 23.5, "pair": 0.16818983480334282, "sign": 0.04452274728100747, "keep": 0.012336176187091041, "leak": 1.1154997846315382, "replacement": 0.4754942487925291} +{"step": 217, "steps": 250, "lr": 7.30369112812219e-06, "grad_norm": 2.570432424545288, "loss_scale": 4.0, "retries": 0, "seconds": 204.3, "max_memory_gb": 23.5, "pair": 0.22869562450796366, "sign": 0.01158165413653478, "keep": 0.0013280081875564065, "leak": 0.423841760201185, "replacement": 0.3411109000444412} +{"step": 218, "steps": 250, "lr": 7.172365221266403e-06, "grad_norm": 5.507616996765137, "loss_scale": 4.0, "retries": 0, "seconds": 205.2, "max_memory_gb": 23.5, "pair": 0.7707657814025879, "sign": 1.3631344065070152, "keep": 0.23286995477974415, "leak": 0.10073117003776133, "replacement": 1.59395070374012} +{"step": 219, "steps": 250, "lr": 7.0447032544387154e-06, "grad_norm": 4.163623809814453, "loss_scale": 4.0, "retries": 0, "seconds": 206.1, "max_memory_gb": 23.5, "pair": 0.280121386051178, "sign": 0.08955626501119696, "keep": 0.008693234820384532, "leak": 0.05363070033490658, "replacement": 0.342653825879097} +{"step": 220, "steps": 250, "lr": 6.920728237979274e-06, "grad_norm": 4.193789958953857, "loss_scale": 4.0, "retries": 0, "seconds": 207.1, "max_memory_gb": 23.5, "pair": 0.3091838210821152, "sign": 0.05628209881251678, "keep": 0.08877278302861669, "leak": 0.08400331204757094, "replacement": 0.402712095528841} +{"step": 221, "steps": 250, "lr": 6.800462517676456e-06, "grad_norm": 3.1668944358825684, "loss_scale": 4.0, "retries": 0, "seconds": 208.0, "max_memory_gb": 23.5, "pair": 0.4022353757172823, "sign": 0.18360121734440327, "keep": 0.010108802234753966, "leak": 0.8214560975320637, "replacement": 0.7044544368982315} +{"step": 222, "steps": 250, "lr": 6.683927770739222e-06, "grad_norm": 2.4526116847991943, "loss_scale": 4.0, "retries": 0, "seconds": 209.0, "max_memory_gb": 23.5, "pair": 0.20838124118745327, "sign": 0.015701519383583218, "keep": 0.04342550827823288, "leak": 0.2312958464026451, "replacement": 0.2957687024027109} +{"step": 223, "steps": 250, "lr": 6.571145001889892e-06, "grad_norm": 6.6106438636779785, "loss_scale": 4.0, "retries": 0, "seconds": 209.9, "max_memory_gb": 23.5, "pair": 0.45659508323296905, "sign": 0.07076830478035845, "keep": 0.011266090659773909, "leak": 0.3019654634408653, "replacement": 0.5731036402285099} +{"step": 224, "steps": 250, "lr": 6.46213453957817e-06, "grad_norm": 2.9970180988311768, "loss_scale": 4.0, "retries": 0, "seconds": 210.9, "max_memory_gb": 23.5, "pair": 0.2624604292213917, "sign": 0.035413987934589386, "keep": 0.00945678561765817, "leak": 0.3263571113348007, "replacement": 0.3664851076900959} +{"step": 225, "steps": 250, "lr": 6.356916032317063e-06, "grad_norm": 5.040408134460449, "loss_scale": 4.0, "retries": 0, "seconds": 211.8, "max_memory_gb": 23.5, "pair": 0.1688397265970707, "sign": 0.02133030234836042, "keep": 0.0021508974987227703, "leak": 1.3703222121112049, "replacement": 0.523160882294178} +{"step": 226, "steps": 250, "lr": 6.255508445141349e-06, "grad_norm": 3.933116912841797, "loss_scale": 4.0, "retries": 0, "seconds": 212.8, "max_memory_gb": 23.5, "pair": 0.10615177545696497, "sign": 0.02340936753898859, "keep": 0.8677920744521543, "leak": 0.5656794346868992, "replacement": 0.6931723393499851} +{"step": 227, "steps": 250, "lr": 6.157930056189228e-06, "grad_norm": 2.4625701904296875, "loss_scale": 4.0, "retries": 0, "seconds": 213.7, "max_memory_gb": 23.5, "pair": 0.1299502938054502, "sign": 0.013071440276689827, "keep": 0.07041355704859598, "leak": 0.3209327161312103, "replacement": 0.25192597322165966} +{"step": 228, "steps": 250, "lr": 6.064198453407796e-06, "grad_norm": 2.1376712322235107, "loss_scale": 4.0, "retries": 0, "seconds": 214.6, "max_memory_gb": 23.5, "pair": 0.15825057262554765, "sign": 0.056596155336592346, "keep": 0.0030068444466451183, "leak": 0.7166885025799274, "replacement": 0.36722419410943985} +{"step": 229, "steps": 250, "lr": 5.974330531382935e-06, "grad_norm": 11.365504264831543, "loss_scale": 4.0, "retries": 0, "seconds": 215.5, "max_memory_gb": 23.5, "pair": 0.20558402314782143, "sign": 0.07724377524573356, "keep": 0.08309052023105146, "leak": 2.19745984547626, "replacement": 0.8351161330938339} +{"step": 230, "steps": 250, "lr": 5.888342488294125e-06, "grad_norm": 12.858081817626953, "loss_scale": 4.0, "retries": 0, "seconds": 216.4, "max_memory_gb": 23.5, "pair": 0.7723374292254448, "sign": 0.5281417109072208, "keep": 0.984033512086171, "leak": 1.0060468707233667, "replacement": 1.7799368053674698} +{"step": 231, "steps": 250, "lr": 5.806249822994849e-06, "grad_norm": 2.992016315460205, "loss_scale": 4.0, "retries": 0, "seconds": 217.3, "max_memory_gb": 23.5, "pair": 0.38202427234500647, "sign": 0.721672025276348, "keep": 0.02158565795400591, "leak": 0.4427891708910465, "replacement": 0.8643504120409489} +{"step": 232, "steps": 250, "lr": 5.728067332218994e-06, "grad_norm": 3.078366756439209, "loss_scale": 4.0, "retries": 0, "seconds": 218.3, "max_memory_gb": 23.5, "pair": 0.11424449644982815, "sign": 0.017823546542786062, "keep": 0.03724151710048318, "leak": 0.9349989891052246, "replacement": 0.3755267709493637} +{"step": 233, "steps": 250, "lr": 5.65380910791383e-06, "grad_norm": 5.180667877197266, "loss_scale": 4.0, "retries": 0, "seconds": 219.2, "max_memory_gb": 23.5, "pair": 0.129950150847435, "sign": 0.005743040703237057, "keep": 0.0049739969545044005, "leak": 1.5184457655996084, "replacement": 0.5149201173335314} +{"step": 234, "steps": 250, "lr": 5.583488534700022e-06, "grad_norm": 5.993764877319336, "loss_scale": 4.0, "retries": 0, "seconds": 220.2, "max_memory_gb": 23.5, "pair": 0.1733337095938623, "sign": 0.021067171619506553, "keep": 0.05146082868918711, "leak": 1.0378300547599792, "replacement": 0.469055213034153} +{"step": 235, "steps": 250, "lr": 5.517118287459118e-06, "grad_norm": 13.91789722442627, "loss_scale": 4.0, "retries": 0, "seconds": 221.1, "max_memory_gb": 23.5, "pair": 0.6980656334199011, "sign": 0.4928800093475729, "keep": 0.9286554333448294, "leak": 2.1531444194260985, "replacement": 1.9471195377409458} +{"step": 236, "steps": 250, "lr": 5.454710329048978e-06, "grad_norm": 2.8260152339935303, "loss_scale": 4.0, "retries": 0, "seconds": 222.0, "max_memory_gb": 23.5, "pair": 0.3085516067221761, "sign": 0.09815530202467926, "keep": 0.004962623992469162, "leak": 0.17029440961778164, "replacement": 0.40268419217318296} +{"step": 237, "steps": 250, "lr": 5.3962759081475436e-06, "grad_norm": 5.613719463348389, "loss_scale": 4.0, "retries": 0, "seconds": 222.9, "max_memory_gb": 23.5, "pair": 0.2603071313351393, "sign": 0.36556589195970446, "keep": 0.15689861500868574, "leak": 0.8834318090230227, "replacement": 0.742397353053093} +{"step": 238, "steps": 250, "lr": 5.341825557225321e-06, "grad_norm": 1.6217684745788574, "loss_scale": 4.0, "retries": 0, "seconds": 223.9, "max_memory_gb": 23.5, "pair": 0.18624384608119726, "sign": 0.014573594846297055, "keep": 0.05446972150821239, "leak": 0.4741312935948372, "replacement": 0.3392983302474022} +{"step": 239, "steps": 250, "lr": 5.291369090646961e-06, "grad_norm": 3.63126277923584, "loss_scale": 4.0, "retries": 0, "seconds": 224.8, "max_memory_gb": 23.5, "pair": 0.437970295548439, "sign": 1.3601209511980414, "keep": 0.15312840034721376, "leak": 0.551393088651821, "replacement": 1.3324432075023651} +{"step": 240, "steps": 250, "lr": 5.244915602902289e-06, "grad_norm": 3.724156618118286, "loss_scale": 4.0, "retries": 0, "seconds": 225.7, "max_memory_gb": 23.5, "pair": 0.28048286959528923, "sign": 0.3974168234854005, "keep": 0.15263433946529403, "leak": 0.24725146680611942, "replacement": 0.617321316152811} +{"step": 241, "steps": 250, "lr": 5.20247346696706e-06, "grad_norm": 6.260414123535156, "loss_scale": 4.0, "retries": 0, "seconds": 226.7, "max_memory_gb": 23.5, "pair": 0.6308572120033205, "sign": 0.5066165977914352, "keep": 0.23950066044744744, "leak": 0.2881203864235431, "replacement": 1.0759459286928177} +{"step": 242, "steps": 250, "lr": 5.164050332793787e-06, "grad_norm": 2.6000447273254395, "loss_scale": 4.0, "retries": 0, "seconds": 227.6, "max_memory_gb": 23.5, "pair": 0.256102473475039, "sign": 0.030715097149368376, "keep": 0.0037998610287104384, "leak": 0.6913017109036446, "replacement": 0.4461853727698326} +{"step": 243, "steps": 250, "lr": 5.129653125932873e-06, "grad_norm": 1.6445056200027466, "loss_scale": 4.0, "retries": 0, "seconds": 228.6, "max_memory_gb": 23.5, "pair": 0.1599548605736345, "sign": 0.01717851616558619, "keep": 0.024663810501806438, "leak": 0.4052627459168434, "replacement": 0.2821917086839676} +{"step": 244, "steps": 250, "lr": 5.099288046284329e-06, "grad_norm": 2.6042678356170654, "loss_scale": 4.0, "retries": 0, "seconds": 229.5, "max_memory_gb": 23.5, "pair": 0.3855668269097805, "sign": 0.192853560205549, "keep": 0.034078802447766066, "leak": 0.8481497839093208, "replacement": 0.7110704779624939} +{"step": 245, "steps": 250, "lr": 5.072960566980277e-06, "grad_norm": 4.801987648010254, "loss_scale": 4.0, "retries": 0, "seconds": 230.4, "max_memory_gb": 23.5, "pair": 0.580885227303952, "sign": 0.4013674828456715, "keep": 0.8131394162774086, "leak": 0.3109374651685357, "replacement": 1.2658730149269104} +{"step": 246, "steps": 250, "lr": 5.050675433398435e-06, "grad_norm": 1.918959617614746, "loss_scale": 4.0, "retries": 0, "seconds": 231.4, "max_memory_gb": 23.5, "pair": 0.15478388569317758, "sign": 0.005035302805481479, "keep": 0.003325324632896809, "leak": 0.10779235698282719, "replacement": 0.1859122859314084} +{"step": 247, "steps": 250, "lr": 5.032436662306811e-06, "grad_norm": 3.0716958045959473, "loss_scale": 4.0, "retries": 0, "seconds": 232.3, "max_memory_gb": 23.5, "pair": 0.1458308519795537, "sign": 0.015828513787710108, "keep": 0.0826823753283179, "leak": 0.7458880078047514, "replacement": 0.3815583065152168} +{"step": 248, "steps": 250, "lr": 5.018247541139684e-06, "grad_norm": 2.055466890335083, "loss_scale": 4.0, "retries": 0, "seconds": 233.3, "max_memory_gb": 23.5, "pair": 0.18706130795180798, "sign": 0.030378783820196986, "keep": 0.0315923710440984, "leak": 0.40186576545238495, "replacement": 0.31851332262158394} +{"step": 249, "steps": 250, "lr": 5.00811062740508e-06, "grad_norm": 5.802879333496094, "loss_scale": 4.0, "retries": 0, "seconds": 234.2, "max_memory_gb": 23.5, "pair": 0.23315476812422276, "sign": 0.09558339044451714, "keep": 0.05538184894248843, "leak": 1.293672613879508, "replacement": 0.6320555433630943} +{"step": 250, "steps": 250, "lr": 5.002027748223778e-06, "grad_norm": 3.8568313121795654, "loss_scale": 4.0, "retries": 0, "seconds": 235.1, "max_memory_gb": 23.5, "pair": 0.3163045644760132, "sign": 0.02467782376334071, "keep": 0.06847052747434645, "leak": 0.10836770237074234, "replacement": 0.38997066766023636} diff --git a/runs/single_internvl/replacement/L11_s13/adapter/README.md b/runs/single_internvl/replacement/L11_s13/adapter/README.md new file mode 100644 index 0000000000000000000000000000000000000000..4f9f856cf0e1531b47e25d266833b2977a19d3ac --- /dev/null +++ b/runs/single_internvl/replacement/L11_s13/adapter/README.md @@ -0,0 +1,207 @@ +--- +base_model: /data/OPSD/assets/models/InternVL3_5-8B-HF +library_name: peft +pipeline_tag: text-generation +tags: +- base_model:adapter:/data/OPSD/assets/models/InternVL3_5-8B-HF +- lora +- transformers +--- + +# Model Card for Model ID + + + + + +## Model Details + +### Model Description + + + + + +- **Developed by:** [More Information Needed] +- **Funded by [optional]:** [More Information Needed] +- **Shared by [optional]:** [More Information Needed] +- **Model type:** [More Information Needed] +- **Language(s) (NLP):** [More Information Needed] +- **License:** [More Information Needed] +- **Finetuned from model [optional]:** [More Information Needed] + +### Model Sources [optional] + + + +- **Repository:** [More Information Needed] +- **Paper [optional]:** [More Information Needed] +- **Demo [optional]:** [More Information Needed] + +## Uses + + + +### Direct Use + + + +[More Information Needed] + +### Downstream Use [optional] + + + +[More Information Needed] + +### Out-of-Scope Use + + + +[More Information Needed] + +## Bias, Risks, and Limitations + + + +[More Information Needed] + +### Recommendations + + + +Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. More information needed for further recommendations. + +## How to Get Started with the Model + +Use the code below to get started with the model. + +[More Information Needed] + +## Training Details + +### Training Data + + + +[More Information Needed] + +### Training Procedure + + + +#### Preprocessing [optional] + +[More Information Needed] + + +#### Training Hyperparameters + +- **Training regime:** [More Information Needed] + +#### Speeds, Sizes, Times [optional] + + + +[More Information Needed] + +## Evaluation + + + +### Testing Data, Factors & Metrics + +#### Testing Data + + + +[More Information Needed] + +#### Factors + + + +[More Information Needed] + +#### Metrics + + + +[More Information Needed] + +### Results + +[More Information Needed] + +#### Summary + + + +## Model Examination [optional] + + + +[More Information Needed] + +## Environmental Impact + + + +Carbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700). + +- **Hardware Type:** [More Information Needed] +- **Hours used:** [More Information Needed] +- **Cloud Provider:** [More Information Needed] +- **Compute Region:** [More Information Needed] +- **Carbon Emitted:** [More Information Needed] + +## Technical Specifications [optional] + +### Model Architecture and Objective + +[More Information Needed] + +### Compute Infrastructure + +[More Information Needed] + +#### Hardware + +[More Information Needed] + +#### Software + +[More Information Needed] + +## Citation [optional] + + + +**BibTeX:** + +[More Information Needed] + +**APA:** + +[More Information Needed] + +## Glossary [optional] + + + +[More Information Needed] + +## More Information [optional] + +[More Information Needed] + +## Model Card Authors [optional] + +[More Information Needed] + +## Model Card Contact + +[More Information Needed] +### Framework versions + +- PEFT 0.20.0 \ No newline at end of file diff --git a/runs/single_internvl/replacement/L11_s13/adapter/adapter_config.json b/runs/single_internvl/replacement/L11_s13/adapter/adapter_config.json new file mode 100644 index 0000000000000000000000000000000000000000..14cec52f4f8929917b2b35f4b4deb56ab178617b --- /dev/null +++ b/runs/single_internvl/replacement/L11_s13/adapter/adapter_config.json @@ -0,0 +1,50 @@ +{ + "alora_invocation_tokens": null, + "alpha_pattern": {}, + "arrow_config": null, + "auto_mapping": null, + "base_model_name_or_path": "/data/OPSD/assets/models/InternVL3_5-8B-HF", + "bias": "none", + "corda_config": null, + "ensure_weight_tying": false, + "eva_config": null, + "exclude_modules": null, + "fan_in_fan_out": false, + "inference_mode": true, + "init_lora_weights": true, + "layer_replication": null, + "layers_pattern": null, + "layers_to_transform": null, + "loftq_config": {}, + "lora_alpha": 32, + "lora_bias": false, + "lora_dropout": 0.05, + "lora_ga_config": null, + "megatron_config": null, + "megatron_core": "megatron.core", + "modules_to_save": null, + "monteclora_config": null, + "peft_type": "LORA", + "peft_version": "0.20.0", + "qalora_group_size": 16, + "r": 16, + "rank_pattern": {}, + "revision": null, + "target_modules": [ + "model.language_model.layers.11.mlp.up_proj", + "model.language_model.layers.11.self_attn.v_proj", + "model.language_model.layers.11.self_attn.k_proj", + "model.language_model.layers.11.mlp.down_proj", + "model.language_model.layers.11.mlp.gate_proj", + "model.language_model.layers.11.self_attn.o_proj", + "model.language_model.layers.11.self_attn.q_proj" + ], + "target_parameters": null, + "task_type": "CAUSAL_LM", + "trainable_token_indices": null, + "use_bdlora": null, + "use_dora": false, + "use_qalora": false, + "use_rslora": false, + "velora_config": null +} \ No newline at end of file diff --git a/runs/single_internvl/replacement/L11_s13/config.json b/runs/single_internvl/replacement/L11_s13/config.json new file mode 100644 index 0000000000000000000000000000000000000000..d2011f9396ec26a624e8926769f1db3f279e92ad --- /dev/null +++ b/runs/single_internvl/replacement/L11_s13/config.json @@ -0,0 +1,24 @@ +{ + "behavior": "replacement", + "seed": 13, + "model": "/data/OPSD/assets/models/InternVL3_5-8B-HF", + "backbone": "internvl", + "train": "/data/OPSD/assets/v3/replacement.jsonl", + "items": 1000, + "steps": 250, + "epochs": 1, + "steps_per_epoch": 250, + "groups_per_step": 4, + "lr": 5e-05, + "scheduler": "cosine", + "warmup_steps": 16, + "final_fraction": 0.1, + "weight_decay": 0.0, + "clip": 1.0, + "initial_loss_scale": 4.0, + "adapter": { + "layer": 11, + "rank": 16, + "trainable_parameters": 1212416 + } +} \ No newline at end of file diff --git a/runs/single_internvl/replacement/L11_s13/steps.jsonl b/runs/single_internvl/replacement/L11_s13/steps.jsonl new file mode 100644 index 0000000000000000000000000000000000000000..7dc8258e14f27f3e20233f5403abc8f2afd23eb6 --- /dev/null +++ b/runs/single_internvl/replacement/L11_s13/steps.jsonl @@ -0,0 +1,250 @@ +{"step": 1, "steps": 250, "lr": 3.125e-06, "grad_norm": 2.1181793212890625, "loss_scale": 4.0, "retries": 0, "seconds": 1.1, "max_memory_gb": 22.38, "pair": 1.0986123085021973, "sign": 0.7328193015418947, "keep": 1.137254401575774, "leak": 0.0, "replacement": 2.0336491465568542} +{"step": 2, "steps": 250, "lr": 6.25e-06, "grad_norm": 1.747008204460144, "loss_scale": 4.0, "retries": 0, "seconds": 2.1, "max_memory_gb": 22.41, "pair": 1.0991760790348053, "sign": 0.1809327519731596, "keep": 1.3697603660548339, "leak": 3.282036141172284e-05, "replacement": 1.8745308816432953} +{"step": 3, "steps": 250, "lr": 9.375000000000001e-06, "grad_norm": 3.5470573902130127, "loss_scale": 4.0, "retries": 0, "seconds": 3.1, "max_memory_gb": 22.41, "pair": 1.0956347286701202, "sign": 0.11495003802701831, "keep": 2.0654870569705963, "leak": 1.8043666386802215e-05, "replacement": 2.1858578622341156} +{"step": 4, "steps": 250, "lr": 1.25e-05, "grad_norm": 2.4380152225494385, "loss_scale": 4.0, "retries": 0, "seconds": 4.1, "max_memory_gb": 22.41, "pair": 1.0918105244636536, "sign": 0.038118366384878755, "keep": 0.869103922334034, "leak": 3.4187671246854734e-05, "replacement": 1.5454302430152893} +{"step": 5, "steps": 250, "lr": 1.5625e-05, "grad_norm": 1.2391257286071777, "loss_scale": 4.0, "retries": 0, "seconds": 5.1, "max_memory_gb": 22.41, "pair": 1.0971755683422089, "sign": 0.11669358005747199, "keep": 0.023221987597935367, "leak": 3.0535063160641585e-05, "replacement": 1.1671409606933594} +{"step": 6, "steps": 250, "lr": 1.8750000000000002e-05, "grad_norm": 1.7409164905548096, "loss_scale": 4.0, "retries": 0, "seconds": 6.1, "max_memory_gb": 22.41, "pair": 1.0951426923274994, "sign": 0.19909089244902134, "keep": 0.6303567353752442, "leak": 3.1215459500799625e-05, "replacement": 1.509874314069748} +{"step": 7, "steps": 250, "lr": 2.1875e-05, "grad_norm": 3.0842018127441406, "loss_scale": 4.0, "retries": 0, "seconds": 7.0, "max_memory_gb": 22.41, "pair": 1.0928565859794617, "sign": 0.08981333952397108, "keep": 1.4940929811746173, "leak": 5.7053063301282236e-05, "replacement": 1.8848240673542023} +{"step": 8, "steps": 250, "lr": 2.5e-05, "grad_norm": 3.447913646697998, "loss_scale": 4.0, "retries": 0, "seconds": 8.0, "max_memory_gb": 22.45, "pair": 1.0907460451126099, "sign": 0.21195872663520277, "keep": 1.4073619022965431, "leak": 0.0003407664116821252, "replacement": 1.9004915654659271} +{"step": 9, "steps": 250, "lr": 2.8125000000000003e-05, "grad_norm": 5.899137020111084, "loss_scale": 4.0, "retries": 0, "seconds": 9.0, "max_memory_gb": 22.45, "pair": 1.0930298566818237, "sign": 0.5762052475474775, "keep": 1.320695623755455, "leak": 0.0003730333410203457, "replacement": 2.0415735840797424} +{"step": 10, "steps": 250, "lr": 3.125e-05, "grad_norm": 1.5211933851242065, "loss_scale": 4.0, "retries": 0, "seconds": 10.0, "max_memory_gb": 22.45, "pair": 1.077620506286621, "sign": 0.08008207171224058, "keep": 0.08669704853673466, "leak": 0.0008381153165828437, "replacement": 1.161219596862793} +{"step": 11, "steps": 250, "lr": 3.4375e-05, "grad_norm": 2.930616617202759, "loss_scale": 4.0, "retries": 0, "seconds": 11.0, "max_memory_gb": 22.45, "pair": 1.0681754052639008, "sign": 0.2036422269884497, "keep": 1.1853102102177218, "leak": 0.00044176728988531977, "replacement": 1.7627621293067932} +{"step": 12, "steps": 250, "lr": 3.7500000000000003e-05, "grad_norm": 2.49308443069458, "loss_scale": 4.0, "retries": 0, "seconds": 12.0, "max_memory_gb": 22.45, "pair": 1.053534984588623, "sign": 0.07451092451810837, "keep": 0.8248271078045946, "leak": 0.0010716725446400233, "replacement": 1.5034719407558441} +{"step": 13, "steps": 250, "lr": 4.0625000000000005e-05, "grad_norm": 2.316953420639038, "loss_scale": 4.0, "retries": 0, "seconds": 12.9, "max_memory_gb": 22.45, "pair": 1.0245705842971802, "sign": 0.05571998795494437, "keep": 0.7447724091543932, "leak": 0.005169128567104053, "replacement": 1.4261090159416199} +{"step": 14, "steps": 250, "lr": 4.375e-05, "grad_norm": 3.7882819175720215, "loss_scale": 4.0, "retries": 0, "seconds": 13.9, "max_memory_gb": 22.45, "pair": 1.0292720198631287, "sign": 0.26250082766637206, "keep": 1.3167209029197693, "leak": 0.00409712124383077, "replacement": 1.819907158613205} +{"step": 15, "steps": 250, "lr": 4.6875e-05, "grad_norm": 2.8912205696105957, "loss_scale": 4.0, "retries": 0, "seconds": 14.8, "max_memory_gb": 22.45, "pair": 1.002750724554062, "sign": 0.20619997172616422, "keep": 0.6935092458970757, "leak": 0.017630936112254858, "replacement": 1.4570130705833435} +{"step": 16, "steps": 250, "lr": 5e-05, "grad_norm": 2.1892874240875244, "loss_scale": 4.0, "retries": 0, "seconds": 15.8, "max_memory_gb": 22.45, "pair": 0.9996028989553452, "sign": 0.2979035726748407, "keep": 0.2620617831416894, "leak": 0.008333420613780618, "replacement": 1.281668946146965} +{"step": 17, "steps": 250, "lr": 5e-05, "grad_norm": 2.0914597511291504, "loss_scale": 4.0, "retries": 0, "seconds": 16.8, "max_memory_gb": 22.45, "pair": 0.9996044337749481, "sign": 0.7443311880342662, "keep": 0.6590149160474539, "leak": 0.019352032337337732, "replacement": 1.7061155438423157} +{"step": 18, "steps": 250, "lr": 4.999797225177623e-05, "grad_norm": 3.115663766860962, "loss_scale": 4.0, "retries": 0, "seconds": 17.7, "max_memory_gb": 22.47, "pair": 0.9451591521501541, "sign": 0.17984150093980134, "keep": 0.5266384501301218, "leak": 0.018837736046407372, "replacement": 1.303108587861061} +{"step": 19, "steps": 250, "lr": 4.999188937259492e-05, "grad_norm": 3.3371427059173584, "loss_scale": 4.0, "retries": 0, "seconds": 18.7, "max_memory_gb": 22.47, "pair": 0.8747403770685196, "sign": 0.03417120222002268, "keep": 0.48839715123176575, "leak": 0.15938481222838163, "replacement": 1.1758707463741302} +{"step": 20, "steps": 250, "lr": 4.998175245886032e-05, "grad_norm": 2.7059295177459717, "loss_scale": 4.0, "retries": 0, "seconds": 19.7, "max_memory_gb": 22.47, "pair": 0.9087682813405991, "sign": 0.02796866837888956, "keep": 0.003384642281162087, "leak": 0.023569850716739893, "replacement": 0.9303374141454697} +{"step": 21, "steps": 250, "lr": 4.996756333769319e-05, "grad_norm": 2.678124189376831, "loss_scale": 4.0, "retries": 0, "seconds": 20.7, "max_memory_gb": 22.47, "pair": 0.8129704147577286, "sign": 0.03907717822585255, "keep": 0.11192337059765123, "leak": 0.1238002423197031, "replacement": 0.9194207489490509} +{"step": 22, "steps": 250, "lr": 4.994932456660157e-05, "grad_norm": 2.438140869140625, "loss_scale": 4.0, "retries": 0, "seconds": 21.7, "max_memory_gb": 22.47, "pair": 0.80018749833107, "sign": 0.058500709012150764, "keep": 0.1961988163238857, "leak": 0.08035207074135542, "replacement": 0.9476252645254135} +{"step": 23, "steps": 250, "lr": 4.992703943301973e-05, "grad_norm": 3.6320035457611084, "loss_scale": 4.0, "retries": 0, "seconds": 22.8, "max_memory_gb": 22.47, "pair": 0.7297145426273346, "sign": 0.028833439340814948, "keep": 0.7133059722336839, "leak": 0.06570523185655475, "replacement": 1.1172105371952057} +{"step": 24, "steps": 250, "lr": 4.9900711953715675e-05, "grad_norm": 2.496598243713379, "loss_scale": 4.0, "retries": 0, "seconds": 23.7, "max_memory_gb": 22.47, "pair": 0.5726217329502106, "sign": 0.012072331854142249, "keep": 0.09043385740369558, "leak": 0.22312631085515022, "replacement": 0.6796563863754272} +{"step": 25, "steps": 250, "lr": 4.987034687406713e-05, "grad_norm": 2.128316879272461, "loss_scale": 4.0, "retries": 0, "seconds": 24.7, "max_memory_gb": 22.47, "pair": 0.6262638419866562, "sign": 0.09849936072714627, "keep": 0.7032660312888765, "leak": 0.2585298493504524, "replacement": 1.091779038310051} +{"step": 26, "steps": 250, "lr": 4.9835949667206215e-05, "grad_norm": 1.6648029088974, "loss_scale": 4.0, "retries": 0, "seconds": 25.7, "max_memory_gb": 22.47, "pair": 0.5096875466406345, "sign": 0.05647747922921553, "keep": 0.06479578418657184, "leak": 0.4004646383691579, "replacement": 0.6704403460025787} +{"step": 27, "steps": 250, "lr": 4.9797526533032945e-05, "grad_norm": 1.6325734853744507, "loss_scale": 4.0, "retries": 0, "seconds": 26.7, "max_memory_gb": 22.47, "pair": 0.5279407799243927, "sign": 0.14427378145046532, "keep": 0.14656902545357298, "leak": 0.4624044578522444, "replacement": 0.7889632731676102} +{"step": 28, "steps": 250, "lr": 4.9755084397097716e-05, "grad_norm": 3.8367762565612793, "loss_scale": 4.0, "retries": 0, "seconds": 27.7, "max_memory_gb": 22.47, "pair": 0.7102968394756317, "sign": 0.41080498648807406, "keep": 1.2007319699041545, "leak": 0.45245337672531605, "replacement": 1.629178762435913} +{"step": 29, "steps": 250, "lr": 4.970863090935304e-05, "grad_norm": 2.4155166149139404, "loss_scale": 4.0, "retries": 0, "seconds": 28.6, "max_memory_gb": 22.47, "pair": 0.7131433337926865, "sign": 0.7326982617378235, "keep": 0.05435425788164139, "leak": 0.25680477311834693, "replacement": 1.1708707958459854} +{"step": 30, "steps": 250, "lr": 4.965817444277468e-05, "grad_norm": 1.7336456775665283, "loss_scale": 4.0, "retries": 0, "seconds": 29.6, "max_memory_gb": 22.47, "pair": 0.20400439202785492, "sign": 0.004412332637002692, "keep": 0.06592847307274496, "leak": 1.1551818028092384, "replacement": 0.5279702544212341} +{"step": 31, "steps": 250, "lr": 4.9603724091852456e-05, "grad_norm": 2.233057975769043, "loss_scale": 4.0, "retries": 0, "seconds": 30.6, "max_memory_gb": 22.47, "pair": 0.4725734554231167, "sign": 0.18380766629707068, "keep": 0.42544106487184763, "leak": 1.1607462046667933, "replacement": 1.0673843771219254} +{"step": 32, "steps": 250, "lr": 4.9545289670951025e-05, "grad_norm": 3.0080368518829346, "loss_scale": 4.0, "retries": 0, "seconds": 31.6, "max_memory_gb": 22.47, "pair": 0.3141705300658941, "sign": 0.09435763463261537, "keep": 0.021163367433473468, "leak": 2.015874058008194, "replacement": 0.8758995532989502} +{"step": 33, "steps": 250, "lr": 4.948288171254089e-05, "grad_norm": 1.8763965368270874, "loss_scale": 4.0, "retries": 0, "seconds": 32.6, "max_memory_gb": 22.47, "pair": 0.6296818256378174, "sign": 0.18208481499459594, "keep": 0.07755745947360992, "leak": 0.2686000873800367, "replacement": 0.8266529589891434} +{"step": 34, "steps": 250, "lr": 4.941651146529998e-05, "grad_norm": 2.206883430480957, "loss_scale": 4.0, "retries": 0, "seconds": 33.6, "max_memory_gb": 22.47, "pair": 0.2820870988070965, "sign": 0.016889105085283518, "keep": 0.004759982108225813, "leak": 1.5293505489826202, "replacement": 0.6752492934465408} +{"step": 35, "steps": 250, "lr": 4.9346190892086174e-05, "grad_norm": 2.0480573177337646, "loss_scale": 4.0, "retries": 0, "seconds": 34.5, "max_memory_gb": 22.47, "pair": 0.31426812149584293, "sign": 0.04239358124323189, "keep": 0.006854727282188833, "leak": 1.1869618371129036, "replacement": 0.6356327384710312} +{"step": 36, "steps": 250, "lr": 4.927193266778101e-05, "grad_norm": 4.562411785125732, "loss_scale": 4.0, "retries": 0, "seconds": 35.6, "max_memory_gb": 22.47, "pair": 0.3033669739961624, "sign": 0.027822130476124585, "keep": 0.1604781405258109, "leak": 1.9439297076314688, "replacement": 0.8834995180368423} +{"step": 37, "steps": 250, "lr": 4.919375017700515e-05, "grad_norm": 2.3277838230133057, "loss_scale": 4.0, "retries": 0, "seconds": 36.5, "max_memory_gb": 22.47, "pair": 0.4659881517291069, "sign": 0.1816174171399325, "keep": 0.4137839809700381, "leak": 1.1382531968411058, "replacement": 1.0482521802186966} +{"step": 38, "steps": 250, "lr": 4.911165751170588e-05, "grad_norm": 4.068784236907959, "loss_scale": 4.0, "retries": 0, "seconds": 37.5, "max_memory_gb": 22.47, "pair": 0.1838555485010147, "sign": 0.07399882923346013, "keep": 0.02751076756976545, "leak": 2.160587877035141, "replacement": 0.7747573181986809} +{"step": 39, "steps": 250, "lr": 4.902566946861708e-05, "grad_norm": 3.005831480026245, "loss_scale": 4.0, "retries": 0, "seconds": 38.5, "max_memory_gb": 22.47, "pair": 0.5486574806272984, "sign": 1.4048132409807295, "keep": 0.21290014296027948, "leak": 0.5316276336088777, "replacement": 1.4904210790991783} +{"step": 40, "steps": 250, "lr": 4.89358015465922e-05, "grad_norm": 1.8836328983306885, "loss_scale": 4.0, "retries": 0, "seconds": 39.4, "max_memory_gb": 22.47, "pair": 0.20623351261019707, "sign": 0.009924400015734136, "keep": 0.008377456470043398, "leak": 1.2266164496541023, "replacement": 0.5220385417342186} +{"step": 41, "steps": 250, "lr": 4.884206994381078e-05, "grad_norm": 4.040044784545898, "loss_scale": 4.0, "retries": 0, "seconds": 40.4, "max_memory_gb": 22.47, "pair": 0.6570723652839661, "sign": 0.9041169937700033, "keep": 1.0253198085119948, "leak": 0.30438696779310703, "replacement": 1.6978875398635864} +{"step": 42, "steps": 250, "lr": 4.874449155485866e-05, "grad_norm": 2.6819229125976562, "loss_scale": 4.0, "retries": 0, "seconds": 41.3, "max_memory_gb": 22.47, "pair": 0.3181272856891155, "sign": 0.03606233175378293, "keep": 0.00188792618791922, "leak": 1.132969118654728, "replacement": 0.6203447133302689} +{"step": 43, "steps": 250, "lr": 4.864308396768294e-05, "grad_norm": 2.9244906902313232, "loss_scale": 4.0, "retries": 0, "seconds": 42.3, "max_memory_gb": 22.47, "pair": 0.3583528730086982, "sign": 0.14586490340298042, "keep": 0.004507413719693432, "leak": 1.7645361945033073, "replacement": 0.8746730834245682} +{"step": 44, "steps": 250, "lr": 4.853786546042184e-05, "grad_norm": 3.290605068206787, "loss_scale": 4.0, "retries": 0, "seconds": 43.3, "max_memory_gb": 22.47, "pair": 0.28759583830833435, "sign": 0.133673824951984, "keep": 0.7677991016407759, "leak": 0.9996977597475052, "replacement": 0.9882567524909973} +{"step": 45, "steps": 250, "lr": 4.842885499811012e-05, "grad_norm": 4.414028167724609, "loss_scale": 4.0, "retries": 0, "seconds": 44.2, "max_memory_gb": 22.47, "pair": 0.6260259672999382, "sign": 0.6819836439099163, "keep": 1.8760437219461892, "leak": 0.3483426681050332, "replacement": 1.9921253100037575} +{"step": 46, "steps": 250, "lr": 4.831607222926078e-05, "grad_norm": 2.5997214317321777, "loss_scale": 4.0, "retries": 0, "seconds": 45.2, "max_memory_gb": 22.47, "pair": 0.494153406471014, "sign": 0.25900528507190756, "keep": 0.4558953787200153, "leak": 0.4743232000619173, "replacement": 0.9701845496892929} +{"step": 47, "steps": 250, "lr": 4.8199537482323545e-05, "grad_norm": 1.2931751012802124, "loss_scale": 4.0, "retries": 0, "seconds": 46.2, "max_memory_gb": 22.47, "pair": 0.40733006969094276, "sign": 0.08677277975948527, "keep": 0.011317621087073348, "leak": 0.7044248001184314, "replacement": 0.6324814707040787} +{"step": 48, "steps": 250, "lr": 4.807927176202073e-05, "grad_norm": 2.3986644744873047, "loss_scale": 4.0, "retries": 0, "seconds": 47.2, "max_memory_gb": 22.47, "pair": 0.25566878728568554, "sign": 0.023763209726894274, "keep": 0.022013326961314306, "leak": 1.2071193903684616, "replacement": 0.5803369134664536} +{"step": 49, "steps": 250, "lr": 4.795529674556129e-05, "grad_norm": 1.0701956748962402, "loss_scale": 4.0, "retries": 0, "seconds": 48.2, "max_memory_gb": 22.47, "pair": 0.299320463091135, "sign": 0.03680013678967953, "keep": 0.028007477656501578, "leak": 0.6365969106554985, "replacement": 0.4908734932541847} +{"step": 50, "steps": 250, "lr": 4.782763477873361e-05, "grad_norm": 1.4721975326538086, "loss_scale": 4.0, "retries": 0, "seconds": 49.1, "max_memory_gb": 22.47, "pair": 0.4147282987833023, "sign": 0.08053645677864552, "keep": 0.23333204865048174, "leak": 0.5286183133721352, "replacement": 0.7038171589374542} +{"step": 51, "steps": 250, "lr": 4.769630887187782e-05, "grad_norm": 3.1966516971588135, "loss_scale": 4.0, "retries": 0, "seconds": 50.1, "max_memory_gb": 22.47, "pair": 0.4075564332306385, "sign": 0.02668271667789668, "keep": 1.080528463590781, "leak": 0.2921326602809131, "replacement": 1.0341951698064804} +{"step": 52, "steps": 250, "lr": 4.756134269573832e-05, "grad_norm": 2.104809522628784, "loss_scale": 4.0, "retries": 0, "seconds": 51.0, "max_memory_gb": 22.47, "pair": 0.49079275876283646, "sign": 0.2625255885068327, "keep": 0.12225586373824626, "leak": 0.2707496266812086, "replacement": 0.7508708909153938} +{"step": 53, "steps": 250, "lr": 4.7422760577197226e-05, "grad_norm": 2.0371813774108887, "loss_scale": 4.0, "retries": 0, "seconds": 52.0, "max_memory_gb": 22.47, "pair": 0.4193918965756893, "sign": 0.08282018726458773, "keep": 0.3044842903036624, "leak": 0.7481655832380056, "replacement": 0.8000855296850204} +{"step": 54, "steps": 250, "lr": 4.728058749488963e-05, "grad_norm": 3.15482234954834, "loss_scale": 4.0, "retries": 0, "seconds": 53.0, "max_memory_gb": 22.47, "pair": 0.46049900352954865, "sign": 0.1279942110995762, "keep": 0.27846754278289154, "leak": 0.6359229404479265, "replacement": 0.8227105960249901} +{"step": 55, "steps": 250, "lr": 4.713484907470133e-05, "grad_norm": 1.591442584991455, "loss_scale": 4.0, "retries": 0, "seconds": 53.9, "max_memory_gb": 22.47, "pair": 0.3687315136194229, "sign": 0.02736759907566011, "keep": 0.014726364723173901, "leak": 0.38916096836328506, "replacement": 0.487068735063076} +{"step": 56, "steps": 250, "lr": 4.6985571585149876e-05, "grad_norm": 1.4455112218856812, "loss_scale": 4.0, "retries": 0, "seconds": 54.9, "max_memory_gb": 22.47, "pair": 0.34569762647151947, "sign": 0.015773925464600325, "keep": 0.021740300580859184, "leak": 0.5276986584067345, "replacement": 0.49637941271066666} +{"step": 57, "steps": 250, "lr": 4.6832781932649884e-05, "grad_norm": 1.9669941663742065, "loss_scale": 4.0, "retries": 0, "seconds": 55.8, "max_memory_gb": 22.47, "pair": 0.38517168536782265, "sign": 0.31683653825894, "keep": 0.008810573945083888, "leak": 0.48969448800198734, "replacement": 0.6704188659787178} +{"step": 58, "steps": 250, "lr": 4.667650765666325e-05, "grad_norm": 1.7499727010726929, "loss_scale": 4.0, "retries": 0, "seconds": 56.8, "max_memory_gb": 22.47, "pair": 0.17144225910305977, "sign": 0.010775470815133303, "keep": 0.01799231404584134, "leak": 1.019098088145256, "replacement": 0.4406006783246994} +{"step": 59, "steps": 250, "lr": 4.651677692473538e-05, "grad_norm": 2.031015634536743, "loss_scale": 4.0, "retries": 0, "seconds": 57.8, "max_memory_gb": 22.47, "pair": 0.4228165131062269, "sign": 0.047037490468937904, "keep": 0.026519646344240755, "leak": 0.44436647999100387, "replacement": 0.5706866979598999} +{"step": 60, "steps": 250, "lr": 4.635361852741808e-05, "grad_norm": 2.0334701538085938, "loss_scale": 4.0, "retries": 0, "seconds": 58.7, "max_memory_gb": 22.47, "pair": 0.3670830652117729, "sign": 0.05175642832182348, "keep": 0.11235483418568037, "leak": 0.4821250746026635, "replacement": 0.5696699693799019} +{"step": 61, "steps": 250, "lr": 4.6187061873080264e-05, "grad_norm": 4.0974202156066895, "loss_scale": 4.0, "retries": 0, "seconds": 59.7, "max_memory_gb": 22.47, "pair": 0.42641996033489704, "sign": 0.24978598672896624, "keep": 0.029894673090893775, "leak": 1.965950921177864, "replacement": 1.0577480494976044} +{"step": 62, "steps": 250, "lr": 4.601713698260727e-05, "grad_norm": 1.223924994468689, "loss_scale": 4.0, "retries": 0, "seconds": 60.7, "max_memory_gb": 22.47, "pair": 0.2934481706470251, "sign": 0.024288140702992678, "keep": 0.048930991237284616, "leak": 0.6768360957503319, "replacement": 0.49926677346229553} +{"step": 63, "steps": 250, "lr": 4.5843874483989746e-05, "grad_norm": 1.8300851583480835, "loss_scale": 4.0, "retries": 0, "seconds": 61.7, "max_memory_gb": 22.47, "pair": 0.34978548251092434, "sign": 0.12106845597736537, "keep": 0.04436213709414005, "leak": 0.22635666746646166, "replacement": 0.4890899658203125} +{"step": 64, "steps": 250, "lr": 4.566730560680316e-05, "grad_norm": 1.4406373500823975, "loss_scale": 4.0, "retries": 0, "seconds": 62.7, "max_memory_gb": 22.47, "pair": 0.27937429398298264, "sign": 0.0176014764001593, "keep": 0.022692125057801604, "leak": 0.4481367114931345, "replacement": 0.4115552604198456} +{"step": 65, "steps": 250, "lr": 4.548746217657878e-05, "grad_norm": 4.981435775756836, "loss_scale": 4.0, "retries": 0, "seconds": 63.7, "max_memory_gb": 22.47, "pair": 0.3713562823832035, "sign": 0.32142779987771064, "keep": 0.25265476666390896, "leak": 1.8152275681495667, "replacement": 1.1122044622898102} +{"step": 66, "steps": 250, "lr": 4.5304376609067396e-05, "grad_norm": 2.642960786819458, "loss_scale": 4.0, "retries": 0, "seconds": 64.7, "max_memory_gb": 22.47, "pair": 0.2843934064731002, "sign": 0.14442169308313169, "keep": 0.33758156493421154, "leak": 0.6960638333112001, "replacement": 0.699411004781723} +{"step": 67, "steps": 250, "lr": 4.5118081904396504e-05, "grad_norm": 3.3718087673187256, "loss_scale": 4.0, "retries": 0, "seconds": 65.6, "max_memory_gb": 22.47, "pair": 0.09838092979043722, "sign": 0.013971138745546341, "keep": 0.010028946388047189, "leak": 1.6106905192136765, "replacement": 0.5130535997450352} +{"step": 68, "steps": 250, "lr": 4.4928611641122225e-05, "grad_norm": 1.4617969989776611, "loss_scale": 4.0, "retries": 0, "seconds": 66.7, "max_memory_gb": 22.47, "pair": 0.2867697961628437, "sign": 0.03428525966592133, "keep": 0.007825075064829434, "leak": 0.6724636182188988, "replacement": 0.47594086825847626} +{"step": 69, "steps": 250, "lr": 4.473599997017701e-05, "grad_norm": 4.374925136566162, "loss_scale": 4.0, "retries": 0, "seconds": 67.6, "max_memory_gb": 22.47, "pair": 0.4829827081412077, "sign": 0.8332525317091495, "keep": 0.7614923171349801, "leak": 1.0028331615030766, "replacement": 1.5310634076595306} +{"step": 70, "steps": 250, "lr": 4.4540281608714053e-05, "grad_norm": 5.395942687988281, "loss_scale": 4.0, "retries": 0, "seconds": 68.6, "max_memory_gb": 22.47, "pair": 0.2506913581164554, "sign": 0.02873915186501108, "keep": 0.00391879107337445, "leak": 2.5940356254577637, "replacement": 0.9155292138457298} +{"step": 71, "steps": 250, "lr": 4.434149183384977e-05, "grad_norm": 2.1665632724761963, "loss_scale": 4.0, "retries": 0, "seconds": 69.6, "max_memory_gb": 22.47, "pair": 0.11307188309729099, "sign": 0.006546748481923714, "keep": 0.017666956679704526, "leak": 1.0047122240066528, "replacement": 0.3763567805290222} +{"step": 72, "steps": 250, "lr": 4.413966647630534e-05, "grad_norm": 3.7986831665039062, "loss_scale": 4.0, "retries": 0, "seconds": 70.6, "max_memory_gb": 22.47, "pair": 0.24807167518883944, "sign": 0.1872993737924844, "keep": 0.03307114681228995, "leak": 1.6059258244931698, "replacement": 0.759738378226757} +{"step": 73, "steps": 250, "lr": 4.3934841913948323e-05, "grad_norm": 2.4193966388702393, "loss_scale": 4.0, "retries": 0, "seconds": 71.6, "max_memory_gb": 22.47, "pair": 0.3141900338232517, "sign": 0.1110600300016813, "keep": 0.5279251211504743, "leak": 0.4674838297069073, "replacement": 0.7505535706877708} +{"step": 74, "steps": 250, "lr": 4.372705506523583e-05, "grad_norm": 3.5810134410858154, "loss_scale": 4.0, "retries": 0, "seconds": 72.5, "max_memory_gb": 22.47, "pair": 0.29664832819253206, "sign": 0.22972045303322375, "keep": 0.12438984581626755, "leak": 1.123958467040211, "replacement": 0.7546931095421314} +{"step": 75, "steps": 250, "lr": 4.351634338256017e-05, "grad_norm": 2.23272442817688, "loss_scale": 4.0, "retries": 0, "seconds": 73.5, "max_memory_gb": 22.47, "pair": 0.6776081919670105, "sign": 0.5080807954072952, "keep": 0.8801386873819865, "leak": 0.12127476419846062, "replacement": 1.4020366370677948} +{"step": 76, "steps": 250, "lr": 4.330274484549825e-05, "grad_norm": 1.0989350080490112, "loss_scale": 4.0, "retries": 0, "seconds": 74.5, "max_memory_gb": 22.47, "pair": 0.23733632825315, "sign": 0.07952532591298223, "keep": 0.015381849414552562, "leak": 0.6323736757040024, "replacement": 0.44288333505392075} +{"step": 77, "steps": 250, "lr": 4.308629795396599e-05, "grad_norm": 1.6755468845367432, "loss_scale": 4.0, "retries": 0, "seconds": 75.5, "max_memory_gb": 22.47, "pair": 0.21803049743175507, "sign": 0.0093165033031255, "keep": 0.005062087046098895, "leak": 1.0693878382444382, "replacement": 0.4925667494535446} +{"step": 78, "steps": 250, "lr": 4.286704172127894e-05, "grad_norm": 2.4542107582092285, "loss_scale": 4.0, "retries": 0, "seconds": 76.4, "max_memory_gb": 22.47, "pair": 0.32268129847943783, "sign": 0.48791392450220883, "keep": 0.0014468430308625102, "leak": 0.4865847136825323, "replacement": 0.689007893204689} +{"step": 79, "steps": 250, "lr": 4.264501566712033e-05, "grad_norm": 2.007857322692871, "loss_scale": 4.0, "retries": 0, "seconds": 77.3, "max_memory_gb": 22.47, "pair": 0.28146689385175705, "sign": 0.0191967444261536, "keep": 0.02300948323454577, "leak": 0.8350571319460869, "replacement": 0.5113342851400375} +{"step": 80, "steps": 250, "lr": 4.2420259810417894e-05, "grad_norm": 1.6659644842147827, "loss_scale": 4.0, "retries": 0, "seconds": 78.3, "max_memory_gb": 22.47, "pair": 0.406850203871727, "sign": 0.09271887433715165, "keep": 0.017530228524265112, "leak": 0.5759947188198566, "replacement": 0.6059734299778938} +{"step": 81, "steps": 250, "lr": 4.219281466213066e-05, "grad_norm": 1.431257724761963, "loss_scale": 4.0, "retries": 0, "seconds": 79.2, "max_memory_gb": 22.47, "pair": 0.26922452077269554, "sign": 0.05498247957439162, "keep": 0.11257132003083825, "leak": 0.6407861970365047, "replacement": 0.5131979882717133} +{"step": 82, "steps": 250, "lr": 4.1962721217947135e-05, "grad_norm": 2.6722936630249023, "loss_scale": 4.0, "retries": 0, "seconds": 80.2, "max_memory_gb": 22.47, "pair": 0.3634904660284519, "sign": 1.0254595937440172, "keep": 0.030658308183774352, "leak": 0.8677275143563747, "replacement": 1.1084812805056572} +{"step": 83, "steps": 250, "lr": 4.173002095089599e-05, "grad_norm": 2.2164998054504395, "loss_scale": 4.0, "retries": 0, "seconds": 81.2, "max_memory_gb": 22.47, "pair": 0.37528627179563046, "sign": 0.03130454596248455, "keep": 0.3527137590572238, "leak": 0.7907123826444149, "replacement": 0.764973521232605} +{"step": 84, "steps": 250, "lr": 4.149475580387084e-05, "grad_norm": 1.4421409368515015, "loss_scale": 4.0, "retries": 0, "seconds": 82.2, "max_memory_gb": 22.47, "pair": 0.21386543661355972, "sign": 0.027411341085098684, "keep": 0.033039310474862305, "leak": 0.43437589704990387, "replacement": 0.35268473252654076} +{"step": 85, "steps": 250, "lr": 4.1256968182070275e-05, "grad_norm": 3.7614872455596924, "loss_scale": 4.0, "retries": 0, "seconds": 83.2, "max_memory_gb": 22.47, "pair": 0.11694046668708324, "sign": 0.027001339360140264, "keep": 0.02827171966566766, "leak": 0.6346080675721169, "replacement": 0.3032290190458298} +{"step": 86, "steps": 250, "lr": 4.101670094535453e-05, "grad_norm": 2.954663038253784, "loss_scale": 4.0, "retries": 0, "seconds": 84.1, "max_memory_gb": 22.47, "pair": 0.3418860696256161, "sign": 0.28488594479858875, "keep": 0.21120923664420843, "leak": 0.3114663492888212, "replacement": 0.6678002402186394} +{"step": 87, "steps": 250, "lr": 4.077399740052027e-05, "grad_norm": 1.3123258352279663, "loss_scale": 4.0, "retries": 0, "seconds": 85.1, "max_memory_gb": 22.47, "pair": 0.29566166922450066, "sign": 0.10055101593025029, "keep": 0.007708890407229774, "leak": 0.7838714569807053, "replacement": 0.5457594841718674} +{"step": 88, "steps": 250, "lr": 4.0528901293494775e-05, "grad_norm": 3.8202130794525146, "loss_scale": 4.0, "retries": 0, "seconds": 86.1, "max_memory_gb": 22.47, "pair": 0.3488040240481496, "sign": 0.21087434701621532, "keep": 0.050272441992262884, "leak": 0.6811246247962117, "replacement": 0.6496585607528687} +{"step": 89, "steps": 250, "lr": 4.0281456801451004e-05, "grad_norm": 3.6769754886627197, "loss_scale": 4.0, "retries": 0, "seconds": 87.0, "max_memory_gb": 22.47, "pair": 0.435259360820055, "sign": 0.15935737208928913, "keep": 0.012386539731096491, "leak": 0.009513043332844973, "replacement": 0.5235095825046301} +{"step": 90, "steps": 250, "lr": 4.0031708524844884e-05, "grad_norm": 2.2221415042877197, "loss_scale": 4.0, "retries": 0, "seconds": 87.9, "max_memory_gb": 22.47, "pair": 0.255153376609087, "sign": 0.0629493790329434, "keep": 0.036095662537263706, "leak": 0.6271671475842595, "replacement": 0.4614676982164383} +{"step": 91, "steps": 250, "lr": 3.977970147937635e-05, "grad_norm": 7.372348785400391, "loss_scale": 4.0, "retries": 0, "seconds": 88.9, "max_memory_gb": 22.47, "pair": 0.6087959706783295, "sign": 0.3167545923497528, "keep": 0.47695974877569824, "leak": 0.5573366209864616, "replacement": 1.1449873000383377} +{"step": 92, "steps": 250, "lr": 3.952548108787553e-05, "grad_norm": 5.8780012130737305, "loss_scale": 4.0, "retries": 0, "seconds": 89.9, "max_memory_gb": 22.47, "pair": 0.2958667743951082, "sign": 0.6186887419898994, "keep": 0.9678093002657988, "leak": 0.4210699461400509, "replacement": 1.1943833231925964} +{"step": 93, "steps": 250, "lr": 3.9269093172115496e-05, "grad_norm": 1.529744029045105, "loss_scale": 4.0, "retries": 0, "seconds": 90.9, "max_memory_gb": 22.47, "pair": 0.22913341503590345, "sign": 0.029217388248071074, "keep": 0.10160012054257095, "leak": 0.4697549639386125, "replacement": 0.41198091208934784} +{"step": 94, "steps": 250, "lr": 3.901058394455323e-05, "grad_norm": 2.693861484527588, "loss_scale": 4.0, "retries": 0, "seconds": 91.9, "max_memory_gb": 22.47, "pair": 0.220508998259902, "sign": 0.02754763513803482, "keep": 0.3975956537760794, "leak": 0.8391760736703873, "replacement": 0.642874650657177} +{"step": 95, "steps": 250, "lr": 3.875e-05, "grad_norm": 3.665132999420166, "loss_scale": 4.0, "retries": 0, "seconds": 92.8, "max_memory_gb": 22.47, "pair": 0.3863615356385708, "sign": 0.3100882776780054, "keep": 1.3536044906359166, "leak": 0.7251694747401558, "replacement": 1.3995002061128616} +{"step": 96, "steps": 250, "lr": 3.8487388307223005e-05, "grad_norm": 2.5585780143737793, "loss_scale": 4.0, "retries": 0, "seconds": 93.8, "max_memory_gb": 22.47, "pair": 0.30770308524370193, "sign": 0.09077915665693581, "keep": 0.9980125861475244, "leak": 0.5815591216087341, "replacement": 0.9974887371063232} +{"step": 97, "steps": 250, "lr": 3.822279620047943e-05, "grad_norm": 2.6842548847198486, "loss_scale": 4.0, "retries": 0, "seconds": 94.8, "max_memory_gb": 22.47, "pair": 0.1930177267640829, "sign": 0.011216133134439588, "keep": 0.014818788177763054, "leak": 0.9589677741751075, "replacement": 0.44577714055776596} +{"step": 98, "steps": 250, "lr": 3.795627137098479e-05, "grad_norm": 1.5132273435592651, "loss_scale": 4.0, "retries": 0, "seconds": 95.8, "max_memory_gb": 22.47, "pair": 0.3117334824055433, "sign": 0.10060231038369238, "keep": 0.02874613995254549, "leak": 0.6650808602571487, "replacement": 0.5426779165863991} +{"step": 99, "steps": 250, "lr": 3.7687861858316844e-05, "grad_norm": 2.3175101280212402, "loss_scale": 4.0, "retries": 0, "seconds": 96.7, "max_memory_gb": 22.47, "pair": 0.4333016723394394, "sign": 0.6381063445587642, "keep": 0.014839030534858466, "leak": 0.5000635026954114, "replacement": 0.8847902864217758} +{"step": 100, "steps": 250, "lr": 3.741761604175668e-05, "grad_norm": 4.512275695800781, "loss_scale": 4.0, "retries": 0, "seconds": 97.7, "max_memory_gb": 22.47, "pair": 0.5867851935327053, "sign": 0.9500229818513617, "keep": 0.31230735778808594, "leak": 1.0799845203291625, "replacement": 1.4879464507102966} +{"step": 101, "steps": 250, "lr": 3.714558263156872e-05, "grad_norm": 8.044857025146484, "loss_scale": 4.0, "retries": 0, "seconds": 98.7, "max_memory_gb": 22.47, "pair": 0.10748547874391079, "sign": 0.006655465986113995, "keep": 0.008965601984527893, "leak": 3.023072600364685, "replacement": 0.8710641711950302} +{"step": 102, "steps": 250, "lr": 3.6871810660220904e-05, "grad_norm": 1.8485231399536133, "loss_scale": 4.0, "retries": 0, "seconds": 99.7, "max_memory_gb": 22.47, "pair": 0.18076730519533157, "sign": 0.015409346437081695, "keep": 0.07454196223989129, "leak": 0.46219844510778785, "replacement": 0.341292567551136} +{"step": 103, "steps": 250, "lr": 3.659634947354686e-05, "grad_norm": 1.6652703285217285, "loss_scale": 4.0, "retries": 0, "seconds": 100.7, "max_memory_gb": 22.47, "pair": 0.17367370799183846, "sign": 0.023678093450143933, "keep": 0.015017390984212398, "leak": 0.8953152634203434, "replacement": 0.4168502688407898} +{"step": 104, "steps": 250, "lr": 3.631924872185169e-05, "grad_norm": 2.7420990467071533, "loss_scale": 4.0, "retries": 0, "seconds": 101.6, "max_memory_gb": 22.47, "pair": 0.1968942703679204, "sign": 0.18089865293586627, "keep": 0.004725344537291676, "leak": 0.32036974281072617, "replacement": 0.3697987049818039} +{"step": 105, "steps": 250, "lr": 3.6040558350962674e-05, "grad_norm": 2.393385887145996, "loss_scale": 4.0, "retries": 0, "seconds": 102.6, "max_memory_gb": 22.47, "pair": 0.29561082273721695, "sign": 0.01798100513406098, "keep": 0.07818719535134733, "leak": 0.6440915130078793, "replacement": 0.5047178119421005} +{"step": 106, "steps": 250, "lr": 3.576032859322687e-05, "grad_norm": 2.1079611778259277, "loss_scale": 4.0, "retries": 0, "seconds": 103.5, "max_memory_gb": 22.47, "pair": 0.2958727292716503, "sign": 0.03122271061874926, "keep": 0.017688920299406163, "leak": 0.317491065710783, "replacement": 0.399701327085495} +{"step": 107, "steps": 250, "lr": 3.5478609958457056e-05, "grad_norm": 5.005122184753418, "loss_scale": 4.0, "retries": 0, "seconds": 104.6, "max_memory_gb": 22.47, "pair": 0.1744301300495863, "sign": 0.259364630561322, "keep": 1.4091492321110763, "leak": 0.4558087522163987, "replacement": 1.1226392686367035} +{"step": 108, "steps": 250, "lr": 3.519545322482755e-05, "grad_norm": 4.6895599365234375, "loss_scale": 4.0, "retries": 0, "seconds": 105.5, "max_memory_gb": 22.47, "pair": 0.41335812769830227, "sign": 0.18695969542022794, "keep": 0.037374097446445376, "leak": 0.4869670197367668, "replacement": 0.6472667530179024} +{"step": 109, "steps": 250, "lr": 3.491090942972179e-05, "grad_norm": 3.9227254390716553, "loss_scale": 4.0, "retries": 0, "seconds": 106.5, "max_memory_gb": 22.47, "pair": 0.3984033167362213, "sign": 0.10039415047504008, "keep": 0.07379021914675832, "leak": 0.6271680174395442, "replacement": 0.6422875225543976} +{"step": 110, "steps": 250, "lr": 3.4625029860533135e-05, "grad_norm": 4.15121603012085, "loss_scale": 4.0, "retries": 0, "seconds": 107.5, "max_memory_gb": 22.47, "pair": 0.2544621303677559, "sign": 0.08380361134186387, "keep": 0.09465229324996471, "leak": 1.3215236030519009, "replacement": 0.6740709915757179} +{"step": 111, "steps": 250, "lr": 3.433786604542056e-05, "grad_norm": 4.744983196258545, "loss_scale": 4.0, "retries": 0, "seconds": 108.5, "max_memory_gb": 22.47, "pair": 0.2680764766409993, "sign": 1.2392094070019084, "keep": 0.0012052970851073042, "leak": 0.805219005793333, "replacement": 1.089588612318039} +{"step": 112, "steps": 250, "lr": 3.404946974402104e-05, "grad_norm": 6.195766925811768, "loss_scale": 4.0, "retries": 0, "seconds": 109.5, "max_memory_gb": 22.47, "pair": 0.08880535699427128, "sign": 0.006964365020394325, "keep": 0.006542768860526849, "leak": 1.6753022372722626, "replacement": 0.514384500682354} +{"step": 113, "steps": 250, "lr": 3.375989293812019e-05, "grad_norm": 1.3672746419906616, "loss_scale": 4.0, "retries": 0, "seconds": 110.5, "max_memory_gb": 22.47, "pair": 0.20282307267189026, "sign": 0.027093791984952986, "keep": 0.011267653142567724, "leak": 0.43197590857744217, "replacement": 0.3299977779388428} +{"step": 114, "steps": 250, "lr": 3.3469187822282766e-05, "grad_norm": 1.9527887105941772, "loss_scale": 4.0, "retries": 0, "seconds": 111.5, "max_memory_gb": 22.47, "pair": 0.24243143387138844, "sign": 0.039039526134729385, "keep": 0.02044201421085745, "leak": 0.9111654758453369, "replacement": 0.49996358156204224} +{"step": 115, "steps": 250, "lr": 3.317740679444499e-05, "grad_norm": 1.6212341785430908, "loss_scale": 4.0, "retries": 0, "seconds": 112.5, "max_memory_gb": 22.47, "pair": 0.19910239055752754, "sign": 0.010186201659962535, "keep": 0.006773726290703053, "leak": 0.370240718126297, "replacement": 0.3001425340771675} +{"step": 116, "steps": 250, "lr": 3.2884602446470056e-05, "grad_norm": 7.0820136070251465, "loss_scale": 4.0, "retries": 0, "seconds": 113.5, "max_memory_gb": 22.47, "pair": 0.18961476907134056, "sign": 0.5849013713304885, "keep": 0.7090014755740413, "leak": 0.31271547079086304, "replacement": 0.9147450588643551} +{"step": 117, "steps": 250, "lr": 3.259082755466876e-05, "grad_norm": 11.197393417358398, "loss_scale": 4.0, "retries": 0, "seconds": 114.5, "max_memory_gb": 22.5, "pair": 0.09451277647167444, "sign": 0.19115784257883206, "keep": 0.6172815099907893, "leak": 2.0907481908798218, "replacement": 1.0214194804430008} +{"step": 118, "steps": 250, "lr": 3.229613507028686e-05, "grad_norm": 2.338719606399536, "loss_scale": 4.0, "retries": 0, "seconds": 115.5, "max_memory_gb": 22.5, "pair": 0.2861657505854964, "sign": 0.33871301473118365, "keep": 0.4727153680942138, "leak": 0.2714526429781472, "replacement": 0.7597431223839521} +{"step": 119, "steps": 250, "lr": 3.2000578109961004e-05, "grad_norm": 4.4538187980651855, "loss_scale": 4.0, "retries": 0, "seconds": 116.5, "max_memory_gb": 22.5, "pair": 0.3208998437039554, "sign": 0.23216941294958815, "keep": 0.016329740033597773, "leak": 0.23511203657835722, "replacement": 0.5039274096488953} +{"step": 120, "steps": 250, "lr": 3.1704209946144616e-05, "grad_norm": 5.839120388031006, "loss_scale": 4.0, "retries": 0, "seconds": 117.5, "max_memory_gb": 22.5, "pair": 0.36453709192574024, "sign": 0.8130034029018134, "keep": 1.6310769692063332, "leak": 0.3062378242611885, "replacement": 1.663136824965477} +{"step": 121, "steps": 250, "lr": 3.140708399750594e-05, "grad_norm": 2.8787877559661865, "loss_scale": 4.0, "retries": 0, "seconds": 118.4, "max_memory_gb": 22.5, "pair": 0.4686766490340233, "sign": 0.3738223838736303, "keep": 0.08084749220233789, "leak": 0.39195364946499467, "replacement": 0.794000007212162} +{"step": 122, "steps": 250, "lr": 3.1109253819299606e-05, "grad_norm": 3.003690719604492, "loss_scale": 4.0, "retries": 0, "seconds": 119.4, "max_memory_gb": 22.5, "pair": 0.31541976146399975, "sign": 0.09646268034703098, "keep": 0.05612604468569771, "leak": 0.5895537175238132, "replacement": 0.5391025580465794} +{"step": 123, "steps": 250, "lr": 3.081077309371361e-05, "grad_norm": 4.162919044494629, "loss_scale": 4.0, "retries": 0, "seconds": 120.4, "max_memory_gb": 22.5, "pair": 0.37097522988915443, "sign": 1.161878328770399, "keep": 0.04920610511908308, "leak": 0.620215502276551, "replacement": 1.131571363657713} +{"step": 124, "steps": 250, "lr": 3.051169562019339e-05, "grad_norm": 3.2433996200561523, "loss_scale": 4.0, "retries": 0, "seconds": 121.4, "max_memory_gb": 22.5, "pair": 0.49688766710460186, "sign": 0.6143439747393131, "keep": 1.9931219025165774, "leak": 0.2758263094583526, "replacement": 1.8695771768689156} +{"step": 125, "steps": 250, "lr": 3.021207530574477e-05, "grad_norm": 5.876475811004639, "loss_scale": 4.0, "retries": 0, "seconds": 122.3, "max_memory_gb": 22.5, "pair": 0.28460093727335334, "sign": 0.23042992502450943, "keep": 0.1892852905871223, "leak": 1.664309149608016, "replacement": 0.9105358198285103} +{"step": 126, "steps": 250, "lr": 2.991196615521758e-05, "grad_norm": 5.641046047210693, "loss_scale": 4.0, "retries": 0, "seconds": 123.3, "max_memory_gb": 22.5, "pair": 0.5863587949424982, "sign": 0.5117933886649553, "keep": 1.1057784976437688, "leak": 1.2851690474635689, "replacement": 1.7164369225502014} +{"step": 127, "steps": 250, "lr": 2.9611422261571536e-05, "grad_norm": 3.6784911155700684, "loss_scale": 4.0, "retries": 0, "seconds": 124.2, "max_memory_gb": 22.5, "pair": 0.14902185276150703, "sign": 0.008755643975746352, "keep": 0.031207511060074467, "leak": 0.6876524174585938, "replacement": 0.34091653674840927} +{"step": 128, "steps": 250, "lr": 2.931049779612634e-05, "grad_norm": 6.067586898803711, "loss_scale": 4.0, "retries": 0, "seconds": 125.2, "max_memory_gb": 22.5, "pair": 0.22712120413780212, "sign": 0.8422537782753352, "keep": 0.8357942475704476, "leak": 0.5290762074291706, "replacement": 1.1984142661094666} +{"step": 129, "steps": 250, "lr": 2.9009246998797557e-05, "grad_norm": 1.8325787782669067, "loss_scale": 4.0, "retries": 0, "seconds": 126.1, "max_memory_gb": 22.5, "pair": 0.2623368063941598, "sign": 0.0792168538318947, "keep": 0.1365291252732277, "leak": 0.5123280938714743, "replacement": 0.4982918053865433} +{"step": 130, "steps": 250, "lr": 2.8707724168320238e-05, "grad_norm": 3.2212419509887695, "loss_scale": 4.0, "retries": 0, "seconds": 127.1, "max_memory_gb": 22.5, "pair": 0.4249061085283756, "sign": 1.5470816027373075, "keep": 0.052335710934130475, "leak": 0.13923338428139687, "replacement": 1.2594231441617012} +{"step": 131, "steps": 250, "lr": 2.840598365246184e-05, "grad_norm": 10.996794700622559, "loss_scale": 4.0, "retries": 0, "seconds": 128.0, "max_memory_gb": 22.5, "pair": 0.6371680232696235, "sign": 0.47313288832083344, "keep": 0.7829487692370094, "leak": 1.1372092319652438, "replacement": 1.5495111644268036} +{"step": 132, "steps": 250, "lr": 2.8104079838226383e-05, "grad_norm": 6.9646711349487305, "loss_scale": 4.0, "retries": 0, "seconds": 129.0, "max_memory_gb": 22.5, "pair": 0.09743285283911973, "sign": 0.012395607132930309, "keep": 0.4333677589470426, "leak": 1.5820913407951593, "replacement": 0.715837387368083} +{"step": 133, "steps": 250, "lr": 2.7802067142051446e-05, "grad_norm": 2.613037347793579, "loss_scale": 4.0, "retries": 0, "seconds": 129.9, "max_memory_gb": 22.5, "pair": 0.48375737108290195, "sign": 0.5459783934056759, "keep": 0.9478803626261652, "leak": 0.5040053278207779, "replacement": 1.3566881120204926} +{"step": 134, "steps": 250, "lr": 2.7500000000000004e-05, "grad_norm": 4.976212501525879, "loss_scale": 4.0, "retries": 0, "seconds": 130.9, "max_memory_gb": 22.5, "pair": 0.19781427551060915, "sign": 0.30731428749277256, "keep": 0.6370428341720071, "leak": 0.051915859803557396, "replacement": 0.682971824426204} +{"step": 135, "steps": 250, "lr": 2.7197932857948566e-05, "grad_norm": 3.374945640563965, "loss_scale": 4.0, "retries": 0, "seconds": 131.9, "max_memory_gb": 22.5, "pair": 0.1558429105207324, "sign": 0.016633686260320246, "keep": 0.07524934789398685, "leak": 0.901101168259629, "replacement": 0.42705971747636795} +{"step": 136, "steps": 250, "lr": 2.689592016177363e-05, "grad_norm": 4.6461944580078125, "loss_scale": 4.0, "retries": 0, "seconds": 132.8, "max_memory_gb": 22.5, "pair": 0.21053970977663994, "sign": 0.35892143729142845, "keep": 0.14493323673377745, "leak": 0.986449234187603, "replacement": 0.7090793251991272} +{"step": 137, "steps": 250, "lr": 2.659401634753817e-05, "grad_norm": 5.946985721588135, "loss_scale": 4.0, "retries": 0, "seconds": 133.8, "max_memory_gb": 22.5, "pair": 0.1277603767812252, "sign": 0.013695260044187307, "keep": 0.03581789384770673, "leak": 1.3949407525360584, "replacement": 0.5012521371245384} +{"step": 138, "steps": 250, "lr": 2.6292275831679764e-05, "grad_norm": 1.6845093965530396, "loss_scale": 4.0, "retries": 0, "seconds": 134.8, "max_memory_gb": 22.5, "pair": 0.1571248099207878, "sign": 0.02561325766146183, "keep": 0.08556065708398819, "leak": 0.43294400721788406, "replacement": 0.3209477625787258} +{"step": 139, "steps": 250, "lr": 2.5990753001202445e-05, "grad_norm": 3.8867716789245605, "loss_scale": 4.0, "retries": 0, "seconds": 135.8, "max_memory_gb": 22.5, "pair": 0.28465647622942924, "sign": 0.5179995505604893, "keep": 0.03809953556628898, "leak": 0.7472638208419085, "replacement": 0.7495220005512238} +{"step": 140, "steps": 250, "lr": 2.5689502203873672e-05, "grad_norm": 2.1735990047454834, "loss_scale": 4.0, "retries": 0, "seconds": 136.8, "max_memory_gb": 22.5, "pair": 0.1783245950937271, "sign": 0.03107497771270573, "keep": 0.02409830642864108, "leak": 0.16208912711590528, "replacement": 0.24643352627754211} +{"step": 141, "steps": 250, "lr": 2.538857773842847e-05, "grad_norm": 2.4767000675201416, "loss_scale": 4.0, "retries": 0, "seconds": 137.9, "max_memory_gb": 22.5, "pair": 0.27633730275556445, "sign": 0.22238430366269313, "keep": 0.008667797839734703, "leak": 0.5381414021467208, "replacement": 0.5263987146317959} +{"step": 142, "steps": 250, "lr": 2.5088033844782426e-05, "grad_norm": 3.630875587463379, "loss_scale": 4.0, "retries": 0, "seconds": 138.8, "max_memory_gb": 22.5, "pair": 0.32886077649891376, "sign": 0.4226692196680233, "keep": 0.34394319250714034, "leak": 0.20796500612050295, "replacement": 0.764158196747303} +{"step": 143, "steps": 250, "lr": 2.4787924694255237e-05, "grad_norm": 5.976757049560547, "loss_scale": 4.0, "retries": 0, "seconds": 139.8, "max_memory_gb": 22.5, "pair": 0.39970335830003023, "sign": 0.19750527542782947, "keep": 1.2432736594560083, "leak": 0.2678792607039213, "replacement": 1.1870626099407673} +{"step": 144, "steps": 250, "lr": 2.4488304379806627e-05, "grad_norm": 4.61199951171875, "loss_scale": 4.0, "retries": 0, "seconds": 140.8, "max_memory_gb": 22.5, "pair": 0.459850013256073, "sign": 0.6861231438815594, "keep": 0.08245279244147241, "leak": 0.32267383293947205, "replacement": 0.9248064756393433} +{"step": 145, "steps": 250, "lr": 2.4189226906286388e-05, "grad_norm": 1.9015763998031616, "loss_scale": 4.0, "retries": 0, "seconds": 141.8, "max_memory_gb": 22.5, "pair": 0.20939296716824174, "sign": 0.026569159323116764, "keep": 0.013734692058278597, "leak": 0.2301524542272091, "replacement": 0.28708299808204174} +{"step": 146, "steps": 250, "lr": 2.3890746180700395e-05, "grad_norm": 4.30116605758667, "loss_scale": 4.0, "retries": 0, "seconds": 142.8, "max_memory_gb": 22.5, "pair": 0.3667667591944337, "sign": 0.08153216796927154, "keep": 0.009740311361383647, "leak": 1.2331737610511482, "replacement": 0.7206964492797852} +{"step": 147, "steps": 250, "lr": 2.359291600249407e-05, "grad_norm": 5.170971393585205, "loss_scale": 4.0, "retries": 0, "seconds": 143.8, "max_memory_gb": 22.5, "pair": 0.5332091376185417, "sign": 1.0504153682850301, "keep": 1.3660024747550779, "leak": 1.2634041011333466, "replacement": 2.057269036769867} +{"step": 148, "steps": 250, "lr": 2.3295790053855392e-05, "grad_norm": 2.296870470046997, "loss_scale": 4.0, "retries": 0, "seconds": 144.8, "max_memory_gb": 22.5, "pair": 0.13106958055868745, "sign": 0.01585920824436471, "keep": 0.0022335037335778907, "leak": 0.4519911970710382, "replacement": 0.25311372336000204} +{"step": 149, "steps": 250, "lr": 2.299942189003901e-05, "grad_norm": 5.343630313873291, "loss_scale": 4.0, "retries": 0, "seconds": 145.7, "max_memory_gb": 22.5, "pair": 0.22235689428634942, "sign": 0.05628933277330361, "keep": 0.015203291492071003, "leak": 1.1449781525880098, "replacement": 0.5443477407097816} +{"step": 150, "steps": 250, "lr": 2.2703864929713144e-05, "grad_norm": 4.437504291534424, "loss_scale": 4.0, "retries": 0, "seconds": 146.7, "max_memory_gb": 22.5, "pair": 0.4093194529414177, "sign": 0.13370518968440592, "keep": 0.19600894022732973, "leak": 0.15212276950478554, "replacement": 0.6122071966528893} +{"step": 151, "steps": 250, "lr": 2.2409172445331257e-05, "grad_norm": 2.7957193851470947, "loss_scale": 4.0, "retries": 0, "seconds": 147.7, "max_memory_gb": 22.5, "pair": 0.2615441642701626, "sign": 0.022663619078230113, "keep": 0.020322406286140904, "leak": 0.0955274747684598, "replacement": 0.30691905319690704} +{"step": 152, "steps": 250, "lr": 2.211539755352996e-05, "grad_norm": 2.4455699920654297, "loss_scale": 4.0, "retries": 0, "seconds": 148.7, "max_memory_gb": 22.5, "pair": 0.27272676676511765, "sign": 0.008984313462860882, "keep": 0.03112951182265533, "leak": 0.31065240129828453, "replacement": 0.3704467974603176} +{"step": 153, "steps": 250, "lr": 2.1822593205555016e-05, "grad_norm": 5.610466003417969, "loss_scale": 4.0, "retries": 0, "seconds": 149.7, "max_memory_gb": 22.5, "pair": 0.15390098321950063, "sign": 0.0062625015852972865, "keep": 0.01859391143443645, "leak": 1.2332613617181778, "replacement": 0.4746445119380951} +{"step": 154, "steps": 250, "lr": 2.1530812177717236e-05, "grad_norm": 2.612229824066162, "loss_scale": 4.0, "retries": 0, "seconds": 150.6, "max_memory_gb": 22.5, "pair": 0.17572044162079692, "sign": 0.00966239400440827, "keep": 0.2004969958215952, "leak": 0.9008550904691219, "replacement": 0.5060139149427414} +{"step": 155, "steps": 250, "lr": 2.1240107061879815e-05, "grad_norm": 1.8858035802841187, "loss_scale": 4.0, "retries": 0, "seconds": 151.6, "max_memory_gb": 22.5, "pair": 0.1338055997621268, "sign": 0.006559516943525523, "keep": 0.5429972677047772, "leak": 0.662713598459959, "replacement": 0.5742623936384916} +{"step": 156, "steps": 250, "lr": 2.095053025597896e-05, "grad_norm": 5.097497463226318, "loss_scale": 4.0, "retries": 0, "seconds": 152.6, "max_memory_gb": 22.5, "pair": 0.1844144999049604, "sign": 0.0034415860136505216, "keep": 0.047935870322589835, "leak": 0.18603163957595825, "replacement": 0.25661114091053605} +{"step": 157, "steps": 250, "lr": 2.066213395457945e-05, "grad_norm": 11.136625289916992, "loss_scale": 4.0, "retries": 0, "seconds": 153.6, "max_memory_gb": 22.5, "pair": 0.14061357406899333, "sign": 1.0994822670472786, "keep": 0.01557294027224998, "leak": 3.908471238799393, "replacement": 1.6752590276300907} +{"step": 158, "steps": 250, "lr": 2.0374970139466877e-05, "grad_norm": 4.131917953491211, "loss_scale": 4.0, "retries": 0, "seconds": 154.6, "max_memory_gb": 22.5, "pair": 0.3694857135415077, "sign": 0.4152790259104222, "keep": 0.4789029818275594, "leak": 0.05845315888291225, "replacement": 0.8311900533735752} +{"step": 159, "steps": 250, "lr": 2.0089090570278213e-05, "grad_norm": 2.7047975063323975, "loss_scale": 4.0, "retries": 0, "seconds": 155.7, "max_memory_gb": 22.53, "pair": 0.21813986636698246, "sign": 0.010077053622808307, "keep": 0.0033888079778989777, "leak": 0.32426691614091396, "replacement": 0.30593952536582947} +{"step": 160, "steps": 250, "lr": 1.9804546775172457e-05, "grad_norm": 5.314532279968262, "loss_scale": 4.0, "retries": 0, "seconds": 156.8, "max_memory_gb": 22.53, "pair": 0.2955189486965537, "sign": 0.1416474762663711, "keep": 0.4929533794056624, "leak": 0.48182072304189205, "replacement": 0.7332745641469955} +{"step": 161, "steps": 250, "lr": 1.9521390041542946e-05, "grad_norm": 2.861464023590088, "loss_scale": 4.0, "retries": 0, "seconds": 157.8, "max_memory_gb": 22.53, "pair": 0.2945169787853956, "sign": 0.11429217923432589, "keep": 0.11937365164249059, "leak": 0.1169971264898777, "replacement": 0.44059917889535427} +{"step": 162, "steps": 250, "lr": 1.923967140677313e-05, "grad_norm": 2.2233662605285645, "loss_scale": 4.0, "retries": 0, "seconds": 158.7, "max_memory_gb": 22.53, "pair": 0.07854915410280228, "sign": 0.013575908727943897, "keep": 0.09123445054615331, "leak": 0.6854388359934092, "replacement": 0.30231403931975365} +{"step": 163, "steps": 250, "lr": 1.8959441649037335e-05, "grad_norm": 6.547972202301025, "loss_scale": 4.0, "retries": 0, "seconds": 159.6, "max_memory_gb": 22.53, "pair": 0.2987097278237343, "sign": 0.31110799172893167, "keep": 0.4264557338319719, "leak": 0.1541528836824, "replacement": 0.7060298100113869} +{"step": 164, "steps": 250, "lr": 1.868075127814832e-05, "grad_norm": 3.204796552658081, "loss_scale": 4.0, "retries": 0, "seconds": 160.6, "max_memory_gb": 22.53, "pair": 0.2654372453689575, "sign": 0.23915082134772092, "keep": 0.03397167532239109, "leak": 0.49260803312063217, "replacement": 0.5251504927873611} +{"step": 165, "steps": 250, "lr": 1.8403650526453143e-05, "grad_norm": 1.2801071405410767, "loss_scale": 4.0, "retries": 0, "seconds": 161.6, "max_memory_gb": 22.53, "pair": 0.09507026034407318, "sign": 0.012797828792827204, "keep": 0.0018601786523504416, "leak": 0.21037291176617146, "replacement": 0.15499249659478664} +{"step": 166, "steps": 250, "lr": 1.812818933977912e-05, "grad_norm": 12.400229454040527, "loss_scale": 4.0, "retries": 0, "seconds": 162.6, "max_memory_gb": 22.53, "pair": 0.6433687172830105, "sign": 0.8587493891245686, "keep": 1.4728613225743175, "leak": 1.274497530888766, "replacement": 2.127798520028591} +{"step": 167, "steps": 250, "lr": 1.785441736843128e-05, "grad_norm": 3.491399049758911, "loss_scale": 4.0, "retries": 0, "seconds": 163.6, "max_memory_gb": 22.53, "pair": 0.23234840855002403, "sign": 0.0429604914970696, "keep": 0.040161655051633716, "leak": 0.972741624340415, "replacement": 0.5170948952436447} +{"step": 168, "steps": 250, "lr": 1.7582383958243314e-05, "grad_norm": 1.5450692176818848, "loss_scale": 4.0, "retries": 0, "seconds": 164.5, "max_memory_gb": 22.53, "pair": 0.15739748813211918, "sign": 0.009642836696002632, "keep": 0.09765097904164577, "leak": 0.34943586215376854, "replacement": 0.2984033636748791} +{"step": 169, "steps": 250, "lr": 1.7312138141683164e-05, "grad_norm": 3.465331554412842, "loss_scale": 4.0, "retries": 0, "seconds": 165.5, "max_memory_gb": 22.53, "pair": 0.3694610074162483, "sign": 0.1347319409251213, "keep": 0.16116139118093997, "leak": 0.4250908214598894, "replacement": 0.6236803755164146} +{"step": 170, "steps": 250, "lr": 1.704372862901521e-05, "grad_norm": 7.257110595703125, "loss_scale": 4.0, "retries": 0, "seconds": 166.5, "max_memory_gb": 22.53, "pair": 0.16016784124076366, "sign": 0.0041293326212326065, "keep": 0.027498099698277656, "leak": 2.0098561327904463, "replacement": 0.6784455850720406} +{"step": 171, "steps": 250, "lr": 1.6777203799520573e-05, "grad_norm": 2.1891047954559326, "loss_scale": 4.0, "retries": 0, "seconds": 167.5, "max_memory_gb": 22.53, "pair": 0.12871663947589695, "sign": 0.07267649185450864, "keep": 0.06261319859913783, "leak": 0.1992087756516412, "replacement": 0.2461636853404343} +{"step": 172, "steps": 250, "lr": 1.6512611692777008e-05, "grad_norm": 1.4912464618682861, "loss_scale": 4.0, "retries": 0, "seconds": 168.5, "max_memory_gb": 22.53, "pair": 0.09384674113243818, "sign": 0.0023398911580443382, "keep": 0.018278107861988246, "leak": 0.4829351268708706, "replacement": 0.22488952055573463} +{"step": 173, "steps": 250, "lr": 1.6250000000000005e-05, "grad_norm": 5.050271987915039, "loss_scale": 4.0, "retries": 0, "seconds": 169.4, "max_memory_gb": 22.53, "pair": 0.0980090576922521, "sign": 0.07715677079977468, "keep": 0.10629267545300536, "leak": 1.2843753956258297, "replacement": 0.5108276270329952} +{"step": 174, "steps": 250, "lr": 1.598941605544678e-05, "grad_norm": 2.402660369873047, "loss_scale": 4.0, "retries": 0, "seconds": 170.3, "max_memory_gb": 22.53, "pair": 0.10108186909928918, "sign": 0.009763973484950839, "keep": 0.0194195129425907, "leak": 0.5381928328424692, "replacement": 0.2502218186855316} +{"step": 175, "steps": 250, "lr": 1.5730906827884506e-05, "grad_norm": 4.44185733795166, "loss_scale": 4.0, "retries": 0, "seconds": 171.2, "max_memory_gb": 22.53, "pair": 0.2827738458290696, "sign": 0.2720015768427402, "keep": 0.5402994584292173, "leak": 0.7225874494761229, "replacement": 0.8695712238550186} +{"step": 176, "steps": 250, "lr": 1.5474518912124476e-05, "grad_norm": 1.6119767427444458, "loss_scale": 4.0, "retries": 0, "seconds": 172.2, "max_memory_gb": 22.53, "pair": 0.09923668019473553, "sign": 0.026477959821932018, "keep": 0.011562577186850831, "leak": 0.41454123705625534, "replacement": 0.2218922507017851} +{"step": 177, "steps": 250, "lr": 1.5220298520623646e-05, "grad_norm": 7.070497989654541, "loss_scale": 4.0, "retries": 0, "seconds": 173.1, "max_memory_gb": 22.53, "pair": 0.1652887990931049, "sign": 0.3553584600886097, "keep": 0.005849178422295154, "leak": 0.9566173702478409, "replacement": 0.5850469507277012} +{"step": 178, "steps": 250, "lr": 1.4968291475155121e-05, "grad_norm": 10.047139167785645, "loss_scale": 4.0, "retries": 0, "seconds": 174.1, "max_memory_gb": 22.53, "pair": 0.057527293218299747, "sign": 0.03331750648794696, "keep": 0.2999057722336147, "leak": 1.255931068211794, "replacement": 0.5381217002868652} +{"step": 179, "steps": 250, "lr": 1.4718543198548998e-05, "grad_norm": 3.4260265827178955, "loss_scale": 4.0, "retries": 0, "seconds": 175.0, "max_memory_gb": 22.53, "pair": 0.5708216708153486, "sign": 0.5341758714057505, "keep": 0.6404592291855806, "leak": 0.0706532862095628, "replacement": 1.1758025055751204} +{"step": 180, "steps": 250, "lr": 1.4471098706505227e-05, "grad_norm": 1.5921117067337036, "loss_scale": 4.0, "retries": 0, "seconds": 175.9, "max_memory_gb": 22.53, "pair": 0.23400237783789635, "sign": 0.0337758130626753, "keep": 0.05467444157693535, "leak": 0.5570891592651606, "replacement": 0.4174997881054878} +{"step": 181, "steps": 250, "lr": 1.4226002599479743e-05, "grad_norm": 13.932480812072754, "loss_scale": 4.0, "retries": 0, "seconds": 176.9, "max_memory_gb": 22.53, "pair": 0.10166081157512963, "sign": 0.005040719595854171, "keep": 0.024204573332099244, "leak": 3.057636469602585, "replacement": 0.8806925714015961} +{"step": 182, "steps": 250, "lr": 1.3983299054645471e-05, "grad_norm": 13.563714027404785, "loss_scale": 4.0, "retries": 0, "seconds": 177.9, "max_memory_gb": 22.53, "pair": 0.3287719823420048, "sign": 0.3316561996471137, "keep": 0.007886752436206734, "leak": 2.7744681276381016, "replacement": 1.1921605076640844} +{"step": 183, "steps": 250, "lr": 1.3743031817929724e-05, "grad_norm": 3.0922045707702637, "loss_scale": 4.0, "retries": 0, "seconds": 178.9, "max_memory_gb": 22.53, "pair": 0.14508807845413685, "sign": 0.048534843255765736, "keep": 0.002741194540703873, "leak": 0.5919121317565441, "replacement": 0.31870412826538086} +{"step": 184, "steps": 250, "lr": 1.3505244196129164e-05, "grad_norm": 5.394189357757568, "loss_scale": 4.0, "retries": 0, "seconds": 179.8, "max_memory_gb": 22.53, "pair": 0.2410735166631639, "sign": 0.0041430178098380566, "keep": 0.00510852674005946, "leak": 1.0773728513158858, "replacement": 0.5150425042957067} +{"step": 185, "steps": 250, "lr": 1.3269979049104014e-05, "grad_norm": 2.4775125980377197, "loss_scale": 4.0, "retries": 0, "seconds": 180.7, "max_memory_gb": 22.53, "pair": 0.16469418443739414, "sign": 0.021653517556842417, "keep": 0.02851138752885163, "leak": 0.4499532208001398, "replacement": 0.3022649474442005} +{"step": 186, "steps": 250, "lr": 1.3037278782052863e-05, "grad_norm": 3.9078528881073, "loss_scale": 4.0, "retries": 0, "seconds": 181.6, "max_memory_gb": 22.53, "pair": 0.16925794014241546, "sign": 0.01814659642695915, "keep": 0.7361593002460722, "leak": 0.5660167932510376, "replacement": 0.6879151035100222} +{"step": 187, "steps": 250, "lr": 1.2807185337869343e-05, "grad_norm": 2.713045835494995, "loss_scale": 4.0, "retries": 0, "seconds": 182.5, "max_memory_gb": 22.53, "pair": 0.2579508414492011, "sign": 0.009531323274131864, "keep": 0.2799482010304928, "leak": 0.5358353836345486, "replacement": 0.5366494506597519} +{"step": 188, "steps": 250, "lr": 1.2579740189582115e-05, "grad_norm": 3.479018211364746, "loss_scale": 4.0, "retries": 0, "seconds": 183.4, "max_memory_gb": 22.53, "pair": 0.5336730624549091, "sign": 0.795332849258557, "keep": 0.7528670790702563, "leak": 0.04051751550287008, "replacement": 1.3179023643024266} +{"step": 189, "steps": 250, "lr": 1.2354984332879668e-05, "grad_norm": 8.980701446533203, "loss_scale": 4.0, "retries": 0, "seconds": 184.3, "max_memory_gb": 22.53, "pair": 0.08518694364465773, "sign": 0.5392164383083582, "keep": 1.1422397346468642, "leak": 2.174583189344048, "replacement": 1.4695608280599117} +{"step": 190, "steps": 250, "lr": 1.2132958278721064e-05, "grad_norm": 4.959600448608398, "loss_scale": 4.0, "retries": 0, "seconds": 185.2, "max_memory_gb": 22.53, "pair": 0.05291673936881125, "sign": 0.004365973640233278, "keep": 0.01044875112711452, "leak": 0.9010126143693924, "replacement": 0.2855772525072098} +{"step": 191, "steps": 250, "lr": 1.1913702046034016e-05, "grad_norm": 22.012659072875977, "loss_scale": 4.0, "retries": 0, "seconds": 186.1, "max_memory_gb": 22.53, "pair": 0.4224998168647289, "sign": 0.6596469193464145, "keep": 1.160039958442212, "leak": 2.0575936771929264, "replacement": 1.8467416502535343} +{"step": 192, "steps": 250, "lr": 1.1697255154501756e-05, "grad_norm": 3.0620689392089844, "loss_scale": 4.0, "retries": 0, "seconds": 187.0, "max_memory_gb": 22.53, "pair": 0.47091774083673954, "sign": 0.47001714527141303, "keep": 1.1253896949347109, "leak": 0.3988053095526993, "replacement": 1.3683225847780704} +{"step": 193, "steps": 250, "lr": 1.1483656617439842e-05, "grad_norm": 5.335739612579346, "loss_scale": 4.0, "retries": 0, "seconds": 187.9, "max_memory_gb": 22.53, "pair": 0.1283910417696461, "sign": 0.017924216022947803, "keep": 0.010598363411190803, "leak": 1.261298045516014, "replacement": 0.45797685091383755} +{"step": 194, "steps": 250, "lr": 1.1272944934764177e-05, "grad_norm": 3.0830986499786377, "loss_scale": 4.0, "retries": 0, "seconds": 188.8, "max_memory_gb": 22.53, "pair": 0.2378688007593155, "sign": 0.18382695247419178, "keep": 0.01995528024053783, "leak": 0.32829800620675087, "replacement": 0.4218344185501337} +{"step": 195, "steps": 250, "lr": 1.1065158086051684e-05, "grad_norm": 4.375974655151367, "loss_scale": 4.0, "retries": 0, "seconds": 189.7, "max_memory_gb": 22.53, "pair": 0.2284661065787077, "sign": 0.0512614046165254, "keep": 0.006690160982543603, "leak": 0.5088755842298269, "replacement": 0.3846607878804207} +{"step": 196, "steps": 250, "lr": 1.086033352369467e-05, "grad_norm": 4.112195014953613, "loss_scale": 4.0, "retries": 0, "seconds": 190.7, "max_memory_gb": 22.53, "pair": 0.14777499437332153, "sign": 0.008476780916680582, "keep": 0.009779758871673039, "leak": 1.2194617085624486, "replacement": 0.4617686732672155} +{"step": 197, "steps": 250, "lr": 1.0658508166150224e-05, "grad_norm": 6.803039073944092, "loss_scale": 4.0, "retries": 0, "seconds": 191.7, "max_memory_gb": 22.53, "pair": 0.11864774720743299, "sign": 0.029510174150345847, "keep": 0.002583708002930507, "leak": 1.5396815191779751, "replacement": 0.5196150727570057} +{"step": 198, "steps": 250, "lr": 1.0459718391285954e-05, "grad_norm": 2.502100706100464, "loss_scale": 4.0, "retries": 0, "seconds": 192.6, "max_memory_gb": 22.53, "pair": 0.17511006211861968, "sign": 0.06695483217481524, "keep": 0.08864327857736498, "leak": 0.36036388017237186, "replacement": 0.3430000729858875} +{"step": 199, "steps": 250, "lr": 1.0264000029822999e-05, "grad_norm": 4.747139930725098, "loss_scale": 4.0, "retries": 0, "seconds": 193.6, "max_memory_gb": 22.53, "pair": 0.1405506341252476, "sign": 0.026336993018048815, "keep": 0.4814461153000593, "leak": 0.8057026099413633, "replacement": 0.5958678424358368} +{"step": 200, "steps": 250, "lr": 1.0071388358877775e-05, "grad_norm": 5.234976768493652, "loss_scale": 4.0, "retries": 0, "seconds": 194.6, "max_memory_gb": 22.53, "pair": 0.3174542039632797, "sign": 0.34291650090017356, "keep": 1.1773175605922006, "leak": 0.7270096282009035, "replacement": 1.259323611855507} +{"step": 201, "steps": 250, "lr": 9.881918095603501e-06, "grad_norm": 5.944198131561279, "loss_scale": 4.0, "retries": 0, "seconds": 195.6, "max_memory_gb": 22.53, "pair": 0.47357360646128654, "sign": 0.26323916646651924, "keep": 0.3047017971985042, "leak": 0.329594137147069, "replacement": 0.8399425968527794} +{"step": 202, "steps": 250, "lr": 9.695623390932612e-06, "grad_norm": 15.639384269714355, "loss_scale": 4.0, "retries": 0, "seconds": 196.5, "max_memory_gb": 22.53, "pair": 0.21157784102251753, "sign": 0.07842010006424971, "keep": 0.014021640612551067, "leak": 3.3280867766588926, "replacement": 1.0898203924298286} +{"step": 203, "steps": 250, "lr": 9.51253782342123e-06, "grad_norm": 2.2346835136413574, "loss_scale": 4.0, "retries": 0, "seconds": 197.5, "max_memory_gb": 22.53, "pair": 0.11013101786375046, "sign": 0.045176027837442234, "keep": 0.0324716167524457, "leak": 0.2570767276920378, "replacement": 0.21322402358055115} +{"step": 204, "steps": 250, "lr": 9.332694393196846e-06, "grad_norm": 2.8964548110961914, "loss_scale": 4.0, "retries": 0, "seconds": 198.5, "max_memory_gb": 22.53, "pair": 0.33793288096785545, "sign": 0.023821283830329776, "keep": 0.05119752371683717, "leak": 0.6471131378784776, "replacement": 0.5372205674648285} +{"step": 205, "steps": 250, "lr": 9.156125516010256e-06, "grad_norm": 2.371436595916748, "loss_scale": 4.0, "retries": 0, "seconds": 199.4, "max_memory_gb": 22.53, "pair": 0.11886560544371605, "sign": 0.0055118270247476175, "keep": 0.00973415412590839, "leak": 0.464252270758152, "replacement": 0.24255166575312614} +{"step": 206, "steps": 250, "lr": 8.982863017392733e-06, "grad_norm": 2.9664711952209473, "loss_scale": 4.0, "retries": 0, "seconds": 200.4, "max_memory_gb": 22.53, "pair": 0.16963036730885506, "sign": 0.011288468027487397, "keep": 0.016181968615683218, "leak": 0.256024107336998, "replacement": 0.24737160839140415} +{"step": 207, "steps": 250, "lr": 8.81293812691974e-06, "grad_norm": 1.4895353317260742, "loss_scale": 4.0, "retries": 0, "seconds": 201.4, "max_memory_gb": 22.53, "pair": 0.2744982047006488, "sign": 0.152241154195508, "keep": 0.08721111848353758, "leak": 0.30091225868090987, "replacement": 0.4694524244405329} +{"step": 208, "steps": 250, "lr": 8.646381472581927e-06, "grad_norm": 1.7171155214309692, "loss_scale": 4.0, "retries": 0, "seconds": 202.4, "max_memory_gb": 22.53, "pair": 0.11653061769902706, "sign": 0.004045929992571473, "keep": 0.010089470188233918, "leak": 0.4136473350226879, "replacement": 0.2270101485773921} +{"step": 209, "steps": 250, "lr": 8.483223075264622e-06, "grad_norm": 5.870275020599365, "loss_scale": 4.0, "retries": 0, "seconds": 203.4, "max_memory_gb": 22.53, "pair": 0.6577718649059534, "sign": 0.42407370393630117, "keep": 0.4362194331479259, "leak": 0.7813612557947636, "replacement": 1.283258780837059} +{"step": 210, "steps": 250, "lr": 8.323492343336747e-06, "grad_norm": 4.6846923828125, "loss_scale": 4.0, "retries": 0, "seconds": 204.4, "max_memory_gb": 22.53, "pair": 0.4984286054968834, "sign": 0.5308636704576202, "keep": 1.517529815279886, "leak": 0.8697943557053804, "replacement": 1.7400739304721355} +{"step": 211, "steps": 250, "lr": 8.167218067350122e-06, "grad_norm": 3.0878870487213135, "loss_scale": 4.0, "retries": 0, "seconds": 205.4, "max_memory_gb": 22.53, "pair": 0.532728049904108, "sign": 0.27753589232452214, "keep": 0.13216611160896719, "leak": 0.18749817460775375, "replacement": 0.7844535820186138} +{"step": 212, "steps": 250, "lr": 8.01442841485013e-06, "grad_norm": 9.439738273620605, "loss_scale": 4.0, "retries": 0, "seconds": 206.4, "max_memory_gb": 22.53, "pair": 0.6013111248612404, "sign": 0.22511387441772968, "keep": 0.09802628454053774, "leak": 1.0062225759029388, "replacement": 1.0144367963075638} +{"step": 213, "steps": 250, "lr": 7.865150925298678e-06, "grad_norm": 5.631274700164795, "loss_scale": 4.0, "retries": 0, "seconds": 207.3, "max_memory_gb": 22.53, "pair": 0.4504523230716586, "sign": 0.3355480899626855, "keep": 0.48489142704056576, "leak": 0.4359807695250879, "replacement": 0.9696672828868032} +{"step": 214, "steps": 250, "lr": 7.719412505110375e-06, "grad_norm": 2.6651113033294678, "loss_scale": 4.0, "retries": 0, "seconds": 208.3, "max_memory_gb": 22.53, "pair": 0.4170527821406722, "sign": 0.37625689006381435, "keep": 1.0542948161355525, "leak": 0.022148114994934076, "replacement": 1.1378656895831227} +{"step": 215, "steps": 250, "lr": 7.57723942280278e-06, "grad_norm": 5.5338053703308105, "loss_scale": 4.0, "retries": 0, "seconds": 209.4, "max_memory_gb": 22.53, "pair": 0.6282319676247425, "sign": 0.8576446528313681, "keep": 1.67916762862842, "leak": 0.12103592232597293, "replacement": 1.9268970935372636} +{"step": 216, "steps": 250, "lr": 7.438657304261687e-06, "grad_norm": 2.930375576019287, "loss_scale": 4.0, "retries": 0, "seconds": 210.4, "max_memory_gb": 22.53, "pair": 0.08686455176211894, "sign": 0.02034628416004125, "keep": 0.004916423196846154, "leak": 1.0355170173570514, "replacement": 0.35837516840547323} +{"step": 217, "steps": 250, "lr": 7.30369112812219e-06, "grad_norm": 2.684013605117798, "loss_scale": 4.0, "retries": 0, "seconds": 211.4, "max_memory_gb": 22.53, "pair": 0.1957970568910241, "sign": 0.005120583111420274, "keep": 0.0014812927875027526, "leak": 0.37114090751856565, "replacement": 0.29188322089612484} +{"step": 218, "steps": 250, "lr": 7.172365221266403e-06, "grad_norm": 8.213654518127441, "loss_scale": 4.0, "retries": 0, "seconds": 212.3, "max_memory_gb": 22.53, "pair": 0.6323689296841621, "sign": 1.210929997265339, "keep": 0.20235112810041755, "leak": 0.17543951468542218, "replacement": 1.3828694075345993} +{"step": 219, "steps": 250, "lr": 7.0447032544387154e-06, "grad_norm": 3.6691880226135254, "loss_scale": 4.0, "retries": 0, "seconds": 213.3, "max_memory_gb": 22.53, "pair": 0.20590457692742348, "sign": 0.06718701159115881, "keep": 0.00710017466917634, "leak": 0.06082199695811141, "replacement": 0.25825368240475655} +{"step": 220, "steps": 250, "lr": 6.920728237979274e-06, "grad_norm": 3.2239818572998047, "loss_scale": 4.0, "retries": 0, "seconds": 214.4, "max_memory_gb": 22.53, "pair": 0.23557222355157137, "sign": 0.05084224135498516, "keep": 0.09493895421792331, "leak": 0.1456389189697802, "replacement": 0.3448725361377001} +{"step": 221, "steps": 250, "lr": 6.800462517676456e-06, "grad_norm": 2.628176212310791, "loss_scale": 4.0, "retries": 0, "seconds": 215.3, "max_memory_gb": 22.53, "pair": 0.3453857973217964, "sign": 0.1760443025850691, "keep": 0.008480206364765763, "leak": 0.888189603574574, "replacement": 0.6596954613924026} +{"step": 222, "steps": 250, "lr": 6.683927770739222e-06, "grad_norm": 2.0822649002075195, "loss_scale": 4.0, "retries": 0, "seconds": 216.3, "max_memory_gb": 22.53, "pair": 0.15588982356712222, "sign": 0.01026017259573564, "keep": 0.04100518486711735, "leak": 0.18159657903015614, "replacement": 0.22692164266481996} +{"step": 223, "steps": 250, "lr": 6.571145001889892e-06, "grad_norm": 13.645058631896973, "loss_scale": 4.0, "retries": 0, "seconds": 217.3, "max_memory_gb": 22.53, "pair": 0.5927910734899342, "sign": 0.13501854733476648, "keep": 0.05913555952429306, "leak": 1.0125232599675655, "replacement": 0.9429989792406559} +{"step": 224, "steps": 250, "lr": 6.46213453957817e-06, "grad_norm": 3.061645030975342, "loss_scale": 4.0, "retries": 0, "seconds": 218.3, "max_memory_gb": 22.53, "pair": 0.2186727076768875, "sign": 0.031063110451214015, "keep": 0.011389250361617087, "leak": 0.5517473760992289, "replacement": 0.37783573009073734} +{"step": 225, "steps": 250, "lr": 6.356916032317063e-06, "grad_norm": 5.6224260330200195, "loss_scale": 4.0, "retries": 0, "seconds": 219.3, "max_memory_gb": 22.53, "pair": 0.15440547047182918, "sign": 0.017613504664041102, "keep": 0.0021507268666027812, "leak": 1.6856550113298, "replacement": 0.5857013538479805} +{"step": 226, "steps": 250, "lr": 6.255508445141349e-06, "grad_norm": 4.151510238647461, "loss_scale": 4.0, "retries": 0, "seconds": 220.3, "max_memory_gb": 22.53, "pair": 0.1131037506274879, "sign": 0.022706733492668718, "keep": 0.6011632254812866, "leak": 0.5771294543519616, "replacement": 0.5693210884928703} +{"step": 227, "steps": 250, "lr": 6.157930056189228e-06, "grad_norm": 2.03643798828125, "loss_scale": 4.0, "retries": 0, "seconds": 221.2, "max_memory_gb": 22.53, "pair": 0.1536616560188122, "sign": 0.012159337878983933, "keep": 0.03088512649628683, "leak": 0.23215575143694878, "replacement": 0.23322283569723368} +{"step": 228, "steps": 250, "lr": 6.064198453407796e-06, "grad_norm": 2.6054861545562744, "loss_scale": 4.0, "retries": 0, "seconds": 222.2, "max_memory_gb": 22.53, "pair": 0.18217993178404868, "sign": 0.06486528587993234, "keep": 0.00468061285209842, "leak": 0.7146375985503255, "replacement": 0.3956122873350978} +{"step": 229, "steps": 250, "lr": 5.974330531382935e-06, "grad_norm": 12.264750480651855, "loss_scale": 4.0, "retries": 0, "seconds": 223.2, "max_memory_gb": 22.53, "pair": 0.17191786132752895, "sign": 0.06815007771365345, "keep": 0.05358619592152358, "leak": 1.816392907872796, "replacement": 0.686884231865406} +{"step": 230, "steps": 250, "lr": 5.888342488294125e-06, "grad_norm": 17.026994705200195, "loss_scale": 4.0, "retries": 0, "seconds": 224.1, "max_memory_gb": 22.53, "pair": 0.8232115730643272, "sign": 0.6914288178086281, "keep": 1.3047747284681463, "leak": 1.3684508185833693, "replacement": 2.163426086306572} +{"step": 231, "steps": 250, "lr": 5.806249822994849e-06, "grad_norm": 2.930006265640259, "loss_scale": 4.0, "retries": 0, "seconds": 225.0, "max_memory_gb": 22.53, "pair": 0.3750723209232092, "sign": 0.6766508294967934, "keep": 0.022170451961301296, "leak": 0.22727843048051, "replacement": 0.7813025247305632} +{"step": 232, "steps": 250, "lr": 5.728067332218994e-06, "grad_norm": 2.4362847805023193, "loss_scale": 4.0, "retries": 0, "seconds": 226.0, "max_memory_gb": 22.53, "pair": 0.13268128782510757, "sign": 0.01791865093400702, "keep": 0.03665004458162002, "leak": 0.9054371193051338, "replacement": 0.3863249197602272} +{"step": 233, "steps": 250, "lr": 5.65380910791383e-06, "grad_norm": 5.739310264587402, "loss_scale": 4.0, "retries": 0, "seconds": 227.0, "max_memory_gb": 22.53, "pair": 0.10389877599664032, "sign": 0.004901801847154275, "keep": 0.003826273954473436, "leak": 1.546828668564558, "replacement": 0.49496997706592083} +{"step": 234, "steps": 250, "lr": 5.583488534700022e-06, "grad_norm": 5.997952938079834, "loss_scale": 4.0, "retries": 0, "seconds": 227.9, "max_memory_gb": 22.53, "pair": 0.09894910705043003, "sign": 0.01490675302920863, "keep": 0.006756080547205556, "leak": 0.8603673111647367, "replacement": 0.32487234845757484} +{"step": 235, "steps": 250, "lr": 5.517118287459118e-06, "grad_norm": 13.413725852966309, "loss_scale": 4.0, "retries": 0, "seconds": 228.9, "max_memory_gb": 22.53, "pair": 0.8339355885982513, "sign": 0.6061311854282394, "keep": 1.1454965393713792, "leak": 1.5655403786076931, "replacement": 2.101134557276964} +{"step": 236, "steps": 250, "lr": 5.454710329048978e-06, "grad_norm": 2.9910056591033936, "loss_scale": 4.0, "retries": 0, "seconds": 229.9, "max_memory_gb": 22.53, "pair": 0.2557313425350003, "sign": 0.07697515792096965, "keep": 0.004764565179357305, "leak": 0.1253105696523562, "replacement": 0.3279288599733263} +{"step": 237, "steps": 250, "lr": 5.3962759081475436e-06, "grad_norm": 4.779725074768066, "loss_scale": 4.0, "retries": 0, "seconds": 230.8, "max_memory_gb": 22.53, "pair": 0.25252140313386917, "sign": 0.31888547842390835, "keep": 0.12512605741414973, "leak": 0.6518917542416602, "replacement": 0.6375001221895218} +{"step": 238, "steps": 250, "lr": 5.341825557225321e-06, "grad_norm": 1.755576491355896, "loss_scale": 4.0, "retries": 0, "seconds": 231.7, "max_memory_gb": 22.53, "pair": 0.1733792731538415, "sign": 0.013275461329612881, "keep": 0.03453286713920534, "leak": 0.23898271471261978, "replacement": 0.25702912732958794} +{"step": 239, "steps": 250, "lr": 5.291369090646961e-06, "grad_norm": 4.199362754821777, "loss_scale": 4.0, "retries": 0, "seconds": 232.7, "max_memory_gb": 22.53, "pair": 0.4280986078083515, "sign": 1.2156945606693625, "keep": 0.11178131436020067, "leak": 0.5505988672375679, "replacement": 1.2294863015413284} +{"step": 240, "steps": 250, "lr": 5.244915602902289e-06, "grad_norm": 3.437533378601074, "loss_scale": 4.0, "retries": 0, "seconds": 233.7, "max_memory_gb": 22.53, "pair": 0.21886535361409187, "sign": 0.35924274544231594, "keep": 0.18635362511849962, "leak": 0.17775477143004537, "replacement": 0.5361022278666496} +{"step": 241, "steps": 250, "lr": 5.20247346696706e-06, "grad_norm": 10.167688369750977, "loss_scale": 4.0, "retries": 0, "seconds": 234.7, "max_memory_gb": 22.53, "pair": 0.6572524337098002, "sign": 0.5343602734064916, "keep": 0.3409922157577512, "leak": 0.28671169182052836, "replacement": 1.16660659853369} +{"step": 242, "steps": 250, "lr": 5.164050332793787e-06, "grad_norm": 1.9687411785125732, "loss_scale": 4.0, "retries": 0, "seconds": 235.6, "max_memory_gb": 22.53, "pair": 0.1473310524597764, "sign": 0.0270000507007353, "keep": 0.0026719040015450446, "leak": 0.782576959580183, "replacement": 0.35781126841902733} +{"step": 243, "steps": 250, "lr": 5.129653125932873e-06, "grad_norm": 1.7874184846878052, "loss_scale": 4.0, "retries": 0, "seconds": 236.6, "max_memory_gb": 22.53, "pair": 0.1796875619329512, "sign": 0.0187979162146803, "keep": 0.02236533467657864, "leak": 0.277864221483469, "replacement": 0.2697352357208729} +{"step": 244, "steps": 250, "lr": 5.099288046284329e-06, "grad_norm": 2.793142795562744, "loss_scale": 4.0, "retries": 0, "seconds": 237.6, "max_memory_gb": 22.53, "pair": 0.29057192616164684, "sign": 0.12266603228636086, "keep": 0.020382012357003987, "leak": 0.6701124124228954, "replacement": 0.5296240523457527} +{"step": 245, "steps": 250, "lr": 5.072960566980277e-06, "grad_norm": 5.876723766326904, "loss_scale": 4.0, "retries": 0, "seconds": 238.6, "max_memory_gb": 22.53, "pair": 0.5606374312192202, "sign": 0.4161803799215704, "keep": 0.8839587466791272, "leak": 0.4097178652882576, "replacement": 1.3131364732980728} +{"step": 246, "steps": 250, "lr": 5.050675433398435e-06, "grad_norm": 1.8574305772781372, "loss_scale": 4.0, "retries": 0, "seconds": 239.6, "max_memory_gb": 22.53, "pair": 0.1329088625498116, "sign": 0.004617735103238374, "keep": 0.003689528617542237, "leak": 0.08636851236224174, "replacement": 0.15865462087094784} +{"step": 247, "steps": 250, "lr": 5.032436662306811e-06, "grad_norm": 11.251906394958496, "loss_scale": 4.0, "retries": 0, "seconds": 240.6, "max_memory_gb": 22.53, "pair": 0.11320886493194848, "sign": 0.012876352593593765, "keep": 0.015145715675117799, "leak": 0.7888036984950304, "replacement": 0.3244208316318691} +{"step": 248, "steps": 250, "lr": 5.018247541139684e-06, "grad_norm": 2.637829303741455, "loss_scale": 4.0, "retries": 0, "seconds": 241.5, "max_memory_gb": 22.53, "pair": 0.1809534039348364, "sign": 0.033328363148029894, "keep": 0.034283970919204876, "leak": 0.3133201866876334, "replacement": 0.293089609593153} +{"step": 249, "steps": 250, "lr": 5.00811062740508e-06, "grad_norm": 4.554159164428711, "loss_scale": 4.0, "retries": 0, "seconds": 242.5, "max_memory_gb": 22.53, "pair": 0.26837552431970835, "sign": 0.0909659843891859, "keep": 0.044395287171937525, "leak": 0.8301713368855417, "replacement": 0.5435989797115326} +{"step": 250, "steps": 250, "lr": 5.002027748223778e-06, "grad_norm": 2.3452818393707275, "loss_scale": 4.0, "retries": 0, "seconds": 243.5, "max_memory_gb": 22.53, "pair": 0.2106462437659502, "sign": 0.01712611224502325, "keep": 0.04521413304973976, "leak": 0.1645621252246201, "replacement": 0.28295689448714256} diff --git a/runs/single_internvl/replacement/L16_s13/adapter/README.md b/runs/single_internvl/replacement/L16_s13/adapter/README.md new file mode 100644 index 0000000000000000000000000000000000000000..4f9f856cf0e1531b47e25d266833b2977a19d3ac --- /dev/null +++ b/runs/single_internvl/replacement/L16_s13/adapter/README.md @@ -0,0 +1,207 @@ +--- +base_model: /data/OPSD/assets/models/InternVL3_5-8B-HF +library_name: peft +pipeline_tag: text-generation +tags: +- base_model:adapter:/data/OPSD/assets/models/InternVL3_5-8B-HF +- lora +- transformers +--- + +# Model Card for Model ID + + + + + +## Model Details + +### Model Description + + + + + +- **Developed by:** [More Information Needed] +- **Funded by [optional]:** [More Information Needed] +- **Shared by [optional]:** [More Information Needed] +- **Model type:** [More Information Needed] +- **Language(s) (NLP):** [More Information Needed] +- **License:** [More Information Needed] +- **Finetuned from model [optional]:** [More Information Needed] + +### Model Sources [optional] + + + +- **Repository:** [More Information Needed] +- **Paper [optional]:** [More Information Needed] +- **Demo [optional]:** [More Information Needed] + +## Uses + + + +### Direct Use + + + +[More Information Needed] + +### Downstream Use [optional] + + + +[More Information Needed] + +### Out-of-Scope Use + + + +[More Information Needed] + +## Bias, Risks, and Limitations + + + +[More Information Needed] + +### Recommendations + + + +Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. More information needed for further recommendations. + +## How to Get Started with the Model + +Use the code below to get started with the model. + +[More Information Needed] + +## Training Details + +### Training Data + + + +[More Information Needed] + +### Training Procedure + + + +#### Preprocessing [optional] + +[More Information Needed] + + +#### Training Hyperparameters + +- **Training regime:** [More Information Needed] + +#### Speeds, Sizes, Times [optional] + + + +[More Information Needed] + +## Evaluation + + + +### Testing Data, Factors & Metrics + +#### Testing Data + + + +[More Information Needed] + +#### Factors + + + +[More Information Needed] + +#### Metrics + + + +[More Information Needed] + +### Results + +[More Information Needed] + +#### Summary + + + +## Model Examination [optional] + + + +[More Information Needed] + +## Environmental Impact + + + +Carbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700). + +- **Hardware Type:** [More Information Needed] +- **Hours used:** [More Information Needed] +- **Cloud Provider:** [More Information Needed] +- **Compute Region:** [More Information Needed] +- **Carbon Emitted:** [More Information Needed] + +## Technical Specifications [optional] + +### Model Architecture and Objective + +[More Information Needed] + +### Compute Infrastructure + +[More Information Needed] + +#### Hardware + +[More Information Needed] + +#### Software + +[More Information Needed] + +## Citation [optional] + + + +**BibTeX:** + +[More Information Needed] + +**APA:** + +[More Information Needed] + +## Glossary [optional] + + + +[More Information Needed] + +## More Information [optional] + +[More Information Needed] + +## Model Card Authors [optional] + +[More Information Needed] + +## Model Card Contact + +[More Information Needed] +### Framework versions + +- PEFT 0.20.0 \ No newline at end of file diff --git a/runs/single_internvl/replacement/L16_s13/adapter/adapter_config.json b/runs/single_internvl/replacement/L16_s13/adapter/adapter_config.json new file mode 100644 index 0000000000000000000000000000000000000000..a63a1ce9d331752cbaf26ab241be476034e06f42 --- /dev/null +++ b/runs/single_internvl/replacement/L16_s13/adapter/adapter_config.json @@ -0,0 +1,50 @@ +{ + "alora_invocation_tokens": null, + "alpha_pattern": {}, + "arrow_config": null, + "auto_mapping": null, + "base_model_name_or_path": "/data/OPSD/assets/models/InternVL3_5-8B-HF", + "bias": "none", + "corda_config": null, + "ensure_weight_tying": false, + "eva_config": null, + "exclude_modules": null, + "fan_in_fan_out": false, + "inference_mode": true, + "init_lora_weights": true, + "layer_replication": null, + "layers_pattern": null, + "layers_to_transform": null, + "loftq_config": {}, + "lora_alpha": 32, + "lora_bias": false, + "lora_dropout": 0.05, + "lora_ga_config": null, + "megatron_config": null, + "megatron_core": "megatron.core", + "modules_to_save": null, + "monteclora_config": null, + "peft_type": "LORA", + "peft_version": "0.20.0", + "qalora_group_size": 16, + "r": 16, + "rank_pattern": {}, + "revision": null, + "target_modules": [ + "model.language_model.layers.16.self_attn.v_proj", + "model.language_model.layers.16.self_attn.q_proj", + "model.language_model.layers.16.mlp.down_proj", + "model.language_model.layers.16.self_attn.k_proj", + "model.language_model.layers.16.mlp.up_proj", + "model.language_model.layers.16.mlp.gate_proj", + "model.language_model.layers.16.self_attn.o_proj" + ], + "target_parameters": null, + "task_type": "CAUSAL_LM", + "trainable_token_indices": null, + "use_bdlora": null, + "use_dora": false, + "use_qalora": false, + "use_rslora": false, + "velora_config": null +} \ No newline at end of file diff --git a/runs/single_internvl/replacement/L16_s13/config.json b/runs/single_internvl/replacement/L16_s13/config.json new file mode 100644 index 0000000000000000000000000000000000000000..0f646936e2579168c0c70b7c742f3616f164be05 --- /dev/null +++ b/runs/single_internvl/replacement/L16_s13/config.json @@ -0,0 +1,24 @@ +{ + "behavior": "replacement", + "seed": 13, + "model": "/data/OPSD/assets/models/InternVL3_5-8B-HF", + "backbone": "internvl", + "train": "/data/OPSD/assets/v3/replacement.jsonl", + "items": 1000, + "steps": 250, + "epochs": 1, + "steps_per_epoch": 250, + "groups_per_step": 4, + "lr": 5e-05, + "scheduler": "cosine", + "warmup_steps": 16, + "final_fraction": 0.1, + "weight_decay": 0.0, + "clip": 1.0, + "initial_loss_scale": 4.0, + "adapter": { + "layer": 16, + "rank": 16, + "trainable_parameters": 1212416 + } +} \ No newline at end of file diff --git a/runs/single_internvl/replacement/L16_s13/steps.jsonl b/runs/single_internvl/replacement/L16_s13/steps.jsonl new file mode 100644 index 0000000000000000000000000000000000000000..731955d160c8fb4ea77362a1af17019cb4d3a911 --- /dev/null +++ b/runs/single_internvl/replacement/L16_s13/steps.jsonl @@ -0,0 +1,250 @@ +{"step": 1, "steps": 250, "lr": 3.125e-06, "grad_norm": 2.3246161937713623, "loss_scale": 4.0, "retries": 0, "seconds": 1.1, "max_memory_gb": 21.19, "pair": 1.0986123085021973, "sign": 0.7328193015418947, "keep": 1.137254401575774, "leak": 0.0, "replacement": 2.0336491465568542} +{"step": 2, "steps": 250, "lr": 6.25e-06, "grad_norm": 2.5453596115112305, "loss_scale": 4.0, "retries": 0, "seconds": 2.0, "max_memory_gb": 21.22, "pair": 1.0998353362083435, "sign": 0.18078364036045969, "keep": 1.3678097140218597, "leak": 3.129068647922395e-05, "replacement": 1.8741397857666016} +{"step": 3, "steps": 250, "lr": 9.375000000000001e-06, "grad_norm": 5.069816589355469, "loss_scale": 4.0, "retries": 0, "seconds": 2.9, "max_memory_gb": 21.23, "pair": 1.0911274552345276, "sign": 0.1159900943748653, "keep": 2.0607174932956696, "leak": 3.558868462505416e-05, "replacement": 2.1794901490211487} +{"step": 4, "steps": 250, "lr": 1.25e-05, "grad_norm": 3.5192432403564453, "loss_scale": 4.0, "retries": 0, "seconds": 3.8, "max_memory_gb": 21.23, "pair": 1.0909902155399323, "sign": 0.03802182455547154, "keep": 0.8680323278385913, "leak": 1.9403593853439816e-05, "replacement": 1.544022113084793} +{"step": 5, "steps": 250, "lr": 1.5625e-05, "grad_norm": 1.8387919664382935, "loss_scale": 4.0, "retries": 0, "seconds": 4.8, "max_memory_gb": 21.23, "pair": 1.0986119210720062, "sign": 0.11665617930702865, "keep": 0.023284009264898486, "leak": 6.075585770304315e-05, "replacement": 1.1685972213745117} +{"step": 6, "steps": 250, "lr": 1.8750000000000002e-05, "grad_norm": 2.467484712600708, "loss_scale": 4.0, "retries": 0, "seconds": 5.6, "max_memory_gb": 21.23, "pair": 1.091438889503479, "sign": 0.19846855197101831, "keep": 0.6248354346025735, "leak": 2.316984172523462e-05, "replacement": 1.5030966699123383} +{"step": 7, "steps": 250, "lr": 2.1875e-05, "grad_norm": 4.5668416023254395, "loss_scale": 4.0, "retries": 0, "seconds": 6.6, "max_memory_gb": 21.23, "pair": 1.0803800821304321, "sign": 0.08911793050356209, "keep": 1.4616508287654142, "leak": 0.0005681090665916599, "replacement": 1.8559065759181976} +{"step": 8, "steps": 250, "lr": 2.5e-05, "grad_norm": 4.507833480834961, "loss_scale": 4.0, "retries": 0, "seconds": 7.5, "max_memory_gb": 21.26, "pair": 1.0737448632717133, "sign": 0.20971342304255813, "keep": 1.3662461452186108, "leak": 0.0007826802466297522, "replacement": 1.861920326948166} +{"step": 9, "steps": 250, "lr": 2.8125000000000003e-05, "grad_norm": 5.287485599517822, "loss_scale": 4.0, "retries": 0, "seconds": 8.4, "max_memory_gb": 21.26, "pair": 1.0802913308143616, "sign": 0.5725770648568869, "keep": 1.2638913169503212, "leak": 0.0004302750458009541, "replacement": 1.998633086681366} +{"step": 10, "steps": 250, "lr": 3.125e-05, "grad_norm": 2.1600089073181152, "loss_scale": 4.0, "retries": 0, "seconds": 9.3, "max_memory_gb": 21.26, "pair": 1.0596312284469604, "sign": 0.07953445659950376, "keep": 0.07890914008021355, "leak": 0.0021208022080827504, "replacement": 1.1393831968307495} +{"step": 11, "steps": 250, "lr": 3.4375e-05, "grad_norm": 4.144200801849365, "loss_scale": 4.0, "retries": 0, "seconds": 10.3, "max_memory_gb": 21.26, "pair": 1.0428882539272308, "sign": 0.20432483544573188, "keep": 1.0911683894810267, "leak": 0.002200506074586883, "replacement": 1.6911849677562714} +{"step": 12, "steps": 250, "lr": 3.7500000000000003e-05, "grad_norm": 3.521724224090576, "loss_scale": 4.0, "retries": 0, "seconds": 11.2, "max_memory_gb": 21.26, "pair": 1.0212091207504272, "sign": 0.07280348287895322, "keep": 0.7423193155118497, "leak": 0.003978717140853405, "replacement": 1.4297651648521423} +{"step": 13, "steps": 250, "lr": 4.0625000000000005e-05, "grad_norm": 3.3862388134002686, "loss_scale": 4.0, "retries": 0, "seconds": 12.0, "max_memory_gb": 21.26, "pair": 0.9765212535858154, "sign": 0.05405826400965452, "keep": 0.6743961041429429, "leak": 0.012984090120880865, "replacement": 1.3439944684505463} +{"step": 14, "steps": 250, "lr": 4.375e-05, "grad_norm": 4.96243143081665, "loss_scale": 4.0, "retries": 0, "seconds": 12.9, "max_memory_gb": 21.26, "pair": 0.98037089407444, "sign": 0.2594081796705723, "keep": 1.1027611941099167, "leak": 0.012156582553870976, "replacement": 1.6644947528839111} +{"step": 15, "steps": 250, "lr": 4.6875e-05, "grad_norm": 3.683072328567505, "loss_scale": 4.0, "retries": 0, "seconds": 13.8, "max_memory_gb": 21.26, "pair": 0.954150378704071, "sign": 0.21092158509418368, "keep": 0.5405612113117968, "leak": 0.03943381831049919, "replacement": 1.3397502303123474} +{"step": 16, "steps": 250, "lr": 5e-05, "grad_norm": 2.973539113998413, "loss_scale": 4.0, "retries": 0, "seconds": 14.7, "max_memory_gb": 21.26, "pair": 0.9310121983289719, "sign": 0.2970041148364544, "keep": 0.19720414743642323, "leak": 0.02512569259852171, "replacement": 1.1843977868556976} +{"step": 17, "steps": 250, "lr": 5e-05, "grad_norm": 2.7836368083953857, "loss_scale": 4.0, "retries": 0, "seconds": 15.7, "max_memory_gb": 21.26, "pair": 0.9375113248825073, "sign": 0.7504861364141107, "keep": 0.5675593595951796, "leak": 0.04704580642282963, "replacement": 1.608295500278473} +{"step": 18, "steps": 250, "lr": 4.999797225177623e-05, "grad_norm": 3.3585617542266846, "loss_scale": 4.0, "retries": 0, "seconds": 16.5, "max_memory_gb": 21.28, "pair": 0.8343035131692886, "sign": 0.17601383361034095, "keep": 0.2560597095725825, "leak": 0.06062548598856665, "replacement": 1.065496638417244} +{"step": 19, "steps": 250, "lr": 4.999188937259492e-05, "grad_norm": 2.7240092754364014, "loss_scale": 4.0, "retries": 0, "seconds": 17.5, "max_memory_gb": 21.28, "pair": 0.7520032227039337, "sign": 0.03177349688485265, "keep": 0.2065592296421528, "leak": 0.3753592916764319, "replacement": 0.9650094360113144} +{"step": 20, "steps": 250, "lr": 4.998175245886032e-05, "grad_norm": 3.372936964035034, "loss_scale": 4.0, "retries": 0, "seconds": 18.4, "max_memory_gb": 21.28, "pair": 0.7334757000207901, "sign": 0.024835461284965277, "keep": 0.0025551835915393895, "leak": 0.11813023127615452, "replacement": 0.7767035812139511} +{"step": 21, "steps": 250, "lr": 4.996756333769319e-05, "grad_norm": 2.675210475921631, "loss_scale": 4.0, "retries": 0, "seconds": 19.3, "max_memory_gb": 21.28, "pair": 0.6115173846483231, "sign": 0.03483244078233838, "keep": 0.026516961091147095, "leak": 0.3629010207951069, "replacement": 0.7329173386096954} +{"step": 22, "steps": 250, "lr": 4.994932456660157e-05, "grad_norm": 2.546886682510376, "loss_scale": 4.0, "retries": 0, "seconds": 20.3, "max_memory_gb": 21.28, "pair": 0.6337757855653763, "sign": 0.052318686386570334, "keep": 0.09720349237613846, "leak": 0.2148254457861185, "replacement": 0.7622432559728622} +{"step": 23, "steps": 250, "lr": 4.992703943301973e-05, "grad_norm": 4.591028213500977, "loss_scale": 4.0, "retries": 0, "seconds": 21.2, "max_memory_gb": 21.28, "pair": 0.475619912147522, "sign": 0.023198247188702226, "keep": 0.6089353505026338, "leak": 0.2687326241284609, "replacement": 0.8588698357343674} +{"step": 24, "steps": 250, "lr": 4.9900711953715675e-05, "grad_norm": 1.0640532970428467, "loss_scale": 4.0, "retries": 0, "seconds": 22.1, "max_memory_gb": 21.28, "pair": 0.3124622367322445, "sign": 0.008417866309173405, "keep": 0.02132368623279035, "leak": 0.7683533728122711, "replacement": 0.519421361386776} +{"step": 25, "steps": 250, "lr": 4.987034687406713e-05, "grad_norm": 1.3736109733581543, "loss_scale": 4.0, "retries": 0, "seconds": 23.0, "max_memory_gb": 21.28, "pair": 0.3784944638609886, "sign": 0.08767771534621716, "keep": 0.6683022081169838, "leak": 0.8198692686855793, "replacement": 0.9614517912268639} +{"step": 26, "steps": 250, "lr": 4.9835949667206215e-05, "grad_norm": 2.669750213623047, "loss_scale": 4.0, "retries": 0, "seconds": 24.0, "max_memory_gb": 21.28, "pair": 0.2891210112720728, "sign": 0.05062205941067077, "keep": 0.0243731876835227, "leak": 1.218104257946834, "replacement": 0.6311446949839592} +{"step": 27, "steps": 250, "lr": 4.9797526533032945e-05, "grad_norm": 1.9612245559692383, "loss_scale": 4.0, "retries": 0, "seconds": 24.9, "max_memory_gb": 21.28, "pair": 0.3037219010293484, "sign": 0.14881762221921235, "keep": 0.1344706760205554, "leak": 1.2115812487900257, "replacement": 0.7482613697648048} +{"step": 28, "steps": 250, "lr": 4.9755084397097716e-05, "grad_norm": 3.634373188018799, "loss_scale": 4.0, "retries": 0, "seconds": 25.8, "max_memory_gb": 21.28, "pair": 0.5575753301382065, "sign": 0.3900362598942593, "keep": 1.1903586174594238, "leak": 0.9292518608272076, "replacement": 1.5800857245922089} +{"step": 29, "steps": 250, "lr": 4.970863090935304e-05, "grad_norm": 3.3767833709716797, "loss_scale": 4.0, "retries": 0, "seconds": 26.7, "max_memory_gb": 21.28, "pair": 0.563149206340313, "sign": 0.7454379536211491, "keep": 0.04369642457459122, "leak": 0.5961774858878925, "replacement": 1.106760747730732} +{"step": 30, "steps": 250, "lr": 4.965817444277468e-05, "grad_norm": 5.280265808105469, "loss_scale": 4.0, "retries": 0, "seconds": 27.6, "max_memory_gb": 21.28, "pair": 0.08362235967069864, "sign": 0.0027028757031075656, "keep": 0.035517162563792226, "leak": 2.0940396189689636, "replacement": 0.6262422800064087} +{"step": 31, "steps": 250, "lr": 4.9603724091852456e-05, "grad_norm": 4.150334358215332, "loss_scale": 4.0, "retries": 0, "seconds": 28.6, "max_memory_gb": 21.28, "pair": 0.4170549139380455, "sign": 0.18727400654461235, "keep": 0.401388147729449, "leak": 1.688918148051016, "replacement": 1.1336155235767365} +{"step": 32, "steps": 250, "lr": 4.9545289670951025e-05, "grad_norm": 4.034589767456055, "loss_scale": 4.0, "retries": 0, "seconds": 29.5, "max_memory_gb": 21.28, "pair": 0.29270096868276596, "sign": 0.0922958912269678, "keep": 0.01222427620086819, "leak": 2.2971081733703613, "replacement": 0.9192380905151367} +{"step": 33, "steps": 250, "lr": 4.948288171254089e-05, "grad_norm": 2.2992753982543945, "loss_scale": 4.0, "retries": 0, "seconds": 30.4, "max_memory_gb": 21.28, "pair": 0.5444852225482464, "sign": 0.18031947268173099, "keep": 0.07918830402195454, "leak": 0.29718767246231437, "replacement": 0.7485360279679298} +{"step": 34, "steps": 250, "lr": 4.941651146529998e-05, "grad_norm": 2.99568247795105, "loss_scale": 4.0, "retries": 0, "seconds": 31.3, "max_memory_gb": 21.28, "pair": 0.22298695892095566, "sign": 0.015417584334500134, "keep": 0.005113365142278781, "leak": 1.676499456167221, "replacement": 0.6523773074150085} +{"step": 35, "steps": 250, "lr": 4.9346190892086174e-05, "grad_norm": 2.18459153175354, "loss_scale": 4.0, "retries": 0, "seconds": 32.2, "max_memory_gb": 21.28, "pair": 0.2740394100546837, "sign": 0.04079060489311814, "keep": 0.004438004340045154, "leak": 1.218797117471695, "replacement": 0.6013530045747757} +{"step": 36, "steps": 250, "lr": 4.927193266778101e-05, "grad_norm": 3.1144840717315674, "loss_scale": 4.0, "retries": 0, "seconds": 33.1, "max_memory_gb": 21.28, "pair": 0.29298336431384087, "sign": 0.027144974912516773, "keep": 0.15062917851173552, "leak": 1.4200892029330134, "replacement": 0.7368927299976349} +{"step": 37, "steps": 250, "lr": 4.919375017700515e-05, "grad_norm": 2.2945687770843506, "loss_scale": 4.0, "retries": 0, "seconds": 34.0, "max_memory_gb": 21.28, "pair": 0.44125975109636784, "sign": 0.16444552957545966, "keep": 0.39091269303753506, "leak": 0.9860941525548697, "replacement": 0.9654623717069626} +{"step": 38, "steps": 250, "lr": 4.911165751170588e-05, "grad_norm": 2.399130344390869, "loss_scale": 4.0, "retries": 0, "seconds": 34.9, "max_memory_gb": 21.28, "pair": 0.2386665903031826, "sign": 0.07381688372697681, "keep": 0.03294443245977163, "leak": 1.5228877067565918, "replacement": 0.672769196331501} +{"step": 39, "steps": 250, "lr": 4.902566946861708e-05, "grad_norm": 3.2829079627990723, "loss_scale": 4.0, "retries": 0, "seconds": 35.8, "max_memory_gb": 21.28, "pair": 0.5406371541321278, "sign": 1.4296650071628392, "keep": 0.20564221579479636, "leak": 0.6033354587852955, "replacement": 1.5091246664524078} +{"step": 40, "steps": 250, "lr": 4.89358015465922e-05, "grad_norm": 1.6568753719329834, "loss_scale": 4.0, "retries": 0, "seconds": 36.7, "max_memory_gb": 21.28, "pair": 0.24302741140127182, "sign": 0.011511262215208262, "keep": 0.01499540778604569, "leak": 1.0666453689336777, "replacement": 0.5229420885443687} +{"step": 41, "steps": 250, "lr": 4.884206994381078e-05, "grad_norm": 3.681837797164917, "loss_scale": 4.0, "retries": 0, "seconds": 37.6, "max_memory_gb": 21.28, "pair": 0.6393745690584183, "sign": 0.8866005148738623, "keep": 1.0194266585167497, "leak": 0.23636072874069214, "replacement": 1.6514784097671509} +{"step": 42, "steps": 250, "lr": 4.874449155485866e-05, "grad_norm": 1.7312637567520142, "loss_scale": 4.0, "retries": 0, "seconds": 38.5, "max_memory_gb": 21.28, "pair": 0.3500712104141712, "sign": 0.03476573945954442, "keep": 0.0031315190135501325, "leak": 0.8261866569519043, "replacement": 0.5755665153264999} +{"step": 43, "steps": 250, "lr": 4.864308396768294e-05, "grad_norm": 2.3616483211517334, "loss_scale": 4.0, "retries": 0, "seconds": 39.4, "max_memory_gb": 21.28, "pair": 0.37108173966407776, "sign": 0.13923505088314414, "keep": 0.007165626873756992, "leak": 1.3307926654815674, "replacement": 0.7769802659749985} +{"step": 44, "steps": 250, "lr": 4.853786546042184e-05, "grad_norm": 3.8176844120025635, "loss_scale": 4.0, "retries": 0, "seconds": 40.3, "max_memory_gb": 21.28, "pair": 0.34628207981586456, "sign": 0.1394189785933122, "keep": 0.6473971704217547, "leak": 0.6688036993145943, "replacement": 0.9068910777568817} +{"step": 45, "steps": 250, "lr": 4.842885499811012e-05, "grad_norm": 5.39283561706543, "loss_scale": 4.0, "retries": 0, "seconds": 41.2, "max_memory_gb": 21.28, "pair": 0.5912055224180222, "sign": 0.6171908916439861, "keep": 1.7810799252183642, "leak": 0.28693721862509847, "replacement": 1.8620753064751625} +{"step": 46, "steps": 250, "lr": 4.831607222926078e-05, "grad_norm": 3.3191351890563965, "loss_scale": 4.0, "retries": 0, "seconds": 42.1, "max_memory_gb": 21.28, "pair": 0.5127335414290428, "sign": 0.23953187954612076, "keep": 0.3333657877519727, "leak": 0.2623925106599927, "replacement": 0.8647804781794548} +{"step": 47, "steps": 250, "lr": 4.8199537482323545e-05, "grad_norm": 2.0024538040161133, "loss_scale": 4.0, "retries": 0, "seconds": 43.0, "max_memory_gb": 21.28, "pair": 0.46394745260477066, "sign": 0.08752642758190632, "keep": 0.01657249881827738, "leak": 0.4020581915974617, "replacement": 0.6165114492177963} +{"step": 48, "steps": 250, "lr": 4.807927176202073e-05, "grad_norm": 1.4417222738265991, "loss_scale": 4.0, "retries": 0, "seconds": 43.9, "max_memory_gb": 21.28, "pair": 0.36231207475066185, "sign": 0.024498372513335198, "keep": 0.03812639130046591, "leak": 0.5626419186592102, "replacement": 0.5342849567532539} +{"step": 49, "steps": 250, "lr": 4.795529674556129e-05, "grad_norm": 1.9419283866882324, "loss_scale": 4.0, "retries": 0, "seconds": 44.8, "max_memory_gb": 21.28, "pair": 0.3733738921582699, "sign": 0.041949406964704394, "keep": 0.02951539877903997, "leak": 0.4492519088089466, "replacement": 0.5214192569255829} +{"step": 50, "steps": 250, "lr": 4.782763477873361e-05, "grad_norm": 1.818345308303833, "loss_scale": 4.0, "retries": 0, "seconds": 45.7, "max_memory_gb": 21.28, "pair": 0.44394949823617935, "sign": 0.08106941729784012, "keep": 0.22840555719903932, "leak": 0.42963746935129166, "replacement": 0.706096351146698} +{"step": 51, "steps": 250, "lr": 4.769630887187782e-05, "grad_norm": 3.5968894958496094, "loss_scale": 4.0, "retries": 0, "seconds": 46.6, "max_memory_gb": 21.28, "pair": 0.35654402524232864, "sign": 0.022637961024884135, "keep": 0.7883921679804189, "leak": 0.42796442401595414, "replacement": 0.8690502196550369} +{"step": 52, "steps": 250, "lr": 4.756134269573832e-05, "grad_norm": 2.1843903064727783, "loss_scale": 4.0, "retries": 0, "seconds": 47.5, "max_memory_gb": 21.28, "pair": 0.4682649001479149, "sign": 0.258952776202932, "keep": 0.140934943126922, "leak": 0.3202000707387924, "replacement": 0.7482587695121765} +{"step": 53, "steps": 250, "lr": 4.7422760577197226e-05, "grad_norm": 1.74850594997406, "loss_scale": 4.0, "retries": 0, "seconds": 48.4, "max_memory_gb": 21.28, "pair": 0.3771284893155098, "sign": 0.08000434207497165, "keep": 0.2103850789135322, "leak": 0.9387374110519886, "replacement": 0.7570075392723083} +{"step": 54, "steps": 250, "lr": 4.728058749488963e-05, "grad_norm": 2.479848623275757, "loss_scale": 4.0, "retries": 0, "seconds": 49.3, "max_memory_gb": 21.28, "pair": 0.36898548156023026, "sign": 0.08219297183677554, "keep": 0.17416745822993107, "leak": 0.7350587286055088, "replacement": 0.6809303760528564} +{"step": 55, "steps": 250, "lr": 4.713484907470133e-05, "grad_norm": 1.5392564535140991, "loss_scale": 4.0, "retries": 0, "seconds": 50.2, "max_memory_gb": 21.28, "pair": 0.2951694540679455, "sign": 0.021704741287976503, "keep": 0.014369185439136345, "leak": 0.46738461405038834, "replacement": 0.4300525635480881} +{"step": 56, "steps": 250, "lr": 4.6985571585149876e-05, "grad_norm": 1.3033981323242188, "loss_scale": 4.0, "retries": 0, "seconds": 51.0, "max_memory_gb": 21.28, "pair": 0.30066005140542984, "sign": 0.014936157269403338, "keep": 0.022138735745102167, "leak": 0.6386716589331627, "replacement": 0.478865422308445} +{"step": 57, "steps": 250, "lr": 4.6832781932649884e-05, "grad_norm": 2.5916025638580322, "loss_scale": 4.0, "retries": 0, "seconds": 51.9, "max_memory_gb": 21.28, "pair": 0.27950418181717396, "sign": 0.28817709477152675, "keep": 0.00804812926435261, "leak": 0.7006075646786485, "replacement": 0.602768711745739} +{"step": 58, "steps": 250, "lr": 4.667650765666325e-05, "grad_norm": 3.762779712677002, "loss_scale": 4.0, "retries": 0, "seconds": 52.9, "max_memory_gb": 21.28, "pair": 0.12204772047698498, "sign": 0.009489417367149144, "keep": 0.011833648237370653, "leak": 1.387240156531334, "replacement": 0.479519285261631} +{"step": 59, "steps": 250, "lr": 4.651677692473538e-05, "grad_norm": 2.332944393157959, "loss_scale": 4.0, "retries": 0, "seconds": 53.8, "max_memory_gb": 21.28, "pair": 0.3590999208390713, "sign": 0.0488014736038167, "keep": 0.025975826720241457, "leak": 0.5210050335153937, "replacement": 0.5267398208379745} +{"step": 60, "steps": 250, "lr": 4.635361852741808e-05, "grad_norm": 2.52349853515625, "loss_scale": 4.0, "retries": 0, "seconds": 54.7, "max_memory_gb": 21.28, "pair": 0.2432661708444357, "sign": 0.0369460741057992, "keep": 0.10603045385505538, "leak": 0.9545275304226379, "replacement": 0.5533863008022308} +{"step": 61, "steps": 250, "lr": 4.6187061873080264e-05, "grad_norm": 5.890431880950928, "loss_scale": 4.0, "retries": 0, "seconds": 55.6, "max_memory_gb": 21.28, "pair": 0.35598795115947723, "sign": 0.2516071933787316, "keep": 0.03559648268856108, "leak": 2.269575485493988, "replacement": 1.0669836699962616} +{"step": 62, "steps": 250, "lr": 4.601713698260727e-05, "grad_norm": 1.6515759229660034, "loss_scale": 4.0, "retries": 0, "seconds": 56.5, "max_memory_gb": 21.28, "pair": 0.18455838784575462, "sign": 0.018366553587839007, "keep": 0.028609926404897124, "leak": 0.9434185177087784, "replacement": 0.44390125572681427} +{"step": 63, "steps": 250, "lr": 4.5843874483989746e-05, "grad_norm": 1.620306134223938, "loss_scale": 4.0, "retries": 0, "seconds": 57.5, "max_memory_gb": 21.28, "pair": 0.29601601511240005, "sign": 0.11061693145893514, "keep": 0.038315639132633805, "leak": 0.3392909746617079, "replacement": 0.4553050398826599} +{"step": 64, "steps": 250, "lr": 4.566730560680316e-05, "grad_norm": 1.138340950012207, "loss_scale": 4.0, "retries": 0, "seconds": 58.4, "max_memory_gb": 21.28, "pair": 0.22476908937096596, "sign": 0.01620496017858386, "keep": 0.019748590886592865, "leak": 0.5882434919476509, "replacement": 0.38980673253536224} +{"step": 65, "steps": 250, "lr": 4.548746217657878e-05, "grad_norm": 5.310763835906982, "loss_scale": 4.0, "retries": 0, "seconds": 59.4, "max_memory_gb": 21.28, "pair": 0.21662492491304874, "sign": 0.17048258567228913, "keep": 0.15028307540342212, "leak": 2.0035809874534607, "replacement": 0.877903014421463} +{"step": 66, "steps": 250, "lr": 4.5304376609067396e-05, "grad_norm": 3.117495059967041, "loss_scale": 4.0, "retries": 0, "seconds": 60.4, "max_memory_gb": 21.28, "pair": 0.2515617813915014, "sign": 0.13482626472250558, "keep": 0.33214291323383094, "leak": 0.7034238614141941, "replacement": 0.6609023548662663} +{"step": 67, "steps": 250, "lr": 4.5118081904396504e-05, "grad_norm": 3.263566017150879, "loss_scale": 4.0, "retries": 0, "seconds": 61.3, "max_memory_gb": 21.28, "pair": 0.11387064214795828, "sign": 0.013492143829353154, "keep": 0.011775712817325257, "leak": 1.58372101187706, "replacement": 0.5224348232150078} +{"step": 68, "steps": 250, "lr": 4.4928611641122225e-05, "grad_norm": 1.4670132398605347, "loss_scale": 4.0, "retries": 0, "seconds": 62.3, "max_memory_gb": 21.28, "pair": 0.22680913284420967, "sign": 0.023126936284825206, "keep": 0.00685008700452272, "leak": 0.6559518873691559, "replacement": 0.40578562766313553} +{"step": 69, "steps": 250, "lr": 4.473599997017701e-05, "grad_norm": 5.811870098114014, "loss_scale": 4.0, "retries": 0, "seconds": 63.2, "max_memory_gb": 21.28, "pair": 0.5911357626318932, "sign": 0.9064351879060268, "keep": 0.9272286787163466, "leak": 1.1156366933137178, "replacement": 1.7868769764900208} +{"step": 70, "steps": 250, "lr": 4.4540281608714053e-05, "grad_norm": 5.414148807525635, "loss_scale": 4.0, "retries": 0, "seconds": 64.2, "max_memory_gb": 21.28, "pair": 0.3193701468408108, "sign": 0.03683451801771298, "keep": 0.004923911939840764, "leak": 1.915306430310011, "replacement": 0.8190759643912315} +{"step": 71, "steps": 250, "lr": 4.434149183384977e-05, "grad_norm": 1.957858920097351, "loss_scale": 4.0, "retries": 0, "seconds": 65.1, "max_memory_gb": 21.28, "pair": 0.13797025568783283, "sign": 0.007118145935237408, "keep": 0.022285579908611908, "leak": 0.9710529446601868, "replacement": 0.3954353556036949} +{"step": 72, "steps": 250, "lr": 4.413966647630534e-05, "grad_norm": 3.0539262294769287, "loss_scale": 4.0, "retries": 0, "seconds": 66.1, "max_memory_gb": 21.28, "pair": 0.2767352182418108, "sign": 0.18773786153178662, "keep": 0.03473736147861928, "leak": 1.1417370364069939, "replacement": 0.6734070852398872} +{"step": 73, "steps": 250, "lr": 4.3934841913948323e-05, "grad_norm": 2.2220308780670166, "loss_scale": 4.0, "retries": 0, "seconds": 67.0, "max_memory_gb": 21.28, "pair": 0.29798005148768425, "sign": 0.10141343099530786, "keep": 0.39931009134670603, "leak": 0.39541179686784744, "replacement": 0.6471947431564331} +{"step": 74, "steps": 250, "lr": 4.372705506523583e-05, "grad_norm": 1.7593739032745361, "loss_scale": 4.0, "retries": 0, "seconds": 67.9, "max_memory_gb": 21.28, "pair": 0.31336712650954723, "sign": 0.19016193901188672, "keep": 0.1306013487292148, "leak": 0.43358511105179787, "replacement": 0.582145057618618} +{"step": 75, "steps": 250, "lr": 4.351634338256017e-05, "grad_norm": 2.5011842250823975, "loss_scale": 4.0, "retries": 0, "seconds": 68.9, "max_memory_gb": 21.28, "pair": 0.6327514797449112, "sign": 0.5259487110888585, "keep": 0.9421524815261364, "leak": 0.20197341218590736, "replacement": 1.4172953814268112} +{"step": 76, "steps": 250, "lr": 4.330274484549825e-05, "grad_norm": 1.5490792989730835, "loss_scale": 4.0, "retries": 0, "seconds": 69.8, "max_memory_gb": 21.28, "pair": 0.2395678348839283, "sign": 0.06293961324263364, "keep": 0.01798149417663808, "leak": 0.5582481268793344, "replacement": 0.41959042847156525} +{"step": 77, "steps": 250, "lr": 4.308629795396599e-05, "grad_norm": 1.7139058113098145, "loss_scale": 4.0, "retries": 0, "seconds": 70.7, "max_memory_gb": 21.28, "pair": 0.2723889425396919, "sign": 0.010762674501165748, "keep": 0.006796340356231667, "leak": 0.8691423926793505, "replacement": 0.4984540417790413} +{"step": 78, "steps": 250, "lr": 4.286704172127894e-05, "grad_norm": 2.784790277481079, "loss_scale": 4.0, "retries": 0, "seconds": 71.7, "max_memory_gb": 21.28, "pair": 0.3224814683198929, "sign": 0.4061262473696843, "keep": 0.002828062053595204, "leak": 0.40176042448729277, "replacement": 0.6273987218737602} +{"step": 79, "steps": 250, "lr": 4.264501566712033e-05, "grad_norm": 2.2725107669830322, "loss_scale": 4.0, "retries": 0, "seconds": 72.6, "max_memory_gb": 21.28, "pair": 0.2891830652952194, "sign": 0.01879276684485376, "keep": 0.019450949997917633, "leak": 0.6279664524081454, "replacement": 0.46529654413461685} +{"step": 80, "steps": 250, "lr": 4.2420259810417894e-05, "grad_norm": 1.4226493835449219, "loss_scale": 4.0, "retries": 0, "seconds": 73.5, "max_memory_gb": 21.28, "pair": 0.3172982223331928, "sign": 0.08576987904962152, "keep": 0.019689439494015915, "leak": 0.6875851713120937, "replacement": 0.5419241823256016} +{"step": 81, "steps": 250, "lr": 4.219281466213066e-05, "grad_norm": 1.7214473485946655, "loss_scale": 4.0, "retries": 0, "seconds": 74.4, "max_memory_gb": 21.28, "pair": 0.30041056126356125, "sign": 0.05758500541560352, "keep": 0.1074622180312872, "leak": 0.5606655515730381, "replacement": 0.5231005549430847} +{"step": 82, "steps": 250, "lr": 4.1962721217947135e-05, "grad_norm": 2.6292402744293213, "loss_scale": 4.0, "retries": 0, "seconds": 75.3, "max_memory_gb": 21.28, "pair": 0.4138267785310745, "sign": 1.0030590750975534, "keep": 0.03713496692944318, "leak": 0.6899854391813278, "replacement": 1.1064201295375824} +{"step": 83, "steps": 250, "lr": 4.173002095089599e-05, "grad_norm": 1.5671101808547974, "loss_scale": 4.0, "retries": 0, "seconds": 76.2, "max_memory_gb": 21.28, "pair": 0.3385962937027216, "sign": 0.02946538748801686, "keep": 0.3585316446842626, "leak": 0.41521965712308884, "replacement": 0.6363997161388397} +{"step": 84, "steps": 250, "lr": 4.149475580387084e-05, "grad_norm": 1.3272294998168945, "loss_scale": 4.0, "retries": 0, "seconds": 77.2, "max_memory_gb": 21.28, "pair": 0.16418926417827606, "sign": 0.02418583451071754, "keep": 0.024498550602743308, "leak": 0.46132298931479454, "replacement": 0.30386220291256905} +{"step": 85, "steps": 250, "lr": 4.1256968182070275e-05, "grad_norm": 3.957383871078491, "loss_scale": 4.0, "retries": 0, "seconds": 78.2, "max_memory_gb": 21.28, "pair": 0.06356408447027206, "sign": 0.011019779965863563, "keep": 0.0018309065615085274, "leak": 0.6387255601584911, "replacement": 0.22967081889510155} +{"step": 86, "steps": 250, "lr": 4.101670094535453e-05, "grad_norm": 2.824700355529785, "loss_scale": 4.0, "retries": 0, "seconds": 79.1, "max_memory_gb": 21.28, "pair": 0.27699729055166245, "sign": 0.20527915563434362, "keep": 0.06891898345202208, "leak": 0.41725376434624195, "replacement": 0.518409788608551} +{"step": 87, "steps": 250, "lr": 4.077399740052027e-05, "grad_norm": 1.9522839784622192, "loss_scale": 4.0, "retries": 0, "seconds": 80.0, "max_memory_gb": 21.28, "pair": 0.24294418096542358, "sign": 0.09923498018179089, "keep": 0.007276841992279515, "leak": 0.9597061574459076, "replacement": 0.5361266434192657} +{"step": 88, "steps": 250, "lr": 4.0528901293494775e-05, "grad_norm": 8.516611099243164, "loss_scale": 4.0, "retries": 0, "seconds": 80.9, "max_memory_gb": 21.28, "pair": 0.4044602564536035, "sign": 0.35490823921281844, "keep": 0.21801711804869228, "leak": 0.6497542392462492, "replacement": 0.853361539542675} +{"step": 89, "steps": 250, "lr": 4.0281456801451004e-05, "grad_norm": 2.890470027923584, "loss_scale": 4.0, "retries": 0, "seconds": 81.8, "max_memory_gb": 21.28, "pair": 0.28955673426389694, "sign": 0.12026975621120073, "keep": 0.008936093257716493, "leak": 0.01822691951019806, "replacement": 0.35871637985110283} +{"step": 90, "steps": 250, "lr": 4.0031708524844884e-05, "grad_norm": 1.9123311042785645, "loss_scale": 4.0, "retries": 0, "seconds": 82.7, "max_memory_gb": 21.28, "pair": 0.24049396626651287, "sign": 0.04380819768994115, "keep": 0.021925262262811884, "leak": 0.586615115404129, "replacement": 0.42001447454094887} +{"step": 91, "steps": 250, "lr": 3.977970147937635e-05, "grad_norm": 15.941851615905762, "loss_scale": 4.0, "retries": 0, "seconds": 83.6, "max_memory_gb": 21.28, "pair": 0.7755090463906527, "sign": 0.46670897281728685, "keep": 0.8167758800555021, "leak": 1.362248957157135, "replacement": 1.757813636213541} +{"step": 92, "steps": 250, "lr": 3.952548108787553e-05, "grad_norm": 7.355874061584473, "loss_scale": 4.0, "retries": 0, "seconds": 84.6, "max_memory_gb": 21.28, "pair": 0.1677514836192131, "sign": 0.4423381056258222, "keep": 0.6530735393002942, "leak": 0.7211956158280373, "replacement": 0.8957561869174242} +{"step": 93, "steps": 250, "lr": 3.9269093172115496e-05, "grad_norm": 1.6230618953704834, "loss_scale": 4.0, "retries": 0, "seconds": 85.5, "max_memory_gb": 21.28, "pair": 0.14391530584543943, "sign": 0.017078503442462534, "keep": 0.06672277871984988, "leak": 0.6859950318466872, "replacement": 0.35731470212340355} +{"step": 94, "steps": 250, "lr": 3.901058394455323e-05, "grad_norm": 2.869311809539795, "loss_scale": 4.0, "retries": 0, "seconds": 86.4, "max_memory_gb": 21.28, "pair": 0.16549440938979387, "sign": 0.020484365988522768, "keep": 0.08791840472258627, "leak": 1.4483654350042343, "replacement": 0.5817871615290642} +{"step": 95, "steps": 250, "lr": 3.875e-05, "grad_norm": 5.9296674728393555, "loss_scale": 4.0, "retries": 0, "seconds": 87.4, "max_memory_gb": 21.28, "pair": 0.34923407062888145, "sign": 0.3233634930802509, "keep": 1.063456338364631, "leak": 0.9065359123051167, "replacement": 1.2692779451608658} +{"step": 96, "steps": 250, "lr": 3.8487388307223005e-05, "grad_norm": 3.5554513931274414, "loss_scale": 4.0, "retries": 0, "seconds": 88.3, "max_memory_gb": 21.28, "pair": 0.2430630959570408, "sign": 0.0739857378648594, "keep": 0.8870850157109089, "leak": 0.7488421555608511, "replacement": 0.9108089506626129} +{"step": 97, "steps": 250, "lr": 3.822279620047943e-05, "grad_norm": 2.749063014984131, "loss_scale": 4.0, "retries": 0, "seconds": 89.2, "max_memory_gb": 21.28, "pair": 0.1348447371274233, "sign": 0.009044756647199392, "keep": 0.008856394935719436, "leak": 0.9005914777517319, "replacement": 0.3689431697130203} +{"step": 98, "steps": 250, "lr": 3.795627137098479e-05, "grad_norm": 1.4394625425338745, "loss_scale": 4.0, "retries": 0, "seconds": 90.2, "max_memory_gb": 21.28, "pair": 0.25055550038814545, "sign": 0.08002458047121763, "keep": 0.03647592633524255, "leak": 0.7661116383969784, "replacement": 0.5003336556255817} +{"step": 99, "steps": 250, "lr": 3.7687861858316844e-05, "grad_norm": 3.2761647701263428, "loss_scale": 4.0, "retries": 0, "seconds": 91.1, "max_memory_gb": 21.28, "pair": 0.3313068337738514, "sign": 0.5851781200617552, "keep": 0.0045349582082963025, "leak": 1.033765021711588, "replacement": 0.8846046254038811} +{"step": 100, "steps": 250, "lr": 3.741761604175668e-05, "grad_norm": 5.246603488922119, "loss_scale": 4.0, "retries": 0, "seconds": 92.0, "max_memory_gb": 21.28, "pair": 0.5274466350674629, "sign": 0.9645754855591804, "keep": 0.12156520341522992, "leak": 1.8056081873364747, "replacement": 1.5219190865755081} +{"step": 101, "steps": 250, "lr": 3.714558263156872e-05, "grad_norm": 9.290416717529297, "loss_scale": 4.0, "retries": 0, "seconds": 92.9, "max_memory_gb": 21.28, "pair": 0.10278894007205963, "sign": 0.005606392631307244, "keep": 0.004453090950846672, "leak": 2.337362229824066, "replacement": 0.6921592205762863} +{"step": 102, "steps": 250, "lr": 3.6871810660220904e-05, "grad_norm": 1.9755359888076782, "loss_scale": 4.0, "retries": 0, "seconds": 93.9, "max_memory_gb": 21.28, "pair": 0.21336006931960583, "sign": 0.015021013095974922, "keep": 0.06277423561550677, "leak": 0.48127930890768766, "replacement": 0.3725775107741356} +{"step": 103, "steps": 250, "lr": 3.659634947354686e-05, "grad_norm": 2.6948862075805664, "loss_scale": 4.0, "retries": 0, "seconds": 94.8, "max_memory_gb": 21.28, "pair": 0.20363958925008774, "sign": 0.02273183545912616, "keep": 0.015278778962965589, "leak": 1.2861363776028156, "replacement": 0.5441789850592613} +{"step": 104, "steps": 250, "lr": 3.631924872185169e-05, "grad_norm": 2.284383773803711, "loss_scale": 4.0, "retries": 0, "seconds": 95.6, "max_memory_gb": 21.28, "pair": 0.21788097638636827, "sign": 0.07861727586714551, "keep": 0.004325032787164673, "leak": 0.16919487348059192, "replacement": 0.30165084078907967} +{"step": 105, "steps": 250, "lr": 3.6040558350962674e-05, "grad_norm": 2.6418843269348145, "loss_scale": 4.0, "retries": 0, "seconds": 96.6, "max_memory_gb": 21.28, "pair": 0.312256820499897, "sign": 0.017624500673264265, "keep": 0.03622741927392781, "leak": 0.649739078071434, "replacement": 0.5016175657510757} +{"step": 106, "steps": 250, "lr": 3.576032859322687e-05, "grad_norm": 3.4562973976135254, "loss_scale": 4.0, "retries": 0, "seconds": 97.5, "max_memory_gb": 21.28, "pair": 0.3632151111960411, "sign": 0.0330829918384552, "keep": 0.018339640060730744, "leak": 0.20873783668503165, "replacement": 0.44111089408397675} +{"step": 107, "steps": 250, "lr": 3.5478609958457056e-05, "grad_norm": 4.657817363739014, "loss_scale": 4.0, "retries": 0, "seconds": 98.4, "max_memory_gb": 21.28, "pair": 0.18101573456078768, "sign": 0.20177667762618512, "keep": 0.4247102763025623, "leak": 0.8070292106131092, "replacement": 0.6960165034979582} +{"step": 108, "steps": 250, "lr": 3.519545322482755e-05, "grad_norm": 12.842348098754883, "loss_scale": 4.0, "retries": 0, "seconds": 99.4, "max_memory_gb": 21.28, "pair": 0.4979218766093254, "sign": 0.11455696739722043, "keep": 0.13358977966072416, "leak": 1.383201453834772, "replacement": 0.9677956458181143} +{"step": 109, "steps": 250, "lr": 3.491090942972179e-05, "grad_norm": 7.443319320678711, "loss_scale": 4.0, "retries": 0, "seconds": 100.3, "max_memory_gb": 21.28, "pair": 0.4866698868572712, "sign": 0.15861993236467242, "keep": 0.15121686295606196, "leak": 0.6939690038561821, "replacement": 0.8150805309414864} +{"step": 110, "steps": 250, "lr": 3.4625029860533135e-05, "grad_norm": 12.193811416625977, "loss_scale": 4.0, "retries": 0, "seconds": 101.3, "max_memory_gb": 21.28, "pair": 0.25004973262548447, "sign": 0.04879198828712106, "keep": 0.054146270267665386, "leak": 3.6696806754916906, "replacement": 1.2189390435814857} +{"step": 111, "steps": 250, "lr": 3.433786604542056e-05, "grad_norm": 5.671184062957764, "loss_scale": 4.0, "retries": 0, "seconds": 102.2, "max_memory_gb": 21.28, "pair": 0.2061264831572771, "sign": 1.0182828482647892, "keep": 0.0021889409772484214, "leak": 0.6136684259399772, "replacement": 0.8697794731706381} +{"step": 112, "steps": 250, "lr": 3.404946974402104e-05, "grad_norm": 2.942440986633301, "loss_scale": 4.0, "retries": 0, "seconds": 103.2, "max_memory_gb": 21.28, "pair": 0.17081857845187187, "sign": 0.009263478918001056, "keep": 0.004755053472763393, "leak": 0.9428103836253285, "replacement": 0.4135304354131222} +{"step": 113, "steps": 250, "lr": 3.375989293812019e-05, "grad_norm": 2.146007537841797, "loss_scale": 4.0, "retries": 0, "seconds": 104.1, "max_memory_gb": 21.28, "pair": 0.23588990792632103, "sign": 0.026317530078813434, "keep": 0.00796801810793113, "leak": 0.283822764409706, "replacement": 0.3239883817732334} +{"step": 114, "steps": 250, "lr": 3.3469187822282766e-05, "grad_norm": 1.913583517074585, "loss_scale": 4.0, "retries": 0, "seconds": 105.1, "max_memory_gb": 21.28, "pair": 0.24838173016905785, "sign": 0.03279356472194195, "keep": 0.013268619542941451, "leak": 0.8523237016052008, "replacement": 0.48449374735355377} +{"step": 115, "steps": 250, "lr": 3.317740679444499e-05, "grad_norm": 2.67612361907959, "loss_scale": 4.0, "retries": 0, "seconds": 106.0, "max_memory_gb": 21.28, "pair": 0.24093009904026985, "sign": 0.01131284050643444, "keep": 0.004894424164376687, "leak": 0.3366349958232604, "replacement": 0.33319248259067535} +{"step": 116, "steps": 250, "lr": 3.2884602446470056e-05, "grad_norm": 5.38584041595459, "loss_scale": 4.0, "retries": 0, "seconds": 106.9, "max_memory_gb": 21.28, "pair": 0.1835957570001483, "sign": 0.4027223990415223, "keep": 0.5067502925940062, "leak": 0.1352937133051455, "replacement": 0.6721555851399899} +{"step": 117, "steps": 250, "lr": 3.259082755466876e-05, "grad_norm": 5.578410625457764, "loss_scale": 4.0, "retries": 0, "seconds": 107.8, "max_memory_gb": 21.29, "pair": 0.11415577307343483, "sign": 0.046706313398317434, "keep": 0.16453848385208403, "leak": 1.094416856765747, "replacement": 0.49338238313794136} +{"step": 118, "steps": 250, "lr": 3.229613507028686e-05, "grad_norm": 4.0698089599609375, "loss_scale": 4.0, "retries": 0, "seconds": 108.8, "max_memory_gb": 21.29, "pair": 0.2885897671803832, "sign": 0.43730664876056835, "keep": 0.7158252566837291, "leak": 0.4604818020015955, "replacement": 0.9802762139588594} +{"step": 119, "steps": 250, "lr": 3.2000578109961004e-05, "grad_norm": 4.014986991882324, "loss_scale": 4.0, "retries": 0, "seconds": 109.7, "max_memory_gb": 21.29, "pair": 0.26893740775994956, "sign": 0.18702644800941925, "keep": 0.011894175757220182, "leak": 0.2163133341819048, "replacement": 0.42247606441378593} +{"step": 120, "steps": 250, "lr": 3.1704209946144616e-05, "grad_norm": 6.995253086090088, "loss_scale": 4.0, "retries": 0, "seconds": 110.7, "max_memory_gb": 21.29, "pair": 0.31153538450598717, "sign": 0.7726057505933568, "keep": 1.5098246667184867, "leak": 0.29952150397002697, "replacement": 1.5276309698820114} +{"step": 121, "steps": 250, "lr": 3.140708399750594e-05, "grad_norm": 3.7054052352905273, "loss_scale": 4.0, "retries": 0, "seconds": 111.6, "max_memory_gb": 21.29, "pair": 0.4221614357084036, "sign": 0.32744732798892073, "keep": 0.1091275293124454, "leak": 0.37987090775277466, "replacement": 0.7354165781289339} +{"step": 122, "steps": 250, "lr": 3.1109253819299606e-05, "grad_norm": 4.972635746002197, "loss_scale": 4.0, "retries": 0, "seconds": 112.5, "max_memory_gb": 21.29, "pair": 0.35779536701738834, "sign": 0.11712550863740034, "keep": 0.10295414912559409, "leak": 0.7067548297345638, "replacement": 0.644523911178112} +{"step": 123, "steps": 250, "lr": 3.081077309371361e-05, "grad_norm": 4.005517482757568, "loss_scale": 4.0, "retries": 0, "seconds": 113.5, "max_memory_gb": 21.29, "pair": 0.33998618921032175, "sign": 0.9868019563145936, "keep": 0.023039056308334693, "leak": 0.41436651023104787, "replacement": 0.9484983161091805} +{"step": 124, "steps": 250, "lr": 3.051169562019339e-05, "grad_norm": 3.3448398113250732, "loss_scale": 4.0, "retries": 0, "seconds": 114.4, "max_memory_gb": 21.29, "pair": 0.4307226464152336, "sign": 0.577943058568053, "keep": 2.0038930240189075, "leak": 0.25139050278812647, "replacement": 1.7844882383942604} +{"step": 125, "steps": 250, "lr": 3.021207530574477e-05, "grad_norm": 4.416796684265137, "loss_scale": 4.0, "retries": 0, "seconds": 115.3, "max_memory_gb": 21.29, "pair": 0.2558937161229551, "sign": 0.2617918390315026, "keep": 0.26093629482465985, "leak": 0.9238680936396122, "replacement": 0.7482247874140739} +{"step": 126, "steps": 250, "lr": 2.991196615521758e-05, "grad_norm": 7.154295444488525, "loss_scale": 4.0, "retries": 0, "seconds": 116.2, "max_memory_gb": 21.29, "pair": 0.6776789333671331, "sign": 0.539754765573889, "keep": 1.2563084904104471, "leak": 1.1373742772266269, "replacement": 1.8600541353225708} +{"step": 127, "steps": 250, "lr": 2.9611422261571536e-05, "grad_norm": 1.515680193901062, "loss_scale": 4.0, "retries": 0, "seconds": 117.1, "max_memory_gb": 21.29, "pair": 0.12622212525457144, "sign": 0.008219519804697484, "keep": 0.020812967030792606, "leak": 0.2721168437274173, "replacement": 0.20876758452504873} +{"step": 128, "steps": 250, "lr": 2.931049779612634e-05, "grad_norm": 6.589985370635986, "loss_scale": 4.0, "retries": 0, "seconds": 118.0, "max_memory_gb": 21.29, "pair": 0.19027190376073122, "sign": 0.7746667338942643, "keep": 0.4140250583877787, "leak": 0.745326042175293, "replacement": 0.9709493070840836} +{"step": 129, "steps": 250, "lr": 2.9009246998797557e-05, "grad_norm": 1.3143759965896606, "loss_scale": 4.0, "retries": 0, "seconds": 119.0, "max_memory_gb": 21.29, "pair": 0.22846772894263268, "sign": 0.07451476418646052, "keep": 0.13454433344304562, "leak": 0.5643926579505205, "replacement": 0.4740954414010048} +{"step": 130, "steps": 250, "lr": 2.8707724168320238e-05, "grad_norm": 4.428045272827148, "loss_scale": 4.0, "retries": 0, "seconds": 119.9, "max_memory_gb": 21.29, "pair": 0.30803724750876427, "sign": 1.3883439358323812, "keep": 0.04913746013698983, "leak": 0.14344637840986252, "replacement": 1.0626395419239998} +{"step": 131, "steps": 250, "lr": 2.840598365246184e-05, "grad_norm": 20.014511108398438, "loss_scale": 4.0, "retries": 0, "seconds": 120.8, "max_memory_gb": 21.29, "pair": 0.6299386895261705, "sign": 0.6717204645974562, "keep": 1.1818866366243128, "leak": 2.0130719542503357, "replacement": 2.0600103065371513} +{"step": 132, "steps": 250, "lr": 2.8104079838226383e-05, "grad_norm": 3.392252206802368, "loss_scale": 4.0, "retries": 0, "seconds": 121.7, "max_memory_gb": 21.29, "pair": 0.0877508707344532, "sign": 0.009953177359420806, "keep": 0.41113449817225955, "leak": 0.8976999260485172, "replacement": 0.522719694301486} +{"step": 133, "steps": 250, "lr": 2.7802067142051446e-05, "grad_norm": 5.199711799621582, "loss_scale": 4.0, "retries": 0, "seconds": 122.6, "max_memory_gb": 21.29, "pair": 0.5395777970552444, "sign": 0.6284527170937508, "keep": 1.1287983474321663, "leak": 1.0227954760193825, "replacement": 1.6739022210240364} +{"step": 134, "steps": 250, "lr": 2.7500000000000004e-05, "grad_norm": 3.2304062843322754, "loss_scale": 4.0, "retries": 0, "seconds": 123.6, "max_memory_gb": 21.29, "pair": 0.15685038035735488, "sign": 0.2743935298130964, "keep": 0.4665936975296745, "leak": 0.03791614985675551, "replacement": 0.5368229888845235} +{"step": 135, "steps": 250, "lr": 2.7197932857948566e-05, "grad_norm": 1.560732126235962, "loss_scale": 4.0, "retries": 0, "seconds": 124.4, "max_memory_gb": 21.29, "pair": 0.18131209909915924, "sign": 0.01430077844997868, "keep": 0.049888293258845806, "leak": 0.564932711655274, "replacement": 0.35463981330394745} +{"step": 136, "steps": 250, "lr": 2.689592016177363e-05, "grad_norm": 3.6128296852111816, "loss_scale": 4.0, "retries": 0, "seconds": 125.3, "max_memory_gb": 21.29, "pair": 0.21255682781338692, "sign": 0.33183740777894855, "keep": 0.0370746847911505, "leak": 0.9581964164972305, "replacement": 0.6365619748830795} +{"step": 137, "steps": 250, "lr": 2.659401634753817e-05, "grad_norm": 4.516939163208008, "loss_scale": 4.0, "retries": 0, "seconds": 126.3, "max_memory_gb": 21.29, "pair": 0.1853133300319314, "sign": 0.01824698073323816, "keep": 0.027178158503375016, "leak": 1.0217855535447598, "replacement": 0.46347227692604065} +{"step": 138, "steps": 250, "lr": 2.6292275831679764e-05, "grad_norm": 1.8840172290802002, "loss_scale": 4.0, "retries": 0, "seconds": 127.3, "max_memory_gb": 21.29, "pair": 0.15149431489408016, "sign": 0.025744167855009437, "keep": 0.06705989665351808, "leak": 0.41854066849919036, "replacement": 0.30253151804208755} +{"step": 139, "steps": 250, "lr": 2.5990753001202445e-05, "grad_norm": 5.225622653961182, "loss_scale": 4.0, "retries": 0, "seconds": 128.2, "max_memory_gb": 21.29, "pair": 0.21321683563292027, "sign": 0.2048908033175394, "keep": 0.015517719206400216, "leak": 0.8825162947177887, "replacement": 0.5440501868724823} +{"step": 140, "steps": 250, "lr": 2.5689502203873672e-05, "grad_norm": 1.4962142705917358, "loss_scale": 4.0, "retries": 0, "seconds": 129.2, "max_memory_gb": 21.29, "pair": 0.168144048191607, "sign": 0.030215676699299365, "keep": 0.01219159438915085, "leak": 0.26484179496765137, "replacement": 0.255558118224144} +{"step": 141, "steps": 250, "lr": 2.538857773842847e-05, "grad_norm": 2.289607524871826, "loss_scale": 4.0, "retries": 0, "seconds": 130.1, "max_memory_gb": 21.29, "pair": 0.22427008231170475, "sign": 0.18507448278251104, "keep": 0.008064225301495753, "leak": 0.22066537896171212, "replacement": 0.37600577622652054} +{"step": 142, "steps": 250, "lr": 2.5088033844782426e-05, "grad_norm": 3.4348483085632324, "loss_scale": 4.0, "retries": 0, "seconds": 131.0, "max_memory_gb": 21.29, "pair": 0.17947224155068398, "sign": 0.1784147581201978, "keep": 0.256788841841626, "leak": 0.46797123178839684, "replacement": 0.5140668302774429} +{"step": 143, "steps": 250, "lr": 2.4787924694255237e-05, "grad_norm": 11.543109893798828, "loss_scale": 4.0, "retries": 0, "seconds": 132.0, "max_memory_gb": 21.29, "pair": 0.3951301199849695, "sign": 0.2704733526916243, "keep": 1.1335551887718793, "leak": 0.8768396656960249, "replacement": 1.3163542933762074} +{"step": 144, "steps": 250, "lr": 2.4488304379806627e-05, "grad_norm": 5.471215724945068, "loss_scale": 4.0, "retries": 0, "seconds": 132.9, "max_memory_gb": 21.29, "pair": 0.2096125353127718, "sign": 0.4613939218688756, "keep": 0.029037504602456465, "leak": 0.5695353837218136, "replacement": 0.597212104126811} +{"step": 145, "steps": 250, "lr": 2.4189226906286388e-05, "grad_norm": 2.392446756362915, "loss_scale": 4.0, "retries": 0, "seconds": 133.9, "max_memory_gb": 21.29, "pair": 0.15556940902024508, "sign": 0.010490910506632645, "keep": 0.011394473722134535, "leak": 0.4579453244805336, "replacement": 0.2809984269551933} +{"step": 146, "steps": 250, "lr": 2.3890746180700395e-05, "grad_norm": 6.749478816986084, "loss_scale": 4.0, "retries": 0, "seconds": 134.8, "max_memory_gb": 21.29, "pair": 0.1740203956142068, "sign": 0.03544080292340368, "keep": 0.00824734218622325, "leak": 1.3491749167442322, "replacement": 0.5331582129001617} +{"step": 147, "steps": 250, "lr": 2.359291600249407e-05, "grad_norm": 9.962739944458008, "loss_scale": 4.0, "retries": 0, "seconds": 135.7, "max_memory_gb": 21.29, "pair": 0.4735791478306055, "sign": 1.249341526068747, "keep": 1.5961178873149038, "leak": 2.3904883712530136, "replacement": 2.4939309135079384} +{"step": 148, "steps": 250, "lr": 2.3295790053855392e-05, "grad_norm": 1.8431702852249146, "loss_scale": 4.0, "retries": 0, "seconds": 136.7, "max_memory_gb": 21.29, "pair": 0.06337902139057405, "sign": 0.010231722262687981, "keep": 0.0007935687608515707, "leak": 0.7470430778339505, "replacement": 0.25565243791788816} +{"step": 149, "steps": 250, "lr": 2.299942189003901e-05, "grad_norm": 2.549145460128784, "loss_scale": 4.0, "retries": 0, "seconds": 137.6, "max_memory_gb": 21.29, "pair": 0.1551406835205853, "sign": 0.02508975547971204, "keep": 0.012102573014999507, "leak": 0.4723814921453595, "replacement": 0.29183221608400345} +{"step": 150, "steps": 250, "lr": 2.2703864929713144e-05, "grad_norm": 3.25675892829895, "loss_scale": 4.0, "retries": 0, "seconds": 138.5, "max_memory_gb": 21.29, "pair": 0.33760693296790123, "sign": 0.11779812222812325, "keep": 0.13750199344940484, "leak": 0.28823568439111114, "replacement": 0.5373159125447273} +{"step": 151, "steps": 250, "lr": 2.2409172445331257e-05, "grad_norm": 2.1494758129119873, "loss_scale": 4.0, "retries": 0, "seconds": 139.5, "max_memory_gb": 21.29, "pair": 0.15612962283194065, "sign": 0.014253141998779029, "keep": 0.01361934254236985, "leak": 0.1178849688731134, "replacement": 0.19953710865229368} +{"step": 152, "steps": 250, "lr": 2.211539755352996e-05, "grad_norm": 2.4637537002563477, "loss_scale": 4.0, "retries": 0, "seconds": 140.4, "max_memory_gb": 21.29, "pair": 0.251396007835865, "sign": 0.008412573020905256, "keep": 0.03545026187202893, "leak": 0.6109141753986478, "replacement": 0.4260559603571892} +{"step": 153, "steps": 250, "lr": 2.1822593205555016e-05, "grad_norm": 3.8043224811553955, "loss_scale": 4.0, "retries": 0, "seconds": 141.3, "max_memory_gb": 21.29, "pair": 0.12366341857705265, "sign": 0.006194989400682971, "keep": 0.012385398147330307, "leak": 0.728226326406002, "replacement": 0.31501019932329655} +{"step": 154, "steps": 250, "lr": 2.1530812177717236e-05, "grad_norm": 2.8766026496887207, "loss_scale": 4.0, "retries": 0, "seconds": 142.3, "max_memory_gb": 21.29, "pair": 0.15278594475239515, "sign": 0.007909908541478217, "keep": 0.0689982536714524, "leak": 1.1926909238100052, "replacement": 0.4894127622246742} +{"step": 155, "steps": 250, "lr": 2.1240107061879815e-05, "grad_norm": 1.4312745332717896, "loss_scale": 4.0, "retries": 0, "seconds": 143.2, "max_memory_gb": 21.29, "pair": 0.12802817579358816, "sign": 0.006450833534472622, "keep": 0.5808918603959228, "leak": 0.4196248390362598, "replacement": 0.5266057197004557} +{"step": 156, "steps": 250, "lr": 2.095053025597896e-05, "grad_norm": 18.576379776000977, "loss_scale": 4.0, "retries": 0, "seconds": 144.2, "max_memory_gb": 21.29, "pair": 0.8967418589163572, "sign": 0.01148242552881129, "keep": 0.025564488630493543, "leak": 0.4081547576934099, "replacement": 1.0173039715737104} +{"step": 157, "steps": 250, "lr": 2.066213395457945e-05, "grad_norm": 11.89239501953125, "loss_scale": 4.0, "retries": 0, "seconds": 145.1, "max_memory_gb": 21.29, "pair": 0.13234339142218232, "sign": 1.0162469865754247, "keep": 0.0038067829882493243, "leak": 2.4534836472012103, "replacement": 1.255741212517023} +{"step": 158, "steps": 250, "lr": 2.0374970139466877e-05, "grad_norm": 6.0519938468933105, "loss_scale": 4.0, "retries": 0, "seconds": 146.0, "max_memory_gb": 21.29, "pair": 0.3757762499153614, "sign": 0.36312922975048423, "keep": 0.4509962381744117, "leak": 0.3175588217563927, "replacement": 0.8622287027537823} +{"step": 159, "steps": 250, "lr": 2.0089090570278213e-05, "grad_norm": 3.297302722930908, "loss_scale": 4.0, "retries": 0, "seconds": 147.0, "max_memory_gb": 21.32, "pair": 0.17118491139262915, "sign": 0.005476348887896165, "keep": 0.005421656824182719, "leak": 0.1625253411475569, "replacement": 0.21726524829864502} +{"step": 160, "steps": 250, "lr": 1.9804546775172457e-05, "grad_norm": 6.151619911193848, "loss_scale": 4.0, "retries": 0, "seconds": 147.9, "max_memory_gb": 21.32, "pair": 0.331625584512949, "sign": 0.15939465368865058, "keep": 0.48802599764894694, "leak": 0.32241179794073105, "replacement": 0.735938873142004} +{"step": 161, "steps": 250, "lr": 1.9521390041542946e-05, "grad_norm": 4.0057692527771, "loss_scale": 4.0, "retries": 0, "seconds": 148.8, "max_memory_gb": 21.32, "pair": 0.24039359111338854, "sign": 0.09280363866128027, "keep": 0.07902011891542315, "leak": 0.47267882619053125, "replacement": 0.4444751888513565} +{"step": 162, "steps": 250, "lr": 1.923967140677313e-05, "grad_norm": 2.728369951248169, "loss_scale": 4.0, "retries": 0, "seconds": 149.9, "max_memory_gb": 21.32, "pair": 0.0848072674125433, "sign": 0.012948561750818044, "keep": 0.06357776968613393, "leak": 0.7776275926735252, "replacement": 0.3174773259088397} +{"step": 163, "steps": 250, "lr": 1.8959441649037335e-05, "grad_norm": 8.223562240600586, "loss_scale": 4.0, "retries": 0, "seconds": 150.8, "max_memory_gb": 21.32, "pair": 0.3301181085407734, "sign": 0.36244754306972027, "keep": 0.49124452332034707, "leak": 0.18248273339122534, "replacement": 0.8025848120450974} +{"step": 164, "steps": 250, "lr": 1.868075127814832e-05, "grad_norm": 3.54201340675354, "loss_scale": 4.0, "retries": 0, "seconds": 151.8, "max_memory_gb": 21.32, "pair": 0.27774448692798615, "sign": 0.2053518679458648, "keep": 0.021151382301468402, "leak": 0.49230677634477615, "replacement": 0.5140728130936623} +{"step": 165, "steps": 250, "lr": 1.8403650526453143e-05, "grad_norm": 1.767743468284607, "loss_scale": 4.0, "retries": 0, "seconds": 152.9, "max_memory_gb": 21.32, "pair": 0.10885633621364832, "sign": 0.011296363547444344, "keep": 0.001365162112051621, "leak": 0.10930376127362251, "replacement": 0.14251303859055042} +{"step": 166, "steps": 250, "lr": 1.812818933977912e-05, "grad_norm": 16.694438934326172, "loss_scale": 4.0, "retries": 0, "seconds": 153.9, "max_memory_gb": 21.32, "pair": 1.1706179403699934, "sign": 1.0668827160843648, "keep": 1.9456151249833056, "leak": 2.4608705108985305, "replacement": 3.2920844815671444} +{"step": 167, "steps": 250, "lr": 1.785441736843128e-05, "grad_norm": 5.452382564544678, "loss_scale": 4.0, "retries": 0, "seconds": 154.8, "max_memory_gb": 21.32, "pair": 0.21113116666674614, "sign": 0.026547277811914682, "keep": 0.05042954551754519, "leak": 1.9552497463300824, "replacement": 0.7384320050477982} +{"step": 168, "steps": 250, "lr": 1.7582383958243314e-05, "grad_norm": 1.6377332210540771, "loss_scale": 4.0, "retries": 0, "seconds": 155.7, "max_memory_gb": 21.32, "pair": 0.12914983555674553, "sign": 0.007614075671881437, "keep": 0.07898455511167413, "leak": 0.43226708844304085, "replacement": 0.2805159203708172} +{"step": 169, "steps": 250, "lr": 1.7312138141683164e-05, "grad_norm": 2.504934310913086, "loss_scale": 4.0, "retries": 0, "seconds": 156.7, "max_memory_gb": 21.32, "pair": 0.2639716677367687, "sign": 0.10572008648887277, "keep": 0.10423350986093283, "leak": 0.6964261569082737, "replacement": 0.5430550128221512} +{"step": 170, "steps": 250, "lr": 1.704372862901521e-05, "grad_norm": 5.352855682373047, "loss_scale": 4.0, "retries": 0, "seconds": 157.5, "max_memory_gb": 21.32, "pair": 0.15934226498939097, "sign": 0.004142646415857598, "keep": 0.01503405721814488, "leak": 1.8155216872692108, "replacement": 0.6228110454976559} +{"step": 171, "steps": 250, "lr": 1.6777203799520573e-05, "grad_norm": 3.018157482147217, "loss_scale": 4.0, "retries": 0, "seconds": 158.5, "max_memory_gb": 21.32, "pair": 0.10325582354562357, "sign": 0.06597374055854743, "keep": 0.061153599351257526, "leak": 0.4751272462308407, "replacement": 0.28560131415724754} +{"step": 172, "steps": 250, "lr": 1.6512611692777008e-05, "grad_norm": 2.4555702209472656, "loss_scale": 4.0, "retries": 0, "seconds": 159.4, "max_memory_gb": 21.32, "pair": 0.09438113495707512, "sign": 0.0025867887889035046, "keep": 0.016894774889806286, "leak": 0.639693271368742, "replacement": 0.26404523849487305} +{"step": 173, "steps": 250, "lr": 1.6250000000000005e-05, "grad_norm": 3.3472890853881836, "loss_scale": 4.0, "retries": 0, "seconds": 160.3, "max_memory_gb": 21.32, "pair": 0.11518965719733387, "sign": 0.09785116191778798, "keep": 0.09921149595174938, "leak": 0.9389026276767254, "replacement": 0.4484466426074505} +{"step": 174, "steps": 250, "lr": 1.598941605544678e-05, "grad_norm": 3.0084640979766846, "loss_scale": 4.0, "retries": 0, "seconds": 161.2, "max_memory_gb": 21.32, "pair": 0.11609150927688461, "sign": 0.01308565332510625, "keep": 0.025993352373909318, "leak": 0.40770064759999514, "replacement": 0.23755617067217827} +{"step": 175, "steps": 250, "lr": 1.5730906827884506e-05, "grad_norm": 3.927100658416748, "loss_scale": 4.0, "retries": 0, "seconds": 162.1, "max_memory_gb": 21.32, "pair": 0.27229546569287777, "sign": 0.2524048872292042, "keep": 0.2079490884207189, "leak": 1.112948140129447, "replacement": 0.7807094790041447} +{"step": 176, "steps": 250, "lr": 1.5474518912124476e-05, "grad_norm": 2.068948268890381, "loss_scale": 4.0, "retries": 0, "seconds": 163.0, "max_memory_gb": 21.32, "pair": 0.12846708856523037, "sign": 0.021384960098657757, "keep": 0.009156824831734411, "leak": 0.4593200106173754, "replacement": 0.2585679814219475} +{"step": 177, "steps": 250, "lr": 1.5220298520623646e-05, "grad_norm": 5.852788925170898, "loss_scale": 4.0, "retries": 0, "seconds": 163.9, "max_memory_gb": 21.32, "pair": 0.17234970297431573, "sign": 0.2564274172327714, "keep": 0.011996759887523467, "leak": 0.5928281545639038, "replacement": 0.45476883836090565} +{"step": 178, "steps": 250, "lr": 1.4968291475155121e-05, "grad_norm": 3.1678247451782227, "loss_scale": 4.0, "retries": 0, "seconds": 164.8, "max_memory_gb": 21.32, "pair": 0.0522017918410711, "sign": 0.026228171089314856, "keep": 0.08890096587128937, "leak": 0.813193389258231, "replacement": 0.31306470884010196} +{"step": 179, "steps": 250, "lr": 1.4718543198548998e-05, "grad_norm": 3.9732587337493896, "loss_scale": 4.0, "retries": 0, "seconds": 165.7, "max_memory_gb": 21.32, "pair": 0.5836156993173063, "sign": 0.5266030168422731, "keep": 0.6703847222950685, "leak": 0.10066408570855856, "replacement": 1.207275576889515} +{"step": 180, "steps": 250, "lr": 1.4471098706505227e-05, "grad_norm": 1.852074146270752, "loss_scale": 4.0, "retries": 0, "seconds": 166.7, "max_memory_gb": 21.32, "pair": 0.19733840972185135, "sign": 0.031080077751539648, "keep": 0.05253644287586212, "leak": 0.6167776621878147, "replacement": 0.3933410868048668} +{"step": 181, "steps": 250, "lr": 1.4226002599479743e-05, "grad_norm": 8.405489921569824, "loss_scale": 4.0, "retries": 0, "seconds": 167.6, "max_memory_gb": 21.32, "pair": 0.12571474770084023, "sign": 0.0075751151307486, "keep": 0.020488377544097602, "leak": 2.0425557792186737, "replacement": 0.6503854244947433} +{"step": 182, "steps": 250, "lr": 1.3983299054645471e-05, "grad_norm": 13.743010520935059, "loss_scale": 4.0, "retries": 0, "seconds": 168.5, "max_memory_gb": 21.32, "pair": 0.3676457500550896, "sign": 0.39733440763666295, "keep": 0.006351513315507873, "leak": 3.307092534378171, "replacement": 1.396261876448989} +{"step": 183, "steps": 250, "lr": 1.3743031817929724e-05, "grad_norm": 4.072021007537842, "loss_scale": 4.0, "retries": 0, "seconds": 169.4, "max_memory_gb": 21.32, "pair": 0.09469899954274297, "sign": 0.007849945890484378, "keep": 0.003147574950158827, "leak": 0.5108563858084381, "replacement": 0.22791185276582837} +{"step": 184, "steps": 250, "lr": 1.3505244196129164e-05, "grad_norm": 2.999692916870117, "loss_scale": 4.0, "retries": 0, "seconds": 170.3, "max_memory_gb": 21.32, "pair": 0.19725113362073898, "sign": 0.004095470154425129, "keep": 0.004062176214347346, "leak": 0.82300556730479, "replacement": 0.40708132926374674} +{"step": 185, "steps": 250, "lr": 1.3269979049104014e-05, "grad_norm": 2.7949821949005127, "loss_scale": 4.0, "retries": 0, "seconds": 171.1, "max_memory_gb": 21.32, "pair": 0.16528471000492573, "sign": 0.013749438861850649, "keep": 0.03293762645625975, "leak": 0.23304785043001175, "replacement": 0.24689020588994026} +{"step": 186, "steps": 250, "lr": 1.3037278782052863e-05, "grad_norm": 4.5505452156066895, "loss_scale": 4.0, "retries": 0, "seconds": 172.0, "max_memory_gb": 21.32, "pair": 0.15981213899794966, "sign": 0.017960071832931135, "keep": 0.38614873955066287, "leak": 0.4991806373000145, "replacement": 0.48666170984506607} +{"step": 187, "steps": 250, "lr": 1.2807185337869343e-05, "grad_norm": 2.11521315574646, "loss_scale": 4.0, "retries": 0, "seconds": 172.9, "max_memory_gb": 21.32, "pair": 0.2300195200368762, "sign": 0.008287431905046105, "keep": 0.24354825953196269, "leak": 0.6074568503536284, "replacement": 0.5078015699982643} +{"step": 188, "steps": 250, "lr": 1.2579740189582115e-05, "grad_norm": 4.557149887084961, "loss_scale": 4.0, "retries": 0, "seconds": 173.8, "max_memory_gb": 21.32, "pair": 0.4722942002117634, "sign": 0.6859247378160944, "keep": 0.7556575003687271, "leak": 0.05451649008318782, "replacement": 1.2067144471220672} +{"step": 189, "steps": 250, "lr": 1.2354984332879668e-05, "grad_norm": 11.68832015991211, "loss_scale": 4.0, "retries": 0, "seconds": 174.7, "max_memory_gb": 21.32, "pair": 0.14539742190390825, "sign": 0.6978421867825091, "keep": 1.5200484952656552, "leak": 2.360363321378827, "replacement": 1.8444335386157036} +{"step": 190, "steps": 250, "lr": 1.2132958278721064e-05, "grad_norm": 4.413784027099609, "loss_scale": 4.0, "retries": 0, "seconds": 175.6, "max_memory_gb": 21.32, "pair": 0.09038518741726875, "sign": 0.005714442580938339, "keep": 0.006623395427595824, "leak": 0.9580170512199402, "replacement": 0.3360583707690239} +{"step": 191, "steps": 250, "lr": 1.1913702046034016e-05, "grad_norm": 22.5040340423584, "loss_scale": 4.0, "retries": 0, "seconds": 176.4, "max_memory_gb": 21.32, "pair": 0.7429432105273008, "sign": 0.8408460188657045, "keep": 1.4604545952934131, "leak": 2.4330550879240036, "replacement": 2.5018573701381683} +{"step": 192, "steps": 250, "lr": 1.1697255154501756e-05, "grad_norm": 2.715571880340576, "loss_scale": 4.0, "retries": 0, "seconds": 177.4, "max_memory_gb": 21.32, "pair": 0.49962586257606745, "sign": 0.4854641635902226, "keep": 1.2054243558086455, "leak": 0.1716916561126709, "replacement": 1.387993048876524} +{"step": 193, "steps": 250, "lr": 1.1483656617439842e-05, "grad_norm": 3.802058219909668, "loss_scale": 4.0, "retries": 0, "seconds": 178.3, "max_memory_gb": 21.32, "pair": 0.1419768063351512, "sign": 0.018271449720487, "keep": 0.014580510759060417, "leak": 1.057793855725322, "replacement": 0.42285124643240124} +{"step": 194, "steps": 250, "lr": 1.1272944934764177e-05, "grad_norm": 3.137625217437744, "loss_scale": 4.0, "retries": 0, "seconds": 179.2, "max_memory_gb": 21.32, "pair": 0.26386322267353535, "sign": 0.16895117377862334, "keep": 0.021262062773303114, "leak": 0.14600970456376672, "replacement": 0.3954722583293915} +{"step": 195, "steps": 250, "lr": 1.1065158086051684e-05, "grad_norm": 4.073507785797119, "loss_scale": 4.0, "retries": 0, "seconds": 180.1, "max_memory_gb": 21.32, "pair": 0.24700016900897026, "sign": 0.04338709427975118, "keep": 0.008512290951330215, "leak": 0.10260215216840152, "replacement": 0.2986003905534744} +{"step": 196, "steps": 250, "lr": 1.086033352369467e-05, "grad_norm": 3.756103038787842, "loss_scale": 4.0, "retries": 0, "seconds": 181.1, "max_memory_gb": 21.32, "pair": 0.17517559346742928, "sign": 0.011128236685181037, "keep": 0.00938020125897765, "leak": 1.059163759011426, "replacement": 0.4502207497134805} +{"step": 197, "steps": 250, "lr": 1.0658508166150224e-05, "grad_norm": 5.165436744689941, "loss_scale": 4.0, "retries": 0, "seconds": 182.0, "max_memory_gb": 21.32, "pair": 0.22508295997977257, "sign": 0.039318624418228865, "keep": 0.003568555388483219, "leak": 0.7882690029218793, "replacement": 0.4435938075184822} +{"step": 198, "steps": 250, "lr": 1.0459718391285954e-05, "grad_norm": 2.4827709197998047, "loss_scale": 4.0, "retries": 0, "seconds": 183.0, "max_memory_gb": 21.32, "pair": 0.21986312046647072, "sign": 0.07020707265473902, "keep": 0.06456780898588477, "leak": 0.317739526508376, "replacement": 0.366685438901186} +{"step": 199, "steps": 250, "lr": 1.0264000029822999e-05, "grad_norm": 2.9833219051361084, "loss_scale": 4.0, "retries": 0, "seconds": 183.9, "max_memory_gb": 21.32, "pair": 0.16596494987607002, "sign": 0.0298693657387048, "keep": 0.4987670758564491, "leak": 0.49592401972040534, "replacement": 0.5542641598731279} +{"step": 200, "steps": 250, "lr": 1.0071388358877775e-05, "grad_norm": 5.335782051086426, "loss_scale": 4.0, "retries": 0, "seconds": 184.8, "max_memory_gb": 21.32, "pair": 0.2635818114504218, "sign": 0.33149263623636216, "keep": 1.166668640114949, "leak": 0.547941979020834, "replacement": 1.1496479660272598} +{"step": 201, "steps": 250, "lr": 9.881918095603501e-06, "grad_norm": 6.542748928070068, "loss_scale": 4.0, "retries": 0, "seconds": 185.7, "max_memory_gb": 21.32, "pair": 0.34726622700691223, "sign": 0.3297190605662763, "keep": 0.4172075928072445, "leak": 0.49375789798796177, "replacement": 0.8441690281033516} +{"step": 202, "steps": 250, "lr": 9.695623390932612e-06, "grad_norm": 20.48324966430664, "loss_scale": 4.0, "retries": 0, "seconds": 186.6, "max_memory_gb": 21.32, "pair": 0.12416189117357135, "sign": 0.03149386636505369, "keep": 0.00918594852535648, "leak": 3.817654687911272, "replacement": 1.0989154502749443} +{"step": 203, "steps": 250, "lr": 9.51253782342123e-06, "grad_norm": 1.9717985391616821, "loss_scale": 4.0, "retries": 0, "seconds": 187.5, "max_memory_gb": 21.32, "pair": 0.07538569252938032, "sign": 0.026957547408528626, "keep": 0.006760153686627746, "leak": 0.4910927973687649, "replacement": 0.21501774713397026} +{"step": 204, "steps": 250, "lr": 9.332694393196846e-06, "grad_norm": 2.209287166595459, "loss_scale": 4.0, "retries": 0, "seconds": 188.4, "max_memory_gb": 21.32, "pair": 0.22353091463446617, "sign": 0.019681483157910407, "keep": 0.03787837619893253, "leak": 1.2008126935688779, "replacement": 0.5525140315294266} +{"step": 205, "steps": 250, "lr": 9.156125516010256e-06, "grad_norm": 2.9636120796203613, "loss_scale": 4.0, "retries": 0, "seconds": 189.3, "max_memory_gb": 21.32, "pair": 0.11337307561188936, "sign": 0.005374882166506723, "keep": 0.0018280181830050424, "leak": 0.6680054538883269, "replacement": 0.2839758824557066} +{"step": 206, "steps": 250, "lr": 8.982863017392733e-06, "grad_norm": 1.855520248413086, "loss_scale": 4.0, "retries": 0, "seconds": 190.3, "max_memory_gb": 21.32, "pair": 0.11676821764558554, "sign": 0.0062163283582776785, "keep": 0.018392649013541984, "leak": 0.5086646508425474, "replacement": 0.2562388670630753} +{"step": 207, "steps": 250, "lr": 8.81293812691974e-06, "grad_norm": 1.9589732885360718, "loss_scale": 4.0, "retries": 0, "seconds": 191.2, "max_memory_gb": 21.32, "pair": 0.23104814731050283, "sign": 0.14225200994405895, "keep": 0.08284830109664654, "leak": 0.4080179068259895, "replacement": 0.44560275925323367} +{"step": 208, "steps": 250, "lr": 8.646381472581927e-06, "grad_norm": 1.5769414901733398, "loss_scale": 4.0, "retries": 0, "seconds": 192.2, "max_memory_gb": 21.32, "pair": 0.14258036331739277, "sign": 0.0042355838741059415, "keep": 0.010322206653640364, "leak": 0.43300530686974525, "replacement": 0.25811058981344104} +{"step": 209, "steps": 250, "lr": 8.483223075264622e-06, "grad_norm": 10.432658195495605, "loss_scale": 4.0, "retries": 0, "seconds": 193.1, "max_memory_gb": 21.32, "pair": 0.5882082157768309, "sign": 0.43250044976593927, "keep": 0.47705525762285106, "leak": 0.9522409364581108, "replacement": 1.2810462415218353} +{"step": 210, "steps": 250, "lr": 8.323492343336747e-06, "grad_norm": 10.182992935180664, "loss_scale": 4.0, "retries": 0, "seconds": 194.1, "max_memory_gb": 21.32, "pair": 0.4188235215842724, "sign": 0.5386610647110501, "keep": 1.8956220047538181, "leak": 2.2498811669647694, "replacement": 2.19843533821404} +{"step": 211, "steps": 250, "lr": 8.167218067350122e-06, "grad_norm": 2.4394853115081787, "loss_scale": 4.0, "retries": 0, "seconds": 195.0, "max_memory_gb": 21.32, "pair": 0.43129250034689903, "sign": 0.272361425566487, "keep": 0.133949133916758, "leak": 0.36500632483512163, "replacement": 0.725699357688427} +{"step": 212, "steps": 250, "lr": 8.01442841485013e-06, "grad_norm": 10.653082847595215, "loss_scale": 4.0, "retries": 0, "seconds": 196.0, "max_memory_gb": 21.32, "pair": 0.482527282088995, "sign": 0.20273916493169963, "keep": 0.075047014397569, "leak": 0.9661877788603306, "replacement": 0.8629672899842262} +{"step": 213, "steps": 250, "lr": 7.865150925298678e-06, "grad_norm": 5.507381916046143, "loss_scale": 4.0, "retries": 0, "seconds": 196.9, "max_memory_gb": 21.32, "pair": 0.41572294756770134, "sign": 0.35595423774793744, "keep": 0.5518497786833905, "leak": 0.5432798145338893, "replacement": 1.0054448898881674} +{"step": 214, "steps": 250, "lr": 7.719412505110375e-06, "grad_norm": 2.7156777381896973, "loss_scale": 4.0, "retries": 0, "seconds": 197.8, "max_memory_gb": 21.32, "pair": 0.388802049215883, "sign": 0.3807845929841278, "keep": 1.0936016659234156, "leak": 0.02922810078598559, "replacement": 1.1333022362086922} +{"step": 215, "steps": 250, "lr": 7.57723942280278e-06, "grad_norm": 7.185884952545166, "loss_scale": 4.0, "retries": 0, "seconds": 198.7, "max_memory_gb": 21.32, "pair": 0.5159709393337835, "sign": 0.8229004397580866, "keep": 1.9421782924046056, "leak": 0.33320347033441067, "replacement": 1.9818111453205347} +{"step": 216, "steps": 250, "lr": 7.438657304261687e-06, "grad_norm": 2.2510056495666504, "loss_scale": 4.0, "retries": 0, "seconds": 199.6, "max_memory_gb": 21.32, "pair": 0.1056160512380302, "sign": 0.025287020165706053, "keep": 0.0040042382388492115, "leak": 0.8012209619046189, "replacement": 0.3205669329036027} +{"step": 217, "steps": 250, "lr": 7.30369112812219e-06, "grad_norm": 2.2829151153564453, "loss_scale": 4.0, "retries": 0, "seconds": 200.6, "max_memory_gb": 21.32, "pair": 0.15316383307799697, "sign": 0.0035388244432397187, "keep": 0.0009875043215288315, "leak": 0.40998244285583496, "replacement": 0.2579226028174162} +{"step": 218, "steps": 250, "lr": 7.172365221266403e-06, "grad_norm": 6.429124355316162, "loss_scale": 4.0, "retries": 0, "seconds": 201.5, "max_memory_gb": 21.32, "pair": 0.4132233504205942, "sign": 0.8288010209798813, "keep": 0.13350218716368545, "leak": 0.24127042142208666, "replacement": 0.954692542552948} +{"step": 219, "steps": 250, "lr": 7.0447032544387154e-06, "grad_norm": 4.5867767333984375, "loss_scale": 4.0, "retries": 0, "seconds": 202.4, "max_memory_gb": 21.32, "pair": 0.16880623996257782, "sign": 0.06098019381170161, "keep": 0.00604592653689906, "leak": 0.10725244600325823, "replacement": 0.22913241386413574} +{"step": 220, "steps": 250, "lr": 6.920728237979274e-06, "grad_norm": 4.499491214752197, "loss_scale": 4.0, "retries": 0, "seconds": 203.4, "max_memory_gb": 21.32, "pair": 0.19474952132441103, "sign": 0.050860355724580586, "keep": 0.09520850506521583, "leak": 0.07225517882034183, "replacement": 0.28584774397313595} +{"step": 221, "steps": 250, "lr": 6.800462517676456e-06, "grad_norm": 3.1349217891693115, "loss_scale": 4.0, "retries": 0, "seconds": 204.3, "max_memory_gb": 21.32, "pair": 0.35735407657921314, "sign": 0.1543563666054979, "keep": 0.006668216898106039, "leak": 0.8467869055457413, "replacement": 0.6495631039142609} +{"step": 222, "steps": 250, "lr": 6.683927770739222e-06, "grad_norm": 2.7504982948303223, "loss_scale": 4.0, "retries": 0, "seconds": 205.2, "max_memory_gb": 21.32, "pair": 0.15710936789400876, "sign": 0.009255517012206838, "keep": 0.03696003974562245, "leak": 0.2575694462284446, "replacement": 0.24460951052606106} +{"step": 223, "steps": 250, "lr": 6.571145001889892e-06, "grad_norm": 31.756738662719727, "loss_scale": 4.0, "retries": 0, "seconds": 206.2, "max_memory_gb": 21.32, "pair": 0.778423429466784, "sign": 0.17317140373052098, "keep": 0.21852532843240624, "leak": 2.227616449818015, "replacement": 1.5311759263277054} +{"step": 224, "steps": 250, "lr": 6.46213453957817e-06, "grad_norm": 4.361721515655518, "loss_scale": 4.0, "retries": 0, "seconds": 207.1, "max_memory_gb": 21.32, "pair": 0.18518462451174855, "sign": 0.029626381437992677, "keep": 0.011808208655054386, "leak": 0.652838945388794, "replacement": 0.36911166133359075} +{"step": 225, "steps": 250, "lr": 6.356916032317063e-06, "grad_norm": 6.133421897888184, "loss_scale": 4.0, "retries": 0, "seconds": 208.1, "max_memory_gb": 21.32, "pair": 0.09472206933423877, "sign": 0.011520811822265387, "keep": 0.002369108064158354, "leak": 1.5792385707609355, "replacement": 0.4964766576886177} +{"step": 226, "steps": 250, "lr": 6.255508445141349e-06, "grad_norm": 4.8083109855651855, "loss_scale": 4.0, "retries": 0, "seconds": 209.0, "max_memory_gb": 21.32, "pair": 0.12834482197649777, "sign": 0.027293091145111248, "keep": 0.6260208918247372, "leak": 0.49212443828582764, "replacement": 0.5780329219996929} +{"step": 227, "steps": 250, "lr": 6.157930056189228e-06, "grad_norm": 3.713937997817993, "loss_scale": 4.0, "retries": 0, "seconds": 209.9, "max_memory_gb": 21.32, "pair": 0.21826887493079994, "sign": 0.01335003605709062, "keep": 0.017290016912966166, "leak": 0.1390921100974083, "replacement": 0.2683619186282158} +{"step": 228, "steps": 250, "lr": 6.064198453407796e-06, "grad_norm": 2.688612699508667, "loss_scale": 4.0, "retries": 0, "seconds": 210.8, "max_memory_gb": 21.32, "pair": 0.1522760996595025, "sign": 0.0539395670639351, "keep": 0.0017952603084268048, "leak": 0.6403499017615104, "replacement": 0.34023099299520254} +{"step": 229, "steps": 250, "lr": 5.974330531382935e-06, "grad_norm": 11.190886497497559, "loss_scale": 4.0, "retries": 0, "seconds": 211.8, "max_memory_gb": 21.32, "pair": 0.15719305258244276, "sign": 0.07610364723950624, "keep": 0.048091230913996696, "leak": 1.2268310813233256, "replacement": 0.5259982794523239} +{"step": 230, "steps": 250, "lr": 5.888342488294125e-06, "grad_norm": 24.45764923095703, "loss_scale": 4.0, "retries": 0, "seconds": 212.7, "max_memory_gb": 21.32, "pair": 0.8712197141721845, "sign": 0.7262305775657296, "keep": 1.4517254848032195, "leak": 1.7438848223537207, "replacement": 2.3961689099669456} +{"step": 231, "steps": 250, "lr": 5.806249822994849e-06, "grad_norm": 3.226576328277588, "loss_scale": 4.0, "retries": 0, "seconds": 213.5, "max_memory_gb": 21.32, "pair": 0.2571797298733145, "sign": 0.5618720060447231, "keep": 0.021741083057122523, "leak": 0.28001561760902405, "replacement": 0.6189901903271675} +{"step": 232, "steps": 250, "lr": 5.728067332218994e-06, "grad_norm": 4.1982526779174805, "loss_scale": 4.0, "retries": 0, "seconds": 214.5, "max_memory_gb": 21.32, "pair": 0.14272519201040268, "sign": 0.019100584322586656, "keep": 0.033726610912708566, "leak": 1.3255463689565659, "replacement": 0.5005254000425339} +{"step": 233, "steps": 250, "lr": 5.65380910791383e-06, "grad_norm": 5.0330047607421875, "loss_scale": 4.0, "retries": 0, "seconds": 215.4, "max_memory_gb": 21.32, "pair": 0.11107745103072375, "sign": 0.004955606593284756, "keep": 0.0035877074842574075, "leak": 1.2734520584344864, "replacement": 0.4337121192365885} +{"step": 234, "steps": 250, "lr": 5.583488534700022e-06, "grad_norm": 2.6937673091888428, "loss_scale": 4.0, "retries": 0, "seconds": 216.3, "max_memory_gb": 21.32, "pair": 0.07513790688244626, "sign": 0.01458758520311676, "keep": 0.009100433598518975, "leak": 0.4118905123323202, "replacement": 0.18995454302057624} +{"step": 235, "steps": 250, "lr": 5.517118287459118e-06, "grad_norm": 16.583768844604492, "loss_scale": 4.0, "retries": 0, "seconds": 217.3, "max_memory_gb": 21.32, "pair": 0.9182407143525779, "sign": 0.625870021933224, "keep": 1.201670241458487, "leak": 1.4915385080967098, "replacement": 2.2048954255878925} +{"step": 236, "steps": 250, "lr": 5.454710329048978e-06, "grad_norm": 2.568700075149536, "loss_scale": 4.0, "retries": 0, "seconds": 218.2, "max_memory_gb": 21.32, "pair": 0.1694607288518455, "sign": 0.047375048969115596, "keep": 0.003856040637856495, "leak": 0.1285431870855973, "replacement": 0.22721207700669765} +{"step": 237, "steps": 250, "lr": 5.3962759081475436e-06, "grad_norm": 4.783168792724609, "loss_scale": 4.0, "retries": 0, "seconds": 219.0, "max_memory_gb": 21.32, "pair": 0.23048159293830395, "sign": 0.2692968505434692, "keep": 0.06917647994120557, "leak": 0.5243764687329531, "replacement": 0.5308123603463173} +{"step": 238, "steps": 250, "lr": 5.341825557225321e-06, "grad_norm": 2.5100314617156982, "loss_scale": 4.0, "retries": 0, "seconds": 219.9, "max_memory_gb": 21.32, "pair": 0.19781668204814196, "sign": 0.01556603517383337, "keep": 0.022498401114717126, "leak": 0.1369756255298853, "replacement": 0.25109281204640865} +{"step": 239, "steps": 250, "lr": 5.291369090646961e-06, "grad_norm": 4.038076877593994, "loss_scale": 4.0, "retries": 0, "seconds": 220.9, "max_memory_gb": 21.32, "pair": 0.4058729335665703, "sign": 0.9752422329038382, "keep": 0.10447753174202035, "leak": 0.5853961436077952, "replacement": 1.0920818448066711} +{"step": 240, "steps": 250, "lr": 5.244915602902289e-06, "grad_norm": 2.2597177028656006, "loss_scale": 4.0, "retries": 0, "seconds": 221.8, "max_memory_gb": 21.32, "pair": 0.23863595305010676, "sign": 0.3029134754324332, "keep": 0.17845756975475524, "leak": 0.0888934675604105, "replacement": 0.5015448434278369} +{"step": 241, "steps": 250, "lr": 5.20247346696706e-06, "grad_norm": 10.864974975585938, "loss_scale": 4.0, "retries": 0, "seconds": 222.7, "max_memory_gb": 21.32, "pair": 0.5567413583048619, "sign": 0.4516750216243963, "keep": 0.3820561410351502, "leak": 0.37462690519168973, "replacement": 1.0672635925002396} +{"step": 242, "steps": 250, "lr": 5.164050332793787e-06, "grad_norm": 2.321166515350342, "loss_scale": 4.0, "retries": 0, "seconds": 223.6, "max_memory_gb": 21.32, "pair": 0.12472782051190734, "sign": 0.02865194558398798, "keep": 0.0027982106328181544, "leak": 0.7985037490725517, "replacement": 0.3400788358412683} +{"step": 243, "steps": 250, "lr": 5.129653125932873e-06, "grad_norm": 1.4436874389648438, "loss_scale": 4.0, "retries": 0, "seconds": 224.6, "max_memory_gb": 21.32, "pair": 0.17301937472075224, "sign": 0.017217043219716288, "keep": 0.02361923799617216, "leak": 0.3494969718158245, "replacement": 0.28081177081912756} +{"step": 244, "steps": 250, "lr": 5.099288046284329e-06, "grad_norm": 2.4920644760131836, "loss_scale": 4.0, "retries": 0, "seconds": 225.5, "max_memory_gb": 21.32, "pair": 0.1967502236366272, "sign": 0.09471726790070534, "keep": 0.023061216110363603, "leak": 0.7186107635498047, "replacement": 0.43529216200113297} +{"step": 245, "steps": 250, "lr": 5.072960566980277e-06, "grad_norm": 6.963828086853027, "loss_scale": 4.0, "retries": 0, "seconds": 226.4, "max_memory_gb": 21.32, "pair": 0.529887530952692, "sign": 0.3964067744091153, "keep": 0.8047807635739446, "leak": 0.7168760718777776, "replacement": 1.309700295329094} +{"step": 246, "steps": 250, "lr": 5.050675433398435e-06, "grad_norm": 1.938320279121399, "loss_scale": 4.0, "retries": 0, "seconds": 227.4, "max_memory_gb": 21.32, "pair": 0.11125925672240555, "sign": 0.004272137186489999, "keep": 0.0037671614627470262, "leak": 0.23258611792698503, "replacement": 0.17342544067651033} +{"step": 247, "steps": 250, "lr": 5.032436662306811e-06, "grad_norm": 4.600336074829102, "loss_scale": 4.0, "retries": 0, "seconds": 228.3, "max_memory_gb": 21.32, "pair": 0.0680644428357482, "sign": 0.00888859662154573, "keep": 0.01069813646608786, "leak": 0.8857368305325508, "replacement": 0.299292009556666} +{"step": 248, "steps": 250, "lr": 5.018247541139684e-06, "grad_norm": 2.8213324546813965, "loss_scale": 4.0, "retries": 0, "seconds": 229.2, "max_memory_gb": 21.32, "pair": 0.1780155450105667, "sign": 0.034082873491570354, "keep": 0.04790341295301914, "leak": 0.2549578527978156, "replacement": 0.28274815157055855} +{"step": 249, "steps": 250, "lr": 5.00811062740508e-06, "grad_norm": 2.4956119060516357, "loss_scale": 4.0, "retries": 0, "seconds": 230.2, "max_memory_gb": 21.32, "pair": 0.2495938055217266, "sign": 0.07715893583372235, "keep": 0.025854389066807926, "leak": 0.5808685459196568, "replacement": 0.4463176131248474} +{"step": 250, "steps": 250, "lr": 5.002027748223778e-06, "grad_norm": 1.8087491989135742, "loss_scale": 4.0, "retries": 0, "seconds": 231.1, "max_memory_gb": 21.32, "pair": 0.12179895862936974, "sign": 0.011752276914194226, "keep": 0.013454332685796544, "leak": 0.41344651603139937, "replacement": 0.23776388727128506} diff --git a/runs/single_internvl/replacement/L29_s13/adapter/README.md b/runs/single_internvl/replacement/L29_s13/adapter/README.md new file mode 100644 index 0000000000000000000000000000000000000000..4f9f856cf0e1531b47e25d266833b2977a19d3ac --- /dev/null +++ b/runs/single_internvl/replacement/L29_s13/adapter/README.md @@ -0,0 +1,207 @@ +--- +base_model: /data/OPSD/assets/models/InternVL3_5-8B-HF +library_name: peft +pipeline_tag: text-generation +tags: +- base_model:adapter:/data/OPSD/assets/models/InternVL3_5-8B-HF +- lora +- transformers +--- + +# Model Card for Model ID + + + + + +## Model Details + +### Model Description + + + + + +- **Developed by:** [More Information Needed] +- **Funded by [optional]:** [More Information Needed] +- **Shared by [optional]:** [More Information Needed] +- **Model type:** [More Information Needed] +- **Language(s) (NLP):** [More Information Needed] +- **License:** [More Information Needed] +- **Finetuned from model [optional]:** [More Information Needed] + +### Model Sources [optional] + + + +- **Repository:** [More Information Needed] +- **Paper [optional]:** [More Information Needed] +- **Demo [optional]:** [More Information Needed] + +## Uses + + + +### Direct Use + + + +[More Information Needed] + +### Downstream Use [optional] + + + +[More Information Needed] + +### Out-of-Scope Use + + + +[More Information Needed] + +## Bias, Risks, and Limitations + + + +[More Information Needed] + +### Recommendations + + + +Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. More information needed for further recommendations. + +## How to Get Started with the Model + +Use the code below to get started with the model. + +[More Information Needed] + +## Training Details + +### Training Data + + + +[More Information Needed] + +### Training Procedure + + + +#### Preprocessing [optional] + +[More Information Needed] + + +#### Training Hyperparameters + +- **Training regime:** [More Information Needed] + +#### Speeds, Sizes, Times [optional] + + + +[More Information Needed] + +## Evaluation + + + +### Testing Data, Factors & Metrics + +#### Testing Data + + + +[More Information Needed] + +#### Factors + + + +[More Information Needed] + +#### Metrics + + + +[More Information Needed] + +### Results + +[More Information Needed] + +#### Summary + + + +## Model Examination [optional] + + + +[More Information Needed] + +## Environmental Impact + + + +Carbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700). + +- **Hardware Type:** [More Information Needed] +- **Hours used:** [More Information Needed] +- **Cloud Provider:** [More Information Needed] +- **Compute Region:** [More Information Needed] +- **Carbon Emitted:** [More Information Needed] + +## Technical Specifications [optional] + +### Model Architecture and Objective + +[More Information Needed] + +### Compute Infrastructure + +[More Information Needed] + +#### Hardware + +[More Information Needed] + +#### Software + +[More Information Needed] + +## Citation [optional] + + + +**BibTeX:** + +[More Information Needed] + +**APA:** + +[More Information Needed] + +## Glossary [optional] + + + +[More Information Needed] + +## More Information [optional] + +[More Information Needed] + +## Model Card Authors [optional] + +[More Information Needed] + +## Model Card Contact + +[More Information Needed] +### Framework versions + +- PEFT 0.20.0 \ No newline at end of file diff --git a/runs/single_internvl/replacement/L29_s13/adapter/adapter_config.json b/runs/single_internvl/replacement/L29_s13/adapter/adapter_config.json new file mode 100644 index 0000000000000000000000000000000000000000..ce0562c2c104d6c14dceb1b44732b50f6af6515e --- /dev/null +++ b/runs/single_internvl/replacement/L29_s13/adapter/adapter_config.json @@ -0,0 +1,50 @@ +{ + "alora_invocation_tokens": null, + "alpha_pattern": {}, + "arrow_config": null, + "auto_mapping": null, + "base_model_name_or_path": "/data/OPSD/assets/models/InternVL3_5-8B-HF", + "bias": "none", + "corda_config": null, + "ensure_weight_tying": false, + "eva_config": null, + "exclude_modules": null, + "fan_in_fan_out": false, + "inference_mode": true, + "init_lora_weights": true, + "layer_replication": null, + "layers_pattern": null, + "layers_to_transform": null, + "loftq_config": {}, + "lora_alpha": 32, + "lora_bias": false, + "lora_dropout": 0.05, + "lora_ga_config": null, + "megatron_config": null, + "megatron_core": "megatron.core", + "modules_to_save": null, + "monteclora_config": null, + "peft_type": "LORA", + "peft_version": "0.20.0", + "qalora_group_size": 16, + "r": 16, + "rank_pattern": {}, + "revision": null, + "target_modules": [ + "model.language_model.layers.29.self_attn.k_proj", + "model.language_model.layers.29.self_attn.q_proj", + "model.language_model.layers.29.mlp.gate_proj", + "model.language_model.layers.29.mlp.up_proj", + "model.language_model.layers.29.self_attn.v_proj", + "model.language_model.layers.29.mlp.down_proj", + "model.language_model.layers.29.self_attn.o_proj" + ], + "target_parameters": null, + "task_type": "CAUSAL_LM", + "trainable_token_indices": null, + "use_bdlora": null, + "use_dora": false, + "use_qalora": false, + "use_rslora": false, + "velora_config": null +} \ No newline at end of file diff --git a/runs/single_internvl/replacement/L29_s13/config.json b/runs/single_internvl/replacement/L29_s13/config.json new file mode 100644 index 0000000000000000000000000000000000000000..fadd267c80b50cc738e516442f4cf73aac912d0d --- /dev/null +++ b/runs/single_internvl/replacement/L29_s13/config.json @@ -0,0 +1,24 @@ +{ + "behavior": "replacement", + "seed": 13, + "model": "/data/OPSD/assets/models/InternVL3_5-8B-HF", + "backbone": "internvl", + "train": "/data/OPSD/assets/v3/replacement.jsonl", + "items": 1000, + "steps": 250, + "epochs": 1, + "steps_per_epoch": 250, + "groups_per_step": 4, + "lr": 5e-05, + "scheduler": "cosine", + "warmup_steps": 16, + "final_fraction": 0.1, + "weight_decay": 0.0, + "clip": 1.0, + "initial_loss_scale": 4.0, + "adapter": { + "layer": 29, + "rank": 16, + "trainable_parameters": 1212416 + } +} \ No newline at end of file diff --git a/runs/single_internvl/replacement/L29_s13/steps.jsonl b/runs/single_internvl/replacement/L29_s13/steps.jsonl new file mode 100644 index 0000000000000000000000000000000000000000..6f6046db190d288f831e56c12e68c26a15166507 --- /dev/null +++ b/runs/single_internvl/replacement/L29_s13/steps.jsonl @@ -0,0 +1,250 @@ +{"step": 1, "steps": 250, "lr": 3.125e-06, "grad_norm": 0.4930037558078766, "loss_scale": 4.0, "retries": 0, "seconds": 1.0, "max_memory_gb": 18.12, "pair": 1.0986123085021973, "sign": 0.7328193015418947, "keep": 1.137254401575774, "leak": 0.0, "replacement": 2.0336491465568542} +{"step": 2, "steps": 250, "lr": 6.25e-06, "grad_norm": 0.7707909345626831, "loss_scale": 4.0, "retries": 0, "seconds": 1.8, "max_memory_gb": 18.15, "pair": 1.0959916412830353, "sign": 0.18021599692292511, "keep": 1.3681343656644458, "leak": 7.469653979796931e-07, "replacement": 1.8701669573783875} +{"step": 3, "steps": 250, "lr": 9.375000000000001e-06, "grad_norm": 1.151118516921997, "loss_scale": 4.0, "retries": 0, "seconds": 2.6, "max_memory_gb": 18.15, "pair": 1.0965837836265564, "sign": 0.11546592926606536, "keep": 2.0607327669858932, "leak": 2.482655281710322e-05, "replacement": 2.1846893429756165} +{"step": 4, "steps": 250, "lr": 1.25e-05, "grad_norm": 0.8758711218833923, "loss_scale": 4.0, "retries": 0, "seconds": 3.5, "max_memory_gb": 18.15, "pair": 1.0938598811626434, "sign": 0.03803216549567878, "keep": 0.872722974861972, "leak": 2.0602021606919152e-05, "replacement": 1.5492425858974457} +{"step": 5, "steps": 250, "lr": 1.5625e-05, "grad_norm": 0.6597382426261902, "loss_scale": 4.0, "retries": 0, "seconds": 4.3, "max_memory_gb": 18.15, "pair": 1.0986437797546387, "sign": 0.11657081497833133, "keep": 0.023346168600255623, "leak": 1.4373494909847295e-05, "replacement": 1.1686058640480042} +{"step": 6, "steps": 250, "lr": 1.8750000000000002e-05, "grad_norm": 0.6996797919273376, "loss_scale": 4.0, "retries": 0, "seconds": 5.2, "max_memory_gb": 18.15, "pair": 1.0963052213191986, "sign": 0.1990567659959197, "keep": 0.637221306678839, "leak": 2.470693461020801e-05, "replacement": 1.5144504308700562} +{"step": 7, "steps": 250, "lr": 2.1875e-05, "grad_norm": 0.9832887649536133, "loss_scale": 4.0, "retries": 0, "seconds": 6.0, "max_memory_gb": 18.15, "pair": 1.0920370817184448, "sign": 0.08932366548106074, "keep": 1.5364361057581846, "leak": 1.586616610893543e-05, "replacement": 1.9049209952354431} +{"step": 8, "steps": 250, "lr": 2.5e-05, "grad_norm": 1.0192021131515503, "loss_scale": 4.0, "retries": 0, "seconds": 6.9, "max_memory_gb": 18.16, "pair": 1.0983522236347198, "sign": 0.21032401244156063, "keep": 1.4832641407847404, "leak": 4.535660173132783e-05, "replacement": 1.9451575875282288} +{"step": 9, "steps": 250, "lr": 2.8125000000000003e-05, "grad_norm": 0.9965993165969849, "loss_scale": 4.0, "retries": 0, "seconds": 7.7, "max_memory_gb": 18.16, "pair": 1.0928905308246613, "sign": 0.5712067366112024, "keep": 1.4101951271295547, "leak": 4.554897395792068e-05, "replacement": 2.0836029052734375} +{"step": 10, "steps": 250, "lr": 3.125e-05, "grad_norm": 0.8277331590652466, "loss_scale": 4.0, "retries": 0, "seconds": 8.5, "max_memory_gb": 18.16, "pair": 1.0904849469661713, "sign": 0.08039979264140129, "keep": 0.10090547040454112, "leak": 0.00024281530204461887, "replacement": 1.181198239326477} +{"step": 11, "steps": 250, "lr": 3.4375e-05, "grad_norm": 1.1336166858673096, "loss_scale": 4.0, "retries": 0, "seconds": 9.4, "max_memory_gb": 18.16, "pair": 1.0838884115219116, "sign": 0.20542556000873446, "keep": 1.3571473883057479, "leak": 0.00037627917299687397, "replacement": 1.8652689754962921} +{"step": 12, "steps": 250, "lr": 3.7500000000000003e-05, "grad_norm": 0.9426317811012268, "loss_scale": 4.0, "retries": 0, "seconds": 10.2, "max_memory_gb": 18.16, "pair": 1.072880119085312, "sign": 0.07503022323362529, "keep": 0.9645189079310512, "leak": 0.0006048024333722424, "replacement": 1.5928059220314026} +{"step": 13, "steps": 250, "lr": 4.0625000000000005e-05, "grad_norm": 1.1738587617874146, "loss_scale": 4.0, "retries": 0, "seconds": 11.0, "max_memory_gb": 18.16, "pair": 1.0635396540164948, "sign": 0.05721096508204937, "keep": 0.8778005866261083, "leak": 0.0010877331988012884, "replacement": 1.5313173234462738} +{"step": 14, "steps": 250, "lr": 4.375e-05, "grad_norm": 1.3754116296768188, "loss_scale": 4.0, "retries": 0, "seconds": 11.8, "max_memory_gb": 18.16, "pair": 1.0746791362762451, "sign": 0.26487979479134083, "keep": 1.694936215877533, "leak": 0.0007230912815430202, "replacement": 2.0547679662704468} +{"step": 15, "steps": 250, "lr": 4.6875e-05, "grad_norm": 1.2029951810836792, "loss_scale": 4.0, "retries": 0, "seconds": 12.7, "max_memory_gb": 18.16, "pair": 1.0536890625953674, "sign": 0.19519441644661129, "keep": 1.0953594267302833, "leak": 0.00359810798545368, "replacement": 1.699865460395813} +{"step": 16, "steps": 250, "lr": 5e-05, "grad_norm": 1.066001534461975, "loss_scale": 4.0, "retries": 0, "seconds": 13.5, "max_memory_gb": 18.16, "pair": 1.0453796982765198, "sign": 0.30525135807693005, "keep": 0.3944958377396688, "leak": 0.003458933235378936, "replacement": 1.396118015050888} +{"step": 17, "steps": 250, "lr": 5e-05, "grad_norm": 1.0700042247772217, "loss_scale": 4.0, "retries": 0, "seconds": 14.3, "max_memory_gb": 18.16, "pair": 1.0474337339401245, "sign": 0.7375287204049528, "keep": 0.8714976775227115, "leak": 0.0032363788050133735, "replacement": 1.8527560532093048} +{"step": 18, "steps": 250, "lr": 4.999797225177623e-05, "grad_norm": 1.4496839046478271, "loss_scale": 4.0, "retries": 0, "seconds": 15.1, "max_memory_gb": 18.17, "pair": 1.0242405384778976, "sign": 0.18712720950134099, "keep": 1.1656320366309956, "leak": 0.005469026131322607, "replacement": 1.7019874453544617} +{"step": 19, "steps": 250, "lr": 4.999188937259492e-05, "grad_norm": 2.2893478870391846, "loss_scale": 4.0, "retries": 0, "seconds": 16.0, "max_memory_gb": 18.17, "pair": 1.008501410484314, "sign": 0.03603973425924778, "keep": 1.5715091228485107, "leak": 0.008587300952058285, "replacement": 1.8144226968288422} +{"step": 20, "steps": 250, "lr": 4.998175245886032e-05, "grad_norm": 1.6101511716842651, "loss_scale": 4.0, "retries": 0, "seconds": 16.8, "max_memory_gb": 18.17, "pair": 0.9644098877906799, "sign": 0.03039924195036292, "keep": 0.003631543588198838, "leak": 0.022875536931678653, "replacement": 0.9871442019939423} +{"step": 21, "steps": 250, "lr": 4.996756333769319e-05, "grad_norm": 1.6942180395126343, "loss_scale": 4.0, "retries": 0, "seconds": 17.6, "max_memory_gb": 18.17, "pair": 0.9640327244997025, "sign": 0.04496057296637446, "keep": 0.7178428612423886, "leak": 0.02157641015946865, "replacement": 1.3508285582065582} +{"step": 22, "steps": 250, "lr": 4.994932456660157e-05, "grad_norm": 1.6478705406188965, "loss_scale": 4.0, "retries": 0, "seconds": 18.5, "max_memory_gb": 18.17, "pair": 0.9688577950000763, "sign": 0.06512382021173835, "keep": 0.8884832771145739, "leak": 0.013835852965712547, "replacement": 1.4491203725337982} +{"step": 23, "steps": 250, "lr": 4.992703943301973e-05, "grad_norm": 2.007434129714966, "loss_scale": 4.0, "retries": 0, "seconds": 19.3, "max_memory_gb": 18.17, "pair": 0.9156039208173752, "sign": 0.03402174776419997, "keep": 1.373101055085499, "leak": 0.03272994188591838, "replacement": 1.6273478865623474} +{"step": 24, "steps": 250, "lr": 4.9900711953715675e-05, "grad_norm": 2.3840160369873047, "loss_scale": 4.0, "retries": 0, "seconds": 20.1, "max_memory_gb": 18.17, "pair": 0.9321347773075104, "sign": 0.016332178143784404, "keep": 0.7516282498836517, "leak": 0.01781778479926288, "replacement": 1.320569485425949} +{"step": 25, "steps": 250, "lr": 4.987034687406713e-05, "grad_norm": 2.1182615756988525, "loss_scale": 4.0, "retries": 0, "seconds": 20.9, "max_memory_gb": 18.17, "pair": 0.926942989230156, "sign": 0.10995381139218807, "keep": 1.5151537842757534, "leak": 0.0354132775682956, "replacement": 1.7483501434326172} +{"step": 26, "steps": 250, "lr": 4.9835949667206215e-05, "grad_norm": 2.5032010078430176, "loss_scale": 4.0, "retries": 0, "seconds": 21.8, "max_memory_gb": 18.17, "pair": 0.903999537229538, "sign": 0.06275922828353941, "keep": 0.8686660528182983, "leak": 0.030999753158539534, "replacement": 1.3774621188640594} +{"step": 27, "steps": 250, "lr": 4.9797526533032945e-05, "grad_norm": 2.160881280899048, "loss_scale": 4.0, "retries": 0, "seconds": 22.6, "max_memory_gb": 18.17, "pair": 0.878705307841301, "sign": 0.1412034237291664, "keep": 1.0408265597361606, "leak": 0.049566023342777044, "replacement": 1.482111781835556} +{"step": 28, "steps": 250, "lr": 4.9755084397097716e-05, "grad_norm": 1.7633713483810425, "loss_scale": 4.0, "retries": 0, "seconds": 23.5, "max_memory_gb": 18.17, "pair": 0.9436296373605728, "sign": 0.415017859544605, "keep": 1.534679263830185, "leak": 0.03299779538065195, "replacement": 1.9267276525497437} +{"step": 29, "steps": 250, "lr": 4.970863090935304e-05, "grad_norm": 1.8775571584701538, "loss_scale": 4.0, "retries": 0, "seconds": 24.3, "max_memory_gb": 18.17, "pair": 0.9455466121435165, "sign": 0.7443315489217639, "keep": 0.8236969448626041, "leak": 0.05259755754377693, "replacement": 1.7427102327346802} +{"step": 30, "steps": 250, "lr": 4.965817444277468e-05, "grad_norm": 2.534353733062744, "loss_scale": 4.0, "retries": 0, "seconds": 25.1, "max_memory_gb": 18.17, "pair": 0.7051541060209274, "sign": 0.010026183910667896, "keep": 0.8290428912296193, "leak": 0.12734435498714447, "replacement": 1.1565247178077698} +{"step": 31, "steps": 250, "lr": 4.9603724091852456e-05, "grad_norm": 2.4484686851501465, "loss_scale": 4.0, "retries": 0, "seconds": 26.0, "max_memory_gb": 18.17, "pair": 0.8741747885942459, "sign": 0.2153393877670169, "keep": 1.6191777288913727, "leak": 0.07626751973293722, "replacement": 1.8105002343654633} +{"step": 32, "steps": 250, "lr": 4.9545289670951025e-05, "grad_norm": 3.3283894062042236, "loss_scale": 4.0, "retries": 0, "seconds": 26.9, "max_memory_gb": 18.17, "pair": 0.7589806765317917, "sign": 0.11313948582392186, "keep": 1.299386277794838, "leak": 0.20222514681518078, "replacement": 1.5157998204231262} +{"step": 33, "steps": 250, "lr": 4.948288171254089e-05, "grad_norm": 1.9011837244033813, "loss_scale": 4.0, "retries": 0, "seconds": 27.7, "max_memory_gb": 18.17, "pair": 0.8756669014692307, "sign": 0.20417362172156572, "keep": 0.6323620583862066, "leak": 0.04275703741586767, "replacement": 1.304624006152153} +{"step": 34, "steps": 250, "lr": 4.941651146529998e-05, "grad_norm": 2.3327085971832275, "loss_scale": 4.0, "retries": 0, "seconds": 28.5, "max_memory_gb": 18.17, "pair": 0.6711076349020004, "sign": 0.02759770560078323, "keep": 0.4339720866555581, "leak": 0.2113686092197895, "replacement": 0.9547346532344818} +{"step": 35, "steps": 250, "lr": 4.9346190892086174e-05, "grad_norm": 2.283285617828369, "loss_scale": 4.0, "retries": 0, "seconds": 29.4, "max_memory_gb": 18.17, "pair": 0.682028517127037, "sign": 0.05970928492024541, "keep": 0.6747471261769533, "leak": 0.22460228949785233, "replacement": 1.1054073572158813} +{"step": 36, "steps": 250, "lr": 4.927193266778101e-05, "grad_norm": 2.32704496383667, "loss_scale": 4.0, "retries": 0, "seconds": 30.2, "max_memory_gb": 18.17, "pair": 0.693606436252594, "sign": 0.04025389428716153, "keep": 0.6375983865582384, "leak": 0.15892688184976578, "replacement": 1.0722642689943314} +{"step": 37, "steps": 250, "lr": 4.919375017700515e-05, "grad_norm": 1.9360190629959106, "loss_scale": 4.0, "retries": 0, "seconds": 31.0, "max_memory_gb": 18.17, "pair": 0.7473393380641937, "sign": 0.20139021542854607, "keep": 0.970503475778969, "leak": 0.18029350374126807, "replacement": 1.378359541296959} +{"step": 38, "steps": 250, "lr": 4.911165751170588e-05, "grad_norm": 2.7681710720062256, "loss_scale": 4.0, "retries": 0, "seconds": 31.9, "max_memory_gb": 18.17, "pair": 0.6226822137832642, "sign": 0.07790456945076585, "keep": 0.49249453097581863, "leak": 0.18636494502425194, "replacement": 0.9544730335474014} +{"step": 39, "steps": 250, "lr": 4.902566946861708e-05, "grad_norm": 1.942903757095337, "loss_scale": 4.0, "retries": 0, "seconds": 32.7, "max_memory_gb": 18.17, "pair": 0.6523708328604698, "sign": 1.4167597215855494, "keep": 0.33942980306892423, "leak": 0.34693725407123566, "replacement": 1.6171999275684357} +{"step": 40, "steps": 250, "lr": 4.89358015465922e-05, "grad_norm": 1.850042700767517, "loss_scale": 4.0, "retries": 0, "seconds": 33.5, "max_memory_gb": 18.17, "pair": 0.43511634320020676, "sign": 0.015210303943604231, "keep": 0.27058940850838553, "leak": 0.6098707802593708, "replacement": 0.7304838970303535} +{"step": 41, "steps": 250, "lr": 4.884206994381078e-05, "grad_norm": 1.6996153593063354, "loss_scale": 4.0, "retries": 0, "seconds": 34.3, "max_memory_gb": 18.17, "pair": 0.8159864097833633, "sign": 0.8635531757026911, "keep": 1.0612370781600475, "leak": 0.13755678292363882, "replacement": 1.812770739197731} +{"step": 42, "steps": 250, "lr": 4.874449155485866e-05, "grad_norm": 2.0740323066711426, "loss_scale": 4.0, "retries": 0, "seconds": 35.1, "max_memory_gb": 18.17, "pair": 0.48433806002140045, "sign": 0.04853421729058027, "keep": 0.3783297354275419, "leak": 0.5252768471837044, "replacement": 0.8290892541408539} +{"step": 43, "steps": 250, "lr": 4.864308396768294e-05, "grad_norm": 2.8877618312835693, "loss_scale": 4.0, "retries": 0, "seconds": 35.9, "max_memory_gb": 18.17, "pair": 0.497408002614975, "sign": 0.1353878416121006, "keep": 0.6223710712656612, "leak": 0.9768253192305565, "replacement": 1.1204937845468521} +{"step": 44, "steps": 250, "lr": 4.853786546042184e-05, "grad_norm": 2.006403923034668, "loss_scale": 4.0, "retries": 0, "seconds": 36.8, "max_memory_gb": 18.17, "pair": 0.3683795630931854, "sign": 0.16850176866864786, "keep": 1.030711434825207, "leak": 0.8778174892067909, "replacement": 1.1874406039714813} +{"step": 45, "steps": 250, "lr": 4.842885499811012e-05, "grad_norm": 2.64003324508667, "loss_scale": 4.0, "retries": 0, "seconds": 37.6, "max_memory_gb": 18.17, "pair": 0.6683556064963341, "sign": 0.6449440170545131, "keep": 2.1421505289617926, "leak": 0.28727335756411776, "replacement": 2.1337211430072784} +{"step": 46, "steps": 250, "lr": 4.831607222926078e-05, "grad_norm": 1.8468530178070068, "loss_scale": 4.0, "retries": 0, "seconds": 38.4, "max_memory_gb": 18.17, "pair": 0.5372498109936714, "sign": 0.25420944136567414, "keep": 0.4740804508328438, "leak": 0.4085653470829129, "replacement": 1.0035361051559448} +{"step": 47, "steps": 250, "lr": 4.8199537482323545e-05, "grad_norm": 2.0243566036224365, "loss_scale": 4.0, "retries": 0, "seconds": 39.2, "max_memory_gb": 18.17, "pair": 0.39692650362849236, "sign": 0.09826578374486417, "keep": 0.1386278719000984, "leak": 1.0445698201656342, "replacement": 0.7765157669782639} +{"step": 48, "steps": 250, "lr": 4.807927176202073e-05, "grad_norm": 2.5490341186523438, "loss_scale": 4.0, "retries": 0, "seconds": 40.1, "max_memory_gb": 18.17, "pair": 0.25220008194446564, "sign": 0.018256249168189242, "keep": 0.40268975344952196, "leak": 1.1812362298369408, "replacement": 0.7579821199178696} +{"step": 49, "steps": 250, "lr": 4.795529674556129e-05, "grad_norm": 2.2557337284088135, "loss_scale": 4.0, "retries": 0, "seconds": 40.9, "max_memory_gb": 18.17, "pair": 0.37141427770256996, "sign": 0.05719513352960348, "keep": 0.40107320162132964, "leak": 0.7652039080858231, "replacement": 0.7918493896722794} +{"step": 50, "steps": 250, "lr": 4.782763477873361e-05, "grad_norm": 2.417966365814209, "loss_scale": 4.0, "retries": 0, "seconds": 41.7, "max_memory_gb": 18.17, "pair": 0.39442865550518036, "sign": 0.10714186914265156, "keep": 0.5707581696624402, "leak": 0.8844037055969238, "replacement": 0.9544796049594879} +{"step": 51, "steps": 250, "lr": 4.769630887187782e-05, "grad_norm": 2.135035991668701, "loss_scale": 4.0, "retries": 0, "seconds": 42.5, "max_memory_gb": 18.17, "pair": 0.2838339991867542, "sign": 0.023709762666840106, "keep": 1.0416201994121366, "leak": 0.9186786524951458, "replacement": 1.0461685955524445} +{"step": 52, "steps": 250, "lr": 4.756134269573832e-05, "grad_norm": 2.1596181392669678, "loss_scale": 4.0, "retries": 0, "seconds": 43.3, "max_memory_gb": 18.17, "pair": 0.4255000166594982, "sign": 0.2735149327199906, "keep": 0.29581568387220614, "leak": 0.8292873241007328, "replacement": 0.9174871891736984} +{"step": 53, "steps": 250, "lr": 4.7422760577197226e-05, "grad_norm": 2.5233728885650635, "loss_scale": 4.0, "retries": 0, "seconds": 44.1, "max_memory_gb": 18.17, "pair": 0.35019319131970406, "sign": 0.091395502560772, "keep": 0.46443353593349457, "leak": 1.1004199385643005, "replacement": 0.9032127112150192} +{"step": 54, "steps": 250, "lr": 4.728058749488963e-05, "grad_norm": 2.4019553661346436, "loss_scale": 4.0, "retries": 0, "seconds": 45.0, "max_memory_gb": 18.17, "pair": 0.35654351860284805, "sign": 0.12659624055959284, "keep": 0.21894688718020916, "leak": 1.0926519185304642, "replacement": 0.8024780601263046} +{"step": 55, "steps": 250, "lr": 4.713484907470133e-05, "grad_norm": 1.56195867061615, "loss_scale": 4.0, "retries": 0, "seconds": 45.8, "max_memory_gb": 18.17, "pair": 0.288498617708683, "sign": 0.030673070752527565, "keep": 0.027763589459937066, "leak": 1.0004479214549065, "replacement": 0.5678289160132408} +{"step": 56, "steps": 250, "lr": 4.6985571585149876e-05, "grad_norm": 1.460221529006958, "loss_scale": 4.0, "retries": 0, "seconds": 46.5, "max_memory_gb": 18.17, "pair": 0.28802402317523956, "sign": 0.014797285199165344, "keep": 0.09990102797746658, "leak": 0.7990852445363998, "replacement": 0.5451444685459137} +{"step": 57, "steps": 250, "lr": 4.6832781932649884e-05, "grad_norm": 2.460616111755371, "loss_scale": 4.0, "retries": 0, "seconds": 47.3, "max_memory_gb": 18.17, "pair": 0.34092192724347115, "sign": 0.3531929641030729, "keep": 0.2155751985119423, "leak": 0.7936236429959536, "replacement": 0.8237119615077972} +{"step": 58, "steps": 250, "lr": 4.667650765666325e-05, "grad_norm": 2.825300693511963, "loss_scale": 4.0, "retries": 0, "seconds": 48.1, "max_memory_gb": 18.17, "pair": 0.18065038323402405, "sign": 0.013297197700012475, "keep": 0.06268567833467387, "leak": 1.4180735647678375, "replacement": 0.5731602162122726} +{"step": 59, "steps": 250, "lr": 4.651677692473538e-05, "grad_norm": 1.4862507581710815, "loss_scale": 4.0, "retries": 0, "seconds": 49.0, "max_memory_gb": 18.17, "pair": 0.34693516977131367, "sign": 0.04724118625745177, "keep": 0.04797638843592722, "leak": 0.5726120844483376, "replacement": 0.5376969799399376} +{"step": 60, "steps": 250, "lr": 4.635361852741808e-05, "grad_norm": 1.9457101821899414, "loss_scale": 4.0, "retries": 0, "seconds": 49.8, "max_memory_gb": 18.17, "pair": 0.3047081343829632, "sign": 0.059899087995290756, "keep": 0.0993755423114635, "leak": 0.8924884237349033, "replacement": 0.6074675470590591} +{"step": 61, "steps": 250, "lr": 4.6187061873080264e-05, "grad_norm": 2.751528024673462, "loss_scale": 4.0, "retries": 0, "seconds": 50.6, "max_memory_gb": 18.17, "pair": 0.44762060791254044, "sign": 0.2555396114476025, "keep": 0.1282631065696478, "leak": 1.2473979223286733, "replacement": 0.9513714760541916} +{"step": 62, "steps": 250, "lr": 4.601713698260727e-05, "grad_norm": 1.3635714054107666, "loss_scale": 4.0, "retries": 0, "seconds": 51.5, "max_memory_gb": 18.17, "pair": 0.3584125190973282, "sign": 0.026506415801122785, "keep": 0.058780148334335536, "leak": 0.707375206053257, "replacement": 0.5778996050357819} +{"step": 63, "steps": 250, "lr": 4.5843874483989746e-05, "grad_norm": 2.0650525093078613, "loss_scale": 4.0, "retries": 0, "seconds": 52.3, "max_memory_gb": 18.17, "pair": 0.37180114537477493, "sign": 0.12135029910132289, "keep": 0.051258682273328304, "leak": 0.45653707534074783, "replacement": 0.5722398981451988} +{"step": 64, "steps": 250, "lr": 4.566730560680316e-05, "grad_norm": 1.6006404161453247, "loss_scale": 4.0, "retries": 0, "seconds": 53.2, "max_memory_gb": 18.17, "pair": 0.25895487889647484, "sign": 0.015216442756354809, "keep": 0.062463223934173584, "leak": 0.7993408888578415, "replacement": 0.49762991815805435} +{"step": 65, "steps": 250, "lr": 4.548746217657878e-05, "grad_norm": 3.201526403427124, "loss_scale": 4.0, "retries": 0, "seconds": 54.0, "max_memory_gb": 18.17, "pair": 0.32399922981858253, "sign": 0.25893525406718254, "keep": 0.3182030748575926, "leak": 1.3732690215110779, "replacement": 0.9558856785297394} +{"step": 66, "steps": 250, "lr": 4.5304376609067396e-05, "grad_norm": 2.2465715408325195, "loss_scale": 4.0, "retries": 0, "seconds": 54.9, "max_memory_gb": 18.17, "pair": 0.3608718663454056, "sign": 0.15342721945489757, "keep": 0.3538660618069098, "leak": 0.7107585966587067, "replacement": 0.7922081649303436} +{"step": 67, "steps": 250, "lr": 4.5118081904396504e-05, "grad_norm": 3.052704095840454, "loss_scale": 4.0, "retries": 0, "seconds": 55.7, "max_memory_gb": 18.17, "pair": 0.1633479930460453, "sign": 0.015896631637588143, "keep": 0.036262479436118156, "leak": 1.4971783310174942, "replacement": 0.5637221187353134} +{"step": 68, "steps": 250, "lr": 4.4928611641122225e-05, "grad_norm": 1.4007630348205566, "loss_scale": 4.0, "retries": 0, "seconds": 56.6, "max_memory_gb": 18.17, "pair": 0.38166917860507965, "sign": 0.051352043403312564, "keep": 0.07318483722156088, "leak": 0.7121932581067085, "replacement": 0.621985912322998} +{"step": 69, "steps": 250, "lr": 4.473599997017701e-05, "grad_norm": 2.8487942218780518, "loss_scale": 4.0, "retries": 0, "seconds": 57.3, "max_memory_gb": 18.17, "pair": 0.5545113794505596, "sign": 0.8813302798662335, "keep": 0.8045517452992499, "leak": 0.8477113954722881, "replacement": 1.6093802452087402} +{"step": 70, "steps": 250, "lr": 4.4540281608714053e-05, "grad_norm": 2.6595406532287598, "loss_scale": 4.0, "retries": 0, "seconds": 58.1, "max_memory_gb": 18.17, "pair": 0.2909915167838335, "sign": 0.030813800636678934, "keep": 0.07652251160470769, "leak": 1.2501225136220455, "replacement": 0.6571902930736542} +{"step": 71, "steps": 250, "lr": 4.434149183384977e-05, "grad_norm": 3.0997774600982666, "loss_scale": 4.0, "retries": 0, "seconds": 58.9, "max_memory_gb": 18.17, "pair": 0.1426145900040865, "sign": 0.007308302563615143, "keep": 0.04392248852036573, "leak": 1.5102331936359406, "replacement": 0.5457882806658745} +{"step": 72, "steps": 250, "lr": 4.413966647630534e-05, "grad_norm": 2.5002007484436035, "loss_scale": 4.0, "retries": 0, "seconds": 59.7, "max_memory_gb": 18.17, "pair": 0.3031403422355652, "sign": 0.19242012011818588, "keep": 0.15098829474300146, "leak": 0.9258439242839813, "replacement": 0.7063055485486984} +{"step": 73, "steps": 250, "lr": 4.3934841913948323e-05, "grad_norm": 1.417062520980835, "loss_scale": 4.0, "retries": 0, "seconds": 60.5, "max_memory_gb": 18.17, "pair": 0.42089029774069786, "sign": 0.12336230371147394, "keep": 0.5497786981432, "leak": 0.4484785944223404, "replacement": 0.8695803955197334} +{"step": 74, "steps": 250, "lr": 4.372705506523583e-05, "grad_norm": 2.502122402191162, "loss_scale": 4.0, "retries": 0, "seconds": 61.3, "max_memory_gb": 18.17, "pair": 0.3575877547264099, "sign": 0.2300567360362038, "keep": 0.15084027676221012, "leak": 0.8459776167874224, "replacement": 0.7595306485891342} +{"step": 75, "steps": 250, "lr": 4.351634338256017e-05, "grad_norm": 1.5423102378845215, "loss_scale": 4.0, "retries": 0, "seconds": 62.2, "max_memory_gb": 18.17, "pair": 0.6723864860832691, "sign": 0.5274754712008871, "keep": 0.9131901315413415, "leak": 0.27731473091989756, "replacement": 1.4620479941368103} +{"step": 76, "steps": 250, "lr": 4.330274484549825e-05, "grad_norm": 2.0626397132873535, "loss_scale": 4.0, "retries": 0, "seconds": 63.0, "max_memory_gb": 18.17, "pair": 0.2973857410252094, "sign": 0.09561600210145116, "keep": 0.049326323613058776, "leak": 0.7638554163277149, "replacement": 0.5608207583427429} +{"step": 77, "steps": 250, "lr": 4.308629795396599e-05, "grad_norm": 2.3754167556762695, "loss_scale": 4.0, "retries": 0, "seconds": 63.8, "max_memory_gb": 18.17, "pair": 0.23215588182210922, "sign": 0.010060922126285732, "keep": 0.016375897288526176, "leak": 1.0655015408992767, "replacement": 0.5117496848106384} +{"step": 78, "steps": 250, "lr": 4.286704172127894e-05, "grad_norm": 2.2314236164093018, "loss_scale": 4.0, "retries": 0, "seconds": 64.6, "max_memory_gb": 18.17, "pair": 0.3406652621924877, "sign": 0.5063187128398567, "keep": 0.0016870326217031106, "leak": 0.4316643252968788, "replacement": 0.7025842219591141} +{"step": 79, "steps": 250, "lr": 4.264501566712033e-05, "grad_norm": 1.6750049591064453, "loss_scale": 4.0, "retries": 0, "seconds": 65.4, "max_memory_gb": 18.17, "pair": 0.34534838795661926, "sign": 0.022354907589033246, "keep": 0.03581179837374293, "leak": 0.6972186304628849, "replacement": 0.5487364009022713} +{"step": 80, "steps": 250, "lr": 4.2420259810417894e-05, "grad_norm": 1.625174641609192, "loss_scale": 4.0, "retries": 0, "seconds": 66.2, "max_memory_gb": 18.17, "pair": 0.37053806707262993, "sign": 0.07666252972558141, "keep": 0.018573656536148064, "leak": 0.7722587212920189, "replacement": 0.6112208440899849} +{"step": 81, "steps": 250, "lr": 4.219281466213066e-05, "grad_norm": 5.389065265655518, "loss_scale": 4.0, "retries": 0, "seconds": 67.0, "max_memory_gb": 18.17, "pair": 0.2793600335717201, "sign": 0.05688135256059468, "keep": 0.12129455548711121, "leak": 1.8782419636845589, "replacement": 0.8380084782838821} +{"step": 82, "steps": 250, "lr": 4.1962721217947135e-05, "grad_norm": 2.1664786338806152, "loss_scale": 4.0, "retries": 0, "seconds": 67.8, "max_memory_gb": 18.17, "pair": 0.42625515535473824, "sign": 0.9615410937694833, "keep": 0.08101894031278789, "leak": 0.7612776607275009, "replacement": 1.1378546133637428} +{"step": 83, "steps": 250, "lr": 4.173002095089599e-05, "grad_norm": 1.921890139579773, "loss_scale": 4.0, "retries": 0, "seconds": 68.6, "max_memory_gb": 18.17, "pair": 0.2368405256420374, "sign": 0.022078357957070693, "keep": 0.318322260864079, "leak": 0.7916181981563568, "replacement": 0.6049453839659691} +{"step": 84, "steps": 250, "lr": 4.149475580387084e-05, "grad_norm": 1.9519962072372437, "loss_scale": 4.0, "retries": 0, "seconds": 69.5, "max_memory_gb": 18.17, "pair": 0.1915804911404848, "sign": 0.024546128028305247, "keep": 0.017287287356111847, "leak": 0.9554132148623466, "replacement": 0.45135049521923065} +{"step": 85, "steps": 250, "lr": 4.1256968182070275e-05, "grad_norm": 2.4401497840881348, "loss_scale": 4.0, "retries": 0, "seconds": 70.4, "max_memory_gb": 18.17, "pair": 0.2344406247138977, "sign": 0.03869649989064783, "keep": 0.03727948942923831, "leak": 0.7657077312469482, "replacement": 0.4638555571436882} +{"step": 86, "steps": 250, "lr": 4.101670094535453e-05, "grad_norm": 3.566443920135498, "loss_scale": 4.0, "retries": 0, "seconds": 71.2, "max_memory_gb": 18.17, "pair": 0.41340063512325287, "sign": 0.27340006455779076, "keep": 0.36913408897817135, "leak": 0.3842466464266181, "replacement": 0.8307293802499771} +{"step": 87, "steps": 250, "lr": 4.077399740052027e-05, "grad_norm": 1.5820107460021973, "loss_scale": 4.0, "retries": 0, "seconds": 72.0, "max_memory_gb": 18.17, "pair": 0.34069687500596046, "sign": 0.08954072871711105, "keep": 0.022341555093589704, "leak": 0.5883921235799789, "replacement": 0.5437360256910324} +{"step": 88, "steps": 250, "lr": 4.0528901293494775e-05, "grad_norm": 2.3838775157928467, "loss_scale": 4.0, "retries": 0, "seconds": 72.8, "max_memory_gb": 18.17, "pair": 0.5553812123835087, "sign": 0.31245615845546126, "keep": 0.0947626261604455, "leak": 0.4698200039565563, "replacement": 0.8764456361532211} +{"step": 89, "steps": 250, "lr": 4.0281456801451004e-05, "grad_norm": 2.231713056564331, "loss_scale": 4.0, "retries": 0, "seconds": 73.6, "max_memory_gb": 18.17, "pair": 0.4936697259545326, "sign": 0.16403938608709723, "keep": 0.012648463197820092, "leak": 0.166826666216366, "replacement": 0.6237203031778336} +{"step": 90, "steps": 250, "lr": 4.0031708524844884e-05, "grad_norm": 1.639687418937683, "loss_scale": 4.0, "retries": 0, "seconds": 74.4, "max_memory_gb": 18.17, "pair": 0.3159483075141907, "sign": 0.06355850998079404, "keep": 0.022011907043633983, "leak": 0.43667763844132423, "replacement": 0.46790292114019394} +{"step": 91, "steps": 250, "lr": 3.977970147937635e-05, "grad_norm": 4.590484142303467, "loss_scale": 4.0, "retries": 0, "seconds": 75.3, "max_memory_gb": 18.17, "pair": 0.5906499102711678, "sign": 0.3077406559023075, "keep": 0.40639011573512107, "leak": 0.6476146038621664, "replacement": 1.1096189767122269} +{"step": 92, "steps": 250, "lr": 3.952548108787553e-05, "grad_norm": 3.350109100341797, "loss_scale": 4.0, "retries": 0, "seconds": 76.1, "max_memory_gb": 18.17, "pair": 0.33302442729473114, "sign": 0.6188254985027015, "keep": 1.0147776128806072, "leak": 0.254008661955595, "replacement": 1.2133281379938126} +{"step": 93, "steps": 250, "lr": 3.9269093172115496e-05, "grad_norm": 3.524967670440674, "loss_scale": 4.0, "retries": 0, "seconds": 76.9, "max_memory_gb": 18.17, "pair": 0.3199533298611641, "sign": 0.03249965503346175, "keep": 0.1396772027364932, "leak": 1.2607371946796775, "replacement": 0.7212260514497757} +{"step": 94, "steps": 250, "lr": 3.901058394455323e-05, "grad_norm": 2.1719720363616943, "loss_scale": 4.0, "retries": 0, "seconds": 77.8, "max_memory_gb": 18.17, "pair": 0.23452629521489143, "sign": 0.02416399330832064, "keep": 0.41575773421209306, "leak": 1.1637294292449951, "replacement": 0.745419517159462} +{"step": 95, "steps": 250, "lr": 3.875e-05, "grad_norm": 3.327775478363037, "loss_scale": 4.0, "retries": 0, "seconds": 78.6, "max_memory_gb": 18.17, "pair": 0.43191106617450714, "sign": 0.31573274033144116, "keep": 1.3775935736484826, "leak": 0.8468556557781994, "replacement": 1.4902881234884262} +{"step": 96, "steps": 250, "lr": 3.8487388307223005e-05, "grad_norm": 2.242480516433716, "loss_scale": 4.0, "retries": 0, "seconds": 79.4, "max_memory_gb": 18.17, "pair": 0.3611767776310444, "sign": 0.10875860368832946, "keep": 0.9899713124614209, "leak": 0.6576469819992781, "replacement": 1.0749534964561462} +{"step": 97, "steps": 250, "lr": 3.822279620047943e-05, "grad_norm": 1.8473832607269287, "loss_scale": 4.0, "retries": 0, "seconds": 80.2, "max_memory_gb": 18.17, "pair": 0.15059569105505943, "sign": 0.0103443005355075, "keep": 0.01082495675291284, "leak": 0.9193672426044941, "replacement": 0.39102213829755783} +{"step": 98, "steps": 250, "lr": 3.795627137098479e-05, "grad_norm": 1.9688389301300049, "loss_scale": 4.0, "retries": 0, "seconds": 81.0, "max_memory_gb": 18.17, "pair": 0.35443735867738724, "sign": 0.14212175749707967, "keep": 0.035180263043002924, "leak": 1.0097357705235481, "replacement": 0.6955223083496094} +{"step": 99, "steps": 250, "lr": 3.7687861858316844e-05, "grad_norm": 3.7333648204803467, "loss_scale": 4.0, "retries": 0, "seconds": 81.9, "max_memory_gb": 18.17, "pair": 0.4362499415874481, "sign": 0.7386243171058595, "keep": 0.010904337772331019, "leak": 1.851680988445878, "replacement": 1.273934505879879} +{"step": 100, "steps": 250, "lr": 3.741761604175668e-05, "grad_norm": 3.3437578678131104, "loss_scale": 4.0, "retries": 0, "seconds": 82.7, "max_memory_gb": 18.17, "pair": 0.5597295463085175, "sign": 0.9279602147871628, "keep": 0.39666370244231075, "leak": 1.09428109228611, "replacement": 1.4956117123365402} +{"step": 101, "steps": 250, "lr": 3.714558263156872e-05, "grad_norm": 3.5701680183410645, "loss_scale": 4.0, "retries": 0, "seconds": 83.6, "max_memory_gb": 18.17, "pair": 0.17073849961161613, "sign": 0.007187618059106171, "keep": 0.04905796455568634, "leak": 1.4277653768658638, "replacement": 0.555802620947361} +{"step": 102, "steps": 250, "lr": 3.6871810660220904e-05, "grad_norm": 1.621874451637268, "loss_scale": 4.0, "retries": 0, "seconds": 84.4, "max_memory_gb": 18.17, "pair": 0.20542125776410103, "sign": 0.01442736922763288, "keep": 0.04515112889930606, "leak": 0.5735124386847019, "replacement": 0.37858860194683075} +{"step": 103, "steps": 250, "lr": 3.659634947354686e-05, "grad_norm": 2.542970657348633, "loss_scale": 4.0, "retries": 0, "seconds": 85.2, "max_memory_gb": 18.17, "pair": 0.2685791477560997, "sign": 0.02387667680159211, "keep": 0.02217749177884798, "leak": 1.388171225786209, "replacement": 0.6386490389704704} +{"step": 104, "steps": 250, "lr": 3.631924872185169e-05, "grad_norm": 2.191150188446045, "loss_scale": 4.0, "retries": 0, "seconds": 86.0, "max_memory_gb": 18.17, "pair": 0.32051808945834637, "sign": 0.20156950555974618, "keep": 0.027911223252885975, "leak": 0.28982645459473133, "replacement": 0.5077150873839855} +{"step": 105, "steps": 250, "lr": 3.6040558350962674e-05, "grad_norm": 2.046018123626709, "loss_scale": 4.0, "retries": 0, "seconds": 86.8, "max_memory_gb": 18.17, "pair": 0.22081071510910988, "sign": 0.018071043770760298, "keep": 0.10475947626400739, "leak": 1.0183388441801071, "replacement": 0.5368106886744499} +{"step": 106, "steps": 250, "lr": 3.576032859322687e-05, "grad_norm": 2.07918119430542, "loss_scale": 4.0, "retries": 0, "seconds": 87.6, "max_memory_gb": 18.17, "pair": 0.22429409995675087, "sign": 0.03221876220777631, "keep": 0.007746034305455396, "leak": 0.8876442909240723, "replacement": 0.46618758887052536} +{"step": 107, "steps": 250, "lr": 3.5478609958457056e-05, "grad_norm": 3.4643290042877197, "loss_scale": 4.0, "retries": 0, "seconds": 88.4, "max_memory_gb": 18.17, "pair": 0.24379804730415344, "sign": 0.28594460792373866, "keep": 1.3153627049123315, "leak": 0.5897563090547919, "replacement": 1.1918907761573792} +{"step": 108, "steps": 250, "lr": 3.519545322482755e-05, "grad_norm": 3.5935497283935547, "loss_scale": 4.0, "retries": 0, "seconds": 89.2, "max_memory_gb": 18.17, "pair": 0.47334009408950806, "sign": 0.2223101642448455, "keep": 0.024688926348062523, "leak": 0.3859858699142933, "replacement": 0.6933361142873764} +{"step": 109, "steps": 250, "lr": 3.491090942972179e-05, "grad_norm": 1.9338059425354004, "loss_scale": 4.0, "retries": 0, "seconds": 90.1, "max_memory_gb": 18.17, "pair": 0.36409879103302956, "sign": 0.10643050284124911, "keep": 0.07968329777941108, "leak": 0.48363771801814437, "replacement": 0.5780650973320007} +{"step": 110, "steps": 250, "lr": 3.4625029860533135e-05, "grad_norm": 1.241780161857605, "loss_scale": 4.0, "retries": 0, "seconds": 90.9, "max_memory_gb": 18.17, "pair": 0.2313193902373314, "sign": 0.12119361897930503, "keep": 0.18059091549366713, "leak": 0.848620630800724, "replacement": 0.5943668186664581} +{"step": 111, "steps": 250, "lr": 3.433786604542056e-05, "grad_norm": 4.3226847648620605, "loss_scale": 4.0, "retries": 0, "seconds": 91.8, "max_memory_gb": 18.17, "pair": 0.27624889090657234, "sign": 1.1993132591451285, "keep": 0.002459886197357264, "leak": 0.84696826338768, "replacement": 1.0888775289058685} +{"step": 112, "steps": 250, "lr": 3.404946974402104e-05, "grad_norm": 2.9510910511016846, "loss_scale": 4.0, "retries": 0, "seconds": 92.6, "max_memory_gb": 18.17, "pair": 0.20717865973711014, "sign": 0.01045088324462995, "keep": 0.005201602256420301, "leak": 0.7841723188757896, "replacement": 0.411047987639904} +{"step": 113, "steps": 250, "lr": 3.375989293812019e-05, "grad_norm": 1.3513751029968262, "loss_scale": 4.0, "retries": 0, "seconds": 93.4, "max_memory_gb": 18.17, "pair": 0.27363210543990135, "sign": 0.02881364186760038, "keep": 0.008159542630892247, "leak": 0.5237103551626205, "replacement": 0.4230462983250618} +{"step": 114, "steps": 250, "lr": 3.3469187822282766e-05, "grad_norm": 2.077953577041626, "loss_scale": 4.0, "retries": 0, "seconds": 94.2, "max_memory_gb": 18.17, "pair": 0.31187446415424347, "sign": 0.041127453092485666, "keep": 0.028813386103138328, "leak": 0.6165926083922386, "replacement": 0.500993050634861} +{"step": 115, "steps": 250, "lr": 3.317740679444499e-05, "grad_norm": 1.4412602186203003, "loss_scale": 4.0, "retries": 0, "seconds": 95.1, "max_memory_gb": 18.17, "pair": 0.17241575941443443, "sign": 0.00887286476790905, "keep": 0.01469637312766281, "leak": 0.4372039996087551, "replacement": 0.2935013771057129} +{"step": 116, "steps": 250, "lr": 3.2884602446470056e-05, "grad_norm": 3.7284889221191406, "loss_scale": 4.0, "retries": 0, "seconds": 95.9, "max_memory_gb": 18.17, "pair": 0.32463992573320866, "sign": 0.6611707573465537, "keep": 0.8257795873823852, "leak": 0.39443931728601456, "replacement": 1.166724856942892} +{"step": 117, "steps": 250, "lr": 3.259082755466876e-05, "grad_norm": 3.6170995235443115, "loss_scale": 4.0, "retries": 0, "seconds": 96.7, "max_memory_gb": 18.17, "pair": 0.25469791516661644, "sign": 0.1903729361947626, "keep": 0.7530714574531885, "leak": 0.8224138021469116, "replacement": 0.9320235438644886} +{"step": 118, "steps": 250, "lr": 3.229613507028686e-05, "grad_norm": 1.6739252805709839, "loss_scale": 4.0, "retries": 0, "seconds": 97.6, "max_memory_gb": 18.17, "pair": 0.3840768039226532, "sign": 0.3596343712415546, "keep": 0.5124670486866307, "leak": 0.34138190746307373, "replacement": 0.9054730087518692} +{"step": 119, "steps": 250, "lr": 3.2000578109961004e-05, "grad_norm": 2.610020875930786, "loss_scale": 4.0, "retries": 0, "seconds": 98.4, "max_memory_gb": 18.17, "pair": 0.3744955472648144, "sign": 0.27143482118844986, "keep": 0.00746975993797605, "leak": 0.501249723136425, "replacement": 0.6392602697014809} +{"step": 120, "steps": 250, "lr": 3.1704209946144616e-05, "grad_norm": 4.472949981689453, "loss_scale": 4.0, "retries": 0, "seconds": 99.2, "max_memory_gb": 18.17, "pair": 0.29177081026136875, "sign": 0.769932220922783, "keep": 1.4909847193048336, "leak": 0.3782142102718353, "replacement": 1.516782820224762} +{"step": 121, "steps": 250, "lr": 3.140708399750594e-05, "grad_norm": 1.7626410722732544, "loss_scale": 4.0, "retries": 0, "seconds": 100.1, "max_memory_gb": 18.17, "pair": 0.49593542143702507, "sign": 0.39271668024593964, "keep": 0.06915804533332448, "leak": 0.26400951331015676, "replacement": 0.7928751520812511} +{"step": 122, "steps": 250, "lr": 3.1109253819299606e-05, "grad_norm": 2.407258987426758, "loss_scale": 4.0, "retries": 0, "seconds": 100.9, "max_memory_gb": 18.17, "pair": 0.4279465191066265, "sign": 0.09567732358118519, "keep": 0.05423723905255429, "leak": 0.3387216981500387, "replacement": 0.5875842273235321} +{"step": 123, "steps": 250, "lr": 3.081077309371361e-05, "grad_norm": 3.602332353591919, "loss_scale": 4.0, "retries": 0, "seconds": 101.8, "max_memory_gb": 18.17, "pair": 0.3643048657104373, "sign": 1.1839100796496496, "keep": 0.07595195109024644, "leak": 0.7805480286478996, "replacement": 1.1893729120492935} +{"step": 124, "steps": 250, "lr": 3.051169562019339e-05, "grad_norm": 2.9297354221343994, "loss_scale": 4.0, "retries": 0, "seconds": 102.6, "max_memory_gb": 18.17, "pair": 0.5652232430875301, "sign": 0.674525368725881, "keep": 1.9331586288826657, "leak": 0.3343222886469448, "replacement": 1.9526458606123924} +{"step": 125, "steps": 250, "lr": 3.021207530574477e-05, "grad_norm": 2.077949047088623, "loss_scale": 4.0, "retries": 0, "seconds": 103.4, "max_memory_gb": 18.17, "pair": 0.37976974435150623, "sign": 0.24292722856625915, "keep": 0.21164456191763747, "leak": 0.8153133550658822, "replacement": 0.8108839616179466} +{"step": 126, "steps": 250, "lr": 2.991196615521758e-05, "grad_norm": 2.63486909866333, "loss_scale": 4.0, "retries": 0, "seconds": 104.2, "max_memory_gb": 18.17, "pair": 0.6457309536635876, "sign": 0.4954267314169556, "keep": 1.2359007284976542, "leak": 0.6390068799519213, "replacement": 1.6711463928222656} +{"step": 127, "steps": 250, "lr": 2.9611422261571536e-05, "grad_norm": 1.136264681816101, "loss_scale": 4.0, "retries": 0, "seconds": 105.0, "max_memory_gb": 18.17, "pair": 0.1588508952409029, "sign": 0.008243468881119043, "keep": 0.016737906958042004, "leak": 0.24388455599546432, "replacement": 0.23231272399425507} +{"step": 128, "steps": 250, "lr": 2.931049779612634e-05, "grad_norm": 4.020803451538086, "loss_scale": 4.0, "retries": 0, "seconds": 105.9, "max_memory_gb": 18.17, "pair": 0.2498484868556261, "sign": 0.7675502269412391, "keep": 0.8525128586916253, "leak": 0.46147918701171875, "replacement": 1.1752498671412468} +{"step": 129, "steps": 250, "lr": 2.9009246998797557e-05, "grad_norm": 2.247441530227661, "loss_scale": 4.0, "retries": 0, "seconds": 106.6, "max_memory_gb": 18.17, "pair": 0.3324233628809452, "sign": 0.08604081650264561, "keep": 0.09886748786084354, "leak": 0.31960031669586897, "replacement": 0.5047776028513908} +{"step": 130, "steps": 250, "lr": 2.8707724168320238e-05, "grad_norm": 3.3935697078704834, "loss_scale": 4.0, "retries": 0, "seconds": 107.4, "max_memory_gb": 18.17, "pair": 0.3902419675141573, "sign": 1.4640270108357072, "keep": 0.06112471984306467, "leak": 0.27525492012500763, "replacement": 1.2216315530240536} +{"step": 131, "steps": 250, "lr": 2.840598365246184e-05, "grad_norm": 4.4288249015808105, "loss_scale": 4.0, "retries": 0, "seconds": 108.1, "max_memory_gb": 18.17, "pair": 0.3466063216328621, "sign": 0.3273714086972177, "keep": 0.4255080118246042, "leak": 0.5441746637225151, "replacement": 0.8590896911919117} +{"step": 132, "steps": 250, "lr": 2.8104079838226383e-05, "grad_norm": 2.3694875240325928, "loss_scale": 4.0, "retries": 0, "seconds": 108.9, "max_memory_gb": 18.17, "pair": 0.2600047253072262, "sign": 0.017355729592964053, "keep": 0.376871984439731, "leak": 0.6452087629586458, "replacement": 0.6184207499027252} +{"step": 133, "steps": 250, "lr": 2.7802067142051446e-05, "grad_norm": 3.2940680980682373, "loss_scale": 4.0, "retries": 0, "seconds": 109.7, "max_memory_gb": 18.17, "pair": 0.551292423158884, "sign": 0.6033767648041248, "keep": 1.0641722440486774, "leak": 0.6916033625602722, "replacement": 1.5579677820205688} +{"step": 134, "steps": 250, "lr": 2.7500000000000004e-05, "grad_norm": 2.488582134246826, "loss_scale": 4.0, "retries": 0, "seconds": 110.5, "max_memory_gb": 18.17, "pair": 0.32608535699546337, "sign": 0.3026792406453751, "keep": 0.5555855481572962, "leak": 0.18186549469828606, "replacement": 0.8006841279566288} +{"step": 135, "steps": 250, "lr": 2.7197932857948566e-05, "grad_norm": 2.1613736152648926, "loss_scale": 4.0, "retries": 0, "seconds": 111.4, "max_memory_gb": 18.17, "pair": 0.20980238541960716, "sign": 0.01984706975053996, "keep": 0.015587902715196833, "leak": 0.7419911175966263, "replacement": 0.4130176603794098} +{"step": 136, "steps": 250, "lr": 2.689592016177363e-05, "grad_norm": 3.301751136779785, "loss_scale": 4.0, "retries": 0, "seconds": 112.3, "max_memory_gb": 18.17, "pair": 0.2230825237929821, "sign": 0.5165968402288854, "keep": 0.1347394577896921, "leak": 0.9632363170385361, "replacement": 0.7895597219467163} +{"step": 137, "steps": 250, "lr": 2.659401634753817e-05, "grad_norm": 2.6249566078186035, "loss_scale": 4.0, "retries": 0, "seconds": 113.3, "max_memory_gb": 18.17, "pair": 0.2557980939745903, "sign": 0.02364984661107883, "keep": 0.005829372108564712, "leak": 0.534868411719799, "replacement": 0.40425480902194977} +{"step": 138, "steps": 250, "lr": 2.6292275831679764e-05, "grad_norm": 1.476804494857788, "loss_scale": 4.0, "retries": 0, "seconds": 114.2, "max_memory_gb": 18.17, "pair": 0.2176252342760563, "sign": 0.02518782007973641, "keep": 0.048900924157351255, "leak": 0.39399056043475866, "replacement": 0.35316723585128784} +{"step": 139, "steps": 250, "lr": 2.5990753001202445e-05, "grad_norm": 3.8413491249084473, "loss_scale": 4.0, "retries": 0, "seconds": 115.2, "max_memory_gb": 18.17, "pair": 0.24913001619279385, "sign": 0.41426775860600173, "keep": 0.004096079122973606, "leak": 1.1370284333825111, "replacement": 0.7425690367817879} +{"step": 140, "steps": 250, "lr": 2.5689502203873672e-05, "grad_norm": 2.679178476333618, "loss_scale": 4.0, "retries": 0, "seconds": 116.1, "max_memory_gb": 18.17, "pair": 0.17222920525819063, "sign": 0.02121372232795693, "keep": 0.007326397138967877, "leak": 0.7770557557232678, "replacement": 0.3807632066309452} +{"step": 141, "steps": 250, "lr": 2.538857773842847e-05, "grad_norm": 2.168240547180176, "loss_scale": 4.0, "retries": 0, "seconds": 117.1, "max_memory_gb": 18.17, "pair": 0.21228148136287928, "sign": 0.1969339725037571, "keep": 0.0026693965337472036, "leak": 0.3456839509308338, "replacement": 0.3985041603446007} +{"step": 142, "steps": 250, "lr": 2.5088033844782426e-05, "grad_norm": 3.3755507469177246, "loss_scale": 4.0, "retries": 0, "seconds": 117.9, "max_memory_gb": 18.17, "pair": 0.3489937409758568, "sign": 0.27624228433705866, "keep": 0.11562990590027766, "leak": 0.7599539719521999, "replacement": 0.7349183335900307} +{"step": 143, "steps": 250, "lr": 2.4787924694255237e-05, "grad_norm": 3.788405179977417, "loss_scale": 4.0, "retries": 0, "seconds": 118.9, "max_memory_gb": 18.17, "pair": 0.38890455663204193, "sign": 0.19361357460729778, "keep": 1.1121899187195368, "leak": 0.5295687317848206, "replacement": 1.174198493361473} +{"step": 144, "steps": 250, "lr": 2.4488304379806627e-05, "grad_norm": 4.047565937042236, "loss_scale": 4.0, "retries": 0, "seconds": 119.7, "max_memory_gb": 18.17, "pair": 0.39430226758122444, "sign": 0.7279972590040416, "keep": 0.194636469357647, "leak": 0.3819065671414137, "replacement": 0.9510957524180412} +{"step": 145, "steps": 250, "lr": 2.4189226906286388e-05, "grad_norm": 3.4369471073150635, "loss_scale": 4.0, "retries": 0, "seconds": 120.6, "max_memory_gb": 18.17, "pair": 0.22641568258404732, "sign": 0.044392478943336755, "keep": 0.0023912369974823378, "leak": 1.2545124180614948, "replacement": 0.5634356401860714} +{"step": 146, "steps": 250, "lr": 2.3890746180700395e-05, "grad_norm": 2.765627384185791, "loss_scale": 4.0, "retries": 0, "seconds": 121.5, "max_memory_gb": 18.17, "pair": 0.2901443652808666, "sign": 0.0758189361076802, "keep": 0.008302090009237872, "leak": 0.47357630729675293, "replacement": 0.45059896260499954} +{"step": 147, "steps": 250, "lr": 2.359291600249407e-05, "grad_norm": 7.378512382507324, "loss_scale": 4.0, "retries": 0, "seconds": 122.3, "max_memory_gb": 18.17, "pair": 0.4628694914281368, "sign": 1.0674471443053335, "keep": 1.5871065904520947, "leak": 2.9845829978585243, "replacement": 2.5362920686602592} +{"step": 148, "steps": 250, "lr": 2.3295790053855392e-05, "grad_norm": 2.8975675106048584, "loss_scale": 4.0, "retries": 0, "seconds": 123.2, "max_memory_gb": 18.17, "pair": 0.13494735769927502, "sign": 0.1046277646382805, "keep": 0.0009363826975388179, "leak": 0.8912645382806659, "replacement": 0.41054558008909225} +{"step": 149, "steps": 250, "lr": 2.299942189003901e-05, "grad_norm": 1.9459480047225952, "loss_scale": 4.0, "retries": 0, "seconds": 124.2, "max_memory_gb": 18.17, "pair": 0.26680903881788254, "sign": 0.05429472913965583, "keep": 0.007059452194880578, "leak": 0.2168248537927866, "replacement": 0.35169233568012714} +{"step": 150, "steps": 250, "lr": 2.2703864929713144e-05, "grad_norm": 3.027594566345215, "loss_scale": 4.0, "retries": 0, "seconds": 125.1, "max_memory_gb": 18.17, "pair": 0.3401784412562847, "sign": 0.1210677131311968, "keep": 0.12321724410867319, "leak": 0.9685640754178166, "replacement": 0.7044619135558605} +{"step": 151, "steps": 250, "lr": 2.2409172445331257e-05, "grad_norm": 1.2988990545272827, "loss_scale": 4.0, "retries": 0, "seconds": 126.1, "max_memory_gb": 18.17, "pair": 0.16386491246521473, "sign": 0.013611229718662798, "keep": 0.012927798425153014, "leak": 0.16970656998455524, "replacement": 0.21956107020378113} +{"step": 152, "steps": 250, "lr": 2.211539755352996e-05, "grad_norm": 1.3889038562774658, "loss_scale": 4.0, "retries": 0, "seconds": 127.2, "max_memory_gb": 18.17, "pair": 0.1976930946111679, "sign": 0.0067695543402805924, "keep": 0.018723626545579464, "leak": 0.33713238686323166, "replacement": 0.2947227954864502} +{"step": 153, "steps": 250, "lr": 2.1822593205555016e-05, "grad_norm": 2.955699920654297, "loss_scale": 4.0, "retries": 0, "seconds": 128.1, "max_memory_gb": 18.17, "pair": 0.12059342488646507, "sign": 0.007706279662670568, "keep": 0.015067788190009423, "leak": 0.8336671758443117, "replacement": 0.34039725363254547} +{"step": 154, "steps": 250, "lr": 2.1530812177717236e-05, "grad_norm": 2.2859716415405273, "loss_scale": 4.0, "retries": 0, "seconds": 128.9, "max_memory_gb": 18.17, "pair": 0.20148942060768604, "sign": 0.012846406316384673, "keep": 0.19426046591252089, "leak": 0.8199440166354179, "replacement": 0.5100288540124893} +{"step": 155, "steps": 250, "lr": 2.1240107061879815e-05, "grad_norm": 2.520376682281494, "loss_scale": 4.0, "retries": 0, "seconds": 129.9, "max_memory_gb": 18.17, "pair": 0.2011000905185938, "sign": 0.007770207652356476, "keep": 0.5701404645503203, "leak": 0.49731742963194847, "replacement": 0.6143847741186619} +{"step": 156, "steps": 250, "lr": 2.095053025597896e-05, "grad_norm": 3.926335334777832, "loss_scale": 4.0, "retries": 0, "seconds": 130.8, "max_memory_gb": 18.17, "pair": 0.3493007766082883, "sign": 0.0053252141224220395, "keep": 0.009176711696852635, "leak": 0.3276120151858777, "replacement": 0.43845473416149616} +{"step": 157, "steps": 250, "lr": 2.066213395457945e-05, "grad_norm": 4.723710060119629, "loss_scale": 4.0, "retries": 0, "seconds": 131.7, "max_memory_gb": 18.17, "pair": 0.25006100442260504, "sign": 0.973065730300732, "keep": 0.009466138828429393, "leak": 0.9866470006236341, "replacement": 0.9879887066781521} +{"step": 158, "steps": 250, "lr": 2.0374970139466877e-05, "grad_norm": 2.6653473377227783, "loss_scale": 4.0, "retries": 0, "seconds": 132.7, "max_memory_gb": 18.17, "pair": 0.4003439340740442, "sign": 0.42802840867079794, "keep": 0.5543095380512568, "leak": 0.06642193952575326, "replacement": 0.9081184212118387} +{"step": 159, "steps": 250, "lr": 2.0089090570278213e-05, "grad_norm": 3.4514646530151367, "loss_scale": 4.0, "retries": 0, "seconds": 133.6, "max_memory_gb": 18.18, "pair": 0.18938837200403214, "sign": 0.009526451816782355, "keep": 0.0018135537666239543, "leak": 0.41383342258632183, "replacement": 0.2985167223960161} +{"step": 160, "steps": 250, "lr": 1.9804546775172457e-05, "grad_norm": 3.022084951400757, "loss_scale": 4.0, "retries": 0, "seconds": 134.6, "max_memory_gb": 18.18, "pair": 0.3234950117766857, "sign": 0.19805126094433945, "keep": 0.5375140904798172, "leak": 0.6774608111009002, "replacement": 0.8606429025530815} +{"step": 161, "steps": 250, "lr": 1.9521390041542946e-05, "grad_norm": 1.8514654636383057, "loss_scale": 4.0, "retries": 0, "seconds": 135.5, "max_memory_gb": 18.18, "pair": 0.2737023290246725, "sign": 0.1043461849913001, "keep": 0.11538725220452761, "leak": 0.21422703936696053, "replacement": 0.4371258057653904} +{"step": 162, "steps": 250, "lr": 1.923967140677313e-05, "grad_norm": 2.0553412437438965, "loss_scale": 4.0, "retries": 0, "seconds": 136.5, "max_memory_gb": 18.18, "pair": 0.19541904889047146, "sign": 0.022291821311227977, "keep": 0.10243765803532057, "leak": 0.4692212911322713, "replacement": 0.3750891201198101} +{"step": 163, "steps": 250, "lr": 1.8959441649037335e-05, "grad_norm": 2.5585038661956787, "loss_scale": 4.0, "retries": 0, "seconds": 137.4, "max_memory_gb": 18.18, "pair": 0.34681466221809387, "sign": 0.2042918661609292, "keep": 0.21754175564274192, "leak": 0.40133913047611713, "replacement": 0.6580662503838539} +{"step": 164, "steps": 250, "lr": 1.868075127814832e-05, "grad_norm": 2.813997507095337, "loss_scale": 4.0, "retries": 0, "seconds": 138.4, "max_memory_gb": 18.18, "pair": 0.42501427233219147, "sign": 0.2725367227103561, "keep": 0.023197046655695885, "leak": 0.5511343190446496, "replacement": 0.7106647565960884} +{"step": 165, "steps": 250, "lr": 1.8403650526453143e-05, "grad_norm": 1.2510071992874146, "loss_scale": 4.0, "retries": 0, "seconds": 139.4, "max_memory_gb": 18.18, "pair": 0.17398105468600988, "sign": 0.014772380294743925, "keep": 0.0011529978673934238, "leak": 0.1548280380666256, "replacement": 0.22065075300633907} +{"step": 166, "steps": 250, "lr": 1.812818933977912e-05, "grad_norm": 10.891448020935059, "loss_scale": 4.0, "retries": 0, "seconds": 140.4, "max_memory_gb": 18.18, "pair": 0.7611041758209467, "sign": 0.980307330028154, "keep": 1.6985982545011211, "leak": 1.8244042317382991, "replacement": 2.5566581524908543} +{"step": 167, "steps": 250, "lr": 1.785441736843128e-05, "grad_norm": 4.833982944488525, "loss_scale": 4.0, "retries": 0, "seconds": 141.3, "max_memory_gb": 18.18, "pair": 0.15664801932871342, "sign": 0.02658693923149258, "keep": 0.030127311882097274, "leak": 1.96218191832304, "replacement": 0.6755506172776222} +{"step": 168, "steps": 250, "lr": 1.7582383958243314e-05, "grad_norm": 3.2712016105651855, "loss_scale": 4.0, "retries": 0, "seconds": 142.3, "max_memory_gb": 18.18, "pair": 0.17375417798757553, "sign": 0.009725368057843298, "keep": 0.028577233482792508, "leak": 0.9844623748213053, "replacement": 0.4390210770070553} +{"step": 169, "steps": 250, "lr": 1.7312138141683164e-05, "grad_norm": 2.2413299083709717, "loss_scale": 4.0, "retries": 0, "seconds": 143.2, "max_memory_gb": 18.18, "pair": 0.3353869616985321, "sign": 0.13286311784759164, "keep": 0.13958958932198584, "leak": 0.26187784294597805, "replacement": 0.537082776427269} +{"step": 170, "steps": 250, "lr": 1.704372862901521e-05, "grad_norm": 3.631307601928711, "loss_scale": 4.0, "retries": 0, "seconds": 144.2, "max_memory_gb": 18.18, "pair": 0.16688433848321438, "sign": 0.00404667662223801, "keep": 0.0052824710292043164, "leak": 1.6117704436182976, "replacement": 0.574491523206234} +{"step": 171, "steps": 250, "lr": 1.6777203799520573e-05, "grad_norm": 2.6663601398468018, "loss_scale": 4.0, "retries": 0, "seconds": 145.1, "max_memory_gb": 18.18, "pair": 0.2331371670588851, "sign": 0.1478004658856662, "keep": 0.08740971868957104, "leak": 0.1351297739893198, "replacement": 0.3845246993005276} +{"step": 172, "steps": 250, "lr": 1.6512611692777008e-05, "grad_norm": 2.8158037662506104, "loss_scale": 4.0, "retries": 0, "seconds": 146.0, "max_memory_gb": 18.18, "pair": 0.15555747039616108, "sign": 0.003869555686833337, "keep": 0.014075430401135236, "leak": 0.7346529494971037, "replacement": 0.3481932058930397} +{"step": 173, "steps": 250, "lr": 1.6250000000000005e-05, "grad_norm": 2.472231864929199, "loss_scale": 4.0, "retries": 0, "seconds": 146.7, "max_memory_gb": 18.18, "pair": 0.20035730302333832, "sign": 0.10940276779001579, "keep": 0.13495339214568958, "leak": 0.794958895072341, "replacement": 0.5212751254439354} +{"step": 174, "steps": 250, "lr": 1.598941605544678e-05, "grad_norm": 2.1909966468811035, "loss_scale": 4.0, "retries": 0, "seconds": 147.5, "max_memory_gb": 18.18, "pair": 0.21330368798226118, "sign": 0.019692130270414054, "keep": 0.046572801132612085, "leak": 0.4052060395479202, "replacement": 0.3477376662194729} +{"step": 175, "steps": 250, "lr": 1.5730906827884506e-05, "grad_norm": 3.252307891845703, "loss_scale": 4.0, "retries": 0, "seconds": 148.3, "max_memory_gb": 18.18, "pair": 0.35323049407452345, "sign": 0.2679137645754963, "keep": 0.5408147778362036, "leak": 0.594284784514457, "replacement": 0.9061659760773182} +{"step": 176, "steps": 250, "lr": 1.5474518912124476e-05, "grad_norm": 2.002955913543701, "loss_scale": 4.0, "retries": 0, "seconds": 149.1, "max_memory_gb": 18.18, "pair": 0.2187667414546013, "sign": 0.03912160824984312, "keep": 0.005688336677849293, "leak": 0.22352596018754411, "replacement": 0.2970532029867172} +{"step": 177, "steps": 250, "lr": 1.5220298520623646e-05, "grad_norm": 2.4322431087493896, "loss_scale": 4.0, "retries": 0, "seconds": 149.9, "max_memory_gb": 18.18, "pair": 0.21616630628705025, "sign": 0.23012778064003214, "keep": 0.012645627137999327, "leak": 0.1513475924730301, "replacement": 0.37538990192115307} +{"step": 178, "steps": 250, "lr": 1.4968291475155121e-05, "grad_norm": 3.0299160480499268, "loss_scale": 4.0, "retries": 0, "seconds": 150.7, "max_memory_gb": 18.18, "pair": 0.13580204197205603, "sign": 0.042864055023528636, "keep": 0.5735547535150545, "leak": 0.4016024237498641, "replacement": 0.5444120354950428} +{"step": 179, "steps": 250, "lr": 1.4718543198548998e-05, "grad_norm": 2.2303996086120605, "loss_scale": 4.0, "retries": 0, "seconds": 151.5, "max_memory_gb": 18.18, "pair": 0.557897019200027, "sign": 0.5222670333751012, "keep": 0.6377332202384878, "leak": 0.18652831949293613, "replacement": 1.1845292411744595} +{"step": 180, "steps": 250, "lr": 1.4471098706505227e-05, "grad_norm": 2.1414177417755127, "loss_scale": 4.0, "retries": 0, "seconds": 152.3, "max_memory_gb": 18.18, "pair": 0.24044358357787132, "sign": 0.032088202075101435, "keep": 0.0329196272068657, "leak": 0.7603357285261154, "replacement": 0.46303143352270126} +{"step": 181, "steps": 250, "lr": 1.4226002599479743e-05, "grad_norm": 3.3565833568573, "loss_scale": 4.0, "retries": 0, "seconds": 153.1, "max_memory_gb": 18.18, "pair": 0.12723097018897533, "sign": 0.012325952528044581, "keep": 0.00878546986496076, "leak": 1.081022396683693, "replacement": 0.4080422893166542} +{"step": 182, "steps": 250, "lr": 1.3983299054645471e-05, "grad_norm": 10.769333839416504, "loss_scale": 4.0, "retries": 0, "seconds": 153.9, "max_memory_gb": 18.18, "pair": 0.28145560063421726, "sign": 0.3383301724679768, "keep": 0.008402600181739217, "leak": 3.3310456573963165, "replacement": 1.2875834312289953} +{"step": 183, "steps": 250, "lr": 1.3743031817929724e-05, "grad_norm": 1.7816402912139893, "loss_scale": 4.0, "retries": 0, "seconds": 154.8, "max_memory_gb": 18.18, "pair": 0.18030390236526728, "sign": 0.036928278277628124, "keep": 0.003721714468298387, "leak": 0.20266653038561344, "replacement": 0.25129553116858006} +{"step": 184, "steps": 250, "lr": 1.3505244196129164e-05, "grad_norm": 2.100111961364746, "loss_scale": 4.0, "retries": 0, "seconds": 155.6, "max_memory_gb": 18.18, "pair": 0.11535651702433825, "sign": 0.0047874595038592815, "keep": 0.0025108720871003243, "leak": 0.5587501004338264, "replacement": 0.2586932089179754} +{"step": 185, "steps": 250, "lr": 1.3269979049104014e-05, "grad_norm": 1.9600849151611328, "loss_scale": 4.0, "retries": 0, "seconds": 156.5, "max_memory_gb": 18.18, "pair": 0.20709475502371788, "sign": 0.020118835207540542, "keep": 0.029319823574041948, "leak": 0.39913122821599245, "replacement": 0.33159690350294113} +{"step": 186, "steps": 250, "lr": 1.3037278782052863e-05, "grad_norm": 2.7473597526550293, "loss_scale": 4.0, "retries": 0, "seconds": 157.3, "max_memory_gb": 18.18, "pair": 0.12356745405122638, "sign": 0.016305036129779182, "keep": 0.4651391138274903, "leak": 0.4762858971953392, "replacement": 0.4833610076457262} +{"step": 187, "steps": 250, "lr": 1.2807185337869343e-05, "grad_norm": 2.2279815673828125, "loss_scale": 4.0, "retries": 0, "seconds": 158.2, "max_memory_gb": 18.18, "pair": 0.20695451274514198, "sign": 0.0077433460392057896, "keep": 0.2616185423103161, "leak": 0.3621305637061596, "replacement": 0.4321680888533592} +{"step": 188, "steps": 250, "lr": 1.2579740189582115e-05, "grad_norm": 2.556016683578491, "loss_scale": 4.0, "retries": 0, "seconds": 159.0, "max_memory_gb": 18.18, "pair": 0.4944359827786684, "sign": 0.7552588424878195, "keep": 0.9127053684526345, "leak": 0.3062024172395468, "replacement": 1.404968736693263} +{"step": 189, "steps": 250, "lr": 1.2354984332879668e-05, "grad_norm": 6.852113246917725, "loss_scale": 4.0, "retries": 0, "seconds": 159.8, "max_memory_gb": 18.18, "pair": 0.24500060826539993, "sign": 0.5271447809645906, "keep": 1.2275033220648766, "leak": 1.6602731235325336, "replacement": 1.537392944097519} +{"step": 190, "steps": 250, "lr": 1.2132958278721064e-05, "grad_norm": 3.9629480838775635, "loss_scale": 4.0, "retries": 0, "seconds": 160.6, "max_memory_gb": 18.18, "pair": 0.14033947698771954, "sign": 0.007611188921146095, "keep": 0.0042522816293057986, "leak": 1.3144909851253033, "replacement": 0.4748939387500286} +{"step": 191, "steps": 250, "lr": 1.1913702046034016e-05, "grad_norm": 11.156750679016113, "loss_scale": 4.0, "retries": 0, "seconds": 161.4, "max_memory_gb": 18.18, "pair": 0.487007075920701, "sign": 0.7670768317766488, "keep": 1.259741266081619, "leak": 1.4150164648890495, "replacement": 1.8541702665388584} +{"step": 192, "steps": 250, "lr": 1.1697255154501756e-05, "grad_norm": 3.0161473751068115, "loss_scale": 4.0, "retries": 0, "seconds": 162.2, "max_memory_gb": 18.18, "pair": 0.491601187735796, "sign": 0.5104104734491557, "keep": 1.357873015338555, "leak": 0.2068392401561141, "replacement": 1.4774526953697205} +{"step": 193, "steps": 250, "lr": 1.1483656617439842e-05, "grad_norm": 2.7697105407714844, "loss_scale": 4.0, "retries": 0, "seconds": 163.1, "max_memory_gb": 18.18, "pair": 0.19445645436644554, "sign": 0.022282492835074663, "keep": 0.006762624455518562, "leak": 1.1520768199115992, "replacement": 0.4969982113689184} +{"step": 194, "steps": 250, "lr": 1.1272944934764177e-05, "grad_norm": 2.692901849746704, "loss_scale": 4.0, "retries": 0, "seconds": 163.9, "max_memory_gb": 18.18, "pair": 0.34328411146998405, "sign": 0.2107921018032357, "keep": 0.03136169617783935, "leak": 0.34757617488503456, "replacement": 0.5512550547719002} +{"step": 195, "steps": 250, "lr": 1.1065158086051684e-05, "grad_norm": 2.3874166011810303, "loss_scale": 4.0, "retries": 0, "seconds": 164.8, "max_memory_gb": 18.18, "pair": 0.22023966908454895, "sign": 0.03831563546555117, "keep": 0.004173834982793778, "leak": 0.1861179880797863, "replacement": 0.28801390156149864} +{"step": 196, "steps": 250, "lr": 1.086033352369467e-05, "grad_norm": 6.870784282684326, "loss_scale": 4.0, "retries": 0, "seconds": 165.7, "max_memory_gb": 18.18, "pair": 0.15403037890791893, "sign": 0.013290757517097518, "keep": 0.0011530818299263501, "leak": 2.178958013653755, "replacement": 0.7059918120503426} +{"step": 197, "steps": 250, "lr": 1.0658508166150224e-05, "grad_norm": 8.837409019470215, "loss_scale": 4.0, "retries": 0, "seconds": 166.6, "max_memory_gb": 18.18, "pair": 0.3216134086251259, "sign": 0.04649335704743862, "keep": 0.008138410616084002, "leak": 1.401717712637037, "replacement": 0.6993587166070938} +{"step": 198, "steps": 250, "lr": 1.0459718391285954e-05, "grad_norm": 2.2862699031829834, "loss_scale": 4.0, "retries": 0, "seconds": 167.4, "max_memory_gb": 18.18, "pair": 0.2342876186594367, "sign": 0.07094873092137277, "keep": 0.0960835719415627, "leak": 0.6859811930917203, "replacement": 0.48929907381534576} +{"step": 199, "steps": 250, "lr": 1.0264000029822999e-05, "grad_norm": 3.2391159534454346, "loss_scale": 4.0, "retries": 0, "seconds": 168.2, "max_memory_gb": 18.18, "pair": 0.168123047798872, "sign": 0.030113778106169775, "keep": 0.3084374592581298, "leak": 0.6829484855406918, "replacement": 0.5081357974559069} +{"step": 200, "steps": 250, "lr": 1.0071388358877775e-05, "grad_norm": 2.4787044525146484, "loss_scale": 4.0, "retries": 0, "seconds": 169.1, "max_memory_gb": 18.18, "pair": 0.35600719414651394, "sign": 0.3492257973412052, "keep": 1.2801849699171726, "leak": 0.46847565472126007, "replacement": 1.287831474095583} +{"step": 201, "steps": 250, "lr": 9.881918095603501e-06, "grad_norm": 1.7436587810516357, "loss_scale": 4.0, "retries": 0, "seconds": 170.0, "max_memory_gb": 18.18, "pair": 0.22126301191747189, "sign": 0.19460208271630108, "keep": 0.16141706419875845, "leak": 0.22791407629847527, "replacement": 0.4562510959804058} +{"step": 202, "steps": 250, "lr": 9.695623390932612e-06, "grad_norm": 9.585650444030762, "loss_scale": 4.0, "retries": 0, "seconds": 170.8, "max_memory_gb": 18.18, "pair": 0.20992597518488765, "sign": 0.0669688849884551, "keep": 0.001193563279088039, "leak": 3.831918343901634, "replacement": 1.2019868046045303} +{"step": 203, "steps": 250, "lr": 9.51253782342123e-06, "grad_norm": 2.2881438732147217, "loss_scale": 4.0, "retries": 0, "seconds": 171.6, "max_memory_gb": 18.18, "pair": 0.09546014945954084, "sign": 0.030164555821102113, "keep": 0.08858729609346483, "leak": 0.4950090767815709, "replacement": 0.2785883489996195} +{"step": 204, "steps": 250, "lr": 9.332694393196846e-06, "grad_norm": 2.561234951019287, "loss_scale": 4.0, "retries": 0, "seconds": 172.4, "max_memory_gb": 18.18, "pair": 0.2591841593384743, "sign": 0.016470822622068226, "keep": 0.02849751489702612, "leak": 0.6062419321388006, "replacement": 0.4332288056612015} +{"step": 205, "steps": 250, "lr": 9.156125516010256e-06, "grad_norm": 3.066831350326538, "loss_scale": 4.0, "retries": 0, "seconds": 173.3, "max_memory_gb": 18.18, "pair": 0.08286705613136292, "sign": 0.005111405334901065, "keep": 0.000749840633943677, "leak": 0.5292644549426768, "replacement": 0.2181137902662158} +{"step": 206, "steps": 250, "lr": 8.982863017392733e-06, "grad_norm": 2.2276692390441895, "loss_scale": 4.0, "retries": 0, "seconds": 174.1, "max_memory_gb": 18.18, "pair": 0.22356140986084938, "sign": 0.025644710927736014, "keep": 0.009465575130874981, "leak": 0.6731955148279667, "replacement": 0.40941542387008667} +{"step": 207, "steps": 250, "lr": 8.81293812691974e-06, "grad_norm": 2.894962787628174, "loss_scale": 4.0, "retries": 0, "seconds": 174.9, "max_memory_gb": 18.18, "pair": 0.21017013676464558, "sign": 0.14214240352157503, "keep": 0.07784321544500017, "leak": 1.1422465997748077, "replacement": 0.6057245768606663} +{"step": 208, "steps": 250, "lr": 8.646381472581927e-06, "grad_norm": 2.649397850036621, "loss_scale": 4.0, "retries": 0, "seconds": 175.7, "max_memory_gb": 18.18, "pair": 0.1301589161157608, "sign": 0.004127142863580957, "keep": 0.0033641502959511627, "leak": 0.7463081830646843, "replacement": 0.32048161141574383} +{"step": 209, "steps": 250, "lr": 8.483223075264622e-06, "grad_norm": 3.356767177581787, "loss_scale": 4.0, "retries": 0, "seconds": 176.5, "max_memory_gb": 18.18, "pair": 0.4572830740362406, "sign": 0.4309210693463683, "keep": 0.41676610172726214, "leak": 0.1909544626250863, "replacement": 0.928865285590291} +{"step": 210, "steps": 250, "lr": 8.323492343336747e-06, "grad_norm": 12.708694458007812, "loss_scale": 4.0, "retries": 0, "seconds": 177.4, "max_memory_gb": 18.18, "pair": 0.35113855078816414, "sign": 0.5794731543282978, "keep": 1.9579854001058257, "leak": 4.555481296032667, "replacement": 2.758738212287426} +{"step": 211, "steps": 250, "lr": 8.167218067350122e-06, "grad_norm": 2.8520302772521973, "loss_scale": 4.0, "retries": 0, "seconds": 178.2, "max_memory_gb": 18.18, "pair": 0.36459522228688, "sign": 0.26243545435136184, "keep": 0.11882463716028724, "leak": 0.8350158729590476, "replacement": 0.7639792403206229} +{"step": 212, "steps": 250, "lr": 8.01442841485013e-06, "grad_norm": 3.9818148612976074, "loss_scale": 4.0, "retries": 0, "seconds": 179.1, "max_memory_gb": 18.18, "pair": 0.2550771590322256, "sign": 0.13839862355962396, "keep": 0.033531850727740675, "leak": 0.9925473248586059, "replacement": 0.5891792215406895} +{"step": 213, "steps": 250, "lr": 7.865150925298678e-06, "grad_norm": 2.5384721755981445, "loss_scale": 4.0, "retries": 0, "seconds": 179.9, "max_memory_gb": 18.18, "pair": 0.36553802993148565, "sign": 0.35978087509283796, "keep": 0.5737828908822848, "leak": 0.33909437706461176, "replacement": 0.9170934986323118} +{"step": 214, "steps": 250, "lr": 7.719412505110375e-06, "grad_norm": 4.724308013916016, "loss_scale": 4.0, "retries": 0, "seconds": 180.8, "max_memory_gb": 18.18, "pair": 0.38725289329886436, "sign": 0.408468130233814, "keep": 1.5734311086706043, "leak": 0.04024737328290939, "replacement": 1.3882643468677998} +{"step": 215, "steps": 250, "lr": 7.57723942280278e-06, "grad_norm": 5.290094375610352, "loss_scale": 4.0, "retries": 0, "seconds": 181.6, "max_memory_gb": 18.18, "pair": 0.5333218667656183, "sign": 0.9165115493815392, "keep": 2.0252569987642346, "leak": 0.4731875613797456, "replacement": 2.1225031167268753} +{"step": 216, "steps": 250, "lr": 7.438657304261687e-06, "grad_norm": 2.2724556922912598, "loss_scale": 4.0, "retries": 0, "seconds": 182.4, "max_memory_gb": 18.18, "pair": 0.07499681133776903, "sign": 0.015620205842424184, "keep": 0.004018888001155574, "leak": 0.823263525031507, "replacement": 0.2906322479248047} +{"step": 217, "steps": 250, "lr": 7.30369112812219e-06, "grad_norm": 1.9056605100631714, "loss_scale": 4.0, "retries": 0, "seconds": 183.3, "max_memory_gb": 18.18, "pair": 0.1614860654808581, "sign": 0.008334160083904862, "keep": 0.0007525190330852638, "leak": 0.23462182097136974, "replacement": 0.22468486055731773} +{"step": 218, "steps": 250, "lr": 7.172365221266403e-06, "grad_norm": 3.5615596771240234, "loss_scale": 4.0, "retries": 0, "seconds": 184.1, "max_memory_gb": 18.18, "pair": 0.4355899393558502, "sign": 1.072705451399088, "keep": 0.172620318189729, "leak": 0.26775142597034574, "replacement": 1.1251907348632812} +{"step": 219, "steps": 250, "lr": 7.0447032544387154e-06, "grad_norm": 1.4854533672332764, "loss_scale": 4.0, "retries": 0, "seconds": 185.0, "max_memory_gb": 18.18, "pair": 0.13132630661129951, "sign": 0.06324507447425276, "keep": 0.008062163189606508, "leak": 0.15573555486116675, "replacement": 0.2059138175100088} +{"step": 220, "steps": 250, "lr": 6.920728237979274e-06, "grad_norm": 2.7956244945526123, "loss_scale": 4.0, "retries": 0, "seconds": 185.8, "max_memory_gb": 18.18, "pair": 0.15082073491066694, "sign": 0.025188116065692157, "keep": 0.0446670988645792, "leak": 0.12689710035920143, "replacement": 0.21747261472046375} +{"step": 221, "steps": 250, "lr": 6.800462517676456e-06, "grad_norm": 3.8738155364990234, "loss_scale": 4.0, "retries": 0, "seconds": 186.6, "max_memory_gb": 18.18, "pair": 0.20359659753739834, "sign": 0.12989134021336213, "keep": 0.007619478681590408, "leak": 1.0848481729626656, "replacement": 0.5435640439391136} +{"step": 222, "steps": 250, "lr": 6.683927770739222e-06, "grad_norm": 1.9058640003204346, "loss_scale": 4.0, "retries": 0, "seconds": 187.5, "max_memory_gb": 18.18, "pair": 0.16219262406229973, "sign": 0.011757121479604393, "keep": 0.017981390814270526, "leak": 0.3373788297176361, "replacement": 0.2614065855741501} +{"step": 223, "steps": 250, "lr": 6.571145001889892e-06, "grad_norm": 8.106989860534668, "loss_scale": 4.0, "retries": 0, "seconds": 188.3, "max_memory_gb": 18.18, "pair": 0.4318421110510826, "sign": 0.10608736865106039, "keep": 0.045414152338707936, "leak": 0.6058220881968737, "replacement": 0.6590483952313662} +{"step": 224, "steps": 250, "lr": 6.46213453957817e-06, "grad_norm": 2.03922176361084, "loss_scale": 4.0, "retries": 0, "seconds": 189.1, "max_memory_gb": 18.18, "pair": 0.14955138973891735, "sign": 0.02474209212232381, "keep": 0.003651249047379679, "leak": 0.2735882700289949, "replacement": 0.23214512690901756} +{"step": 225, "steps": 250, "lr": 6.356916032317063e-06, "grad_norm": 2.8207852840423584, "loss_scale": 4.0, "retries": 0, "seconds": 190.0, "max_memory_gb": 18.18, "pair": 0.15382160805165768, "sign": 0.018330892315134406, "keep": 0.0036564295878633857, "leak": 1.397035711677745, "replacement": 0.5140742138028145} +{"step": 226, "steps": 250, "lr": 6.255508445141349e-06, "grad_norm": 3.873556613922119, "loss_scale": 4.0, "retries": 0, "seconds": 190.8, "max_memory_gb": 18.18, "pair": 0.13171869050711393, "sign": 0.03805404179729521, "keep": 0.8901553412142675, "leak": 1.1481649354100227, "replacement": 0.8828646019101143} +{"step": 227, "steps": 250, "lr": 6.157930056189228e-06, "grad_norm": 2.100571870803833, "loss_scale": 4.0, "retries": 0, "seconds": 191.6, "max_memory_gb": 18.18, "pair": 0.19812155421823263, "sign": 0.009568508889060467, "keep": 0.04725418755560895, "leak": 0.06999612340587191, "replacement": 0.24403193406760693} +{"step": 228, "steps": 250, "lr": 6.064198453407796e-06, "grad_norm": 3.284339427947998, "loss_scale": 4.0, "retries": 0, "seconds": 192.4, "max_memory_gb": 18.18, "pair": 0.08626778237521648, "sign": 0.035680862434674054, "keep": 0.001176396324808593, "leak": 0.8624542058096267, "replacement": 0.32030997052788734} +{"step": 229, "steps": 250, "lr": 5.974330531382935e-06, "grad_norm": 2.7185685634613037, "loss_scale": 4.0, "retries": 0, "seconds": 193.3, "max_memory_gb": 18.18, "pair": 0.1394540909677744, "sign": 0.07538883987581357, "keep": 0.0481115979487754, "leak": 0.5690707638859749, "replacement": 0.3434720039367676} +{"step": 230, "steps": 250, "lr": 5.888342488294125e-06, "grad_norm": 15.95473575592041, "loss_scale": 4.0, "retries": 0, "seconds": 194.1, "max_memory_gb": 18.18, "pair": 0.9216250404715538, "sign": 0.6827229782938957, "keep": 1.4094787407647118, "leak": 1.821876692585647, "replacement": 2.4231951534748077} +{"step": 231, "steps": 250, "lr": 5.806249822994849e-06, "grad_norm": 2.308171510696411, "loss_scale": 4.0, "retries": 0, "seconds": 194.9, "max_memory_gb": 18.18, "pair": 0.3453449234366417, "sign": 0.5429016666021198, "keep": 0.017939460521986916, "leak": 0.1900415327399969, "replacement": 0.6732758730649948} +{"step": 232, "steps": 250, "lr": 5.728067332218994e-06, "grad_norm": 4.518896102905273, "loss_scale": 4.0, "retries": 0, "seconds": 195.7, "max_memory_gb": 18.18, "pair": 0.23412730544805527, "sign": 0.02255940210307017, "keep": 0.011612977017648518, "leak": 1.963917225599289, "replacement": 0.7421927973628044} +{"step": 233, "steps": 250, "lr": 5.65380910791383e-06, "grad_norm": 2.838899850845337, "loss_scale": 4.0, "retries": 0, "seconds": 196.5, "max_memory_gb": 18.18, "pair": 0.15919001679867506, "sign": 0.00610144657548517, "keep": 0.002822516136802733, "leak": 0.9147790484130383, "replacement": 0.3923467546701431} +{"step": 234, "steps": 250, "lr": 5.583488534700022e-06, "grad_norm": 3.345038652420044, "loss_scale": 4.0, "retries": 0, "seconds": 197.4, "max_memory_gb": 18.18, "pair": 0.054736068937927485, "sign": 0.011453405895736068, "keep": 0.005153888260338135, "leak": 1.191459703084547, "replacement": 0.36090464517474174} +{"step": 235, "steps": 250, "lr": 5.517118287459118e-06, "grad_norm": 9.870991706848145, "loss_scale": 4.0, "retries": 0, "seconds": 198.2, "max_memory_gb": 18.18, "pair": 0.7340509025380015, "sign": 0.48768236630712636, "keep": 1.0446061840557377, "leak": 0.9078043196477665, "replacement": 1.7271463079378009} +{"step": 236, "steps": 250, "lr": 5.454710329048978e-06, "grad_norm": 1.792175054550171, "loss_scale": 4.0, "retries": 0, "seconds": 199.0, "max_memory_gb": 18.18, "pair": 0.2022514594718814, "sign": 0.06251722076558508, "keep": 0.004020740665509948, "leak": 0.15766596648609266, "replacement": 0.2749369200319052} +{"step": 237, "steps": 250, "lr": 5.3962759081475436e-06, "grad_norm": 3.2495474815368652, "loss_scale": 4.0, "retries": 0, "seconds": 199.8, "max_memory_gb": 18.18, "pair": 0.43394751101732254, "sign": 0.3417001748457551, "keep": 0.10545535013076801, "leak": 0.4857258531264961, "replacement": 0.7789566963911057} +{"step": 238, "steps": 250, "lr": 5.341825557225321e-06, "grad_norm": 2.2462449073791504, "loss_scale": 4.0, "retries": 0, "seconds": 200.5, "max_memory_gb": 18.18, "pair": 0.1641683727502823, "sign": 0.01832158665638417, "keep": 0.040335924160899594, "leak": 0.5803040415048599, "replacement": 0.3385731428861618} +{"step": 239, "steps": 250, "lr": 5.291369090646961e-06, "grad_norm": 4.637610912322998, "loss_scale": 4.0, "retries": 0, "seconds": 201.3, "max_memory_gb": 18.18, "pair": 0.3533824756741524, "sign": 1.0710281557403505, "keep": 0.15605465299358912, "leak": 0.5965206041000783, "replacement": 1.1160540021955967} +{"step": 240, "steps": 250, "lr": 5.244915602902289e-06, "grad_norm": 2.190053701400757, "loss_scale": 4.0, "retries": 0, "seconds": 202.2, "max_memory_gb": 18.18, "pair": 0.28526263358071446, "sign": 0.28692609269637614, "keep": 0.08026548768248176, "leak": 0.2308136047795415, "replacement": 0.5265618469566107} +{"step": 241, "steps": 250, "lr": 5.20247346696706e-06, "grad_norm": 5.183825969696045, "loss_scale": 4.0, "retries": 0, "seconds": 203.0, "max_memory_gb": 18.18, "pair": 0.5315504632890224, "sign": 0.47180354614101816, "keep": 0.2881506886187424, "leak": 0.3255191554217731, "replacement": 0.9929073825478554} +{"step": 242, "steps": 250, "lr": 5.164050332793787e-06, "grad_norm": 3.4826138019561768, "loss_scale": 4.0, "retries": 0, "seconds": 203.9, "max_memory_gb": 18.18, "pair": 0.13972512260079384, "sign": 0.027274301159195602, "keep": 0.006482921053247992, "leak": 1.4536148626357317, "replacement": 0.5200074333697557} +{"step": 243, "steps": 250, "lr": 5.129653125932873e-06, "grad_norm": 1.9780852794647217, "loss_scale": 4.0, "retries": 0, "seconds": 204.7, "max_memory_gb": 18.18, "pair": 0.1719268667511642, "sign": 0.015672725276090205, "keep": 0.007647274498594925, "leak": 0.3803628347814083, "replacement": 0.2786775715649128} +{"step": 244, "steps": 250, "lr": 5.099288046284329e-06, "grad_norm": 3.2221691608428955, "loss_scale": 4.0, "retries": 0, "seconds": 205.5, "max_memory_gb": 18.18, "pair": 0.23940463922917843, "sign": 0.14083606516942382, "keep": 0.00755217632104177, "leak": 0.8863602764904499, "replacement": 0.535188838839531} +{"step": 245, "steps": 250, "lr": 5.072960566980277e-06, "grad_norm": 4.35159969329834, "loss_scale": 4.0, "retries": 0, "seconds": 206.4, "max_memory_gb": 18.18, "pair": 0.5930556952953339, "sign": 0.3933242610655725, "keep": 0.8816566129680723, "leak": 0.767126775579527, "replacement": 1.422327846288681} +{"step": 246, "steps": 250, "lr": 5.050675433398435e-06, "grad_norm": 1.3895236253738403, "loss_scale": 4.0, "retries": 0, "seconds": 207.2, "max_memory_gb": 18.18, "pair": 0.16371620446443558, "sign": 0.0046841808361932635, "keep": 0.0009013429637434456, "leak": 0.2625413239002228, "replacement": 0.2321442998945713} +{"step": 247, "steps": 250, "lr": 5.032436662306811e-06, "grad_norm": 3.9528446197509766, "loss_scale": 4.0, "retries": 0, "seconds": 208.1, "max_memory_gb": 18.18, "pair": 0.1245504254475236, "sign": 0.011365329177351668, "keep": 0.00745488033862074, "leak": 1.1395485643297434, "replacement": 0.41884766705334187} +{"step": 248, "steps": 250, "lr": 5.018247541139684e-06, "grad_norm": 2.4687368869781494, "loss_scale": 4.0, "retries": 0, "seconds": 208.9, "max_memory_gb": 18.18, "pair": 0.16064042411744595, "sign": 0.029055493127088994, "keep": 0.030237113021939876, "leak": 0.7155264373868704, "replacement": 0.36916833743453026} +{"step": 249, "steps": 250, "lr": 5.00811062740508e-06, "grad_norm": 3.2205522060394287, "loss_scale": 4.0, "retries": 0, "seconds": 209.7, "max_memory_gb": 18.18, "pair": 0.3801560718566179, "sign": 0.1223823232576251, "keep": 0.04856686055427417, "leak": 0.23980306647717953, "replacement": 0.5255814157426357} +{"step": 250, "steps": 250, "lr": 5.002027748223778e-06, "grad_norm": 2.503981351852417, "loss_scale": 4.0, "retries": 0, "seconds": 210.5, "max_memory_gb": 18.18, "pair": 0.15866538509726524, "sign": 0.01934841403272003, "keep": 0.030124984736971783, "leak": 0.9876317754387856, "replacement": 0.43031002953648567} diff --git a/runs/single_llava/removal/L01_s13/adapter/README.md b/runs/single_llava/removal/L01_s13/adapter/README.md new file mode 100644 index 0000000000000000000000000000000000000000..ae587920ca55972d94e38f783837f71be14f1dda --- /dev/null +++ b/runs/single_llava/removal/L01_s13/adapter/README.md @@ -0,0 +1,207 @@ +--- +base_model: /data/OPSD/assets/models/llava-1.5-7b-hf +library_name: peft +pipeline_tag: text-generation +tags: +- base_model:adapter:/data/OPSD/assets/models/llava-1.5-7b-hf +- lora +- transformers +--- + +# Model Card for Model ID + + + + + +## Model Details + +### Model Description + + + + + +- **Developed by:** [More Information Needed] +- **Funded by [optional]:** [More Information Needed] +- **Shared by [optional]:** [More Information Needed] +- **Model type:** [More Information Needed] +- **Language(s) (NLP):** [More Information Needed] +- **License:** [More Information Needed] +- **Finetuned from model [optional]:** [More Information Needed] + +### Model Sources [optional] + + + +- **Repository:** [More Information Needed] +- **Paper [optional]:** [More Information Needed] +- **Demo [optional]:** [More Information Needed] + +## Uses + + + +### Direct Use + + + +[More Information Needed] + +### Downstream Use [optional] + + + +[More Information Needed] + +### Out-of-Scope Use + + + +[More Information Needed] + +## Bias, Risks, and Limitations + + + +[More Information Needed] + +### Recommendations + + + +Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. More information needed for further recommendations. + +## How to Get Started with the Model + +Use the code below to get started with the model. + +[More Information Needed] + +## Training Details + +### Training Data + + + +[More Information Needed] + +### Training Procedure + + + +#### Preprocessing [optional] + +[More Information Needed] + + +#### Training Hyperparameters + +- **Training regime:** [More Information Needed] + +#### Speeds, Sizes, Times [optional] + + + +[More Information Needed] + +## Evaluation + + + +### Testing Data, Factors & Metrics + +#### Testing Data + + + +[More Information Needed] + +#### Factors + + + +[More Information Needed] + +#### Metrics + + + +[More Information Needed] + +### Results + +[More Information Needed] + +#### Summary + + + +## Model Examination [optional] + + + +[More Information Needed] + +## Environmental Impact + + + +Carbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700). + +- **Hardware Type:** [More Information Needed] +- **Hours used:** [More Information Needed] +- **Cloud Provider:** [More Information Needed] +- **Compute Region:** [More Information Needed] +- **Carbon Emitted:** [More Information Needed] + +## Technical Specifications [optional] + +### Model Architecture and Objective + +[More Information Needed] + +### Compute Infrastructure + +[More Information Needed] + +#### Hardware + +[More Information Needed] + +#### Software + +[More Information Needed] + +## Citation [optional] + + + +**BibTeX:** + +[More Information Needed] + +**APA:** + +[More Information Needed] + +## Glossary [optional] + + + +[More Information Needed] + +## More Information [optional] + +[More Information Needed] + +## Model Card Authors [optional] + +[More Information Needed] + +## Model Card Contact + +[More Information Needed] +### Framework versions + +- PEFT 0.20.0 \ No newline at end of file diff --git a/runs/single_llava/removal/L01_s13/adapter/adapter_config.json b/runs/single_llava/removal/L01_s13/adapter/adapter_config.json new file mode 100644 index 0000000000000000000000000000000000000000..7fe8ea2a47be778a28c0d04a3c9f659a9fa8de02 --- /dev/null +++ b/runs/single_llava/removal/L01_s13/adapter/adapter_config.json @@ -0,0 +1,50 @@ +{ + "alora_invocation_tokens": null, + "alpha_pattern": {}, + "arrow_config": null, + "auto_mapping": null, + "base_model_name_or_path": "/data/OPSD/assets/models/llava-1.5-7b-hf", + "bias": "none", + "corda_config": null, + "ensure_weight_tying": false, + "eva_config": null, + "exclude_modules": null, + "fan_in_fan_out": false, + "inference_mode": true, + "init_lora_weights": true, + "layer_replication": null, + "layers_pattern": null, + "layers_to_transform": null, + "loftq_config": {}, + "lora_alpha": 32, + "lora_bias": false, + "lora_dropout": 0.05, + "lora_ga_config": null, + "megatron_config": null, + "megatron_core": "megatron.core", + "modules_to_save": null, + "monteclora_config": null, + "peft_type": "LORA", + "peft_version": "0.20.0", + "qalora_group_size": 16, + "r": 16, + "rank_pattern": {}, + "revision": null, + "target_modules": [ + "model.language_model.layers.1.self_attn.q_proj", + "model.language_model.layers.1.self_attn.v_proj", + "model.language_model.layers.1.self_attn.o_proj", + "model.language_model.layers.1.self_attn.k_proj", + "model.language_model.layers.1.mlp.gate_proj", + "model.language_model.layers.1.mlp.down_proj", + "model.language_model.layers.1.mlp.up_proj" + ], + "target_parameters": null, + "task_type": "CAUSAL_LM", + "trainable_token_indices": null, + "use_bdlora": null, + "use_dora": false, + "use_qalora": false, + "use_rslora": false, + "velora_config": null +} \ No newline at end of file diff --git a/runs/single_llava/removal/L01_s13/config.json b/runs/single_llava/removal/L01_s13/config.json new file mode 100644 index 0000000000000000000000000000000000000000..35f2d41389a12b26f963c6c360b5768ed7249e36 --- /dev/null +++ b/runs/single_llava/removal/L01_s13/config.json @@ -0,0 +1,24 @@ +{ + "behavior": "removal", + "seed": 13, + "model": "/data/OPSD/assets/models/llava-1.5-7b-hf", + "backbone": "llava", + "train": "/data/OPSD/assets/v3/removal.jsonl", + "items": 1000, + "steps": 500, + "epochs": 2, + "steps_per_epoch": 250, + "groups_per_step": 4, + "lr": 2e-05, + "scheduler": "constant", + "warmup_steps": 0, + "final_fraction": 1.0, + "weight_decay": 0.0, + "clip": 1.0, + "initial_loss_scale": 4.0, + "adapter": { + "layer": 1, + "rank": 16, + "trainable_parameters": 1249280 + } +} \ No newline at end of file diff --git a/runs/single_llava/removal/L01_s13/steps.jsonl b/runs/single_llava/removal/L01_s13/steps.jsonl new file mode 100644 index 0000000000000000000000000000000000000000..ca8ee185783c3e452cb5146cee3db9ffa61ead56 --- /dev/null +++ b/runs/single_llava/removal/L01_s13/steps.jsonl @@ -0,0 +1,500 @@ +{"step": 1, "steps": 500, "lr": 2e-05, "grad_norm": 0.5907116532325745, "loss_scale": 4.0, "retries": 0, "seconds": 3.4, "max_memory_gb": 18.27, "ce": 1.6287612388841808, "rank": 0.13243506848812103} +{"step": 2, "steps": 500, "lr": 2e-05, "grad_norm": 0.9236217141151428, "loss_scale": 4.0, "retries": 0, "seconds": 5.4, "max_memory_gb": 18.29, "ce": 2.6737076211720705, "rank": 0.033110618591308594} +{"step": 3, "steps": 500, "lr": 2e-05, "grad_norm": 2.0792934894561768, "loss_scale": 4.0, "retries": 0, "seconds": 7.3, "max_memory_gb": 18.29, "ce": 5.265624014660716, "rank": 0.14771413803100586} +{"step": 4, "steps": 500, "lr": 2e-05, "grad_norm": 1.305083990097046, "loss_scale": 4.0, "retries": 0, "seconds": 9.2, "max_memory_gb": 18.29, "ce": 3.792689300724305, "rank": 0.1235727071762085} +{"step": 5, "steps": 500, "lr": 2e-05, "grad_norm": 0.16869200766086578, "loss_scale": 4.0, "retries": 0, "seconds": 11.1, "max_memory_gb": 18.29, "ce": 0.13963695825077593, "rank": 0.0} +{"step": 6, "steps": 500, "lr": 2e-05, "grad_norm": 0.6338127255439758, "loss_scale": 4.0, "retries": 0, "seconds": 12.9, "max_memory_gb": 18.29, "ce": 1.5061200023628771, "rank": 0.0} +{"step": 7, "steps": 500, "lr": 2e-05, "grad_norm": 1.5288686752319336, "loss_scale": 4.0, "retries": 0, "seconds": 14.8, "max_memory_gb": 18.29, "ce": 3.4382379532326013, "rank": 0.020269274711608887} +{"step": 8, "steps": 500, "lr": 2e-05, "grad_norm": 2.0091044902801514, "loss_scale": 4.0, "retries": 0, "seconds": 16.7, "max_memory_gb": 18.29, "ce": 3.6769178742542863, "rank": 0.19329845905303955} +{"step": 9, "steps": 500, "lr": 2e-05, "grad_norm": 1.4587239027023315, "loss_scale": 4.0, "retries": 0, "seconds": 18.5, "max_memory_gb": 18.29, "ce": 2.8115794053301215, "rank": 0.037418246269226074} +{"step": 10, "steps": 500, "lr": 2e-05, "grad_norm": 0.7088631987571716, "loss_scale": 4.0, "retries": 0, "seconds": 20.4, "max_memory_gb": 18.3, "ce": 1.6082657109946012, "rank": 0.0} +{"step": 11, "steps": 500, "lr": 2e-05, "grad_norm": 1.9194141626358032, "loss_scale": 4.0, "retries": 0, "seconds": 22.2, "max_memory_gb": 18.3, "ce": 2.5886380234733224, "rank": 0.20417650043964386} +{"step": 12, "steps": 500, "lr": 2e-05, "grad_norm": 1.1685280799865723, "loss_scale": 4.0, "retries": 0, "seconds": 24.1, "max_memory_gb": 18.3, "ce": 1.4920770335011184, "rank": 0.03692519664764404} +{"step": 13, "steps": 500, "lr": 2e-05, "grad_norm": 1.7867581844329834, "loss_scale": 4.0, "retries": 0, "seconds": 25.9, "max_memory_gb": 18.3, "ce": 2.6353311522398144, "rank": 0.0} +{"step": 14, "steps": 500, "lr": 2e-05, "grad_norm": 2.5501890182495117, "loss_scale": 4.0, "retries": 0, "seconds": 27.8, "max_memory_gb": 18.3, "ce": 3.815641699358821, "rank": 0.14619731903076172} +{"step": 15, "steps": 500, "lr": 2e-05, "grad_norm": 2.063767433166504, "loss_scale": 4.0, "retries": 0, "seconds": 29.6, "max_memory_gb": 18.3, "ce": 2.5525243625743315, "rank": 0.17034339904785156} +{"step": 16, "steps": 500, "lr": 2e-05, "grad_norm": 1.2227256298065186, "loss_scale": 4.0, "retries": 0, "seconds": 31.5, "max_memory_gb": 18.3, "ce": 1.3910900448681787, "rank": 0.0} +{"step": 17, "steps": 500, "lr": 2e-05, "grad_norm": 1.388386845588684, "loss_scale": 4.0, "retries": 0, "seconds": 33.3, "max_memory_gb": 18.3, "ce": 1.5640577669255435, "rank": 0.03625047206878662} +{"step": 18, "steps": 500, "lr": 2e-05, "grad_norm": 3.1044139862060547, "loss_scale": 4.0, "retries": 0, "seconds": 35.2, "max_memory_gb": 18.31, "ce": 3.5192906400188804, "rank": 0.19143366813659668} +{"step": 19, "steps": 500, "lr": 2e-05, "grad_norm": 4.962343692779541, "loss_scale": 4.0, "retries": 0, "seconds": 37.1, "max_memory_gb": 18.31, "ce": 5.15336842648685, "rank": 0.2927870750427246} +{"step": 20, "steps": 500, "lr": 2e-05, "grad_norm": 0.07561888545751572, "loss_scale": 4.0, "retries": 0, "seconds": 38.9, "max_memory_gb": 18.31, "ce": 0.06515129143372178, "rank": 0.10957713425159454} +{"step": 21, "steps": 500, "lr": 2e-05, "grad_norm": 3.0470468997955322, "loss_scale": 4.0, "retries": 0, "seconds": 40.8, "max_memory_gb": 18.31, "ce": 2.6647789638955146, "rank": 0.14167630672454834} +{"step": 22, "steps": 500, "lr": 2e-05, "grad_norm": 2.431511163711548, "loss_scale": 4.0, "retries": 0, "seconds": 42.7, "max_memory_gb": 18.31, "ce": 2.8754324093461037, "rank": 0.18636083602905273} +{"step": 23, "steps": 500, "lr": 2e-05, "grad_norm": 3.0678305625915527, "loss_scale": 4.0, "retries": 0, "seconds": 44.6, "max_memory_gb": 18.31, "ce": 2.6678539105923846, "rank": 0.11187744140625} +{"step": 24, "steps": 500, "lr": 2e-05, "grad_norm": 4.57044792175293, "loss_scale": 4.0, "retries": 0, "seconds": 46.4, "max_memory_gb": 18.31, "ce": 4.861695949919522, "rank": 0.26302504539489746} +{"step": 25, "steps": 500, "lr": 2e-05, "grad_norm": 5.270484924316406, "loss_scale": 4.0, "retries": 0, "seconds": 48.5, "max_memory_gb": 18.31, "ce": 3.9660124606452882, "rank": 0.014870405197143555} +{"step": 26, "steps": 500, "lr": 2e-05, "grad_norm": 7.225626468658447, "loss_scale": 4.0, "retries": 0, "seconds": 50.5, "max_memory_gb": 18.31, "ce": 4.953056436963379, "rank": 0.05689537525177002} +{"step": 27, "steps": 500, "lr": 2e-05, "grad_norm": 5.515596866607666, "loss_scale": 4.0, "retries": 0, "seconds": 52.7, "max_memory_gb": 18.31, "ce": 4.0285657700151205, "rank": 0.49215030670166016} +{"step": 28, "steps": 500, "lr": 2e-05, "grad_norm": 3.504291296005249, "loss_scale": 4.0, "retries": 0, "seconds": 54.8, "max_memory_gb": 18.31, "ce": 2.662933203391731, "rank": 0.0} +{"step": 29, "steps": 500, "lr": 2e-05, "grad_norm": 3.424813985824585, "loss_scale": 4.0, "retries": 0, "seconds": 56.7, "max_memory_gb": 18.31, "ce": 2.5710455330554396, "rank": 0.0} +{"step": 30, "steps": 500, "lr": 2e-05, "grad_norm": 4.549286842346191, "loss_scale": 4.0, "retries": 0, "seconds": 58.8, "max_memory_gb": 18.31, "ce": 2.590659955632873, "rank": 0.0} +{"step": 31, "steps": 500, "lr": 2e-05, "grad_norm": 4.715498447418213, "loss_scale": 4.0, "retries": 0, "seconds": 61.0, "max_memory_gb": 18.31, "ce": 2.657009322196245, "rank": 0.1393907070159912} +{"step": 32, "steps": 500, "lr": 2e-05, "grad_norm": 9.005624771118164, "loss_scale": 4.0, "retries": 0, "seconds": 63.1, "max_memory_gb": 18.31, "ce": 4.983460009098053, "rank": 0.198403000831604} +{"step": 33, "steps": 500, "lr": 2e-05, "grad_norm": 3.395764112472534, "loss_scale": 4.0, "retries": 0, "seconds": 64.9, "max_memory_gb": 18.31, "ce": 2.3966050134040415, "rank": 0.007236853241920471} +{"step": 34, "steps": 500, "lr": 2e-05, "grad_norm": 3.6256637573242188, "loss_scale": 4.0, "retries": 0, "seconds": 66.8, "max_memory_gb": 18.31, "ce": 2.5958533209050074, "rank": 0.0} +{"step": 35, "steps": 500, "lr": 2e-05, "grad_norm": 6.430530071258545, "loss_scale": 4.0, "retries": 0, "seconds": 68.7, "max_memory_gb": 18.31, "ce": 2.483442618511617, "rank": 0.11629104614257812} +{"step": 36, "steps": 500, "lr": 2e-05, "grad_norm": 4.5747504234313965, "loss_scale": 4.0, "retries": 0, "seconds": 70.6, "max_memory_gb": 18.31, "ce": 2.428677624091506, "rank": 0.0} +{"step": 37, "steps": 500, "lr": 2e-05, "grad_norm": 5.442377090454102, "loss_scale": 4.0, "retries": 0, "seconds": 72.4, "max_memory_gb": 18.31, "ce": 2.4251954844221473, "rank": 0.0} +{"step": 38, "steps": 500, "lr": 2e-05, "grad_norm": 9.654544830322266, "loss_scale": 4.0, "retries": 0, "seconds": 74.3, "max_memory_gb": 18.31, "ce": 4.495698665268719, "rank": 0.11075735092163086} +{"step": 39, "steps": 500, "lr": 2e-05, "grad_norm": 4.38059663772583, "loss_scale": 4.0, "retries": 0, "seconds": 76.2, "max_memory_gb": 18.31, "ce": 1.3001402271911502, "rank": 0.11279569566249847} +{"step": 40, "steps": 500, "lr": 2e-05, "grad_norm": 4.923818588256836, "loss_scale": 4.0, "retries": 0, "seconds": 78.0, "max_memory_gb": 18.31, "ce": 2.486141710076481, "rank": 0.0} +{"step": 41, "steps": 500, "lr": 2e-05, "grad_norm": 2.8017773628234863, "loss_scale": 4.0, "retries": 0, "seconds": 79.9, "max_memory_gb": 18.31, "ce": 1.4240965442731977, "rank": 0.0} +{"step": 42, "steps": 500, "lr": 2e-05, "grad_norm": 6.0435004234313965, "loss_scale": 4.0, "retries": 0, "seconds": 81.8, "max_memory_gb": 18.31, "ce": 2.199653811287135, "rank": 0.0} +{"step": 43, "steps": 500, "lr": 2e-05, "grad_norm": 8.668405532836914, "loss_scale": 4.0, "retries": 0, "seconds": 83.6, "max_memory_gb": 18.31, "ce": 2.162314834538847, "rank": 0.0} +{"step": 44, "steps": 500, "lr": 2e-05, "grad_norm": 3.9950668811798096, "loss_scale": 4.0, "retries": 0, "seconds": 85.5, "max_memory_gb": 18.31, "ce": 1.2953834796790034, "rank": 0.0} +{"step": 45, "steps": 500, "lr": 2e-05, "grad_norm": 4.589493274688721, "loss_scale": 4.0, "retries": 0, "seconds": 87.3, "max_memory_gb": 18.31, "ce": 1.2703814241103828, "rank": 0.0} +{"step": 46, "steps": 500, "lr": 2e-05, "grad_norm": 7.407657146453857, "loss_scale": 4.0, "retries": 0, "seconds": 89.2, "max_memory_gb": 18.31, "ce": 2.0924230637028813, "rank": 0.0} +{"step": 47, "steps": 500, "lr": 2e-05, "grad_norm": 8.682641983032227, "loss_scale": 4.0, "retries": 0, "seconds": 91.1, "max_memory_gb": 18.31, "ce": 1.7674370273016393, "rank": 0.0} +{"step": 48, "steps": 500, "lr": 2e-05, "grad_norm": 12.951146125793457, "loss_scale": 4.0, "retries": 0, "seconds": 93.0, "max_memory_gb": 18.31, "ce": 3.2515446268953383, "rank": 0.026900291442871094} +{"step": 49, "steps": 500, "lr": 2e-05, "grad_norm": 18.631267547607422, "loss_scale": 4.0, "retries": 0, "seconds": 94.9, "max_memory_gb": 18.31, "ce": 2.850562514271587, "rank": 0.0} +{"step": 50, "steps": 500, "lr": 2e-05, "grad_norm": 14.574987411499023, "loss_scale": 4.0, "retries": 0, "seconds": 96.8, "max_memory_gb": 18.31, "ce": 3.1313680593157187, "rank": 0.0} +{"step": 51, "steps": 500, "lr": 2e-05, "grad_norm": 6.523275375366211, "loss_scale": 4.0, "retries": 0, "seconds": 98.6, "max_memory_gb": 18.31, "ce": 1.1424778499640524, "rank": 0.0} +{"step": 52, "steps": 500, "lr": 2e-05, "grad_norm": 11.5653657913208, "loss_scale": 4.0, "retries": 0, "seconds": 100.5, "max_memory_gb": 18.31, "ce": 1.877055644057691, "rank": 0.0} +{"step": 53, "steps": 500, "lr": 2e-05, "grad_norm": 20.103260040283203, "loss_scale": 4.0, "retries": 0, "seconds": 102.4, "max_memory_gb": 18.31, "ce": 2.911471029743552, "rank": 0.0} +{"step": 54, "steps": 500, "lr": 2e-05, "grad_norm": 23.988109588623047, "loss_scale": 4.0, "retries": 0, "seconds": 104.2, "max_memory_gb": 18.31, "ce": 2.795452995225787, "rank": 0.0} +{"step": 55, "steps": 500, "lr": 2e-05, "grad_norm": 13.490427017211914, "loss_scale": 4.0, "retries": 0, "seconds": 106.1, "max_memory_gb": 18.31, "ce": 1.6942013433435932, "rank": 0.0} +{"step": 56, "steps": 500, "lr": 2e-05, "grad_norm": 31.970895767211914, "loss_scale": 4.0, "retries": 0, "seconds": 108.0, "max_memory_gb": 18.31, "ce": 3.090365666896105, "rank": 0.0} +{"step": 57, "steps": 500, "lr": 2e-05, "grad_norm": 7.539858341217041, "loss_scale": 4.0, "retries": 0, "seconds": 109.8, "max_memory_gb": 18.31, "ce": 0.8325130459852517, "rank": 0.01601308584213257} +{"step": 58, "steps": 500, "lr": 2e-05, "grad_norm": 14.194149017333984, "loss_scale": 4.0, "retries": 0, "seconds": 111.7, "max_memory_gb": 18.31, "ce": 1.435679837479256, "rank": 0.0} +{"step": 59, "steps": 500, "lr": 2e-05, "grad_norm": 6.139357566833496, "loss_scale": 4.0, "retries": 0, "seconds": 113.5, "max_memory_gb": 18.31, "ce": 1.0893552550114691, "rank": 0.0} +{"step": 60, "steps": 500, "lr": 2e-05, "grad_norm": 12.863786697387695, "loss_scale": 4.0, "retries": 0, "seconds": 115.4, "max_memory_gb": 18.31, "ce": 1.544598538428545, "rank": 0.0} +{"step": 61, "steps": 500, "lr": 2e-05, "grad_norm": 19.194074630737305, "loss_scale": 4.0, "retries": 0, "seconds": 117.3, "max_memory_gb": 18.31, "ce": 1.56645745690912, "rank": 0.0} +{"step": 62, "steps": 500, "lr": 2e-05, "grad_norm": 8.571046829223633, "loss_scale": 4.0, "retries": 0, "seconds": 119.1, "max_memory_gb": 18.31, "ce": 0.8185488225426525, "rank": 0.0} +{"step": 63, "steps": 500, "lr": 2e-05, "grad_norm": 13.703704833984375, "loss_scale": 4.0, "retries": 0, "seconds": 121.0, "max_memory_gb": 18.31, "ce": 1.8470755950547755, "rank": 0.0031706690788269043} +{"step": 64, "steps": 500, "lr": 2e-05, "grad_norm": 18.977340698242188, "loss_scale": 4.0, "retries": 0, "seconds": 122.9, "max_memory_gb": 18.31, "ce": 1.7552154674194753, "rank": 0.0} +{"step": 65, "steps": 500, "lr": 2e-05, "grad_norm": 14.50256061553955, "loss_scale": 4.0, "retries": 0, "seconds": 124.8, "max_memory_gb": 18.31, "ce": 1.8330026669427752, "rank": 0.0} +{"step": 66, "steps": 500, "lr": 2e-05, "grad_norm": 4.429299354553223, "loss_scale": 4.0, "retries": 0, "seconds": 126.7, "max_memory_gb": 18.31, "ce": 0.6157788990531117, "rank": 0.28296662867069244} +{"step": 67, "steps": 500, "lr": 2e-05, "grad_norm": 9.103622436523438, "loss_scale": 4.0, "retries": 0, "seconds": 128.7, "max_memory_gb": 18.31, "ce": 1.1013804720714688, "rank": 0.0} +{"step": 68, "steps": 500, "lr": 2e-05, "grad_norm": 6.175860404968262, "loss_scale": 4.0, "retries": 0, "seconds": 130.7, "max_memory_gb": 18.31, "ce": 0.5955070555210114, "rank": 0.0} +{"step": 69, "steps": 500, "lr": 2e-05, "grad_norm": 6.400826930999756, "loss_scale": 4.0, "retries": 0, "seconds": 132.7, "max_memory_gb": 18.31, "ce": 0.7843049028888345, "rank": 0.0} +{"step": 70, "steps": 500, "lr": 2e-05, "grad_norm": 4.728668212890625, "loss_scale": 4.0, "retries": 0, "seconds": 134.7, "max_memory_gb": 18.31, "ce": 0.6862623179331422, "rank": 0.0} +{"step": 71, "steps": 500, "lr": 2e-05, "grad_norm": 7.120515823364258, "loss_scale": 4.0, "retries": 0, "seconds": 136.7, "max_memory_gb": 18.31, "ce": 1.0068872007541358, "rank": 0.0} +{"step": 72, "steps": 500, "lr": 2e-05, "grad_norm": 8.064395904541016, "loss_scale": 4.0, "retries": 0, "seconds": 138.8, "max_memory_gb": 18.31, "ce": 1.1816633958369493, "rank": 0.0} +{"step": 73, "steps": 500, "lr": 2e-05, "grad_norm": 2.1528589725494385, "loss_scale": 4.0, "retries": 0, "seconds": 140.7, "max_memory_gb": 18.31, "ce": 0.3400706681422889, "rank": 0.0} +{"step": 74, "steps": 500, "lr": 2e-05, "grad_norm": 2.5153260231018066, "loss_scale": 4.0, "retries": 0, "seconds": 142.7, "max_memory_gb": 18.31, "ce": 0.6348998863250017, "rank": 0.0} +{"step": 75, "steps": 500, "lr": 2e-05, "grad_norm": 2.03932523727417, "loss_scale": 4.0, "retries": 0, "seconds": 144.7, "max_memory_gb": 18.31, "ce": 0.8953217640519142, "rank": 0.27959585189819336} +{"step": 76, "steps": 500, "lr": 2e-05, "grad_norm": 1.9097927808761597, "loss_scale": 4.0, "retries": 0, "seconds": 146.7, "max_memory_gb": 18.31, "ce": 0.6977278548292816, "rank": 0.0} +{"step": 77, "steps": 500, "lr": 2e-05, "grad_norm": 4.003260135650635, "loss_scale": 4.0, "retries": 0, "seconds": 148.7, "max_memory_gb": 18.31, "ce": 1.0293851336464286, "rank": 0.07186374068260193} +{"step": 78, "steps": 500, "lr": 2e-05, "grad_norm": 2.466580390930176, "loss_scale": 4.0, "retries": 0, "seconds": 150.6, "max_memory_gb": 18.31, "ce": 0.5736435353755951, "rank": 0.0} +{"step": 79, "steps": 500, "lr": 2e-05, "grad_norm": 2.3239526748657227, "loss_scale": 4.0, "retries": 0, "seconds": 152.6, "max_memory_gb": 18.31, "ce": 0.4409197848290205, "rank": 0.0} +{"step": 80, "steps": 500, "lr": 2e-05, "grad_norm": 2.719022750854492, "loss_scale": 4.0, "retries": 0, "seconds": 154.6, "max_memory_gb": 18.31, "ce": 0.5195091711357236, "rank": 0.0} +{"step": 81, "steps": 500, "lr": 2e-05, "grad_norm": 3.1825428009033203, "loss_scale": 4.0, "retries": 0, "seconds": 156.5, "max_memory_gb": 18.31, "ce": 0.8663215972483158, "rank": 0.0} +{"step": 82, "steps": 500, "lr": 2e-05, "grad_norm": 1.9953919649124146, "loss_scale": 4.0, "retries": 0, "seconds": 158.5, "max_memory_gb": 18.31, "ce": 0.620103731751442, "rank": 0.0} +{"step": 83, "steps": 500, "lr": 2e-05, "grad_norm": 2.311074733734131, "loss_scale": 4.0, "retries": 0, "seconds": 160.5, "max_memory_gb": 18.31, "ce": 0.7384721478447318, "rank": 0.0} +{"step": 84, "steps": 500, "lr": 2e-05, "grad_norm": 1.3279569149017334, "loss_scale": 4.0, "retries": 0, "seconds": 162.5, "max_memory_gb": 18.31, "ce": 0.35136326402425766, "rank": 0.0} +{"step": 85, "steps": 500, "lr": 2e-05, "grad_norm": 1.6746559143066406, "loss_scale": 4.0, "retries": 0, "seconds": 164.5, "max_memory_gb": 18.31, "ce": 0.19913136260583997, "rank": 0.0} +{"step": 86, "steps": 500, "lr": 2e-05, "grad_norm": 2.7358782291412354, "loss_scale": 4.0, "retries": 0, "seconds": 166.6, "max_memory_gb": 18.31, "ce": 0.741773103363812, "rank": 0.0} +{"step": 87, "steps": 500, "lr": 2e-05, "grad_norm": 2.1710057258605957, "loss_scale": 4.0, "retries": 0, "seconds": 168.5, "max_memory_gb": 18.31, "ce": 0.7230345620773733, "rank": 0.0} +{"step": 88, "steps": 500, "lr": 2e-05, "grad_norm": 2.1764044761657715, "loss_scale": 4.0, "retries": 0, "seconds": 170.5, "max_memory_gb": 18.31, "ce": 0.6845906008966267, "rank": 0.0} +{"step": 89, "steps": 500, "lr": 2e-05, "grad_norm": 0.7646390199661255, "loss_scale": 4.0, "retries": 0, "seconds": 172.5, "max_memory_gb": 18.31, "ce": 0.17540770093910396, "rank": 0.12289862334728241} +{"step": 90, "steps": 500, "lr": 2e-05, "grad_norm": 2.3174071311950684, "loss_scale": 4.0, "retries": 0, "seconds": 174.5, "max_memory_gb": 18.31, "ce": 0.5618009343743324, "rank": 0.0} +{"step": 91, "steps": 500, "lr": 2e-05, "grad_norm": 1.147375226020813, "loss_scale": 4.0, "retries": 0, "seconds": 176.5, "max_memory_gb": 18.31, "ce": 0.39560356084257364, "rank": 0.0} +{"step": 92, "steps": 500, "lr": 2e-05, "grad_norm": 1.7355773448944092, "loss_scale": 4.0, "retries": 0, "seconds": 178.5, "max_memory_gb": 18.31, "ce": 0.33899620547890663, "rank": 0.0} +{"step": 93, "steps": 500, "lr": 2e-05, "grad_norm": 3.8717920780181885, "loss_scale": 4.0, "retries": 0, "seconds": 180.5, "max_memory_gb": 18.31, "ce": 0.5396818423178047, "rank": 0.0} +{"step": 94, "steps": 500, "lr": 2e-05, "grad_norm": 4.078088760375977, "loss_scale": 4.0, "retries": 0, "seconds": 182.5, "max_memory_gb": 18.31, "ce": 0.6120377890765667, "rank": 0.0} +{"step": 95, "steps": 500, "lr": 2e-05, "grad_norm": 2.3936588764190674, "loss_scale": 4.0, "retries": 0, "seconds": 184.5, "max_memory_gb": 18.31, "ce": 0.6536250147037208, "rank": 0.0} +{"step": 96, "steps": 500, "lr": 2e-05, "grad_norm": 4.50222110748291, "loss_scale": 4.0, "retries": 0, "seconds": 186.5, "max_memory_gb": 18.31, "ce": 0.7336782226338983, "rank": 0.023739680647850037} +{"step": 97, "steps": 500, "lr": 2e-05, "grad_norm": 1.3198878765106201, "loss_scale": 4.0, "retries": 0, "seconds": 188.5, "max_memory_gb": 18.31, "ce": 0.38497006380930543, "rank": 0.0} +{"step": 98, "steps": 500, "lr": 2e-05, "grad_norm": 2.128443479537964, "loss_scale": 4.0, "retries": 0, "seconds": 190.5, "max_memory_gb": 18.31, "ce": 0.3863756670616567, "rank": 0.0} +{"step": 99, "steps": 500, "lr": 2e-05, "grad_norm": 1.4639595746994019, "loss_scale": 4.0, "retries": 0, "seconds": 192.5, "max_memory_gb": 18.31, "ce": 0.240547088207677, "rank": 0.01625332236289978} +{"step": 100, "steps": 500, "lr": 2e-05, "grad_norm": 3.803046226501465, "loss_scale": 4.0, "retries": 0, "seconds": 194.5, "max_memory_gb": 18.31, "ce": 0.6496609062887728, "rank": 0.0} +{"step": 101, "steps": 500, "lr": 2e-05, "grad_norm": 2.4584639072418213, "loss_scale": 4.0, "retries": 0, "seconds": 196.6, "max_memory_gb": 18.31, "ce": 0.6494545508176088, "rank": 0.0} +{"step": 102, "steps": 500, "lr": 2e-05, "grad_norm": 3.7477335929870605, "loss_scale": 4.0, "retries": 0, "seconds": 198.6, "max_memory_gb": 18.31, "ce": 0.6789464191533625, "rank": 0.0} +{"step": 103, "steps": 500, "lr": 2e-05, "grad_norm": 1.0988794565200806, "loss_scale": 4.0, "retries": 0, "seconds": 200.6, "max_memory_gb": 18.31, "ce": 0.34171728137880564, "rank": 0.12793219089508057} +{"step": 104, "steps": 500, "lr": 2e-05, "grad_norm": 1.3566032648086548, "loss_scale": 4.0, "retries": 0, "seconds": 202.5, "max_memory_gb": 18.31, "ce": 0.28219386911951005, "rank": 0.09601322561502457} +{"step": 105, "steps": 500, "lr": 2e-05, "grad_norm": 1.898671269416809, "loss_scale": 4.0, "retries": 0, "seconds": 204.6, "max_memory_gb": 18.31, "ce": 0.561103982385248, "rank": 0.0} +{"step": 106, "steps": 500, "lr": 2e-05, "grad_norm": 1.757282018661499, "loss_scale": 4.0, "retries": 0, "seconds": 206.5, "max_memory_gb": 18.31, "ce": 0.5320590939372778, "rank": 0.06822144240140915} +{"step": 107, "steps": 500, "lr": 2e-05, "grad_norm": 1.1326117515563965, "loss_scale": 4.0, "retries": 0, "seconds": 208.6, "max_memory_gb": 18.31, "ce": 0.31290564232040197, "rank": 0.0} +{"step": 108, "steps": 500, "lr": 2e-05, "grad_norm": 0.9427939653396606, "loss_scale": 4.0, "retries": 0, "seconds": 210.5, "max_memory_gb": 18.31, "ce": 0.31829749164171517, "rank": 0.0} +{"step": 109, "steps": 500, "lr": 2e-05, "grad_norm": 1.7820696830749512, "loss_scale": 4.0, "retries": 0, "seconds": 212.6, "max_memory_gb": 18.31, "ce": 0.6713440143503249, "rank": 0.0} +{"step": 110, "steps": 500, "lr": 2e-05, "grad_norm": 1.5101172924041748, "loss_scale": 4.0, "retries": 0, "seconds": 214.6, "max_memory_gb": 18.31, "ce": 0.4734259080141783, "rank": 0.012196511030197144} +{"step": 111, "steps": 500, "lr": 2e-05, "grad_norm": 1.9143060445785522, "loss_scale": 4.0, "retries": 0, "seconds": 216.6, "max_memory_gb": 18.31, "ce": 0.4309189706109464, "rank": 0.0} +{"step": 112, "steps": 500, "lr": 2e-05, "grad_norm": 1.198463797569275, "loss_scale": 4.0, "retries": 0, "seconds": 218.6, "max_memory_gb": 18.31, "ce": 0.3346054498106241, "rank": 0.0} +{"step": 113, "steps": 500, "lr": 2e-05, "grad_norm": 1.0988796949386597, "loss_scale": 4.0, "retries": 0, "seconds": 220.6, "max_memory_gb": 18.31, "ce": 0.4105471484363079, "rank": 0.0} +{"step": 114, "steps": 500, "lr": 2e-05, "grad_norm": 3.679158926010132, "loss_scale": 4.0, "retries": 0, "seconds": 222.6, "max_memory_gb": 18.31, "ce": 0.8094872483052313, "rank": 0.18026883900165558} +{"step": 115, "steps": 500, "lr": 2e-05, "grad_norm": 0.890186071395874, "loss_scale": 4.0, "retries": 0, "seconds": 224.7, "max_memory_gb": 18.31, "ce": 0.4595544310286641, "rank": 0.03654561936855316} +{"step": 116, "steps": 500, "lr": 2e-05, "grad_norm": 0.34266144037246704, "loss_scale": 4.0, "retries": 0, "seconds": 226.7, "max_memory_gb": 18.31, "ce": 0.14965362451039255, "rank": 0.0} +{"step": 117, "steps": 500, "lr": 2e-05, "grad_norm": 5.043933868408203, "loss_scale": 4.0, "retries": 0, "seconds": 228.6, "max_memory_gb": 18.31, "ce": 0.5412510588066652, "rank": 0.0} +{"step": 118, "steps": 500, "lr": 2e-05, "grad_norm": 0.5552418828010559, "loss_scale": 4.0, "retries": 0, "seconds": 230.7, "max_memory_gb": 18.31, "ce": 0.2965145199559629, "rank": 0.0} +{"step": 119, "steps": 500, "lr": 2e-05, "grad_norm": 1.4427558183670044, "loss_scale": 4.0, "retries": 0, "seconds": 232.7, "max_memory_gb": 18.31, "ce": 0.2903419937938452, "rank": 0.0} +{"step": 120, "steps": 500, "lr": 2e-05, "grad_norm": 2.8447482585906982, "loss_scale": 4.0, "retries": 0, "seconds": 234.7, "max_memory_gb": 18.31, "ce": 0.4719014912843704, "rank": 0.0} +{"step": 121, "steps": 500, "lr": 2e-05, "grad_norm": 2.27293062210083, "loss_scale": 4.0, "retries": 0, "seconds": 236.7, "max_memory_gb": 18.31, "ce": 0.376802516868338, "rank": 0.0} +{"step": 122, "steps": 500, "lr": 2e-05, "grad_norm": 1.1381886005401611, "loss_scale": 4.0, "retries": 0, "seconds": 238.7, "max_memory_gb": 18.31, "ce": 0.379904743982479, "rank": 0.0} +{"step": 123, "steps": 500, "lr": 2e-05, "grad_norm": 1.4442660808563232, "loss_scale": 4.0, "retries": 0, "seconds": 240.7, "max_memory_gb": 18.31, "ce": 0.47882878954987973, "rank": 0.013811826705932617} +{"step": 124, "steps": 500, "lr": 2e-05, "grad_norm": 1.4994014501571655, "loss_scale": 4.0, "retries": 0, "seconds": 242.7, "max_memory_gb": 18.31, "ce": 0.40811988338828087, "rank": 0.037844330072402954} +{"step": 125, "steps": 500, "lr": 2e-05, "grad_norm": 0.5964733362197876, "loss_scale": 4.0, "retries": 0, "seconds": 244.6, "max_memory_gb": 18.31, "ce": 0.2981872702948749, "rank": 0.0} +{"step": 126, "steps": 500, "lr": 2e-05, "grad_norm": 2.819124221801758, "loss_scale": 4.0, "retries": 0, "seconds": 246.6, "max_memory_gb": 18.31, "ce": 0.6039483747445047, "rank": 0.0} +{"step": 127, "steps": 500, "lr": 2e-05, "grad_norm": 0.668684184551239, "loss_scale": 4.0, "retries": 0, "seconds": 248.6, "max_memory_gb": 18.31, "ce": 0.1792061243322678, "rank": 0.0} +{"step": 128, "steps": 500, "lr": 2e-05, "grad_norm": 1.4646505117416382, "loss_scale": 4.0, "retries": 0, "seconds": 250.6, "max_memory_gb": 18.31, "ce": 0.4265053898561746, "rank": 0.0} +{"step": 129, "steps": 500, "lr": 2e-05, "grad_norm": 1.650977611541748, "loss_scale": 4.0, "retries": 0, "seconds": 252.6, "max_memory_gb": 18.31, "ce": 0.4978488301858306, "rank": 0.0} +{"step": 130, "steps": 500, "lr": 2e-05, "grad_norm": 3.770195245742798, "loss_scale": 4.0, "retries": 0, "seconds": 254.6, "max_memory_gb": 18.31, "ce": 0.5723220584914088, "rank": 0.0} +{"step": 131, "steps": 500, "lr": 2e-05, "grad_norm": 1.0827420949935913, "loss_scale": 4.0, "retries": 0, "seconds": 256.6, "max_memory_gb": 18.31, "ce": 0.38402725383639336, "rank": 0.0} +{"step": 132, "steps": 500, "lr": 2e-05, "grad_norm": 1.4594990015029907, "loss_scale": 4.0, "retries": 0, "seconds": 258.6, "max_memory_gb": 18.31, "ce": 0.5260983879561536, "rank": 0.0} +{"step": 133, "steps": 500, "lr": 2e-05, "grad_norm": 1.6845076084136963, "loss_scale": 4.0, "retries": 0, "seconds": 260.6, "max_memory_gb": 18.31, "ce": 0.5249764895997941, "rank": 0.09709364175796509} +{"step": 134, "steps": 500, "lr": 2e-05, "grad_norm": 0.923667311668396, "loss_scale": 4.0, "retries": 0, "seconds": 262.6, "max_memory_gb": 18.31, "ce": 0.20775700930971652, "rank": 0.0} +{"step": 135, "steps": 500, "lr": 2e-05, "grad_norm": 1.776752233505249, "loss_scale": 4.0, "retries": 0, "seconds": 264.6, "max_memory_gb": 18.31, "ce": 0.4456454822793603, "rank": 0.021580949425697327} +{"step": 136, "steps": 500, "lr": 2e-05, "grad_norm": 1.321881651878357, "loss_scale": 4.0, "retries": 0, "seconds": 266.6, "max_memory_gb": 18.31, "ce": 0.3977982571814209, "rank": 0.0} +{"step": 137, "steps": 500, "lr": 2e-05, "grad_norm": 0.9321845173835754, "loss_scale": 4.0, "retries": 0, "seconds": 268.6, "max_memory_gb": 18.31, "ce": 0.3401366490870714, "rank": 0.0} +{"step": 138, "steps": 500, "lr": 2e-05, "grad_norm": 1.2699286937713623, "loss_scale": 4.0, "retries": 0, "seconds": 270.7, "max_memory_gb": 18.31, "ce": 0.32382293278351426, "rank": 0.0036299526691436768} +{"step": 139, "steps": 500, "lr": 2e-05, "grad_norm": 7.783802509307861, "loss_scale": 4.0, "retries": 0, "seconds": 272.7, "max_memory_gb": 18.31, "ce": 0.6174359177239239, "rank": 0.0} +{"step": 140, "steps": 500, "lr": 2e-05, "grad_norm": 1.3773435354232788, "loss_scale": 4.0, "retries": 0, "seconds": 274.7, "max_memory_gb": 18.31, "ce": 0.2844881978817284, "rank": 0.0} +{"step": 141, "steps": 500, "lr": 2e-05, "grad_norm": 0.908558189868927, "loss_scale": 4.0, "retries": 0, "seconds": 276.8, "max_memory_gb": 18.31, "ce": 0.3653662479482591, "rank": 0.0} +{"step": 142, "steps": 500, "lr": 2e-05, "grad_norm": 1.001021146774292, "loss_scale": 4.0, "retries": 0, "seconds": 278.8, "max_memory_gb": 18.31, "ce": 0.3245869716629386, "rank": 0.0} +{"step": 143, "steps": 500, "lr": 2e-05, "grad_norm": 0.38515833020210266, "loss_scale": 4.0, "retries": 0, "seconds": 280.8, "max_memory_gb": 18.31, "ce": 0.17508949770126492, "rank": 0.0} +{"step": 144, "steps": 500, "lr": 2e-05, "grad_norm": 1.9876083135604858, "loss_scale": 4.0, "retries": 0, "seconds": 282.9, "max_memory_gb": 18.31, "ce": 0.39984282304067165, "rank": 0.0} +{"step": 145, "steps": 500, "lr": 2e-05, "grad_norm": 0.7622568011283875, "loss_scale": 4.0, "retries": 0, "seconds": 284.9, "max_memory_gb": 18.31, "ce": 0.21851173852337524, "rank": 0.0} +{"step": 146, "steps": 500, "lr": 2e-05, "grad_norm": 0.9072088599205017, "loss_scale": 4.0, "retries": 0, "seconds": 286.9, "max_memory_gb": 18.31, "ce": 0.3161202729679644, "rank": 0.0} +{"step": 147, "steps": 500, "lr": 2e-05, "grad_norm": 1.2531086206436157, "loss_scale": 4.0, "retries": 0, "seconds": 289.0, "max_memory_gb": 18.31, "ce": 0.5131501925643533, "rank": 0.0} +{"step": 148, "steps": 500, "lr": 2e-05, "grad_norm": 1.0978169441223145, "loss_scale": 4.0, "retries": 0, "seconds": 291.0, "max_memory_gb": 18.31, "ce": 0.14383666717913002, "rank": 0.0} +{"step": 149, "steps": 500, "lr": 2e-05, "grad_norm": 1.1800119876861572, "loss_scale": 4.0, "retries": 0, "seconds": 293.0, "max_memory_gb": 18.31, "ce": 0.23353971145115793, "rank": 0.13681399822235107} +{"step": 150, "steps": 500, "lr": 2e-05, "grad_norm": 2.1084680557250977, "loss_scale": 4.0, "retries": 0, "seconds": 295.0, "max_memory_gb": 18.31, "ce": 0.4581875652074814, "rank": 0.02456572651863098} +{"step": 151, "steps": 500, "lr": 2e-05, "grad_norm": 3.576277017593384, "loss_scale": 4.0, "retries": 0, "seconds": 297.0, "max_memory_gb": 18.31, "ce": 0.2701233578263782, "rank": 0.0} +{"step": 152, "steps": 500, "lr": 2e-05, "grad_norm": 1.9431639909744263, "loss_scale": 4.0, "retries": 0, "seconds": 299.0, "max_memory_gb": 18.31, "ce": 0.4342640438117087, "rank": 0.0} +{"step": 153, "steps": 500, "lr": 2e-05, "grad_norm": 3.2998740673065186, "loss_scale": 4.0, "retries": 0, "seconds": 301.0, "max_memory_gb": 18.31, "ce": 0.36602868558838964, "rank": 0.0} +{"step": 154, "steps": 500, "lr": 2e-05, "grad_norm": 2.0778794288635254, "loss_scale": 4.0, "retries": 0, "seconds": 303.1, "max_memory_gb": 18.31, "ce": 0.6142933629453182, "rank": 0.1311507374048233} +{"step": 155, "steps": 500, "lr": 2e-05, "grad_norm": 1.893209457397461, "loss_scale": 4.0, "retries": 0, "seconds": 305.1, "max_memory_gb": 18.31, "ce": 0.5771302902139723, "rank": 0.0} +{"step": 156, "steps": 500, "lr": 2e-05, "grad_norm": 1.5132277011871338, "loss_scale": 4.0, "retries": 0, "seconds": 307.1, "max_memory_gb": 18.31, "ce": 0.26909288857132196, "rank": 0.0} +{"step": 157, "steps": 500, "lr": 2e-05, "grad_norm": 0.9409525394439697, "loss_scale": 4.0, "retries": 0, "seconds": 309.1, "max_memory_gb": 18.31, "ce": 0.2618532197084278, "rank": 0.0} +{"step": 158, "steps": 500, "lr": 2e-05, "grad_norm": 1.67660653591156, "loss_scale": 4.0, "retries": 0, "seconds": 311.1, "max_memory_gb": 18.31, "ce": 0.43257842562161386, "rank": 0.0} +{"step": 159, "steps": 500, "lr": 2e-05, "grad_norm": 2.4746923446655273, "loss_scale": 4.0, "retries": 0, "seconds": 313.1, "max_memory_gb": 18.32, "ce": 0.2755411786492914, "rank": 0.0} +{"step": 160, "steps": 500, "lr": 2e-05, "grad_norm": 1.5353223085403442, "loss_scale": 4.0, "retries": 0, "seconds": 315.2, "max_memory_gb": 18.32, "ce": 0.33609702764078975, "rank": 0.031847938895225525} +{"step": 161, "steps": 500, "lr": 2e-05, "grad_norm": 0.9892170429229736, "loss_scale": 4.0, "retries": 0, "seconds": 317.2, "max_memory_gb": 18.32, "ce": 0.32294549979269505, "rank": 0.0} +{"step": 162, "steps": 500, "lr": 2e-05, "grad_norm": 0.7242389917373657, "loss_scale": 4.0, "retries": 0, "seconds": 319.1, "max_memory_gb": 18.32, "ce": 0.2140504214912653, "rank": 0.0} +{"step": 163, "steps": 500, "lr": 2e-05, "grad_norm": 1.747138500213623, "loss_scale": 4.0, "retries": 0, "seconds": 321.1, "max_memory_gb": 18.32, "ce": 0.3358354961965233, "rank": 0.0} +{"step": 164, "steps": 500, "lr": 2e-05, "grad_norm": 1.7143757343292236, "loss_scale": 4.0, "retries": 0, "seconds": 323.1, "max_memory_gb": 18.32, "ce": 0.384681835770607, "rank": 0.062122225761413574} +{"step": 165, "steps": 500, "lr": 2e-05, "grad_norm": 1.3346211910247803, "loss_scale": 4.0, "retries": 0, "seconds": 325.2, "max_memory_gb": 18.32, "ce": 0.23971152491867542, "rank": 0.0} +{"step": 166, "steps": 500, "lr": 2e-05, "grad_norm": 2.513136386871338, "loss_scale": 4.0, "retries": 0, "seconds": 327.2, "max_memory_gb": 18.32, "ce": 0.35448462003841996, "rank": 0.0} +{"step": 167, "steps": 500, "lr": 2e-05, "grad_norm": 2.785365104675293, "loss_scale": 4.0, "retries": 0, "seconds": 329.3, "max_memory_gb": 18.32, "ce": 0.36417938489466906, "rank": 0.0} +{"step": 168, "steps": 500, "lr": 2e-05, "grad_norm": 2.5202860832214355, "loss_scale": 4.0, "retries": 0, "seconds": 331.2, "max_memory_gb": 18.32, "ce": 0.24004781059920788, "rank": 0.0} +{"step": 169, "steps": 500, "lr": 2e-05, "grad_norm": 4.060373306274414, "loss_scale": 4.0, "retries": 0, "seconds": 333.3, "max_memory_gb": 18.32, "ce": 0.7449459475465119, "rank": 0.0} +{"step": 170, "steps": 500, "lr": 2e-05, "grad_norm": 1.6521494388580322, "loss_scale": 4.0, "retries": 0, "seconds": 335.3, "max_memory_gb": 18.32, "ce": 0.2598539567552507, "rank": 0.0} +{"step": 171, "steps": 500, "lr": 2e-05, "grad_norm": 1.3624235391616821, "loss_scale": 4.0, "retries": 0, "seconds": 337.3, "max_memory_gb": 18.32, "ce": 0.25540945172542706, "rank": 0.0} +{"step": 172, "steps": 500, "lr": 2e-05, "grad_norm": 2.8468756675720215, "loss_scale": 4.0, "retries": 0, "seconds": 339.3, "max_memory_gb": 18.32, "ce": 0.3391500471625477, "rank": 0.0} +{"step": 173, "steps": 500, "lr": 2e-05, "grad_norm": 2.089695692062378, "loss_scale": 4.0, "retries": 0, "seconds": 341.3, "max_memory_gb": 18.32, "ce": 0.3949899417348206, "rank": 0.0} +{"step": 174, "steps": 500, "lr": 2e-05, "grad_norm": 1.6668647527694702, "loss_scale": 4.0, "retries": 0, "seconds": 343.3, "max_memory_gb": 18.32, "ce": 0.2942879308247939, "rank": 0.0} +{"step": 175, "steps": 500, "lr": 2e-05, "grad_norm": 1.9302988052368164, "loss_scale": 4.0, "retries": 0, "seconds": 345.3, "max_memory_gb": 18.32, "ce": 0.36531672766432166, "rank": 0.0} +{"step": 176, "steps": 500, "lr": 2e-05, "grad_norm": 2.175701141357422, "loss_scale": 4.0, "retries": 0, "seconds": 347.3, "max_memory_gb": 18.32, "ce": 0.27086852164939046, "rank": 0.0} +{"step": 177, "steps": 500, "lr": 2e-05, "grad_norm": 2.6504766941070557, "loss_scale": 4.0, "retries": 0, "seconds": 349.3, "max_memory_gb": 18.32, "ce": 0.3295565026346594, "rank": 0.12686584889888763} +{"step": 178, "steps": 500, "lr": 2e-05, "grad_norm": 1.332173466682434, "loss_scale": 4.0, "retries": 0, "seconds": 351.3, "max_memory_gb": 18.32, "ce": 0.24071971734520048, "rank": 0.0} +{"step": 179, "steps": 500, "lr": 2e-05, "grad_norm": 1.4591366052627563, "loss_scale": 4.0, "retries": 0, "seconds": 353.3, "max_memory_gb": 18.32, "ce": 0.43351836380315945, "rank": 0.0} +{"step": 180, "steps": 500, "lr": 2e-05, "grad_norm": 2.5533788204193115, "loss_scale": 4.0, "retries": 0, "seconds": 355.3, "max_memory_gb": 18.32, "ce": 0.5309699508361518, "rank": 0.0} +{"step": 181, "steps": 500, "lr": 2e-05, "grad_norm": 4.072080612182617, "loss_scale": 4.0, "retries": 0, "seconds": 357.3, "max_memory_gb": 18.32, "ce": 0.30300918500870466, "rank": 0.0} +{"step": 182, "steps": 500, "lr": 2e-05, "grad_norm": 2.0116679668426514, "loss_scale": 4.0, "retries": 0, "seconds": 359.3, "max_memory_gb": 18.32, "ce": 0.26240852172486484, "rank": 0.0} +{"step": 183, "steps": 500, "lr": 2e-05, "grad_norm": 1.0292670726776123, "loss_scale": 4.0, "retries": 0, "seconds": 361.3, "max_memory_gb": 18.32, "ce": 0.1327290665358305, "rank": 0.0} +{"step": 184, "steps": 500, "lr": 2e-05, "grad_norm": 1.2900819778442383, "loss_scale": 4.0, "retries": 0, "seconds": 363.4, "max_memory_gb": 18.32, "ce": 0.20606197929009795, "rank": 0.11525106430053711} +{"step": 185, "steps": 500, "lr": 2e-05, "grad_norm": 1.8766381740570068, "loss_scale": 4.0, "retries": 0, "seconds": 365.4, "max_memory_gb": 18.32, "ce": 0.22911954997107387, "rank": 0.0} +{"step": 186, "steps": 500, "lr": 2e-05, "grad_norm": 1.6797308921813965, "loss_scale": 4.0, "retries": 0, "seconds": 367.4, "max_memory_gb": 18.32, "ce": 0.1801077526761219, "rank": 0.0} +{"step": 187, "steps": 500, "lr": 2e-05, "grad_norm": 2.4304990768432617, "loss_scale": 4.0, "retries": 0, "seconds": 369.5, "max_memory_gb": 18.32, "ce": 0.3412913763895631, "rank": 0.12550145387649536} +{"step": 188, "steps": 500, "lr": 2e-05, "grad_norm": 0.47059109807014465, "loss_scale": 4.0, "retries": 0, "seconds": 371.4, "max_memory_gb": 18.32, "ce": 0.1551526619878132, "rank": 0.0} +{"step": 189, "steps": 500, "lr": 2e-05, "grad_norm": 2.5109164714813232, "loss_scale": 4.0, "retries": 0, "seconds": 373.4, "max_memory_gb": 18.32, "ce": 0.2213273054221645, "rank": 0.0} +{"step": 190, "steps": 500, "lr": 2e-05, "grad_norm": 3.3735220432281494, "loss_scale": 4.0, "retries": 0, "seconds": 375.4, "max_memory_gb": 18.32, "ce": 0.29795574210584164, "rank": 0.0} +{"step": 191, "steps": 500, "lr": 2e-05, "grad_norm": 1.173689603805542, "loss_scale": 4.0, "retries": 0, "seconds": 377.4, "max_memory_gb": 18.32, "ce": 0.31396467331796885, "rank": 0.0} +{"step": 192, "steps": 500, "lr": 2e-05, "grad_norm": 2.343683958053589, "loss_scale": 4.0, "retries": 0, "seconds": 379.4, "max_memory_gb": 18.32, "ce": 0.3063203482888639, "rank": 0.0} +{"step": 193, "steps": 500, "lr": 2e-05, "grad_norm": 2.4738810062408447, "loss_scale": 4.0, "retries": 0, "seconds": 381.4, "max_memory_gb": 18.32, "ce": 0.2174482379341498, "rank": 0.12554316222667694} +{"step": 194, "steps": 500, "lr": 2e-05, "grad_norm": 2.400810956954956, "loss_scale": 4.0, "retries": 0, "seconds": 383.4, "max_memory_gb": 18.32, "ce": 0.3675345336087048, "rank": 0.0} +{"step": 195, "steps": 500, "lr": 2e-05, "grad_norm": 2.767277956008911, "loss_scale": 4.0, "retries": 0, "seconds": 385.4, "max_memory_gb": 18.32, "ce": 0.2930086317937821, "rank": 0.0} +{"step": 196, "steps": 500, "lr": 2e-05, "grad_norm": 2.7841901779174805, "loss_scale": 4.0, "retries": 0, "seconds": 387.5, "max_memory_gb": 18.32, "ce": 0.19711516512325034, "rank": 0.0} +{"step": 197, "steps": 500, "lr": 2e-05, "grad_norm": 1.427781581878662, "loss_scale": 4.0, "retries": 0, "seconds": 389.5, "max_memory_gb": 18.32, "ce": 0.19190263538621366, "rank": 0.07084564119577408} +{"step": 198, "steps": 500, "lr": 2e-05, "grad_norm": 1.8207627534866333, "loss_scale": 4.0, "retries": 0, "seconds": 391.5, "max_memory_gb": 18.32, "ce": 0.2631106667686254, "rank": 0.0} +{"step": 199, "steps": 500, "lr": 2e-05, "grad_norm": 5.317220687866211, "loss_scale": 4.0, "retries": 0, "seconds": 393.5, "max_memory_gb": 18.32, "ce": 0.2872769245877862, "rank": 0.0} +{"step": 200, "steps": 500, "lr": 2e-05, "grad_norm": 1.2221707105636597, "loss_scale": 4.0, "retries": 0, "seconds": 395.5, "max_memory_gb": 18.32, "ce": 0.28807949693873525, "rank": 0.0} +{"step": 201, "steps": 500, "lr": 2e-05, "grad_norm": 1.8158386945724487, "loss_scale": 4.0, "retries": 0, "seconds": 397.5, "max_memory_gb": 18.32, "ce": 0.1826290120370686, "rank": 0.0} +{"step": 202, "steps": 500, "lr": 2e-05, "grad_norm": 1.3955881595611572, "loss_scale": 4.0, "retries": 0, "seconds": 399.5, "max_memory_gb": 18.32, "ce": 0.08993440741323866, "rank": 0.056428566575050354} +{"step": 203, "steps": 500, "lr": 2e-05, "grad_norm": 2.133164644241333, "loss_scale": 4.0, "retries": 0, "seconds": 401.5, "max_memory_gb": 18.32, "ce": 0.24907873035408556, "rank": 0.0} +{"step": 204, "steps": 500, "lr": 2e-05, "grad_norm": 4.337616920471191, "loss_scale": 4.0, "retries": 0, "seconds": 403.5, "max_memory_gb": 18.32, "ce": 0.3255832127761096, "rank": 0.0} +{"step": 205, "steps": 500, "lr": 2e-05, "grad_norm": 4.148371696472168, "loss_scale": 4.0, "retries": 0, "seconds": 405.5, "max_memory_gb": 18.32, "ce": 0.26461037664557807, "rank": 0.0} +{"step": 206, "steps": 500, "lr": 2e-05, "grad_norm": 2.5777621269226074, "loss_scale": 4.0, "retries": 0, "seconds": 407.5, "max_memory_gb": 18.32, "ce": 0.22059691924368963, "rank": 0.0} +{"step": 207, "steps": 500, "lr": 2e-05, "grad_norm": 1.839280605316162, "loss_scale": 4.0, "retries": 0, "seconds": 409.5, "max_memory_gb": 18.32, "ce": 0.4814087301492691, "rank": 0.15789270401000977} +{"step": 208, "steps": 500, "lr": 2e-05, "grad_norm": 2.49075984954834, "loss_scale": 4.0, "retries": 0, "seconds": 411.6, "max_memory_gb": 18.32, "ce": 0.10898236118373461, "rank": 0.0} +{"step": 209, "steps": 500, "lr": 2e-05, "grad_norm": 1.3977385759353638, "loss_scale": 4.0, "retries": 0, "seconds": 413.6, "max_memory_gb": 18.32, "ce": 0.24111256003379822, "rank": 0.12396179884672165} +{"step": 210, "steps": 500, "lr": 2e-05, "grad_norm": 1.0134658813476562, "loss_scale": 4.0, "retries": 0, "seconds": 415.6, "max_memory_gb": 18.32, "ce": 0.286324639339, "rank": 0.055821627378463745} +{"step": 211, "steps": 500, "lr": 2e-05, "grad_norm": 2.2529854774475098, "loss_scale": 4.0, "retries": 0, "seconds": 417.6, "max_memory_gb": 18.32, "ce": 0.25869105476886034, "rank": 0.0} +{"step": 212, "steps": 500, "lr": 2e-05, "grad_norm": 2.289865732192993, "loss_scale": 4.0, "retries": 0, "seconds": 419.7, "max_memory_gb": 18.32, "ce": 0.3713175421580672, "rank": 0.0} +{"step": 213, "steps": 500, "lr": 2e-05, "grad_norm": 1.5843509435653687, "loss_scale": 4.0, "retries": 0, "seconds": 421.7, "max_memory_gb": 18.32, "ce": 0.31659865472465754, "rank": 0.0} +{"step": 214, "steps": 500, "lr": 2e-05, "grad_norm": 2.297292470932007, "loss_scale": 4.0, "retries": 0, "seconds": 423.7, "max_memory_gb": 18.32, "ce": 0.3234924638236407, "rank": 0.0} +{"step": 215, "steps": 500, "lr": 2e-05, "grad_norm": 3.0949015617370605, "loss_scale": 4.0, "retries": 0, "seconds": 425.7, "max_memory_gb": 18.32, "ce": 0.4358470746083185, "rank": 0.0} +{"step": 216, "steps": 500, "lr": 2e-05, "grad_norm": 2.599602699279785, "loss_scale": 4.0, "retries": 0, "seconds": 427.7, "max_memory_gb": 18.32, "ce": 0.2234219026286155, "rank": 0.0} +{"step": 217, "steps": 500, "lr": 2e-05, "grad_norm": 2.4495997428894043, "loss_scale": 4.0, "retries": 0, "seconds": 429.8, "max_memory_gb": 18.32, "ce": 0.10048736399039626, "rank": 0.0} +{"step": 218, "steps": 500, "lr": 2e-05, "grad_norm": 21.67994499206543, "loss_scale": 4.0, "retries": 0, "seconds": 431.7, "max_memory_gb": 18.32, "ce": 0.5236764145083725, "rank": 0.0} +{"step": 219, "steps": 500, "lr": 2e-05, "grad_norm": 7.746950149536133, "loss_scale": 4.0, "retries": 0, "seconds": 433.8, "max_memory_gb": 18.32, "ce": 0.180508483259473, "rank": 0.0} +{"step": 220, "steps": 500, "lr": 2e-05, "grad_norm": 2.176785469055176, "loss_scale": 4.0, "retries": 0, "seconds": 435.8, "max_memory_gb": 18.32, "ce": 0.11054199415957555, "rank": 0.0} +{"step": 221, "steps": 500, "lr": 2e-05, "grad_norm": 14.946172714233398, "loss_scale": 4.0, "retries": 0, "seconds": 437.8, "max_memory_gb": 18.32, "ce": 0.2178103239275515, "rank": 0.0} +{"step": 222, "steps": 500, "lr": 2e-05, "grad_norm": 4.015596389770508, "loss_scale": 4.0, "retries": 0, "seconds": 439.9, "max_memory_gb": 18.32, "ce": 0.30489544570446014, "rank": 0.0} +{"step": 223, "steps": 500, "lr": 2e-05, "grad_norm": 0.8564186692237854, "loss_scale": 4.0, "retries": 0, "seconds": 441.9, "max_memory_gb": 18.32, "ce": 0.17889671307057142, "rank": 0.0} +{"step": 224, "steps": 500, "lr": 2e-05, "grad_norm": 1.5045019388198853, "loss_scale": 4.0, "retries": 0, "seconds": 443.9, "max_memory_gb": 18.32, "ce": 0.13163260486908257, "rank": 0.09911451488733292} +{"step": 225, "steps": 500, "lr": 2e-05, "grad_norm": 2.2971243858337402, "loss_scale": 4.0, "retries": 0, "seconds": 445.9, "max_memory_gb": 18.32, "ce": 0.1121503934264183, "rank": 0.0} +{"step": 226, "steps": 500, "lr": 2e-05, "grad_norm": 5.133442401885986, "loss_scale": 4.0, "retries": 0, "seconds": 447.9, "max_memory_gb": 18.32, "ce": 0.14204797067213804, "rank": 0.0} +{"step": 227, "steps": 500, "lr": 2e-05, "grad_norm": 2.4756317138671875, "loss_scale": 4.0, "retries": 0, "seconds": 449.9, "max_memory_gb": 18.32, "ce": 0.07707347595714964, "rank": 0.0} +{"step": 228, "steps": 500, "lr": 2e-05, "grad_norm": 1.5356462001800537, "loss_scale": 4.0, "retries": 0, "seconds": 451.9, "max_memory_gb": 18.32, "ce": 0.08972584502771497, "rank": 0.0} +{"step": 229, "steps": 500, "lr": 2e-05, "grad_norm": 1.0881917476654053, "loss_scale": 4.0, "retries": 0, "seconds": 453.9, "max_memory_gb": 18.32, "ce": 0.18908716249279678, "rank": 0.0} +{"step": 230, "steps": 500, "lr": 2e-05, "grad_norm": 1.4090659618377686, "loss_scale": 4.0, "retries": 0, "seconds": 455.8, "max_memory_gb": 18.32, "ce": 0.24706766569579486, "rank": 0.0} +{"step": 231, "steps": 500, "lr": 2e-05, "grad_norm": 1.2007348537445068, "loss_scale": 4.0, "retries": 0, "seconds": 457.8, "max_memory_gb": 18.32, "ce": 0.26608334759657737, "rank": 0.0} +{"step": 232, "steps": 500, "lr": 2e-05, "grad_norm": 7.163333415985107, "loss_scale": 4.0, "retries": 0, "seconds": 459.8, "max_memory_gb": 18.32, "ce": 0.16421741619706154, "rank": 0.0} +{"step": 233, "steps": 500, "lr": 2e-05, "grad_norm": 8.177688598632812, "loss_scale": 4.0, "retries": 0, "seconds": 461.8, "max_memory_gb": 18.32, "ce": 0.2870282621588558, "rank": 0.0} +{"step": 234, "steps": 500, "lr": 2e-05, "grad_norm": 1.8827890157699585, "loss_scale": 4.0, "retries": 0, "seconds": 463.8, "max_memory_gb": 18.32, "ce": 0.18968787149060518, "rank": 0.0} +{"step": 235, "steps": 500, "lr": 2e-05, "grad_norm": 1.5486903190612793, "loss_scale": 4.0, "retries": 0, "seconds": 465.9, "max_memory_gb": 18.32, "ce": 0.33304805075749755, "rank": 0.0} +{"step": 236, "steps": 500, "lr": 2e-05, "grad_norm": 0.9766548275947571, "loss_scale": 4.0, "retries": 0, "seconds": 467.8, "max_memory_gb": 18.32, "ce": 0.13248877634759992, "rank": 0.0} +{"step": 237, "steps": 500, "lr": 2e-05, "grad_norm": 3.9870340824127197, "loss_scale": 4.0, "retries": 0, "seconds": 469.8, "max_memory_gb": 18.32, "ce": 0.29324307770002633, "rank": 0.0} +{"step": 238, "steps": 500, "lr": 2e-05, "grad_norm": 1.6978802680969238, "loss_scale": 4.0, "retries": 0, "seconds": 471.8, "max_memory_gb": 18.32, "ce": 0.09984770603477955, "rank": 0.0} +{"step": 239, "steps": 500, "lr": 2e-05, "grad_norm": 2.052016258239746, "loss_scale": 4.0, "retries": 0, "seconds": 473.8, "max_memory_gb": 18.32, "ce": 0.17549941455945373, "rank": 0.0} +{"step": 240, "steps": 500, "lr": 2e-05, "grad_norm": 1.555354118347168, "loss_scale": 4.0, "retries": 0, "seconds": 475.8, "max_memory_gb": 18.32, "ce": 0.1274777398793958, "rank": 0.0} +{"step": 241, "steps": 500, "lr": 2e-05, "grad_norm": 1.207617163658142, "loss_scale": 4.0, "retries": 0, "seconds": 477.8, "max_memory_gb": 18.32, "ce": 0.1541355245281011, "rank": 0.0} +{"step": 242, "steps": 500, "lr": 2e-05, "grad_norm": 1.1374077796936035, "loss_scale": 4.0, "retries": 0, "seconds": 479.8, "max_memory_gb": 18.32, "ce": 0.1256361790583469, "rank": 0.0} +{"step": 243, "steps": 500, "lr": 2e-05, "grad_norm": 7.983183860778809, "loss_scale": 4.0, "retries": 0, "seconds": 481.8, "max_memory_gb": 18.32, "ce": 0.22932414716342464, "rank": 0.0} +{"step": 244, "steps": 500, "lr": 2e-05, "grad_norm": 2.4249684810638428, "loss_scale": 4.0, "retries": 0, "seconds": 483.8, "max_memory_gb": 18.32, "ce": 0.28142980084521696, "rank": 0.0} +{"step": 245, "steps": 500, "lr": 2e-05, "grad_norm": 1.8929789066314697, "loss_scale": 4.0, "retries": 0, "seconds": 485.8, "max_memory_gb": 18.32, "ce": 0.3735011559911072, "rank": 0.0} +{"step": 246, "steps": 500, "lr": 2e-05, "grad_norm": 0.7038154006004333, "loss_scale": 4.0, "retries": 0, "seconds": 487.9, "max_memory_gb": 18.32, "ce": 0.04011519253253937, "rank": 0.0} +{"step": 247, "steps": 500, "lr": 2e-05, "grad_norm": 17.508787155151367, "loss_scale": 4.0, "retries": 0, "seconds": 489.9, "max_memory_gb": 18.32, "ce": 0.214592304546386, "rank": 0.0} +{"step": 248, "steps": 500, "lr": 2e-05, "grad_norm": 1.5168944597244263, "loss_scale": 4.0, "retries": 0, "seconds": 491.9, "max_memory_gb": 18.32, "ce": 0.11050715344026685, "rank": 0.0} +{"step": 249, "steps": 500, "lr": 2e-05, "grad_norm": 6.776597499847412, "loss_scale": 4.0, "retries": 0, "seconds": 493.9, "max_memory_gb": 18.32, "ce": 0.18714017228921875, "rank": 0.0} +{"step": 250, "steps": 500, "lr": 2e-05, "grad_norm": 1.4693256616592407, "loss_scale": 4.0, "retries": 0, "seconds": 495.9, "max_memory_gb": 18.32, "ce": 0.07148750388296321, "rank": 0.0} +{"step": 251, "steps": 500, "lr": 2e-05, "grad_norm": 1.4010014533996582, "loss_scale": 4.0, "retries": 0, "seconds": 497.9, "max_memory_gb": 18.32, "ce": 0.14796934725018218, "rank": 0.0} +{"step": 252, "steps": 500, "lr": 2e-05, "grad_norm": 1.284011721611023, "loss_scale": 4.0, "retries": 0, "seconds": 499.9, "max_memory_gb": 18.32, "ce": 0.06443219061475247, "rank": 0.0} +{"step": 253, "steps": 500, "lr": 2e-05, "grad_norm": 1.3931092023849487, "loss_scale": 4.0, "retries": 0, "seconds": 502.0, "max_memory_gb": 18.32, "ce": 0.12935581061174162, "rank": 0.0} +{"step": 254, "steps": 500, "lr": 2e-05, "grad_norm": 2.567643880844116, "loss_scale": 4.0, "retries": 0, "seconds": 503.9, "max_memory_gb": 18.32, "ce": 0.1776261607883498, "rank": 0.0} +{"step": 255, "steps": 500, "lr": 2e-05, "grad_norm": 0.9756045341491699, "loss_scale": 4.0, "retries": 0, "seconds": 505.9, "max_memory_gb": 18.32, "ce": 0.20571815373841673, "rank": 0.0} +{"step": 256, "steps": 500, "lr": 2e-05, "grad_norm": 0.741580069065094, "loss_scale": 4.0, "retries": 0, "seconds": 507.9, "max_memory_gb": 18.32, "ce": 0.09114173770649359, "rank": 0.0} +{"step": 257, "steps": 500, "lr": 2e-05, "grad_norm": 0.7456820607185364, "loss_scale": 4.0, "retries": 0, "seconds": 509.9, "max_memory_gb": 18.32, "ce": 0.05943756767373998, "rank": 0.0} +{"step": 258, "steps": 500, "lr": 2e-05, "grad_norm": 2.327420949935913, "loss_scale": 4.0, "retries": 0, "seconds": 511.9, "max_memory_gb": 18.32, "ce": 0.11494681850308552, "rank": 0.0} +{"step": 259, "steps": 500, "lr": 2e-05, "grad_norm": 0.6644304990768433, "loss_scale": 4.0, "retries": 0, "seconds": 513.9, "max_memory_gb": 18.32, "ce": 0.04915992810856551, "rank": 0.0} +{"step": 260, "steps": 500, "lr": 2e-05, "grad_norm": 23.179018020629883, "loss_scale": 4.0, "retries": 0, "seconds": 515.9, "max_memory_gb": 18.32, "ce": 0.17842590413056314, "rank": 0.0} +{"step": 261, "steps": 500, "lr": 2e-05, "grad_norm": 1.6525487899780273, "loss_scale": 4.0, "retries": 0, "seconds": 517.9, "max_memory_gb": 18.32, "ce": 0.14685230585746467, "rank": 0.0} +{"step": 262, "steps": 500, "lr": 2e-05, "grad_norm": 2.9605510234832764, "loss_scale": 4.0, "retries": 0, "seconds": 519.9, "max_memory_gb": 18.32, "ce": 0.21012901212088764, "rank": 0.0} +{"step": 263, "steps": 500, "lr": 2e-05, "grad_norm": 1.919085144996643, "loss_scale": 4.0, "retries": 0, "seconds": 522.0, "max_memory_gb": 18.32, "ce": 0.18336342077236623, "rank": 0.12379581481218338} +{"step": 264, "steps": 500, "lr": 2e-05, "grad_norm": 2.9141290187835693, "loss_scale": 4.0, "retries": 0, "seconds": 524.0, "max_memory_gb": 18.32, "ce": 0.17527730772417272, "rank": 0.0} +{"step": 265, "steps": 500, "lr": 2e-05, "grad_norm": 1.3224388360977173, "loss_scale": 4.0, "retries": 0, "seconds": 526.0, "max_memory_gb": 18.32, "ce": 0.1001119181746617, "rank": 0.0} +{"step": 266, "steps": 500, "lr": 2e-05, "grad_norm": 6.44838809967041, "loss_scale": 4.0, "retries": 0, "seconds": 528.0, "max_memory_gb": 18.32, "ce": 0.3841035859659314, "rank": 0.0} +{"step": 267, "steps": 500, "lr": 2e-05, "grad_norm": 0.8809218406677246, "loss_scale": 4.0, "retries": 0, "seconds": 530.0, "max_memory_gb": 18.32, "ce": 0.1502567313145846, "rank": 0.0} +{"step": 268, "steps": 500, "lr": 2e-05, "grad_norm": 0.8541676998138428, "loss_scale": 4.0, "retries": 0, "seconds": 532.0, "max_memory_gb": 18.32, "ce": 0.2394499988295138, "rank": 0.0} +{"step": 269, "steps": 500, "lr": 2e-05, "grad_norm": 7.899562835693359, "loss_scale": 4.0, "retries": 0, "seconds": 534.0, "max_memory_gb": 18.32, "ce": 0.4617759274551645, "rank": 0.0} +{"step": 270, "steps": 500, "lr": 2e-05, "grad_norm": 5.356691837310791, "loss_scale": 4.0, "retries": 0, "seconds": 536.1, "max_memory_gb": 18.32, "ce": 0.23170086462050676, "rank": 0.0} +{"step": 271, "steps": 500, "lr": 2e-05, "grad_norm": 3.047585964202881, "loss_scale": 4.0, "retries": 0, "seconds": 538.0, "max_memory_gb": 18.32, "ce": 0.0860612440155819, "rank": 0.09109179675579071} +{"step": 272, "steps": 500, "lr": 2e-05, "grad_norm": 2.132499933242798, "loss_scale": 4.0, "retries": 0, "seconds": 540.0, "max_memory_gb": 18.32, "ce": 0.12114530906546861, "rank": 0.0} +{"step": 273, "steps": 500, "lr": 2e-05, "grad_norm": 0.6615415215492249, "loss_scale": 4.0, "retries": 0, "seconds": 542.0, "max_memory_gb": 18.32, "ce": 0.052398961284779944, "rank": 0.12503622472286224} +{"step": 274, "steps": 500, "lr": 2e-05, "grad_norm": 1.1102609634399414, "loss_scale": 4.0, "retries": 0, "seconds": 544.0, "max_memory_gb": 18.32, "ce": 0.12226250092498958, "rank": 0.0} +{"step": 275, "steps": 500, "lr": 2e-05, "grad_norm": 43.46424102783203, "loss_scale": 4.0, "retries": 0, "seconds": 546.1, "max_memory_gb": 18.32, "ce": 1.0293165014591068, "rank": 0.07682323455810547} +{"step": 276, "steps": 500, "lr": 2e-05, "grad_norm": 1.5033987760543823, "loss_scale": 4.0, "retries": 0, "seconds": 548.1, "max_memory_gb": 18.32, "ce": 0.036253019934520125, "rank": 0.0} +{"step": 277, "steps": 500, "lr": 2e-05, "grad_norm": 3.335287094116211, "loss_scale": 4.0, "retries": 0, "seconds": 550.1, "max_memory_gb": 18.32, "ce": 0.05895753559889272, "rank": 0.0} +{"step": 278, "steps": 500, "lr": 2e-05, "grad_norm": 33.64385986328125, "loss_scale": 4.0, "retries": 0, "seconds": 552.1, "max_memory_gb": 18.32, "ce": 0.37040861777495593, "rank": 0.0} +{"step": 279, "steps": 500, "lr": 2e-05, "grad_norm": 53.118534088134766, "loss_scale": 4.0, "retries": 0, "seconds": 554.1, "max_memory_gb": 18.32, "ce": 0.29125072772148997, "rank": 0.0} +{"step": 280, "steps": 500, "lr": 2e-05, "grad_norm": 61.373924255371094, "loss_scale": 4.0, "retries": 0, "seconds": 556.1, "max_memory_gb": 18.32, "ce": 0.34438719393801875, "rank": 0.0} +{"step": 281, "steps": 500, "lr": 2e-05, "grad_norm": 6.790961265563965, "loss_scale": 4.0, "retries": 0, "seconds": 558.1, "max_memory_gb": 18.32, "ce": 0.2408371628553141, "rank": 0.0} +{"step": 282, "steps": 500, "lr": 2e-05, "grad_norm": 0.890654981136322, "loss_scale": 4.0, "retries": 0, "seconds": 560.1, "max_memory_gb": 18.32, "ce": 0.09055286422517383, "rank": 0.0} +{"step": 283, "steps": 500, "lr": 2e-05, "grad_norm": 3.5075361728668213, "loss_scale": 4.0, "retries": 0, "seconds": 562.2, "max_memory_gb": 18.32, "ce": 0.13951554369123187, "rank": 0.0} +{"step": 284, "steps": 500, "lr": 2e-05, "grad_norm": 3.0522372722625732, "loss_scale": 4.0, "retries": 0, "seconds": 564.2, "max_memory_gb": 18.32, "ce": 0.07730247039580718, "rank": 0.0} +{"step": 285, "steps": 500, "lr": 2e-05, "grad_norm": 2.9920759201049805, "loss_scale": 4.0, "retries": 0, "seconds": 566.2, "max_memory_gb": 18.32, "ce": 0.12970330461394042, "rank": 0.0} +{"step": 286, "steps": 500, "lr": 2e-05, "grad_norm": 2.25154185295105, "loss_scale": 4.0, "retries": 0, "seconds": 568.2, "max_memory_gb": 18.32, "ce": 0.06998041772749275, "rank": 0.0} +{"step": 287, "steps": 500, "lr": 2e-05, "grad_norm": 2.5444040298461914, "loss_scale": 4.0, "retries": 0, "seconds": 570.3, "max_memory_gb": 18.32, "ce": 0.13019050130969845, "rank": 0.0} +{"step": 288, "steps": 500, "lr": 2e-05, "grad_norm": 0.3378034830093384, "loss_scale": 4.0, "retries": 0, "seconds": 572.3, "max_memory_gb": 18.32, "ce": 0.02530945466423873, "rank": 0.0} +{"step": 289, "steps": 500, "lr": 2e-05, "grad_norm": 3.257451295852661, "loss_scale": 4.0, "retries": 0, "seconds": 574.3, "max_memory_gb": 18.32, "ce": 0.08624821019475348, "rank": 0.0} +{"step": 290, "steps": 500, "lr": 2e-05, "grad_norm": 1.3711682558059692, "loss_scale": 4.0, "retries": 0, "seconds": 576.3, "max_memory_gb": 18.32, "ce": 0.12799341380014084, "rank": 0.0} +{"step": 291, "steps": 500, "lr": 2e-05, "grad_norm": 0.9788954854011536, "loss_scale": 4.0, "retries": 0, "seconds": 578.3, "max_memory_gb": 18.32, "ce": 0.030357816751347855, "rank": 0.0} +{"step": 292, "steps": 500, "lr": 2e-05, "grad_norm": 19.548234939575195, "loss_scale": 4.0, "retries": 0, "seconds": 580.3, "max_memory_gb": 18.32, "ce": 0.2111393862433033, "rank": 0.0} +{"step": 293, "steps": 500, "lr": 2e-05, "grad_norm": 1.3747295141220093, "loss_scale": 4.0, "retries": 0, "seconds": 582.3, "max_memory_gb": 18.32, "ce": 0.07292757916729897, "rank": 0.0} +{"step": 294, "steps": 500, "lr": 2e-05, "grad_norm": 2.763075351715088, "loss_scale": 4.0, "retries": 0, "seconds": 584.3, "max_memory_gb": 18.32, "ce": 0.06022011331515387, "rank": 0.0} +{"step": 295, "steps": 500, "lr": 2e-05, "grad_norm": 60.32217025756836, "loss_scale": 4.0, "retries": 0, "seconds": 586.3, "max_memory_gb": 18.32, "ce": 0.49701421344070695, "rank": 0.0} +{"step": 296, "steps": 500, "lr": 2e-05, "grad_norm": 68.851806640625, "loss_scale": 4.0, "retries": 0, "seconds": 588.2, "max_memory_gb": 18.32, "ce": 0.4960729400627315, "rank": 0.0} +{"step": 297, "steps": 500, "lr": 2e-05, "grad_norm": 1.6820083856582642, "loss_scale": 4.0, "retries": 0, "seconds": 590.2, "max_memory_gb": 18.32, "ce": 0.1663223776849918, "rank": 0.0} +{"step": 298, "steps": 500, "lr": 2e-05, "grad_norm": 2.904385805130005, "loss_scale": 4.0, "retries": 0, "seconds": 592.2, "max_memory_gb": 18.32, "ce": 0.44581962694064714, "rank": 0.0} +{"step": 299, "steps": 500, "lr": 2e-05, "grad_norm": 1.6823995113372803, "loss_scale": 4.0, "retries": 0, "seconds": 594.2, "max_memory_gb": 18.32, "ce": 0.057210352490074, "rank": 0.0} +{"step": 300, "steps": 500, "lr": 2e-05, "grad_norm": 1.4572477340698242, "loss_scale": 4.0, "retries": 0, "seconds": 596.2, "max_memory_gb": 18.32, "ce": 0.27223619207507, "rank": 0.0} +{"step": 301, "steps": 500, "lr": 2e-05, "grad_norm": 1.3229433298110962, "loss_scale": 4.0, "retries": 0, "seconds": 598.2, "max_memory_gb": 18.32, "ce": 0.13075313704030123, "rank": 0.0} +{"step": 302, "steps": 500, "lr": 2e-05, "grad_norm": 0.5924894213676453, "loss_scale": 4.0, "retries": 0, "seconds": 600.2, "max_memory_gb": 18.32, "ce": 0.03640441919560544, "rank": 0.0} +{"step": 303, "steps": 500, "lr": 2e-05, "grad_norm": 1.0950736999511719, "loss_scale": 4.0, "retries": 0, "seconds": 602.2, "max_memory_gb": 18.32, "ce": 0.04003966005620896, "rank": 0.0} +{"step": 304, "steps": 500, "lr": 2e-05, "grad_norm": 7.408455848693848, "loss_scale": 4.0, "retries": 0, "seconds": 604.2, "max_memory_gb": 18.32, "ce": 0.07769866230955813, "rank": 0.1067851185798645} +{"step": 305, "steps": 500, "lr": 2e-05, "grad_norm": 4.971889972686768, "loss_scale": 4.0, "retries": 0, "seconds": 606.2, "max_memory_gb": 18.32, "ce": 0.16973267099092482, "rank": 0.0} +{"step": 306, "steps": 500, "lr": 2e-05, "grad_norm": 0.9173895716667175, "loss_scale": 4.0, "retries": 0, "seconds": 608.2, "max_memory_gb": 18.32, "ce": 0.04437214523204602, "rank": 0.0} +{"step": 307, "steps": 500, "lr": 2e-05, "grad_norm": 3.007894515991211, "loss_scale": 4.0, "retries": 0, "seconds": 610.2, "max_memory_gb": 18.32, "ce": 0.14089927815075498, "rank": 0.0} +{"step": 308, "steps": 500, "lr": 2e-05, "grad_norm": 65.13755798339844, "loss_scale": 4.0, "retries": 0, "seconds": 612.2, "max_memory_gb": 18.32, "ce": 0.36032689416606445, "rank": 0.0} +{"step": 309, "steps": 500, "lr": 2e-05, "grad_norm": 2.015979051589966, "loss_scale": 4.0, "retries": 0, "seconds": 614.2, "max_memory_gb": 18.32, "ce": 0.2137067253352143, "rank": 0.0} +{"step": 310, "steps": 500, "lr": 2e-05, "grad_norm": 1.9053159952163696, "loss_scale": 4.0, "retries": 0, "seconds": 616.2, "max_memory_gb": 18.32, "ce": 0.06184243036841508, "rank": 0.0} +{"step": 311, "steps": 500, "lr": 2e-05, "grad_norm": 1.0135700702667236, "loss_scale": 4.0, "retries": 0, "seconds": 618.1, "max_memory_gb": 18.32, "ce": 0.1311225546960486, "rank": 0.0} +{"step": 312, "steps": 500, "lr": 2e-05, "grad_norm": 0.8575417399406433, "loss_scale": 4.0, "retries": 0, "seconds": 620.1, "max_memory_gb": 18.32, "ce": 0.03214822869631462, "rank": 0.0} +{"step": 313, "steps": 500, "lr": 2e-05, "grad_norm": 16.46809959411621, "loss_scale": 4.0, "retries": 0, "seconds": 622.2, "max_memory_gb": 18.32, "ce": 0.23890154020045884, "rank": 0.0} +{"step": 314, "steps": 500, "lr": 2e-05, "grad_norm": 17.276445388793945, "loss_scale": 4.0, "retries": 0, "seconds": 624.2, "max_memory_gb": 18.32, "ce": 0.27784919695113786, "rank": 0.0} +{"step": 315, "steps": 500, "lr": 2e-05, "grad_norm": 5.693838596343994, "loss_scale": 4.0, "retries": 0, "seconds": 626.2, "max_memory_gb": 18.32, "ce": 0.08221500365471002, "rank": 0.09389559924602509} +{"step": 316, "steps": 500, "lr": 2e-05, "grad_norm": 9.82160472869873, "loss_scale": 4.0, "retries": 0, "seconds": 628.2, "max_memory_gb": 18.32, "ce": 0.3835287732654251, "rank": 0.0} +{"step": 317, "steps": 500, "lr": 2e-05, "grad_norm": 3.6861209869384766, "loss_scale": 4.0, "retries": 0, "seconds": 630.1, "max_memory_gb": 18.32, "ce": 0.11388032326067332, "rank": 0.0} +{"step": 318, "steps": 500, "lr": 2e-05, "grad_norm": 31.482683181762695, "loss_scale": 4.0, "retries": 0, "seconds": 632.1, "max_memory_gb": 18.32, "ce": 0.46032727765850723, "rank": 0.0} +{"step": 319, "steps": 500, "lr": 2e-05, "grad_norm": 2.0273122787475586, "loss_scale": 4.0, "retries": 0, "seconds": 634.1, "max_memory_gb": 18.32, "ce": 0.07779021468013525, "rank": 0.0} +{"step": 320, "steps": 500, "lr": 2e-05, "grad_norm": 1.6777294874191284, "loss_scale": 4.0, "retries": 0, "seconds": 636.1, "max_memory_gb": 18.32, "ce": 0.11142793094040826, "rank": 0.0} +{"step": 321, "steps": 500, "lr": 2e-05, "grad_norm": 1.4816319942474365, "loss_scale": 4.0, "retries": 0, "seconds": 638.1, "max_memory_gb": 18.32, "ce": 0.10001544406986795, "rank": 0.12575285136699677} +{"step": 322, "steps": 500, "lr": 2e-05, "grad_norm": 1.2917604446411133, "loss_scale": 4.0, "retries": 0, "seconds": 640.1, "max_memory_gb": 18.32, "ce": 0.04548872302984819, "rank": 0.0} +{"step": 323, "steps": 500, "lr": 2e-05, "grad_norm": 3.5057780742645264, "loss_scale": 4.0, "retries": 0, "seconds": 642.1, "max_memory_gb": 18.32, "ce": 0.1471512431744486, "rank": 0.0} +{"step": 324, "steps": 500, "lr": 2e-05, "grad_norm": 5.817412853240967, "loss_scale": 4.0, "retries": 0, "seconds": 644.1, "max_memory_gb": 18.32, "ce": 0.10299634213151876, "rank": 0.0} +{"step": 325, "steps": 500, "lr": 2e-05, "grad_norm": 1.128263235092163, "loss_scale": 4.0, "retries": 0, "seconds": 646.1, "max_memory_gb": 18.32, "ce": 0.21889595792163163, "rank": 0.0} +{"step": 326, "steps": 500, "lr": 2e-05, "grad_norm": 0.8604462742805481, "loss_scale": 4.0, "retries": 0, "seconds": 648.1, "max_memory_gb": 18.32, "ce": 0.06622055145999184, "rank": 0.10452809184789658} +{"step": 327, "steps": 500, "lr": 2e-05, "grad_norm": 5.341226100921631, "loss_scale": 4.0, "retries": 0, "seconds": 650.1, "max_memory_gb": 18.32, "ce": 0.25203566427808255, "rank": 0.0} +{"step": 328, "steps": 500, "lr": 2e-05, "grad_norm": 1.7442829608917236, "loss_scale": 4.0, "retries": 0, "seconds": 652.1, "max_memory_gb": 18.32, "ce": 0.15894845320144668, "rank": 0.0} +{"step": 329, "steps": 500, "lr": 2e-05, "grad_norm": 1.6621477603912354, "loss_scale": 4.0, "retries": 0, "seconds": 654.1, "max_memory_gb": 18.32, "ce": 0.1872725116481888, "rank": 0.0} +{"step": 330, "steps": 500, "lr": 2e-05, "grad_norm": 1.5630313158035278, "loss_scale": 4.0, "retries": 0, "seconds": 656.1, "max_memory_gb": 18.32, "ce": 0.15750927975750528, "rank": 0.0} +{"step": 331, "steps": 500, "lr": 2e-05, "grad_norm": 2.775925874710083, "loss_scale": 4.0, "retries": 0, "seconds": 658.0, "max_memory_gb": 18.32, "ce": 0.3368079394858796, "rank": 0.0} +{"step": 332, "steps": 500, "lr": 2e-05, "grad_norm": 5.554131031036377, "loss_scale": 4.0, "retries": 0, "seconds": 660.1, "max_memory_gb": 18.32, "ce": 0.19765218649990857, "rank": 0.0} +{"step": 333, "steps": 500, "lr": 2e-05, "grad_norm": 1.5927715301513672, "loss_scale": 4.0, "retries": 0, "seconds": 662.0, "max_memory_gb": 18.32, "ce": 0.25135208386927843, "rank": 0.0} +{"step": 334, "steps": 500, "lr": 2e-05, "grad_norm": 1.410292387008667, "loss_scale": 4.0, "retries": 0, "seconds": 664.0, "max_memory_gb": 18.32, "ce": 0.13025387172820047, "rank": 0.0} +{"step": 335, "steps": 500, "lr": 2e-05, "grad_norm": 10.402628898620605, "loss_scale": 4.0, "retries": 0, "seconds": 666.0, "max_memory_gb": 18.32, "ce": 0.3246606072425493, "rank": 0.0} +{"step": 336, "steps": 500, "lr": 2e-05, "grad_norm": 1.8436369895935059, "loss_scale": 4.0, "retries": 0, "seconds": 668.0, "max_memory_gb": 18.32, "ce": 0.05145364985219203, "rank": 0.0} +{"step": 337, "steps": 500, "lr": 2e-05, "grad_norm": 7.5842695236206055, "loss_scale": 4.0, "retries": 0, "seconds": 670.1, "max_memory_gb": 18.32, "ce": 0.271076601697132, "rank": 0.0} +{"step": 338, "steps": 500, "lr": 2e-05, "grad_norm": 2.865802526473999, "loss_scale": 4.0, "retries": 0, "seconds": 672.1, "max_memory_gb": 18.32, "ce": 0.25586786659550853, "rank": 0.0} +{"step": 339, "steps": 500, "lr": 2e-05, "grad_norm": 1.2594376802444458, "loss_scale": 4.0, "retries": 0, "seconds": 674.1, "max_memory_gb": 18.32, "ce": 0.13141067899414338, "rank": 0.0} +{"step": 340, "steps": 500, "lr": 2e-05, "grad_norm": 2.6403961181640625, "loss_scale": 4.0, "retries": 0, "seconds": 676.0, "max_memory_gb": 18.32, "ce": 0.24465272229281254, "rank": 0.0} +{"step": 341, "steps": 500, "lr": 2e-05, "grad_norm": 12.2113618850708, "loss_scale": 4.0, "retries": 0, "seconds": 678.0, "max_memory_gb": 18.32, "ce": 0.22307374617957976, "rank": 0.0} +{"step": 342, "steps": 500, "lr": 2e-05, "grad_norm": 3.150752305984497, "loss_scale": 4.0, "retries": 0, "seconds": 680.0, "max_memory_gb": 18.32, "ce": 0.5051003613043576, "rank": 0.0} +{"step": 343, "steps": 500, "lr": 2e-05, "grad_norm": 3.2979371547698975, "loss_scale": 4.0, "retries": 0, "seconds": 681.8, "max_memory_gb": 18.32, "ce": 0.10471107915509492, "rank": 0.0} +{"step": 344, "steps": 500, "lr": 2e-05, "grad_norm": 1.3369977474212646, "loss_scale": 4.0, "retries": 0, "seconds": 683.7, "max_memory_gb": 18.32, "ce": 0.05076371552422643, "rank": 0.0} +{"step": 345, "steps": 500, "lr": 2e-05, "grad_norm": 1.3543477058410645, "loss_scale": 4.0, "retries": 0, "seconds": 685.5, "max_memory_gb": 18.32, "ce": 0.2025974443749874, "rank": 0.019968107342720032} +{"step": 346, "steps": 500, "lr": 2e-05, "grad_norm": 2.530484914779663, "loss_scale": 4.0, "retries": 0, "seconds": 687.4, "max_memory_gb": 18.32, "ce": 0.229909234229126, "rank": 0.0} +{"step": 347, "steps": 500, "lr": 2e-05, "grad_norm": 3.0163064002990723, "loss_scale": 4.0, "retries": 0, "seconds": 689.2, "max_memory_gb": 18.32, "ce": 0.19166542659513652, "rank": 0.0} +{"step": 348, "steps": 500, "lr": 2e-05, "grad_norm": 0.4379607141017914, "loss_scale": 4.0, "retries": 0, "seconds": 691.1, "max_memory_gb": 18.32, "ce": 0.01797970276675187, "rank": 0.0} +{"step": 349, "steps": 500, "lr": 2e-05, "grad_norm": 0.6398317813873291, "loss_scale": 4.0, "retries": 0, "seconds": 693.0, "max_memory_gb": 18.32, "ce": 0.11512496731302235, "rank": 0.0} +{"step": 350, "steps": 500, "lr": 2e-05, "grad_norm": 1.6669639348983765, "loss_scale": 4.0, "retries": 0, "seconds": 694.8, "max_memory_gb": 18.32, "ce": 0.12067151872906834, "rank": 0.0} +{"step": 351, "steps": 500, "lr": 2e-05, "grad_norm": 4.468695163726807, "loss_scale": 4.0, "retries": 0, "seconds": 696.7, "max_memory_gb": 18.32, "ce": 0.10998644372739363, "rank": 0.0} +{"step": 352, "steps": 500, "lr": 2e-05, "grad_norm": 3.4215071201324463, "loss_scale": 4.0, "retries": 0, "seconds": 698.5, "max_memory_gb": 18.32, "ce": 0.2782285982684698, "rank": 0.0} +{"step": 353, "steps": 500, "lr": 2e-05, "grad_norm": 3.024064779281616, "loss_scale": 4.0, "retries": 0, "seconds": 700.4, "max_memory_gb": 18.32, "ce": 0.14642713347711833, "rank": 0.0} +{"step": 354, "steps": 500, "lr": 2e-05, "grad_norm": 1.2665882110595703, "loss_scale": 4.0, "retries": 0, "seconds": 702.2, "max_memory_gb": 18.32, "ce": 0.16182726650004042, "rank": 0.0} +{"step": 355, "steps": 500, "lr": 2e-05, "grad_norm": 1.0752941370010376, "loss_scale": 4.0, "retries": 0, "seconds": 704.1, "max_memory_gb": 18.32, "ce": 0.12034337850491283, "rank": 0.0} +{"step": 356, "steps": 500, "lr": 2e-05, "grad_norm": 2.3838489055633545, "loss_scale": 4.0, "retries": 0, "seconds": 705.9, "max_memory_gb": 18.32, "ce": 0.05827347660670057, "rank": 0.0} +{"step": 357, "steps": 500, "lr": 2e-05, "grad_norm": 1.2075207233428955, "loss_scale": 4.0, "retries": 0, "seconds": 707.8, "max_memory_gb": 18.32, "ce": 0.05657466965203639, "rank": 0.0} +{"step": 358, "steps": 500, "lr": 2e-05, "grad_norm": 2.046558141708374, "loss_scale": 4.0, "retries": 0, "seconds": 709.7, "max_memory_gb": 18.32, "ce": 0.18638969451421872, "rank": 0.0} +{"step": 359, "steps": 500, "lr": 2e-05, "grad_norm": 2.7547359466552734, "loss_scale": 4.0, "retries": 0, "seconds": 711.5, "max_memory_gb": 18.32, "ce": 0.0648777317837812, "rank": 0.0} +{"step": 360, "steps": 500, "lr": 2e-05, "grad_norm": 1.9320251941680908, "loss_scale": 4.0, "retries": 0, "seconds": 713.4, "max_memory_gb": 18.32, "ce": 0.06102611064852681, "rank": 0.10968498885631561} +{"step": 361, "steps": 500, "lr": 2e-05, "grad_norm": 4.618383884429932, "loss_scale": 4.0, "retries": 0, "seconds": 715.2, "max_memory_gb": 18.32, "ce": 0.13439757454034407, "rank": 0.0} +{"step": 362, "steps": 500, "lr": 2e-05, "grad_norm": 5.323235988616943, "loss_scale": 4.0, "retries": 0, "seconds": 717.1, "max_memory_gb": 18.32, "ce": 0.3148766116646584, "rank": 0.0} +{"step": 363, "steps": 500, "lr": 2e-05, "grad_norm": 1.283575177192688, "loss_scale": 4.0, "retries": 0, "seconds": 718.9, "max_memory_gb": 18.32, "ce": 0.04856027342611924, "rank": 0.0} +{"step": 364, "steps": 500, "lr": 2e-05, "grad_norm": 7.804623603820801, "loss_scale": 4.0, "retries": 0, "seconds": 720.8, "max_memory_gb": 18.32, "ce": 0.10310437015505158, "rank": 0.0} +{"step": 365, "steps": 500, "lr": 2e-05, "grad_norm": 1.3779720067977905, "loss_scale": 4.0, "retries": 0, "seconds": 722.6, "max_memory_gb": 18.32, "ce": 0.12523666008200962, "rank": 0.0} +{"step": 366, "steps": 500, "lr": 2e-05, "grad_norm": 4.366084098815918, "loss_scale": 4.0, "retries": 0, "seconds": 724.5, "max_memory_gb": 18.32, "ce": 0.09374825104896445, "rank": 0.0} +{"step": 367, "steps": 500, "lr": 2e-05, "grad_norm": 1.087996482849121, "loss_scale": 4.0, "retries": 0, "seconds": 726.3, "max_memory_gb": 18.32, "ce": 0.10912680534238461, "rank": 0.0} +{"step": 368, "steps": 500, "lr": 2e-05, "grad_norm": 0.6387014389038086, "loss_scale": 4.0, "retries": 0, "seconds": 728.2, "max_memory_gb": 18.32, "ce": 0.016632158047286794, "rank": 0.0} +{"step": 369, "steps": 500, "lr": 2e-05, "grad_norm": 1.0095981359481812, "loss_scale": 4.0, "retries": 0, "seconds": 730.0, "max_memory_gb": 18.32, "ce": 0.04998865629022475, "rank": 0.11367464810609818} +{"step": 370, "steps": 500, "lr": 2e-05, "grad_norm": 4.9268975257873535, "loss_scale": 4.0, "retries": 0, "seconds": 731.9, "max_memory_gb": 18.32, "ce": 0.3413566849194467, "rank": 0.0} +{"step": 371, "steps": 500, "lr": 2e-05, "grad_norm": 5.254430770874023, "loss_scale": 4.0, "retries": 0, "seconds": 733.8, "max_memory_gb": 18.32, "ce": 0.09901338009512983, "rank": 0.0} +{"step": 372, "steps": 500, "lr": 2e-05, "grad_norm": 1.5094739198684692, "loss_scale": 4.0, "retries": 0, "seconds": 735.6, "max_memory_gb": 18.32, "ce": 0.04182223450334277, "rank": 0.0} +{"step": 373, "steps": 500, "lr": 2e-05, "grad_norm": 1.2158433198928833, "loss_scale": 4.0, "retries": 0, "seconds": 737.5, "max_memory_gb": 18.32, "ce": 0.08686413202667609, "rank": 0.0} +{"step": 374, "steps": 500, "lr": 2e-05, "grad_norm": 1.3456889390945435, "loss_scale": 4.0, "retries": 0, "seconds": 739.3, "max_memory_gb": 18.32, "ce": 0.14972648757247953, "rank": 0.0} +{"step": 375, "steps": 500, "lr": 2e-05, "grad_norm": 1.132346510887146, "loss_scale": 4.0, "retries": 0, "seconds": 741.2, "max_memory_gb": 18.32, "ce": 0.16159306094050407, "rank": 0.0} +{"step": 376, "steps": 500, "lr": 2e-05, "grad_norm": 1.330492615699768, "loss_scale": 4.0, "retries": 0, "seconds": 743.0, "max_memory_gb": 18.32, "ce": 0.05152647625800455, "rank": 0.0} +{"step": 377, "steps": 500, "lr": 2e-05, "grad_norm": 1.1090335845947266, "loss_scale": 4.0, "retries": 0, "seconds": 744.9, "max_memory_gb": 18.32, "ce": 0.08714492178114597, "rank": 0.0} +{"step": 378, "steps": 500, "lr": 2e-05, "grad_norm": 1.901809811592102, "loss_scale": 4.0, "retries": 0, "seconds": 746.7, "max_memory_gb": 18.32, "ce": 0.05313673146883957, "rank": 0.0} +{"step": 379, "steps": 500, "lr": 2e-05, "grad_norm": 1.045945644378662, "loss_scale": 4.0, "retries": 0, "seconds": 748.6, "max_memory_gb": 18.32, "ce": 0.10021686802792829, "rank": 0.0} +{"step": 380, "steps": 500, "lr": 2e-05, "grad_norm": 20.051198959350586, "loss_scale": 4.0, "retries": 0, "seconds": 750.4, "max_memory_gb": 18.32, "ce": 0.3179383509705076, "rank": 0.0} +{"step": 381, "steps": 500, "lr": 2e-05, "grad_norm": 1.1882368326187134, "loss_scale": 4.0, "retries": 0, "seconds": 752.3, "max_memory_gb": 18.32, "ce": 0.2264277936774306, "rank": 0.0} +{"step": 382, "steps": 500, "lr": 2e-05, "grad_norm": 4.196444988250732, "loss_scale": 4.0, "retries": 0, "seconds": 754.1, "max_memory_gb": 18.32, "ce": 0.36449498355796095, "rank": 0.0} +{"step": 383, "steps": 500, "lr": 2e-05, "grad_norm": 0.6328665018081665, "loss_scale": 4.0, "retries": 0, "seconds": 756.0, "max_memory_gb": 18.32, "ce": 0.08901365607744083, "rank": 0.0} +{"step": 384, "steps": 500, "lr": 2e-05, "grad_norm": 4.118411064147949, "loss_scale": 4.0, "retries": 0, "seconds": 757.9, "max_memory_gb": 18.32, "ce": 0.029990170296514407, "rank": 0.0} +{"step": 385, "steps": 500, "lr": 2e-05, "grad_norm": 1.0091607570648193, "loss_scale": 4.0, "retries": 0, "seconds": 759.7, "max_memory_gb": 18.32, "ce": 0.14212490541831357, "rank": 0.0} +{"step": 386, "steps": 500, "lr": 2e-05, "grad_norm": 0.8326770067214966, "loss_scale": 4.0, "retries": 0, "seconds": 761.6, "max_memory_gb": 18.32, "ce": 0.042933885677484795, "rank": 0.0} +{"step": 387, "steps": 500, "lr": 2e-05, "grad_norm": 0.42950236797332764, "loss_scale": 4.0, "retries": 0, "seconds": 763.4, "max_memory_gb": 18.32, "ce": 0.05067586198128993, "rank": 0.0} +{"step": 388, "steps": 500, "lr": 2e-05, "grad_norm": 4.6786699295043945, "loss_scale": 4.0, "retries": 0, "seconds": 765.3, "max_memory_gb": 18.32, "ce": 0.05711905470161582, "rank": 0.0} +{"step": 389, "steps": 500, "lr": 2e-05, "grad_norm": 1.5348392724990845, "loss_scale": 4.0, "retries": 0, "seconds": 767.2, "max_memory_gb": 18.32, "ce": 0.010998683676007204, "rank": 0.0} +{"step": 390, "steps": 500, "lr": 2e-05, "grad_norm": 24.87162971496582, "loss_scale": 4.0, "retries": 0, "seconds": 769.1, "max_memory_gb": 18.32, "ce": 0.09694546183163766, "rank": 0.0} +{"step": 391, "steps": 500, "lr": 2e-05, "grad_norm": 2.0554301738739014, "loss_scale": 4.0, "retries": 0, "seconds": 770.9, "max_memory_gb": 18.32, "ce": 0.23140355321811512, "rank": 0.0} +{"step": 392, "steps": 500, "lr": 2e-05, "grad_norm": 0.13607467710971832, "loss_scale": 4.0, "retries": 0, "seconds": 772.8, "max_memory_gb": 18.32, "ce": 0.0026805151828739326, "rank": 0.12059521675109863} +{"step": 393, "steps": 500, "lr": 2e-05, "grad_norm": 15.830114364624023, "loss_scale": 4.0, "retries": 0, "seconds": 774.7, "max_memory_gb": 18.32, "ce": 0.19604541402077302, "rank": 0.0} +{"step": 394, "steps": 500, "lr": 2e-05, "grad_norm": 2.276203155517578, "loss_scale": 4.0, "retries": 0, "seconds": 776.5, "max_memory_gb": 18.32, "ce": 0.12806066725897836, "rank": 0.0} +{"step": 395, "steps": 500, "lr": 2e-05, "grad_norm": 0.6366990804672241, "loss_scale": 4.0, "retries": 0, "seconds": 778.4, "max_memory_gb": 18.32, "ce": 0.012263593293027952, "rank": 0.0} +{"step": 396, "steps": 500, "lr": 2e-05, "grad_norm": 9.411765098571777, "loss_scale": 4.0, "retries": 0, "seconds": 780.2, "max_memory_gb": 18.32, "ce": 0.10159707104321569, "rank": 0.0} +{"step": 397, "steps": 500, "lr": 2e-05, "grad_norm": 0.8766034245491028, "loss_scale": 4.0, "retries": 0, "seconds": 782.0, "max_memory_gb": 18.32, "ce": 0.10967553619411774, "rank": 0.0} +{"step": 398, "steps": 500, "lr": 2e-05, "grad_norm": 0.6679569482803345, "loss_scale": 4.0, "retries": 0, "seconds": 783.9, "max_memory_gb": 18.32, "ce": 0.06599630247364985, "rank": 0.0} +{"step": 399, "steps": 500, "lr": 2e-05, "grad_norm": 3.6386590003967285, "loss_scale": 4.0, "retries": 0, "seconds": 785.8, "max_memory_gb": 18.32, "ce": 0.18678404362435685, "rank": 0.0} +{"step": 400, "steps": 500, "lr": 2e-05, "grad_norm": 63.70691680908203, "loss_scale": 4.0, "retries": 0, "seconds": 787.7, "max_memory_gb": 18.32, "ce": 0.4351371344382642, "rank": 0.0} +{"step": 401, "steps": 500, "lr": 2e-05, "grad_norm": 2.915407180786133, "loss_scale": 4.0, "retries": 0, "seconds": 789.6, "max_memory_gb": 18.32, "ce": 0.20375454462191556, "rank": 0.0} +{"step": 402, "steps": 500, "lr": 2e-05, "grad_norm": 0.9583258032798767, "loss_scale": 4.0, "retries": 0, "seconds": 791.5, "max_memory_gb": 18.32, "ce": 0.10514037807297427, "rank": 0.09727439284324646} +{"step": 403, "steps": 500, "lr": 2e-05, "grad_norm": 1.274003028869629, "loss_scale": 4.0, "retries": 0, "seconds": 793.3, "max_memory_gb": 18.32, "ce": 0.008250388127635233, "rank": 0.0} +{"step": 404, "steps": 500, "lr": 2e-05, "grad_norm": 2.223012685775757, "loss_scale": 4.0, "retries": 0, "seconds": 795.2, "max_memory_gb": 18.32, "ce": 0.16167420326382853, "rank": 0.0} +{"step": 405, "steps": 500, "lr": 2e-05, "grad_norm": 0.5756174325942993, "loss_scale": 4.0, "retries": 0, "seconds": 797.0, "max_memory_gb": 18.32, "ce": 0.020599196323018987, "rank": 0.0} +{"step": 406, "steps": 500, "lr": 2e-05, "grad_norm": 1.2654783725738525, "loss_scale": 4.0, "retries": 0, "seconds": 798.9, "max_memory_gb": 18.32, "ce": 0.12463225612737006, "rank": 0.0} +{"step": 407, "steps": 500, "lr": 2e-05, "grad_norm": 3.573075532913208, "loss_scale": 4.0, "retries": 0, "seconds": 800.7, "max_memory_gb": 18.32, "ce": 0.7087287303293124, "rank": 0.0011374354362487793} +{"step": 408, "steps": 500, "lr": 2e-05, "grad_norm": 2.3064234256744385, "loss_scale": 4.0, "retries": 0, "seconds": 802.6, "max_memory_gb": 18.32, "ce": 0.32573863239667844, "rank": 0.00253865122795105} +{"step": 409, "steps": 500, "lr": 2e-05, "grad_norm": 1.1342191696166992, "loss_scale": 4.0, "retries": 0, "seconds": 804.5, "max_memory_gb": 18.32, "ce": 0.14280446498014499, "rank": 0.0} +{"step": 410, "steps": 500, "lr": 2e-05, "grad_norm": 5.95489501953125, "loss_scale": 4.0, "retries": 0, "seconds": 806.3, "max_memory_gb": 18.32, "ce": 0.0710164344200166, "rank": 0.0} +{"step": 411, "steps": 500, "lr": 2e-05, "grad_norm": 6.5235724449157715, "loss_scale": 4.0, "retries": 0, "seconds": 808.2, "max_memory_gb": 18.32, "ce": 0.23412458106031409, "rank": 0.0} +{"step": 412, "steps": 500, "lr": 2e-05, "grad_norm": 0.7048159241676331, "loss_scale": 4.0, "retries": 0, "seconds": 810.1, "max_memory_gb": 18.32, "ce": 0.15214904741151258, "rank": 0.1248396560549736} +{"step": 413, "steps": 500, "lr": 2e-05, "grad_norm": 0.9268092513084412, "loss_scale": 4.0, "retries": 0, "seconds": 811.9, "max_memory_gb": 18.32, "ce": 0.030789581200224347, "rank": 0.0} +{"step": 414, "steps": 500, "lr": 2e-05, "grad_norm": 3.0443146228790283, "loss_scale": 4.0, "retries": 0, "seconds": 813.8, "max_memory_gb": 18.32, "ce": 0.2093742019496858, "rank": 0.0} +{"step": 415, "steps": 500, "lr": 2e-05, "grad_norm": 8.308897972106934, "loss_scale": 4.0, "retries": 0, "seconds": 815.7, "max_memory_gb": 18.32, "ce": 0.3444577584923536, "rank": 0.0} +{"step": 416, "steps": 500, "lr": 2e-05, "grad_norm": 1.0446562767028809, "loss_scale": 4.0, "retries": 0, "seconds": 817.5, "max_memory_gb": 18.32, "ce": 0.012820026597182732, "rank": 0.0} +{"step": 417, "steps": 500, "lr": 2e-05, "grad_norm": 6.361216068267822, "loss_scale": 4.0, "retries": 0, "seconds": 819.4, "max_memory_gb": 18.32, "ce": 0.4757749334385153, "rank": 0.1301356703042984} +{"step": 418, "steps": 500, "lr": 2e-05, "grad_norm": 2.547196388244629, "loss_scale": 4.0, "retries": 0, "seconds": 821.2, "max_memory_gb": 18.32, "ce": 0.055346002482110634, "rank": 0.0} +{"step": 419, "steps": 500, "lr": 2e-05, "grad_norm": 0.6184739470481873, "loss_scale": 4.0, "retries": 0, "seconds": 823.0, "max_memory_gb": 18.32, "ce": 0.0711411957454402, "rank": 0.0} +{"step": 420, "steps": 500, "lr": 2e-05, "grad_norm": 81.20502471923828, "loss_scale": 4.0, "retries": 0, "seconds": 824.9, "max_memory_gb": 18.32, "ce": 0.8358905529603362, "rank": 0.0} +{"step": 421, "steps": 500, "lr": 2e-05, "grad_norm": 0.7460228204727173, "loss_scale": 4.0, "retries": 0, "seconds": 826.7, "max_memory_gb": 18.32, "ce": 0.031110698848351603, "rank": 0.12137903273105621} +{"step": 422, "steps": 500, "lr": 2e-05, "grad_norm": 111.32431030273438, "loss_scale": 4.0, "retries": 0, "seconds": 828.6, "max_memory_gb": 18.32, "ce": 0.5768692329329497, "rank": 0.0} +{"step": 423, "steps": 500, "lr": 2e-05, "grad_norm": 11.707991600036621, "loss_scale": 4.0, "retries": 0, "seconds": 830.5, "max_memory_gb": 18.32, "ce": 0.3487719491822645, "rank": 0.0} +{"step": 424, "steps": 500, "lr": 2e-05, "grad_norm": 0.8171172142028809, "loss_scale": 4.0, "retries": 0, "seconds": 832.4, "max_memory_gb": 18.32, "ce": 0.031738551162561635, "rank": 0.0} +{"step": 425, "steps": 500, "lr": 2e-05, "grad_norm": 0.9364499449729919, "loss_scale": 4.0, "retries": 0, "seconds": 834.2, "max_memory_gb": 18.32, "ce": 0.03251645948330406, "rank": 0.0} +{"step": 426, "steps": 500, "lr": 2e-05, "grad_norm": 2.2963128089904785, "loss_scale": 4.0, "retries": 0, "seconds": 836.1, "max_memory_gb": 18.32, "ce": 0.20323645905591547, "rank": 0.0} +{"step": 427, "steps": 500, "lr": 2e-05, "grad_norm": 1.5781452655792236, "loss_scale": 4.0, "retries": 0, "seconds": 838.0, "max_memory_gb": 18.32, "ce": 0.0645869856889476, "rank": 0.0} +{"step": 428, "steps": 500, "lr": 2e-05, "grad_norm": 1.9776644706726074, "loss_scale": 4.0, "retries": 0, "seconds": 839.9, "max_memory_gb": 18.32, "ce": 0.06647106927630375, "rank": 0.0} +{"step": 429, "steps": 500, "lr": 2e-05, "grad_norm": 0.857224702835083, "loss_scale": 4.0, "retries": 0, "seconds": 841.8, "max_memory_gb": 18.32, "ce": 0.019384552571864333, "rank": 0.0} +{"step": 430, "steps": 500, "lr": 2e-05, "grad_norm": 1.2609111070632935, "loss_scale": 4.0, "retries": 0, "seconds": 843.6, "max_memory_gb": 18.32, "ce": 0.10478501736724866, "rank": 0.0} +{"step": 431, "steps": 500, "lr": 2e-05, "grad_norm": 7.5254645347595215, "loss_scale": 4.0, "retries": 0, "seconds": 845.5, "max_memory_gb": 18.32, "ce": 0.2305755913912435, "rank": 0.0} +{"step": 432, "steps": 500, "lr": 2e-05, "grad_norm": 0.9913309216499329, "loss_scale": 4.0, "retries": 0, "seconds": 847.4, "max_memory_gb": 18.32, "ce": 0.07458538375794888, "rank": 0.0} +{"step": 433, "steps": 500, "lr": 2e-05, "grad_norm": 3.683805465698242, "loss_scale": 4.0, "retries": 0, "seconds": 849.2, "max_memory_gb": 18.32, "ce": 0.07348276400443865, "rank": 0.0} +{"step": 434, "steps": 500, "lr": 2e-05, "grad_norm": 0.24994565546512604, "loss_scale": 4.0, "retries": 0, "seconds": 851.1, "max_memory_gb": 18.32, "ce": 0.0038632935829809867, "rank": 0.0} +{"step": 435, "steps": 500, "lr": 2e-05, "grad_norm": 2.9372305870056152, "loss_scale": 4.0, "retries": 0, "seconds": 852.9, "max_memory_gb": 18.32, "ce": 0.16044287073236774, "rank": 0.0} +{"step": 436, "steps": 500, "lr": 2e-05, "grad_norm": 3.7193641662597656, "loss_scale": 4.0, "retries": 0, "seconds": 854.8, "max_memory_gb": 18.32, "ce": 0.14555093147646403, "rank": 0.0} +{"step": 437, "steps": 500, "lr": 2e-05, "grad_norm": 11.26116943359375, "loss_scale": 4.0, "retries": 0, "seconds": 856.7, "max_memory_gb": 18.32, "ce": 0.34988730886834674, "rank": 0.09015437960624695} +{"step": 438, "steps": 500, "lr": 2e-05, "grad_norm": 2.5579915046691895, "loss_scale": 4.0, "retries": 0, "seconds": 858.6, "max_memory_gb": 18.32, "ce": 0.047992777504987316, "rank": 0.0} +{"step": 439, "steps": 500, "lr": 2e-05, "grad_norm": 2.7298905849456787, "loss_scale": 4.0, "retries": 0, "seconds": 860.4, "max_memory_gb": 18.32, "ce": 0.18207845788856503, "rank": 0.0} +{"step": 440, "steps": 500, "lr": 2e-05, "grad_norm": 0.5450428128242493, "loss_scale": 4.0, "retries": 0, "seconds": 862.3, "max_memory_gb": 18.32, "ce": 0.0318335528827447, "rank": 0.0} +{"step": 441, "steps": 500, "lr": 2e-05, "grad_norm": 15.294699668884277, "loss_scale": 4.0, "retries": 0, "seconds": 864.2, "max_memory_gb": 18.32, "ce": 0.36830080379513674, "rank": 0.0} +{"step": 442, "steps": 500, "lr": 2e-05, "grad_norm": 100.43207550048828, "loss_scale": 4.0, "retries": 0, "seconds": 866.0, "max_memory_gb": 18.32, "ce": 0.6196058047826227, "rank": 0.0} +{"step": 443, "steps": 500, "lr": 2e-05, "grad_norm": 1.0022412538528442, "loss_scale": 4.0, "retries": 0, "seconds": 867.9, "max_memory_gb": 18.32, "ce": 0.05571861167845782, "rank": 0.0} +{"step": 444, "steps": 500, "lr": 2e-05, "grad_norm": 1.5451533794403076, "loss_scale": 4.0, "retries": 0, "seconds": 869.7, "max_memory_gb": 18.32, "ce": 0.10726810334745096, "rank": 0.0} +{"step": 445, "steps": 500, "lr": 2e-05, "grad_norm": 2.522768974304199, "loss_scale": 4.0, "retries": 0, "seconds": 871.6, "max_memory_gb": 18.32, "ce": 0.3847708854736993, "rank": 0.08791611343622208} +{"step": 446, "steps": 500, "lr": 2e-05, "grad_norm": 13.316143035888672, "loss_scale": 4.0, "retries": 0, "seconds": 873.4, "max_memory_gb": 18.32, "ce": 0.16765869465234573, "rank": 0.0} +{"step": 447, "steps": 500, "lr": 2e-05, "grad_norm": 1.2975585460662842, "loss_scale": 4.0, "retries": 0, "seconds": 875.3, "max_memory_gb": 18.32, "ce": 0.18280871264141751, "rank": 0.0} +{"step": 448, "steps": 500, "lr": 2e-05, "grad_norm": 4.786685466766357, "loss_scale": 4.0, "retries": 0, "seconds": 877.1, "max_memory_gb": 18.32, "ce": 0.16042338998886407, "rank": 0.0} +{"step": 449, "steps": 500, "lr": 2e-05, "grad_norm": 0.7866418957710266, "loss_scale": 4.0, "retries": 0, "seconds": 878.9, "max_memory_gb": 18.32, "ce": 0.012104147670470411, "rank": 0.0} +{"step": 450, "steps": 500, "lr": 2e-05, "grad_norm": 4.129992485046387, "loss_scale": 4.0, "retries": 0, "seconds": 880.8, "max_memory_gb": 18.32, "ce": 0.09430640132268309, "rank": 0.0} +{"step": 451, "steps": 500, "lr": 2e-05, "grad_norm": 1.896761417388916, "loss_scale": 4.0, "retries": 0, "seconds": 882.6, "max_memory_gb": 18.32, "ce": 0.0500697109146131, "rank": 0.0} +{"step": 452, "steps": 500, "lr": 2e-05, "grad_norm": 2.034409284591675, "loss_scale": 4.0, "retries": 0, "seconds": 884.5, "max_memory_gb": 18.32, "ce": 0.13700743823392258, "rank": 0.0} +{"step": 453, "steps": 500, "lr": 2e-05, "grad_norm": 1.9996522665023804, "loss_scale": 4.0, "retries": 0, "seconds": 886.3, "max_memory_gb": 18.32, "ce": 0.031737541967231664, "rank": 0.0} +{"step": 454, "steps": 500, "lr": 2e-05, "grad_norm": 5.249238967895508, "loss_scale": 4.0, "retries": 0, "seconds": 888.2, "max_memory_gb": 18.32, "ce": 0.12203576967658591, "rank": 0.0} +{"step": 455, "steps": 500, "lr": 2e-05, "grad_norm": 7.47859525680542, "loss_scale": 4.0, "retries": 0, "seconds": 890.0, "max_memory_gb": 18.32, "ce": 0.10704889037515386, "rank": 0.0} +{"step": 456, "steps": 500, "lr": 2e-05, "grad_norm": 0.6267043352127075, "loss_scale": 4.0, "retries": 0, "seconds": 891.9, "max_memory_gb": 18.32, "ce": 0.02209775357187027, "rank": 0.0} +{"step": 457, "steps": 500, "lr": 2e-05, "grad_norm": 1.6139395236968994, "loss_scale": 4.0, "retries": 0, "seconds": 893.7, "max_memory_gb": 18.32, "ce": 0.04166128612632747, "rank": 0.06664017587900162} +{"step": 458, "steps": 500, "lr": 2e-05, "grad_norm": 1.888211965560913, "loss_scale": 4.0, "retries": 0, "seconds": 895.6, "max_memory_gb": 18.32, "ce": 0.18846716669213492, "rank": 0.0} +{"step": 459, "steps": 500, "lr": 2e-05, "grad_norm": 3.091186761856079, "loss_scale": 4.0, "retries": 0, "seconds": 897.4, "max_memory_gb": 18.32, "ce": 0.1135401524916233, "rank": 0.0} +{"step": 460, "steps": 500, "lr": 2e-05, "grad_norm": 0.41429588198661804, "loss_scale": 4.0, "retries": 0, "seconds": 899.3, "max_memory_gb": 18.32, "ce": 0.016816362611280056, "rank": 0.0} +{"step": 461, "steps": 500, "lr": 2e-05, "grad_norm": 4.045022010803223, "loss_scale": 4.0, "retries": 0, "seconds": 901.2, "max_memory_gb": 18.32, "ce": 0.2337250613127253, "rank": 0.0} +{"step": 462, "steps": 500, "lr": 2e-05, "grad_norm": 0.412291020154953, "loss_scale": 4.0, "retries": 0, "seconds": 903.0, "max_memory_gb": 18.32, "ce": 0.020483696036535548, "rank": 0.0} +{"step": 463, "steps": 500, "lr": 2e-05, "grad_norm": 0.6212554574012756, "loss_scale": 4.0, "retries": 0, "seconds": 904.8, "max_memory_gb": 18.32, "ce": 0.018838394942576997, "rank": 0.0} +{"step": 464, "steps": 500, "lr": 2e-05, "grad_norm": 0.08977308124303818, "loss_scale": 4.0, "retries": 0, "seconds": 906.7, "max_memory_gb": 18.32, "ce": 0.0017174440108647104, "rank": 0.0} +{"step": 465, "steps": 500, "lr": 2e-05, "grad_norm": 3.6625192165374756, "loss_scale": 4.0, "retries": 0, "seconds": 908.6, "max_memory_gb": 18.32, "ce": 0.23089391088069533, "rank": 0.0} +{"step": 466, "steps": 500, "lr": 2e-05, "grad_norm": 5.116847038269043, "loss_scale": 4.0, "retries": 0, "seconds": 910.5, "max_memory_gb": 18.32, "ce": 0.11398743780091536, "rank": 0.0} +{"step": 467, "steps": 500, "lr": 2e-05, "grad_norm": 0.686377227306366, "loss_scale": 4.0, "retries": 0, "seconds": 912.3, "max_memory_gb": 18.32, "ce": 0.027694734915712615, "rank": 0.0} +{"step": 468, "steps": 500, "lr": 2e-05, "grad_norm": 20.606060028076172, "loss_scale": 4.0, "retries": 0, "seconds": 914.2, "max_memory_gb": 18.32, "ce": 0.05036721492979268, "rank": 0.0} +{"step": 469, "steps": 500, "lr": 2e-05, "grad_norm": 6.442201137542725, "loss_scale": 4.0, "retries": 0, "seconds": 916.0, "max_memory_gb": 18.32, "ce": 0.19145178466715151, "rank": 0.0} +{"step": 470, "steps": 500, "lr": 2e-05, "grad_norm": 3.791067600250244, "loss_scale": 4.0, "retries": 0, "seconds": 917.9, "max_memory_gb": 18.32, "ce": 0.08856109444968752, "rank": 0.0} +{"step": 471, "steps": 500, "lr": 2e-05, "grad_norm": 2.406858205795288, "loss_scale": 4.0, "retries": 0, "seconds": 919.8, "max_memory_gb": 18.32, "ce": 0.21193032752125873, "rank": 0.0} +{"step": 472, "steps": 500, "lr": 2e-05, "grad_norm": 3.8990447521209717, "loss_scale": 4.0, "retries": 0, "seconds": 921.6, "max_memory_gb": 18.32, "ce": 0.039716320414299844, "rank": 0.0} +{"step": 473, "steps": 500, "lr": 2e-05, "grad_norm": 1.1250369548797607, "loss_scale": 4.0, "retries": 0, "seconds": 923.5, "max_memory_gb": 18.32, "ce": 0.020112353857257403, "rank": 0.13018907606601715} +{"step": 474, "steps": 500, "lr": 2e-05, "grad_norm": 1.6822078227996826, "loss_scale": 4.0, "retries": 0, "seconds": 925.4, "max_memory_gb": 18.32, "ce": 0.10308574510418111, "rank": 0.0} +{"step": 475, "steps": 500, "lr": 2e-05, "grad_norm": 2.197598934173584, "loss_scale": 4.0, "retries": 0, "seconds": 927.2, "max_memory_gb": 18.32, "ce": 0.1576354868739145, "rank": 0.0} +{"step": 476, "steps": 500, "lr": 2e-05, "grad_norm": 0.8005141019821167, "loss_scale": 4.0, "retries": 0, "seconds": 929.1, "max_memory_gb": 18.32, "ce": 0.012933484467794187, "rank": 0.0} +{"step": 477, "steps": 500, "lr": 2e-05, "grad_norm": 2.8066327571868896, "loss_scale": 4.0, "retries": 0, "seconds": 930.9, "max_memory_gb": 18.32, "ce": 0.05280273025528004, "rank": 0.0} +{"step": 478, "steps": 500, "lr": 2e-05, "grad_norm": 1.42909836769104, "loss_scale": 4.0, "retries": 0, "seconds": 932.8, "max_memory_gb": 18.32, "ce": 0.07217364780626667, "rank": 0.0} +{"step": 479, "steps": 500, "lr": 2e-05, "grad_norm": 7.7293524742126465, "loss_scale": 4.0, "retries": 0, "seconds": 934.7, "max_memory_gb": 18.32, "ce": 0.3064040273493447, "rank": 0.03196847438812256} +{"step": 480, "steps": 500, "lr": 2e-05, "grad_norm": 1.1562916040420532, "loss_scale": 4.0, "retries": 0, "seconds": 936.6, "max_memory_gb": 18.32, "ce": 0.14306367603785475, "rank": 0.0} +{"step": 481, "steps": 500, "lr": 2e-05, "grad_norm": 30.019399642944336, "loss_scale": 4.0, "retries": 0, "seconds": 938.4, "max_memory_gb": 18.32, "ce": 0.08255708100659831, "rank": 0.0} +{"step": 482, "steps": 500, "lr": 2e-05, "grad_norm": 48.690364837646484, "loss_scale": 4.0, "retries": 0, "seconds": 940.3, "max_memory_gb": 18.32, "ce": 0.19119226153088675, "rank": 0.0} +{"step": 483, "steps": 500, "lr": 2e-05, "grad_norm": 14.279749870300293, "loss_scale": 4.0, "retries": 0, "seconds": 942.1, "max_memory_gb": 18.32, "ce": 0.6669459512359026, "rank": 0.0} +{"step": 484, "steps": 500, "lr": 2e-05, "grad_norm": 3.214646577835083, "loss_scale": 4.0, "retries": 0, "seconds": 944.0, "max_memory_gb": 18.32, "ce": 0.45599966888767085, "rank": 0.0} +{"step": 485, "steps": 500, "lr": 2e-05, "grad_norm": 1.3820898532867432, "loss_scale": 4.0, "retries": 0, "seconds": 945.9, "max_memory_gb": 18.32, "ce": 0.05658761411177693, "rank": 0.0} +{"step": 486, "steps": 500, "lr": 2e-05, "grad_norm": 5.373387813568115, "loss_scale": 4.0, "retries": 0, "seconds": 947.7, "max_memory_gb": 18.32, "ce": 0.1835441409421037, "rank": 0.0} +{"step": 487, "steps": 500, "lr": 2e-05, "grad_norm": 13.985042572021484, "loss_scale": 4.0, "retries": 0, "seconds": 949.5, "max_memory_gb": 18.32, "ce": 0.09237184717858327, "rank": 0.0} +{"step": 488, "steps": 500, "lr": 2e-05, "grad_norm": 1.3058735132217407, "loss_scale": 4.0, "retries": 0, "seconds": 951.4, "max_memory_gb": 18.32, "ce": 0.07405358527830685, "rank": 0.0} +{"step": 489, "steps": 500, "lr": 2e-05, "grad_norm": 10.30135726928711, "loss_scale": 4.0, "retries": 0, "seconds": 953.3, "max_memory_gb": 18.32, "ce": 0.3614045844642533, "rank": 0.0} +{"step": 490, "steps": 500, "lr": 2e-05, "grad_norm": 0.9895980954170227, "loss_scale": 4.0, "retries": 0, "seconds": 955.1, "max_memory_gb": 18.32, "ce": 0.028921001006892766, "rank": 0.0} +{"step": 491, "steps": 500, "lr": 2e-05, "grad_norm": 1.212921142578125, "loss_scale": 4.0, "retries": 0, "seconds": 957.0, "max_memory_gb": 18.32, "ce": 0.1251104861003114, "rank": 0.0} +{"step": 492, "steps": 500, "lr": 2e-05, "grad_norm": 3.3107752799987793, "loss_scale": 4.0, "retries": 0, "seconds": 958.8, "max_memory_gb": 18.32, "ce": 0.01887664281821344, "rank": 0.0} +{"step": 493, "steps": 500, "lr": 2e-05, "grad_norm": 2.2011163234710693, "loss_scale": 4.0, "retries": 0, "seconds": 960.7, "max_memory_gb": 18.32, "ce": 0.06549296791854431, "rank": 0.0} +{"step": 494, "steps": 500, "lr": 2e-05, "grad_norm": 2.4016616344451904, "loss_scale": 4.0, "retries": 0, "seconds": 962.6, "max_memory_gb": 18.32, "ce": 0.08476316474479972, "rank": 0.0} +{"step": 495, "steps": 500, "lr": 2e-05, "grad_norm": 1.0660969018936157, "loss_scale": 4.0, "retries": 0, "seconds": 964.4, "max_memory_gb": 18.32, "ce": 0.022967322507611243, "rank": 0.0} +{"step": 496, "steps": 500, "lr": 2e-05, "grad_norm": 1.285075306892395, "loss_scale": 4.0, "retries": 0, "seconds": 966.3, "max_memory_gb": 18.32, "ce": 0.020212087934851297, "rank": 0.0} +{"step": 497, "steps": 500, "lr": 2e-05, "grad_norm": 2.948519468307495, "loss_scale": 4.0, "retries": 0, "seconds": 968.2, "max_memory_gb": 18.32, "ce": 0.1690991309187666, "rank": 0.0} +{"step": 498, "steps": 500, "lr": 2e-05, "grad_norm": 2.992276668548584, "loss_scale": 4.0, "retries": 0, "seconds": 970.0, "max_memory_gb": 18.32, "ce": 0.14590169633811456, "rank": 0.0} +{"step": 499, "steps": 500, "lr": 2e-05, "grad_norm": 3.8788278102874756, "loss_scale": 4.0, "retries": 0, "seconds": 971.9, "max_memory_gb": 18.32, "ce": 0.20928585755609674, "rank": 0.0} +{"step": 500, "steps": 500, "lr": 2e-05, "grad_norm": 3.140929937362671, "loss_scale": 4.0, "retries": 0, "seconds": 973.7, "max_memory_gb": 18.32, "ce": 0.21855640826470335, "rank": 0.0} diff --git a/runs/single_llava/removal/L02_s13/adapter/README.md b/runs/single_llava/removal/L02_s13/adapter/README.md new file mode 100644 index 0000000000000000000000000000000000000000..ae587920ca55972d94e38f783837f71be14f1dda --- /dev/null +++ b/runs/single_llava/removal/L02_s13/adapter/README.md @@ -0,0 +1,207 @@ +--- +base_model: /data/OPSD/assets/models/llava-1.5-7b-hf +library_name: peft +pipeline_tag: text-generation +tags: +- base_model:adapter:/data/OPSD/assets/models/llava-1.5-7b-hf +- lora +- transformers +--- + +# Model Card for Model ID + + + + + +## Model Details + +### Model Description + + + + + +- **Developed by:** [More Information Needed] +- **Funded by [optional]:** [More Information Needed] +- **Shared by [optional]:** [More Information Needed] +- **Model type:** [More Information Needed] +- **Language(s) (NLP):** [More Information Needed] +- **License:** [More Information Needed] +- **Finetuned from model [optional]:** [More Information Needed] + +### Model Sources [optional] + + + +- **Repository:** [More Information Needed] +- **Paper [optional]:** [More Information Needed] +- **Demo [optional]:** [More Information Needed] + +## Uses + + + +### Direct Use + + + +[More Information Needed] + +### Downstream Use [optional] + + + +[More Information Needed] + +### Out-of-Scope Use + + + +[More Information Needed] + +## Bias, Risks, and Limitations + + + +[More Information Needed] + +### Recommendations + + + +Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. More information needed for further recommendations. + +## How to Get Started with the Model + +Use the code below to get started with the model. + +[More Information Needed] + +## Training Details + +### Training Data + + + +[More Information Needed] + +### Training Procedure + + + +#### Preprocessing [optional] + +[More Information Needed] + + +#### Training Hyperparameters + +- **Training regime:** [More Information Needed] + +#### Speeds, Sizes, Times [optional] + + + +[More Information Needed] + +## Evaluation + + + +### Testing Data, Factors & Metrics + +#### Testing Data + + + +[More Information Needed] + +#### Factors + + + +[More Information Needed] + +#### Metrics + + + +[More Information Needed] + +### Results + +[More Information Needed] + +#### Summary + + + +## Model Examination [optional] + + + +[More Information Needed] + +## Environmental Impact + + + +Carbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700). + +- **Hardware Type:** [More Information Needed] +- **Hours used:** [More Information Needed] +- **Cloud Provider:** [More Information Needed] +- **Compute Region:** [More Information Needed] +- **Carbon Emitted:** [More Information Needed] + +## Technical Specifications [optional] + +### Model Architecture and Objective + +[More Information Needed] + +### Compute Infrastructure + +[More Information Needed] + +#### Hardware + +[More Information Needed] + +#### Software + +[More Information Needed] + +## Citation [optional] + + + +**BibTeX:** + +[More Information Needed] + +**APA:** + +[More Information Needed] + +## Glossary [optional] + + + +[More Information Needed] + +## More Information [optional] + +[More Information Needed] + +## Model Card Authors [optional] + +[More Information Needed] + +## Model Card Contact + +[More Information Needed] +### Framework versions + +- PEFT 0.20.0 \ No newline at end of file diff --git a/runs/single_llava/removal/L02_s13/adapter/adapter_config.json b/runs/single_llava/removal/L02_s13/adapter/adapter_config.json new file mode 100644 index 0000000000000000000000000000000000000000..47d53563b5a747a8b706b98e7c63097bfee74fe4 --- /dev/null +++ b/runs/single_llava/removal/L02_s13/adapter/adapter_config.json @@ -0,0 +1,50 @@ +{ + "alora_invocation_tokens": null, + "alpha_pattern": {}, + "arrow_config": null, + "auto_mapping": null, + "base_model_name_or_path": "/data/OPSD/assets/models/llava-1.5-7b-hf", + "bias": "none", + "corda_config": null, + "ensure_weight_tying": false, + "eva_config": null, + "exclude_modules": null, + "fan_in_fan_out": false, + "inference_mode": true, + "init_lora_weights": true, + "layer_replication": null, + "layers_pattern": null, + "layers_to_transform": null, + "loftq_config": {}, + "lora_alpha": 32, + "lora_bias": false, + "lora_dropout": 0.05, + "lora_ga_config": null, + "megatron_config": null, + "megatron_core": "megatron.core", + "modules_to_save": null, + "monteclora_config": null, + "peft_type": "LORA", + "peft_version": "0.20.0", + "qalora_group_size": 16, + "r": 16, + "rank_pattern": {}, + "revision": null, + "target_modules": [ + "model.language_model.layers.2.self_attn.v_proj", + "model.language_model.layers.2.mlp.down_proj", + "model.language_model.layers.2.self_attn.o_proj", + "model.language_model.layers.2.self_attn.q_proj", + "model.language_model.layers.2.self_attn.k_proj", + "model.language_model.layers.2.mlp.up_proj", + "model.language_model.layers.2.mlp.gate_proj" + ], + "target_parameters": null, + "task_type": "CAUSAL_LM", + "trainable_token_indices": null, + "use_bdlora": null, + "use_dora": false, + "use_qalora": false, + "use_rslora": false, + "velora_config": null +} \ No newline at end of file diff --git a/runs/single_llava/removal/L02_s13/config.json b/runs/single_llava/removal/L02_s13/config.json new file mode 100644 index 0000000000000000000000000000000000000000..b22e2608ba8aff8dbd4e784402edd5c8f644c956 --- /dev/null +++ b/runs/single_llava/removal/L02_s13/config.json @@ -0,0 +1,24 @@ +{ + "behavior": "removal", + "seed": 13, + "model": "/data/OPSD/assets/models/llava-1.5-7b-hf", + "backbone": "llava", + "train": "/data/OPSD/assets/v3/removal.jsonl", + "items": 1000, + "steps": 500, + "epochs": 2, + "steps_per_epoch": 250, + "groups_per_step": 4, + "lr": 2e-05, + "scheduler": "constant", + "warmup_steps": 0, + "final_fraction": 1.0, + "weight_decay": 0.0, + "clip": 1.0, + "initial_loss_scale": 4.0, + "adapter": { + "layer": 2, + "rank": 16, + "trainable_parameters": 1249280 + } +} \ No newline at end of file diff --git a/runs/single_llava/removal/L02_s13/steps.jsonl b/runs/single_llava/removal/L02_s13/steps.jsonl new file mode 100644 index 0000000000000000000000000000000000000000..9adfa913d80dd90d57d735c1853c03feb5e08a6f --- /dev/null +++ b/runs/single_llava/removal/L02_s13/steps.jsonl @@ -0,0 +1,500 @@ +{"step": 1, "steps": 500, "lr": 2e-05, "grad_norm": 0.1418166607618332, "loss_scale": 4.0, "retries": 0, "seconds": 3.0, "max_memory_gb": 18.13, "ce": 1.6287612388841808, "rank": 0.13243506848812103} +{"step": 2, "steps": 500, "lr": 2e-05, "grad_norm": 0.2339147925376892, "loss_scale": 4.0, "retries": 0, "seconds": 5.0, "max_memory_gb": 18.15, "ce": 2.674810584168881, "rank": 0.03323686122894287} +{"step": 3, "steps": 500, "lr": 2e-05, "grad_norm": 0.528289794921875, "loss_scale": 4.0, "retries": 0, "seconds": 7.0, "max_memory_gb": 18.15, "ce": 5.2766911424696445, "rank": 0.14882218837738037} +{"step": 4, "steps": 500, "lr": 2e-05, "grad_norm": 0.3117513656616211, "loss_scale": 4.0, "retries": 0, "seconds": 9.1, "max_memory_gb": 18.15, "ce": 3.802738910075277, "rank": 0.12371551990509033} +{"step": 5, "steps": 500, "lr": 2e-05, "grad_norm": 0.04703275486826897, "loss_scale": 4.0, "retries": 0, "seconds": 11.0, "max_memory_gb": 18.15, "ce": 0.1396179653238505, "rank": 0.0} +{"step": 6, "steps": 500, "lr": 2e-05, "grad_norm": 0.13196274638175964, "loss_scale": 4.0, "retries": 0, "seconds": 13.0, "max_memory_gb": 18.15, "ce": 1.5126409735530615, "rank": 0.0} +{"step": 7, "steps": 500, "lr": 2e-05, "grad_norm": 0.3665620982646942, "loss_scale": 4.0, "retries": 0, "seconds": 15.1, "max_memory_gb": 18.15, "ce": 3.4593343855813146, "rank": 0.02066648006439209} +{"step": 8, "steps": 500, "lr": 2e-05, "grad_norm": 0.42078593373298645, "loss_scale": 4.0, "retries": 0, "seconds": 17.1, "max_memory_gb": 18.15, "ce": 3.7041693991050124, "rank": 0.2063145637512207} +{"step": 9, "steps": 500, "lr": 2e-05, "grad_norm": 0.32003554701805115, "loss_scale": 4.0, "retries": 0, "seconds": 19.1, "max_memory_gb": 18.15, "ce": 2.833598689176142, "rank": 0.04105401039123535} +{"step": 10, "steps": 500, "lr": 2e-05, "grad_norm": 0.16453413665294647, "loss_scale": 4.0, "retries": 0, "seconds": 21.0, "max_memory_gb": 18.15, "ce": 1.6168475029990077, "rank": 0.0} +{"step": 11, "steps": 500, "lr": 2e-05, "grad_norm": 0.4007941782474518, "loss_scale": 4.0, "retries": 0, "seconds": 23.0, "max_memory_gb": 18.15, "ce": 2.620367611758411, "rank": 0.21386121213436127} +{"step": 12, "steps": 500, "lr": 2e-05, "grad_norm": 0.20720481872558594, "loss_scale": 4.0, "retries": 0, "seconds": 25.0, "max_memory_gb": 18.15, "ce": 1.5088886159937829, "rank": 0.05079936981201172} +{"step": 13, "steps": 500, "lr": 2e-05, "grad_norm": 0.32248398661613464, "loss_scale": 4.0, "retries": 0, "seconds": 27.0, "max_memory_gb": 18.15, "ce": 2.667711005313322, "rank": 0.0} +{"step": 14, "steps": 500, "lr": 2e-05, "grad_norm": 0.5156862139701843, "loss_scale": 4.0, "retries": 0, "seconds": 28.9, "max_memory_gb": 18.15, "ce": 3.863085229881108, "rank": 0.15020287036895752} +{"step": 15, "steps": 500, "lr": 2e-05, "grad_norm": 0.3706808388233185, "loss_scale": 4.0, "retries": 0, "seconds": 30.9, "max_memory_gb": 18.15, "ce": 2.5925897766137496, "rank": 0.1719440221786499} +{"step": 16, "steps": 500, "lr": 2e-05, "grad_norm": 0.23645441234111786, "loss_scale": 4.0, "retries": 0, "seconds": 32.9, "max_memory_gb": 18.15, "ce": 1.4121987005346455, "rank": 0.0} +{"step": 17, "steps": 500, "lr": 2e-05, "grad_norm": 0.2507833242416382, "loss_scale": 4.0, "retries": 0, "seconds": 34.8, "max_memory_gb": 18.15, "ce": 1.5899563818238676, "rank": 0.03712666034698486} +{"step": 18, "steps": 500, "lr": 2e-05, "grad_norm": 0.49215012788772583, "loss_scale": 4.0, "retries": 0, "seconds": 36.8, "max_memory_gb": 18.16, "ce": 3.5763300955295563, "rank": 0.20132631063461304} +{"step": 19, "steps": 500, "lr": 2e-05, "grad_norm": 0.8655421137809753, "loss_scale": 4.0, "retries": 0, "seconds": 38.8, "max_memory_gb": 18.16, "ce": 5.250508800148964, "rank": 0.3163415193557739} +{"step": 20, "steps": 500, "lr": 2e-05, "grad_norm": 0.02635851316154003, "loss_scale": 4.0, "retries": 0, "seconds": 40.8, "max_memory_gb": 18.16, "ce": 0.0655368443694897, "rank": 0.10911641269922256} +{"step": 21, "steps": 500, "lr": 2e-05, "grad_norm": 0.5025784969329834, "loss_scale": 4.0, "retries": 0, "seconds": 42.8, "max_memory_gb": 18.16, "ce": 2.7218507040524855, "rank": 0.1621769666671753} +{"step": 22, "steps": 500, "lr": 2e-05, "grad_norm": 0.512218713760376, "loss_scale": 4.0, "retries": 0, "seconds": 44.9, "max_memory_gb": 18.16, "ce": 2.925056705251336, "rank": 0.18648791313171387} +{"step": 23, "steps": 500, "lr": 2e-05, "grad_norm": 0.5335860848426819, "loss_scale": 4.0, "retries": 0, "seconds": 47.0, "max_memory_gb": 18.16, "ce": 2.725272564101033, "rank": 0.1221691370010376} +{"step": 24, "steps": 500, "lr": 2e-05, "grad_norm": 0.9281453490257263, "loss_scale": 4.0, "retries": 0, "seconds": 49.2, "max_memory_gb": 18.16, "ce": 4.953263263218105, "rank": 0.2587096691131592} +{"step": 25, "steps": 500, "lr": 2e-05, "grad_norm": 0.9439056515693665, "loss_scale": 4.0, "retries": 0, "seconds": 51.2, "max_memory_gb": 18.16, "ce": 4.0730010420084, "rank": 0.04280400276184082} +{"step": 26, "steps": 500, "lr": 2e-05, "grad_norm": 1.2163044214248657, "loss_scale": 4.0, "retries": 0, "seconds": 53.3, "max_memory_gb": 18.16, "ce": 5.092586786486208, "rank": 0.07343077659606934} +{"step": 27, "steps": 500, "lr": 2e-05, "grad_norm": 0.9689059257507324, "loss_scale": 4.0, "retries": 0, "seconds": 55.4, "max_memory_gb": 18.16, "ce": 4.139717862010002, "rank": 0.4836317300796509} +{"step": 28, "steps": 500, "lr": 2e-05, "grad_norm": 0.5826318264007568, "loss_scale": 4.0, "retries": 0, "seconds": 57.6, "max_memory_gb": 18.16, "ce": 2.7301323506981134, "rank": 0.0} +{"step": 29, "steps": 500, "lr": 2e-05, "grad_norm": 0.5387762188911438, "loss_scale": 4.0, "retries": 0, "seconds": 59.6, "max_memory_gb": 18.16, "ce": 2.637924848590046, "rank": 0.0} +{"step": 30, "steps": 500, "lr": 2e-05, "grad_norm": 0.915666937828064, "loss_scale": 4.0, "retries": 0, "seconds": 61.6, "max_memory_gb": 18.16, "ce": 2.6818429846316576, "rank": 0.0} +{"step": 31, "steps": 500, "lr": 2e-05, "grad_norm": 0.9199861884117126, "loss_scale": 4.0, "retries": 0, "seconds": 63.6, "max_memory_gb": 18.16, "ce": 2.743006281554699, "rank": 0.1493159532546997} +{"step": 32, "steps": 500, "lr": 2e-05, "grad_norm": 1.6474896669387817, "loss_scale": 4.0, "retries": 0, "seconds": 65.6, "max_memory_gb": 18.16, "ce": 5.158201524056494, "rank": 0.23656630516052246} +{"step": 33, "steps": 500, "lr": 2e-05, "grad_norm": 0.790155291557312, "loss_scale": 4.0, "retries": 0, "seconds": 67.6, "max_memory_gb": 18.16, "ce": 2.460213240701705, "rank": 0.011545553803443909} +{"step": 34, "steps": 500, "lr": 2e-05, "grad_norm": 0.8684725761413574, "loss_scale": 4.0, "retries": 0, "seconds": 69.6, "max_memory_gb": 18.16, "ce": 2.6633344447473064, "rank": 0.0} +{"step": 35, "steps": 500, "lr": 2e-05, "grad_norm": 1.4342767000198364, "loss_scale": 4.0, "retries": 0, "seconds": 71.6, "max_memory_gb": 18.16, "ce": 2.597513075452298, "rank": 0.13342702388763428} +{"step": 36, "steps": 500, "lr": 2e-05, "grad_norm": 0.8576323390007019, "loss_scale": 4.0, "retries": 0, "seconds": 73.6, "max_memory_gb": 18.16, "ce": 2.5146903982385993, "rank": 0.0} +{"step": 37, "steps": 500, "lr": 2e-05, "grad_norm": 1.0420292615890503, "loss_scale": 4.0, "retries": 0, "seconds": 75.5, "max_memory_gb": 18.16, "ce": 2.529502478428185, "rank": 0.0} +{"step": 38, "steps": 500, "lr": 2e-05, "grad_norm": 1.9365692138671875, "loss_scale": 4.0, "retries": 0, "seconds": 77.5, "max_memory_gb": 18.16, "ce": 4.666010550223291, "rank": 0.1426539421081543} +{"step": 39, "steps": 500, "lr": 2e-05, "grad_norm": 1.0172405242919922, "loss_scale": 4.0, "retries": 0, "seconds": 79.5, "max_memory_gb": 18.16, "ce": 1.3709572076331824, "rank": 0.15499325096607208} +{"step": 40, "steps": 500, "lr": 2e-05, "grad_norm": 1.589140772819519, "loss_scale": 4.0, "retries": 0, "seconds": 81.5, "max_memory_gb": 18.16, "ce": 2.5522642792202532, "rank": 0.0012933015823364258} +{"step": 41, "steps": 500, "lr": 2e-05, "grad_norm": 0.6005830764770508, "loss_scale": 4.0, "retries": 0, "seconds": 83.4, "max_memory_gb": 18.16, "ce": 1.4711913531646132, "rank": 0.0} +{"step": 42, "steps": 500, "lr": 2e-05, "grad_norm": 1.366169810295105, "loss_scale": 4.0, "retries": 0, "seconds": 85.4, "max_memory_gb": 18.16, "ce": 2.2907950337976217, "rank": 0.0} +{"step": 43, "steps": 500, "lr": 2e-05, "grad_norm": 2.2000749111175537, "loss_scale": 4.0, "retries": 0, "seconds": 87.4, "max_memory_gb": 18.16, "ce": 2.3044486006256193, "rank": 0.0} +{"step": 44, "steps": 500, "lr": 2e-05, "grad_norm": 1.100877046585083, "loss_scale": 4.0, "retries": 0, "seconds": 89.4, "max_memory_gb": 18.16, "ce": 1.356782790273428, "rank": 0.0} +{"step": 45, "steps": 500, "lr": 2e-05, "grad_norm": 1.285048007965088, "loss_scale": 4.0, "retries": 0, "seconds": 91.3, "max_memory_gb": 18.16, "ce": 1.3301279991865158, "rank": 0.0} +{"step": 46, "steps": 500, "lr": 2e-05, "grad_norm": 2.061511993408203, "loss_scale": 4.0, "retries": 0, "seconds": 93.3, "max_memory_gb": 18.16, "ce": 2.2104112654924393, "rank": 0.0} +{"step": 47, "steps": 500, "lr": 2e-05, "grad_norm": 2.6820578575134277, "loss_scale": 4.0, "retries": 0, "seconds": 95.3, "max_memory_gb": 18.16, "ce": 1.876661135116592, "rank": 0.0} +{"step": 48, "steps": 500, "lr": 2e-05, "grad_norm": 4.4866132736206055, "loss_scale": 4.0, "retries": 0, "seconds": 97.3, "max_memory_gb": 18.16, "ce": 3.414862311910838, "rank": 0.05034768581390381} +{"step": 49, "steps": 500, "lr": 2e-05, "grad_norm": 5.957362174987793, "loss_scale": 4.0, "retries": 0, "seconds": 99.2, "max_memory_gb": 18.16, "ce": 3.120094741694629, "rank": 0.0} +{"step": 50, "steps": 500, "lr": 2e-05, "grad_norm": 4.329732894897461, "loss_scale": 4.0, "retries": 0, "seconds": 101.2, "max_memory_gb": 18.16, "ce": 3.3263377872062847, "rank": 0.0} +{"step": 51, "steps": 500, "lr": 2e-05, "grad_norm": 2.007018566131592, "loss_scale": 4.0, "retries": 0, "seconds": 103.2, "max_memory_gb": 18.16, "ce": 1.238118805224076, "rank": 0.0} +{"step": 52, "steps": 500, "lr": 2e-05, "grad_norm": 4.202111721038818, "loss_scale": 4.0, "retries": 0, "seconds": 105.0, "max_memory_gb": 18.16, "ce": 2.021534482948482, "rank": 0.0} +{"step": 53, "steps": 500, "lr": 2e-05, "grad_norm": 6.390848636627197, "loss_scale": 4.0, "retries": 0, "seconds": 106.8, "max_memory_gb": 18.16, "ce": 3.2098374105989933, "rank": 0.03165543079376221} +{"step": 54, "steps": 500, "lr": 2e-05, "grad_norm": 8.6488618850708, "loss_scale": 4.0, "retries": 0, "seconds": 108.7, "max_memory_gb": 18.16, "ce": 3.0829142490401864, "rank": 0.0} +{"step": 55, "steps": 500, "lr": 2e-05, "grad_norm": 6.132368564605713, "loss_scale": 4.0, "retries": 0, "seconds": 110.5, "max_memory_gb": 18.16, "ce": 1.800432427204214, "rank": 0.0} +{"step": 56, "steps": 500, "lr": 2e-05, "grad_norm": 12.248745918273926, "loss_scale": 4.0, "retries": 0, "seconds": 112.3, "max_memory_gb": 18.16, "ce": 3.5862517608329654, "rank": 0.0} +{"step": 57, "steps": 500, "lr": 2e-05, "grad_norm": 3.056069850921631, "loss_scale": 4.0, "retries": 0, "seconds": 114.1, "max_memory_gb": 18.16, "ce": 0.9793431744910777, "rank": 0.017887920141220093} +{"step": 58, "steps": 500, "lr": 2e-05, "grad_norm": 5.628541469573975, "loss_scale": 4.0, "retries": 0, "seconds": 116.0, "max_memory_gb": 18.16, "ce": 1.6515091890469193, "rank": 0.0} +{"step": 59, "steps": 500, "lr": 2e-05, "grad_norm": 2.7234556674957275, "loss_scale": 4.0, "retries": 0, "seconds": 117.8, "max_memory_gb": 18.16, "ce": 1.1800261037424207, "rank": 0.0} +{"step": 60, "steps": 500, "lr": 2e-05, "grad_norm": 5.051931381225586, "loss_scale": 4.0, "retries": 0, "seconds": 119.6, "max_memory_gb": 18.16, "ce": 1.8054974400438368, "rank": 0.0} +{"step": 61, "steps": 500, "lr": 2e-05, "grad_norm": 8.829627990722656, "loss_scale": 4.0, "retries": 0, "seconds": 121.5, "max_memory_gb": 18.16, "ce": 1.9241525949910283, "rank": 0.0} +{"step": 62, "steps": 500, "lr": 2e-05, "grad_norm": 3.7403528690338135, "loss_scale": 4.0, "retries": 0, "seconds": 123.3, "max_memory_gb": 18.16, "ce": 0.938575919251889, "rank": 0.0} +{"step": 63, "steps": 500, "lr": 2e-05, "grad_norm": 6.925289630889893, "loss_scale": 4.0, "retries": 0, "seconds": 125.2, "max_memory_gb": 18.16, "ce": 2.149437464075163, "rank": 0.0} +{"step": 64, "steps": 500, "lr": 2e-05, "grad_norm": 9.067178726196289, "loss_scale": 4.0, "retries": 0, "seconds": 127.0, "max_memory_gb": 18.16, "ce": 2.0607856849674135, "rank": 0.0} +{"step": 65, "steps": 500, "lr": 2e-05, "grad_norm": 8.466085433959961, "loss_scale": 4.0, "retries": 0, "seconds": 128.9, "max_memory_gb": 18.16, "ce": 2.0879512187093496, "rank": 0.0} +{"step": 66, "steps": 500, "lr": 2e-05, "grad_norm": 2.049421548843384, "loss_scale": 4.0, "retries": 0, "seconds": 130.7, "max_memory_gb": 18.16, "ce": 0.674270526622422, "rank": 0.28506288677453995} +{"step": 67, "steps": 500, "lr": 2e-05, "grad_norm": 5.748128890991211, "loss_scale": 4.0, "retries": 0, "seconds": 132.6, "max_memory_gb": 18.16, "ce": 1.3272194461897016, "rank": 0.0} +{"step": 68, "steps": 500, "lr": 2e-05, "grad_norm": 2.898158311843872, "loss_scale": 4.0, "retries": 0, "seconds": 134.4, "max_memory_gb": 18.16, "ce": 0.6724911404307932, "rank": 0.0} +{"step": 69, "steps": 500, "lr": 2e-05, "grad_norm": 4.5337958335876465, "loss_scale": 4.0, "retries": 0, "seconds": 136.2, "max_memory_gb": 18.16, "ce": 0.9932928392663598, "rank": 0.0} +{"step": 70, "steps": 500, "lr": 2e-05, "grad_norm": 3.70516300201416, "loss_scale": 4.0, "retries": 0, "seconds": 138.1, "max_memory_gb": 18.16, "ce": 0.8470701687037945, "rank": 0.0} +{"step": 71, "steps": 500, "lr": 2e-05, "grad_norm": 6.06073522567749, "loss_scale": 4.0, "retries": 0, "seconds": 139.9, "max_memory_gb": 18.16, "ce": 1.2459674291312695, "rank": 0.0} +{"step": 72, "steps": 500, "lr": 2e-05, "grad_norm": 7.696965217590332, "loss_scale": 4.0, "retries": 0, "seconds": 141.8, "max_memory_gb": 18.16, "ce": 1.5182847445830703, "rank": 0.0} +{"step": 73, "steps": 500, "lr": 2e-05, "grad_norm": 0.8033193349838257, "loss_scale": 4.0, "retries": 0, "seconds": 143.5, "max_memory_gb": 18.16, "ce": 0.3051873983349651, "rank": 0.0} +{"step": 74, "steps": 500, "lr": 2e-05, "grad_norm": 2.188652992248535, "loss_scale": 4.0, "retries": 0, "seconds": 145.3, "max_memory_gb": 18.16, "ce": 0.7151041775941849, "rank": 0.0} +{"step": 75, "steps": 500, "lr": 2e-05, "grad_norm": 2.441580057144165, "loss_scale": 4.0, "retries": 0, "seconds": 147.2, "max_memory_gb": 18.16, "ce": 0.9527793088927865, "rank": 0.21720772981643677} +{"step": 76, "steps": 500, "lr": 2e-05, "grad_norm": 2.0374224185943604, "loss_scale": 4.0, "retries": 0, "seconds": 149.0, "max_memory_gb": 18.16, "ce": 0.7527499049901962, "rank": 0.0} +{"step": 77, "steps": 500, "lr": 2e-05, "grad_norm": 4.128293991088867, "loss_scale": 4.0, "retries": 0, "seconds": 150.8, "max_memory_gb": 18.16, "ce": 1.1658408166840672, "rank": 0.07024618983268738} +{"step": 78, "steps": 500, "lr": 2e-05, "grad_norm": 2.2211928367614746, "loss_scale": 4.0, "retries": 0, "seconds": 152.6, "max_memory_gb": 18.16, "ce": 0.6445214189589024, "rank": 0.0} +{"step": 79, "steps": 500, "lr": 2e-05, "grad_norm": 1.3883723020553589, "loss_scale": 4.0, "retries": 0, "seconds": 154.5, "max_memory_gb": 18.16, "ce": 0.4578545093536377, "rank": 0.0} +{"step": 80, "steps": 500, "lr": 2e-05, "grad_norm": 1.7948821783065796, "loss_scale": 4.0, "retries": 0, "seconds": 156.3, "max_memory_gb": 18.16, "ce": 0.5555140608921647, "rank": 0.0} +{"step": 81, "steps": 500, "lr": 2e-05, "grad_norm": 2.4921982288360596, "loss_scale": 4.0, "retries": 0, "seconds": 158.1, "max_memory_gb": 18.16, "ce": 0.9330494627356529, "rank": 0.0} +{"step": 82, "steps": 500, "lr": 2e-05, "grad_norm": 1.8125303983688354, "loss_scale": 4.0, "retries": 0, "seconds": 159.9, "max_memory_gb": 18.16, "ce": 0.68371507152915, "rank": 0.0} +{"step": 83, "steps": 500, "lr": 2e-05, "grad_norm": 1.7988767623901367, "loss_scale": 4.0, "retries": 0, "seconds": 161.8, "max_memory_gb": 18.16, "ce": 0.8154228408820927, "rank": 0.0} +{"step": 84, "steps": 500, "lr": 2e-05, "grad_norm": 1.6349077224731445, "loss_scale": 4.0, "retries": 0, "seconds": 163.6, "max_memory_gb": 18.16, "ce": 0.4237370193004608, "rank": 0.0} +{"step": 85, "steps": 500, "lr": 2e-05, "grad_norm": 2.018035411834717, "loss_scale": 4.0, "retries": 0, "seconds": 165.5, "max_memory_gb": 18.16, "ce": 0.2451303913258016, "rank": 0.0} +{"step": 86, "steps": 500, "lr": 2e-05, "grad_norm": 2.946509599685669, "loss_scale": 4.0, "retries": 0, "seconds": 167.3, "max_memory_gb": 18.16, "ce": 0.8720276998355985, "rank": 0.0} +{"step": 87, "steps": 500, "lr": 2e-05, "grad_norm": 1.471139907836914, "loss_scale": 4.0, "retries": 0, "seconds": 169.1, "max_memory_gb": 18.16, "ce": 0.7875635409727693, "rank": 0.0} +{"step": 88, "steps": 500, "lr": 2e-05, "grad_norm": 1.953900694847107, "loss_scale": 4.0, "retries": 0, "seconds": 171.0, "max_memory_gb": 18.16, "ce": 0.766096746083349, "rank": 0.0} +{"step": 89, "steps": 500, "lr": 2e-05, "grad_norm": 1.4697993993759155, "loss_scale": 4.0, "retries": 0, "seconds": 172.8, "max_memory_gb": 18.16, "ce": 0.21647998760454357, "rank": 0.1198773980140686} +{"step": 90, "steps": 500, "lr": 2e-05, "grad_norm": 1.6810567378997803, "loss_scale": 4.0, "retries": 0, "seconds": 174.6, "max_memory_gb": 18.16, "ce": 0.6524388333782554, "rank": 0.0} +{"step": 91, "steps": 500, "lr": 2e-05, "grad_norm": 1.1634409427642822, "loss_scale": 4.0, "retries": 0, "seconds": 176.4, "max_memory_gb": 18.16, "ce": 0.4517845516093075, "rank": 0.0} +{"step": 92, "steps": 500, "lr": 2e-05, "grad_norm": 1.0302501916885376, "loss_scale": 4.0, "retries": 0, "seconds": 178.3, "max_memory_gb": 18.16, "ce": 0.3757760445587337, "rank": 0.0} +{"step": 93, "steps": 500, "lr": 2e-05, "grad_norm": 1.3448406457901, "loss_scale": 4.0, "retries": 0, "seconds": 180.1, "max_memory_gb": 18.16, "ce": 0.5727731022052467, "rank": 0.0} +{"step": 94, "steps": 500, "lr": 2e-05, "grad_norm": 1.671282410621643, "loss_scale": 4.0, "retries": 0, "seconds": 182.0, "max_memory_gb": 18.16, "ce": 0.6982000740244985, "rank": 0.0} +{"step": 95, "steps": 500, "lr": 2e-05, "grad_norm": 1.3807413578033447, "loss_scale": 4.0, "retries": 0, "seconds": 183.8, "max_memory_gb": 18.16, "ce": 0.712691580876708, "rank": 0.0} +{"step": 96, "steps": 500, "lr": 2e-05, "grad_norm": 1.5692007541656494, "loss_scale": 4.0, "retries": 0, "seconds": 185.6, "max_memory_gb": 18.16, "ce": 0.7978450325317681, "rank": 0.06287875771522522} +{"step": 97, "steps": 500, "lr": 2e-05, "grad_norm": 0.9860432744026184, "loss_scale": 4.0, "retries": 0, "seconds": 187.4, "max_memory_gb": 18.16, "ce": 0.44117384031414986, "rank": 0.009839251637458801} +{"step": 98, "steps": 500, "lr": 2e-05, "grad_norm": 1.0055245161056519, "loss_scale": 4.0, "retries": 0, "seconds": 189.3, "max_memory_gb": 18.16, "ce": 0.44612169987522066, "rank": 0.0} +{"step": 99, "steps": 500, "lr": 2e-05, "grad_norm": 1.0698121786117554, "loss_scale": 4.0, "retries": 0, "seconds": 191.1, "max_memory_gb": 18.16, "ce": 0.28376206243410707, "rank": 0.0649753212928772} +{"step": 100, "steps": 500, "lr": 2e-05, "grad_norm": 1.8001651763916016, "loss_scale": 4.0, "retries": 0, "seconds": 193.0, "max_memory_gb": 18.16, "ce": 0.7073246031068265, "rank": 0.0} +{"step": 101, "steps": 500, "lr": 2e-05, "grad_norm": 1.4813556671142578, "loss_scale": 4.0, "retries": 0, "seconds": 194.8, "max_memory_gb": 18.16, "ce": 0.7022381294518709, "rank": 0.0} +{"step": 102, "steps": 500, "lr": 2e-05, "grad_norm": 2.3253324031829834, "loss_scale": 4.0, "retries": 0, "seconds": 196.7, "max_memory_gb": 18.16, "ce": 0.7686393368057907, "rank": 0.0} +{"step": 103, "steps": 500, "lr": 2e-05, "grad_norm": 0.8800620436668396, "loss_scale": 4.0, "retries": 0, "seconds": 198.5, "max_memory_gb": 18.16, "ce": 0.3750937203876674, "rank": 0.1392141580581665} +{"step": 104, "steps": 500, "lr": 2e-05, "grad_norm": 0.6842507123947144, "loss_scale": 4.0, "retries": 0, "seconds": 200.3, "max_memory_gb": 18.16, "ce": 0.30021531437523663, "rank": 0.07472890615463257} +{"step": 105, "steps": 500, "lr": 2e-05, "grad_norm": 1.2021734714508057, "loss_scale": 4.0, "retries": 0, "seconds": 202.1, "max_memory_gb": 18.16, "ce": 0.6173380119726062, "rank": 0.007114902138710022} +{"step": 106, "steps": 500, "lr": 2e-05, "grad_norm": 2.94785213470459, "loss_scale": 4.0, "retries": 0, "seconds": 204.0, "max_memory_gb": 18.16, "ce": 0.6185914492234588, "rank": 0.05101858079433441} +{"step": 107, "steps": 500, "lr": 2e-05, "grad_norm": 0.4934125244617462, "loss_scale": 4.0, "retries": 0, "seconds": 205.8, "max_memory_gb": 18.16, "ce": 0.3270048485137522, "rank": 0.0} +{"step": 108, "steps": 500, "lr": 2e-05, "grad_norm": 0.6252261400222778, "loss_scale": 4.0, "retries": 0, "seconds": 207.6, "max_memory_gb": 18.16, "ce": 0.3466661835554987, "rank": 0.0} +{"step": 109, "steps": 500, "lr": 2e-05, "grad_norm": 2.0653913021087646, "loss_scale": 4.0, "retries": 0, "seconds": 209.5, "max_memory_gb": 18.16, "ce": 0.7249364689923823, "rank": 0.0} +{"step": 110, "steps": 500, "lr": 2e-05, "grad_norm": 1.3912837505340576, "loss_scale": 4.0, "retries": 0, "seconds": 211.3, "max_memory_gb": 18.16, "ce": 0.5286518624052405, "rank": 0.007406964898109436} +{"step": 111, "steps": 500, "lr": 2e-05, "grad_norm": 1.5145913362503052, "loss_scale": 4.0, "retries": 0, "seconds": 213.2, "max_memory_gb": 18.16, "ce": 0.46859642351046205, "rank": 0.0} +{"step": 112, "steps": 500, "lr": 2e-05, "grad_norm": 0.6761723160743713, "loss_scale": 4.0, "retries": 0, "seconds": 215.0, "max_memory_gb": 18.16, "ce": 0.3749286630190909, "rank": 0.0} +{"step": 113, "steps": 500, "lr": 2e-05, "grad_norm": 1.3734079599380493, "loss_scale": 4.0, "retries": 0, "seconds": 216.8, "max_memory_gb": 18.16, "ce": 0.4392743152566254, "rank": 0.0} +{"step": 114, "steps": 500, "lr": 2e-05, "grad_norm": 1.9997810125350952, "loss_scale": 4.0, "retries": 0, "seconds": 218.7, "max_memory_gb": 18.16, "ce": 0.8412886247970164, "rank": 0.16932563483715057} +{"step": 115, "steps": 500, "lr": 2e-05, "grad_norm": 0.8191068768501282, "loss_scale": 4.0, "retries": 0, "seconds": 220.5, "max_memory_gb": 18.16, "ce": 0.4971650755032897, "rank": 0.00931800901889801} +{"step": 116, "steps": 500, "lr": 2e-05, "grad_norm": 0.4119715988636017, "loss_scale": 4.0, "retries": 0, "seconds": 222.3, "max_memory_gb": 18.16, "ce": 0.16144527937285602, "rank": 0.0} +{"step": 117, "steps": 500, "lr": 2e-05, "grad_norm": 4.313961982727051, "loss_scale": 4.0, "retries": 0, "seconds": 224.2, "max_memory_gb": 18.16, "ce": 0.6655187023570761, "rank": 0.0} +{"step": 118, "steps": 500, "lr": 2e-05, "grad_norm": 0.3936131000518799, "loss_scale": 4.0, "retries": 0, "seconds": 226.0, "max_memory_gb": 18.16, "ce": 0.31328837503679097, "rank": 0.0} +{"step": 119, "steps": 500, "lr": 2e-05, "grad_norm": 0.505397379398346, "loss_scale": 4.0, "retries": 0, "seconds": 227.8, "max_memory_gb": 18.16, "ce": 0.3142714104615152, "rank": 0.0} +{"step": 120, "steps": 500, "lr": 2e-05, "grad_norm": 0.9362943768501282, "loss_scale": 4.0, "retries": 0, "seconds": 229.7, "max_memory_gb": 18.16, "ce": 0.5403827070258558, "rank": 0.0} +{"step": 121, "steps": 500, "lr": 2e-05, "grad_norm": 0.7435526847839355, "loss_scale": 4.0, "retries": 0, "seconds": 231.5, "max_memory_gb": 18.16, "ce": 0.43559910054318607, "rank": 0.0} +{"step": 122, "steps": 500, "lr": 2e-05, "grad_norm": 0.8177743554115295, "loss_scale": 4.0, "retries": 0, "seconds": 233.4, "max_memory_gb": 18.16, "ce": 0.4236517755780369, "rank": 0.0} +{"step": 123, "steps": 500, "lr": 2e-05, "grad_norm": 1.0591157674789429, "loss_scale": 4.0, "retries": 0, "seconds": 235.2, "max_memory_gb": 18.16, "ce": 0.4908280516974628, "rank": 0.007964923977851868} +{"step": 124, "steps": 500, "lr": 2e-05, "grad_norm": 1.1868057250976562, "loss_scale": 4.0, "retries": 0, "seconds": 237.0, "max_memory_gb": 18.16, "ce": 0.42646180279552937, "rank": 0.024517208337783813} +{"step": 125, "steps": 500, "lr": 2e-05, "grad_norm": 0.36202871799468994, "loss_scale": 4.0, "retries": 0, "seconds": 238.8, "max_memory_gb": 18.16, "ce": 0.32424792228266597, "rank": 0.0} +{"step": 126, "steps": 500, "lr": 2e-05, "grad_norm": 1.144975185394287, "loss_scale": 4.0, "retries": 0, "seconds": 240.6, "max_memory_gb": 18.16, "ce": 0.6210831375792623, "rank": 0.0} +{"step": 127, "steps": 500, "lr": 2e-05, "grad_norm": 0.36804094910621643, "loss_scale": 4.0, "retries": 0, "seconds": 242.4, "max_memory_gb": 18.16, "ce": 0.20658079267013818, "rank": 0.0} +{"step": 128, "steps": 500, "lr": 2e-05, "grad_norm": 1.0458228588104248, "loss_scale": 4.0, "retries": 0, "seconds": 244.3, "max_memory_gb": 18.16, "ce": 0.4736693133600056, "rank": 0.0} +{"step": 129, "steps": 500, "lr": 2e-05, "grad_norm": 1.7914948463439941, "loss_scale": 4.0, "retries": 0, "seconds": 246.1, "max_memory_gb": 18.16, "ce": 0.571044459939003, "rank": 0.0} +{"step": 130, "steps": 500, "lr": 2e-05, "grad_norm": 0.9364795088768005, "loss_scale": 4.0, "retries": 0, "seconds": 247.9, "max_memory_gb": 18.16, "ce": 0.6161915115080774, "rank": 0.0} +{"step": 131, "steps": 500, "lr": 2e-05, "grad_norm": 1.1521047353744507, "loss_scale": 4.0, "retries": 0, "seconds": 249.8, "max_memory_gb": 18.16, "ce": 0.4214415717869997, "rank": 0.0} +{"step": 132, "steps": 500, "lr": 2e-05, "grad_norm": 0.9765698909759521, "loss_scale": 4.0, "retries": 0, "seconds": 251.6, "max_memory_gb": 18.16, "ce": 0.5772117340238765, "rank": 0.0} +{"step": 133, "steps": 500, "lr": 2e-05, "grad_norm": 1.686145544052124, "loss_scale": 4.0, "retries": 0, "seconds": 253.4, "max_memory_gb": 18.16, "ce": 0.5572482319548726, "rank": 0.08719335496425629} +{"step": 134, "steps": 500, "lr": 2e-05, "grad_norm": 0.4892100691795349, "loss_scale": 4.0, "retries": 0, "seconds": 255.3, "max_memory_gb": 18.16, "ce": 0.22229433245956898, "rank": 0.0} +{"step": 135, "steps": 500, "lr": 2e-05, "grad_norm": 1.8737554550170898, "loss_scale": 4.0, "retries": 0, "seconds": 257.1, "max_memory_gb": 18.16, "ce": 0.520506726577878, "rank": 0.005907967686653137} +{"step": 136, "steps": 500, "lr": 2e-05, "grad_norm": 1.7174776792526245, "loss_scale": 4.0, "retries": 0, "seconds": 258.9, "max_memory_gb": 18.16, "ce": 0.4404677404090762, "rank": 0.0} +{"step": 137, "steps": 500, "lr": 2e-05, "grad_norm": 0.5528613924980164, "loss_scale": 4.0, "retries": 0, "seconds": 260.7, "max_memory_gb": 18.16, "ce": 0.36692886939272285, "rank": 0.0} +{"step": 138, "steps": 500, "lr": 2e-05, "grad_norm": 0.49483829736709595, "loss_scale": 4.0, "retries": 0, "seconds": 262.6, "max_memory_gb": 18.16, "ce": 0.3573495550081134, "rank": 0.0} +{"step": 139, "steps": 500, "lr": 2e-05, "grad_norm": 5.442861080169678, "loss_scale": 4.0, "retries": 0, "seconds": 264.4, "max_memory_gb": 18.16, "ce": 0.6398202273994684, "rank": 0.0} +{"step": 140, "steps": 500, "lr": 2e-05, "grad_norm": 0.49270129203796387, "loss_scale": 4.0, "retries": 0, "seconds": 266.3, "max_memory_gb": 18.16, "ce": 0.3284030288923532, "rank": 0.0} +{"step": 141, "steps": 500, "lr": 2e-05, "grad_norm": 0.5482137203216553, "loss_scale": 4.0, "retries": 0, "seconds": 268.1, "max_memory_gb": 18.16, "ce": 0.38723316695541143, "rank": 0.0} +{"step": 142, "steps": 500, "lr": 2e-05, "grad_norm": 0.41316694021224976, "loss_scale": 4.0, "retries": 0, "seconds": 270.0, "max_memory_gb": 18.16, "ce": 0.3682372528128326, "rank": 0.0} +{"step": 143, "steps": 500, "lr": 2e-05, "grad_norm": 0.27403920888900757, "loss_scale": 4.0, "retries": 0, "seconds": 271.8, "max_memory_gb": 18.16, "ce": 0.18407835683319718, "rank": 0.0} +{"step": 144, "steps": 500, "lr": 2e-05, "grad_norm": 1.0187562704086304, "loss_scale": 4.0, "retries": 0, "seconds": 273.6, "max_memory_gb": 18.16, "ce": 0.45139201218262315, "rank": 0.0} +{"step": 145, "steps": 500, "lr": 2e-05, "grad_norm": 0.668590247631073, "loss_scale": 4.0, "retries": 0, "seconds": 275.5, "max_memory_gb": 18.16, "ce": 0.24910954001825303, "rank": 0.0} +{"step": 146, "steps": 500, "lr": 2e-05, "grad_norm": 0.6843160390853882, "loss_scale": 4.0, "retries": 0, "seconds": 277.5, "max_memory_gb": 18.16, "ce": 0.35098226787522435, "rank": 0.0} +{"step": 147, "steps": 500, "lr": 2e-05, "grad_norm": 0.9648283123970032, "loss_scale": 4.0, "retries": 0, "seconds": 279.3, "max_memory_gb": 18.16, "ce": 0.5232707823161036, "rank": 0.0} +{"step": 148, "steps": 500, "lr": 2e-05, "grad_norm": 0.6752365231513977, "loss_scale": 4.0, "retries": 0, "seconds": 281.2, "max_memory_gb": 18.16, "ce": 0.1725556117016822, "rank": 0.0} +{"step": 149, "steps": 500, "lr": 2e-05, "grad_norm": 1.1715155839920044, "loss_scale": 4.0, "retries": 0, "seconds": 283.0, "max_memory_gb": 18.16, "ce": 0.2795561084058136, "rank": 0.13423098623752594} +{"step": 150, "steps": 500, "lr": 2e-05, "grad_norm": 0.937803328037262, "loss_scale": 4.0, "retries": 0, "seconds": 284.8, "max_memory_gb": 18.16, "ce": 0.5299814529716969, "rank": 0.05887627601623535} +{"step": 151, "steps": 500, "lr": 2e-05, "grad_norm": 5.189846992492676, "loss_scale": 4.0, "retries": 0, "seconds": 286.7, "max_memory_gb": 18.16, "ce": 0.3262155541451648, "rank": 0.0} +{"step": 152, "steps": 500, "lr": 2e-05, "grad_norm": 0.607385516166687, "loss_scale": 4.0, "retries": 0, "seconds": 288.6, "max_memory_gb": 18.16, "ce": 0.48332189209759235, "rank": 0.0} +{"step": 153, "steps": 500, "lr": 2e-05, "grad_norm": 2.437366008758545, "loss_scale": 4.0, "retries": 0, "seconds": 290.4, "max_memory_gb": 18.16, "ce": 0.3952176596503705, "rank": 0.0} +{"step": 154, "steps": 500, "lr": 2e-05, "grad_norm": 1.0426819324493408, "loss_scale": 4.0, "retries": 0, "seconds": 292.2, "max_memory_gb": 18.16, "ce": 0.693914083763957, "rank": 0.14587882161140442} +{"step": 155, "steps": 500, "lr": 2e-05, "grad_norm": 2.120931386947632, "loss_scale": 4.0, "retries": 0, "seconds": 294.1, "max_memory_gb": 18.16, "ce": 0.6705321365734562, "rank": 0.0476360023021698} +{"step": 156, "steps": 500, "lr": 2e-05, "grad_norm": 1.0381050109863281, "loss_scale": 4.0, "retries": 0, "seconds": 295.9, "max_memory_gb": 18.16, "ce": 0.3093556867679581, "rank": 0.0} +{"step": 157, "steps": 500, "lr": 2e-05, "grad_norm": 0.5594546794891357, "loss_scale": 4.0, "retries": 0, "seconds": 297.8, "max_memory_gb": 18.16, "ce": 0.3032393050380051, "rank": 0.0} +{"step": 158, "steps": 500, "lr": 2e-05, "grad_norm": 0.6328936815261841, "loss_scale": 4.0, "retries": 0, "seconds": 299.6, "max_memory_gb": 18.16, "ce": 0.4902345947921276, "rank": 0.0} +{"step": 159, "steps": 500, "lr": 2e-05, "grad_norm": 1.2986464500427246, "loss_scale": 4.0, "retries": 0, "seconds": 301.5, "max_memory_gb": 18.17, "ce": 0.3216555758845061, "rank": 0.0} +{"step": 160, "steps": 500, "lr": 2e-05, "grad_norm": 0.631023645401001, "loss_scale": 4.0, "retries": 0, "seconds": 303.3, "max_memory_gb": 18.17, "ce": 0.39982602559030056, "rank": 0.055580660700798035} +{"step": 161, "steps": 500, "lr": 2e-05, "grad_norm": 0.44302457571029663, "loss_scale": 4.0, "retries": 0, "seconds": 305.2, "max_memory_gb": 18.17, "ce": 0.33147446520160884, "rank": 0.0} +{"step": 162, "steps": 500, "lr": 2e-05, "grad_norm": 0.3379290699958801, "loss_scale": 4.0, "retries": 0, "seconds": 307.0, "max_memory_gb": 18.17, "ce": 0.2344513339921832, "rank": 0.0} +{"step": 163, "steps": 500, "lr": 2e-05, "grad_norm": 1.1493223905563354, "loss_scale": 4.0, "retries": 0, "seconds": 308.8, "max_memory_gb": 18.17, "ce": 0.4036097240168601, "rank": 0.0} +{"step": 164, "steps": 500, "lr": 2e-05, "grad_norm": 1.603000521659851, "loss_scale": 4.0, "retries": 0, "seconds": 310.7, "max_memory_gb": 18.17, "ce": 0.4452881026081741, "rank": 0.06870701909065247} +{"step": 165, "steps": 500, "lr": 2e-05, "grad_norm": 0.7100290060043335, "loss_scale": 4.0, "retries": 0, "seconds": 312.6, "max_memory_gb": 18.17, "ce": 0.2899041506461799, "rank": 0.0} +{"step": 166, "steps": 500, "lr": 2e-05, "grad_norm": 1.121715784072876, "loss_scale": 4.0, "retries": 0, "seconds": 314.4, "max_memory_gb": 18.17, "ce": 0.40177660854533315, "rank": 0.0} +{"step": 167, "steps": 500, "lr": 2e-05, "grad_norm": 1.579033374786377, "loss_scale": 4.0, "retries": 0, "seconds": 316.3, "max_memory_gb": 18.17, "ce": 0.47442664159461856, "rank": 0.0} +{"step": 168, "steps": 500, "lr": 2e-05, "grad_norm": 1.75412118434906, "loss_scale": 4.0, "retries": 0, "seconds": 318.2, "max_memory_gb": 18.17, "ce": 0.3245944632217288, "rank": 0.0} +{"step": 169, "steps": 500, "lr": 2e-05, "grad_norm": 3.5280728340148926, "loss_scale": 4.0, "retries": 0, "seconds": 320.0, "max_memory_gb": 18.17, "ce": 0.8402972416952252, "rank": 0.0} +{"step": 170, "steps": 500, "lr": 2e-05, "grad_norm": 0.9511498212814331, "loss_scale": 4.0, "retries": 0, "seconds": 321.9, "max_memory_gb": 18.17, "ce": 0.34136381722055376, "rank": 0.0} +{"step": 171, "steps": 500, "lr": 2e-05, "grad_norm": 0.7533251047134399, "loss_scale": 4.0, "retries": 0, "seconds": 323.7, "max_memory_gb": 18.17, "ce": 0.3065480161458254, "rank": 0.006984129548072815} +{"step": 172, "steps": 500, "lr": 2e-05, "grad_norm": 0.866163432598114, "loss_scale": 4.0, "retries": 0, "seconds": 325.6, "max_memory_gb": 18.17, "ce": 0.42593014193698764, "rank": 0.0} +{"step": 173, "steps": 500, "lr": 2e-05, "grad_norm": 1.171199917793274, "loss_scale": 4.0, "retries": 0, "seconds": 327.4, "max_memory_gb": 18.17, "ce": 0.48596165236085653, "rank": 0.021351933479309082} +{"step": 174, "steps": 500, "lr": 2e-05, "grad_norm": 0.9480156898498535, "loss_scale": 4.0, "retries": 0, "seconds": 329.3, "max_memory_gb": 18.17, "ce": 0.34901907690800726, "rank": 0.0} +{"step": 175, "steps": 500, "lr": 2e-05, "grad_norm": 1.1038864850997925, "loss_scale": 4.0, "retries": 0, "seconds": 331.1, "max_memory_gb": 18.17, "ce": 0.4440535493195057, "rank": 0.0} +{"step": 176, "steps": 500, "lr": 2e-05, "grad_norm": 1.1950567960739136, "loss_scale": 4.0, "retries": 0, "seconds": 333.0, "max_memory_gb": 18.17, "ce": 0.3857529740780592, "rank": 0.010530486702919006} +{"step": 177, "steps": 500, "lr": 2e-05, "grad_norm": 1.8490725755691528, "loss_scale": 4.0, "retries": 0, "seconds": 334.8, "max_memory_gb": 18.17, "ce": 0.36011964874342084, "rank": 0.12782232463359833} +{"step": 178, "steps": 500, "lr": 2e-05, "grad_norm": 0.6883350610733032, "loss_scale": 4.0, "retries": 0, "seconds": 336.6, "max_memory_gb": 18.17, "ce": 0.31934839743189514, "rank": 0.0} +{"step": 179, "steps": 500, "lr": 2e-05, "grad_norm": 1.8225866556167603, "loss_scale": 4.0, "retries": 0, "seconds": 338.4, "max_memory_gb": 18.17, "ce": 0.4779519133735448, "rank": 0.0} +{"step": 180, "steps": 500, "lr": 2e-05, "grad_norm": 3.8070995807647705, "loss_scale": 4.0, "retries": 0, "seconds": 340.3, "max_memory_gb": 18.17, "ce": 0.616258667781949, "rank": 0.0} +{"step": 181, "steps": 500, "lr": 2e-05, "grad_norm": 5.267152786254883, "loss_scale": 4.0, "retries": 0, "seconds": 342.2, "max_memory_gb": 18.17, "ce": 0.44422344770282507, "rank": 0.04852387309074402} +{"step": 182, "steps": 500, "lr": 2e-05, "grad_norm": 0.5882893800735474, "loss_scale": 4.0, "retries": 0, "seconds": 344.0, "max_memory_gb": 18.17, "ce": 0.24934808595571667, "rank": 0.0} +{"step": 183, "steps": 500, "lr": 2e-05, "grad_norm": 0.5040888786315918, "loss_scale": 4.0, "retries": 0, "seconds": 345.9, "max_memory_gb": 18.17, "ce": 0.19047003920422867, "rank": 0.0} +{"step": 184, "steps": 500, "lr": 2e-05, "grad_norm": 0.6595212817192078, "loss_scale": 4.0, "retries": 0, "seconds": 347.7, "max_memory_gb": 18.17, "ce": 0.2522300769924186, "rank": 0.11321593075990677} +{"step": 185, "steps": 500, "lr": 2e-05, "grad_norm": 0.9096860289573669, "loss_scale": 4.0, "retries": 0, "seconds": 349.6, "max_memory_gb": 18.17, "ce": 0.31948510836809874, "rank": 0.0} +{"step": 186, "steps": 500, "lr": 2e-05, "grad_norm": 0.8229278922080994, "loss_scale": 4.0, "retries": 0, "seconds": 351.4, "max_memory_gb": 18.17, "ce": 0.23776489595184103, "rank": 0.0} +{"step": 187, "steps": 500, "lr": 2e-05, "grad_norm": 2.9132537841796875, "loss_scale": 4.0, "retries": 0, "seconds": 353.3, "max_memory_gb": 18.17, "ce": 0.5102179944515228, "rank": 0.10975103825330734} +{"step": 188, "steps": 500, "lr": 2e-05, "grad_norm": 0.20856575667858124, "loss_scale": 4.0, "retries": 0, "seconds": 355.1, "max_memory_gb": 18.17, "ce": 0.17002726517966948, "rank": 0.0} +{"step": 189, "steps": 500, "lr": 2e-05, "grad_norm": 3.0389227867126465, "loss_scale": 4.0, "retries": 0, "seconds": 356.9, "max_memory_gb": 18.17, "ce": 0.3207782569807023, "rank": 0.0} +{"step": 190, "steps": 500, "lr": 2e-05, "grad_norm": 2.0603528022766113, "loss_scale": 4.0, "retries": 0, "seconds": 358.8, "max_memory_gb": 18.17, "ce": 0.4413899593055248, "rank": 0.0} +{"step": 191, "steps": 500, "lr": 2e-05, "grad_norm": 1.1466776132583618, "loss_scale": 4.0, "retries": 0, "seconds": 360.6, "max_memory_gb": 18.17, "ce": 0.36448355484753847, "rank": 0.0} +{"step": 192, "steps": 500, "lr": 2e-05, "grad_norm": 0.8674055933952332, "loss_scale": 4.0, "retries": 0, "seconds": 362.5, "max_memory_gb": 18.17, "ce": 0.3637557033216581, "rank": 0.0} +{"step": 193, "steps": 500, "lr": 2e-05, "grad_norm": 1.213284969329834, "loss_scale": 4.0, "retries": 0, "seconds": 364.3, "max_memory_gb": 18.17, "ce": 0.2827411247417331, "rank": 0.12564261257648468} +{"step": 194, "steps": 500, "lr": 2e-05, "grad_norm": 1.0590864419937134, "loss_scale": 4.0, "retries": 0, "seconds": 366.2, "max_memory_gb": 18.17, "ce": 0.3699285304173827, "rank": 0.0} +{"step": 195, "steps": 500, "lr": 2e-05, "grad_norm": 1.2699840068817139, "loss_scale": 4.0, "retries": 0, "seconds": 368.0, "max_memory_gb": 18.17, "ce": 0.3721770327538252, "rank": 0.0} +{"step": 196, "steps": 500, "lr": 2e-05, "grad_norm": 0.9999151229858398, "loss_scale": 4.0, "retries": 0, "seconds": 369.9, "max_memory_gb": 18.17, "ce": 0.27504780306480825, "rank": 0.0} +{"step": 197, "steps": 500, "lr": 2e-05, "grad_norm": 0.9024571180343628, "loss_scale": 4.0, "retries": 0, "seconds": 371.7, "max_memory_gb": 18.17, "ce": 0.2760153429117054, "rank": 0.08556802570819855} +{"step": 198, "steps": 500, "lr": 2e-05, "grad_norm": 0.8210251927375793, "loss_scale": 4.0, "retries": 0, "seconds": 373.6, "max_memory_gb": 18.17, "ce": 0.3514081174507737, "rank": 0.0} +{"step": 199, "steps": 500, "lr": 2e-05, "grad_norm": 4.625514030456543, "loss_scale": 4.0, "retries": 0, "seconds": 375.4, "max_memory_gb": 18.17, "ce": 0.4259987899567932, "rank": 0.0} +{"step": 200, "steps": 500, "lr": 2e-05, "grad_norm": 0.8023871779441833, "loss_scale": 4.0, "retries": 0, "seconds": 377.3, "max_memory_gb": 18.17, "ce": 0.3405531463213265, "rank": 0.015649303793907166} +{"step": 201, "steps": 500, "lr": 2e-05, "grad_norm": 0.7467019557952881, "loss_scale": 4.0, "retries": 0, "seconds": 379.1, "max_memory_gb": 18.17, "ce": 0.2789114071056247, "rank": 0.0} +{"step": 202, "steps": 500, "lr": 2e-05, "grad_norm": 1.3867003917694092, "loss_scale": 4.0, "retries": 0, "seconds": 380.9, "max_memory_gb": 18.17, "ce": 0.12546445277985185, "rank": 0.045582398772239685} +{"step": 203, "steps": 500, "lr": 2e-05, "grad_norm": 0.9519790410995483, "loss_scale": 4.0, "retries": 0, "seconds": 382.8, "max_memory_gb": 18.17, "ce": 0.3179127920884639, "rank": 0.0} +{"step": 204, "steps": 500, "lr": 2e-05, "grad_norm": 8.088081359863281, "loss_scale": 4.0, "retries": 0, "seconds": 384.6, "max_memory_gb": 18.17, "ce": 0.4921891903504729, "rank": 0.0} +{"step": 205, "steps": 500, "lr": 2e-05, "grad_norm": 2.495427370071411, "loss_scale": 4.0, "retries": 0, "seconds": 386.5, "max_memory_gb": 18.17, "ce": 0.36808970346464776, "rank": 0.0} +{"step": 206, "steps": 500, "lr": 2e-05, "grad_norm": 4.526402473449707, "loss_scale": 4.0, "retries": 0, "seconds": 388.3, "max_memory_gb": 18.17, "ce": 0.3496905234642327, "rank": 0.0} +{"step": 207, "steps": 500, "lr": 2e-05, "grad_norm": 1.4709172248840332, "loss_scale": 4.0, "retries": 0, "seconds": 390.2, "max_memory_gb": 18.17, "ce": 0.4712314313510433, "rank": 0.15923911333084106} +{"step": 208, "steps": 500, "lr": 2e-05, "grad_norm": 1.1312158107757568, "loss_scale": 4.0, "retries": 0, "seconds": 392.0, "max_memory_gb": 18.17, "ce": 0.21376518212491646, "rank": 0.0} +{"step": 209, "steps": 500, "lr": 2e-05, "grad_norm": 1.1533665657043457, "loss_scale": 4.0, "retries": 0, "seconds": 393.9, "max_memory_gb": 18.17, "ce": 0.3321946281939745, "rank": 0.1253482848405838} +{"step": 210, "steps": 500, "lr": 2e-05, "grad_norm": 0.9656633138656616, "loss_scale": 4.0, "retries": 0, "seconds": 395.7, "max_memory_gb": 18.17, "ce": 0.3289892808534205, "rank": 0.04423344135284424} +{"step": 211, "steps": 500, "lr": 2e-05, "grad_norm": 1.3096563816070557, "loss_scale": 4.0, "retries": 0, "seconds": 397.6, "max_memory_gb": 18.17, "ce": 0.3300741300918162, "rank": 0.0} +{"step": 212, "steps": 500, "lr": 2e-05, "grad_norm": 2.109851360321045, "loss_scale": 4.0, "retries": 0, "seconds": 399.5, "max_memory_gb": 18.17, "ce": 0.48607944091781974, "rank": 0.0} +{"step": 213, "steps": 500, "lr": 2e-05, "grad_norm": 0.9658203721046448, "loss_scale": 4.0, "retries": 0, "seconds": 401.3, "max_memory_gb": 18.17, "ce": 0.3727518664672971, "rank": 0.0} +{"step": 214, "steps": 500, "lr": 2e-05, "grad_norm": 1.1403379440307617, "loss_scale": 4.0, "retries": 0, "seconds": 403.2, "max_memory_gb": 18.17, "ce": 0.3963618420821149, "rank": 0.0} +{"step": 215, "steps": 500, "lr": 2e-05, "grad_norm": 1.4430944919586182, "loss_scale": 4.0, "retries": 0, "seconds": 405.0, "max_memory_gb": 18.17, "ce": 0.4136714010965079, "rank": 0.02601759135723114} +{"step": 216, "steps": 500, "lr": 2e-05, "grad_norm": 1.0069817304611206, "loss_scale": 4.0, "retries": 0, "seconds": 406.9, "max_memory_gb": 18.17, "ce": 0.2904877728433348, "rank": 0.0} +{"step": 217, "steps": 500, "lr": 2e-05, "grad_norm": 0.6251105666160583, "loss_scale": 4.0, "retries": 0, "seconds": 408.7, "max_memory_gb": 18.17, "ce": 0.17601277632638812, "rank": 0.0} +{"step": 218, "steps": 500, "lr": 2e-05, "grad_norm": 4.43351936340332, "loss_scale": 4.0, "retries": 0, "seconds": 410.5, "max_memory_gb": 18.17, "ce": 0.3923408428672701, "rank": 0.0} +{"step": 219, "steps": 500, "lr": 2e-05, "grad_norm": 4.957935810089111, "loss_scale": 4.0, "retries": 0, "seconds": 412.4, "max_memory_gb": 18.17, "ce": 0.23793604655656964, "rank": 0.0} +{"step": 220, "steps": 500, "lr": 2e-05, "grad_norm": 1.4455927610397339, "loss_scale": 4.0, "retries": 0, "seconds": 414.2, "max_memory_gb": 18.17, "ce": 0.16766774677671492, "rank": 0.0} +{"step": 221, "steps": 500, "lr": 2e-05, "grad_norm": 1.8546887636184692, "loss_scale": 4.0, "retries": 0, "seconds": 416.1, "max_memory_gb": 18.17, "ce": 0.3315672962926328, "rank": 0.0} +{"step": 222, "steps": 500, "lr": 2e-05, "grad_norm": 4.6856842041015625, "loss_scale": 4.0, "retries": 0, "seconds": 417.9, "max_memory_gb": 18.17, "ce": 0.37729620561003685, "rank": 0.0} +{"step": 223, "steps": 500, "lr": 2e-05, "grad_norm": 0.4084673523902893, "loss_scale": 4.0, "retries": 0, "seconds": 419.8, "max_memory_gb": 18.17, "ce": 0.2056438132422045, "rank": 0.0} +{"step": 224, "steps": 500, "lr": 2e-05, "grad_norm": 1.4970661401748657, "loss_scale": 4.0, "retries": 0, "seconds": 421.6, "max_memory_gb": 18.17, "ce": 0.222198152448982, "rank": 0.09860088676214218} +{"step": 225, "steps": 500, "lr": 2e-05, "grad_norm": 1.8189018964767456, "loss_scale": 4.0, "retries": 0, "seconds": 423.5, "max_memory_gb": 18.17, "ce": 0.19285846268758178, "rank": 0.0} +{"step": 226, "steps": 500, "lr": 2e-05, "grad_norm": 1.990311861038208, "loss_scale": 4.0, "retries": 0, "seconds": 425.3, "max_memory_gb": 18.17, "ce": 0.21435921906959265, "rank": 0.0} +{"step": 227, "steps": 500, "lr": 2e-05, "grad_norm": 0.9250087738037109, "loss_scale": 4.0, "retries": 0, "seconds": 427.2, "max_memory_gb": 18.17, "ce": 0.13880681412410922, "rank": 0.0} +{"step": 228, "steps": 500, "lr": 2e-05, "grad_norm": 1.1336268186569214, "loss_scale": 4.0, "retries": 0, "seconds": 429.0, "max_memory_gb": 18.17, "ce": 0.15804847609251738, "rank": 0.0} +{"step": 229, "steps": 500, "lr": 2e-05, "grad_norm": 0.8880356550216675, "loss_scale": 4.0, "retries": 0, "seconds": 430.8, "max_memory_gb": 18.17, "ce": 0.2479379065334797, "rank": 0.0} +{"step": 230, "steps": 500, "lr": 2e-05, "grad_norm": 0.8933231830596924, "loss_scale": 4.0, "retries": 0, "seconds": 432.6, "max_memory_gb": 18.17, "ce": 0.2758224886201788, "rank": 0.0} +{"step": 231, "steps": 500, "lr": 2e-05, "grad_norm": 0.8338634371757507, "loss_scale": 4.0, "retries": 0, "seconds": 434.4, "max_memory_gb": 18.17, "ce": 0.2679590735351667, "rank": 0.0} +{"step": 232, "steps": 500, "lr": 2e-05, "grad_norm": 3.5330076217651367, "loss_scale": 4.0, "retries": 0, "seconds": 436.3, "max_memory_gb": 18.17, "ce": 0.21891334420070052, "rank": 0.0} +{"step": 233, "steps": 500, "lr": 2e-05, "grad_norm": 5.7518534660339355, "loss_scale": 4.0, "retries": 0, "seconds": 438.2, "max_memory_gb": 18.17, "ce": 0.3448429047130048, "rank": 0.0} +{"step": 234, "steps": 500, "lr": 2e-05, "grad_norm": 0.8920125365257263, "loss_scale": 4.0, "retries": 0, "seconds": 440.0, "max_memory_gb": 18.17, "ce": 0.2523936623474583, "rank": 0.0} +{"step": 235, "steps": 500, "lr": 2e-05, "grad_norm": 1.179348349571228, "loss_scale": 4.0, "retries": 0, "seconds": 441.8, "max_memory_gb": 18.17, "ce": 0.3065619010012597, "rank": 0.0} +{"step": 236, "steps": 500, "lr": 2e-05, "grad_norm": 1.3278857469558716, "loss_scale": 4.0, "retries": 0, "seconds": 443.7, "max_memory_gb": 18.17, "ce": 0.15871074463939294, "rank": 0.0} +{"step": 237, "steps": 500, "lr": 2e-05, "grad_norm": 3.472806692123413, "loss_scale": 4.0, "retries": 0, "seconds": 445.5, "max_memory_gb": 18.17, "ce": 0.3019354710704647, "rank": 0.0} +{"step": 238, "steps": 500, "lr": 2e-05, "grad_norm": 1.3845562934875488, "loss_scale": 4.0, "retries": 0, "seconds": 447.3, "max_memory_gb": 18.17, "ce": 0.13723037391901016, "rank": 0.0} +{"step": 239, "steps": 500, "lr": 2e-05, "grad_norm": 3.3141324520111084, "loss_scale": 4.0, "retries": 0, "seconds": 449.2, "max_memory_gb": 18.17, "ce": 0.2561354984063655, "rank": 0.0} +{"step": 240, "steps": 500, "lr": 2e-05, "grad_norm": 1.9616369009017944, "loss_scale": 4.0, "retries": 0, "seconds": 451.0, "max_memory_gb": 18.17, "ce": 0.1935816309414804, "rank": 0.0} +{"step": 241, "steps": 500, "lr": 2e-05, "grad_norm": 0.6753836274147034, "loss_scale": 4.0, "retries": 0, "seconds": 452.8, "max_memory_gb": 18.17, "ce": 0.190121412393637, "rank": 0.0} +{"step": 242, "steps": 500, "lr": 2e-05, "grad_norm": 2.7424356937408447, "loss_scale": 4.0, "retries": 0, "seconds": 454.7, "max_memory_gb": 18.17, "ce": 0.17824332253076136, "rank": 0.0} +{"step": 243, "steps": 500, "lr": 2e-05, "grad_norm": 3.451084852218628, "loss_scale": 4.0, "retries": 0, "seconds": 456.5, "max_memory_gb": 18.17, "ce": 0.22523496649228036, "rank": 0.0} +{"step": 244, "steps": 500, "lr": 2e-05, "grad_norm": 1.5524768829345703, "loss_scale": 4.0, "retries": 0, "seconds": 458.4, "max_memory_gb": 18.17, "ce": 0.3044751681154594, "rank": 0.0} +{"step": 245, "steps": 500, "lr": 2e-05, "grad_norm": 2.213240623474121, "loss_scale": 4.0, "retries": 0, "seconds": 460.2, "max_memory_gb": 18.17, "ce": 0.39896540297195315, "rank": 0.0} +{"step": 246, "steps": 500, "lr": 2e-05, "grad_norm": 3.0805776119232178, "loss_scale": 4.0, "retries": 0, "seconds": 462.1, "max_memory_gb": 18.17, "ce": 0.06445833115139976, "rank": 0.0} +{"step": 247, "steps": 500, "lr": 2e-05, "grad_norm": 3.275238513946533, "loss_scale": 4.0, "retries": 0, "seconds": 463.9, "max_memory_gb": 18.17, "ce": 0.16522069793427363, "rank": 0.0} +{"step": 248, "steps": 500, "lr": 2e-05, "grad_norm": 2.3051369190216064, "loss_scale": 4.0, "retries": 0, "seconds": 465.8, "max_memory_gb": 18.17, "ce": 0.14883325225673616, "rank": 0.0} +{"step": 249, "steps": 500, "lr": 2e-05, "grad_norm": 5.629601955413818, "loss_scale": 4.0, "retries": 0, "seconds": 467.6, "max_memory_gb": 18.17, "ce": 0.2445538236061111, "rank": 0.0} +{"step": 250, "steps": 500, "lr": 2e-05, "grad_norm": 1.7883679866790771, "loss_scale": 4.0, "retries": 0, "seconds": 469.5, "max_memory_gb": 18.17, "ce": 0.08429140262887813, "rank": 0.0} +{"step": 251, "steps": 500, "lr": 2e-05, "grad_norm": 0.7549654841423035, "loss_scale": 4.0, "retries": 0, "seconds": 471.3, "max_memory_gb": 18.17, "ce": 0.16128307092003524, "rank": 0.0} +{"step": 252, "steps": 500, "lr": 2e-05, "grad_norm": 1.6791713237762451, "loss_scale": 4.0, "retries": 0, "seconds": 473.2, "max_memory_gb": 18.17, "ce": 0.07924475480103865, "rank": 0.0} +{"step": 253, "steps": 500, "lr": 2e-05, "grad_norm": 0.9395402073860168, "loss_scale": 4.0, "retries": 0, "seconds": 475.0, "max_memory_gb": 18.17, "ce": 0.13164671760750934, "rank": 0.0} +{"step": 254, "steps": 500, "lr": 2e-05, "grad_norm": 26.54413604736328, "loss_scale": 4.0, "retries": 0, "seconds": 476.8, "max_memory_gb": 18.17, "ce": 0.2950153313577175, "rank": 0.0} +{"step": 255, "steps": 500, "lr": 2e-05, "grad_norm": 0.8281362056732178, "loss_scale": 4.0, "retries": 0, "seconds": 478.6, "max_memory_gb": 18.17, "ce": 0.25186843250412494, "rank": 0.0} +{"step": 256, "steps": 500, "lr": 2e-05, "grad_norm": 0.5768480896949768, "loss_scale": 4.0, "retries": 0, "seconds": 480.5, "max_memory_gb": 18.17, "ce": 0.10918986611068249, "rank": 0.0} +{"step": 257, "steps": 500, "lr": 2e-05, "grad_norm": 0.6418483257293701, "loss_scale": 4.0, "retries": 0, "seconds": 482.3, "max_memory_gb": 18.17, "ce": 0.0936252287356183, "rank": 0.0} +{"step": 258, "steps": 500, "lr": 2e-05, "grad_norm": 2.8840930461883545, "loss_scale": 4.0, "retries": 0, "seconds": 484.2, "max_memory_gb": 18.17, "ce": 0.1758249513222836, "rank": 0.0} +{"step": 259, "steps": 500, "lr": 2e-05, "grad_norm": 0.32826757431030273, "loss_scale": 4.0, "retries": 0, "seconds": 486.0, "max_memory_gb": 18.17, "ce": 0.05710457626264542, "rank": 0.0} +{"step": 260, "steps": 500, "lr": 2e-05, "grad_norm": 5.348422050476074, "loss_scale": 4.0, "retries": 0, "seconds": 487.9, "max_memory_gb": 18.17, "ce": 0.22164977341890335, "rank": 0.0} +{"step": 261, "steps": 500, "lr": 2e-05, "grad_norm": 3.149973154067993, "loss_scale": 4.0, "retries": 0, "seconds": 489.7, "max_memory_gb": 18.17, "ce": 0.16075515002012253, "rank": 0.0} +{"step": 262, "steps": 500, "lr": 2e-05, "grad_norm": 3.5020275115966797, "loss_scale": 4.0, "retries": 0, "seconds": 491.5, "max_memory_gb": 18.17, "ce": 0.23930403613485396, "rank": 0.0} +{"step": 263, "steps": 500, "lr": 2e-05, "grad_norm": 1.3472362756729126, "loss_scale": 4.0, "retries": 0, "seconds": 493.4, "max_memory_gb": 18.17, "ce": 0.22925331944134086, "rank": 0.12594865262508392} +{"step": 264, "steps": 500, "lr": 2e-05, "grad_norm": 1.316794991493225, "loss_scale": 4.0, "retries": 0, "seconds": 495.2, "max_memory_gb": 18.17, "ce": 0.166027192666661, "rank": 0.0} +{"step": 265, "steps": 500, "lr": 2e-05, "grad_norm": 1.3043718338012695, "loss_scale": 4.0, "retries": 0, "seconds": 497.1, "max_memory_gb": 18.17, "ce": 0.13768867764156312, "rank": 0.0} +{"step": 266, "steps": 500, "lr": 2e-05, "grad_norm": 6.505734920501709, "loss_scale": 4.0, "retries": 0, "seconds": 498.9, "max_memory_gb": 18.17, "ce": 0.3143903942545876, "rank": 0.0} +{"step": 267, "steps": 500, "lr": 2e-05, "grad_norm": 3.439303159713745, "loss_scale": 4.0, "retries": 0, "seconds": 500.8, "max_memory_gb": 18.17, "ce": 0.16492429288337007, "rank": 0.0} +{"step": 268, "steps": 500, "lr": 2e-05, "grad_norm": 0.5791064500808716, "loss_scale": 4.0, "retries": 0, "seconds": 502.6, "max_memory_gb": 18.17, "ce": 0.2631251228740439, "rank": 0.0} +{"step": 269, "steps": 500, "lr": 2e-05, "grad_norm": 5.955136775970459, "loss_scale": 4.0, "retries": 0, "seconds": 504.5, "max_memory_gb": 18.17, "ce": 0.46524939965456724, "rank": 0.0} +{"step": 270, "steps": 500, "lr": 2e-05, "grad_norm": 2.6546971797943115, "loss_scale": 4.0, "retries": 0, "seconds": 506.3, "max_memory_gb": 18.17, "ce": 0.17678714753128588, "rank": 0.0} +{"step": 271, "steps": 500, "lr": 2e-05, "grad_norm": 2.1258292198181152, "loss_scale": 4.0, "retries": 0, "seconds": 508.2, "max_memory_gb": 18.17, "ce": 0.10044115467462689, "rank": 0.06478782743215561} +{"step": 272, "steps": 500, "lr": 2e-05, "grad_norm": 7.206804275512695, "loss_scale": 4.0, "retries": 0, "seconds": 510.0, "max_memory_gb": 18.17, "ce": 0.15568211319623515, "rank": 0.0} +{"step": 273, "steps": 500, "lr": 2e-05, "grad_norm": 1.5142884254455566, "loss_scale": 4.0, "retries": 0, "seconds": 511.8, "max_memory_gb": 18.17, "ce": 0.06842996808700264, "rank": 0.12530362606048584} +{"step": 274, "steps": 500, "lr": 2e-05, "grad_norm": 1.0422825813293457, "loss_scale": 4.0, "retries": 0, "seconds": 513.7, "max_memory_gb": 18.17, "ce": 0.15124822803772986, "rank": 0.0} +{"step": 275, "steps": 500, "lr": 2e-05, "grad_norm": 5.756484508514404, "loss_scale": 4.0, "retries": 0, "seconds": 515.6, "max_memory_gb": 18.17, "ce": 0.6458257085178047, "rank": 0.0026091337203979492} +{"step": 276, "steps": 500, "lr": 2e-05, "grad_norm": 5.7189621925354, "loss_scale": 4.0, "retries": 0, "seconds": 517.4, "max_memory_gb": 18.17, "ce": 0.05916479125153273, "rank": 0.0} +{"step": 277, "steps": 500, "lr": 2e-05, "grad_norm": 2.7739765644073486, "loss_scale": 4.0, "retries": 0, "seconds": 519.3, "max_memory_gb": 18.17, "ce": 0.07343737184419297, "rank": 0.0} +{"step": 278, "steps": 500, "lr": 2e-05, "grad_norm": 3.320782423019409, "loss_scale": 4.0, "retries": 0, "seconds": 521.1, "max_memory_gb": 18.17, "ce": 0.20895804378233152, "rank": 0.0033371448516845703} +{"step": 279, "steps": 500, "lr": 2e-05, "grad_norm": 2.2320568561553955, "loss_scale": 4.0, "retries": 0, "seconds": 522.9, "max_memory_gb": 18.17, "ce": 0.100009580492042, "rank": 0.0} +{"step": 280, "steps": 500, "lr": 2e-05, "grad_norm": 0.7671866416931152, "loss_scale": 4.0, "retries": 0, "seconds": 524.8, "max_memory_gb": 18.17, "ce": 0.0472992408904247, "rank": 0.0} +{"step": 281, "steps": 500, "lr": 2e-05, "grad_norm": 3.326784133911133, "loss_scale": 4.0, "retries": 0, "seconds": 526.6, "max_memory_gb": 18.17, "ce": 0.24000732795684598, "rank": 0.0} +{"step": 282, "steps": 500, "lr": 2e-05, "grad_norm": 0.6514006853103638, "loss_scale": 4.0, "retries": 0, "seconds": 528.5, "max_memory_gb": 18.17, "ce": 0.11251818265009206, "rank": 0.0} +{"step": 283, "steps": 500, "lr": 2e-05, "grad_norm": 4.352471828460693, "loss_scale": 4.0, "retries": 0, "seconds": 530.3, "max_memory_gb": 18.17, "ce": 0.13319466140819713, "rank": 0.0} +{"step": 284, "steps": 500, "lr": 2e-05, "grad_norm": 8.775145530700684, "loss_scale": 4.0, "retries": 0, "seconds": 532.2, "max_memory_gb": 18.17, "ce": 0.13562426788848825, "rank": 0.0} +{"step": 285, "steps": 500, "lr": 2e-05, "grad_norm": 3.33146595954895, "loss_scale": 4.0, "retries": 0, "seconds": 534.0, "max_memory_gb": 18.17, "ce": 0.14548924716655165, "rank": 0.0} +{"step": 286, "steps": 500, "lr": 2e-05, "grad_norm": 1.3774371147155762, "loss_scale": 4.0, "retries": 0, "seconds": 535.9, "max_memory_gb": 18.17, "ce": 0.0748013217234984, "rank": 0.0} +{"step": 287, "steps": 500, "lr": 2e-05, "grad_norm": 6.862326622009277, "loss_scale": 4.0, "retries": 0, "seconds": 537.8, "max_memory_gb": 18.17, "ce": 0.1631076299236156, "rank": 0.0} +{"step": 288, "steps": 500, "lr": 2e-05, "grad_norm": 0.28242605924606323, "loss_scale": 4.0, "retries": 0, "seconds": 539.6, "max_memory_gb": 18.17, "ce": 0.03218087850837037, "rank": 0.0} +{"step": 289, "steps": 500, "lr": 2e-05, "grad_norm": 6.712115287780762, "loss_scale": 4.0, "retries": 0, "seconds": 541.5, "max_memory_gb": 18.17, "ce": 0.14249396324157715, "rank": 0.0} +{"step": 290, "steps": 500, "lr": 2e-05, "grad_norm": 1.4776952266693115, "loss_scale": 4.0, "retries": 0, "seconds": 543.3, "max_memory_gb": 18.17, "ce": 0.14318839405314066, "rank": 0.0} +{"step": 291, "steps": 500, "lr": 2e-05, "grad_norm": 1.8704453706741333, "loss_scale": 4.0, "retries": 0, "seconds": 545.2, "max_memory_gb": 18.17, "ce": 0.03594899435120169, "rank": 0.0} +{"step": 292, "steps": 500, "lr": 2e-05, "grad_norm": 10.96335506439209, "loss_scale": 4.0, "retries": 0, "seconds": 547.0, "max_memory_gb": 18.17, "ce": 0.2114589098491706, "rank": 0.0} +{"step": 293, "steps": 500, "lr": 2e-05, "grad_norm": 0.9361065626144409, "loss_scale": 4.0, "retries": 0, "seconds": 548.9, "max_memory_gb": 18.17, "ce": 0.08605646289652213, "rank": 0.0} +{"step": 294, "steps": 500, "lr": 2e-05, "grad_norm": 6.840836048126221, "loss_scale": 4.0, "retries": 0, "seconds": 550.7, "max_memory_gb": 18.17, "ce": 0.09042963449610397, "rank": 0.0} +{"step": 295, "steps": 500, "lr": 2e-05, "grad_norm": 71.57910919189453, "loss_scale": 4.0, "retries": 0, "seconds": 552.5, "max_memory_gb": 18.17, "ce": 0.4744026130065322, "rank": 0.0} +{"step": 296, "steps": 500, "lr": 2e-05, "grad_norm": 30.02859115600586, "loss_scale": 4.0, "retries": 0, "seconds": 554.4, "max_memory_gb": 18.17, "ce": 0.3453582909423858, "rank": 0.0} +{"step": 297, "steps": 500, "lr": 2e-05, "grad_norm": 1.2018812894821167, "loss_scale": 4.0, "retries": 0, "seconds": 556.2, "max_memory_gb": 18.17, "ce": 0.1681317129987292, "rank": 0.0} +{"step": 298, "steps": 500, "lr": 2e-05, "grad_norm": 7.656444549560547, "loss_scale": 4.0, "retries": 0, "seconds": 558.0, "max_memory_gb": 18.17, "ce": 0.5243908331904095, "rank": 0.0} +{"step": 299, "steps": 500, "lr": 2e-05, "grad_norm": 4.734877109527588, "loss_scale": 4.0, "retries": 0, "seconds": 559.9, "max_memory_gb": 18.17, "ce": 0.0888716047629714, "rank": 0.0} +{"step": 300, "steps": 500, "lr": 2e-05, "grad_norm": 0.7638217806816101, "loss_scale": 4.0, "retries": 0, "seconds": 561.7, "max_memory_gb": 18.17, "ce": 0.2674721145594958, "rank": 0.0} +{"step": 301, "steps": 500, "lr": 2e-05, "grad_norm": 0.8696886301040649, "loss_scale": 4.0, "retries": 0, "seconds": 563.5, "max_memory_gb": 18.17, "ce": 0.124376509222202, "rank": 0.0} +{"step": 302, "steps": 500, "lr": 2e-05, "grad_norm": 1.5765416622161865, "loss_scale": 4.0, "retries": 0, "seconds": 565.4, "max_memory_gb": 18.17, "ce": 0.04255091602681205, "rank": 0.0} +{"step": 303, "steps": 500, "lr": 2e-05, "grad_norm": 44.052799224853516, "loss_scale": 4.0, "retries": 0, "seconds": 567.2, "max_memory_gb": 18.17, "ce": 0.11358820192981511, "rank": 0.0} +{"step": 304, "steps": 500, "lr": 2e-05, "grad_norm": 8.664243698120117, "loss_scale": 4.0, "retries": 0, "seconds": 569.1, "max_memory_gb": 18.17, "ce": 0.11807891284115613, "rank": 0.10438686609268188} +{"step": 305, "steps": 500, "lr": 2e-05, "grad_norm": 16.1278076171875, "loss_scale": 4.0, "retries": 0, "seconds": 571.0, "max_memory_gb": 18.17, "ce": 0.2581215335521847, "rank": 0.0} +{"step": 306, "steps": 500, "lr": 2e-05, "grad_norm": 3.9710919857025146, "loss_scale": 4.0, "retries": 0, "seconds": 572.8, "max_memory_gb": 18.17, "ce": 0.0687141028320184, "rank": 0.0} +{"step": 307, "steps": 500, "lr": 2e-05, "grad_norm": 3.484255313873291, "loss_scale": 4.0, "retries": 0, "seconds": 574.7, "max_memory_gb": 18.17, "ce": 0.14325139843276702, "rank": 0.0} +{"step": 308, "steps": 500, "lr": 2e-05, "grad_norm": 7.476601600646973, "loss_scale": 4.0, "retries": 0, "seconds": 576.6, "max_memory_gb": 18.17, "ce": 0.23407218103238847, "rank": 0.0} +{"step": 309, "steps": 500, "lr": 2e-05, "grad_norm": 0.8228901028633118, "loss_scale": 4.0, "retries": 0, "seconds": 578.4, "max_memory_gb": 18.17, "ce": 0.19929463986045448, "rank": 0.0} +{"step": 310, "steps": 500, "lr": 2e-05, "grad_norm": 1.970963478088379, "loss_scale": 4.0, "retries": 0, "seconds": 580.2, "max_memory_gb": 18.17, "ce": 0.06732066796394065, "rank": 0.0} +{"step": 311, "steps": 500, "lr": 2e-05, "grad_norm": 1.1479547023773193, "loss_scale": 4.0, "retries": 0, "seconds": 582.0, "max_memory_gb": 18.17, "ce": 0.12208871939219534, "rank": 0.0} +{"step": 312, "steps": 500, "lr": 2e-05, "grad_norm": 1.898148775100708, "loss_scale": 4.0, "retries": 0, "seconds": 583.9, "max_memory_gb": 18.17, "ce": 0.05074623064137995, "rank": 0.0} +{"step": 313, "steps": 500, "lr": 2e-05, "grad_norm": 10.250323295593262, "loss_scale": 4.0, "retries": 0, "seconds": 585.7, "max_memory_gb": 18.17, "ce": 0.19776888890191913, "rank": 0.0} +{"step": 314, "steps": 500, "lr": 2e-05, "grad_norm": 2.197197675704956, "loss_scale": 4.0, "retries": 0, "seconds": 587.7, "max_memory_gb": 18.17, "ce": 0.22245932830264792, "rank": 0.0} +{"step": 315, "steps": 500, "lr": 2e-05, "grad_norm": 1.770250916481018, "loss_scale": 4.0, "retries": 0, "seconds": 589.5, "max_memory_gb": 18.17, "ce": 0.08834576164372265, "rank": 0.09182214736938477} +{"step": 316, "steps": 500, "lr": 2e-05, "grad_norm": 19.32260513305664, "loss_scale": 4.0, "retries": 0, "seconds": 591.4, "max_memory_gb": 18.17, "ce": 0.4129464755533263, "rank": 0.0} +{"step": 317, "steps": 500, "lr": 2e-05, "grad_norm": 3.8475441932678223, "loss_scale": 4.0, "retries": 0, "seconds": 593.2, "max_memory_gb": 18.17, "ce": 0.1489269791636616, "rank": 0.0} +{"step": 318, "steps": 500, "lr": 2e-05, "grad_norm": 27.890474319458008, "loss_scale": 4.0, "retries": 0, "seconds": 595.0, "max_memory_gb": 18.17, "ce": 0.44651309511391446, "rank": 0.0} +{"step": 319, "steps": 500, "lr": 2e-05, "grad_norm": 6.2882981300354, "loss_scale": 4.0, "retries": 0, "seconds": 596.9, "max_memory_gb": 18.17, "ce": 0.10630160802975297, "rank": 0.0} +{"step": 320, "steps": 500, "lr": 2e-05, "grad_norm": 3.0731585025787354, "loss_scale": 4.0, "retries": 0, "seconds": 598.7, "max_memory_gb": 18.17, "ce": 0.14897616731468588, "rank": 0.0} +{"step": 321, "steps": 500, "lr": 2e-05, "grad_norm": 0.9186423420906067, "loss_scale": 4.0, "retries": 0, "seconds": 600.5, "max_memory_gb": 18.17, "ce": 0.12288579333107919, "rank": 0.12605313956737518} +{"step": 322, "steps": 500, "lr": 2e-05, "grad_norm": 1.763225793838501, "loss_scale": 4.0, "retries": 0, "seconds": 602.4, "max_memory_gb": 18.17, "ce": 0.059073525306303054, "rank": 0.0} +{"step": 323, "steps": 500, "lr": 2e-05, "grad_norm": 5.915005683898926, "loss_scale": 4.0, "retries": 0, "seconds": 604.2, "max_memory_gb": 18.17, "ce": 0.16753336257534102, "rank": 0.0} +{"step": 324, "steps": 500, "lr": 2e-05, "grad_norm": 6.958407878875732, "loss_scale": 4.0, "retries": 0, "seconds": 606.1, "max_memory_gb": 18.17, "ce": 0.09940689218637999, "rank": 0.0} +{"step": 325, "steps": 500, "lr": 2e-05, "grad_norm": 0.8229695558547974, "loss_scale": 4.0, "retries": 0, "seconds": 607.9, "max_memory_gb": 18.17, "ce": 0.22166190658754203, "rank": 0.0} +{"step": 326, "steps": 500, "lr": 2e-05, "grad_norm": 0.4753699004650116, "loss_scale": 4.0, "retries": 0, "seconds": 609.8, "max_memory_gb": 18.17, "ce": 0.07702458590938477, "rank": 0.0999685674905777} +{"step": 327, "steps": 500, "lr": 2e-05, "grad_norm": 15.04274845123291, "loss_scale": 4.0, "retries": 0, "seconds": 611.7, "max_memory_gb": 18.17, "ce": 0.34322882404376287, "rank": 0.0} +{"step": 328, "steps": 500, "lr": 2e-05, "grad_norm": 3.2221219539642334, "loss_scale": 4.0, "retries": 0, "seconds": 613.5, "max_memory_gb": 18.17, "ce": 0.19556531695707235, "rank": 0.0} +{"step": 329, "steps": 500, "lr": 2e-05, "grad_norm": 0.7852571606636047, "loss_scale": 4.0, "retries": 0, "seconds": 615.3, "max_memory_gb": 18.17, "ce": 0.17600741318892688, "rank": 0.0} +{"step": 330, "steps": 500, "lr": 2e-05, "grad_norm": 16.660640716552734, "loss_scale": 4.0, "retries": 0, "seconds": 617.2, "max_memory_gb": 18.17, "ce": 0.1805495161388535, "rank": 0.0} +{"step": 331, "steps": 500, "lr": 2e-05, "grad_norm": 5.794269561767578, "loss_scale": 4.0, "retries": 0, "seconds": 619.0, "max_memory_gb": 18.17, "ce": 0.3415280419867486, "rank": 0.0} +{"step": 332, "steps": 500, "lr": 2e-05, "grad_norm": 3.797287702560425, "loss_scale": 4.0, "retries": 0, "seconds": 620.9, "max_memory_gb": 18.17, "ce": 0.1358041614294052, "rank": 0.0} +{"step": 333, "steps": 500, "lr": 2e-05, "grad_norm": 1.486878752708435, "loss_scale": 4.0, "retries": 0, "seconds": 622.7, "max_memory_gb": 18.17, "ce": 0.2720249443082139, "rank": 0.0} +{"step": 334, "steps": 500, "lr": 2e-05, "grad_norm": 1.6331740617752075, "loss_scale": 4.0, "retries": 0, "seconds": 624.6, "max_memory_gb": 18.17, "ce": 0.14458410485531203, "rank": 0.0} +{"step": 335, "steps": 500, "lr": 2e-05, "grad_norm": 4.863023281097412, "loss_scale": 4.0, "retries": 0, "seconds": 626.4, "max_memory_gb": 18.17, "ce": 0.3450025772617664, "rank": 0.0} +{"step": 336, "steps": 500, "lr": 2e-05, "grad_norm": 5.8095269203186035, "loss_scale": 4.0, "retries": 0, "seconds": 628.3, "max_memory_gb": 18.17, "ce": 0.08643359070993029, "rank": 0.0} +{"step": 337, "steps": 500, "lr": 2e-05, "grad_norm": 3.5736751556396484, "loss_scale": 4.0, "retries": 0, "seconds": 630.1, "max_memory_gb": 18.17, "ce": 0.2638394118985161, "rank": 0.0} +{"step": 338, "steps": 500, "lr": 2e-05, "grad_norm": 1.5724406242370605, "loss_scale": 4.0, "retries": 0, "seconds": 632.0, "max_memory_gb": 18.17, "ce": 0.2247715578705538, "rank": 0.0} +{"step": 339, "steps": 500, "lr": 2e-05, "grad_norm": 3.674265146255493, "loss_scale": 4.0, "retries": 0, "seconds": 633.8, "max_memory_gb": 18.17, "ce": 0.14927357851411216, "rank": 0.0} +{"step": 340, "steps": 500, "lr": 2e-05, "grad_norm": 4.729248523712158, "loss_scale": 4.0, "retries": 0, "seconds": 635.6, "max_memory_gb": 18.17, "ce": 0.22998168016783893, "rank": 0.0} +{"step": 341, "steps": 500, "lr": 2e-05, "grad_norm": 2.372666597366333, "loss_scale": 4.0, "retries": 0, "seconds": 637.5, "max_memory_gb": 18.17, "ce": 0.1832552685518749, "rank": 0.0} +{"step": 342, "steps": 500, "lr": 2e-05, "grad_norm": 2.8438668251037598, "loss_scale": 4.0, "retries": 0, "seconds": 639.3, "max_memory_gb": 18.17, "ce": 0.5080258885864168, "rank": 0.0} +{"step": 343, "steps": 500, "lr": 2e-05, "grad_norm": 4.042470932006836, "loss_scale": 4.0, "retries": 0, "seconds": 641.2, "max_memory_gb": 18.17, "ce": 0.12117551093979273, "rank": 0.0} +{"step": 344, "steps": 500, "lr": 2e-05, "grad_norm": 0.5666807293891907, "loss_scale": 4.0, "retries": 0, "seconds": 643.0, "max_memory_gb": 18.17, "ce": 0.05025787465274334, "rank": 0.0} +{"step": 345, "steps": 500, "lr": 2e-05, "grad_norm": 0.5611703991889954, "loss_scale": 4.0, "retries": 0, "seconds": 644.8, "max_memory_gb": 18.17, "ce": 0.20962057744327467, "rank": 0.0} +{"step": 346, "steps": 500, "lr": 2e-05, "grad_norm": 1.2970689535140991, "loss_scale": 4.0, "retries": 0, "seconds": 646.7, "max_memory_gb": 18.17, "ce": 0.20819764063344337, "rank": 0.0} +{"step": 347, "steps": 500, "lr": 2e-05, "grad_norm": 4.049383640289307, "loss_scale": 4.0, "retries": 0, "seconds": 648.5, "max_memory_gb": 18.17, "ce": 0.179124369926285, "rank": 0.0} +{"step": 348, "steps": 500, "lr": 2e-05, "grad_norm": 1.3990139961242676, "loss_scale": 4.0, "retries": 0, "seconds": 650.4, "max_memory_gb": 18.17, "ce": 0.03508557174063753, "rank": 0.0} +{"step": 349, "steps": 500, "lr": 2e-05, "grad_norm": 3.3678219318389893, "loss_scale": 4.0, "retries": 0, "seconds": 652.2, "max_memory_gb": 18.17, "ce": 0.14167416575946845, "rank": 0.0} +{"step": 350, "steps": 500, "lr": 2e-05, "grad_norm": 1.5664384365081787, "loss_scale": 4.0, "retries": 0, "seconds": 654.0, "max_memory_gb": 18.17, "ce": 0.18154043331742287, "rank": 0.0} +{"step": 351, "steps": 500, "lr": 2e-05, "grad_norm": 2.086874008178711, "loss_scale": 4.0, "retries": 0, "seconds": 655.9, "max_memory_gb": 18.17, "ce": 0.08276878175092861, "rank": 0.0} +{"step": 352, "steps": 500, "lr": 2e-05, "grad_norm": 4.032617568969727, "loss_scale": 4.0, "retries": 0, "seconds": 657.7, "max_memory_gb": 18.17, "ce": 0.299609688896453, "rank": 0.0} +{"step": 353, "steps": 500, "lr": 2e-05, "grad_norm": 5.426660060882568, "loss_scale": 4.0, "retries": 0, "seconds": 659.5, "max_memory_gb": 18.17, "ce": 0.20554258124320768, "rank": 0.0} +{"step": 354, "steps": 500, "lr": 2e-05, "grad_norm": 3.298055648803711, "loss_scale": 4.0, "retries": 0, "seconds": 661.4, "max_memory_gb": 18.17, "ce": 0.2084075345483143, "rank": 0.0} +{"step": 355, "steps": 500, "lr": 2e-05, "grad_norm": 3.832282781600952, "loss_scale": 4.0, "retries": 0, "seconds": 663.2, "max_memory_gb": 18.17, "ce": 0.16608664531668182, "rank": 0.0} +{"step": 356, "steps": 500, "lr": 2e-05, "grad_norm": 4.504091739654541, "loss_scale": 4.0, "retries": 0, "seconds": 665.1, "max_memory_gb": 18.17, "ce": 0.07757969309750479, "rank": 0.0} +{"step": 357, "steps": 500, "lr": 2e-05, "grad_norm": 0.6909552812576294, "loss_scale": 4.0, "retries": 0, "seconds": 667.0, "max_memory_gb": 18.17, "ce": 0.06952157246996649, "rank": 0.0} +{"step": 358, "steps": 500, "lr": 2e-05, "grad_norm": 1.3461816310882568, "loss_scale": 4.0, "retries": 0, "seconds": 668.8, "max_memory_gb": 18.17, "ce": 0.23715008446015418, "rank": 0.0} +{"step": 359, "steps": 500, "lr": 2e-05, "grad_norm": 22.17321014404297, "loss_scale": 4.0, "retries": 0, "seconds": 670.6, "max_memory_gb": 18.17, "ce": 0.121746318298392, "rank": 0.0} +{"step": 360, "steps": 500, "lr": 2e-05, "grad_norm": 2.0308303833007812, "loss_scale": 4.0, "retries": 0, "seconds": 672.5, "max_memory_gb": 18.17, "ce": 0.05230529139225837, "rank": 0.050535865128040314} +{"step": 361, "steps": 500, "lr": 2e-05, "grad_norm": 3.3125550746917725, "loss_scale": 4.0, "retries": 0, "seconds": 674.3, "max_memory_gb": 18.17, "ce": 0.14781093546480406, "rank": 0.0} +{"step": 362, "steps": 500, "lr": 2e-05, "grad_norm": 5.254147529602051, "loss_scale": 4.0, "retries": 0, "seconds": 676.1, "max_memory_gb": 18.17, "ce": 0.3563809924235102, "rank": 0.0} +{"step": 363, "steps": 500, "lr": 2e-05, "grad_norm": 16.792476654052734, "loss_scale": 4.0, "retries": 0, "seconds": 678.0, "max_memory_gb": 18.17, "ce": 0.12830875147483312, "rank": 0.0} +{"step": 364, "steps": 500, "lr": 2e-05, "grad_norm": 1.693246841430664, "loss_scale": 4.0, "retries": 0, "seconds": 679.8, "max_memory_gb": 18.17, "ce": 0.0914934923639521, "rank": 0.0} +{"step": 365, "steps": 500, "lr": 2e-05, "grad_norm": 0.6741579174995422, "loss_scale": 4.0, "retries": 0, "seconds": 681.7, "max_memory_gb": 18.17, "ce": 0.14916455309139565, "rank": 0.0} +{"step": 366, "steps": 500, "lr": 2e-05, "grad_norm": 6.241971969604492, "loss_scale": 4.0, "retries": 0, "seconds": 683.5, "max_memory_gb": 18.17, "ce": 0.13410162247600965, "rank": 0.0} +{"step": 367, "steps": 500, "lr": 2e-05, "grad_norm": 3.618396282196045, "loss_scale": 4.0, "retries": 0, "seconds": 685.3, "max_memory_gb": 18.17, "ce": 0.1622450565773761, "rank": 0.0} +{"step": 368, "steps": 500, "lr": 2e-05, "grad_norm": 0.6516332030296326, "loss_scale": 4.0, "retries": 0, "seconds": 687.1, "max_memory_gb": 18.17, "ce": 0.035342997798579745, "rank": 0.0} +{"step": 369, "steps": 500, "lr": 2e-05, "grad_norm": 0.7039018869400024, "loss_scale": 4.0, "retries": 0, "seconds": 689.0, "max_memory_gb": 18.17, "ce": 0.07054633022926282, "rank": 0.11341027915477753} +{"step": 370, "steps": 500, "lr": 2e-05, "grad_norm": 2.1821765899658203, "loss_scale": 4.0, "retries": 0, "seconds": 690.8, "max_memory_gb": 18.17, "ce": 0.2925555561669171, "rank": 0.0} +{"step": 371, "steps": 500, "lr": 2e-05, "grad_norm": 5.6143293380737305, "loss_scale": 4.0, "retries": 0, "seconds": 692.7, "max_memory_gb": 18.17, "ce": 0.20142777130240574, "rank": 0.0} +{"step": 372, "steps": 500, "lr": 2e-05, "grad_norm": 3.680074691772461, "loss_scale": 4.0, "retries": 0, "seconds": 694.5, "max_memory_gb": 18.17, "ce": 0.09741189662599936, "rank": 0.0} +{"step": 373, "steps": 500, "lr": 2e-05, "grad_norm": 0.6496761441230774, "loss_scale": 4.0, "retries": 0, "seconds": 696.3, "max_memory_gb": 18.17, "ce": 0.1175357969768811, "rank": 0.0} +{"step": 374, "steps": 500, "lr": 2e-05, "grad_norm": 0.6110003590583801, "loss_scale": 4.0, "retries": 0, "seconds": 698.2, "max_memory_gb": 18.17, "ce": 0.1575149761629291, "rank": 0.0} +{"step": 375, "steps": 500, "lr": 2e-05, "grad_norm": 0.8474687337875366, "loss_scale": 4.0, "retries": 0, "seconds": 700.0, "max_memory_gb": 18.17, "ce": 0.18244409690669272, "rank": 0.0} +{"step": 376, "steps": 500, "lr": 2e-05, "grad_norm": 0.7110260128974915, "loss_scale": 4.0, "retries": 0, "seconds": 701.8, "max_memory_gb": 18.17, "ce": 0.08187109666323522, "rank": 0.0} +{"step": 377, "steps": 500, "lr": 2e-05, "grad_norm": 2.505767345428467, "loss_scale": 4.0, "retries": 0, "seconds": 703.7, "max_memory_gb": 18.17, "ce": 0.1420202989684185, "rank": 0.0} +{"step": 378, "steps": 500, "lr": 2e-05, "grad_norm": 5.644346714019775, "loss_scale": 4.0, "retries": 0, "seconds": 705.5, "max_memory_gb": 18.17, "ce": 0.0813139442470856, "rank": 0.0} +{"step": 379, "steps": 500, "lr": 2e-05, "grad_norm": 0.7308553457260132, "loss_scale": 4.0, "retries": 0, "seconds": 707.3, "max_memory_gb": 18.17, "ce": 0.1201493994158227, "rank": 0.0} +{"step": 380, "steps": 500, "lr": 2e-05, "grad_norm": 16.79548454284668, "loss_scale": 4.0, "retries": 0, "seconds": 709.2, "max_memory_gb": 18.17, "ce": 0.2795345511549385, "rank": 0.0} +{"step": 381, "steps": 500, "lr": 2e-05, "grad_norm": 1.0819698572158813, "loss_scale": 4.0, "retries": 0, "seconds": 711.0, "max_memory_gb": 18.17, "ce": 0.26654489239444956, "rank": 0.0} +{"step": 382, "steps": 500, "lr": 2e-05, "grad_norm": 3.001842737197876, "loss_scale": 4.0, "retries": 0, "seconds": 712.8, "max_memory_gb": 18.17, "ce": 0.3647759665036574, "rank": 0.0} +{"step": 383, "steps": 500, "lr": 2e-05, "grad_norm": 1.1643455028533936, "loss_scale": 4.0, "retries": 0, "seconds": 714.7, "max_memory_gb": 18.17, "ce": 0.11582552516483702, "rank": 0.0} +{"step": 384, "steps": 500, "lr": 2e-05, "grad_norm": 2.321894645690918, "loss_scale": 4.0, "retries": 0, "seconds": 716.5, "max_memory_gb": 18.17, "ce": 0.05662562279030681, "rank": 0.0} +{"step": 385, "steps": 500, "lr": 2e-05, "grad_norm": 0.7227669954299927, "loss_scale": 4.0, "retries": 0, "seconds": 718.4, "max_memory_gb": 18.17, "ce": 0.1666227568202885, "rank": 0.0} +{"step": 386, "steps": 500, "lr": 2e-05, "grad_norm": 0.6124075055122375, "loss_scale": 4.0, "retries": 0, "seconds": 720.2, "max_memory_gb": 18.17, "ce": 0.06595927278976887, "rank": 0.0} +{"step": 387, "steps": 500, "lr": 2e-05, "grad_norm": 0.44190704822540283, "loss_scale": 4.0, "retries": 0, "seconds": 722.1, "max_memory_gb": 18.17, "ce": 0.08258380895131268, "rank": 0.0} +{"step": 388, "steps": 500, "lr": 2e-05, "grad_norm": 5.321372032165527, "loss_scale": 4.0, "retries": 0, "seconds": 723.9, "max_memory_gb": 18.17, "ce": 0.09018325546639971, "rank": 0.0} +{"step": 389, "steps": 500, "lr": 2e-05, "grad_norm": 3.811685800552368, "loss_scale": 4.0, "retries": 0, "seconds": 725.8, "max_memory_gb": 18.17, "ce": 0.04275461292127147, "rank": 0.0} +{"step": 390, "steps": 500, "lr": 2e-05, "grad_norm": 3.988795518875122, "loss_scale": 4.0, "retries": 0, "seconds": 727.6, "max_memory_gb": 18.17, "ce": 0.06728041410678998, "rank": 0.0} +{"step": 391, "steps": 500, "lr": 2e-05, "grad_norm": 2.374720811843872, "loss_scale": 4.0, "retries": 0, "seconds": 729.5, "max_memory_gb": 18.17, "ce": 0.2731023294618353, "rank": 0.0} +{"step": 392, "steps": 500, "lr": 2e-05, "grad_norm": 0.3555435836315155, "loss_scale": 4.0, "retries": 0, "seconds": 731.3, "max_memory_gb": 18.17, "ce": 0.02623169412072457, "rank": 0.11858414113521576} +{"step": 393, "steps": 500, "lr": 2e-05, "grad_norm": 4.506961822509766, "loss_scale": 4.0, "retries": 0, "seconds": 733.2, "max_memory_gb": 18.17, "ce": 0.1753736769896932, "rank": 0.0} +{"step": 394, "steps": 500, "lr": 2e-05, "grad_norm": 2.4847664833068848, "loss_scale": 4.0, "retries": 0, "seconds": 735.0, "max_memory_gb": 18.17, "ce": 0.12878174720390234, "rank": 0.0} +{"step": 395, "steps": 500, "lr": 2e-05, "grad_norm": 0.8440645933151245, "loss_scale": 4.0, "retries": 0, "seconds": 736.8, "max_memory_gb": 18.17, "ce": 0.044488796243967954, "rank": 0.0} +{"step": 396, "steps": 500, "lr": 2e-05, "grad_norm": 13.771637916564941, "loss_scale": 4.0, "retries": 0, "seconds": 738.7, "max_memory_gb": 18.17, "ce": 0.10704880533376127, "rank": 0.0} +{"step": 397, "steps": 500, "lr": 2e-05, "grad_norm": 0.4344596862792969, "loss_scale": 4.0, "retries": 0, "seconds": 740.5, "max_memory_gb": 18.17, "ce": 0.150197685463354, "rank": 0.0} +{"step": 398, "steps": 500, "lr": 2e-05, "grad_norm": 1.2694936990737915, "loss_scale": 4.0, "retries": 0, "seconds": 742.3, "max_memory_gb": 18.17, "ce": 0.09711323637748137, "rank": 0.0} +{"step": 399, "steps": 500, "lr": 2e-05, "grad_norm": 1.3010939359664917, "loss_scale": 4.0, "retries": 0, "seconds": 744.2, "max_memory_gb": 18.17, "ce": 0.15889496631552902, "rank": 0.0} +{"step": 400, "steps": 500, "lr": 2e-05, "grad_norm": 28.878093719482422, "loss_scale": 4.0, "retries": 0, "seconds": 746.0, "max_memory_gb": 18.17, "ce": 0.23647282761521637, "rank": 0.0} +{"step": 401, "steps": 500, "lr": 2e-05, "grad_norm": 3.8545782566070557, "loss_scale": 4.0, "retries": 0, "seconds": 747.9, "max_memory_gb": 18.17, "ce": 0.1958964994410053, "rank": 0.0} +{"step": 402, "steps": 500, "lr": 2e-05, "grad_norm": 0.6186238527297974, "loss_scale": 4.0, "retries": 0, "seconds": 749.7, "max_memory_gb": 18.17, "ce": 0.13383020155015402, "rank": 0.09749872982501984} +{"step": 403, "steps": 500, "lr": 2e-05, "grad_norm": 1.7608319520950317, "loss_scale": 4.0, "retries": 0, "seconds": 751.6, "max_memory_gb": 18.17, "ce": 0.027688609225151595, "rank": 0.0} +{"step": 404, "steps": 500, "lr": 2e-05, "grad_norm": 1.130793809890747, "loss_scale": 4.0, "retries": 0, "seconds": 753.4, "max_memory_gb": 18.17, "ce": 0.17876633384730667, "rank": 0.0} +{"step": 405, "steps": 500, "lr": 2e-05, "grad_norm": 0.37538620829582214, "loss_scale": 4.0, "retries": 0, "seconds": 755.2, "max_memory_gb": 18.17, "ce": 0.04578254243460833, "rank": 0.0} +{"step": 406, "steps": 500, "lr": 2e-05, "grad_norm": 0.8831842541694641, "loss_scale": 4.0, "retries": 0, "seconds": 757.1, "max_memory_gb": 18.17, "ce": 0.15267978773772484, "rank": 0.0} +{"step": 407, "steps": 500, "lr": 2e-05, "grad_norm": 2.8330469131469727, "loss_scale": 4.0, "retries": 0, "seconds": 758.9, "max_memory_gb": 18.17, "ce": 0.6851239247334888, "rank": 0.10318619012832642} +{"step": 408, "steps": 500, "lr": 2e-05, "grad_norm": 3.743734359741211, "loss_scale": 4.0, "retries": 0, "seconds": 760.7, "max_memory_gb": 18.17, "ce": 0.366079476341838, "rank": 0.0} +{"step": 409, "steps": 500, "lr": 2e-05, "grad_norm": 1.5957590341567993, "loss_scale": 4.0, "retries": 0, "seconds": 762.6, "max_memory_gb": 18.17, "ce": 0.17757937840360682, "rank": 0.0} +{"step": 410, "steps": 500, "lr": 2e-05, "grad_norm": 7.464216709136963, "loss_scale": 4.0, "retries": 0, "seconds": 764.4, "max_memory_gb": 18.17, "ce": 0.10378722878704139, "rank": 0.0} +{"step": 411, "steps": 500, "lr": 2e-05, "grad_norm": 33.33526611328125, "loss_scale": 4.0, "retries": 0, "seconds": 766.3, "max_memory_gb": 18.17, "ce": 0.5225244432367617, "rank": 0.0} +{"step": 412, "steps": 500, "lr": 2e-05, "grad_norm": 0.44758903980255127, "loss_scale": 4.0, "retries": 0, "seconds": 768.1, "max_memory_gb": 18.17, "ce": 0.16788151866785483, "rank": 0.12500590085983276} +{"step": 413, "steps": 500, "lr": 2e-05, "grad_norm": 0.5702083110809326, "loss_scale": 4.0, "retries": 0, "seconds": 769.9, "max_memory_gb": 18.17, "ce": 0.049485267605632544, "rank": 0.0} +{"step": 414, "steps": 500, "lr": 2e-05, "grad_norm": 5.7333574295043945, "loss_scale": 4.0, "retries": 0, "seconds": 771.8, "max_memory_gb": 18.17, "ce": 0.241478760195605, "rank": 0.0} +{"step": 415, "steps": 500, "lr": 2e-05, "grad_norm": 6.329236030578613, "loss_scale": 4.0, "retries": 0, "seconds": 773.6, "max_memory_gb": 18.17, "ce": 0.40119029812922236, "rank": 0.0} +{"step": 416, "steps": 500, "lr": 2e-05, "grad_norm": 3.437922954559326, "loss_scale": 4.0, "retries": 0, "seconds": 775.4, "max_memory_gb": 18.17, "ce": 0.044861297097668285, "rank": 0.0} +{"step": 417, "steps": 500, "lr": 2e-05, "grad_norm": 3.3668034076690674, "loss_scale": 4.0, "retries": 0, "seconds": 777.2, "max_memory_gb": 18.17, "ce": 0.42242934479145333, "rank": 0.1311703622341156} +{"step": 418, "steps": 500, "lr": 2e-05, "grad_norm": 1.4987531900405884, "loss_scale": 4.0, "retries": 0, "seconds": 779.1, "max_memory_gb": 18.17, "ce": 0.09331682469201041, "rank": 0.0} +{"step": 419, "steps": 500, "lr": 2e-05, "grad_norm": 0.4090936481952667, "loss_scale": 4.0, "retries": 0, "seconds": 780.8, "max_memory_gb": 18.17, "ce": 0.09880441805944429, "rank": 0.0} +{"step": 420, "steps": 500, "lr": 2e-05, "grad_norm": 3.2684857845306396, "loss_scale": 4.0, "retries": 0, "seconds": 782.7, "max_memory_gb": 18.17, "ce": 0.34156044156407006, "rank": 0.0} +{"step": 421, "steps": 500, "lr": 2e-05, "grad_norm": 0.44602951407432556, "loss_scale": 4.0, "retries": 0, "seconds": 784.5, "max_memory_gb": 18.17, "ce": 0.05426090991386445, "rank": 0.12621161341667175} +{"step": 422, "steps": 500, "lr": 2e-05, "grad_norm": 8.220174789428711, "loss_scale": 4.0, "retries": 0, "seconds": 786.4, "max_memory_gb": 18.17, "ce": 0.2613979318393831, "rank": 0.0} +{"step": 423, "steps": 500, "lr": 2e-05, "grad_norm": 4.61680269241333, "loss_scale": 4.0, "retries": 0, "seconds": 788.2, "max_memory_gb": 18.17, "ce": 0.36213063687318936, "rank": 0.0} +{"step": 424, "steps": 500, "lr": 2e-05, "grad_norm": 1.8204104900360107, "loss_scale": 4.0, "retries": 0, "seconds": 790.1, "max_memory_gb": 18.17, "ce": 0.0645215120021021, "rank": 0.0} +{"step": 425, "steps": 500, "lr": 2e-05, "grad_norm": 0.6469238996505737, "loss_scale": 4.0, "retries": 0, "seconds": 791.9, "max_memory_gb": 18.17, "ce": 0.0643669232740649, "rank": 0.0} +{"step": 426, "steps": 500, "lr": 2e-05, "grad_norm": 1.3933717012405396, "loss_scale": 4.0, "retries": 0, "seconds": 793.8, "max_memory_gb": 18.17, "ce": 0.20742961327778175, "rank": 0.0} +{"step": 427, "steps": 500, "lr": 2e-05, "grad_norm": 0.7734333276748657, "loss_scale": 4.0, "retries": 0, "seconds": 795.7, "max_memory_gb": 18.17, "ce": 0.0796898385087843, "rank": 0.0} +{"step": 428, "steps": 500, "lr": 2e-05, "grad_norm": 1.448158621788025, "loss_scale": 4.0, "retries": 0, "seconds": 797.6, "max_memory_gb": 18.17, "ce": 0.06606947799446061, "rank": 0.0} +{"step": 429, "steps": 500, "lr": 2e-05, "grad_norm": 9.815559387207031, "loss_scale": 4.0, "retries": 0, "seconds": 799.5, "max_memory_gb": 18.17, "ce": 0.07276971538522048, "rank": 0.0} +{"step": 430, "steps": 500, "lr": 2e-05, "grad_norm": 0.6026278138160706, "loss_scale": 4.0, "retries": 0, "seconds": 801.3, "max_memory_gb": 18.17, "ce": 0.1176922598388046, "rank": 0.0} +{"step": 431, "steps": 500, "lr": 2e-05, "grad_norm": 6.9050798416137695, "loss_scale": 4.0, "retries": 0, "seconds": 803.2, "max_memory_gb": 18.17, "ce": 0.20865773096011253, "rank": 0.0} +{"step": 432, "steps": 500, "lr": 2e-05, "grad_norm": 0.714657723903656, "loss_scale": 4.0, "retries": 0, "seconds": 805.0, "max_memory_gb": 18.17, "ce": 0.07322577136801556, "rank": 0.0} +{"step": 433, "steps": 500, "lr": 2e-05, "grad_norm": 44.525718688964844, "loss_scale": 4.0, "retries": 0, "seconds": 806.8, "max_memory_gb": 18.17, "ce": 0.19763954333029687, "rank": 0.0} +{"step": 434, "steps": 500, "lr": 2e-05, "grad_norm": 3.1371421813964844, "loss_scale": 4.0, "retries": 0, "seconds": 808.7, "max_memory_gb": 18.17, "ce": 0.030938036070438102, "rank": 0.0} +{"step": 435, "steps": 500, "lr": 2e-05, "grad_norm": 2.0382776260375977, "loss_scale": 4.0, "retries": 0, "seconds": 810.5, "max_memory_gb": 18.17, "ce": 0.17324311550328275, "rank": 0.0} +{"step": 436, "steps": 500, "lr": 2e-05, "grad_norm": 3.4953389167785645, "loss_scale": 4.0, "retries": 0, "seconds": 812.4, "max_memory_gb": 18.17, "ce": 0.17141001633717678, "rank": 0.0} +{"step": 437, "steps": 500, "lr": 2e-05, "grad_norm": 15.333885192871094, "loss_scale": 4.0, "retries": 0, "seconds": 814.2, "max_memory_gb": 18.17, "ce": 0.36350296105229063, "rank": 0.1082497090101242} +{"step": 438, "steps": 500, "lr": 2e-05, "grad_norm": 5.420368671417236, "loss_scale": 4.0, "retries": 0, "seconds": 816.1, "max_memory_gb": 18.17, "ce": 0.07968567926218384, "rank": 0.0} +{"step": 439, "steps": 500, "lr": 2e-05, "grad_norm": 2.3901100158691406, "loss_scale": 4.0, "retries": 0, "seconds": 817.9, "max_memory_gb": 18.17, "ce": 0.16222849767655134, "rank": 0.0} +{"step": 440, "steps": 500, "lr": 2e-05, "grad_norm": 0.5715795755386353, "loss_scale": 4.0, "retries": 0, "seconds": 819.8, "max_memory_gb": 18.17, "ce": 0.053189247199043166, "rank": 0.0} +{"step": 441, "steps": 500, "lr": 2e-05, "grad_norm": 18.09577751159668, "loss_scale": 4.0, "retries": 0, "seconds": 821.6, "max_memory_gb": 18.17, "ce": 0.3384271640679799, "rank": 0.0} +{"step": 442, "steps": 500, "lr": 2e-05, "grad_norm": 24.043132781982422, "loss_scale": 4.0, "retries": 0, "seconds": 823.4, "max_memory_gb": 18.17, "ce": 0.14441030964371748, "rank": 0.0} +{"step": 443, "steps": 500, "lr": 2e-05, "grad_norm": 0.8122451305389404, "loss_scale": 4.0, "retries": 0, "seconds": 825.2, "max_memory_gb": 18.17, "ce": 0.0798473526010639, "rank": 0.0} +{"step": 444, "steps": 500, "lr": 2e-05, "grad_norm": 1.3609867095947266, "loss_scale": 4.0, "retries": 0, "seconds": 827.1, "max_memory_gb": 18.17, "ce": 0.13022125561838038, "rank": 0.0} +{"step": 445, "steps": 500, "lr": 2e-05, "grad_norm": 2.358550786972046, "loss_scale": 4.0, "retries": 0, "seconds": 828.9, "max_memory_gb": 18.17, "ce": 0.37846090333914617, "rank": 0.06438900530338287} +{"step": 446, "steps": 500, "lr": 2e-05, "grad_norm": 24.17572021484375, "loss_scale": 4.0, "retries": 0, "seconds": 830.7, "max_memory_gb": 18.17, "ce": 0.24604925900348462, "rank": 0.2908337116241455} +{"step": 447, "steps": 500, "lr": 2e-05, "grad_norm": 1.1628661155700684, "loss_scale": 4.0, "retries": 0, "seconds": 832.6, "max_memory_gb": 18.17, "ce": 0.1935384688731574, "rank": 0.0} +{"step": 448, "steps": 500, "lr": 2e-05, "grad_norm": 6.094340801239014, "loss_scale": 4.0, "retries": 0, "seconds": 834.4, "max_memory_gb": 18.17, "ce": 0.23826602916233242, "rank": 0.0} +{"step": 449, "steps": 500, "lr": 2e-05, "grad_norm": 3.821119785308838, "loss_scale": 4.0, "retries": 0, "seconds": 836.2, "max_memory_gb": 18.17, "ce": 0.04455979968770407, "rank": 0.0} +{"step": 450, "steps": 500, "lr": 2e-05, "grad_norm": 7.767148494720459, "loss_scale": 4.0, "retries": 0, "seconds": 838.0, "max_memory_gb": 18.17, "ce": 0.11767913975199917, "rank": 0.0} +{"step": 451, "steps": 500, "lr": 2e-05, "grad_norm": 3.430349111557007, "loss_scale": 4.0, "retries": 0, "seconds": 839.9, "max_memory_gb": 18.17, "ce": 0.0635889916538872, "rank": 0.0} +{"step": 452, "steps": 500, "lr": 2e-05, "grad_norm": 6.7299485206604, "loss_scale": 4.0, "retries": 0, "seconds": 841.7, "max_memory_gb": 18.17, "ce": 0.17593276931438595, "rank": 0.0} +{"step": 453, "steps": 500, "lr": 2e-05, "grad_norm": 2.9080188274383545, "loss_scale": 4.0, "retries": 0, "seconds": 843.6, "max_memory_gb": 18.17, "ce": 0.05633642202883493, "rank": 0.0} +{"step": 454, "steps": 500, "lr": 2e-05, "grad_norm": 4.2830119132995605, "loss_scale": 4.0, "retries": 0, "seconds": 845.4, "max_memory_gb": 18.17, "ce": 0.15742443365888903, "rank": 0.0} +{"step": 455, "steps": 500, "lr": 2e-05, "grad_norm": 5.893167495727539, "loss_scale": 4.0, "retries": 0, "seconds": 847.3, "max_memory_gb": 18.17, "ce": 0.155458542983979, "rank": 0.0} +{"step": 456, "steps": 500, "lr": 2e-05, "grad_norm": 0.8825544714927673, "loss_scale": 4.0, "retries": 0, "seconds": 849.1, "max_memory_gb": 18.17, "ce": 0.04118299332913011, "rank": 0.0} +{"step": 457, "steps": 500, "lr": 2e-05, "grad_norm": 1.874404788017273, "loss_scale": 4.0, "retries": 0, "seconds": 850.9, "max_memory_gb": 18.17, "ce": 0.06941323266437394, "rank": 0.09367558360099792} +{"step": 458, "steps": 500, "lr": 2e-05, "grad_norm": 29.870105743408203, "loss_scale": 4.0, "retries": 0, "seconds": 852.8, "max_memory_gb": 18.17, "ce": 0.3141038757748902, "rank": 0.0} +{"step": 459, "steps": 500, "lr": 2e-05, "grad_norm": 2.8671016693115234, "loss_scale": 4.0, "retries": 0, "seconds": 854.6, "max_memory_gb": 18.17, "ce": 0.1130270988214761, "rank": 0.0} +{"step": 460, "steps": 500, "lr": 2e-05, "grad_norm": 9.060452461242676, "loss_scale": 4.0, "retries": 0, "seconds": 856.5, "max_memory_gb": 18.17, "ce": 0.059298071209923364, "rank": 0.0} +{"step": 461, "steps": 500, "lr": 2e-05, "grad_norm": 17.109169006347656, "loss_scale": 4.0, "retries": 0, "seconds": 858.3, "max_memory_gb": 18.17, "ce": 0.2960368641652167, "rank": 0.0} +{"step": 462, "steps": 500, "lr": 2e-05, "grad_norm": 0.25951704382896423, "loss_scale": 4.0, "retries": 0, "seconds": 860.1, "max_memory_gb": 18.17, "ce": 0.043805480334413005, "rank": 0.0} +{"step": 463, "steps": 500, "lr": 2e-05, "grad_norm": 0.45407623052597046, "loss_scale": 4.0, "retries": 0, "seconds": 861.9, "max_memory_gb": 18.17, "ce": 0.03981038038909901, "rank": 0.0} +{"step": 464, "steps": 500, "lr": 2e-05, "grad_norm": 0.3236713111400604, "loss_scale": 4.0, "retries": 0, "seconds": 863.8, "max_memory_gb": 18.17, "ce": 0.021787445053632837, "rank": 0.0} +{"step": 465, "steps": 500, "lr": 2e-05, "grad_norm": 2.3994908332824707, "loss_scale": 4.0, "retries": 0, "seconds": 865.7, "max_memory_gb": 18.17, "ce": 0.2333701354727964, "rank": 0.0} +{"step": 466, "steps": 500, "lr": 2e-05, "grad_norm": 2.9595963954925537, "loss_scale": 4.0, "retries": 0, "seconds": 867.5, "max_memory_gb": 18.17, "ce": 0.11366311539677554, "rank": 0.0} +{"step": 467, "steps": 500, "lr": 2e-05, "grad_norm": 3.1457831859588623, "loss_scale": 4.0, "retries": 0, "seconds": 869.4, "max_memory_gb": 18.17, "ce": 0.08351727529952768, "rank": 0.0} +{"step": 468, "steps": 500, "lr": 2e-05, "grad_norm": 0.809034526348114, "loss_scale": 4.0, "retries": 0, "seconds": 871.2, "max_memory_gb": 18.17, "ce": 0.036649007866799366, "rank": 0.0} +{"step": 469, "steps": 500, "lr": 2e-05, "grad_norm": 8.82933521270752, "loss_scale": 4.0, "retries": 0, "seconds": 873.0, "max_memory_gb": 18.17, "ce": 0.21273193899833132, "rank": 0.0} +{"step": 470, "steps": 500, "lr": 2e-05, "grad_norm": 3.4787943363189697, "loss_scale": 4.0, "retries": 0, "seconds": 874.9, "max_memory_gb": 18.17, "ce": 0.112237125478714, "rank": 0.0} +{"step": 471, "steps": 500, "lr": 2e-05, "grad_norm": 2.1723530292510986, "loss_scale": 4.0, "retries": 0, "seconds": 876.7, "max_memory_gb": 18.17, "ce": 0.19516852607193869, "rank": 0.0} +{"step": 472, "steps": 500, "lr": 2e-05, "grad_norm": 3.261326313018799, "loss_scale": 4.0, "retries": 0, "seconds": 878.5, "max_memory_gb": 18.17, "ce": 0.06243787684070412, "rank": 0.0} +{"step": 473, "steps": 500, "lr": 2e-05, "grad_norm": 4.38939094543457, "loss_scale": 4.0, "retries": 0, "seconds": 880.4, "max_memory_gb": 18.17, "ce": 0.06378060173301492, "rank": 0.15118584036827087} +{"step": 474, "steps": 500, "lr": 2e-05, "grad_norm": 2.013585090637207, "loss_scale": 4.0, "retries": 0, "seconds": 882.2, "max_memory_gb": 18.17, "ce": 0.12228500098353834, "rank": 0.0} +{"step": 475, "steps": 500, "lr": 2e-05, "grad_norm": 1.0680681467056274, "loss_scale": 4.0, "retries": 0, "seconds": 884.0, "max_memory_gb": 18.17, "ce": 0.14551504695918993, "rank": 0.0} +{"step": 476, "steps": 500, "lr": 2e-05, "grad_norm": 1.3557065725326538, "loss_scale": 4.0, "retries": 0, "seconds": 885.9, "max_memory_gb": 18.17, "ce": 0.04271441233868245, "rank": 0.0} +{"step": 477, "steps": 500, "lr": 2e-05, "grad_norm": 26.688791275024414, "loss_scale": 4.0, "retries": 0, "seconds": 887.7, "max_memory_gb": 18.17, "ce": 0.09475365627440624, "rank": 0.0} +{"step": 478, "steps": 500, "lr": 2e-05, "grad_norm": 1.3810722827911377, "loss_scale": 4.0, "retries": 0, "seconds": 889.5, "max_memory_gb": 18.17, "ce": 0.10070970162632875, "rank": 0.0} +{"step": 479, "steps": 500, "lr": 2e-05, "grad_norm": 7.9332404136657715, "loss_scale": 4.0, "retries": 0, "seconds": 891.4, "max_memory_gb": 18.17, "ce": 0.3388389586471021, "rank": 0.040682196617126465} +{"step": 480, "steps": 500, "lr": 2e-05, "grad_norm": 0.7038509249687195, "loss_scale": 4.0, "retries": 0, "seconds": 893.2, "max_memory_gb": 18.17, "ce": 0.17002621534629725, "rank": 0.0} +{"step": 481, "steps": 500, "lr": 2e-05, "grad_norm": 11.8507080078125, "loss_scale": 4.0, "retries": 0, "seconds": 895.1, "max_memory_gb": 18.17, "ce": 0.07199242347269319, "rank": 0.0} +{"step": 482, "steps": 500, "lr": 2e-05, "grad_norm": 9.653359413146973, "loss_scale": 4.0, "retries": 0, "seconds": 896.9, "max_memory_gb": 18.17, "ce": 0.19316648246604018, "rank": 0.0} +{"step": 483, "steps": 500, "lr": 2e-05, "grad_norm": 23.656892776489258, "loss_scale": 4.0, "retries": 0, "seconds": 898.7, "max_memory_gb": 18.17, "ce": 0.5796149270609021, "rank": 0.0} +{"step": 484, "steps": 500, "lr": 2e-05, "grad_norm": 2.103292226791382, "loss_scale": 4.0, "retries": 0, "seconds": 900.6, "max_memory_gb": 18.17, "ce": 0.4410845737438649, "rank": 0.0} +{"step": 485, "steps": 500, "lr": 2e-05, "grad_norm": 1.004713773727417, "loss_scale": 4.0, "retries": 0, "seconds": 902.4, "max_memory_gb": 18.17, "ce": 0.08189704251708463, "rank": 0.0} +{"step": 486, "steps": 500, "lr": 2e-05, "grad_norm": 5.809371471405029, "loss_scale": 4.0, "retries": 0, "seconds": 904.2, "max_memory_gb": 18.17, "ce": 0.2635218717623502, "rank": 0.0} +{"step": 487, "steps": 500, "lr": 2e-05, "grad_norm": 2.2857248783111572, "loss_scale": 4.0, "retries": 0, "seconds": 906.0, "max_memory_gb": 18.17, "ce": 0.08637815299880458, "rank": 0.0} +{"step": 488, "steps": 500, "lr": 2e-05, "grad_norm": 2.1292011737823486, "loss_scale": 4.0, "retries": 0, "seconds": 907.9, "max_memory_gb": 18.17, "ce": 0.1263488312251866, "rank": 0.0} +{"step": 489, "steps": 500, "lr": 2e-05, "grad_norm": 4.4388556480407715, "loss_scale": 4.0, "retries": 0, "seconds": 909.7, "max_memory_gb": 18.17, "ce": 0.38963942455666256, "rank": 0.0} +{"step": 490, "steps": 500, "lr": 2e-05, "grad_norm": 1.1034055948257446, "loss_scale": 4.0, "retries": 0, "seconds": 911.6, "max_memory_gb": 18.17, "ce": 0.056816027899913024, "rank": 0.0} +{"step": 491, "steps": 500, "lr": 2e-05, "grad_norm": 0.8115298748016357, "loss_scale": 4.0, "retries": 0, "seconds": 913.4, "max_memory_gb": 18.17, "ce": 0.12054986259499856, "rank": 0.0} +{"step": 492, "steps": 500, "lr": 2e-05, "grad_norm": 0.5456184148788452, "loss_scale": 4.0, "retries": 0, "seconds": 915.2, "max_memory_gb": 18.17, "ce": 0.02355215128045529, "rank": 0.0} +{"step": 493, "steps": 500, "lr": 2e-05, "grad_norm": 3.115030288696289, "loss_scale": 4.0, "retries": 0, "seconds": 917.1, "max_memory_gb": 18.17, "ce": 0.11984676045540255, "rank": 0.0} +{"step": 494, "steps": 500, "lr": 2e-05, "grad_norm": 1.9629757404327393, "loss_scale": 4.0, "retries": 0, "seconds": 918.9, "max_memory_gb": 18.17, "ce": 0.1265249231728376, "rank": 0.0} +{"step": 495, "steps": 500, "lr": 2e-05, "grad_norm": 0.883539617061615, "loss_scale": 4.0, "retries": 0, "seconds": 920.8, "max_memory_gb": 18.17, "ce": 0.041951868537580594, "rank": 0.0} +{"step": 496, "steps": 500, "lr": 2e-05, "grad_norm": 0.7359150052070618, "loss_scale": 4.0, "retries": 0, "seconds": 922.6, "max_memory_gb": 18.17, "ce": 0.031746729830047116, "rank": 0.0} +{"step": 497, "steps": 500, "lr": 2e-05, "grad_norm": 6.998233795166016, "loss_scale": 4.0, "retries": 0, "seconds": 924.4, "max_memory_gb": 18.17, "ce": 0.2637621805915842, "rank": 0.0} +{"step": 498, "steps": 500, "lr": 2e-05, "grad_norm": 2.0739939212799072, "loss_scale": 4.0, "retries": 0, "seconds": 926.2, "max_memory_gb": 18.17, "ce": 0.13397932724183192, "rank": 0.0} +{"step": 499, "steps": 500, "lr": 2e-05, "grad_norm": 4.382185935974121, "loss_scale": 4.0, "retries": 0, "seconds": 928.1, "max_memory_gb": 18.17, "ce": 0.18566119041861384, "rank": 0.0} +{"step": 500, "steps": 500, "lr": 2e-05, "grad_norm": 2.1597933769226074, "loss_scale": 4.0, "retries": 0, "seconds": 929.9, "max_memory_gb": 18.17, "ce": 0.26583348394524364, "rank": 0.0} diff --git a/runs/single_llava/removal/L06_s13/adapter/README.md b/runs/single_llava/removal/L06_s13/adapter/README.md new file mode 100644 index 0000000000000000000000000000000000000000..ae587920ca55972d94e38f783837f71be14f1dda --- /dev/null +++ b/runs/single_llava/removal/L06_s13/adapter/README.md @@ -0,0 +1,207 @@ +--- +base_model: /data/OPSD/assets/models/llava-1.5-7b-hf +library_name: peft +pipeline_tag: text-generation +tags: +- base_model:adapter:/data/OPSD/assets/models/llava-1.5-7b-hf +- lora +- transformers +--- + +# Model Card for Model ID + + + + + +## Model Details + +### Model Description + + + + + +- **Developed by:** [More Information Needed] +- **Funded by [optional]:** [More Information Needed] +- **Shared by [optional]:** [More Information Needed] +- **Model type:** [More Information Needed] +- **Language(s) (NLP):** [More Information Needed] +- **License:** [More Information Needed] +- **Finetuned from model [optional]:** [More Information Needed] + +### Model Sources [optional] + + + +- **Repository:** [More Information Needed] +- **Paper [optional]:** [More Information Needed] +- **Demo [optional]:** [More Information Needed] + +## Uses + + + +### Direct Use + + + +[More Information Needed] + +### Downstream Use [optional] + + + +[More Information Needed] + +### Out-of-Scope Use + + + +[More Information Needed] + +## Bias, Risks, and Limitations + + + +[More Information Needed] + +### Recommendations + + + +Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. More information needed for further recommendations. + +## How to Get Started with the Model + +Use the code below to get started with the model. + +[More Information Needed] + +## Training Details + +### Training Data + + + +[More Information Needed] + +### Training Procedure + + + +#### Preprocessing [optional] + +[More Information Needed] + + +#### Training Hyperparameters + +- **Training regime:** [More Information Needed] + +#### Speeds, Sizes, Times [optional] + + + +[More Information Needed] + +## Evaluation + + + +### Testing Data, Factors & Metrics + +#### Testing Data + + + +[More Information Needed] + +#### Factors + + + +[More Information Needed] + +#### Metrics + + + +[More Information Needed] + +### Results + +[More Information Needed] + +#### Summary + + + +## Model Examination [optional] + + + +[More Information Needed] + +## Environmental Impact + + + +Carbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700). + +- **Hardware Type:** [More Information Needed] +- **Hours used:** [More Information Needed] +- **Cloud Provider:** [More Information Needed] +- **Compute Region:** [More Information Needed] +- **Carbon Emitted:** [More Information Needed] + +## Technical Specifications [optional] + +### Model Architecture and Objective + +[More Information Needed] + +### Compute Infrastructure + +[More Information Needed] + +#### Hardware + +[More Information Needed] + +#### Software + +[More Information Needed] + +## Citation [optional] + + + +**BibTeX:** + +[More Information Needed] + +**APA:** + +[More Information Needed] + +## Glossary [optional] + + + +[More Information Needed] + +## More Information [optional] + +[More Information Needed] + +## Model Card Authors [optional] + +[More Information Needed] + +## Model Card Contact + +[More Information Needed] +### Framework versions + +- PEFT 0.20.0 \ No newline at end of file diff --git a/runs/single_llava/removal/L06_s13/adapter/adapter_config.json b/runs/single_llava/removal/L06_s13/adapter/adapter_config.json new file mode 100644 index 0000000000000000000000000000000000000000..a8b91dd7d7f630ac5e3c93c2481986e960ef6802 --- /dev/null +++ b/runs/single_llava/removal/L06_s13/adapter/adapter_config.json @@ -0,0 +1,50 @@ +{ + "alora_invocation_tokens": null, + "alpha_pattern": {}, + "arrow_config": null, + "auto_mapping": null, + "base_model_name_or_path": "/data/OPSD/assets/models/llava-1.5-7b-hf", + "bias": "none", + "corda_config": null, + "ensure_weight_tying": false, + "eva_config": null, + "exclude_modules": null, + "fan_in_fan_out": false, + "inference_mode": true, + "init_lora_weights": true, + "layer_replication": null, + "layers_pattern": null, + "layers_to_transform": null, + "loftq_config": {}, + "lora_alpha": 32, + "lora_bias": false, + "lora_dropout": 0.05, + "lora_ga_config": null, + "megatron_config": null, + "megatron_core": "megatron.core", + "modules_to_save": null, + "monteclora_config": null, + "peft_type": "LORA", + "peft_version": "0.20.0", + "qalora_group_size": 16, + "r": 16, + "rank_pattern": {}, + "revision": null, + "target_modules": [ + "model.language_model.layers.6.mlp.gate_proj", + "model.language_model.layers.6.mlp.up_proj", + "model.language_model.layers.6.self_attn.v_proj", + "model.language_model.layers.6.mlp.down_proj", + "model.language_model.layers.6.self_attn.q_proj", + "model.language_model.layers.6.self_attn.o_proj", + "model.language_model.layers.6.self_attn.k_proj" + ], + "target_parameters": null, + "task_type": "CAUSAL_LM", + "trainable_token_indices": null, + "use_bdlora": null, + "use_dora": false, + "use_qalora": false, + "use_rslora": false, + "velora_config": null +} \ No newline at end of file diff --git a/runs/single_llava/removal/L06_s13/config.json b/runs/single_llava/removal/L06_s13/config.json new file mode 100644 index 0000000000000000000000000000000000000000..8e11abd709fd02dfc3e863fe65fa22f9a386cd26 --- /dev/null +++ b/runs/single_llava/removal/L06_s13/config.json @@ -0,0 +1,24 @@ +{ + "behavior": "removal", + "seed": 13, + "model": "/data/OPSD/assets/models/llava-1.5-7b-hf", + "backbone": "llava", + "train": "/data/OPSD/assets/v3/removal.jsonl", + "items": 1000, + "steps": 500, + "epochs": 2, + "steps_per_epoch": 250, + "groups_per_step": 4, + "lr": 2e-05, + "scheduler": "constant", + "warmup_steps": 0, + "final_fraction": 1.0, + "weight_decay": 0.0, + "clip": 1.0, + "initial_loss_scale": 4.0, + "adapter": { + "layer": 6, + "rank": 16, + "trainable_parameters": 1249280 + } +} \ No newline at end of file diff --git a/runs/single_llava/removal/L06_s13/steps.jsonl b/runs/single_llava/removal/L06_s13/steps.jsonl new file mode 100644 index 0000000000000000000000000000000000000000..66050cd30032bdf2bac989e19de23316f1063511 --- /dev/null +++ b/runs/single_llava/removal/L06_s13/steps.jsonl @@ -0,0 +1,500 @@ +{"step": 1, "steps": 500, "lr": 2e-05, "grad_norm": 0.20203010737895966, "loss_scale": 4.0, "retries": 0, "seconds": 3.0, "max_memory_gb": 17.58, "ce": 1.6287612388841808, "rank": 0.13243506848812103} +{"step": 2, "steps": 500, "lr": 2e-05, "grad_norm": 0.30311885476112366, "loss_scale": 4.0, "retries": 0, "seconds": 4.8, "max_memory_gb": 17.6, "ce": 2.6747625801945105, "rank": 0.032688260078430176} +{"step": 3, "steps": 500, "lr": 2e-05, "grad_norm": 0.6770873069763184, "loss_scale": 4.0, "retries": 0, "seconds": 6.7, "max_memory_gb": 17.6, "ce": 5.276225603185594, "rank": 0.14902842044830322} +{"step": 4, "steps": 500, "lr": 2e-05, "grad_norm": 0.4268968105316162, "loss_scale": 4.0, "retries": 0, "seconds": 8.5, "max_memory_gb": 17.6, "ce": 3.7991889053955674, "rank": 0.12350702285766602} +{"step": 5, "steps": 500, "lr": 2e-05, "grad_norm": 0.06990338116884232, "loss_scale": 4.0, "retries": 0, "seconds": 10.3, "max_memory_gb": 17.6, "ce": 0.1398294421378523, "rank": 0.0} +{"step": 6, "steps": 500, "lr": 2e-05, "grad_norm": 0.19180069863796234, "loss_scale": 4.0, "retries": 0, "seconds": 12.2, "max_memory_gb": 17.6, "ce": 1.5120405466295779, "rank": 0.0} +{"step": 7, "steps": 500, "lr": 2e-05, "grad_norm": 0.4887597858905792, "loss_scale": 4.0, "retries": 0, "seconds": 14.1, "max_memory_gb": 17.6, "ce": 3.4529846906661987, "rank": 0.020332932472229004} +{"step": 8, "steps": 500, "lr": 2e-05, "grad_norm": 0.5591391921043396, "loss_scale": 4.0, "retries": 0, "seconds": 16.0, "max_memory_gb": 17.6, "ce": 3.696428645402193, "rank": 0.20374524593353271} +{"step": 9, "steps": 500, "lr": 2e-05, "grad_norm": 0.45174524188041687, "loss_scale": 4.0, "retries": 0, "seconds": 17.9, "max_memory_gb": 17.6, "ce": 2.8256307914853096, "rank": 0.03829920291900635} +{"step": 10, "steps": 500, "lr": 2e-05, "grad_norm": 0.23649846017360687, "loss_scale": 4.0, "retries": 0, "seconds": 19.8, "max_memory_gb": 17.6, "ce": 1.6156268203631043, "rank": 0.0} +{"step": 11, "steps": 500, "lr": 2e-05, "grad_norm": 0.5697635412216187, "loss_scale": 4.0, "retries": 0, "seconds": 21.7, "max_memory_gb": 17.6, "ce": 2.610296191647649, "rank": 0.20689041912555695} +{"step": 12, "steps": 500, "lr": 2e-05, "grad_norm": 0.31066927313804626, "loss_scale": 4.0, "retries": 0, "seconds": 23.6, "max_memory_gb": 17.6, "ce": 1.502004976850003, "rank": 0.04898989200592041} +{"step": 13, "steps": 500, "lr": 2e-05, "grad_norm": 0.48047006130218506, "loss_scale": 4.0, "retries": 0, "seconds": 25.5, "max_memory_gb": 17.6, "ce": 2.6541133186547086, "rank": 0.0} +{"step": 14, "steps": 500, "lr": 2e-05, "grad_norm": 0.7694302201271057, "loss_scale": 4.0, "retries": 0, "seconds": 27.3, "max_memory_gb": 17.6, "ce": 3.8427018262445927, "rank": 0.1439368724822998} +{"step": 15, "steps": 500, "lr": 2e-05, "grad_norm": 0.5608099102973938, "loss_scale": 4.0, "retries": 0, "seconds": 29.2, "max_memory_gb": 17.6, "ce": 2.575459626270458, "rank": 0.1696547269821167} +{"step": 16, "steps": 500, "lr": 2e-05, "grad_norm": 0.36811432242393494, "loss_scale": 4.0, "retries": 0, "seconds": 31.1, "max_memory_gb": 17.6, "ce": 1.4035872151725926, "rank": 0.0} +{"step": 17, "steps": 500, "lr": 2e-05, "grad_norm": 0.40189605951309204, "loss_scale": 4.0, "retries": 0, "seconds": 33.0, "max_memory_gb": 17.6, "ce": 1.5759962253505364, "rank": 0.032262444496154785} +{"step": 18, "steps": 500, "lr": 2e-05, "grad_norm": 0.7840573787689209, "loss_scale": 4.0, "retries": 0, "seconds": 34.9, "max_memory_gb": 17.61, "ce": 3.547807613387704, "rank": 0.19495952129364014} +{"step": 19, "steps": 500, "lr": 2e-05, "grad_norm": 1.2873191833496094, "loss_scale": 4.0, "retries": 0, "seconds": 36.8, "max_memory_gb": 17.61, "ce": 5.203304274939001, "rank": 0.30709993839263916} +{"step": 20, "steps": 500, "lr": 2e-05, "grad_norm": 0.026754437014460564, "loss_scale": 4.0, "retries": 0, "seconds": 38.7, "max_memory_gb": 17.61, "ce": 0.0657071553869173, "rank": 0.1093277782201767} +{"step": 21, "steps": 500, "lr": 2e-05, "grad_norm": 0.8024663329124451, "loss_scale": 4.0, "retries": 0, "seconds": 40.6, "max_memory_gb": 17.61, "ce": 2.691887469496578, "rank": 0.1420578956604004} +{"step": 22, "steps": 500, "lr": 2e-05, "grad_norm": 0.7994015216827393, "loss_scale": 4.0, "retries": 0, "seconds": 42.5, "max_memory_gb": 17.61, "ce": 2.8916168147698045, "rank": 0.18069469928741455} +{"step": 23, "steps": 500, "lr": 2e-05, "grad_norm": 0.8256050944328308, "loss_scale": 4.0, "retries": 0, "seconds": 44.5, "max_memory_gb": 17.61, "ce": 2.689604517072439, "rank": 0.11605536937713623} +{"step": 24, "steps": 500, "lr": 2e-05, "grad_norm": 1.444665551185608, "loss_scale": 4.0, "retries": 0, "seconds": 46.4, "max_memory_gb": 17.61, "ce": 4.881436741910875, "rank": 0.25840532779693604} +{"step": 25, "steps": 500, "lr": 2e-05, "grad_norm": 1.5452593564987183, "loss_scale": 4.0, "retries": 0, "seconds": 48.5, "max_memory_gb": 17.61, "ce": 3.9972427948378026, "rank": 0.02014482021331787} +{"step": 26, "steps": 500, "lr": 2e-05, "grad_norm": 2.023425579071045, "loss_scale": 4.0, "retries": 0, "seconds": 50.6, "max_memory_gb": 17.61, "ce": 4.9797075018286705, "rank": 0.04693305492401123} +{"step": 27, "steps": 500, "lr": 2e-05, "grad_norm": 1.6534640789031982, "loss_scale": 4.0, "retries": 0, "seconds": 52.5, "max_memory_gb": 17.61, "ce": 4.048987148329616, "rank": 0.4503868818283081} +{"step": 28, "steps": 500, "lr": 2e-05, "grad_norm": 0.9953721761703491, "loss_scale": 4.0, "retries": 0, "seconds": 54.5, "max_memory_gb": 17.61, "ce": 2.666598392650485, "rank": 0.0} +{"step": 29, "steps": 500, "lr": 2e-05, "grad_norm": 0.9621303081512451, "loss_scale": 4.0, "retries": 0, "seconds": 56.6, "max_memory_gb": 17.61, "ce": 2.5770995342172682, "rank": 0.0} +{"step": 30, "steps": 500, "lr": 2e-05, "grad_norm": 1.4263747930526733, "loss_scale": 4.0, "retries": 0, "seconds": 58.6, "max_memory_gb": 17.61, "ce": 2.5957279548747465, "rank": 0.0} +{"step": 31, "steps": 500, "lr": 2e-05, "grad_norm": 1.4544503688812256, "loss_scale": 4.0, "retries": 0, "seconds": 60.5, "max_memory_gb": 17.61, "ce": 2.6526609929278493, "rank": 0.1314300298690796} +{"step": 32, "steps": 500, "lr": 2e-05, "grad_norm": 2.649026870727539, "loss_scale": 4.0, "retries": 0, "seconds": 62.5, "max_memory_gb": 17.61, "ce": 4.988147655501962, "rank": 0.20023345947265625} +{"step": 33, "steps": 500, "lr": 2e-05, "grad_norm": 1.2981410026550293, "loss_scale": 4.0, "retries": 0, "seconds": 64.4, "max_memory_gb": 17.61, "ce": 2.374734818469733, "rank": 0.00839550793170929} +{"step": 34, "steps": 500, "lr": 2e-05, "grad_norm": 1.569150447845459, "loss_scale": 4.0, "retries": 0, "seconds": 66.3, "max_memory_gb": 17.61, "ce": 2.545370294013992, "rank": 0.0} +{"step": 35, "steps": 500, "lr": 2e-05, "grad_norm": 2.2868990898132324, "loss_scale": 4.0, "retries": 0, "seconds": 68.1, "max_memory_gb": 17.61, "ce": 2.4546556561253965, "rank": 0.10327494144439697} +{"step": 36, "steps": 500, "lr": 2e-05, "grad_norm": 1.415490746498108, "loss_scale": 4.0, "retries": 0, "seconds": 70.0, "max_memory_gb": 17.61, "ce": 2.409682455472648, "rank": 0.0} +{"step": 37, "steps": 500, "lr": 2e-05, "grad_norm": 1.7234646081924438, "loss_scale": 4.0, "retries": 0, "seconds": 71.9, "max_memory_gb": 17.61, "ce": 2.4046906400471926, "rank": 0.0} +{"step": 38, "steps": 500, "lr": 2e-05, "grad_norm": 3.4503543376922607, "loss_scale": 4.0, "retries": 0, "seconds": 73.8, "max_memory_gb": 17.61, "ce": 4.41400262247771, "rank": 0.11091578006744385} +{"step": 39, "steps": 500, "lr": 2e-05, "grad_norm": 1.553192377090454, "loss_scale": 4.0, "retries": 0, "seconds": 75.7, "max_memory_gb": 17.61, "ce": 1.2623596291523427, "rank": 0.0945308580994606} +{"step": 40, "steps": 500, "lr": 2e-05, "grad_norm": 2.4313230514526367, "loss_scale": 4.0, "retries": 0, "seconds": 77.6, "max_memory_gb": 17.61, "ce": 2.387072009500116, "rank": 0.0} +{"step": 41, "steps": 500, "lr": 2e-05, "grad_norm": 1.1900062561035156, "loss_scale": 4.0, "retries": 0, "seconds": 79.5, "max_memory_gb": 17.61, "ce": 1.3909084452316165, "rank": 0.0} +{"step": 42, "steps": 500, "lr": 2e-05, "grad_norm": 2.6031839847564697, "loss_scale": 4.0, "retries": 0, "seconds": 81.4, "max_memory_gb": 17.61, "ce": 2.068495230283588, "rank": 0.0} +{"step": 43, "steps": 500, "lr": 2e-05, "grad_norm": 4.145608425140381, "loss_scale": 4.0, "retries": 0, "seconds": 83.2, "max_memory_gb": 17.61, "ce": 2.0298888408578932, "rank": 0.0} +{"step": 44, "steps": 500, "lr": 2e-05, "grad_norm": 2.1167404651641846, "loss_scale": 4.0, "retries": 0, "seconds": 85.1, "max_memory_gb": 17.61, "ce": 1.2381775777321309, "rank": 0.0} +{"step": 45, "steps": 500, "lr": 2e-05, "grad_norm": 2.200315237045288, "loss_scale": 4.0, "retries": 0, "seconds": 87.0, "max_memory_gb": 17.61, "ce": 1.1681629340164363, "rank": 0.0} +{"step": 46, "steps": 500, "lr": 2e-05, "grad_norm": 2.989485025405884, "loss_scale": 4.0, "retries": 0, "seconds": 88.9, "max_memory_gb": 17.61, "ce": 2.0124415140599012, "rank": 0.0} +{"step": 47, "steps": 500, "lr": 2e-05, "grad_norm": 3.833693504333496, "loss_scale": 4.0, "retries": 0, "seconds": 90.7, "max_memory_gb": 17.61, "ce": 1.607157735619694, "rank": 0.0} +{"step": 48, "steps": 500, "lr": 2e-05, "grad_norm": 7.6488471031188965, "loss_scale": 4.0, "retries": 0, "seconds": 92.7, "max_memory_gb": 17.61, "ce": 2.8609338020905852, "rank": 0.0} +{"step": 49, "steps": 500, "lr": 2e-05, "grad_norm": 7.215600967407227, "loss_scale": 4.0, "retries": 0, "seconds": 94.5, "max_memory_gb": 17.61, "ce": 2.591270613949746, "rank": 0.0} +{"step": 50, "steps": 500, "lr": 2e-05, "grad_norm": 5.468171119689941, "loss_scale": 4.0, "retries": 0, "seconds": 96.4, "max_memory_gb": 17.61, "ce": 2.8634820923907682, "rank": 0.0} +{"step": 51, "steps": 500, "lr": 2e-05, "grad_norm": 2.28688645362854, "loss_scale": 4.0, "retries": 0, "seconds": 98.3, "max_memory_gb": 17.61, "ce": 1.0588089723605663, "rank": 0.0} +{"step": 52, "steps": 500, "lr": 2e-05, "grad_norm": 4.135217189788818, "loss_scale": 4.0, "retries": 0, "seconds": 100.2, "max_memory_gb": 17.61, "ce": 1.7029349468648434, "rank": 0.0} +{"step": 53, "steps": 500, "lr": 2e-05, "grad_norm": 7.5441203117370605, "loss_scale": 4.0, "retries": 0, "seconds": 102.1, "max_memory_gb": 17.61, "ce": 2.5871867882087827, "rank": 0.0} +{"step": 54, "steps": 500, "lr": 2e-05, "grad_norm": 8.577696800231934, "loss_scale": 4.0, "retries": 0, "seconds": 104.0, "max_memory_gb": 17.61, "ce": 2.485894400626421, "rank": 0.0} +{"step": 55, "steps": 500, "lr": 2e-05, "grad_norm": 4.762803077697754, "loss_scale": 4.0, "retries": 0, "seconds": 105.8, "max_memory_gb": 17.61, "ce": 1.5293828214053065, "rank": 0.0} +{"step": 56, "steps": 500, "lr": 2e-05, "grad_norm": 10.513915061950684, "loss_scale": 4.0, "retries": 0, "seconds": 107.7, "max_memory_gb": 17.61, "ce": 2.924038083292544, "rank": 0.0} +{"step": 57, "steps": 500, "lr": 2e-05, "grad_norm": 2.640911102294922, "loss_scale": 4.0, "retries": 0, "seconds": 109.6, "max_memory_gb": 17.61, "ce": 0.8357979394495487, "rank": 0.023054644465446472} +{"step": 58, "steps": 500, "lr": 2e-05, "grad_norm": 5.313722133636475, "loss_scale": 4.0, "retries": 0, "seconds": 111.4, "max_memory_gb": 17.61, "ce": 1.2849127445369959, "rank": 0.0} +{"step": 59, "steps": 500, "lr": 2e-05, "grad_norm": 2.1898059844970703, "loss_scale": 4.0, "retries": 0, "seconds": 113.3, "max_memory_gb": 17.61, "ce": 1.0914642810821533, "rank": 0.0} +{"step": 60, "steps": 500, "lr": 2e-05, "grad_norm": 4.633935928344727, "loss_scale": 4.0, "retries": 0, "seconds": 115.2, "max_memory_gb": 17.61, "ce": 1.5588603089563549, "rank": 0.0} +{"step": 61, "steps": 500, "lr": 2e-05, "grad_norm": 7.970447063446045, "loss_scale": 4.0, "retries": 0, "seconds": 117.1, "max_memory_gb": 17.61, "ce": 1.609144888818264, "rank": 0.0} +{"step": 62, "steps": 500, "lr": 2e-05, "grad_norm": 3.1368157863616943, "loss_scale": 4.0, "retries": 0, "seconds": 119.0, "max_memory_gb": 17.61, "ce": 0.7100166636519134, "rank": 0.0} +{"step": 63, "steps": 500, "lr": 2e-05, "grad_norm": 8.379268646240234, "loss_scale": 4.0, "retries": 0, "seconds": 120.9, "max_memory_gb": 17.61, "ce": 1.9371642912738025, "rank": 0.10271143913269043} +{"step": 64, "steps": 500, "lr": 2e-05, "grad_norm": 8.602184295654297, "loss_scale": 4.0, "retries": 0, "seconds": 122.8, "max_memory_gb": 17.61, "ce": 1.7435487140901387, "rank": 0.0} +{"step": 65, "steps": 500, "lr": 2e-05, "grad_norm": 7.378983497619629, "loss_scale": 4.0, "retries": 0, "seconds": 124.7, "max_memory_gb": 17.61, "ce": 1.8725062906742096, "rank": 0.0119820237159729} +{"step": 66, "steps": 500, "lr": 2e-05, "grad_norm": 1.684448003768921, "loss_scale": 4.0, "retries": 0, "seconds": 126.6, "max_memory_gb": 17.61, "ce": 0.6150828320533037, "rank": 0.2887139096856117} +{"step": 67, "steps": 500, "lr": 2e-05, "grad_norm": 5.3135271072387695, "loss_scale": 4.0, "retries": 0, "seconds": 128.6, "max_memory_gb": 17.61, "ce": 1.2655450124293566, "rank": 0.0} +{"step": 68, "steps": 500, "lr": 2e-05, "grad_norm": 2.570868730545044, "loss_scale": 4.0, "retries": 0, "seconds": 130.5, "max_memory_gb": 17.61, "ce": 0.5560756498016417, "rank": 0.015602827072143555} +{"step": 69, "steps": 500, "lr": 2e-05, "grad_norm": 4.005553245544434, "loss_scale": 4.0, "retries": 0, "seconds": 132.4, "max_memory_gb": 17.61, "ce": 0.8415561863221228, "rank": 0.007128298282623291} +{"step": 70, "steps": 500, "lr": 2e-05, "grad_norm": 3.4456961154937744, "loss_scale": 4.0, "retries": 0, "seconds": 134.3, "max_memory_gb": 17.61, "ce": 0.7589482832700014, "rank": 0.0074616968631744385} +{"step": 71, "steps": 500, "lr": 2e-05, "grad_norm": 5.067704677581787, "loss_scale": 4.0, "retries": 0, "seconds": 136.2, "max_memory_gb": 17.61, "ce": 1.064785714726895, "rank": 0.0} +{"step": 72, "steps": 500, "lr": 2e-05, "grad_norm": 6.209153175354004, "loss_scale": 4.0, "retries": 0, "seconds": 138.1, "max_memory_gb": 17.61, "ce": 1.253060419112444, "rank": 0.0} +{"step": 73, "steps": 500, "lr": 2e-05, "grad_norm": 1.758890986442566, "loss_scale": 4.0, "retries": 0, "seconds": 140.0, "max_memory_gb": 17.61, "ce": 0.3420882811769843, "rank": 0.0} +{"step": 74, "steps": 500, "lr": 2e-05, "grad_norm": 1.9980697631835938, "loss_scale": 4.0, "retries": 0, "seconds": 141.9, "max_memory_gb": 17.61, "ce": 0.6990197217091918, "rank": 0.0} +{"step": 75, "steps": 500, "lr": 2e-05, "grad_norm": 2.000347137451172, "loss_scale": 4.0, "retries": 0, "seconds": 143.8, "max_memory_gb": 17.61, "ce": 0.9489462170749903, "rank": 0.24180907011032104} +{"step": 76, "steps": 500, "lr": 2e-05, "grad_norm": 1.8116854429244995, "loss_scale": 4.0, "retries": 0, "seconds": 145.7, "max_memory_gb": 17.61, "ce": 0.7471342645585537, "rank": 0.0} +{"step": 77, "steps": 500, "lr": 2e-05, "grad_norm": 3.023693561553955, "loss_scale": 4.0, "retries": 0, "seconds": 147.6, "max_memory_gb": 17.61, "ce": 1.0803726930171251, "rank": 0.04874533414840698} +{"step": 78, "steps": 500, "lr": 2e-05, "grad_norm": 3.447742223739624, "loss_scale": 4.0, "retries": 0, "seconds": 149.4, "max_memory_gb": 17.61, "ce": 0.7029129648581147, "rank": 0.0} +{"step": 79, "steps": 500, "lr": 2e-05, "grad_norm": 1.7532559633255005, "loss_scale": 4.0, "retries": 0, "seconds": 151.3, "max_memory_gb": 17.61, "ce": 0.45481486711651087, "rank": 0.0} +{"step": 80, "steps": 500, "lr": 2e-05, "grad_norm": 2.0567617416381836, "loss_scale": 4.0, "retries": 0, "seconds": 153.2, "max_memory_gb": 17.61, "ce": 0.5676070405170321, "rank": 0.0} +{"step": 81, "steps": 500, "lr": 2e-05, "grad_norm": 2.2517788410186768, "loss_scale": 4.0, "retries": 0, "seconds": 155.0, "max_memory_gb": 17.61, "ce": 0.8403578586876392, "rank": 0.0} +{"step": 82, "steps": 500, "lr": 2e-05, "grad_norm": 1.8420042991638184, "loss_scale": 4.0, "retries": 0, "seconds": 157.0, "max_memory_gb": 17.61, "ce": 0.6888204077258706, "rank": 0.0} +{"step": 83, "steps": 500, "lr": 2e-05, "grad_norm": 1.9777992963790894, "loss_scale": 4.0, "retries": 0, "seconds": 158.9, "max_memory_gb": 17.61, "ce": 0.8103516465052962, "rank": 0.0} +{"step": 84, "steps": 500, "lr": 2e-05, "grad_norm": 1.293627381324768, "loss_scale": 4.0, "retries": 0, "seconds": 160.8, "max_memory_gb": 17.61, "ce": 0.40081478632055223, "rank": 0.0} +{"step": 85, "steps": 500, "lr": 2e-05, "grad_norm": 1.6811423301696777, "loss_scale": 4.0, "retries": 0, "seconds": 162.7, "max_memory_gb": 17.61, "ce": 0.20698917284607887, "rank": 0.0} +{"step": 86, "steps": 500, "lr": 2e-05, "grad_norm": 2.0835301876068115, "loss_scale": 4.0, "retries": 0, "seconds": 164.6, "max_memory_gb": 17.61, "ce": 0.7739124931395054, "rank": 0.0} +{"step": 87, "steps": 500, "lr": 2e-05, "grad_norm": 2.5669262409210205, "loss_scale": 4.0, "retries": 0, "seconds": 166.4, "max_memory_gb": 17.61, "ce": 0.7640720494091511, "rank": 0.0} +{"step": 88, "steps": 500, "lr": 2e-05, "grad_norm": 1.791384220123291, "loss_scale": 4.0, "retries": 0, "seconds": 168.4, "max_memory_gb": 17.61, "ce": 0.6814777925610542, "rank": 0.0} +{"step": 89, "steps": 500, "lr": 2e-05, "grad_norm": 0.7354233860969543, "loss_scale": 4.0, "retries": 0, "seconds": 170.2, "max_memory_gb": 17.61, "ce": 0.19106744066812098, "rank": 0.12273287773132324} +{"step": 90, "steps": 500, "lr": 2e-05, "grad_norm": 1.4418784379959106, "loss_scale": 4.0, "retries": 0, "seconds": 172.1, "max_memory_gb": 17.61, "ce": 0.556664532981813, "rank": 0.0} +{"step": 91, "steps": 500, "lr": 2e-05, "grad_norm": 0.9443396329879761, "loss_scale": 4.0, "retries": 0, "seconds": 174.0, "max_memory_gb": 17.61, "ce": 0.4141117031686008, "rank": 0.0} +{"step": 92, "steps": 500, "lr": 2e-05, "grad_norm": 0.8620626330375671, "loss_scale": 4.0, "retries": 0, "seconds": 175.9, "max_memory_gb": 17.61, "ce": 0.35130131943151355, "rank": 0.0} +{"step": 93, "steps": 500, "lr": 2e-05, "grad_norm": 1.3010214567184448, "loss_scale": 4.0, "retries": 0, "seconds": 177.8, "max_memory_gb": 17.61, "ce": 0.5240423511713743, "rank": 0.0} +{"step": 94, "steps": 500, "lr": 2e-05, "grad_norm": 1.600634217262268, "loss_scale": 4.0, "retries": 0, "seconds": 179.7, "max_memory_gb": 17.61, "ce": 0.6492173317819834, "rank": 0.0} +{"step": 95, "steps": 500, "lr": 2e-05, "grad_norm": 1.2217636108398438, "loss_scale": 4.0, "retries": 0, "seconds": 181.6, "max_memory_gb": 17.61, "ce": 0.6495559136383235, "rank": 0.0} +{"step": 96, "steps": 500, "lr": 2e-05, "grad_norm": 1.8732231855392456, "loss_scale": 4.0, "retries": 0, "seconds": 183.5, "max_memory_gb": 17.61, "ce": 0.7581873992457986, "rank": 0.03846222162246704} +{"step": 97, "steps": 500, "lr": 2e-05, "grad_norm": 0.7975355386734009, "loss_scale": 4.0, "retries": 0, "seconds": 185.4, "max_memory_gb": 17.61, "ce": 0.3996682991273701, "rank": 0.01232251524925232} +{"step": 98, "steps": 500, "lr": 2e-05, "grad_norm": 1.2424615621566772, "loss_scale": 4.0, "retries": 0, "seconds": 187.3, "max_memory_gb": 17.61, "ce": 0.4069217077922076, "rank": 0.0} +{"step": 99, "steps": 500, "lr": 2e-05, "grad_norm": 0.754963755607605, "loss_scale": 4.0, "retries": 0, "seconds": 189.2, "max_memory_gb": 17.61, "ce": 0.2651486140675843, "rank": 0.05555222928524017} +{"step": 100, "steps": 500, "lr": 2e-05, "grad_norm": 1.7404526472091675, "loss_scale": 4.0, "retries": 0, "seconds": 191.1, "max_memory_gb": 17.61, "ce": 0.6575042018666863, "rank": 0.0} +{"step": 101, "steps": 500, "lr": 2e-05, "grad_norm": 1.5651428699493408, "loss_scale": 4.0, "retries": 0, "seconds": 193.0, "max_memory_gb": 17.61, "ce": 0.625669558532536, "rank": 0.0} +{"step": 102, "steps": 500, "lr": 2e-05, "grad_norm": 1.9816718101501465, "loss_scale": 4.0, "retries": 0, "seconds": 194.9, "max_memory_gb": 17.61, "ce": 0.7656032899394631, "rank": 0.0} +{"step": 103, "steps": 500, "lr": 2e-05, "grad_norm": 0.7000796794891357, "loss_scale": 4.0, "retries": 0, "seconds": 196.8, "max_memory_gb": 17.61, "ce": 0.36002194008324295, "rank": 0.1151496022939682} +{"step": 104, "steps": 500, "lr": 2e-05, "grad_norm": 0.5642911791801453, "loss_scale": 4.0, "retries": 0, "seconds": 198.7, "max_memory_gb": 17.61, "ce": 0.2807578848442063, "rank": 0.07318967580795288} +{"step": 105, "steps": 500, "lr": 2e-05, "grad_norm": 1.1457170248031616, "loss_scale": 4.0, "retries": 0, "seconds": 200.6, "max_memory_gb": 17.61, "ce": 0.5495631042867899, "rank": 0.05371904373168945} +{"step": 106, "steps": 500, "lr": 2e-05, "grad_norm": 0.9647225737571716, "loss_scale": 4.0, "retries": 0, "seconds": 202.5, "max_memory_gb": 17.61, "ce": 0.532035724259913, "rank": 0.07082043588161469} +{"step": 107, "steps": 500, "lr": 2e-05, "grad_norm": 0.38151195645332336, "loss_scale": 4.0, "retries": 0, "seconds": 204.4, "max_memory_gb": 17.61, "ce": 0.31246878020465374, "rank": 0.0} +{"step": 108, "steps": 500, "lr": 2e-05, "grad_norm": 0.39796245098114014, "loss_scale": 4.0, "retries": 0, "seconds": 206.3, "max_memory_gb": 17.61, "ce": 0.305632037576288, "rank": 0.0} +{"step": 109, "steps": 500, "lr": 2e-05, "grad_norm": 1.4034068584442139, "loss_scale": 4.0, "retries": 0, "seconds": 208.2, "max_memory_gb": 17.61, "ce": 0.6815853044390678, "rank": 0.0} +{"step": 110, "steps": 500, "lr": 2e-05, "grad_norm": 0.8951611518859863, "loss_scale": 4.0, "retries": 0, "seconds": 210.1, "max_memory_gb": 17.61, "ce": 0.4697133060544729, "rank": 0.02540796995162964} +{"step": 111, "steps": 500, "lr": 2e-05, "grad_norm": 1.026528239250183, "loss_scale": 4.0, "retries": 0, "seconds": 212.0, "max_memory_gb": 17.61, "ce": 0.401528735877946, "rank": 0.0} +{"step": 112, "steps": 500, "lr": 2e-05, "grad_norm": 0.5353949069976807, "loss_scale": 4.0, "retries": 0, "seconds": 213.9, "max_memory_gb": 17.61, "ce": 0.34729003557004035, "rank": 0.0} +{"step": 113, "steps": 500, "lr": 2e-05, "grad_norm": 0.6864220499992371, "loss_scale": 4.0, "retries": 0, "seconds": 215.8, "max_memory_gb": 17.61, "ce": 0.39123945496976376, "rank": 0.0} +{"step": 114, "steps": 500, "lr": 2e-05, "grad_norm": 2.5015690326690674, "loss_scale": 4.0, "retries": 0, "seconds": 217.8, "max_memory_gb": 17.61, "ce": 0.8047444792464375, "rank": 0.1853102743625641} +{"step": 115, "steps": 500, "lr": 2e-05, "grad_norm": 0.5072789192199707, "loss_scale": 4.0, "retries": 0, "seconds": 219.7, "max_memory_gb": 17.61, "ce": 0.4649037541821599, "rank": 0.04113910347223282} +{"step": 116, "steps": 500, "lr": 2e-05, "grad_norm": 0.16721050441265106, "loss_scale": 4.0, "retries": 0, "seconds": 221.6, "max_memory_gb": 17.61, "ce": 0.15388388745486736, "rank": 0.0} +{"step": 117, "steps": 500, "lr": 2e-05, "grad_norm": 0.8910313844680786, "loss_scale": 4.0, "retries": 0, "seconds": 223.4, "max_memory_gb": 17.61, "ce": 0.5012393940705806, "rank": 0.0} +{"step": 118, "steps": 500, "lr": 2e-05, "grad_norm": 0.3909754455089569, "loss_scale": 4.0, "retries": 0, "seconds": 225.4, "max_memory_gb": 17.61, "ce": 0.293410254875198, "rank": 0.0} +{"step": 119, "steps": 500, "lr": 2e-05, "grad_norm": 0.37189024686813354, "loss_scale": 4.0, "retries": 0, "seconds": 227.2, "max_memory_gb": 17.61, "ce": 0.2754748477600515, "rank": 0.0} +{"step": 120, "steps": 500, "lr": 2e-05, "grad_norm": 0.8739343881607056, "loss_scale": 4.0, "retries": 0, "seconds": 229.2, "max_memory_gb": 17.61, "ce": 0.453954987693578, "rank": 0.0} +{"step": 121, "steps": 500, "lr": 2e-05, "grad_norm": 0.6375969052314758, "loss_scale": 4.0, "retries": 0, "seconds": 231.1, "max_memory_gb": 17.61, "ce": 0.38968784851022065, "rank": 0.0} +{"step": 122, "steps": 500, "lr": 2e-05, "grad_norm": 0.401009738445282, "loss_scale": 4.0, "retries": 0, "seconds": 233.0, "max_memory_gb": 17.61, "ce": 0.3952163301873952, "rank": 0.0} +{"step": 123, "steps": 500, "lr": 2e-05, "grad_norm": 0.9340283870697021, "loss_scale": 4.0, "retries": 0, "seconds": 234.9, "max_memory_gb": 17.61, "ce": 0.4745215973816812, "rank": 0.018634289503097534} +{"step": 124, "steps": 500, "lr": 2e-05, "grad_norm": 0.5984257459640503, "loss_scale": 4.0, "retries": 0, "seconds": 236.8, "max_memory_gb": 17.61, "ce": 0.3823630539700389, "rank": 0.01803545653820038} +{"step": 125, "steps": 500, "lr": 2e-05, "grad_norm": 0.24426497519016266, "loss_scale": 4.0, "retries": 0, "seconds": 238.6, "max_memory_gb": 17.61, "ce": 0.2993541341274977, "rank": 0.0} +{"step": 126, "steps": 500, "lr": 2e-05, "grad_norm": 0.5666425228118896, "loss_scale": 4.0, "retries": 0, "seconds": 240.5, "max_memory_gb": 17.61, "ce": 0.5549223627895117, "rank": 0.0} +{"step": 127, "steps": 500, "lr": 2e-05, "grad_norm": 0.3029765188694, "loss_scale": 4.0, "retries": 0, "seconds": 242.4, "max_memory_gb": 17.61, "ce": 0.1883907694136724, "rank": 0.0} +{"step": 128, "steps": 500, "lr": 2e-05, "grad_norm": 0.5453979969024658, "loss_scale": 4.0, "retries": 0, "seconds": 244.3, "max_memory_gb": 17.61, "ce": 0.43761829496361315, "rank": 0.0} +{"step": 129, "steps": 500, "lr": 2e-05, "grad_norm": 0.9197272658348083, "loss_scale": 4.0, "retries": 0, "seconds": 246.2, "max_memory_gb": 17.61, "ce": 0.5120241660624743, "rank": 0.0} +{"step": 130, "steps": 500, "lr": 2e-05, "grad_norm": 0.6572425365447998, "loss_scale": 4.0, "retries": 0, "seconds": 248.1, "max_memory_gb": 17.61, "ce": 0.5895254681818187, "rank": 0.0} +{"step": 131, "steps": 500, "lr": 2e-05, "grad_norm": 0.5543351173400879, "loss_scale": 4.0, "retries": 0, "seconds": 249.9, "max_memory_gb": 17.61, "ce": 0.36973869102075696, "rank": 0.0} +{"step": 132, "steps": 500, "lr": 2e-05, "grad_norm": 1.0365861654281616, "loss_scale": 4.0, "retries": 0, "seconds": 251.8, "max_memory_gb": 17.61, "ce": 0.5438667148118839, "rank": 0.021390527486801147} +{"step": 133, "steps": 500, "lr": 2e-05, "grad_norm": 0.9824712872505188, "loss_scale": 4.0, "retries": 0, "seconds": 253.7, "max_memory_gb": 17.61, "ce": 0.5077559929341078, "rank": 0.09833615273237228} +{"step": 134, "steps": 500, "lr": 2e-05, "grad_norm": 0.4038294851779938, "loss_scale": 4.0, "retries": 0, "seconds": 255.6, "max_memory_gb": 17.61, "ce": 0.22374399658292532, "rank": 0.0} +{"step": 135, "steps": 500, "lr": 2e-05, "grad_norm": 0.7028629779815674, "loss_scale": 4.0, "retries": 0, "seconds": 257.5, "max_memory_gb": 17.61, "ce": 0.44555578008294106, "rank": 0.07784059643745422} +{"step": 136, "steps": 500, "lr": 2e-05, "grad_norm": 0.6274380087852478, "loss_scale": 4.0, "retries": 0, "seconds": 259.4, "max_memory_gb": 17.61, "ce": 0.40258049429394305, "rank": 0.0} +{"step": 137, "steps": 500, "lr": 2e-05, "grad_norm": 0.40021049976348877, "loss_scale": 4.0, "retries": 0, "seconds": 261.3, "max_memory_gb": 17.61, "ce": 0.3444468593224883, "rank": 0.0} +{"step": 138, "steps": 500, "lr": 2e-05, "grad_norm": 0.5365908145904541, "loss_scale": 4.0, "retries": 0, "seconds": 263.2, "max_memory_gb": 17.61, "ce": 0.35641899099573493, "rank": 0.03670495003461838} +{"step": 139, "steps": 500, "lr": 2e-05, "grad_norm": 1.1201881170272827, "loss_scale": 4.0, "retries": 0, "seconds": 265.1, "max_memory_gb": 17.61, "ce": 0.5218862001784146, "rank": 0.0} +{"step": 140, "steps": 500, "lr": 2e-05, "grad_norm": 0.5095705389976501, "loss_scale": 4.0, "retries": 0, "seconds": 267.0, "max_memory_gb": 17.61, "ce": 0.2800367169547826, "rank": 0.0} +{"step": 141, "steps": 500, "lr": 2e-05, "grad_norm": 0.4837019741535187, "loss_scale": 4.0, "retries": 0, "seconds": 269.0, "max_memory_gb": 17.61, "ce": 0.3607970755547285, "rank": 0.0} +{"step": 142, "steps": 500, "lr": 2e-05, "grad_norm": 0.4444679915904999, "loss_scale": 4.0, "retries": 0, "seconds": 270.8, "max_memory_gb": 17.61, "ce": 0.3279167111031711, "rank": 0.0} +{"step": 143, "steps": 500, "lr": 2e-05, "grad_norm": 0.25087475776672363, "loss_scale": 4.0, "retries": 0, "seconds": 272.7, "max_memory_gb": 17.61, "ce": 0.18391004885779694, "rank": 0.0} +{"step": 144, "steps": 500, "lr": 2e-05, "grad_norm": 0.6566535234451294, "loss_scale": 4.0, "retries": 0, "seconds": 274.7, "max_memory_gb": 17.61, "ce": 0.3868259999435395, "rank": 0.0} +{"step": 145, "steps": 500, "lr": 2e-05, "grad_norm": 0.45754915475845337, "loss_scale": 4.0, "retries": 0, "seconds": 276.6, "max_memory_gb": 17.61, "ce": 0.23353316116845235, "rank": 0.0} +{"step": 146, "steps": 500, "lr": 2e-05, "grad_norm": 0.5185317397117615, "loss_scale": 4.0, "retries": 0, "seconds": 278.6, "max_memory_gb": 17.61, "ce": 0.31523694144561887, "rank": 0.0} +{"step": 147, "steps": 500, "lr": 2e-05, "grad_norm": 0.6379430294036865, "loss_scale": 4.0, "retries": 0, "seconds": 280.5, "max_memory_gb": 17.61, "ce": 0.5102244589943439, "rank": 0.0} +{"step": 148, "steps": 500, "lr": 2e-05, "grad_norm": 0.29972460865974426, "loss_scale": 4.0, "retries": 0, "seconds": 282.4, "max_memory_gb": 17.61, "ce": 0.14532663661520928, "rank": 0.0} +{"step": 149, "steps": 500, "lr": 2e-05, "grad_norm": 0.4036838710308075, "loss_scale": 4.0, "retries": 0, "seconds": 284.2, "max_memory_gb": 17.61, "ce": 0.2531642052344978, "rank": 0.13155466318130493} +{"step": 150, "steps": 500, "lr": 2e-05, "grad_norm": 0.7730251550674438, "loss_scale": 4.0, "retries": 0, "seconds": 286.2, "max_memory_gb": 17.61, "ce": 0.4493866916745901, "rank": 0.04267212748527527} +{"step": 151, "steps": 500, "lr": 2e-05, "grad_norm": 0.5209571719169617, "loss_scale": 4.0, "retries": 0, "seconds": 288.1, "max_memory_gb": 17.61, "ce": 0.24439725204138085, "rank": 0.0} +{"step": 152, "steps": 500, "lr": 2e-05, "grad_norm": 0.8851020336151123, "loss_scale": 4.0, "retries": 0, "seconds": 290.0, "max_memory_gb": 17.61, "ce": 0.45626596082001925, "rank": 0.0} +{"step": 153, "steps": 500, "lr": 2e-05, "grad_norm": 0.6572021245956421, "loss_scale": 4.0, "retries": 0, "seconds": 291.9, "max_memory_gb": 17.61, "ce": 0.344332629814744, "rank": 0.0} +{"step": 154, "steps": 500, "lr": 2e-05, "grad_norm": 1.20832097530365, "loss_scale": 4.0, "retries": 0, "seconds": 293.8, "max_memory_gb": 17.61, "ce": 0.6075219237245619, "rank": 0.13175323605537415} +{"step": 155, "steps": 500, "lr": 2e-05, "grad_norm": 1.5944589376449585, "loss_scale": 4.0, "retries": 0, "seconds": 295.7, "max_memory_gb": 17.61, "ce": 0.6088062521303073, "rank": 0.033988773822784424} +{"step": 156, "steps": 500, "lr": 2e-05, "grad_norm": 0.5579633712768555, "loss_scale": 4.0, "retries": 0, "seconds": 297.6, "max_memory_gb": 17.61, "ce": 0.2775837172521278, "rank": 0.0} +{"step": 157, "steps": 500, "lr": 2e-05, "grad_norm": 0.4807203412055969, "loss_scale": 4.0, "retries": 0, "seconds": 299.5, "max_memory_gb": 17.61, "ce": 0.2803621341008693, "rank": 0.0} +{"step": 158, "steps": 500, "lr": 2e-05, "grad_norm": 0.7144293785095215, "loss_scale": 4.0, "retries": 0, "seconds": 301.4, "max_memory_gb": 17.61, "ce": 0.47121773613616824, "rank": 0.0} +{"step": 159, "steps": 500, "lr": 2e-05, "grad_norm": 0.47321832180023193, "loss_scale": 4.0, "retries": 0, "seconds": 303.3, "max_memory_gb": 17.62, "ce": 0.27339688618667424, "rank": 0.0} +{"step": 160, "steps": 500, "lr": 2e-05, "grad_norm": 0.7141905426979065, "loss_scale": 4.0, "retries": 0, "seconds": 305.2, "max_memory_gb": 17.62, "ce": 0.35417798021808267, "rank": 0.05431901663541794} +{"step": 161, "steps": 500, "lr": 2e-05, "grad_norm": 0.4005620777606964, "loss_scale": 4.0, "retries": 0, "seconds": 307.1, "max_memory_gb": 17.62, "ce": 0.32245117297861725, "rank": 0.0} +{"step": 162, "steps": 500, "lr": 2e-05, "grad_norm": 0.3361662030220032, "loss_scale": 4.0, "retries": 0, "seconds": 308.9, "max_memory_gb": 17.62, "ce": 0.21991487592458725, "rank": 0.0} +{"step": 163, "steps": 500, "lr": 2e-05, "grad_norm": 0.7118457555770874, "loss_scale": 4.0, "retries": 0, "seconds": 310.8, "max_memory_gb": 17.62, "ce": 0.31911395164206624, "rank": 0.0} +{"step": 164, "steps": 500, "lr": 2e-05, "grad_norm": 0.7815670371055603, "loss_scale": 4.0, "retries": 0, "seconds": 312.7, "max_memory_gb": 17.62, "ce": 0.4045450543053448, "rank": 0.03378383815288544} +{"step": 165, "steps": 500, "lr": 2e-05, "grad_norm": 0.467705100774765, "loss_scale": 4.0, "retries": 0, "seconds": 314.7, "max_memory_gb": 17.62, "ce": 0.2577891335822642, "rank": 0.0} +{"step": 166, "steps": 500, "lr": 2e-05, "grad_norm": 0.681520938873291, "loss_scale": 4.0, "retries": 0, "seconds": 316.6, "max_memory_gb": 17.62, "ce": 0.3542156000621617, "rank": 0.0} +{"step": 167, "steps": 500, "lr": 2e-05, "grad_norm": 0.7736896276473999, "loss_scale": 4.0, "retries": 0, "seconds": 318.5, "max_memory_gb": 17.62, "ce": 0.38189835706725717, "rank": 0.0} +{"step": 168, "steps": 500, "lr": 2e-05, "grad_norm": 0.5341899394989014, "loss_scale": 4.0, "retries": 0, "seconds": 320.4, "max_memory_gb": 17.62, "ce": 0.25371910841204226, "rank": 0.0} +{"step": 169, "steps": 500, "lr": 2e-05, "grad_norm": 1.7604292631149292, "loss_scale": 4.0, "retries": 0, "seconds": 322.4, "max_memory_gb": 17.62, "ce": 0.8023899430409074, "rank": 0.06497898697853088} +{"step": 170, "steps": 500, "lr": 2e-05, "grad_norm": 0.7336684465408325, "loss_scale": 4.0, "retries": 0, "seconds": 324.2, "max_memory_gb": 17.62, "ce": 0.2569984889123589, "rank": 0.0} +{"step": 171, "steps": 500, "lr": 2e-05, "grad_norm": 0.895559549331665, "loss_scale": 4.0, "retries": 0, "seconds": 326.2, "max_memory_gb": 17.62, "ce": 0.30403738416498527, "rank": 0.008571013808250427} +{"step": 172, "steps": 500, "lr": 2e-05, "grad_norm": 0.9203730225563049, "loss_scale": 4.0, "retries": 0, "seconds": 328.1, "max_memory_gb": 17.62, "ce": 0.3496646052226424, "rank": 0.0} +{"step": 173, "steps": 500, "lr": 2e-05, "grad_norm": 1.4512323141098022, "loss_scale": 4.0, "retries": 0, "seconds": 330.0, "max_memory_gb": 17.62, "ce": 0.42401905125007033, "rank": 0.045229315757751465} +{"step": 174, "steps": 500, "lr": 2e-05, "grad_norm": 1.1134357452392578, "loss_scale": 4.0, "retries": 0, "seconds": 331.9, "max_memory_gb": 17.62, "ce": 0.29826906422385946, "rank": 0.0} +{"step": 175, "steps": 500, "lr": 2e-05, "grad_norm": 0.6933056712150574, "loss_scale": 4.0, "retries": 0, "seconds": 333.8, "max_memory_gb": 17.62, "ce": 0.38816945953294635, "rank": 0.0} +{"step": 176, "steps": 500, "lr": 2e-05, "grad_norm": 1.10580313205719, "loss_scale": 4.0, "retries": 0, "seconds": 335.7, "max_memory_gb": 17.62, "ce": 0.30417350586503744, "rank": 0.0} +{"step": 177, "steps": 500, "lr": 2e-05, "grad_norm": 0.9005101919174194, "loss_scale": 4.0, "retries": 0, "seconds": 337.5, "max_memory_gb": 17.62, "ce": 0.304688852513209, "rank": 0.12740643322467804} +{"step": 178, "steps": 500, "lr": 2e-05, "grad_norm": 0.5563676357269287, "loss_scale": 4.0, "retries": 0, "seconds": 339.4, "max_memory_gb": 17.62, "ce": 0.2811438653152436, "rank": 0.0} +{"step": 179, "steps": 500, "lr": 2e-05, "grad_norm": 0.7238121628761292, "loss_scale": 4.0, "retries": 0, "seconds": 341.3, "max_memory_gb": 17.62, "ce": 0.4443615079508163, "rank": 0.0} +{"step": 180, "steps": 500, "lr": 2e-05, "grad_norm": 1.4325282573699951, "loss_scale": 4.0, "retries": 0, "seconds": 343.2, "max_memory_gb": 17.62, "ce": 0.44973368290811777, "rank": 0.0} +{"step": 181, "steps": 500, "lr": 2e-05, "grad_norm": 1.199471354484558, "loss_scale": 4.0, "retries": 0, "seconds": 345.1, "max_memory_gb": 17.62, "ce": 0.33192519284784794, "rank": 0.01670973002910614} +{"step": 182, "steps": 500, "lr": 2e-05, "grad_norm": 0.755095899105072, "loss_scale": 4.0, "retries": 0, "seconds": 347.0, "max_memory_gb": 17.62, "ce": 0.24139902426395565, "rank": 0.0} +{"step": 183, "steps": 500, "lr": 2e-05, "grad_norm": 0.4570786654949188, "loss_scale": 4.0, "retries": 0, "seconds": 348.9, "max_memory_gb": 17.62, "ce": 0.1420357613824308, "rank": 0.0} +{"step": 184, "steps": 500, "lr": 2e-05, "grad_norm": 0.8897966146469116, "loss_scale": 4.0, "retries": 0, "seconds": 350.8, "max_memory_gb": 17.62, "ce": 0.2160217969212681, "rank": 0.11676619201898575} +{"step": 185, "steps": 500, "lr": 2e-05, "grad_norm": 0.8840776085853577, "loss_scale": 4.0, "retries": 0, "seconds": 352.8, "max_memory_gb": 17.62, "ce": 0.25948365475051105, "rank": 0.0} +{"step": 186, "steps": 500, "lr": 2e-05, "grad_norm": 0.6464740633964539, "loss_scale": 4.0, "retries": 0, "seconds": 354.7, "max_memory_gb": 17.62, "ce": 0.1773920910200104, "rank": 0.0} +{"step": 187, "steps": 500, "lr": 2e-05, "grad_norm": 1.8266085386276245, "loss_scale": 4.0, "retries": 0, "seconds": 356.6, "max_memory_gb": 17.62, "ce": 0.4123955345712602, "rank": 0.15068164467811584} +{"step": 188, "steps": 500, "lr": 2e-05, "grad_norm": 0.18801350891590118, "loss_scale": 4.0, "retries": 0, "seconds": 358.5, "max_memory_gb": 17.62, "ce": 0.16659495839849114, "rank": 0.0} +{"step": 189, "steps": 500, "lr": 2e-05, "grad_norm": 0.9979846477508545, "loss_scale": 4.0, "retries": 0, "seconds": 360.4, "max_memory_gb": 17.62, "ce": 0.22874652175232768, "rank": 0.0} +{"step": 190, "steps": 500, "lr": 2e-05, "grad_norm": 1.2999645471572876, "loss_scale": 4.0, "retries": 0, "seconds": 362.3, "max_memory_gb": 17.62, "ce": 0.31790949450805783, "rank": 0.0} +{"step": 191, "steps": 500, "lr": 2e-05, "grad_norm": 0.5393891334533691, "loss_scale": 4.0, "retries": 0, "seconds": 364.1, "max_memory_gb": 17.62, "ce": 0.3246500282548368, "rank": 0.0} +{"step": 192, "steps": 500, "lr": 2e-05, "grad_norm": 0.4970167279243469, "loss_scale": 4.0, "retries": 0, "seconds": 366.0, "max_memory_gb": 17.62, "ce": 0.30794338090345263, "rank": 0.0} +{"step": 193, "steps": 500, "lr": 2e-05, "grad_norm": 1.0006989240646362, "loss_scale": 4.0, "retries": 0, "seconds": 367.9, "max_memory_gb": 17.62, "ce": 0.22077365207951516, "rank": 0.16559460014104843} +{"step": 194, "steps": 500, "lr": 2e-05, "grad_norm": 0.9253080487251282, "loss_scale": 4.0, "retries": 0, "seconds": 369.8, "max_memory_gb": 17.62, "ce": 0.3546040584333241, "rank": 0.0} +{"step": 195, "steps": 500, "lr": 2e-05, "grad_norm": 0.8479958176612854, "loss_scale": 4.0, "retries": 0, "seconds": 371.7, "max_memory_gb": 17.62, "ce": 0.3137611201964319, "rank": 0.0} +{"step": 196, "steps": 500, "lr": 2e-05, "grad_norm": 0.9587298631668091, "loss_scale": 4.0, "retries": 0, "seconds": 373.7, "max_memory_gb": 17.62, "ce": 0.1908664555521682, "rank": 0.0} +{"step": 197, "steps": 500, "lr": 2e-05, "grad_norm": 0.6623421311378479, "loss_scale": 4.0, "retries": 0, "seconds": 375.6, "max_memory_gb": 17.62, "ce": 0.2226924393326044, "rank": 0.06954431533813477} +{"step": 198, "steps": 500, "lr": 2e-05, "grad_norm": 0.7524804472923279, "loss_scale": 4.0, "retries": 0, "seconds": 377.5, "max_memory_gb": 17.62, "ce": 0.28734448133036494, "rank": 0.0} +{"step": 199, "steps": 500, "lr": 2e-05, "grad_norm": 1.0935531854629517, "loss_scale": 4.0, "retries": 0, "seconds": 379.4, "max_memory_gb": 17.62, "ce": 0.2916911148931831, "rank": 0.0} +{"step": 200, "steps": 500, "lr": 2e-05, "grad_norm": 0.7065804600715637, "loss_scale": 4.0, "retries": 0, "seconds": 381.3, "max_memory_gb": 17.62, "ce": 0.31239595590159297, "rank": 0.02497398853302002} +{"step": 201, "steps": 500, "lr": 2e-05, "grad_norm": 0.7745880484580994, "loss_scale": 4.0, "retries": 0, "seconds": 383.2, "max_memory_gb": 17.62, "ce": 0.1976900501176715, "rank": 0.0} +{"step": 202, "steps": 500, "lr": 2e-05, "grad_norm": 0.7082841396331787, "loss_scale": 4.0, "retries": 0, "seconds": 385.0, "max_memory_gb": 17.62, "ce": 0.08647937735076994, "rank": 0.03899924457073212} +{"step": 203, "steps": 500, "lr": 2e-05, "grad_norm": 0.9472269415855408, "loss_scale": 4.0, "retries": 0, "seconds": 386.9, "max_memory_gb": 17.62, "ce": 0.24586137710139155, "rank": 0.0} +{"step": 204, "steps": 500, "lr": 2e-05, "grad_norm": 2.9441375732421875, "loss_scale": 4.0, "retries": 0, "seconds": 388.9, "max_memory_gb": 17.62, "ce": 0.3046845984645188, "rank": 0.0} +{"step": 205, "steps": 500, "lr": 2e-05, "grad_norm": 1.613334059715271, "loss_scale": 4.0, "retries": 0, "seconds": 390.7, "max_memory_gb": 17.62, "ce": 0.2680449197650887, "rank": 0.0} +{"step": 206, "steps": 500, "lr": 2e-05, "grad_norm": 2.6974422931671143, "loss_scale": 4.0, "retries": 0, "seconds": 392.7, "max_memory_gb": 17.62, "ce": 0.24792829871876165, "rank": 0.0} +{"step": 207, "steps": 500, "lr": 2e-05, "grad_norm": 1.2713772058486938, "loss_scale": 4.0, "retries": 0, "seconds": 394.6, "max_memory_gb": 17.62, "ce": 0.46825971687212586, "rank": 0.20317721366882324} +{"step": 208, "steps": 500, "lr": 2e-05, "grad_norm": 0.8331745266914368, "loss_scale": 4.0, "retries": 0, "seconds": 396.5, "max_memory_gb": 17.62, "ce": 0.11437043722253293, "rank": 0.0} +{"step": 209, "steps": 500, "lr": 2e-05, "grad_norm": 0.6892129778862, "loss_scale": 4.0, "retries": 0, "seconds": 398.4, "max_memory_gb": 17.62, "ce": 0.26959741162136197, "rank": 0.1256023794412613} +{"step": 210, "steps": 500, "lr": 2e-05, "grad_norm": 0.7698301672935486, "loss_scale": 4.0, "retries": 0, "seconds": 400.3, "max_memory_gb": 17.62, "ce": 0.28073993290308863, "rank": 0.04982566833496094} +{"step": 211, "steps": 500, "lr": 2e-05, "grad_norm": 0.8257426619529724, "loss_scale": 4.0, "retries": 0, "seconds": 402.2, "max_memory_gb": 17.62, "ce": 0.2638981754425913, "rank": 0.0} +{"step": 212, "steps": 500, "lr": 2e-05, "grad_norm": 1.665658950805664, "loss_scale": 4.0, "retries": 0, "seconds": 404.2, "max_memory_gb": 17.62, "ce": 0.42338212253525853, "rank": 0.0} +{"step": 213, "steps": 500, "lr": 2e-05, "grad_norm": 0.7128347158432007, "loss_scale": 4.0, "retries": 0, "seconds": 406.1, "max_memory_gb": 17.62, "ce": 0.3225014074705541, "rank": 0.0} +{"step": 214, "steps": 500, "lr": 2e-05, "grad_norm": 1.079816222190857, "loss_scale": 4.0, "retries": 0, "seconds": 408.0, "max_memory_gb": 17.62, "ce": 0.37770184691180475, "rank": 0.0} +{"step": 215, "steps": 500, "lr": 2e-05, "grad_norm": 1.40104341506958, "loss_scale": 4.0, "retries": 0, "seconds": 409.9, "max_memory_gb": 17.62, "ce": 0.4112990403082222, "rank": 0.030567705631256104} +{"step": 216, "steps": 500, "lr": 2e-05, "grad_norm": 1.0757101774215698, "loss_scale": 4.0, "retries": 0, "seconds": 411.8, "max_memory_gb": 17.62, "ce": 0.24220225639874116, "rank": 0.0} +{"step": 217, "steps": 500, "lr": 2e-05, "grad_norm": 0.6254862546920776, "loss_scale": 4.0, "retries": 0, "seconds": 413.7, "max_memory_gb": 17.62, "ce": 0.10821730480529368, "rank": 0.0} +{"step": 218, "steps": 500, "lr": 2e-05, "grad_norm": 2.151399850845337, "loss_scale": 4.0, "retries": 0, "seconds": 415.6, "max_memory_gb": 17.62, "ce": 0.3289494013879448, "rank": 0.0} +{"step": 219, "steps": 500, "lr": 2e-05, "grad_norm": 2.6935718059539795, "loss_scale": 4.0, "retries": 0, "seconds": 417.5, "max_memory_gb": 17.62, "ce": 0.2102205923292786, "rank": 0.0} +{"step": 220, "steps": 500, "lr": 2e-05, "grad_norm": 1.3443914651870728, "loss_scale": 4.0, "retries": 0, "seconds": 419.4, "max_memory_gb": 17.62, "ce": 0.16590331413317472, "rank": 0.0} +{"step": 221, "steps": 500, "lr": 2e-05, "grad_norm": 1.9009417295455933, "loss_scale": 4.0, "retries": 0, "seconds": 421.3, "max_memory_gb": 17.62, "ce": 0.1911298311315477, "rank": 0.0} +{"step": 222, "steps": 500, "lr": 2e-05, "grad_norm": 3.043616771697998, "loss_scale": 4.0, "retries": 0, "seconds": 423.3, "max_memory_gb": 17.62, "ce": 0.271740491501987, "rank": 0.0} +{"step": 223, "steps": 500, "lr": 2e-05, "grad_norm": 0.5573574304580688, "loss_scale": 4.0, "retries": 0, "seconds": 425.2, "max_memory_gb": 17.62, "ce": 0.21853111952077597, "rank": 0.0} +{"step": 224, "steps": 500, "lr": 2e-05, "grad_norm": 1.1032178401947021, "loss_scale": 4.0, "retries": 0, "seconds": 427.1, "max_memory_gb": 17.62, "ce": 0.16088081640191376, "rank": 0.08981519937515259} +{"step": 225, "steps": 500, "lr": 2e-05, "grad_norm": 0.9057798981666565, "loss_scale": 4.0, "retries": 0, "seconds": 429.0, "max_memory_gb": 17.62, "ce": 0.1312090172432363, "rank": 0.0} +{"step": 226, "steps": 500, "lr": 2e-05, "grad_norm": 0.8210506439208984, "loss_scale": 4.0, "retries": 0, "seconds": 430.9, "max_memory_gb": 17.62, "ce": 0.1390044221188873, "rank": 0.0} +{"step": 227, "steps": 500, "lr": 2e-05, "grad_norm": 0.8283258080482483, "loss_scale": 4.0, "retries": 0, "seconds": 432.8, "max_memory_gb": 17.62, "ce": 0.0861221412778832, "rank": 0.0} +{"step": 228, "steps": 500, "lr": 2e-05, "grad_norm": 0.7205401659011841, "loss_scale": 4.0, "retries": 0, "seconds": 434.7, "max_memory_gb": 17.62, "ce": 0.106266338378191, "rank": 0.0} +{"step": 229, "steps": 500, "lr": 2e-05, "grad_norm": 0.6602327227592468, "loss_scale": 4.0, "retries": 0, "seconds": 436.6, "max_memory_gb": 17.62, "ce": 0.22284982982091606, "rank": 0.0} +{"step": 230, "steps": 500, "lr": 2e-05, "grad_norm": 0.6882899403572083, "loss_scale": 4.0, "retries": 0, "seconds": 438.4, "max_memory_gb": 17.62, "ce": 0.22888859128579497, "rank": 0.0} +{"step": 231, "steps": 500, "lr": 2e-05, "grad_norm": 0.6447615027427673, "loss_scale": 4.0, "retries": 0, "seconds": 440.3, "max_memory_gb": 17.62, "ce": 0.2567627936950885, "rank": 0.0} +{"step": 232, "steps": 500, "lr": 2e-05, "grad_norm": 2.6465604305267334, "loss_scale": 4.0, "retries": 0, "seconds": 442.2, "max_memory_gb": 17.62, "ce": 0.13651965372264385, "rank": 0.0} +{"step": 233, "steps": 500, "lr": 2e-05, "grad_norm": 5.521836757659912, "loss_scale": 4.0, "retries": 0, "seconds": 444.1, "max_memory_gb": 17.62, "ce": 0.32314199116081, "rank": 0.0} +{"step": 234, "steps": 500, "lr": 2e-05, "grad_norm": 0.7032188773155212, "loss_scale": 4.0, "retries": 0, "seconds": 446.0, "max_memory_gb": 17.62, "ce": 0.20149191783275455, "rank": 0.0} +{"step": 235, "steps": 500, "lr": 2e-05, "grad_norm": 0.9486814141273499, "loss_scale": 4.0, "retries": 0, "seconds": 447.9, "max_memory_gb": 17.62, "ce": 0.30909429932944477, "rank": 0.0} +{"step": 236, "steps": 500, "lr": 2e-05, "grad_norm": 0.5095072984695435, "loss_scale": 4.0, "retries": 0, "seconds": 449.8, "max_memory_gb": 17.62, "ce": 0.12919066566973925, "rank": 0.0} +{"step": 237, "steps": 500, "lr": 2e-05, "grad_norm": 4.221158027648926, "loss_scale": 4.0, "retries": 0, "seconds": 451.7, "max_memory_gb": 17.62, "ce": 0.3050267891958356, "rank": 0.0} +{"step": 238, "steps": 500, "lr": 2e-05, "grad_norm": 1.3790534734725952, "loss_scale": 4.0, "retries": 0, "seconds": 453.6, "max_memory_gb": 17.62, "ce": 0.11197590141091496, "rank": 0.0} +{"step": 239, "steps": 500, "lr": 2e-05, "grad_norm": 0.7569031119346619, "loss_scale": 4.0, "retries": 0, "seconds": 455.5, "max_memory_gb": 17.62, "ce": 0.20131050376221538, "rank": 0.0} +{"step": 240, "steps": 500, "lr": 2e-05, "grad_norm": 1.6168992519378662, "loss_scale": 4.0, "retries": 0, "seconds": 457.4, "max_memory_gb": 17.62, "ce": 0.12429728690767661, "rank": 0.0} +{"step": 241, "steps": 500, "lr": 2e-05, "grad_norm": 0.45873314142227173, "loss_scale": 4.0, "retries": 0, "seconds": 459.3, "max_memory_gb": 17.62, "ce": 0.1757502118125558, "rank": 0.0} +{"step": 242, "steps": 500, "lr": 2e-05, "grad_norm": 1.153788685798645, "loss_scale": 4.0, "retries": 0, "seconds": 461.2, "max_memory_gb": 17.62, "ce": 0.16293895326089114, "rank": 0.0} +{"step": 243, "steps": 500, "lr": 2e-05, "grad_norm": 1.716958999633789, "loss_scale": 4.0, "retries": 0, "seconds": 463.1, "max_memory_gb": 17.62, "ce": 0.17897794797318056, "rank": 0.0} +{"step": 244, "steps": 500, "lr": 2e-05, "grad_norm": 1.2809215784072876, "loss_scale": 4.0, "retries": 0, "seconds": 465.0, "max_memory_gb": 17.62, "ce": 0.31730082153808326, "rank": 0.0} +{"step": 245, "steps": 500, "lr": 2e-05, "grad_norm": 0.8333182334899902, "loss_scale": 4.0, "retries": 0, "seconds": 466.9, "max_memory_gb": 17.62, "ce": 0.3720476625021547, "rank": 0.0} +{"step": 246, "steps": 500, "lr": 2e-05, "grad_norm": 1.2894991636276245, "loss_scale": 4.0, "retries": 0, "seconds": 468.9, "max_memory_gb": 17.62, "ce": 0.05412317576701753, "rank": 0.0} +{"step": 247, "steps": 500, "lr": 2e-05, "grad_norm": 2.256230115890503, "loss_scale": 4.0, "retries": 0, "seconds": 470.8, "max_memory_gb": 17.62, "ce": 0.17175111832330003, "rank": 0.0} +{"step": 248, "steps": 500, "lr": 2e-05, "grad_norm": 1.1605101823806763, "loss_scale": 4.0, "retries": 0, "seconds": 472.7, "max_memory_gb": 17.62, "ce": 0.08746472746133804, "rank": 0.0} +{"step": 249, "steps": 500, "lr": 2e-05, "grad_norm": 2.6533079147338867, "loss_scale": 4.0, "retries": 0, "seconds": 474.6, "max_memory_gb": 17.62, "ce": 0.17202382232062519, "rank": 0.0} +{"step": 250, "steps": 500, "lr": 2e-05, "grad_norm": 0.936753511428833, "loss_scale": 4.0, "retries": 0, "seconds": 476.5, "max_memory_gb": 17.62, "ce": 0.07635705528082326, "rank": 0.0} +{"step": 251, "steps": 500, "lr": 2e-05, "grad_norm": 0.45464587211608887, "loss_scale": 4.0, "retries": 0, "seconds": 478.4, "max_memory_gb": 17.62, "ce": 0.1417350860137958, "rank": 0.0} +{"step": 252, "steps": 500, "lr": 2e-05, "grad_norm": 1.48917555809021, "loss_scale": 4.0, "retries": 0, "seconds": 480.3, "max_memory_gb": 17.62, "ce": 0.0770907083642669, "rank": 0.0} +{"step": 253, "steps": 500, "lr": 2e-05, "grad_norm": 0.5363450050354004, "loss_scale": 4.0, "retries": 0, "seconds": 482.2, "max_memory_gb": 17.62, "ce": 0.12305683473823592, "rank": 0.0} +{"step": 254, "steps": 500, "lr": 2e-05, "grad_norm": 13.440760612487793, "loss_scale": 4.0, "retries": 0, "seconds": 484.1, "max_memory_gb": 17.62, "ce": 0.21516344812698662, "rank": 0.0} +{"step": 255, "steps": 500, "lr": 2e-05, "grad_norm": 0.5209751129150391, "loss_scale": 4.0, "retries": 0, "seconds": 486.0, "max_memory_gb": 17.62, "ce": 0.22307167737744749, "rank": 0.0} +{"step": 256, "steps": 500, "lr": 2e-05, "grad_norm": 0.34386512637138367, "loss_scale": 4.0, "retries": 0, "seconds": 487.9, "max_memory_gb": 17.62, "ce": 0.09392368083354086, "rank": 0.0} +{"step": 257, "steps": 500, "lr": 2e-05, "grad_norm": 0.6589995622634888, "loss_scale": 4.0, "retries": 0, "seconds": 489.8, "max_memory_gb": 17.62, "ce": 0.10401969478698447, "rank": 0.0} +{"step": 258, "steps": 500, "lr": 2e-05, "grad_norm": 3.607322931289673, "loss_scale": 4.0, "retries": 0, "seconds": 491.7, "max_memory_gb": 17.62, "ce": 0.17667194426758215, "rank": 0.0} +{"step": 259, "steps": 500, "lr": 2e-05, "grad_norm": 0.33489832282066345, "loss_scale": 4.0, "retries": 0, "seconds": 493.6, "max_memory_gb": 17.62, "ce": 0.046106520225293934, "rank": 0.0} +{"step": 260, "steps": 500, "lr": 2e-05, "grad_norm": 2.2975151538848877, "loss_scale": 4.0, "retries": 0, "seconds": 495.5, "max_memory_gb": 17.62, "ce": 0.19042521133087575, "rank": 0.0} +{"step": 261, "steps": 500, "lr": 2e-05, "grad_norm": 1.2779772281646729, "loss_scale": 4.0, "retries": 0, "seconds": 497.4, "max_memory_gb": 17.62, "ce": 0.14599506882950664, "rank": 0.0} +{"step": 262, "steps": 500, "lr": 2e-05, "grad_norm": 1.1592679023742676, "loss_scale": 4.0, "retries": 0, "seconds": 499.3, "max_memory_gb": 17.62, "ce": 0.19057882437482476, "rank": 0.0} +{"step": 263, "steps": 500, "lr": 2e-05, "grad_norm": 1.369299054145813, "loss_scale": 4.0, "retries": 0, "seconds": 501.2, "max_memory_gb": 17.62, "ce": 0.1978448781883344, "rank": 0.12447528541088104} +{"step": 264, "steps": 500, "lr": 2e-05, "grad_norm": 0.988897979259491, "loss_scale": 4.0, "retries": 0, "seconds": 503.1, "max_memory_gb": 17.62, "ce": 0.118045205323142, "rank": 0.0} +{"step": 265, "steps": 500, "lr": 2e-05, "grad_norm": 0.7014011740684509, "loss_scale": 4.0, "retries": 0, "seconds": 505.0, "max_memory_gb": 17.62, "ce": 0.1187131637125276, "rank": 0.0} +{"step": 266, "steps": 500, "lr": 2e-05, "grad_norm": 5.390406608581543, "loss_scale": 4.0, "retries": 0, "seconds": 507.0, "max_memory_gb": 17.62, "ce": 0.4836606863536872, "rank": 0.0} +{"step": 267, "steps": 500, "lr": 2e-05, "grad_norm": 0.45719200372695923, "loss_scale": 4.0, "retries": 0, "seconds": 508.9, "max_memory_gb": 17.62, "ce": 0.13964541343739256, "rank": 0.0} +{"step": 268, "steps": 500, "lr": 2e-05, "grad_norm": 0.4122008979320526, "loss_scale": 4.0, "retries": 0, "seconds": 510.8, "max_memory_gb": 17.62, "ce": 0.25214609125396237, "rank": 0.0} +{"step": 269, "steps": 500, "lr": 2e-05, "grad_norm": 6.637139320373535, "loss_scale": 4.0, "retries": 0, "seconds": 512.7, "max_memory_gb": 17.62, "ce": 0.613956309389323, "rank": 0.0} +{"step": 270, "steps": 500, "lr": 2e-05, "grad_norm": 2.2396366596221924, "loss_scale": 4.0, "retries": 0, "seconds": 514.6, "max_memory_gb": 17.62, "ce": 0.15643965292838402, "rank": 0.0} +{"step": 271, "steps": 500, "lr": 2e-05, "grad_norm": 3.254228115081787, "loss_scale": 4.0, "retries": 0, "seconds": 516.5, "max_memory_gb": 17.62, "ce": 0.04913675331044942, "rank": 0.013137586414813995} +{"step": 272, "steps": 500, "lr": 2e-05, "grad_norm": 2.8278729915618896, "loss_scale": 4.0, "retries": 0, "seconds": 518.4, "max_memory_gb": 17.62, "ce": 0.16295039164833724, "rank": 0.0} +{"step": 273, "steps": 500, "lr": 2e-05, "grad_norm": 2.0512783527374268, "loss_scale": 4.0, "retries": 0, "seconds": 520.3, "max_memory_gb": 17.62, "ce": 0.07359782018465921, "rank": 0.12533412873744965} +{"step": 274, "steps": 500, "lr": 2e-05, "grad_norm": 0.6757294535636902, "loss_scale": 4.0, "retries": 0, "seconds": 522.2, "max_memory_gb": 17.62, "ce": 0.13551623991224915, "rank": 0.0007549077272415161} +{"step": 275, "steps": 500, "lr": 2e-05, "grad_norm": 51.84917449951172, "loss_scale": 4.0, "retries": 0, "seconds": 524.1, "max_memory_gb": 17.62, "ce": 1.0702974572777748, "rank": 0.13439655303955078} +{"step": 276, "steps": 500, "lr": 2e-05, "grad_norm": 0.8469386696815491, "loss_scale": 4.0, "retries": 0, "seconds": 526.0, "max_memory_gb": 17.62, "ce": 0.021779978123959154, "rank": 0.0} +{"step": 277, "steps": 500, "lr": 2e-05, "grad_norm": 27.23138999938965, "loss_scale": 4.0, "retries": 0, "seconds": 528.0, "max_memory_gb": 17.62, "ce": 0.16717840201454237, "rank": 0.0} +{"step": 278, "steps": 500, "lr": 2e-05, "grad_norm": 4.122521877288818, "loss_scale": 4.0, "retries": 0, "seconds": 529.8, "max_memory_gb": 17.62, "ce": 0.25251774572825525, "rank": 0.0} +{"step": 279, "steps": 500, "lr": 2e-05, "grad_norm": 7.519510269165039, "loss_scale": 4.0, "retries": 0, "seconds": 531.7, "max_memory_gb": 17.62, "ce": 0.0763168124249205, "rank": 0.0} +{"step": 280, "steps": 500, "lr": 2e-05, "grad_norm": 4.628920078277588, "loss_scale": 4.0, "retries": 0, "seconds": 533.6, "max_memory_gb": 17.62, "ce": 0.0480146529735066, "rank": 0.0} +{"step": 281, "steps": 500, "lr": 2e-05, "grad_norm": 2.964615821838379, "loss_scale": 4.0, "retries": 0, "seconds": 535.5, "max_memory_gb": 17.62, "ce": 0.1493057635088917, "rank": 0.0} +{"step": 282, "steps": 500, "lr": 2e-05, "grad_norm": 0.9586161375045776, "loss_scale": 4.0, "retries": 0, "seconds": 537.4, "max_memory_gb": 17.62, "ce": 0.09713441805797629, "rank": 0.0} +{"step": 283, "steps": 500, "lr": 2e-05, "grad_norm": 4.145000457763672, "loss_scale": 4.0, "retries": 0, "seconds": 539.4, "max_memory_gb": 17.62, "ce": 0.10628162024659105, "rank": 0.0} +{"step": 284, "steps": 500, "lr": 2e-05, "grad_norm": 4.904512882232666, "loss_scale": 4.0, "retries": 0, "seconds": 541.3, "max_memory_gb": 17.62, "ce": 0.07547723120660521, "rank": 0.0} +{"step": 285, "steps": 500, "lr": 2e-05, "grad_norm": 2.477877616882324, "loss_scale": 4.0, "retries": 0, "seconds": 543.2, "max_memory_gb": 17.62, "ce": 0.14945990417618304, "rank": 0.0} +{"step": 286, "steps": 500, "lr": 2e-05, "grad_norm": 1.152693271636963, "loss_scale": 4.0, "retries": 0, "seconds": 545.1, "max_memory_gb": 17.62, "ce": 0.05583481566281989, "rank": 0.0} +{"step": 287, "steps": 500, "lr": 2e-05, "grad_norm": 1.2608788013458252, "loss_scale": 4.0, "retries": 0, "seconds": 547.1, "max_memory_gb": 17.62, "ce": 0.11822729813866317, "rank": 0.0} +{"step": 288, "steps": 500, "lr": 2e-05, "grad_norm": 0.2923955023288727, "loss_scale": 4.0, "retries": 0, "seconds": 549.0, "max_memory_gb": 17.62, "ce": 0.010194953181780875, "rank": 0.0} +{"step": 289, "steps": 500, "lr": 2e-05, "grad_norm": 2.2747480869293213, "loss_scale": 4.0, "retries": 0, "seconds": 550.9, "max_memory_gb": 17.62, "ce": 0.08908284673816524, "rank": 0.0} +{"step": 290, "steps": 500, "lr": 2e-05, "grad_norm": 0.4809015393257141, "loss_scale": 4.0, "retries": 0, "seconds": 552.8, "max_memory_gb": 17.62, "ce": 0.11150833748251898, "rank": 0.0} +{"step": 291, "steps": 500, "lr": 2e-05, "grad_norm": 0.3887788951396942, "loss_scale": 4.0, "retries": 0, "seconds": 554.7, "max_memory_gb": 17.62, "ce": 0.013656817289302126, "rank": 0.0} +{"step": 292, "steps": 500, "lr": 2e-05, "grad_norm": 25.096914291381836, "loss_scale": 4.0, "retries": 0, "seconds": 556.6, "max_memory_gb": 17.62, "ce": 0.2770511374110356, "rank": 0.0} +{"step": 293, "steps": 500, "lr": 2e-05, "grad_norm": 0.9263749718666077, "loss_scale": 4.0, "retries": 0, "seconds": 558.5, "max_memory_gb": 17.62, "ce": 0.08105682148016058, "rank": 0.0} +{"step": 294, "steps": 500, "lr": 2e-05, "grad_norm": 79.17929077148438, "loss_scale": 4.0, "retries": 0, "seconds": 560.4, "max_memory_gb": 17.62, "ce": 0.5073777819343377, "rank": 0.0} +{"step": 295, "steps": 500, "lr": 2e-05, "grad_norm": 6.02178955078125, "loss_scale": 4.0, "retries": 0, "seconds": 562.3, "max_memory_gb": 17.62, "ce": 0.15510724767227657, "rank": 0.0} +{"step": 296, "steps": 500, "lr": 2e-05, "grad_norm": 16.466930389404297, "loss_scale": 4.0, "retries": 0, "seconds": 564.2, "max_memory_gb": 17.62, "ce": 0.3171171527937986, "rank": 0.0} +{"step": 297, "steps": 500, "lr": 2e-05, "grad_norm": 0.8278719782829285, "loss_scale": 4.0, "retries": 0, "seconds": 566.0, "max_memory_gb": 17.62, "ce": 0.1536320211016573, "rank": 0.0} +{"step": 298, "steps": 500, "lr": 2e-05, "grad_norm": 7.4210686683654785, "loss_scale": 4.0, "retries": 0, "seconds": 567.8, "max_memory_gb": 17.62, "ce": 0.47374628120451234, "rank": 0.0} +{"step": 299, "steps": 500, "lr": 2e-05, "grad_norm": 2.2971222400665283, "loss_scale": 4.0, "retries": 0, "seconds": 569.6, "max_memory_gb": 17.62, "ce": 0.03790689782181289, "rank": 0.0} +{"step": 300, "steps": 500, "lr": 2e-05, "grad_norm": 1.2598848342895508, "loss_scale": 4.0, "retries": 0, "seconds": 571.4, "max_memory_gb": 17.62, "ce": 0.2793174057005672, "rank": 0.0} +{"step": 301, "steps": 500, "lr": 2e-05, "grad_norm": 0.7296242713928223, "loss_scale": 4.0, "retries": 0, "seconds": 573.1, "max_memory_gb": 17.62, "ce": 0.10341901780702756, "rank": 0.0} +{"step": 302, "steps": 500, "lr": 2e-05, "grad_norm": 2.476818799972534, "loss_scale": 4.0, "retries": 0, "seconds": 574.9, "max_memory_gb": 17.62, "ce": 0.02490816795034334, "rank": 0.0} +{"step": 303, "steps": 500, "lr": 2e-05, "grad_norm": 2.304229497909546, "loss_scale": 4.0, "retries": 0, "seconds": 576.7, "max_memory_gb": 17.62, "ce": 0.03329369082348421, "rank": 0.0} +{"step": 304, "steps": 500, "lr": 2e-05, "grad_norm": 3.9162437915802, "loss_scale": 4.0, "retries": 0, "seconds": 578.4, "max_memory_gb": 17.62, "ce": 0.11100301981787197, "rank": 0.12039791792631149} +{"step": 305, "steps": 500, "lr": 2e-05, "grad_norm": 4.983860015869141, "loss_scale": 4.0, "retries": 0, "seconds": 580.2, "max_memory_gb": 17.62, "ce": 0.14665721641358687, "rank": 0.0} +{"step": 306, "steps": 500, "lr": 2e-05, "grad_norm": 2.3819963932037354, "loss_scale": 4.0, "retries": 0, "seconds": 582.0, "max_memory_gb": 17.62, "ce": 0.04883553106628824, "rank": 0.0} +{"step": 307, "steps": 500, "lr": 2e-05, "grad_norm": 2.2535650730133057, "loss_scale": 4.0, "retries": 0, "seconds": 583.8, "max_memory_gb": 17.62, "ce": 0.17083366059523541, "rank": 0.0} +{"step": 308, "steps": 500, "lr": 2e-05, "grad_norm": 4.511019229888916, "loss_scale": 4.0, "retries": 0, "seconds": 585.6, "max_memory_gb": 17.62, "ce": 0.274120976915583, "rank": 0.0} +{"step": 309, "steps": 500, "lr": 2e-05, "grad_norm": 0.7272554039955139, "loss_scale": 4.0, "retries": 0, "seconds": 587.3, "max_memory_gb": 17.62, "ce": 0.18379999269382097, "rank": 0.0} +{"step": 310, "steps": 500, "lr": 2e-05, "grad_norm": 1.4710146188735962, "loss_scale": 4.0, "retries": 0, "seconds": 589.1, "max_memory_gb": 17.62, "ce": 0.03943112996057607, "rank": 0.0} +{"step": 311, "steps": 500, "lr": 2e-05, "grad_norm": 0.39176103472709656, "loss_scale": 4.0, "retries": 0, "seconds": 590.8, "max_memory_gb": 17.62, "ce": 0.10882693299208768, "rank": 0.0} +{"step": 312, "steps": 500, "lr": 2e-05, "grad_norm": 56.99782943725586, "loss_scale": 4.0, "retries": 0, "seconds": 592.6, "max_memory_gb": 17.62, "ce": 0.1629442902194569, "rank": 0.0} +{"step": 313, "steps": 500, "lr": 2e-05, "grad_norm": 16.5665340423584, "loss_scale": 4.0, "retries": 0, "seconds": 594.3, "max_memory_gb": 17.62, "ce": 0.23281725755077787, "rank": 0.0} +{"step": 314, "steps": 500, "lr": 2e-05, "grad_norm": 52.83463668823242, "loss_scale": 4.0, "retries": 0, "seconds": 596.1, "max_memory_gb": 17.62, "ce": 0.6349818401795346, "rank": 0.0} +{"step": 315, "steps": 500, "lr": 2e-05, "grad_norm": 81.95610809326172, "loss_scale": 4.0, "retries": 0, "seconds": 597.9, "max_memory_gb": 17.62, "ce": 0.33149355617933907, "rank": 0.0} +{"step": 316, "steps": 500, "lr": 2e-05, "grad_norm": 4.267809867858887, "loss_scale": 4.0, "retries": 0, "seconds": 599.7, "max_memory_gb": 17.62, "ce": 0.32605273387162015, "rank": 0.0} +{"step": 317, "steps": 500, "lr": 2e-05, "grad_norm": 11.912063598632812, "loss_scale": 4.0, "retries": 0, "seconds": 601.4, "max_memory_gb": 17.62, "ce": 0.13357530202483758, "rank": 0.0} +{"step": 318, "steps": 500, "lr": 2e-05, "grad_norm": 106.60529327392578, "loss_scale": 4.0, "retries": 0, "seconds": 603.2, "max_memory_gb": 17.62, "ce": 0.9116149324108846, "rank": 0.0} +{"step": 319, "steps": 500, "lr": 2e-05, "grad_norm": 1.9193674325942993, "loss_scale": 4.0, "retries": 0, "seconds": 605.0, "max_memory_gb": 17.62, "ce": 0.07163622169173323, "rank": 0.0} +{"step": 320, "steps": 500, "lr": 2e-05, "grad_norm": 1.384717345237732, "loss_scale": 4.0, "retries": 0, "seconds": 606.7, "max_memory_gb": 17.62, "ce": 0.11552793199371081, "rank": 0.0} +{"step": 321, "steps": 500, "lr": 2e-05, "grad_norm": 0.7205500602722168, "loss_scale": 4.0, "retries": 0, "seconds": 608.5, "max_memory_gb": 17.62, "ce": 0.08581620342738461, "rank": 0.12607529759407043} +{"step": 322, "steps": 500, "lr": 2e-05, "grad_norm": 1.1793211698532104, "loss_scale": 4.0, "retries": 0, "seconds": 610.2, "max_memory_gb": 17.62, "ce": 0.03350607828178909, "rank": 0.0} +{"step": 323, "steps": 500, "lr": 2e-05, "grad_norm": 5.064142227172852, "loss_scale": 4.0, "retries": 0, "seconds": 612.0, "max_memory_gb": 17.62, "ce": 0.08587601213366725, "rank": 0.0} +{"step": 324, "steps": 500, "lr": 2e-05, "grad_norm": 2.6709282398223877, "loss_scale": 4.0, "retries": 0, "seconds": 613.8, "max_memory_gb": 17.62, "ce": 0.05953135612071492, "rank": 0.0} +{"step": 325, "steps": 500, "lr": 2e-05, "grad_norm": 0.6215255260467529, "loss_scale": 4.0, "retries": 0, "seconds": 615.6, "max_memory_gb": 17.62, "ce": 0.22100625658640638, "rank": 0.0} +{"step": 326, "steps": 500, "lr": 2e-05, "grad_norm": 0.4033887982368469, "loss_scale": 4.0, "retries": 0, "seconds": 617.3, "max_memory_gb": 17.62, "ce": 0.05476840266783256, "rank": 0.09921230375766754} +{"step": 327, "steps": 500, "lr": 2e-05, "grad_norm": 8.558540344238281, "loss_scale": 4.0, "retries": 0, "seconds": 619.1, "max_memory_gb": 17.62, "ce": 0.3230949570934172, "rank": 0.0} +{"step": 328, "steps": 500, "lr": 2e-05, "grad_norm": 0.7998552322387695, "loss_scale": 4.0, "retries": 0, "seconds": 620.9, "max_memory_gb": 17.62, "ce": 0.15941339035634883, "rank": 0.0} +{"step": 329, "steps": 500, "lr": 2e-05, "grad_norm": 0.82587730884552, "loss_scale": 4.0, "retries": 0, "seconds": 622.7, "max_memory_gb": 17.62, "ce": 0.17455980768136214, "rank": 0.0} +{"step": 330, "steps": 500, "lr": 2e-05, "grad_norm": 1.8684265613555908, "loss_scale": 4.0, "retries": 0, "seconds": 624.4, "max_memory_gb": 17.62, "ce": 0.1454790173265792, "rank": 0.0} +{"step": 331, "steps": 500, "lr": 2e-05, "grad_norm": 5.613901615142822, "loss_scale": 4.0, "retries": 0, "seconds": 626.2, "max_memory_gb": 17.62, "ce": 0.395012631022837, "rank": 0.0} +{"step": 332, "steps": 500, "lr": 2e-05, "grad_norm": 3.4131391048431396, "loss_scale": 4.0, "retries": 0, "seconds": 628.0, "max_memory_gb": 17.62, "ce": 0.0917693768569734, "rank": 0.0} +{"step": 333, "steps": 500, "lr": 2e-05, "grad_norm": 1.186760425567627, "loss_scale": 4.0, "retries": 0, "seconds": 629.7, "max_memory_gb": 17.62, "ce": 0.27301532999263145, "rank": 0.0} +{"step": 334, "steps": 500, "lr": 2e-05, "grad_norm": 0.8051325678825378, "loss_scale": 4.0, "retries": 0, "seconds": 631.5, "max_memory_gb": 17.62, "ce": 0.12830205979116727, "rank": 0.0} +{"step": 335, "steps": 500, "lr": 2e-05, "grad_norm": 3.974026918411255, "loss_scale": 4.0, "retries": 0, "seconds": 633.3, "max_memory_gb": 17.62, "ce": 0.25128502398729324, "rank": 0.0} +{"step": 336, "steps": 500, "lr": 2e-05, "grad_norm": 1.165844202041626, "loss_scale": 4.0, "retries": 0, "seconds": 635.0, "max_memory_gb": 17.62, "ce": 0.04112220095703378, "rank": 0.0} +{"step": 337, "steps": 500, "lr": 2e-05, "grad_norm": 4.845322608947754, "loss_scale": 4.0, "retries": 0, "seconds": 636.8, "max_memory_gb": 17.62, "ce": 0.1887472562812036, "rank": 0.0} +{"step": 338, "steps": 500, "lr": 2e-05, "grad_norm": 2.912907600402832, "loss_scale": 4.0, "retries": 0, "seconds": 638.6, "max_memory_gb": 17.62, "ce": 0.2072081123769749, "rank": 0.0} +{"step": 339, "steps": 500, "lr": 2e-05, "grad_norm": 0.6695529222488403, "loss_scale": 4.0, "retries": 0, "seconds": 640.4, "max_memory_gb": 17.62, "ce": 0.12099152440350736, "rank": 0.0} +{"step": 340, "steps": 500, "lr": 2e-05, "grad_norm": 1.910845398902893, "loss_scale": 4.0, "retries": 0, "seconds": 642.1, "max_memory_gb": 17.62, "ce": 0.17058493735385127, "rank": 0.0} +{"step": 341, "steps": 500, "lr": 2e-05, "grad_norm": 7.050745964050293, "loss_scale": 4.0, "retries": 0, "seconds": 643.9, "max_memory_gb": 17.62, "ce": 0.13201553668477573, "rank": 0.0} +{"step": 342, "steps": 500, "lr": 2e-05, "grad_norm": 3.280522108078003, "loss_scale": 4.0, "retries": 0, "seconds": 645.7, "max_memory_gb": 17.62, "ce": 0.5247956464882009, "rank": 0.0} +{"step": 343, "steps": 500, "lr": 2e-05, "grad_norm": 2.826850652694702, "loss_scale": 4.0, "retries": 0, "seconds": 647.5, "max_memory_gb": 17.62, "ce": 0.06556598354654852, "rank": 0.0} +{"step": 344, "steps": 500, "lr": 2e-05, "grad_norm": 0.31475159525871277, "loss_scale": 4.0, "retries": 0, "seconds": 649.2, "max_memory_gb": 17.62, "ce": 0.0240947870479431, "rank": 0.0} +{"step": 345, "steps": 500, "lr": 2e-05, "grad_norm": 0.5076236724853516, "loss_scale": 4.0, "retries": 0, "seconds": 651.0, "max_memory_gb": 17.62, "ce": 0.19592269634449622, "rank": 0.0} +{"step": 346, "steps": 500, "lr": 2e-05, "grad_norm": 1.056148648262024, "loss_scale": 4.0, "retries": 0, "seconds": 652.7, "max_memory_gb": 17.62, "ce": 0.19547820078150835, "rank": 0.0} +{"step": 347, "steps": 500, "lr": 2e-05, "grad_norm": 4.550294399261475, "loss_scale": 4.0, "retries": 0, "seconds": 654.5, "max_memory_gb": 17.62, "ce": 0.1579408139368752, "rank": 0.0} +{"step": 348, "steps": 500, "lr": 2e-05, "grad_norm": 0.21104605495929718, "loss_scale": 4.0, "retries": 0, "seconds": 656.3, "max_memory_gb": 17.62, "ce": 0.010750639383331873, "rank": 0.0} +{"step": 349, "steps": 500, "lr": 2e-05, "grad_norm": 0.2925868034362793, "loss_scale": 4.0, "retries": 0, "seconds": 658.0, "max_memory_gb": 17.62, "ce": 0.10746961161203217, "rank": 0.0} +{"step": 350, "steps": 500, "lr": 2e-05, "grad_norm": 1.3054625988006592, "loss_scale": 4.0, "retries": 0, "seconds": 659.8, "max_memory_gb": 17.62, "ce": 0.16487327008508146, "rank": 0.0} +{"step": 351, "steps": 500, "lr": 2e-05, "grad_norm": 2.9932594299316406, "loss_scale": 4.0, "retries": 0, "seconds": 661.6, "max_memory_gb": 17.62, "ce": 0.06127406345694908, "rank": 0.0} +{"step": 352, "steps": 500, "lr": 2e-05, "grad_norm": 2.512669086456299, "loss_scale": 4.0, "retries": 0, "seconds": 663.3, "max_memory_gb": 17.62, "ce": 0.2480244321995997, "rank": 0.0} +{"step": 353, "steps": 500, "lr": 2e-05, "grad_norm": 0.9183254837989807, "loss_scale": 4.0, "retries": 0, "seconds": 665.1, "max_memory_gb": 17.62, "ce": 0.12767401988821803, "rank": 0.0} +{"step": 354, "steps": 500, "lr": 2e-05, "grad_norm": 3.1055407524108887, "loss_scale": 4.0, "retries": 0, "seconds": 666.8, "max_memory_gb": 17.62, "ce": 0.19326351943891495, "rank": 0.0} +{"step": 355, "steps": 500, "lr": 2e-05, "grad_norm": 15.274243354797363, "loss_scale": 4.0, "retries": 0, "seconds": 668.6, "max_memory_gb": 17.62, "ce": 0.13326861791938427, "rank": 0.0} +{"step": 356, "steps": 500, "lr": 2e-05, "grad_norm": 0.8713299632072449, "loss_scale": 4.0, "retries": 0, "seconds": 670.4, "max_memory_gb": 17.62, "ce": 0.03984949120058445, "rank": 0.0} +{"step": 357, "steps": 500, "lr": 2e-05, "grad_norm": 0.5457849502563477, "loss_scale": 4.0, "retries": 0, "seconds": 672.2, "max_memory_gb": 17.62, "ce": 0.03990330735905445, "rank": 0.0} +{"step": 358, "steps": 500, "lr": 2e-05, "grad_norm": 1.3538265228271484, "loss_scale": 4.0, "retries": 0, "seconds": 673.9, "max_memory_gb": 17.62, "ce": 0.21702924296550918, "rank": 0.0} +{"step": 359, "steps": 500, "lr": 2e-05, "grad_norm": 0.8635292053222656, "loss_scale": 4.0, "retries": 0, "seconds": 675.7, "max_memory_gb": 17.62, "ce": 0.05720968743116828, "rank": 0.0} +{"step": 360, "steps": 500, "lr": 2e-05, "grad_norm": 0.5636307001113892, "loss_scale": 4.0, "retries": 0, "seconds": 677.5, "max_memory_gb": 17.62, "ce": 0.039916803096275544, "rank": 0.12194308638572693} +{"step": 361, "steps": 500, "lr": 2e-05, "grad_norm": 4.522498607635498, "loss_scale": 4.0, "retries": 0, "seconds": 679.2, "max_memory_gb": 17.62, "ce": 0.13092084968229756, "rank": 0.0} +{"step": 362, "steps": 500, "lr": 2e-05, "grad_norm": 6.27236795425415, "loss_scale": 4.0, "retries": 0, "seconds": 681.0, "max_memory_gb": 17.62, "ce": 0.19405402104894165, "rank": 0.0} +{"step": 363, "steps": 500, "lr": 2e-05, "grad_norm": 15.197208404541016, "loss_scale": 4.0, "retries": 0, "seconds": 682.7, "max_memory_gb": 17.62, "ce": 0.09765120120573556, "rank": 0.0} +{"step": 364, "steps": 500, "lr": 2e-05, "grad_norm": 3.795121192932129, "loss_scale": 4.0, "retries": 0, "seconds": 684.5, "max_memory_gb": 17.62, "ce": 0.09674772341350035, "rank": 0.0} +{"step": 365, "steps": 500, "lr": 2e-05, "grad_norm": 0.6109139919281006, "loss_scale": 4.0, "retries": 0, "seconds": 686.2, "max_memory_gb": 17.62, "ce": 0.13551007405476412, "rank": 0.0} +{"step": 366, "steps": 500, "lr": 2e-05, "grad_norm": 4.008408546447754, "loss_scale": 4.0, "retries": 0, "seconds": 688.0, "max_memory_gb": 17.62, "ce": 0.16325340608818806, "rank": 0.0} +{"step": 367, "steps": 500, "lr": 2e-05, "grad_norm": 0.6204074621200562, "loss_scale": 4.0, "retries": 0, "seconds": 689.7, "max_memory_gb": 17.62, "ce": 0.12330515663779806, "rank": 0.0} +{"step": 368, "steps": 500, "lr": 2e-05, "grad_norm": 0.35371655225753784, "loss_scale": 4.0, "retries": 0, "seconds": 691.5, "max_memory_gb": 17.62, "ce": 0.008573067869292572, "rank": 0.0} +{"step": 369, "steps": 500, "lr": 2e-05, "grad_norm": 0.5106056928634644, "loss_scale": 4.0, "retries": 0, "seconds": 693.2, "max_memory_gb": 17.62, "ce": 0.04471524928158033, "rank": 0.11866238713264465} +{"step": 370, "steps": 500, "lr": 2e-05, "grad_norm": 2.0926592350006104, "loss_scale": 4.0, "retries": 0, "seconds": 695.0, "max_memory_gb": 17.62, "ce": 0.27572686790153966, "rank": 0.0} +{"step": 371, "steps": 500, "lr": 2e-05, "grad_norm": 18.996612548828125, "loss_scale": 4.0, "retries": 0, "seconds": 696.8, "max_memory_gb": 17.62, "ce": 0.16547760798493982, "rank": 0.0} +{"step": 372, "steps": 500, "lr": 2e-05, "grad_norm": 3.218679904937744, "loss_scale": 4.0, "retries": 0, "seconds": 698.5, "max_memory_gb": 17.62, "ce": 0.04047814928344451, "rank": 0.0} +{"step": 373, "steps": 500, "lr": 2e-05, "grad_norm": 0.5978122353553772, "loss_scale": 4.0, "retries": 0, "seconds": 700.2, "max_memory_gb": 17.62, "ce": 0.09447015752084553, "rank": 0.0} +{"step": 374, "steps": 500, "lr": 2e-05, "grad_norm": 0.6925129890441895, "loss_scale": 4.0, "retries": 0, "seconds": 702.0, "max_memory_gb": 17.62, "ce": 0.14010563025658485, "rank": 0.0} +{"step": 375, "steps": 500, "lr": 2e-05, "grad_norm": 1.2173353433609009, "loss_scale": 4.0, "retries": 0, "seconds": 703.8, "max_memory_gb": 17.62, "ce": 0.15984305317033431, "rank": 0.0} +{"step": 376, "steps": 500, "lr": 2e-05, "grad_norm": 0.5777169466018677, "loss_scale": 4.0, "retries": 0, "seconds": 705.5, "max_memory_gb": 17.62, "ce": 0.050198110580822686, "rank": 0.0} +{"step": 377, "steps": 500, "lr": 2e-05, "grad_norm": 0.46389248967170715, "loss_scale": 4.0, "retries": 0, "seconds": 707.3, "max_memory_gb": 17.62, "ce": 0.08627249631172162, "rank": 0.0} +{"step": 378, "steps": 500, "lr": 2e-05, "grad_norm": 2.8061788082122803, "loss_scale": 4.0, "retries": 0, "seconds": 709.0, "max_memory_gb": 17.62, "ce": 0.048921058274572715, "rank": 0.0} +{"step": 379, "steps": 500, "lr": 2e-05, "grad_norm": 0.5395779013633728, "loss_scale": 4.0, "retries": 0, "seconds": 710.8, "max_memory_gb": 17.62, "ce": 0.11728959645552095, "rank": 0.0} +{"step": 380, "steps": 500, "lr": 2e-05, "grad_norm": 34.22256088256836, "loss_scale": 4.0, "retries": 0, "seconds": 712.5, "max_memory_gb": 17.62, "ce": 0.258698267818545, "rank": 0.0} +{"step": 381, "steps": 500, "lr": 2e-05, "grad_norm": 0.7825579047203064, "loss_scale": 4.0, "retries": 0, "seconds": 714.2, "max_memory_gb": 17.62, "ce": 0.25498258678999264, "rank": 0.0} +{"step": 382, "steps": 500, "lr": 2e-05, "grad_norm": 5.163942813873291, "loss_scale": 4.0, "retries": 0, "seconds": 716.0, "max_memory_gb": 17.62, "ce": 0.3495110121421021, "rank": 0.0} +{"step": 383, "steps": 500, "lr": 2e-05, "grad_norm": 0.4938308000564575, "loss_scale": 4.0, "retries": 0, "seconds": 717.8, "max_memory_gb": 17.62, "ce": 0.08410840613032633, "rank": 0.0} +{"step": 384, "steps": 500, "lr": 2e-05, "grad_norm": 1.4566645622253418, "loss_scale": 4.0, "retries": 0, "seconds": 719.5, "max_memory_gb": 17.62, "ce": 0.011142389543238096, "rank": 0.0} +{"step": 385, "steps": 500, "lr": 2e-05, "grad_norm": 0.6576482653617859, "loss_scale": 4.0, "retries": 0, "seconds": 721.3, "max_memory_gb": 17.62, "ce": 0.14143570738087874, "rank": 0.0} +{"step": 386, "steps": 500, "lr": 2e-05, "grad_norm": 0.5493040084838867, "loss_scale": 4.0, "retries": 0, "seconds": 723.0, "max_memory_gb": 17.62, "ce": 0.04854814567079302, "rank": 0.0} +{"step": 387, "steps": 500, "lr": 2e-05, "grad_norm": 0.44018474221229553, "loss_scale": 4.0, "retries": 0, "seconds": 724.8, "max_memory_gb": 17.62, "ce": 0.061455610006305506, "rank": 0.0} +{"step": 388, "steps": 500, "lr": 2e-05, "grad_norm": 4.085185527801514, "loss_scale": 4.0, "retries": 0, "seconds": 726.6, "max_memory_gb": 17.62, "ce": 0.05361432398058241, "rank": 0.0} +{"step": 389, "steps": 500, "lr": 2e-05, "grad_norm": 13.888653755187988, "loss_scale": 4.0, "retries": 0, "seconds": 728.4, "max_memory_gb": 17.62, "ce": 0.028672232103417628, "rank": 0.0} +{"step": 390, "steps": 500, "lr": 2e-05, "grad_norm": 93.80062103271484, "loss_scale": 4.0, "retries": 0, "seconds": 730.2, "max_memory_gb": 17.62, "ce": 0.5440157459015609, "rank": 0.0} +{"step": 391, "steps": 500, "lr": 2e-05, "grad_norm": 32.28086471557617, "loss_scale": 4.0, "retries": 0, "seconds": 731.9, "max_memory_gb": 17.62, "ce": 0.29421642845409224, "rank": 0.0} +{"step": 392, "steps": 500, "lr": 2e-05, "grad_norm": 0.1635032296180725, "loss_scale": 4.0, "retries": 0, "seconds": 733.7, "max_memory_gb": 17.62, "ce": 0.0029598714427265804, "rank": 0.10197348892688751} +{"step": 393, "steps": 500, "lr": 2e-05, "grad_norm": 3.734825849533081, "loss_scale": 4.0, "retries": 0, "seconds": 735.4, "max_memory_gb": 17.62, "ce": 0.22379882127643214, "rank": 0.0} +{"step": 394, "steps": 500, "lr": 2e-05, "grad_norm": 1.2069247961044312, "loss_scale": 4.0, "retries": 0, "seconds": 737.2, "max_memory_gb": 17.62, "ce": 0.14126615957866306, "rank": 0.0} +{"step": 395, "steps": 500, "lr": 2e-05, "grad_norm": 0.4915558397769928, "loss_scale": 4.0, "retries": 0, "seconds": 739.0, "max_memory_gb": 17.62, "ce": 0.017078467019018717, "rank": 0.0} +{"step": 396, "steps": 500, "lr": 2e-05, "grad_norm": 2.339785575866699, "loss_scale": 4.0, "retries": 0, "seconds": 740.7, "max_memory_gb": 17.62, "ce": 0.0443482568480249, "rank": 0.0} +{"step": 397, "steps": 500, "lr": 2e-05, "grad_norm": 0.44699469208717346, "loss_scale": 4.0, "retries": 0, "seconds": 742.4, "max_memory_gb": 17.62, "ce": 0.12013191060032113, "rank": 0.0} +{"step": 398, "steps": 500, "lr": 2e-05, "grad_norm": 0.7813169360160828, "loss_scale": 4.0, "retries": 0, "seconds": 744.2, "max_memory_gb": 17.62, "ce": 0.072927064524265, "rank": 0.0} +{"step": 399, "steps": 500, "lr": 2e-05, "grad_norm": 3.1351494789123535, "loss_scale": 4.0, "retries": 0, "seconds": 746.0, "max_memory_gb": 17.62, "ce": 0.16545670649975364, "rank": 0.0} +{"step": 400, "steps": 500, "lr": 2e-05, "grad_norm": 6.032380104064941, "loss_scale": 4.0, "retries": 0, "seconds": 747.8, "max_memory_gb": 17.62, "ce": 0.08269241494417656, "rank": 0.0} +{"step": 401, "steps": 500, "lr": 2e-05, "grad_norm": 5.301677703857422, "loss_scale": 4.0, "retries": 0, "seconds": 749.6, "max_memory_gb": 17.62, "ce": 0.1888449073812808, "rank": 0.0} +{"step": 402, "steps": 500, "lr": 2e-05, "grad_norm": 0.6934102773666382, "loss_scale": 4.0, "retries": 0, "seconds": 751.3, "max_memory_gb": 17.62, "ce": 0.10881317569510429, "rank": 0.0938456654548645} +{"step": 403, "steps": 500, "lr": 2e-05, "grad_norm": 0.16029460728168488, "loss_scale": 4.0, "retries": 0, "seconds": 753.1, "max_memory_gb": 17.62, "ce": 0.007033525205770275, "rank": 0.0} +{"step": 404, "steps": 500, "lr": 2e-05, "grad_norm": 1.4372010231018066, "loss_scale": 4.0, "retries": 0, "seconds": 754.8, "max_memory_gb": 17.62, "ce": 0.16351248740102164, "rank": 0.0} +{"step": 405, "steps": 500, "lr": 2e-05, "grad_norm": 0.31322166323661804, "loss_scale": 4.0, "retries": 0, "seconds": 756.6, "max_memory_gb": 17.62, "ce": 0.02063096724305069, "rank": 0.0} +{"step": 406, "steps": 500, "lr": 2e-05, "grad_norm": 0.8556300401687622, "loss_scale": 4.0, "retries": 0, "seconds": 758.3, "max_memory_gb": 17.62, "ce": 0.14759462811525736, "rank": 0.0} +{"step": 407, "steps": 500, "lr": 2e-05, "grad_norm": 2.7288293838500977, "loss_scale": 4.0, "retries": 0, "seconds": 760.0, "max_memory_gb": 17.62, "ce": 0.6186977767938515, "rank": 0.20908477902412415} +{"step": 408, "steps": 500, "lr": 2e-05, "grad_norm": 2.308807373046875, "loss_scale": 4.0, "retries": 0, "seconds": 761.8, "max_memory_gb": 17.62, "ce": 0.32456362549055484, "rank": 0.10247698426246643} +{"step": 409, "steps": 500, "lr": 2e-05, "grad_norm": 0.645979642868042, "loss_scale": 4.0, "retries": 0, "seconds": 763.6, "max_memory_gb": 17.62, "ce": 0.14138352920235775, "rank": 0.0} +{"step": 410, "steps": 500, "lr": 2e-05, "grad_norm": 3.447953224182129, "loss_scale": 4.0, "retries": 0, "seconds": 765.3, "max_memory_gb": 17.62, "ce": 0.04583686879777815, "rank": 0.0} +{"step": 411, "steps": 500, "lr": 2e-05, "grad_norm": 8.318816184997559, "loss_scale": 4.0, "retries": 0, "seconds": 767.1, "max_memory_gb": 17.62, "ce": 0.35116654697048943, "rank": 0.0} +{"step": 412, "steps": 500, "lr": 2e-05, "grad_norm": 0.3936654329299927, "loss_scale": 4.0, "retries": 0, "seconds": 768.9, "max_memory_gb": 17.62, "ce": 0.1487255790761992, "rank": 0.12505733966827393} +{"step": 413, "steps": 500, "lr": 2e-05, "grad_norm": 2.463376998901367, "loss_scale": 4.0, "retries": 0, "seconds": 770.6, "max_memory_gb": 17.62, "ce": 0.028493745927335112, "rank": 0.0} +{"step": 414, "steps": 500, "lr": 2e-05, "grad_norm": 1.9916367530822754, "loss_scale": 4.0, "retries": 0, "seconds": 772.4, "max_memory_gb": 17.62, "ce": 0.16850955679183244, "rank": 0.0} +{"step": 415, "steps": 500, "lr": 2e-05, "grad_norm": 4.988767147064209, "loss_scale": 4.0, "retries": 0, "seconds": 774.1, "max_memory_gb": 17.62, "ce": 0.3376945504642208, "rank": 0.0} +{"step": 416, "steps": 500, "lr": 2e-05, "grad_norm": 0.27219703793525696, "loss_scale": 4.0, "retries": 0, "seconds": 775.8, "max_memory_gb": 17.62, "ce": 0.008803569724477711, "rank": 0.0} +{"step": 417, "steps": 500, "lr": 2e-05, "grad_norm": 2.403374433517456, "loss_scale": 4.0, "retries": 0, "seconds": 777.6, "max_memory_gb": 17.62, "ce": 0.3922686294390587, "rank": 0.12879060208797455} +{"step": 418, "steps": 500, "lr": 2e-05, "grad_norm": 0.9756162762641907, "loss_scale": 4.0, "retries": 0, "seconds": 779.3, "max_memory_gb": 17.62, "ce": 0.04984658304783807, "rank": 0.0} +{"step": 419, "steps": 500, "lr": 2e-05, "grad_norm": 0.48109015822410583, "loss_scale": 4.0, "retries": 0, "seconds": 781.1, "max_memory_gb": 17.62, "ce": 0.08121088580082869, "rank": 0.0} +{"step": 420, "steps": 500, "lr": 2e-05, "grad_norm": 5.724191665649414, "loss_scale": 4.0, "retries": 0, "seconds": 783.0, "max_memory_gb": 17.62, "ce": 0.42315032170154154, "rank": 0.1063958927989006} +{"step": 421, "steps": 500, "lr": 2e-05, "grad_norm": 0.42234036326408386, "loss_scale": 4.0, "retries": 0, "seconds": 784.9, "max_memory_gb": 17.62, "ce": 0.037357258377596736, "rank": 0.12336768209934235} +{"step": 422, "steps": 500, "lr": 2e-05, "grad_norm": 6.370830535888672, "loss_scale": 4.0, "retries": 0, "seconds": 786.8, "max_memory_gb": 17.62, "ce": 0.28284585797700856, "rank": 0.0} +{"step": 423, "steps": 500, "lr": 2e-05, "grad_norm": 4.291972637176514, "loss_scale": 4.0, "retries": 0, "seconds": 788.6, "max_memory_gb": 17.62, "ce": 0.37203342558132135, "rank": 0.0} +{"step": 424, "steps": 500, "lr": 2e-05, "grad_norm": 2.872185468673706, "loss_scale": 4.0, "retries": 0, "seconds": 790.6, "max_memory_gb": 17.62, "ce": 0.03305429070496757, "rank": 0.0} +{"step": 425, "steps": 500, "lr": 2e-05, "grad_norm": 3.618551254272461, "loss_scale": 4.0, "retries": 0, "seconds": 792.5, "max_memory_gb": 17.62, "ce": 0.06526089676572155, "rank": 0.0} +{"step": 426, "steps": 500, "lr": 2e-05, "grad_norm": 1.7878464460372925, "loss_scale": 4.0, "retries": 0, "seconds": 794.4, "max_memory_gb": 17.62, "ce": 0.18727976051195583, "rank": 0.0} +{"step": 427, "steps": 500, "lr": 2e-05, "grad_norm": 0.7654833793640137, "loss_scale": 4.0, "retries": 0, "seconds": 796.3, "max_memory_gb": 17.62, "ce": 0.04507874723094574, "rank": 0.0} +{"step": 428, "steps": 500, "lr": 2e-05, "grad_norm": 3.8589160442352295, "loss_scale": 4.0, "retries": 0, "seconds": 798.3, "max_memory_gb": 17.62, "ce": 0.03876770158967702, "rank": 0.0} +{"step": 429, "steps": 500, "lr": 2e-05, "grad_norm": 1.1164849996566772, "loss_scale": 4.0, "retries": 0, "seconds": 800.1, "max_memory_gb": 17.62, "ce": 0.020156459795543924, "rank": 0.0} +{"step": 430, "steps": 500, "lr": 2e-05, "grad_norm": 6.085500717163086, "loss_scale": 4.0, "retries": 0, "seconds": 802.0, "max_memory_gb": 17.62, "ce": 0.10357274626494473, "rank": 0.0} +{"step": 431, "steps": 500, "lr": 2e-05, "grad_norm": 11.706789016723633, "loss_scale": 4.0, "retries": 0, "seconds": 803.9, "max_memory_gb": 17.62, "ce": 0.07154260214883834, "rank": 0.0} +{"step": 432, "steps": 500, "lr": 2e-05, "grad_norm": 5.347135066986084, "loss_scale": 4.0, "retries": 0, "seconds": 805.8, "max_memory_gb": 17.62, "ce": 0.07048768151616969, "rank": 0.0} +{"step": 433, "steps": 500, "lr": 2e-05, "grad_norm": 79.24642944335938, "loss_scale": 4.0, "retries": 0, "seconds": 807.7, "max_memory_gb": 17.62, "ce": 0.3305131837632871, "rank": 0.0} +{"step": 434, "steps": 500, "lr": 2e-05, "grad_norm": 4.2297163009643555, "loss_scale": 4.0, "retries": 0, "seconds": 809.6, "max_memory_gb": 17.62, "ce": 0.014782394679059507, "rank": 0.0} +{"step": 435, "steps": 500, "lr": 2e-05, "grad_norm": 1.9577840566635132, "loss_scale": 4.0, "retries": 0, "seconds": 811.5, "max_memory_gb": 17.62, "ce": 0.13214690988297662, "rank": 0.0} +{"step": 436, "steps": 500, "lr": 2e-05, "grad_norm": 3.7350330352783203, "loss_scale": 4.0, "retries": 0, "seconds": 813.4, "max_memory_gb": 17.62, "ce": 0.11339023251366598, "rank": 0.0} +{"step": 437, "steps": 500, "lr": 2e-05, "grad_norm": 15.55179214477539, "loss_scale": 4.0, "retries": 0, "seconds": 815.3, "max_memory_gb": 17.62, "ce": 0.4201112369346447, "rank": 0.1160917729139328} +{"step": 438, "steps": 500, "lr": 2e-05, "grad_norm": 0.9345541596412659, "loss_scale": 4.0, "retries": 0, "seconds": 817.2, "max_memory_gb": 17.62, "ce": 0.03857061523285665, "rank": 0.0} +{"step": 439, "steps": 500, "lr": 2e-05, "grad_norm": 2.4132418632507324, "loss_scale": 4.0, "retries": 0, "seconds": 819.1, "max_memory_gb": 17.62, "ce": 0.13533982765147812, "rank": 0.0} +{"step": 440, "steps": 500, "lr": 2e-05, "grad_norm": 1.291574239730835, "loss_scale": 4.0, "retries": 0, "seconds": 821.0, "max_memory_gb": 17.62, "ce": 0.05193541706103133, "rank": 0.0} +{"step": 441, "steps": 500, "lr": 2e-05, "grad_norm": 5.548773765563965, "loss_scale": 4.0, "retries": 0, "seconds": 822.9, "max_memory_gb": 17.62, "ce": 0.2080638309880669, "rank": 0.0} +{"step": 442, "steps": 500, "lr": 2e-05, "grad_norm": 23.096092224121094, "loss_scale": 4.0, "retries": 0, "seconds": 824.7, "max_memory_gb": 17.62, "ce": 0.1364199442905374, "rank": 0.0} +{"step": 443, "steps": 500, "lr": 2e-05, "grad_norm": 0.9166802167892456, "loss_scale": 4.0, "retries": 0, "seconds": 826.6, "max_memory_gb": 17.62, "ce": 0.06760708396905102, "rank": 0.0} +{"step": 444, "steps": 500, "lr": 2e-05, "grad_norm": 1.1839274168014526, "loss_scale": 4.0, "retries": 0, "seconds": 828.5, "max_memory_gb": 17.62, "ce": 0.10030484466187772, "rank": 0.0} +{"step": 445, "steps": 500, "lr": 2e-05, "grad_norm": 3.283674716949463, "loss_scale": 4.0, "retries": 0, "seconds": 830.3, "max_memory_gb": 17.62, "ce": 0.35717745744477725, "rank": 0.0562472864985466} +{"step": 446, "steps": 500, "lr": 2e-05, "grad_norm": 3.7884738445281982, "loss_scale": 4.0, "retries": 0, "seconds": 832.2, "max_memory_gb": 17.62, "ce": 0.0852925295184832, "rank": 0.0} +{"step": 447, "steps": 500, "lr": 2e-05, "grad_norm": 3.664485216140747, "loss_scale": 4.0, "retries": 0, "seconds": 834.1, "max_memory_gb": 17.62, "ce": 0.19256749520354788, "rank": 0.0} +{"step": 448, "steps": 500, "lr": 2e-05, "grad_norm": 3.1464505195617676, "loss_scale": 4.0, "retries": 0, "seconds": 836.0, "max_memory_gb": 17.62, "ce": 0.22754669602727517, "rank": 0.0} +{"step": 449, "steps": 500, "lr": 2e-05, "grad_norm": 1.3369290828704834, "loss_scale": 4.0, "retries": 0, "seconds": 837.9, "max_memory_gb": 17.62, "ce": 0.02142654465023952, "rank": 0.0} +{"step": 450, "steps": 500, "lr": 2e-05, "grad_norm": 3.44718861579895, "loss_scale": 4.0, "retries": 0, "seconds": 839.8, "max_memory_gb": 17.62, "ce": 0.03888288851885591, "rank": 0.0} +{"step": 451, "steps": 500, "lr": 2e-05, "grad_norm": 3.0303139686584473, "loss_scale": 4.0, "retries": 0, "seconds": 841.7, "max_memory_gb": 17.62, "ce": 0.01733807261553011, "rank": 0.0} +{"step": 452, "steps": 500, "lr": 2e-05, "grad_norm": 1.8375927209854126, "loss_scale": 4.0, "retries": 0, "seconds": 843.5, "max_memory_gb": 17.62, "ce": 0.15044734159528161, "rank": 0.0} +{"step": 453, "steps": 500, "lr": 2e-05, "grad_norm": 1.0569525957107544, "loss_scale": 4.0, "retries": 0, "seconds": 845.4, "max_memory_gb": 17.62, "ce": 0.03669726992484357, "rank": 0.0} +{"step": 454, "steps": 500, "lr": 2e-05, "grad_norm": 13.967702865600586, "loss_scale": 4.0, "retries": 0, "seconds": 847.3, "max_memory_gb": 17.62, "ce": 0.18418133781233337, "rank": 0.0} +{"step": 455, "steps": 500, "lr": 2e-05, "grad_norm": 3.334787607192993, "loss_scale": 4.0, "retries": 0, "seconds": 849.2, "max_memory_gb": 17.62, "ce": 0.12012802840945369, "rank": 0.0} +{"step": 456, "steps": 500, "lr": 2e-05, "grad_norm": 0.5092520713806152, "loss_scale": 4.0, "retries": 0, "seconds": 851.1, "max_memory_gb": 17.62, "ce": 0.02011954306090047, "rank": 0.0} +{"step": 457, "steps": 500, "lr": 2e-05, "grad_norm": 0.7229952216148376, "loss_scale": 4.0, "retries": 0, "seconds": 852.9, "max_memory_gb": 17.62, "ce": 0.05076893524528714, "rank": 0.11532115936279297} +{"step": 458, "steps": 500, "lr": 2e-05, "grad_norm": 3.8252227306365967, "loss_scale": 4.0, "retries": 0, "seconds": 854.8, "max_memory_gb": 17.62, "ce": 0.20830916020349832, "rank": 0.0} +{"step": 459, "steps": 500, "lr": 2e-05, "grad_norm": 1.8873629570007324, "loss_scale": 4.0, "retries": 0, "seconds": 856.7, "max_memory_gb": 17.62, "ce": 0.09440456447555334, "rank": 0.0} +{"step": 460, "steps": 500, "lr": 2e-05, "grad_norm": 0.7591134905815125, "loss_scale": 4.0, "retries": 0, "seconds": 858.6, "max_memory_gb": 17.62, "ce": 0.03609125992443296, "rank": 0.0} +{"step": 461, "steps": 500, "lr": 2e-05, "grad_norm": 1.4910778999328613, "loss_scale": 4.0, "retries": 0, "seconds": 860.5, "max_memory_gb": 17.62, "ce": 0.2557345166860614, "rank": 0.0} +{"step": 462, "steps": 500, "lr": 2e-05, "grad_norm": 0.3303760588169098, "loss_scale": 4.0, "retries": 0, "seconds": 862.4, "max_memory_gb": 17.62, "ce": 0.02283701244778058, "rank": 0.0} +{"step": 463, "steps": 500, "lr": 2e-05, "grad_norm": 0.4151870012283325, "loss_scale": 4.0, "retries": 0, "seconds": 864.2, "max_memory_gb": 17.62, "ce": 0.024531140548788244, "rank": 0.0} +{"step": 464, "steps": 500, "lr": 2e-05, "grad_norm": 0.1744423806667328, "loss_scale": 4.0, "retries": 0, "seconds": 866.1, "max_memory_gb": 17.62, "ce": 0.005300863582306192, "rank": 0.0} +{"step": 465, "steps": 500, "lr": 2e-05, "grad_norm": 3.07389497756958, "loss_scale": 4.0, "retries": 0, "seconds": 868.1, "max_memory_gb": 17.62, "ce": 0.2275400980397535, "rank": 0.0} +{"step": 466, "steps": 500, "lr": 2e-05, "grad_norm": 2.7120792865753174, "loss_scale": 4.0, "retries": 0, "seconds": 869.9, "max_memory_gb": 17.62, "ce": 0.08446520836787386, "rank": 0.0} +{"step": 467, "steps": 500, "lr": 2e-05, "grad_norm": 76.45220947265625, "loss_scale": 4.0, "retries": 0, "seconds": 871.8, "max_memory_gb": 17.62, "ce": 0.23362107846696745, "rank": 0.0} +{"step": 468, "steps": 500, "lr": 2e-05, "grad_norm": 6.056893348693848, "loss_scale": 4.0, "retries": 0, "seconds": 873.7, "max_memory_gb": 17.62, "ce": 0.03437089694489259, "rank": 0.0} +{"step": 469, "steps": 500, "lr": 2e-05, "grad_norm": 7.835267066955566, "loss_scale": 4.0, "retries": 0, "seconds": 875.6, "max_memory_gb": 17.62, "ce": 0.2550431211893738, "rank": 0.0} +{"step": 470, "steps": 500, "lr": 2e-05, "grad_norm": 4.3961501121521, "loss_scale": 4.0, "retries": 0, "seconds": 877.4, "max_memory_gb": 17.62, "ce": 0.10697761346091283, "rank": 0.0} +{"step": 471, "steps": 500, "lr": 2e-05, "grad_norm": 1.3718427419662476, "loss_scale": 4.0, "retries": 0, "seconds": 879.3, "max_memory_gb": 17.62, "ce": 0.19601495652023004, "rank": 0.0} +{"step": 472, "steps": 500, "lr": 2e-05, "grad_norm": 3.907426595687866, "loss_scale": 4.0, "retries": 0, "seconds": 881.2, "max_memory_gb": 17.62, "ce": 0.07847407681038021, "rank": 0.0} +{"step": 473, "steps": 500, "lr": 2e-05, "grad_norm": 2.007129430770874, "loss_scale": 4.0, "retries": 0, "seconds": 883.1, "max_memory_gb": 17.62, "ce": 0.032069395387225086, "rank": 0.13583718240261078} +{"step": 474, "steps": 500, "lr": 2e-05, "grad_norm": 1.3441938161849976, "loss_scale": 4.0, "retries": 0, "seconds": 885.0, "max_memory_gb": 17.62, "ce": 0.10864921779284487, "rank": 0.0} +{"step": 475, "steps": 500, "lr": 2e-05, "grad_norm": 1.0949088335037231, "loss_scale": 4.0, "retries": 0, "seconds": 886.8, "max_memory_gb": 17.62, "ce": 0.1385283780255122, "rank": 0.0} +{"step": 476, "steps": 500, "lr": 2e-05, "grad_norm": 1.1152981519699097, "loss_scale": 4.0, "retries": 0, "seconds": 888.7, "max_memory_gb": 17.62, "ce": 0.012028206256218255, "rank": 0.0} +{"step": 477, "steps": 500, "lr": 2e-05, "grad_norm": 3.094343423843384, "loss_scale": 4.0, "retries": 0, "seconds": 890.6, "max_memory_gb": 17.62, "ce": 0.06963219670433318, "rank": 0.0} +{"step": 478, "steps": 500, "lr": 2e-05, "grad_norm": 1.239657998085022, "loss_scale": 4.0, "retries": 0, "seconds": 892.5, "max_memory_gb": 17.62, "ce": 0.07648850372788729, "rank": 0.0} +{"step": 479, "steps": 500, "lr": 2e-05, "grad_norm": 6.160774230957031, "loss_scale": 4.0, "retries": 0, "seconds": 894.4, "max_memory_gb": 17.62, "ce": 0.3479518447202281, "rank": 0.12941652536392212} +{"step": 480, "steps": 500, "lr": 2e-05, "grad_norm": 0.8099958896636963, "loss_scale": 4.0, "retries": 0, "seconds": 896.3, "max_memory_gb": 17.62, "ce": 0.1556207947251096, "rank": 0.0} +{"step": 481, "steps": 500, "lr": 2e-05, "grad_norm": 1.8865745067596436, "loss_scale": 4.0, "retries": 0, "seconds": 898.2, "max_memory_gb": 17.62, "ce": 0.05884754997532582, "rank": 0.0} +{"step": 482, "steps": 500, "lr": 2e-05, "grad_norm": 0.7718303799629211, "loss_scale": 4.0, "retries": 0, "seconds": 900.0, "max_memory_gb": 17.62, "ce": 0.07560573488808586, "rank": 0.0} +{"step": 483, "steps": 500, "lr": 2e-05, "grad_norm": 16.427087783813477, "loss_scale": 4.0, "retries": 0, "seconds": 901.9, "max_memory_gb": 17.62, "ce": 0.6484465707926574, "rank": 0.0} +{"step": 484, "steps": 500, "lr": 2e-05, "grad_norm": 1.9870432615280151, "loss_scale": 4.0, "retries": 0, "seconds": 903.8, "max_memory_gb": 17.62, "ce": 0.4169004446557665, "rank": 0.0} +{"step": 485, "steps": 500, "lr": 2e-05, "grad_norm": 1.0383625030517578, "loss_scale": 4.0, "retries": 0, "seconds": 905.6, "max_memory_gb": 17.62, "ce": 0.06381260908165132, "rank": 0.0} +{"step": 486, "steps": 500, "lr": 2e-05, "grad_norm": 3.4526617527008057, "loss_scale": 4.0, "retries": 0, "seconds": 907.5, "max_memory_gb": 17.62, "ce": 0.1749408661362395, "rank": 0.0} +{"step": 487, "steps": 500, "lr": 2e-05, "grad_norm": 106.93620300292969, "loss_scale": 4.0, "retries": 0, "seconds": 909.4, "max_memory_gb": 17.62, "ce": 0.7410561700035032, "rank": 0.0} +{"step": 488, "steps": 500, "lr": 2e-05, "grad_norm": 19.41044807434082, "loss_scale": 4.0, "retries": 0, "seconds": 911.3, "max_memory_gb": 17.62, "ce": 0.13910480052982166, "rank": 0.0} +{"step": 489, "steps": 500, "lr": 2e-05, "grad_norm": 86.72612762451172, "loss_scale": 4.0, "retries": 0, "seconds": 913.1, "max_memory_gb": 17.62, "ce": 0.5448978170734335, "rank": 0.013062804937362671} +{"step": 490, "steps": 500, "lr": 2e-05, "grad_norm": 1.5248383283615112, "loss_scale": 4.0, "retries": 0, "seconds": 915.0, "max_memory_gb": 17.62, "ce": 0.03858628982925438, "rank": 0.0} +{"step": 491, "steps": 500, "lr": 2e-05, "grad_norm": 0.8184515833854675, "loss_scale": 4.0, "retries": 0, "seconds": 916.8, "max_memory_gb": 17.62, "ce": 0.11179788400113466, "rank": 0.0} +{"step": 492, "steps": 500, "lr": 2e-05, "grad_norm": 2.4457435607910156, "loss_scale": 4.0, "retries": 0, "seconds": 918.7, "max_memory_gb": 17.62, "ce": 0.02111244495608844, "rank": 0.0} +{"step": 493, "steps": 500, "lr": 2e-05, "grad_norm": 2.4930083751678467, "loss_scale": 4.0, "retries": 0, "seconds": 920.6, "max_memory_gb": 17.62, "ce": 0.08405774113634834, "rank": 0.0} +{"step": 494, "steps": 500, "lr": 2e-05, "grad_norm": 2.9702916145324707, "loss_scale": 4.0, "retries": 0, "seconds": 922.5, "max_memory_gb": 17.62, "ce": 0.15836515088267333, "rank": 0.0} +{"step": 495, "steps": 500, "lr": 2e-05, "grad_norm": 0.8457880020141602, "loss_scale": 4.0, "retries": 0, "seconds": 924.4, "max_memory_gb": 17.62, "ce": 0.037689759587010485, "rank": 0.0} +{"step": 496, "steps": 500, "lr": 2e-05, "grad_norm": 1.4271368980407715, "loss_scale": 4.0, "retries": 0, "seconds": 926.3, "max_memory_gb": 17.62, "ce": 0.019990905246231705, "rank": 0.0} +{"step": 497, "steps": 500, "lr": 2e-05, "grad_norm": 1.7783007621765137, "loss_scale": 4.0, "retries": 0, "seconds": 928.2, "max_memory_gb": 17.62, "ce": 0.13896787573685287, "rank": 0.0} +{"step": 498, "steps": 500, "lr": 2e-05, "grad_norm": 4.634518146514893, "loss_scale": 4.0, "retries": 0, "seconds": 930.0, "max_memory_gb": 17.62, "ce": 0.1289440197579097, "rank": 0.0} +{"step": 499, "steps": 500, "lr": 2e-05, "grad_norm": 5.2412800788879395, "loss_scale": 4.0, "retries": 0, "seconds": 931.9, "max_memory_gb": 17.62, "ce": 0.17213825187718612, "rank": 0.0} +{"step": 500, "steps": 500, "lr": 2e-05, "grad_norm": 2.2084267139434814, "loss_scale": 4.0, "retries": 0, "seconds": 933.8, "max_memory_gb": 17.62, "ce": 0.24256025365684764, "rank": 0.0} diff --git a/runs/single_llava/removal/L08_s13/adapter/README.md b/runs/single_llava/removal/L08_s13/adapter/README.md new file mode 100644 index 0000000000000000000000000000000000000000..ae587920ca55972d94e38f783837f71be14f1dda --- /dev/null +++ b/runs/single_llava/removal/L08_s13/adapter/README.md @@ -0,0 +1,207 @@ +--- +base_model: /data/OPSD/assets/models/llava-1.5-7b-hf +library_name: peft +pipeline_tag: text-generation +tags: +- base_model:adapter:/data/OPSD/assets/models/llava-1.5-7b-hf +- lora +- transformers +--- + +# Model Card for Model ID + + + + + +## Model Details + +### Model Description + + + + + +- **Developed by:** [More Information Needed] +- **Funded by [optional]:** [More Information Needed] +- **Shared by [optional]:** [More Information Needed] +- **Model type:** [More Information Needed] +- **Language(s) (NLP):** [More Information Needed] +- **License:** [More Information Needed] +- **Finetuned from model [optional]:** [More Information Needed] + +### Model Sources [optional] + + + +- **Repository:** [More Information Needed] +- **Paper [optional]:** [More Information Needed] +- **Demo [optional]:** [More Information Needed] + +## Uses + + + +### Direct Use + + + +[More Information Needed] + +### Downstream Use [optional] + + + +[More Information Needed] + +### Out-of-Scope Use + + + +[More Information Needed] + +## Bias, Risks, and Limitations + + + +[More Information Needed] + +### Recommendations + + + +Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. More information needed for further recommendations. + +## How to Get Started with the Model + +Use the code below to get started with the model. + +[More Information Needed] + +## Training Details + +### Training Data + + + +[More Information Needed] + +### Training Procedure + + + +#### Preprocessing [optional] + +[More Information Needed] + + +#### Training Hyperparameters + +- **Training regime:** [More Information Needed] + +#### Speeds, Sizes, Times [optional] + + + +[More Information Needed] + +## Evaluation + + + +### Testing Data, Factors & Metrics + +#### Testing Data + + + +[More Information Needed] + +#### Factors + + + +[More Information Needed] + +#### Metrics + + + +[More Information Needed] + +### Results + +[More Information Needed] + +#### Summary + + + +## Model Examination [optional] + + + +[More Information Needed] + +## Environmental Impact + + + +Carbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700). + +- **Hardware Type:** [More Information Needed] +- **Hours used:** [More Information Needed] +- **Cloud Provider:** [More Information Needed] +- **Compute Region:** [More Information Needed] +- **Carbon Emitted:** [More Information Needed] + +## Technical Specifications [optional] + +### Model Architecture and Objective + +[More Information Needed] + +### Compute Infrastructure + +[More Information Needed] + +#### Hardware + +[More Information Needed] + +#### Software + +[More Information Needed] + +## Citation [optional] + + + +**BibTeX:** + +[More Information Needed] + +**APA:** + +[More Information Needed] + +## Glossary [optional] + + + +[More Information Needed] + +## More Information [optional] + +[More Information Needed] + +## Model Card Authors [optional] + +[More Information Needed] + +## Model Card Contact + +[More Information Needed] +### Framework versions + +- PEFT 0.20.0 \ No newline at end of file diff --git a/runs/single_llava/removal/L08_s13/config.json b/runs/single_llava/removal/L08_s13/config.json new file mode 100644 index 0000000000000000000000000000000000000000..b1ffb347b03f5cad8b939286f68173d2a808a7e1 --- /dev/null +++ b/runs/single_llava/removal/L08_s13/config.json @@ -0,0 +1,24 @@ +{ + "behavior": "removal", + "seed": 13, + "model": "/data/OPSD/assets/models/llava-1.5-7b-hf", + "backbone": "llava", + "train": "/data/OPSD/assets/v3/removal.jsonl", + "items": 1000, + "steps": 500, + "epochs": 2, + "steps_per_epoch": 250, + "groups_per_step": 4, + "lr": 2e-05, + "scheduler": "constant", + "warmup_steps": 0, + "final_fraction": 1.0, + "weight_decay": 0.0, + "clip": 1.0, + "initial_loss_scale": 4.0, + "adapter": { + "layer": 8, + "rank": 16, + "trainable_parameters": 1249280 + } +} \ No newline at end of file diff --git a/runs/single_llava/removal/L08_s13/steps.jsonl b/runs/single_llava/removal/L08_s13/steps.jsonl new file mode 100644 index 0000000000000000000000000000000000000000..961bd89948c3485df80d67ce0cb0284cf3f59acb --- /dev/null +++ b/runs/single_llava/removal/L08_s13/steps.jsonl @@ -0,0 +1,500 @@ +{"step": 1, "steps": 500, "lr": 2e-05, "grad_norm": 0.24176135659217834, "loss_scale": 4.0, "retries": 0, "seconds": 3.3, "max_memory_gb": 17.3, "ce": 1.6287612388841808, "rank": 0.13243506848812103} +{"step": 2, "steps": 500, "lr": 2e-05, "grad_norm": 0.3097337484359741, "loss_scale": 4.0, "retries": 0, "seconds": 5.1, "max_memory_gb": 17.32, "ce": 2.67563838721253, "rank": 0.03308701515197754} +{"step": 3, "steps": 500, "lr": 2e-05, "grad_norm": 0.7056853771209717, "loss_scale": 4.0, "retries": 0, "seconds": 7.0, "max_memory_gb": 17.32, "ce": 5.274260751903057, "rank": 0.1507800817489624} +{"step": 4, "steps": 500, "lr": 2e-05, "grad_norm": 0.45862317085266113, "loss_scale": 4.0, "retries": 0, "seconds": 8.9, "max_memory_gb": 17.32, "ce": 3.7993304044939578, "rank": 0.12462747097015381} +{"step": 5, "steps": 500, "lr": 2e-05, "grad_norm": 0.0886424109339714, "loss_scale": 4.0, "retries": 0, "seconds": 10.7, "max_memory_gb": 17.32, "ce": 0.1395322026219219, "rank": 0.0} +{"step": 6, "steps": 500, "lr": 2e-05, "grad_norm": 0.23511646687984467, "loss_scale": 4.0, "retries": 0, "seconds": 12.5, "max_memory_gb": 17.32, "ce": 1.5111454245634377, "rank": 0.0} +{"step": 7, "steps": 500, "lr": 2e-05, "grad_norm": 0.5145045518875122, "loss_scale": 4.0, "retries": 0, "seconds": 14.4, "max_memory_gb": 17.32, "ce": 3.4526670733466744, "rank": 0.020557045936584473} +{"step": 8, "steps": 500, "lr": 2e-05, "grad_norm": 0.6052871942520142, "loss_scale": 4.0, "retries": 0, "seconds": 16.3, "max_memory_gb": 17.32, "ce": 3.6944724554196, "rank": 0.20227301120758057} +{"step": 9, "steps": 500, "lr": 2e-05, "grad_norm": 0.48256534337997437, "loss_scale": 4.0, "retries": 0, "seconds": 18.1, "max_memory_gb": 17.32, "ce": 2.824452214874327, "rank": 0.039731383323669434} +{"step": 10, "steps": 500, "lr": 2e-05, "grad_norm": 0.2808643579483032, "loss_scale": 4.0, "retries": 0, "seconds": 20.0, "max_memory_gb": 17.33, "ce": 1.6126700714230537, "rank": 0.0} +{"step": 11, "steps": 500, "lr": 2e-05, "grad_norm": 0.6018903255462646, "loss_scale": 4.0, "retries": 0, "seconds": 21.8, "max_memory_gb": 17.33, "ce": 2.6085353903472424, "rank": 0.2063031792640686} +{"step": 12, "steps": 500, "lr": 2e-05, "grad_norm": 0.3934643864631653, "loss_scale": 4.0, "retries": 0, "seconds": 23.7, "max_memory_gb": 17.33, "ce": 1.4995027887634933, "rank": 0.04831898212432861} +{"step": 13, "steps": 500, "lr": 2e-05, "grad_norm": 0.5295979380607605, "loss_scale": 4.0, "retries": 0, "seconds": 25.5, "max_memory_gb": 17.33, "ce": 2.652324458118528, "rank": 0.0} +{"step": 14, "steps": 500, "lr": 2e-05, "grad_norm": 0.8410614132881165, "loss_scale": 4.0, "retries": 0, "seconds": 27.3, "max_memory_gb": 17.33, "ce": 3.8369220783933997, "rank": 0.14312744140625} +{"step": 15, "steps": 500, "lr": 2e-05, "grad_norm": 0.6187682151794434, "loss_scale": 4.0, "retries": 0, "seconds": 29.1, "max_memory_gb": 17.33, "ce": 2.5697093167109415, "rank": 0.16878247261047363} +{"step": 16, "steps": 500, "lr": 2e-05, "grad_norm": 0.41405370831489563, "loss_scale": 4.0, "retries": 0, "seconds": 31.0, "max_memory_gb": 17.33, "ce": 1.4004178085015155, "rank": 0.0} +{"step": 17, "steps": 500, "lr": 2e-05, "grad_norm": 0.43599605560302734, "loss_scale": 4.0, "retries": 0, "seconds": 32.8, "max_memory_gb": 17.33, "ce": 1.5728472584160045, "rank": 0.03293323516845703} +{"step": 18, "steps": 500, "lr": 2e-05, "grad_norm": 0.9039011001586914, "loss_scale": 4.0, "retries": 0, "seconds": 34.6, "max_memory_gb": 17.34, "ce": 3.5359937185421586, "rank": 0.19019711017608643} +{"step": 19, "steps": 500, "lr": 2e-05, "grad_norm": 1.3887903690338135, "loss_scale": 4.0, "retries": 0, "seconds": 36.5, "max_memory_gb": 17.34, "ce": 5.193546824157238, "rank": 0.31073832511901855} +{"step": 20, "steps": 500, "lr": 2e-05, "grad_norm": 0.030554424971342087, "loss_scale": 4.0, "retries": 0, "seconds": 38.3, "max_memory_gb": 17.34, "ce": 0.06554607418365777, "rank": 0.10980142652988434} +{"step": 21, "steps": 500, "lr": 2e-05, "grad_norm": 0.8697973489761353, "loss_scale": 4.0, "retries": 0, "seconds": 40.2, "max_memory_gb": 17.34, "ce": 2.683093619882129, "rank": 0.1456083059310913} +{"step": 22, "steps": 500, "lr": 2e-05, "grad_norm": 0.8851677179336548, "loss_scale": 4.0, "retries": 0, "seconds": 42.0, "max_memory_gb": 17.34, "ce": 2.878552853129804, "rank": 0.18462061882019043} +{"step": 23, "steps": 500, "lr": 2e-05, "grad_norm": 0.8723111152648926, "loss_scale": 4.0, "retries": 0, "seconds": 43.9, "max_memory_gb": 17.34, "ce": 2.683539856923744, "rank": 0.11857223510742188} +{"step": 24, "steps": 500, "lr": 2e-05, "grad_norm": 1.537123441696167, "loss_scale": 4.0, "retries": 0, "seconds": 45.8, "max_memory_gb": 17.34, "ce": 4.868989282287657, "rank": 0.25790905952453613} +{"step": 25, "steps": 500, "lr": 2e-05, "grad_norm": 1.6349176168441772, "loss_scale": 4.0, "retries": 0, "seconds": 47.7, "max_memory_gb": 17.34, "ce": 3.984785752836615, "rank": 0.023704171180725098} +{"step": 26, "steps": 500, "lr": 2e-05, "grad_norm": 2.1568734645843506, "loss_scale": 4.0, "retries": 0, "seconds": 49.8, "max_memory_gb": 17.34, "ce": 4.954357312992215, "rank": 0.0406113862991333} +{"step": 27, "steps": 500, "lr": 2e-05, "grad_norm": 1.7699730396270752, "loss_scale": 4.0, "retries": 0, "seconds": 51.8, "max_memory_gb": 17.34, "ce": 4.027610904537141, "rank": 0.4574793577194214} +{"step": 28, "steps": 500, "lr": 2e-05, "grad_norm": 1.0972484350204468, "loss_scale": 4.0, "retries": 0, "seconds": 53.7, "max_memory_gb": 17.34, "ce": 2.646693766117096, "rank": 0.0} +{"step": 29, "steps": 500, "lr": 2e-05, "grad_norm": 1.0088015794754028, "loss_scale": 4.0, "retries": 0, "seconds": 55.6, "max_memory_gb": 17.34, "ce": 2.568077955278568, "rank": 0.0} +{"step": 30, "steps": 500, "lr": 2e-05, "grad_norm": 1.4885315895080566, "loss_scale": 4.0, "retries": 0, "seconds": 57.6, "max_memory_gb": 17.34, "ce": 2.5785966184921563, "rank": 0.0} +{"step": 31, "steps": 500, "lr": 2e-05, "grad_norm": 1.5429198741912842, "loss_scale": 4.0, "retries": 0, "seconds": 59.6, "max_memory_gb": 17.34, "ce": 2.6348100462928414, "rank": 0.13172948360443115} +{"step": 32, "steps": 500, "lr": 2e-05, "grad_norm": 2.7783923149108887, "loss_scale": 4.0, "retries": 0, "seconds": 61.5, "max_memory_gb": 17.34, "ce": 4.958924552425742, "rank": 0.20267367362976074} +{"step": 33, "steps": 500, "lr": 2e-05, "grad_norm": 1.3132416009902954, "loss_scale": 4.0, "retries": 0, "seconds": 63.4, "max_memory_gb": 17.34, "ce": 2.3623947114683688, "rank": 0.006592094898223877} +{"step": 34, "steps": 500, "lr": 2e-05, "grad_norm": 1.5797131061553955, "loss_scale": 4.0, "retries": 0, "seconds": 65.2, "max_memory_gb": 17.34, "ce": 2.5301388105144724, "rank": 0.0} +{"step": 35, "steps": 500, "lr": 2e-05, "grad_norm": 2.0715065002441406, "loss_scale": 4.0, "retries": 0, "seconds": 67.0, "max_memory_gb": 17.34, "ce": 2.450880232732743, "rank": 0.12478303909301758} +{"step": 36, "steps": 500, "lr": 2e-05, "grad_norm": 1.4328404664993286, "loss_scale": 4.0, "retries": 0, "seconds": 68.9, "max_memory_gb": 17.34, "ce": 2.3888896964490414, "rank": 0.0} +{"step": 37, "steps": 500, "lr": 2e-05, "grad_norm": 1.6641669273376465, "loss_scale": 4.0, "retries": 0, "seconds": 70.7, "max_memory_gb": 17.34, "ce": 2.3869716664776206, "rank": 0.0} +{"step": 38, "steps": 500, "lr": 2e-05, "grad_norm": 3.211677074432373, "loss_scale": 4.0, "retries": 0, "seconds": 72.6, "max_memory_gb": 17.34, "ce": 4.388543339446187, "rank": 0.1259615421295166} +{"step": 39, "steps": 500, "lr": 2e-05, "grad_norm": 1.6379787921905518, "loss_scale": 4.0, "retries": 0, "seconds": 74.4, "max_memory_gb": 17.34, "ce": 1.2463010256178677, "rank": 0.09955359995365143} +{"step": 40, "steps": 500, "lr": 2e-05, "grad_norm": 2.14471697807312, "loss_scale": 4.0, "retries": 0, "seconds": 76.2, "max_memory_gb": 17.34, "ce": 2.3771423771977425, "rank": 0.0} +{"step": 41, "steps": 500, "lr": 2e-05, "grad_norm": 0.9798232316970825, "loss_scale": 4.0, "retries": 0, "seconds": 78.0, "max_memory_gb": 17.34, "ce": 1.3915541423484683, "rank": 0.0} +{"step": 42, "steps": 500, "lr": 2e-05, "grad_norm": 2.402458667755127, "loss_scale": 4.0, "retries": 0, "seconds": 79.9, "max_memory_gb": 17.34, "ce": 2.0452525787986815, "rank": 0.0} +{"step": 43, "steps": 500, "lr": 2e-05, "grad_norm": 2.8314452171325684, "loss_scale": 4.0, "retries": 0, "seconds": 81.7, "max_memory_gb": 17.34, "ce": 2.081079768948257, "rank": 0.0} +{"step": 44, "steps": 500, "lr": 2e-05, "grad_norm": 1.3500491380691528, "loss_scale": 4.0, "retries": 0, "seconds": 83.5, "max_memory_gb": 17.34, "ce": 1.2657712267246097, "rank": 0.0} +{"step": 45, "steps": 500, "lr": 2e-05, "grad_norm": 1.6121418476104736, "loss_scale": 4.0, "retries": 0, "seconds": 85.3, "max_memory_gb": 17.34, "ce": 1.2085123849101365, "rank": 0.0} +{"step": 46, "steps": 500, "lr": 2e-05, "grad_norm": 2.416421890258789, "loss_scale": 4.0, "retries": 0, "seconds": 87.2, "max_memory_gb": 17.34, "ce": 2.014447048306465, "rank": 0.0} +{"step": 47, "steps": 500, "lr": 2e-05, "grad_norm": 2.774829387664795, "loss_scale": 4.0, "retries": 0, "seconds": 89.0, "max_memory_gb": 17.34, "ce": 1.6670501134358346, "rank": 0.0} +{"step": 48, "steps": 500, "lr": 2e-05, "grad_norm": 6.317294120788574, "loss_scale": 4.0, "retries": 0, "seconds": 90.9, "max_memory_gb": 17.34, "ce": 2.9623494120314717, "rank": 0.003445744514465332} +{"step": 49, "steps": 500, "lr": 2e-05, "grad_norm": 6.4185991287231445, "loss_scale": 4.0, "retries": 0, "seconds": 92.7, "max_memory_gb": 17.34, "ce": 2.771866184193641, "rank": 0.0} +{"step": 50, "steps": 500, "lr": 2e-05, "grad_norm": 4.8315582275390625, "loss_scale": 4.0, "retries": 0, "seconds": 94.6, "max_memory_gb": 17.34, "ce": 2.890579945524223, "rank": 0.0} +{"step": 51, "steps": 500, "lr": 2e-05, "grad_norm": 2.088808059692383, "loss_scale": 4.0, "retries": 0, "seconds": 96.4, "max_memory_gb": 17.34, "ce": 1.1052802347112447, "rank": 0.0} +{"step": 52, "steps": 500, "lr": 2e-05, "grad_norm": 4.528656959533691, "loss_scale": 4.0, "retries": 0, "seconds": 98.2, "max_memory_gb": 17.34, "ce": 1.742985099554062, "rank": 0.0} +{"step": 53, "steps": 500, "lr": 2e-05, "grad_norm": 8.544352531433105, "loss_scale": 4.0, "retries": 0, "seconds": 100.1, "max_memory_gb": 17.34, "ce": 2.7595019033178687, "rank": 0.0} +{"step": 54, "steps": 500, "lr": 2e-05, "grad_norm": 9.849609375, "loss_scale": 4.0, "retries": 0, "seconds": 101.9, "max_memory_gb": 17.34, "ce": 2.613226531073451, "rank": 0.0} +{"step": 55, "steps": 500, "lr": 2e-05, "grad_norm": 6.243852138519287, "loss_scale": 4.0, "retries": 0, "seconds": 103.7, "max_memory_gb": 17.34, "ce": 1.5800056781154126, "rank": 0.0} +{"step": 56, "steps": 500, "lr": 2e-05, "grad_norm": 13.824606895446777, "loss_scale": 4.0, "retries": 0, "seconds": 105.5, "max_memory_gb": 17.34, "ce": 2.927625677548349, "rank": 0.0} +{"step": 57, "steps": 500, "lr": 2e-05, "grad_norm": 3.7510108947753906, "loss_scale": 4.0, "retries": 0, "seconds": 107.4, "max_memory_gb": 17.34, "ce": 0.8216518955305219, "rank": 0.005446568131446838} +{"step": 58, "steps": 500, "lr": 2e-05, "grad_norm": 6.93580436706543, "loss_scale": 4.0, "retries": 0, "seconds": 109.2, "max_memory_gb": 17.34, "ce": 1.2932693575276062, "rank": 0.0} +{"step": 59, "steps": 500, "lr": 2e-05, "grad_norm": 2.842679977416992, "loss_scale": 4.0, "retries": 0, "seconds": 111.0, "max_memory_gb": 17.34, "ce": 1.110278032720089, "rank": 0.0} +{"step": 60, "steps": 500, "lr": 2e-05, "grad_norm": 6.099782943725586, "loss_scale": 4.0, "retries": 0, "seconds": 112.9, "max_memory_gb": 17.34, "ce": 1.3463162570260465, "rank": 0.0} +{"step": 61, "steps": 500, "lr": 2e-05, "grad_norm": 10.023585319519043, "loss_scale": 4.0, "retries": 0, "seconds": 114.7, "max_memory_gb": 17.34, "ce": 1.4609523881226778, "rank": 0.0} +{"step": 62, "steps": 500, "lr": 2e-05, "grad_norm": 3.269993543624878, "loss_scale": 4.0, "retries": 0, "seconds": 116.5, "max_memory_gb": 17.34, "ce": 0.7041023671627045, "rank": 0.0} +{"step": 63, "steps": 500, "lr": 2e-05, "grad_norm": 9.020225524902344, "loss_scale": 4.0, "retries": 0, "seconds": 118.4, "max_memory_gb": 17.34, "ce": 1.7282179107423872, "rank": 0.03400975465774536} +{"step": 64, "steps": 500, "lr": 2e-05, "grad_norm": 9.46395206451416, "loss_scale": 4.0, "retries": 0, "seconds": 120.3, "max_memory_gb": 17.34, "ce": 1.558552564587444, "rank": 0.0} +{"step": 65, "steps": 500, "lr": 2e-05, "grad_norm": 7.7949042320251465, "loss_scale": 4.0, "retries": 0, "seconds": 122.1, "max_memory_gb": 17.34, "ce": 1.6859642760828137, "rank": 0.0} +{"step": 66, "steps": 500, "lr": 2e-05, "grad_norm": 2.7619965076446533, "loss_scale": 4.0, "retries": 0, "seconds": 123.9, "max_memory_gb": 17.34, "ce": 0.6019593116361648, "rank": 0.287388414144516} +{"step": 67, "steps": 500, "lr": 2e-05, "grad_norm": 5.1615142822265625, "loss_scale": 4.0, "retries": 0, "seconds": 125.8, "max_memory_gb": 17.34, "ce": 1.128297376446426, "rank": 0.0} +{"step": 68, "steps": 500, "lr": 2e-05, "grad_norm": 3.22772479057312, "loss_scale": 4.0, "retries": 0, "seconds": 127.7, "max_memory_gb": 17.34, "ce": 0.47235896717756987, "rank": 0.0} +{"step": 69, "steps": 500, "lr": 2e-05, "grad_norm": 4.193150520324707, "loss_scale": 4.0, "retries": 0, "seconds": 129.6, "max_memory_gb": 17.34, "ce": 0.7279859082773328, "rank": 0.0} +{"step": 70, "steps": 500, "lr": 2e-05, "grad_norm": 3.487277030944824, "loss_scale": 4.0, "retries": 0, "seconds": 131.4, "max_memory_gb": 17.34, "ce": 0.6675147386267781, "rank": 0.01150086522102356} +{"step": 71, "steps": 500, "lr": 2e-05, "grad_norm": 4.912594318389893, "loss_scale": 4.0, "retries": 0, "seconds": 133.3, "max_memory_gb": 17.34, "ce": 0.8540216332767159, "rank": 0.0} +{"step": 72, "steps": 500, "lr": 2e-05, "grad_norm": 5.557231903076172, "loss_scale": 4.0, "retries": 0, "seconds": 135.1, "max_memory_gb": 17.34, "ce": 1.0710841324180365, "rank": 0.0} +{"step": 73, "steps": 500, "lr": 2e-05, "grad_norm": 1.3501522541046143, "loss_scale": 4.0, "retries": 0, "seconds": 136.9, "max_memory_gb": 17.34, "ce": 0.31309322104789317, "rank": 0.0} +{"step": 74, "steps": 500, "lr": 2e-05, "grad_norm": 1.9615143537521362, "loss_scale": 4.0, "retries": 0, "seconds": 138.8, "max_memory_gb": 17.34, "ce": 0.6209592190571129, "rank": 0.0} +{"step": 75, "steps": 500, "lr": 2e-05, "grad_norm": 1.6654322147369385, "loss_scale": 4.0, "retries": 0, "seconds": 140.6, "max_memory_gb": 17.34, "ce": 0.909344376064837, "rank": 0.2642558515071869} +{"step": 76, "steps": 500, "lr": 2e-05, "grad_norm": 1.7835737466812134, "loss_scale": 4.0, "retries": 0, "seconds": 142.5, "max_memory_gb": 17.34, "ce": 0.6467451569624245, "rank": 0.0} +{"step": 77, "steps": 500, "lr": 2e-05, "grad_norm": 2.4726316928863525, "loss_scale": 4.0, "retries": 0, "seconds": 144.3, "max_memory_gb": 17.34, "ce": 0.9007809944450855, "rank": 0.0} +{"step": 78, "steps": 500, "lr": 2e-05, "grad_norm": 2.481137752532959, "loss_scale": 4.0, "retries": 0, "seconds": 146.1, "max_memory_gb": 17.34, "ce": 0.5918539743870497, "rank": 0.0} +{"step": 79, "steps": 500, "lr": 2e-05, "grad_norm": 1.697904348373413, "loss_scale": 4.0, "retries": 0, "seconds": 148.0, "max_memory_gb": 17.34, "ce": 0.4003100134432316, "rank": 0.0} +{"step": 80, "steps": 500, "lr": 2e-05, "grad_norm": 1.732640027999878, "loss_scale": 4.0, "retries": 0, "seconds": 149.8, "max_memory_gb": 17.34, "ce": 0.46377215860411525, "rank": 0.0} +{"step": 81, "steps": 500, "lr": 2e-05, "grad_norm": 2.1611640453338623, "loss_scale": 4.0, "retries": 0, "seconds": 151.6, "max_memory_gb": 17.34, "ce": 0.7833280079066753, "rank": 0.0} +{"step": 82, "steps": 500, "lr": 2e-05, "grad_norm": 1.7521005868911743, "loss_scale": 4.0, "retries": 0, "seconds": 153.5, "max_memory_gb": 17.34, "ce": 0.6054010456427932, "rank": 0.0} +{"step": 83, "steps": 500, "lr": 2e-05, "grad_norm": 1.8167004585266113, "loss_scale": 4.0, "retries": 0, "seconds": 155.4, "max_memory_gb": 17.34, "ce": 0.7046401998959482, "rank": 0.0} +{"step": 84, "steps": 500, "lr": 2e-05, "grad_norm": 1.0957136154174805, "loss_scale": 4.0, "retries": 0, "seconds": 157.2, "max_memory_gb": 17.34, "ce": 0.33506036968901753, "rank": 0.0} +{"step": 85, "steps": 500, "lr": 2e-05, "grad_norm": 1.0861154794692993, "loss_scale": 4.0, "retries": 0, "seconds": 159.1, "max_memory_gb": 17.34, "ce": 0.16664183977991343, "rank": 0.0} +{"step": 86, "steps": 500, "lr": 2e-05, "grad_norm": 1.611312985420227, "loss_scale": 4.0, "retries": 0, "seconds": 160.9, "max_memory_gb": 17.34, "ce": 0.6412815200164914, "rank": 0.0} +{"step": 87, "steps": 500, "lr": 2e-05, "grad_norm": 1.8008615970611572, "loss_scale": 4.0, "retries": 0, "seconds": 162.8, "max_memory_gb": 17.34, "ce": 0.6324054137803614, "rank": 0.0} +{"step": 88, "steps": 500, "lr": 2e-05, "grad_norm": 2.008326530456543, "loss_scale": 4.0, "retries": 0, "seconds": 164.6, "max_memory_gb": 17.34, "ce": 0.5891312079038471, "rank": 0.0} +{"step": 89, "steps": 500, "lr": 2e-05, "grad_norm": 0.5539577603340149, "loss_scale": 4.0, "retries": 0, "seconds": 166.4, "max_memory_gb": 17.34, "ce": 0.17621691874228418, "rank": 0.12540726363658905} +{"step": 90, "steps": 500, "lr": 2e-05, "grad_norm": 1.055309772491455, "loss_scale": 4.0, "retries": 0, "seconds": 168.3, "max_memory_gb": 17.34, "ce": 0.4566812734119594, "rank": 0.0} +{"step": 91, "steps": 500, "lr": 2e-05, "grad_norm": 0.7156156897544861, "loss_scale": 4.0, "retries": 0, "seconds": 170.1, "max_memory_gb": 17.34, "ce": 0.38253894774243236, "rank": 0.0} +{"step": 92, "steps": 500, "lr": 2e-05, "grad_norm": 0.4905855655670166, "loss_scale": 4.0, "retries": 0, "seconds": 172.0, "max_memory_gb": 17.34, "ce": 0.30325164983514696, "rank": 0.0} +{"step": 93, "steps": 500, "lr": 2e-05, "grad_norm": 1.3400225639343262, "loss_scale": 4.0, "retries": 0, "seconds": 173.8, "max_memory_gb": 17.34, "ce": 0.4530157716944814, "rank": 0.0} +{"step": 94, "steps": 500, "lr": 2e-05, "grad_norm": 0.913673996925354, "loss_scale": 4.0, "retries": 0, "seconds": 175.7, "max_memory_gb": 17.34, "ce": 0.5404089689254761, "rank": 0.012807890772819519} +{"step": 95, "steps": 500, "lr": 2e-05, "grad_norm": 0.9875110387802124, "loss_scale": 4.0, "retries": 0, "seconds": 177.5, "max_memory_gb": 17.34, "ce": 0.5348724224604666, "rank": 0.0} +{"step": 96, "steps": 500, "lr": 2e-05, "grad_norm": 1.0996694564819336, "loss_scale": 4.0, "retries": 0, "seconds": 179.4, "max_memory_gb": 17.34, "ce": 0.6055750716477633, "rank": 0.018270105123519897} +{"step": 97, "steps": 500, "lr": 2e-05, "grad_norm": 0.7057777643203735, "loss_scale": 4.0, "retries": 0, "seconds": 181.2, "max_memory_gb": 17.34, "ce": 0.336971715092659, "rank": 0.024963364005088806} +{"step": 98, "steps": 500, "lr": 2e-05, "grad_norm": 0.6135924458503723, "loss_scale": 4.0, "retries": 0, "seconds": 183.1, "max_memory_gb": 17.34, "ce": 0.3352437927387655, "rank": 0.001840353012084961} +{"step": 99, "steps": 500, "lr": 2e-05, "grad_norm": 0.7690953016281128, "loss_scale": 4.0, "retries": 0, "seconds": 184.9, "max_memory_gb": 17.34, "ce": 0.24317629379220307, "rank": 0.01948697865009308} +{"step": 100, "steps": 500, "lr": 2e-05, "grad_norm": 0.7686785459518433, "loss_scale": 4.0, "retries": 0, "seconds": 186.8, "max_memory_gb": 17.34, "ce": 0.5377555340528488, "rank": 0.0} +{"step": 101, "steps": 500, "lr": 2e-05, "grad_norm": 0.7198999524116516, "loss_scale": 4.0, "retries": 0, "seconds": 188.7, "max_memory_gb": 17.34, "ce": 0.49706004187464714, "rank": 0.0} +{"step": 102, "steps": 500, "lr": 2e-05, "grad_norm": 1.1081600189208984, "loss_scale": 4.0, "retries": 0, "seconds": 190.5, "max_memory_gb": 17.34, "ce": 0.5232335785403848, "rank": 0.0} +{"step": 103, "steps": 500, "lr": 2e-05, "grad_norm": 0.36986345052719116, "loss_scale": 4.0, "retries": 0, "seconds": 192.3, "max_memory_gb": 17.34, "ce": 0.3128368059406057, "rank": 0.12410037964582443} +{"step": 104, "steps": 500, "lr": 2e-05, "grad_norm": 0.31842997670173645, "loss_scale": 4.0, "retries": 0, "seconds": 194.1, "max_memory_gb": 17.34, "ce": 0.23392807698110119, "rank": 0.08974498510360718} +{"step": 105, "steps": 500, "lr": 2e-05, "grad_norm": 0.6614657640457153, "loss_scale": 4.0, "retries": 0, "seconds": 196.0, "max_memory_gb": 17.34, "ce": 0.4861956248059869, "rank": 0.0} +{"step": 106, "steps": 500, "lr": 2e-05, "grad_norm": 0.5524685382843018, "loss_scale": 4.0, "retries": 0, "seconds": 197.8, "max_memory_gb": 17.34, "ce": 0.4718133918941021, "rank": 0.09006102383136749} +{"step": 107, "steps": 500, "lr": 2e-05, "grad_norm": 0.3639889359474182, "loss_scale": 4.0, "retries": 0, "seconds": 199.7, "max_memory_gb": 17.34, "ce": 0.28976128390058875, "rank": 0.0} +{"step": 108, "steps": 500, "lr": 2e-05, "grad_norm": 0.4260652959346771, "loss_scale": 4.0, "retries": 0, "seconds": 201.5, "max_memory_gb": 17.34, "ce": 0.28442806866951287, "rank": 0.0} +{"step": 109, "steps": 500, "lr": 2e-05, "grad_norm": 0.6435948014259338, "loss_scale": 4.0, "retries": 0, "seconds": 203.4, "max_memory_gb": 17.34, "ce": 0.6179638803005219, "rank": 0.0} +{"step": 110, "steps": 500, "lr": 2e-05, "grad_norm": 0.6991602182388306, "loss_scale": 4.0, "retries": 0, "seconds": 205.3, "max_memory_gb": 17.34, "ce": 0.40586599614471197, "rank": 0.04566193372011185} +{"step": 111, "steps": 500, "lr": 2e-05, "grad_norm": 0.46297794580459595, "loss_scale": 4.0, "retries": 0, "seconds": 207.1, "max_memory_gb": 17.34, "ce": 0.3456305209547281, "rank": 0.0} +{"step": 112, "steps": 500, "lr": 2e-05, "grad_norm": 0.5678306221961975, "loss_scale": 4.0, "retries": 0, "seconds": 209.0, "max_memory_gb": 17.34, "ce": 0.32480523525737226, "rank": 0.017912447452545166} +{"step": 113, "steps": 500, "lr": 2e-05, "grad_norm": 0.39809080958366394, "loss_scale": 4.0, "retries": 0, "seconds": 210.8, "max_memory_gb": 17.34, "ce": 0.3737763436511159, "rank": 0.0} +{"step": 114, "steps": 500, "lr": 2e-05, "grad_norm": 0.8472577333450317, "loss_scale": 4.0, "retries": 0, "seconds": 212.7, "max_memory_gb": 17.34, "ce": 0.7400926072150469, "rank": 0.16142195463180542} +{"step": 115, "steps": 500, "lr": 2e-05, "grad_norm": 0.4343844950199127, "loss_scale": 4.0, "retries": 0, "seconds": 214.6, "max_memory_gb": 17.34, "ce": 0.4508802080526948, "rank": 0.08629003912210464} +{"step": 116, "steps": 500, "lr": 2e-05, "grad_norm": 0.21647951006889343, "loss_scale": 4.0, "retries": 0, "seconds": 216.4, "max_memory_gb": 17.34, "ce": 0.15229713590815663, "rank": 0.0} +{"step": 117, "steps": 500, "lr": 2e-05, "grad_norm": 0.8111744523048401, "loss_scale": 4.0, "retries": 0, "seconds": 218.3, "max_memory_gb": 17.34, "ce": 0.462206082069315, "rank": 0.0} +{"step": 118, "steps": 500, "lr": 2e-05, "grad_norm": 0.2705690860748291, "loss_scale": 4.0, "retries": 0, "seconds": 220.1, "max_memory_gb": 17.34, "ce": 0.2925189514644444, "rank": 0.0} +{"step": 119, "steps": 500, "lr": 2e-05, "grad_norm": 0.342858225107193, "loss_scale": 4.0, "retries": 0, "seconds": 222.0, "max_memory_gb": 17.34, "ce": 0.2628589286468923, "rank": 0.0} +{"step": 120, "steps": 500, "lr": 2e-05, "grad_norm": 0.7357234358787537, "loss_scale": 4.0, "retries": 0, "seconds": 223.8, "max_memory_gb": 17.34, "ce": 0.42576196242589504, "rank": 0.0} +{"step": 121, "steps": 500, "lr": 2e-05, "grad_norm": 0.6206809282302856, "loss_scale": 4.0, "retries": 0, "seconds": 225.7, "max_memory_gb": 17.34, "ce": 0.363963428651914, "rank": 0.0} +{"step": 122, "steps": 500, "lr": 2e-05, "grad_norm": 0.37195828557014465, "loss_scale": 4.0, "retries": 0, "seconds": 227.6, "max_memory_gb": 17.34, "ce": 0.36994031129870564, "rank": 0.0} +{"step": 123, "steps": 500, "lr": 2e-05, "grad_norm": 1.028823733329773, "loss_scale": 4.0, "retries": 0, "seconds": 229.4, "max_memory_gb": 17.34, "ce": 0.4532492713769898, "rank": 0.0071530938148498535} +{"step": 124, "steps": 500, "lr": 2e-05, "grad_norm": 0.5933623909950256, "loss_scale": 4.0, "retries": 0, "seconds": 231.3, "max_memory_gb": 17.34, "ce": 0.38639546115882695, "rank": 0.060962677001953125} +{"step": 125, "steps": 500, "lr": 2e-05, "grad_norm": 0.27419134974479675, "loss_scale": 4.0, "retries": 0, "seconds": 233.1, "max_memory_gb": 17.34, "ce": 0.3001829986460507, "rank": 0.0} +{"step": 126, "steps": 500, "lr": 2e-05, "grad_norm": 0.6709418892860413, "loss_scale": 4.0, "retries": 0, "seconds": 234.9, "max_memory_gb": 17.34, "ce": 0.5574380992911756, "rank": 0.0} +{"step": 127, "steps": 500, "lr": 2e-05, "grad_norm": 0.2593696415424347, "loss_scale": 4.0, "retries": 0, "seconds": 236.8, "max_memory_gb": 17.34, "ce": 0.17106277705170214, "rank": 0.0} +{"step": 128, "steps": 500, "lr": 2e-05, "grad_norm": 0.5580383539199829, "loss_scale": 4.0, "retries": 0, "seconds": 238.6, "max_memory_gb": 17.34, "ce": 0.41846209403593093, "rank": 0.0} +{"step": 129, "steps": 500, "lr": 2e-05, "grad_norm": 0.6308526396751404, "loss_scale": 4.0, "retries": 0, "seconds": 240.5, "max_memory_gb": 17.34, "ce": 0.49207643046975136, "rank": 0.0} +{"step": 130, "steps": 500, "lr": 2e-05, "grad_norm": 0.7809165120124817, "loss_scale": 4.0, "retries": 0, "seconds": 242.3, "max_memory_gb": 17.34, "ce": 0.5541061866097152, "rank": 0.0} +{"step": 131, "steps": 500, "lr": 2e-05, "grad_norm": 0.7817592024803162, "loss_scale": 4.0, "retries": 0, "seconds": 244.2, "max_memory_gb": 17.34, "ce": 0.38683519000187516, "rank": 0.0} +{"step": 132, "steps": 500, "lr": 2e-05, "grad_norm": 0.9439342617988586, "loss_scale": 4.0, "retries": 0, "seconds": 246.1, "max_memory_gb": 17.34, "ce": 0.5411914943251759, "rank": 0.06650236248970032} +{"step": 133, "steps": 500, "lr": 2e-05, "grad_norm": 0.984186053276062, "loss_scale": 4.0, "retries": 0, "seconds": 247.9, "max_memory_gb": 17.34, "ce": 0.5085477819666266, "rank": 0.08126857131719589} +{"step": 134, "steps": 500, "lr": 2e-05, "grad_norm": 0.5840386748313904, "loss_scale": 4.0, "retries": 0, "seconds": 249.8, "max_memory_gb": 17.34, "ce": 0.22009602014441043, "rank": 0.0} +{"step": 135, "steps": 500, "lr": 2e-05, "grad_norm": 0.5671910643577576, "loss_scale": 4.0, "retries": 0, "seconds": 251.6, "max_memory_gb": 17.34, "ce": 0.42591370176523924, "rank": 0.03089126944541931} +{"step": 136, "steps": 500, "lr": 2e-05, "grad_norm": 0.5983936786651611, "loss_scale": 4.0, "retries": 0, "seconds": 253.5, "max_memory_gb": 17.34, "ce": 0.4247374855913222, "rank": 0.0} +{"step": 137, "steps": 500, "lr": 2e-05, "grad_norm": 0.5264595150947571, "loss_scale": 4.0, "retries": 0, "seconds": 255.3, "max_memory_gb": 17.34, "ce": 0.35364956548437476, "rank": 0.0} +{"step": 138, "steps": 500, "lr": 2e-05, "grad_norm": 0.510338306427002, "loss_scale": 4.0, "retries": 0, "seconds": 257.2, "max_memory_gb": 17.34, "ce": 0.3294811355881393, "rank": 0.0005962401628494263} +{"step": 139, "steps": 500, "lr": 2e-05, "grad_norm": 1.3204140663146973, "loss_scale": 4.0, "retries": 0, "seconds": 259.1, "max_memory_gb": 17.34, "ce": 0.523496572393924, "rank": 0.0} +{"step": 140, "steps": 500, "lr": 2e-05, "grad_norm": 0.4503270089626312, "loss_scale": 4.0, "retries": 0, "seconds": 261.0, "max_memory_gb": 17.34, "ce": 0.27347363461740315, "rank": 0.0} +{"step": 141, "steps": 500, "lr": 2e-05, "grad_norm": 0.47767436504364014, "loss_scale": 4.0, "retries": 0, "seconds": 262.9, "max_memory_gb": 17.34, "ce": 0.36702672322280705, "rank": 0.0} +{"step": 142, "steps": 500, "lr": 2e-05, "grad_norm": 0.3802909255027771, "loss_scale": 4.0, "retries": 0, "seconds": 264.7, "max_memory_gb": 17.34, "ce": 0.3370741819962859, "rank": 0.0} +{"step": 143, "steps": 500, "lr": 2e-05, "grad_norm": 0.21434248983860016, "loss_scale": 4.0, "retries": 0, "seconds": 266.6, "max_memory_gb": 17.34, "ce": 0.17433077580062672, "rank": 0.0} +{"step": 144, "steps": 500, "lr": 2e-05, "grad_norm": 0.6179916858673096, "loss_scale": 4.0, "retries": 0, "seconds": 268.5, "max_memory_gb": 17.34, "ce": 0.3652825236786157, "rank": 0.0} +{"step": 145, "steps": 500, "lr": 2e-05, "grad_norm": 0.36642804741859436, "loss_scale": 4.0, "retries": 0, "seconds": 270.3, "max_memory_gb": 17.34, "ce": 0.22953221038915217, "rank": 0.0} +{"step": 146, "steps": 500, "lr": 2e-05, "grad_norm": 0.5141533613204956, "loss_scale": 4.0, "retries": 0, "seconds": 272.2, "max_memory_gb": 17.34, "ce": 0.3126876261085272, "rank": 0.0} +{"step": 147, "steps": 500, "lr": 2e-05, "grad_norm": 0.7647861838340759, "loss_scale": 4.0, "retries": 0, "seconds": 274.1, "max_memory_gb": 17.34, "ce": 0.5399182558758184, "rank": 0.0} +{"step": 148, "steps": 500, "lr": 2e-05, "grad_norm": 0.2615666091442108, "loss_scale": 4.0, "retries": 0, "seconds": 275.9, "max_memory_gb": 17.34, "ce": 0.1473073357483372, "rank": 0.0} +{"step": 149, "steps": 500, "lr": 2e-05, "grad_norm": 0.4098558723926544, "loss_scale": 4.0, "retries": 0, "seconds": 277.7, "max_memory_gb": 17.34, "ce": 0.24119182000868022, "rank": 0.1340540647506714} +{"step": 150, "steps": 500, "lr": 2e-05, "grad_norm": 0.790276050567627, "loss_scale": 4.0, "retries": 0, "seconds": 279.7, "max_memory_gb": 17.34, "ce": 0.46055363956838846, "rank": 0.022324830293655396} +{"step": 151, "steps": 500, "lr": 2e-05, "grad_norm": 0.42819416522979736, "loss_scale": 4.0, "retries": 0, "seconds": 281.6, "max_memory_gb": 17.34, "ce": 0.23110536497551948, "rank": 0.0} +{"step": 152, "steps": 500, "lr": 2e-05, "grad_norm": 0.8866899609565735, "loss_scale": 4.0, "retries": 0, "seconds": 283.4, "max_memory_gb": 17.34, "ce": 0.44601561315357685, "rank": 0.0} +{"step": 153, "steps": 500, "lr": 2e-05, "grad_norm": 0.6201188564300537, "loss_scale": 4.0, "retries": 0, "seconds": 285.3, "max_memory_gb": 17.34, "ce": 0.3391284419922158, "rank": 0.0} +{"step": 154, "steps": 500, "lr": 2e-05, "grad_norm": 1.1066838502883911, "loss_scale": 4.0, "retries": 0, "seconds": 287.2, "max_memory_gb": 17.34, "ce": 0.5793533567339182, "rank": 0.14037568867206573} +{"step": 155, "steps": 500, "lr": 2e-05, "grad_norm": 1.2863378524780273, "loss_scale": 4.0, "retries": 0, "seconds": 289.0, "max_memory_gb": 17.34, "ce": 0.5334343076683581, "rank": 0.0} +{"step": 156, "steps": 500, "lr": 2e-05, "grad_norm": 0.486722856760025, "loss_scale": 4.0, "retries": 0, "seconds": 290.9, "max_memory_gb": 17.34, "ce": 0.2727886321954429, "rank": 0.0} +{"step": 157, "steps": 500, "lr": 2e-05, "grad_norm": 0.4099171757698059, "loss_scale": 4.0, "retries": 0, "seconds": 292.7, "max_memory_gb": 17.34, "ce": 0.25925178173929453, "rank": 0.0} +{"step": 158, "steps": 500, "lr": 2e-05, "grad_norm": 0.7266494631767273, "loss_scale": 4.0, "retries": 0, "seconds": 294.5, "max_memory_gb": 17.34, "ce": 0.4575706892646849, "rank": 0.0} +{"step": 159, "steps": 500, "lr": 2e-05, "grad_norm": 0.4864291846752167, "loss_scale": 4.0, "retries": 0, "seconds": 296.4, "max_memory_gb": 17.34, "ce": 0.2734670464415103, "rank": 0.0} +{"step": 160, "steps": 500, "lr": 2e-05, "grad_norm": 0.7668102383613586, "loss_scale": 4.0, "retries": 0, "seconds": 298.3, "max_memory_gb": 17.34, "ce": 0.36382291512563825, "rank": 0.04839296638965607} +{"step": 161, "steps": 500, "lr": 2e-05, "grad_norm": 0.5953884124755859, "loss_scale": 4.0, "retries": 0, "seconds": 300.1, "max_memory_gb": 17.34, "ce": 0.3383690875489265, "rank": 0.0} +{"step": 162, "steps": 500, "lr": 2e-05, "grad_norm": 0.3692464530467987, "loss_scale": 4.0, "retries": 0, "seconds": 301.9, "max_memory_gb": 17.34, "ce": 0.2247372311539948, "rank": 0.0} +{"step": 163, "steps": 500, "lr": 2e-05, "grad_norm": 0.7543410658836365, "loss_scale": 4.0, "retries": 0, "seconds": 303.7, "max_memory_gb": 17.34, "ce": 0.33460321580059826, "rank": 0.0} +{"step": 164, "steps": 500, "lr": 2e-05, "grad_norm": 0.8098766207695007, "loss_scale": 4.0, "retries": 0, "seconds": 305.6, "max_memory_gb": 17.34, "ce": 0.4057140424847603, "rank": 0.027532100677490234} +{"step": 165, "steps": 500, "lr": 2e-05, "grad_norm": 0.4599829614162445, "loss_scale": 4.0, "retries": 0, "seconds": 307.5, "max_memory_gb": 17.34, "ce": 0.2309956904500723, "rank": 0.0} +{"step": 166, "steps": 500, "lr": 2e-05, "grad_norm": 0.8146171569824219, "loss_scale": 4.0, "retries": 0, "seconds": 309.3, "max_memory_gb": 17.34, "ce": 0.33056111820042133, "rank": 0.0} +{"step": 167, "steps": 500, "lr": 2e-05, "grad_norm": 0.9097492694854736, "loss_scale": 4.0, "retries": 0, "seconds": 311.3, "max_memory_gb": 17.34, "ce": 0.38601739052683115, "rank": 0.0} +{"step": 168, "steps": 500, "lr": 2e-05, "grad_norm": 0.6141856908798218, "loss_scale": 4.0, "retries": 0, "seconds": 313.1, "max_memory_gb": 17.34, "ce": 0.24773850641213357, "rank": 0.0} +{"step": 169, "steps": 500, "lr": 2e-05, "grad_norm": 1.9414047002792358, "loss_scale": 4.0, "retries": 0, "seconds": 315.0, "max_memory_gb": 17.34, "ce": 0.7693972638808191, "rank": 0.0} +{"step": 170, "steps": 500, "lr": 2e-05, "grad_norm": 0.8773718476295471, "loss_scale": 4.0, "retries": 0, "seconds": 316.8, "max_memory_gb": 17.34, "ce": 0.25344414287246764, "rank": 0.0} +{"step": 171, "steps": 500, "lr": 2e-05, "grad_norm": 0.947774350643158, "loss_scale": 4.0, "retries": 0, "seconds": 318.7, "max_memory_gb": 17.34, "ce": 0.270555098191835, "rank": 0.0} +{"step": 172, "steps": 500, "lr": 2e-05, "grad_norm": 1.138973593711853, "loss_scale": 4.0, "retries": 0, "seconds": 320.5, "max_memory_gb": 17.34, "ce": 0.31797429849393666, "rank": 0.0} +{"step": 173, "steps": 500, "lr": 2e-05, "grad_norm": 1.6571425199508667, "loss_scale": 4.0, "retries": 0, "seconds": 322.4, "max_memory_gb": 17.34, "ce": 0.4447808456607163, "rank": 0.0} +{"step": 174, "steps": 500, "lr": 2e-05, "grad_norm": 1.892441987991333, "loss_scale": 4.0, "retries": 0, "seconds": 324.2, "max_memory_gb": 17.34, "ce": 0.29080428852466866, "rank": 0.0} +{"step": 175, "steps": 500, "lr": 2e-05, "grad_norm": 1.1577613353729248, "loss_scale": 4.0, "retries": 0, "seconds": 326.0, "max_memory_gb": 17.34, "ce": 0.3677772884257138, "rank": 0.0} +{"step": 176, "steps": 500, "lr": 2e-05, "grad_norm": 1.9785497188568115, "loss_scale": 4.0, "retries": 0, "seconds": 327.9, "max_memory_gb": 17.34, "ce": 0.3014901615679264, "rank": 0.0} +{"step": 177, "steps": 500, "lr": 2e-05, "grad_norm": 0.8998209238052368, "loss_scale": 4.0, "retries": 0, "seconds": 329.7, "max_memory_gb": 17.34, "ce": 0.3085697724018246, "rank": 0.12715068459510803} +{"step": 178, "steps": 500, "lr": 2e-05, "grad_norm": 0.5893502235412598, "loss_scale": 4.0, "retries": 0, "seconds": 331.6, "max_memory_gb": 17.34, "ce": 0.24098566372413188, "rank": 0.0} +{"step": 179, "steps": 500, "lr": 2e-05, "grad_norm": 0.49053171277046204, "loss_scale": 4.0, "retries": 0, "seconds": 333.4, "max_memory_gb": 17.34, "ce": 0.4119155812077224, "rank": 0.0} +{"step": 180, "steps": 500, "lr": 2e-05, "grad_norm": 1.5250962972640991, "loss_scale": 4.0, "retries": 0, "seconds": 335.2, "max_memory_gb": 17.34, "ce": 0.43157938960939646, "rank": 0.0} +{"step": 181, "steps": 500, "lr": 2e-05, "grad_norm": 1.1743109226226807, "loss_scale": 4.0, "retries": 0, "seconds": 337.1, "max_memory_gb": 17.34, "ce": 0.30286247562617064, "rank": 0.0} +{"step": 182, "steps": 500, "lr": 2e-05, "grad_norm": 1.0158960819244385, "loss_scale": 4.0, "retries": 0, "seconds": 338.9, "max_memory_gb": 17.34, "ce": 0.24460338824428618, "rank": 0.0} +{"step": 183, "steps": 500, "lr": 2e-05, "grad_norm": 0.5640507936477661, "loss_scale": 4.0, "retries": 0, "seconds": 340.8, "max_memory_gb": 17.34, "ce": 0.12254069664049894, "rank": 0.0} +{"step": 184, "steps": 500, "lr": 2e-05, "grad_norm": 0.7847194075584412, "loss_scale": 4.0, "retries": 0, "seconds": 342.6, "max_memory_gb": 17.34, "ce": 0.1910036593908444, "rank": 0.11665014177560806} +{"step": 185, "steps": 500, "lr": 2e-05, "grad_norm": 1.0794775485992432, "loss_scale": 4.0, "retries": 0, "seconds": 344.5, "max_memory_gb": 17.34, "ce": 0.2135641893837601, "rank": 0.0} +{"step": 186, "steps": 500, "lr": 2e-05, "grad_norm": 0.7400692701339722, "loss_scale": 4.0, "retries": 0, "seconds": 346.4, "max_memory_gb": 17.34, "ce": 0.15583302319282666, "rank": 0.0} +{"step": 187, "steps": 500, "lr": 2e-05, "grad_norm": 1.3642981052398682, "loss_scale": 4.0, "retries": 0, "seconds": 348.3, "max_memory_gb": 17.34, "ce": 0.3440498895943165, "rank": 0.12978045642375946} +{"step": 188, "steps": 500, "lr": 2e-05, "grad_norm": 0.23895570635795593, "loss_scale": 4.0, "retries": 0, "seconds": 350.1, "max_memory_gb": 17.34, "ce": 0.1513524444308132, "rank": 0.0} +{"step": 189, "steps": 500, "lr": 2e-05, "grad_norm": 1.0107115507125854, "loss_scale": 4.0, "retries": 0, "seconds": 351.9, "max_memory_gb": 17.34, "ce": 0.1766729997470975, "rank": 0.0} +{"step": 190, "steps": 500, "lr": 2e-05, "grad_norm": 1.524193286895752, "loss_scale": 4.0, "retries": 0, "seconds": 353.8, "max_memory_gb": 17.34, "ce": 0.27447924111038446, "rank": 0.0} +{"step": 191, "steps": 500, "lr": 2e-05, "grad_norm": 0.6565166711807251, "loss_scale": 4.0, "retries": 0, "seconds": 355.5, "max_memory_gb": 17.34, "ce": 0.29807682731188834, "rank": 0.0} +{"step": 192, "steps": 500, "lr": 2e-05, "grad_norm": 0.7756606340408325, "loss_scale": 4.0, "retries": 0, "seconds": 357.4, "max_memory_gb": 17.34, "ce": 0.256171234883368, "rank": 0.0} +{"step": 193, "steps": 500, "lr": 2e-05, "grad_norm": 1.3401292562484741, "loss_scale": 4.0, "retries": 0, "seconds": 359.3, "max_memory_gb": 17.34, "ce": 0.1832860220456496, "rank": 0.12553638219833374} +{"step": 194, "steps": 500, "lr": 2e-05, "grad_norm": 1.386344075202942, "loss_scale": 4.0, "retries": 0, "seconds": 361.1, "max_memory_gb": 17.34, "ce": 0.34559836657717824, "rank": 0.0} +{"step": 195, "steps": 500, "lr": 2e-05, "grad_norm": 1.2668352127075195, "loss_scale": 4.0, "retries": 0, "seconds": 363.0, "max_memory_gb": 17.34, "ce": 0.2796035229694098, "rank": 0.0} +{"step": 196, "steps": 500, "lr": 2e-05, "grad_norm": 0.9861289262771606, "loss_scale": 4.0, "retries": 0, "seconds": 364.9, "max_memory_gb": 17.34, "ce": 0.1205839809263125, "rank": 0.0} +{"step": 197, "steps": 500, "lr": 2e-05, "grad_norm": 0.8410735726356506, "loss_scale": 4.0, "retries": 0, "seconds": 366.8, "max_memory_gb": 17.34, "ce": 0.1827125372365117, "rank": 0.06431664526462555} +{"step": 198, "steps": 500, "lr": 2e-05, "grad_norm": 0.7416362166404724, "loss_scale": 4.0, "retries": 0, "seconds": 368.6, "max_memory_gb": 17.34, "ce": 0.22903733258135617, "rank": 0.0} +{"step": 199, "steps": 500, "lr": 2e-05, "grad_norm": 2.502167224884033, "loss_scale": 4.0, "retries": 0, "seconds": 370.4, "max_memory_gb": 17.34, "ce": 0.235089342109859, "rank": 0.0} +{"step": 200, "steps": 500, "lr": 2e-05, "grad_norm": 1.0829073190689087, "loss_scale": 4.0, "retries": 0, "seconds": 372.3, "max_memory_gb": 17.34, "ce": 0.2789253043010831, "rank": 0.0} +{"step": 201, "steps": 500, "lr": 2e-05, "grad_norm": 0.8034685850143433, "loss_scale": 4.0, "retries": 0, "seconds": 374.2, "max_memory_gb": 17.34, "ce": 0.14277003076858819, "rank": 0.0} +{"step": 202, "steps": 500, "lr": 2e-05, "grad_norm": 0.9931910634040833, "loss_scale": 4.0, "retries": 0, "seconds": 375.9, "max_memory_gb": 17.34, "ce": 0.0595718365511857, "rank": 0.03137776255607605} +{"step": 203, "steps": 500, "lr": 2e-05, "grad_norm": 1.3835891485214233, "loss_scale": 4.0, "retries": 0, "seconds": 377.8, "max_memory_gb": 17.34, "ce": 0.19390899478457868, "rank": 0.0} +{"step": 204, "steps": 500, "lr": 2e-05, "grad_norm": 3.6708104610443115, "loss_scale": 4.0, "retries": 0, "seconds": 379.6, "max_memory_gb": 17.34, "ce": 0.2200972440186888, "rank": 0.0} +{"step": 205, "steps": 500, "lr": 2e-05, "grad_norm": 2.9087822437286377, "loss_scale": 4.0, "retries": 0, "seconds": 381.5, "max_memory_gb": 17.34, "ce": 0.24393037194386125, "rank": 0.0} +{"step": 206, "steps": 500, "lr": 2e-05, "grad_norm": 1.2892146110534668, "loss_scale": 4.0, "retries": 0, "seconds": 383.4, "max_memory_gb": 17.34, "ce": 0.129349582886789, "rank": 0.0} +{"step": 207, "steps": 500, "lr": 2e-05, "grad_norm": 1.8529707193374634, "loss_scale": 4.0, "retries": 0, "seconds": 385.2, "max_memory_gb": 17.34, "ce": 0.4761245851404965, "rank": 0.2198195457458496} +{"step": 208, "steps": 500, "lr": 2e-05, "grad_norm": 1.0772104263305664, "loss_scale": 4.0, "retries": 0, "seconds": 387.1, "max_memory_gb": 17.34, "ce": 0.06416087900288403, "rank": 0.0} +{"step": 209, "steps": 500, "lr": 2e-05, "grad_norm": 0.7448029518127441, "loss_scale": 4.0, "retries": 0, "seconds": 388.9, "max_memory_gb": 17.34, "ce": 0.21333458297885954, "rank": 0.12600569427013397} +{"step": 210, "steps": 500, "lr": 2e-05, "grad_norm": 0.7573855519294739, "loss_scale": 4.0, "retries": 0, "seconds": 390.8, "max_memory_gb": 17.34, "ce": 0.2585004387656227, "rank": 0.04663825035095215} +{"step": 211, "steps": 500, "lr": 2e-05, "grad_norm": 0.9080575704574585, "loss_scale": 4.0, "retries": 0, "seconds": 392.7, "max_memory_gb": 17.34, "ce": 0.2026229351758957, "rank": 0.0} +{"step": 212, "steps": 500, "lr": 2e-05, "grad_norm": 2.317002296447754, "loss_scale": 4.0, "retries": 0, "seconds": 394.5, "max_memory_gb": 17.34, "ce": 0.36710766004398465, "rank": 0.0} +{"step": 213, "steps": 500, "lr": 2e-05, "grad_norm": 0.8601547479629517, "loss_scale": 4.0, "retries": 0, "seconds": 396.4, "max_memory_gb": 17.34, "ce": 0.28829611954279244, "rank": 0.0} +{"step": 214, "steps": 500, "lr": 2e-05, "grad_norm": 1.518054723739624, "loss_scale": 4.0, "retries": 0, "seconds": 398.2, "max_memory_gb": 17.34, "ce": 0.31423384803929366, "rank": 0.0} +{"step": 215, "steps": 500, "lr": 2e-05, "grad_norm": 2.070561408996582, "loss_scale": 4.0, "retries": 0, "seconds": 400.1, "max_memory_gb": 17.34, "ce": 0.45855614077299833, "rank": 0.02138388156890869} +{"step": 216, "steps": 500, "lr": 2e-05, "grad_norm": 1.3538434505462646, "loss_scale": 4.0, "retries": 0, "seconds": 402.0, "max_memory_gb": 17.34, "ce": 0.1921133149880916, "rank": 0.0} +{"step": 217, "steps": 500, "lr": 2e-05, "grad_norm": 0.9765594005584717, "loss_scale": 4.0, "retries": 0, "seconds": 403.8, "max_memory_gb": 17.34, "ce": 0.0641074402956292, "rank": 0.0} +{"step": 218, "steps": 500, "lr": 2e-05, "grad_norm": 5.603617191314697, "loss_scale": 4.0, "retries": 0, "seconds": 405.6, "max_memory_gb": 17.34, "ce": 0.4095348829869181, "rank": 0.0} +{"step": 219, "steps": 500, "lr": 2e-05, "grad_norm": 3.348454475402832, "loss_scale": 4.0, "retries": 0, "seconds": 407.5, "max_memory_gb": 17.34, "ce": 0.17677055107196793, "rank": 0.0} +{"step": 220, "steps": 500, "lr": 2e-05, "grad_norm": 2.048604726791382, "loss_scale": 4.0, "retries": 0, "seconds": 409.3, "max_memory_gb": 17.34, "ce": 0.08304121671244502, "rank": 0.0} +{"step": 221, "steps": 500, "lr": 2e-05, "grad_norm": 3.259324073791504, "loss_scale": 4.0, "retries": 0, "seconds": 411.2, "max_memory_gb": 17.34, "ce": 0.14381992095150054, "rank": 0.0} +{"step": 222, "steps": 500, "lr": 2e-05, "grad_norm": 6.075520992279053, "loss_scale": 4.0, "retries": 0, "seconds": 413.1, "max_memory_gb": 17.34, "ce": 0.3210786384297535, "rank": 0.0} +{"step": 223, "steps": 500, "lr": 2e-05, "grad_norm": 0.749915599822998, "loss_scale": 4.0, "retries": 0, "seconds": 414.9, "max_memory_gb": 17.34, "ce": 0.1928064947715029, "rank": 0.0} +{"step": 224, "steps": 500, "lr": 2e-05, "grad_norm": 0.9201546311378479, "loss_scale": 4.0, "retries": 0, "seconds": 416.8, "max_memory_gb": 17.34, "ce": 0.10398166324011981, "rank": 0.08562833815813065} +{"step": 225, "steps": 500, "lr": 2e-05, "grad_norm": 1.3372266292572021, "loss_scale": 4.0, "retries": 0, "seconds": 418.6, "max_memory_gb": 17.34, "ce": 0.10742412344552577, "rank": 0.0} +{"step": 226, "steps": 500, "lr": 2e-05, "grad_norm": 0.6734748482704163, "loss_scale": 4.0, "retries": 0, "seconds": 420.5, "max_memory_gb": 17.34, "ce": 0.09767546609509736, "rank": 0.0} +{"step": 227, "steps": 500, "lr": 2e-05, "grad_norm": 0.8668078184127808, "loss_scale": 4.0, "retries": 0, "seconds": 422.3, "max_memory_gb": 17.34, "ce": 0.03858027423848398, "rank": 0.0} +{"step": 228, "steps": 500, "lr": 2e-05, "grad_norm": 0.6280785202980042, "loss_scale": 4.0, "retries": 0, "seconds": 424.2, "max_memory_gb": 17.34, "ce": 0.06973055959679186, "rank": 0.0} +{"step": 229, "steps": 500, "lr": 2e-05, "grad_norm": 0.8973547220230103, "loss_scale": 4.0, "retries": 0, "seconds": 426.0, "max_memory_gb": 17.34, "ce": 0.20045694592408836, "rank": 0.0} +{"step": 230, "steps": 500, "lr": 2e-05, "grad_norm": 0.6408532857894897, "loss_scale": 4.0, "retries": 0, "seconds": 427.8, "max_memory_gb": 17.34, "ce": 0.19323536139563657, "rank": 0.0} +{"step": 231, "steps": 500, "lr": 2e-05, "grad_norm": 0.7407932877540588, "loss_scale": 4.0, "retries": 0, "seconds": 429.6, "max_memory_gb": 17.34, "ce": 0.2092447051545605, "rank": 0.0} +{"step": 232, "steps": 500, "lr": 2e-05, "grad_norm": 1.3765215873718262, "loss_scale": 4.0, "retries": 0, "seconds": 431.5, "max_memory_gb": 17.34, "ce": 0.06123459408991039, "rank": 0.0} +{"step": 233, "steps": 500, "lr": 2e-05, "grad_norm": 9.919699668884277, "loss_scale": 4.0, "retries": 0, "seconds": 433.3, "max_memory_gb": 17.34, "ce": 0.2553487945115194, "rank": 0.0} +{"step": 234, "steps": 500, "lr": 2e-05, "grad_norm": 0.8200515508651733, "loss_scale": 4.0, "retries": 0, "seconds": 435.2, "max_memory_gb": 17.34, "ce": 0.16875017253914848, "rank": 0.0} +{"step": 235, "steps": 500, "lr": 2e-05, "grad_norm": 0.9340640306472778, "loss_scale": 4.0, "retries": 0, "seconds": 437.0, "max_memory_gb": 17.34, "ce": 0.29026115662418306, "rank": 0.0} +{"step": 236, "steps": 500, "lr": 2e-05, "grad_norm": 0.5853433609008789, "loss_scale": 4.0, "retries": 0, "seconds": 438.9, "max_memory_gb": 17.34, "ce": 0.09430809423793107, "rank": 0.0} +{"step": 237, "steps": 500, "lr": 2e-05, "grad_norm": 4.415462970733643, "loss_scale": 4.0, "retries": 0, "seconds": 440.7, "max_memory_gb": 17.34, "ce": 0.27535548782907426, "rank": 0.0} +{"step": 238, "steps": 500, "lr": 2e-05, "grad_norm": 1.6751765012741089, "loss_scale": 4.0, "retries": 0, "seconds": 442.5, "max_memory_gb": 17.34, "ce": 0.07592779136030003, "rank": 0.0} +{"step": 239, "steps": 500, "lr": 2e-05, "grad_norm": 0.8056338429450989, "loss_scale": 4.0, "retries": 0, "seconds": 444.3, "max_memory_gb": 17.34, "ce": 0.16119330358924344, "rank": 0.0} +{"step": 240, "steps": 500, "lr": 2e-05, "grad_norm": 2.0005123615264893, "loss_scale": 4.0, "retries": 0, "seconds": 446.2, "max_memory_gb": 17.34, "ce": 0.08855933760059997, "rank": 0.0} +{"step": 241, "steps": 500, "lr": 2e-05, "grad_norm": 0.6085051894187927, "loss_scale": 4.0, "retries": 0, "seconds": 448.0, "max_memory_gb": 17.34, "ce": 0.14193129423074424, "rank": 0.0} +{"step": 242, "steps": 500, "lr": 2e-05, "grad_norm": 0.9644417762756348, "loss_scale": 4.0, "retries": 0, "seconds": 449.9, "max_memory_gb": 17.34, "ce": 0.11349932989105582, "rank": 0.0} +{"step": 243, "steps": 500, "lr": 2e-05, "grad_norm": 3.619361639022827, "loss_scale": 4.0, "retries": 0, "seconds": 451.7, "max_memory_gb": 17.34, "ce": 0.17687902558827773, "rank": 0.0} +{"step": 244, "steps": 500, "lr": 2e-05, "grad_norm": 1.916208267211914, "loss_scale": 4.0, "retries": 0, "seconds": 453.6, "max_memory_gb": 17.34, "ce": 0.33709391657612287, "rank": 0.0} +{"step": 245, "steps": 500, "lr": 2e-05, "grad_norm": 0.7491885423660278, "loss_scale": 4.0, "retries": 0, "seconds": 455.4, "max_memory_gb": 17.34, "ce": 0.32795043187798, "rank": 0.0} +{"step": 246, "steps": 500, "lr": 2e-05, "grad_norm": 0.5591325163841248, "loss_scale": 4.0, "retries": 0, "seconds": 457.3, "max_memory_gb": 17.34, "ce": 0.016870082530658692, "rank": 0.0} +{"step": 247, "steps": 500, "lr": 2e-05, "grad_norm": 1.133212685585022, "loss_scale": 4.0, "retries": 0, "seconds": 459.2, "max_memory_gb": 17.34, "ce": 0.1262634122685995, "rank": 0.0} +{"step": 248, "steps": 500, "lr": 2e-05, "grad_norm": 1.107168197631836, "loss_scale": 4.0, "retries": 0, "seconds": 461.0, "max_memory_gb": 17.34, "ce": 0.0508955163822975, "rank": 0.0} +{"step": 249, "steps": 500, "lr": 2e-05, "grad_norm": 7.382717132568359, "loss_scale": 4.0, "retries": 0, "seconds": 462.9, "max_memory_gb": 17.34, "ce": 0.1840465494460659, "rank": 0.0} +{"step": 250, "steps": 500, "lr": 2e-05, "grad_norm": 2.2250843048095703, "loss_scale": 4.0, "retries": 0, "seconds": 464.7, "max_memory_gb": 17.34, "ce": 0.059715360184782185, "rank": 0.0} +{"step": 251, "steps": 500, "lr": 2e-05, "grad_norm": 0.3975410759449005, "loss_scale": 4.0, "retries": 0, "seconds": 466.6, "max_memory_gb": 17.34, "ce": 0.11052465728425886, "rank": 0.0} +{"step": 252, "steps": 500, "lr": 2e-05, "grad_norm": 1.1271790266036987, "loss_scale": 4.0, "retries": 0, "seconds": 468.4, "max_memory_gb": 17.34, "ce": 0.04752440105949063, "rank": 0.0} +{"step": 253, "steps": 500, "lr": 2e-05, "grad_norm": 0.63056880235672, "loss_scale": 4.0, "retries": 0, "seconds": 470.3, "max_memory_gb": 17.34, "ce": 0.10740775827434845, "rank": 0.0} +{"step": 254, "steps": 500, "lr": 2e-05, "grad_norm": 2.1917672157287598, "loss_scale": 4.0, "retries": 0, "seconds": 472.1, "max_memory_gb": 17.34, "ce": 0.10622823423909722, "rank": 0.0} +{"step": 255, "steps": 500, "lr": 2e-05, "grad_norm": 0.6396930813789368, "loss_scale": 4.0, "retries": 0, "seconds": 473.9, "max_memory_gb": 17.34, "ce": 0.18318049730442, "rank": 0.0} +{"step": 256, "steps": 500, "lr": 2e-05, "grad_norm": 0.5349525809288025, "loss_scale": 4.0, "retries": 0, "seconds": 475.8, "max_memory_gb": 17.34, "ce": 0.07598263745603617, "rank": 0.0} +{"step": 257, "steps": 500, "lr": 2e-05, "grad_norm": 0.9571841359138489, "loss_scale": 4.0, "retries": 0, "seconds": 477.6, "max_memory_gb": 17.34, "ce": 0.07427141077641863, "rank": 0.0} +{"step": 258, "steps": 500, "lr": 2e-05, "grad_norm": 2.497164487838745, "loss_scale": 4.0, "retries": 0, "seconds": 479.5, "max_memory_gb": 17.34, "ce": 0.09499312964908313, "rank": 0.0} +{"step": 259, "steps": 500, "lr": 2e-05, "grad_norm": 0.3720521926879883, "loss_scale": 4.0, "retries": 0, "seconds": 481.4, "max_memory_gb": 17.34, "ce": 0.02206908361404203, "rank": 0.0} +{"step": 260, "steps": 500, "lr": 2e-05, "grad_norm": 4.499895095825195, "loss_scale": 4.0, "retries": 0, "seconds": 483.2, "max_memory_gb": 17.34, "ce": 0.13419194641755894, "rank": 0.0} +{"step": 261, "steps": 500, "lr": 2e-05, "grad_norm": 1.2586864233016968, "loss_scale": 4.0, "retries": 0, "seconds": 485.1, "max_memory_gb": 17.34, "ce": 0.11472400341153843, "rank": 0.0} +{"step": 262, "steps": 500, "lr": 2e-05, "grad_norm": 2.1562652587890625, "loss_scale": 4.0, "retries": 0, "seconds": 486.9, "max_memory_gb": 17.34, "ce": 0.15164433862810256, "rank": 0.0} +{"step": 263, "steps": 500, "lr": 2e-05, "grad_norm": 5.148926258087158, "loss_scale": 4.0, "retries": 0, "seconds": 488.8, "max_memory_gb": 17.34, "ce": 0.21795749412558507, "rank": 0.1253519505262375} +{"step": 264, "steps": 500, "lr": 2e-05, "grad_norm": 1.2106125354766846, "loss_scale": 4.0, "retries": 0, "seconds": 490.6, "max_memory_gb": 17.34, "ce": 0.10453557528671809, "rank": 0.0} +{"step": 265, "steps": 500, "lr": 2e-05, "grad_norm": 0.8371900916099548, "loss_scale": 4.0, "retries": 0, "seconds": 492.5, "max_memory_gb": 17.34, "ce": 0.07557235169224441, "rank": 0.0} +{"step": 266, "steps": 500, "lr": 2e-05, "grad_norm": 6.7926130294799805, "loss_scale": 4.0, "retries": 0, "seconds": 494.3, "max_memory_gb": 17.34, "ce": 0.3915744062687736, "rank": 0.0} +{"step": 267, "steps": 500, "lr": 2e-05, "grad_norm": 0.49840545654296875, "loss_scale": 4.0, "retries": 0, "seconds": 496.2, "max_memory_gb": 17.34, "ce": 0.11562258111007395, "rank": 0.0} +{"step": 268, "steps": 500, "lr": 2e-05, "grad_norm": 0.5838438868522644, "loss_scale": 4.0, "retries": 0, "seconds": 498.0, "max_memory_gb": 17.34, "ce": 0.2157898852819926, "rank": 0.0} +{"step": 269, "steps": 500, "lr": 2e-05, "grad_norm": 8.16740894317627, "loss_scale": 4.0, "retries": 0, "seconds": 499.9, "max_memory_gb": 17.34, "ce": 0.5175845124558691, "rank": 0.0} +{"step": 270, "steps": 500, "lr": 2e-05, "grad_norm": 1.2925175428390503, "loss_scale": 4.0, "retries": 0, "seconds": 501.8, "max_memory_gb": 17.34, "ce": 0.13708406248042593, "rank": 0.0} +{"step": 271, "steps": 500, "lr": 2e-05, "grad_norm": 1.416272521018982, "loss_scale": 4.0, "retries": 0, "seconds": 503.6, "max_memory_gb": 17.34, "ce": 0.029431994234982994, "rank": 0.1072622686624527} +{"step": 272, "steps": 500, "lr": 2e-05, "grad_norm": 4.045141220092773, "loss_scale": 4.0, "retries": 0, "seconds": 505.4, "max_memory_gb": 17.34, "ce": 0.1555324305609247, "rank": 0.0} +{"step": 273, "steps": 500, "lr": 2e-05, "grad_norm": 16.555349349975586, "loss_scale": 4.0, "retries": 0, "seconds": 507.3, "max_memory_gb": 17.34, "ce": 0.07884757677675225, "rank": 0.12513667345046997} +{"step": 274, "steps": 500, "lr": 2e-05, "grad_norm": 1.0424755811691284, "loss_scale": 4.0, "retries": 0, "seconds": 509.1, "max_memory_gb": 17.34, "ce": 0.11081407456367742, "rank": 0.0} +{"step": 275, "steps": 500, "lr": 2e-05, "grad_norm": 53.072654724121094, "loss_scale": 4.0, "retries": 0, "seconds": 511.0, "max_memory_gb": 17.34, "ce": 1.2756557458415045, "rank": 0.060036540031433105} +{"step": 276, "steps": 500, "lr": 2e-05, "grad_norm": 0.392535537481308, "loss_scale": 4.0, "retries": 0, "seconds": 512.9, "max_memory_gb": 17.34, "ce": 0.006663260348432232, "rank": 0.0} +{"step": 277, "steps": 500, "lr": 2e-05, "grad_norm": 0.9178724884986877, "loss_scale": 4.0, "retries": 0, "seconds": 514.8, "max_memory_gb": 17.34, "ce": 0.018523387866480334, "rank": 0.0} +{"step": 278, "steps": 500, "lr": 2e-05, "grad_norm": 4.422728538513184, "loss_scale": 4.0, "retries": 0, "seconds": 516.6, "max_memory_gb": 17.34, "ce": 0.19309814904772793, "rank": 0.011425398290157318} +{"step": 279, "steps": 500, "lr": 2e-05, "grad_norm": 5.5496110916137695, "loss_scale": 4.0, "retries": 0, "seconds": 518.4, "max_memory_gb": 17.34, "ce": 0.047316996302470216, "rank": 0.0} +{"step": 280, "steps": 500, "lr": 2e-05, "grad_norm": 11.122288703918457, "loss_scale": 4.0, "retries": 0, "seconds": 520.3, "max_memory_gb": 17.34, "ce": 0.0435213052569452, "rank": 0.0} +{"step": 281, "steps": 500, "lr": 2e-05, "grad_norm": 1.8719722032546997, "loss_scale": 4.0, "retries": 0, "seconds": 522.1, "max_memory_gb": 17.34, "ce": 0.08953112626659276, "rank": 0.0} +{"step": 282, "steps": 500, "lr": 2e-05, "grad_norm": 0.44522956013679504, "loss_scale": 4.0, "retries": 0, "seconds": 524.0, "max_memory_gb": 17.34, "ce": 0.07045710649981629, "rank": 0.0} +{"step": 283, "steps": 500, "lr": 2e-05, "grad_norm": 2.8256335258483887, "loss_scale": 4.0, "retries": 0, "seconds": 525.9, "max_memory_gb": 17.34, "ce": 0.09898956472898135, "rank": 0.0} +{"step": 284, "steps": 500, "lr": 2e-05, "grad_norm": 3.0379891395568848, "loss_scale": 4.0, "retries": 0, "seconds": 527.7, "max_memory_gb": 17.34, "ce": 0.042860426618062775, "rank": 0.0} +{"step": 285, "steps": 500, "lr": 2e-05, "grad_norm": 2.806476593017578, "loss_scale": 4.0, "retries": 0, "seconds": 529.6, "max_memory_gb": 17.34, "ce": 0.12136647516308585, "rank": 0.0} +{"step": 286, "steps": 500, "lr": 2e-05, "grad_norm": 2.364327907562256, "loss_scale": 4.0, "retries": 0, "seconds": 531.5, "max_memory_gb": 17.34, "ce": 0.05102733644162072, "rank": 0.0} +{"step": 287, "steps": 500, "lr": 2e-05, "grad_norm": 3.122936248779297, "loss_scale": 4.0, "retries": 0, "seconds": 533.3, "max_memory_gb": 17.34, "ce": 0.12425178834564576, "rank": 0.0} +{"step": 288, "steps": 500, "lr": 2e-05, "grad_norm": 0.18813881278038025, "loss_scale": 4.0, "retries": 0, "seconds": 535.2, "max_memory_gb": 17.34, "ce": 0.004462519507796969, "rank": 0.0} +{"step": 289, "steps": 500, "lr": 2e-05, "grad_norm": 4.382970809936523, "loss_scale": 4.0, "retries": 0, "seconds": 537.1, "max_memory_gb": 17.34, "ce": 0.06989790933403128, "rank": 0.0} +{"step": 290, "steps": 500, "lr": 2e-05, "grad_norm": 0.8087422847747803, "loss_scale": 4.0, "retries": 0, "seconds": 538.9, "max_memory_gb": 17.34, "ce": 0.09983858262785361, "rank": 0.0} +{"step": 291, "steps": 500, "lr": 2e-05, "grad_norm": 2.2746241092681885, "loss_scale": 4.0, "retries": 0, "seconds": 540.8, "max_memory_gb": 17.34, "ce": 0.017319369135293528, "rank": 0.0} +{"step": 292, "steps": 500, "lr": 2e-05, "grad_norm": 21.2104549407959, "loss_scale": 4.0, "retries": 0, "seconds": 542.6, "max_memory_gb": 17.34, "ce": 0.2930839591617769, "rank": 0.0} +{"step": 293, "steps": 500, "lr": 2e-05, "grad_norm": 1.2876665592193604, "loss_scale": 4.0, "retries": 0, "seconds": 544.5, "max_memory_gb": 17.34, "ce": 0.056275898615240294, "rank": 0.0} +{"step": 294, "steps": 500, "lr": 2e-05, "grad_norm": 60.69575500488281, "loss_scale": 4.0, "retries": 0, "seconds": 546.3, "max_memory_gb": 17.34, "ce": 0.5484157470109494, "rank": 0.0} +{"step": 295, "steps": 500, "lr": 2e-05, "grad_norm": 84.977294921875, "loss_scale": 4.0, "retries": 0, "seconds": 548.1, "max_memory_gb": 17.34, "ce": 0.608986486106005, "rank": 0.0} +{"step": 296, "steps": 500, "lr": 2e-05, "grad_norm": 60.758331298828125, "loss_scale": 4.0, "retries": 0, "seconds": 550.0, "max_memory_gb": 17.34, "ce": 0.9311609644282726, "rank": 0.0} +{"step": 297, "steps": 500, "lr": 2e-05, "grad_norm": 1.3921319246292114, "loss_scale": 4.0, "retries": 0, "seconds": 551.8, "max_memory_gb": 17.34, "ce": 0.15032585952576483, "rank": 0.0} +{"step": 298, "steps": 500, "lr": 2e-05, "grad_norm": 7.869202136993408, "loss_scale": 4.0, "retries": 0, "seconds": 553.6, "max_memory_gb": 17.34, "ce": 0.4766404645924922, "rank": 0.0} +{"step": 299, "steps": 500, "lr": 2e-05, "grad_norm": 1.5491396188735962, "loss_scale": 4.0, "retries": 0, "seconds": 555.5, "max_memory_gb": 17.34, "ce": 0.01577224631637364, "rank": 0.0} +{"step": 300, "steps": 500, "lr": 2e-05, "grad_norm": 1.17289137840271, "loss_scale": 4.0, "retries": 0, "seconds": 557.3, "max_memory_gb": 17.34, "ce": 0.23615606129169464, "rank": 0.0} +{"step": 301, "steps": 500, "lr": 2e-05, "grad_norm": 0.761552095413208, "loss_scale": 4.0, "retries": 0, "seconds": 559.2, "max_memory_gb": 17.34, "ce": 0.08242276933378889, "rank": 0.0} +{"step": 302, "steps": 500, "lr": 2e-05, "grad_norm": 1.2307236194610596, "loss_scale": 4.0, "retries": 0, "seconds": 561.0, "max_memory_gb": 17.34, "ce": 0.021799718151669367, "rank": 0.0} +{"step": 303, "steps": 500, "lr": 2e-05, "grad_norm": 6.098043918609619, "loss_scale": 4.0, "retries": 0, "seconds": 562.9, "max_memory_gb": 17.34, "ce": 0.027307792792271357, "rank": 0.0} +{"step": 304, "steps": 500, "lr": 2e-05, "grad_norm": 10.1454439163208, "loss_scale": 4.0, "retries": 0, "seconds": 564.7, "max_memory_gb": 17.34, "ce": 0.04455410531409143, "rank": 0.03820393979549408} +{"step": 305, "steps": 500, "lr": 2e-05, "grad_norm": 5.6122846603393555, "loss_scale": 4.0, "retries": 0, "seconds": 566.6, "max_memory_gb": 17.34, "ce": 0.15553770636870468, "rank": 0.0} +{"step": 306, "steps": 500, "lr": 2e-05, "grad_norm": 2.2831172943115234, "loss_scale": 4.0, "retries": 0, "seconds": 568.4, "max_memory_gb": 17.34, "ce": 0.0809460513828526, "rank": 0.0} +{"step": 307, "steps": 500, "lr": 2e-05, "grad_norm": 1.7317625284194946, "loss_scale": 4.0, "retries": 0, "seconds": 570.3, "max_memory_gb": 17.34, "ce": 0.15506869851196825, "rank": 0.0} +{"step": 308, "steps": 500, "lr": 2e-05, "grad_norm": 5.008003234863281, "loss_scale": 4.0, "retries": 0, "seconds": 572.1, "max_memory_gb": 17.34, "ce": 0.29717080075351987, "rank": 0.0} +{"step": 309, "steps": 500, "lr": 2e-05, "grad_norm": 0.94606614112854, "loss_scale": 4.0, "retries": 0, "seconds": 573.9, "max_memory_gb": 17.34, "ce": 0.1582852971041575, "rank": 0.0} +{"step": 310, "steps": 500, "lr": 2e-05, "grad_norm": 0.8098495602607727, "loss_scale": 4.0, "retries": 0, "seconds": 575.8, "max_memory_gb": 17.34, "ce": 0.025597538025976974, "rank": 0.0} +{"step": 311, "steps": 500, "lr": 2e-05, "grad_norm": 0.5561753511428833, "loss_scale": 4.0, "retries": 0, "seconds": 577.5, "max_memory_gb": 17.34, "ce": 0.08626311526313657, "rank": 0.0} +{"step": 312, "steps": 500, "lr": 2e-05, "grad_norm": 0.18906624615192413, "loss_scale": 4.0, "retries": 0, "seconds": 579.4, "max_memory_gb": 17.34, "ce": 0.007878393225837499, "rank": 0.0} +{"step": 313, "steps": 500, "lr": 2e-05, "grad_norm": 2.631101131439209, "loss_scale": 4.0, "retries": 0, "seconds": 581.2, "max_memory_gb": 17.34, "ce": 0.2237603427220165, "rank": 0.0} +{"step": 314, "steps": 500, "lr": 2e-05, "grad_norm": 2.2495946884155273, "loss_scale": 4.0, "retries": 0, "seconds": 583.1, "max_memory_gb": 17.34, "ce": 0.23612044449509995, "rank": 0.0} +{"step": 315, "steps": 500, "lr": 2e-05, "grad_norm": 0.5563194751739502, "loss_scale": 4.0, "retries": 0, "seconds": 585.0, "max_memory_gb": 17.34, "ce": 0.046680157216542284, "rank": 0.11732988804578781} +{"step": 316, "steps": 500, "lr": 2e-05, "grad_norm": 2.9285848140716553, "loss_scale": 4.0, "retries": 0, "seconds": 586.8, "max_memory_gb": 17.34, "ce": 0.36076541422971786, "rank": 0.0} +{"step": 317, "steps": 500, "lr": 2e-05, "grad_norm": 0.9908443093299866, "loss_scale": 4.0, "retries": 0, "seconds": 588.6, "max_memory_gb": 17.34, "ce": 0.10166161117194861, "rank": 0.0} +{"step": 318, "steps": 500, "lr": 2e-05, "grad_norm": 46.586490631103516, "loss_scale": 4.0, "retries": 0, "seconds": 590.4, "max_memory_gb": 17.34, "ce": 0.49281153862466454, "rank": 0.0} +{"step": 319, "steps": 500, "lr": 2e-05, "grad_norm": 1.1926636695861816, "loss_scale": 4.0, "retries": 0, "seconds": 592.3, "max_memory_gb": 17.34, "ce": 0.06434003099457186, "rank": 0.0} +{"step": 320, "steps": 500, "lr": 2e-05, "grad_norm": 1.0834991931915283, "loss_scale": 4.0, "retries": 0, "seconds": 594.1, "max_memory_gb": 17.34, "ce": 0.08863621097498253, "rank": 0.0} +{"step": 321, "steps": 500, "lr": 2e-05, "grad_norm": 0.9659472107887268, "loss_scale": 4.0, "retries": 0, "seconds": 595.9, "max_memory_gb": 17.34, "ce": 0.08624778025841806, "rank": 0.12597067654132843} +{"step": 322, "steps": 500, "lr": 2e-05, "grad_norm": 0.9580357074737549, "loss_scale": 4.0, "retries": 0, "seconds": 597.7, "max_memory_gb": 17.34, "ce": 0.02657139447364898, "rank": 0.0} +{"step": 323, "steps": 500, "lr": 2e-05, "grad_norm": 5.051473140716553, "loss_scale": 4.0, "retries": 0, "seconds": 599.6, "max_memory_gb": 17.34, "ce": 0.036204482363245916, "rank": 0.0} +{"step": 324, "steps": 500, "lr": 2e-05, "grad_norm": 2.878511667251587, "loss_scale": 4.0, "retries": 0, "seconds": 601.4, "max_memory_gb": 17.34, "ce": 0.0374292377182428, "rank": 0.0} +{"step": 325, "steps": 500, "lr": 2e-05, "grad_norm": 0.9266024231910706, "loss_scale": 4.0, "retries": 0, "seconds": 603.2, "max_memory_gb": 17.34, "ce": 0.20401870049681747, "rank": 0.0} +{"step": 326, "steps": 500, "lr": 2e-05, "grad_norm": 0.5136226415634155, "loss_scale": 4.0, "retries": 0, "seconds": 605.1, "max_memory_gb": 17.34, "ce": 0.043212126149228425, "rank": 0.10066397488117218} +{"step": 327, "steps": 500, "lr": 2e-05, "grad_norm": 22.724140167236328, "loss_scale": 4.0, "retries": 0, "seconds": 607.0, "max_memory_gb": 17.34, "ce": 0.2863911049935268, "rank": 0.0} +{"step": 328, "steps": 500, "lr": 2e-05, "grad_norm": 1.1266201734542847, "loss_scale": 4.0, "retries": 0, "seconds": 608.8, "max_memory_gb": 17.34, "ce": 0.12858885750938498, "rank": 0.0} +{"step": 329, "steps": 500, "lr": 2e-05, "grad_norm": 1.0545539855957031, "loss_scale": 4.0, "retries": 0, "seconds": 610.6, "max_memory_gb": 17.34, "ce": 0.1680998750271101, "rank": 0.0} +{"step": 330, "steps": 500, "lr": 2e-05, "grad_norm": 7.459719181060791, "loss_scale": 4.0, "retries": 0, "seconds": 612.4, "max_memory_gb": 17.34, "ce": 0.1558542766924802, "rank": 0.0} +{"step": 331, "steps": 500, "lr": 2e-05, "grad_norm": 5.221827983856201, "loss_scale": 4.0, "retries": 0, "seconds": 614.3, "max_memory_gb": 17.34, "ce": 0.38593208943166246, "rank": 0.0} +{"step": 332, "steps": 500, "lr": 2e-05, "grad_norm": 4.293094635009766, "loss_scale": 4.0, "retries": 0, "seconds": 616.1, "max_memory_gb": 17.34, "ce": 0.1225420109813058, "rank": 0.0} +{"step": 333, "steps": 500, "lr": 2e-05, "grad_norm": 1.1566202640533447, "loss_scale": 4.0, "retries": 0, "seconds": 617.9, "max_memory_gb": 17.34, "ce": 0.22898772483131324, "rank": 0.0} +{"step": 334, "steps": 500, "lr": 2e-05, "grad_norm": 1.1266816854476929, "loss_scale": 4.0, "retries": 0, "seconds": 619.8, "max_memory_gb": 17.34, "ce": 0.11324565091672412, "rank": 0.0} +{"step": 335, "steps": 500, "lr": 2e-05, "grad_norm": 4.130053997039795, "loss_scale": 4.0, "retries": 0, "seconds": 621.6, "max_memory_gb": 17.34, "ce": 0.2840561842749594, "rank": 0.0} +{"step": 336, "steps": 500, "lr": 2e-05, "grad_norm": 0.3788195252418518, "loss_scale": 4.0, "retries": 0, "seconds": 623.5, "max_memory_gb": 17.34, "ce": 0.016324217976944055, "rank": 0.0} +{"step": 337, "steps": 500, "lr": 2e-05, "grad_norm": 4.982776165008545, "loss_scale": 4.0, "retries": 0, "seconds": 625.3, "max_memory_gb": 17.34, "ce": 0.2175947538789842, "rank": 0.0} +{"step": 338, "steps": 500, "lr": 2e-05, "grad_norm": 2.22872257232666, "loss_scale": 4.0, "retries": 0, "seconds": 627.1, "max_memory_gb": 17.34, "ce": 0.18712365610917914, "rank": 0.0} +{"step": 339, "steps": 500, "lr": 2e-05, "grad_norm": 0.8601744771003723, "loss_scale": 4.0, "retries": 0, "seconds": 629.0, "max_memory_gb": 17.34, "ce": 0.12051211105881521, "rank": 0.0} +{"step": 340, "steps": 500, "lr": 2e-05, "grad_norm": 2.3584420680999756, "loss_scale": 4.0, "retries": 0, "seconds": 630.8, "max_memory_gb": 17.34, "ce": 0.18905080757758697, "rank": 0.0} +{"step": 341, "steps": 500, "lr": 2e-05, "grad_norm": 6.58430814743042, "loss_scale": 4.0, "retries": 0, "seconds": 632.7, "max_memory_gb": 17.34, "ce": 0.14559729354914452, "rank": 0.0} +{"step": 342, "steps": 500, "lr": 2e-05, "grad_norm": 2.487567901611328, "loss_scale": 4.0, "retries": 0, "seconds": 634.5, "max_memory_gb": 17.34, "ce": 0.5585892506815071, "rank": 0.0} +{"step": 343, "steps": 500, "lr": 2e-05, "grad_norm": 2.6451046466827393, "loss_scale": 4.0, "retries": 0, "seconds": 636.3, "max_memory_gb": 17.34, "ce": 0.08528993225263548, "rank": 0.0} +{"step": 344, "steps": 500, "lr": 2e-05, "grad_norm": 0.3764834403991699, "loss_scale": 4.0, "retries": 0, "seconds": 638.1, "max_memory_gb": 17.34, "ce": 0.0147041233467462, "rank": 0.0} +{"step": 345, "steps": 500, "lr": 2e-05, "grad_norm": 0.7249931693077087, "loss_scale": 4.0, "retries": 0, "seconds": 640.0, "max_memory_gb": 17.34, "ce": 0.17721394154978043, "rank": 0.0} +{"step": 346, "steps": 500, "lr": 2e-05, "grad_norm": 1.293809175491333, "loss_scale": 4.0, "retries": 0, "seconds": 641.8, "max_memory_gb": 17.34, "ce": 0.21396889188326895, "rank": 0.0} +{"step": 347, "steps": 500, "lr": 2e-05, "grad_norm": 1.286513328552246, "loss_scale": 4.0, "retries": 0, "seconds": 643.6, "max_memory_gb": 17.34, "ce": 0.13031332765331172, "rank": 0.0} +{"step": 348, "steps": 500, "lr": 2e-05, "grad_norm": 1.1888200044631958, "loss_scale": 4.0, "retries": 0, "seconds": 645.4, "max_memory_gb": 17.34, "ce": 0.009017147347549326, "rank": 0.0} +{"step": 349, "steps": 500, "lr": 2e-05, "grad_norm": 0.2805410921573639, "loss_scale": 4.0, "retries": 0, "seconds": 647.3, "max_memory_gb": 17.34, "ce": 0.08921992167779536, "rank": 0.0} +{"step": 350, "steps": 500, "lr": 2e-05, "grad_norm": 1.0018231868743896, "loss_scale": 4.0, "retries": 0, "seconds": 649.1, "max_memory_gb": 17.34, "ce": 0.10929493565708981, "rank": 0.0} +{"step": 351, "steps": 500, "lr": 2e-05, "grad_norm": 3.1314282417297363, "loss_scale": 4.0, "retries": 0, "seconds": 650.9, "max_memory_gb": 17.34, "ce": 0.06526177510568232, "rank": 0.0} +{"step": 352, "steps": 500, "lr": 2e-05, "grad_norm": 2.9470858573913574, "loss_scale": 4.0, "retries": 0, "seconds": 652.7, "max_memory_gb": 17.34, "ce": 0.29648990483838134, "rank": 0.0} +{"step": 353, "steps": 500, "lr": 2e-05, "grad_norm": 0.8833968043327332, "loss_scale": 4.0, "retries": 0, "seconds": 654.6, "max_memory_gb": 17.34, "ce": 0.1242236015882554, "rank": 0.0} +{"step": 354, "steps": 500, "lr": 2e-05, "grad_norm": 1.5674148797988892, "loss_scale": 4.0, "retries": 0, "seconds": 656.4, "max_memory_gb": 17.34, "ce": 0.15354665950599156, "rank": 0.0} +{"step": 355, "steps": 500, "lr": 2e-05, "grad_norm": 0.4132229685783386, "loss_scale": 4.0, "retries": 0, "seconds": 658.2, "max_memory_gb": 17.34, "ce": 0.08631379464759448, "rank": 0.0} +{"step": 356, "steps": 500, "lr": 2e-05, "grad_norm": 1.1045359373092651, "loss_scale": 4.0, "retries": 0, "seconds": 660.1, "max_memory_gb": 17.34, "ce": 0.04585838828188571, "rank": 0.0} +{"step": 357, "steps": 500, "lr": 2e-05, "grad_norm": 0.7388846278190613, "loss_scale": 4.0, "retries": 0, "seconds": 662.0, "max_memory_gb": 17.34, "ce": 0.057269901658855815, "rank": 0.0} +{"step": 358, "steps": 500, "lr": 2e-05, "grad_norm": 1.5023192167282104, "loss_scale": 4.0, "retries": 0, "seconds": 663.8, "max_memory_gb": 17.34, "ce": 0.16696107587813458, "rank": 0.0} +{"step": 359, "steps": 500, "lr": 2e-05, "grad_norm": 14.603419303894043, "loss_scale": 4.0, "retries": 0, "seconds": 665.6, "max_memory_gb": 17.34, "ce": 0.08479294035760176, "rank": 0.0} +{"step": 360, "steps": 500, "lr": 2e-05, "grad_norm": 0.9804427623748779, "loss_scale": 4.0, "retries": 0, "seconds": 667.5, "max_memory_gb": 17.34, "ce": 0.047250377441741875, "rank": 0.12457840144634247} +{"step": 361, "steps": 500, "lr": 2e-05, "grad_norm": 3.7710492610931396, "loss_scale": 4.0, "retries": 0, "seconds": 669.3, "max_memory_gb": 17.34, "ce": 0.13946018855676812, "rank": 0.0} +{"step": 362, "steps": 500, "lr": 2e-05, "grad_norm": 8.205804824829102, "loss_scale": 4.0, "retries": 0, "seconds": 671.1, "max_memory_gb": 17.34, "ce": 0.3010010850625804, "rank": 0.0} +{"step": 363, "steps": 500, "lr": 2e-05, "grad_norm": 27.575010299682617, "loss_scale": 4.0, "retries": 0, "seconds": 672.9, "max_memory_gb": 17.34, "ce": 0.09809517811254409, "rank": 0.0} +{"step": 364, "steps": 500, "lr": 2e-05, "grad_norm": 21.586624145507812, "loss_scale": 4.0, "retries": 0, "seconds": 674.7, "max_memory_gb": 17.34, "ce": 0.29736104024641463, "rank": 0.0} +{"step": 365, "steps": 500, "lr": 2e-05, "grad_norm": 1.1286147832870483, "loss_scale": 4.0, "retries": 0, "seconds": 676.6, "max_memory_gb": 17.34, "ce": 0.11450276153027517, "rank": 0.0} +{"step": 366, "steps": 500, "lr": 2e-05, "grad_norm": 25.17925453186035, "loss_scale": 4.0, "retries": 0, "seconds": 678.5, "max_memory_gb": 17.34, "ce": 0.21160627226799988, "rank": 0.0} +{"step": 367, "steps": 500, "lr": 2e-05, "grad_norm": 0.6911321878433228, "loss_scale": 4.0, "retries": 0, "seconds": 680.3, "max_memory_gb": 17.34, "ce": 0.08471948381520633, "rank": 0.0} +{"step": 368, "steps": 500, "lr": 2e-05, "grad_norm": 1.1357589960098267, "loss_scale": 4.0, "retries": 0, "seconds": 682.0, "max_memory_gb": 17.34, "ce": 0.011735296021470276, "rank": 0.0} +{"step": 369, "steps": 500, "lr": 2e-05, "grad_norm": 0.6007646918296814, "loss_scale": 4.0, "retries": 0, "seconds": 683.9, "max_memory_gb": 17.34, "ce": 0.03278919929471158, "rank": 0.1169460266828537} +{"step": 370, "steps": 500, "lr": 2e-05, "grad_norm": 3.383709192276001, "loss_scale": 4.0, "retries": 0, "seconds": 685.7, "max_memory_gb": 17.34, "ce": 0.3557596093846769, "rank": 0.0} +{"step": 371, "steps": 500, "lr": 2e-05, "grad_norm": 9.005438804626465, "loss_scale": 4.0, "retries": 0, "seconds": 687.6, "max_memory_gb": 17.34, "ce": 0.13069418693794432, "rank": 0.0} +{"step": 372, "steps": 500, "lr": 2e-05, "grad_norm": 2.2397701740264893, "loss_scale": 4.0, "retries": 0, "seconds": 689.4, "max_memory_gb": 17.34, "ce": 0.039532598142614006, "rank": 0.0} +{"step": 373, "steps": 500, "lr": 2e-05, "grad_norm": 1.0411195755004883, "loss_scale": 4.0, "retries": 0, "seconds": 691.2, "max_memory_gb": 17.34, "ce": 0.08632211439726234, "rank": 0.0} +{"step": 374, "steps": 500, "lr": 2e-05, "grad_norm": 1.1083598136901855, "loss_scale": 4.0, "retries": 0, "seconds": 693.1, "max_memory_gb": 17.34, "ce": 0.14653713198913465, "rank": 0.0} +{"step": 375, "steps": 500, "lr": 2e-05, "grad_norm": 1.010820746421814, "loss_scale": 4.0, "retries": 0, "seconds": 694.9, "max_memory_gb": 17.34, "ce": 0.16094035880269075, "rank": 0.0} +{"step": 376, "steps": 500, "lr": 2e-05, "grad_norm": 0.6996742486953735, "loss_scale": 4.0, "retries": 0, "seconds": 696.7, "max_memory_gb": 17.34, "ce": 0.04582622005318626, "rank": 0.0} +{"step": 377, "steps": 500, "lr": 2e-05, "grad_norm": 0.8388099074363708, "loss_scale": 4.0, "retries": 0, "seconds": 698.5, "max_memory_gb": 17.34, "ce": 0.09166907857070328, "rank": 0.0} +{"step": 378, "steps": 500, "lr": 2e-05, "grad_norm": 3.096721887588501, "loss_scale": 4.0, "retries": 0, "seconds": 700.4, "max_memory_gb": 17.34, "ce": 0.05795493190680645, "rank": 0.0} +{"step": 379, "steps": 500, "lr": 2e-05, "grad_norm": 0.7953013181686401, "loss_scale": 4.0, "retries": 0, "seconds": 702.2, "max_memory_gb": 17.34, "ce": 0.10537371197005996, "rank": 0.0} +{"step": 380, "steps": 500, "lr": 2e-05, "grad_norm": 4.804739952087402, "loss_scale": 4.0, "retries": 0, "seconds": 704.0, "max_memory_gb": 17.34, "ce": 0.1802366774541042, "rank": 0.0} +{"step": 381, "steps": 500, "lr": 2e-05, "grad_norm": 0.9733048677444458, "loss_scale": 4.0, "retries": 0, "seconds": 705.8, "max_memory_gb": 17.34, "ce": 0.21195004818810048, "rank": 0.0} +{"step": 382, "steps": 500, "lr": 2e-05, "grad_norm": 2.997715711593628, "loss_scale": 4.0, "retries": 0, "seconds": 707.6, "max_memory_gb": 17.34, "ce": 0.3279915170237473, "rank": 0.0} +{"step": 383, "steps": 500, "lr": 2e-05, "grad_norm": 0.3352365493774414, "loss_scale": 4.0, "retries": 0, "seconds": 709.5, "max_memory_gb": 17.34, "ce": 0.07151329098587667, "rank": 0.0} +{"step": 384, "steps": 500, "lr": 2e-05, "grad_norm": 3.6418612003326416, "loss_scale": 4.0, "retries": 0, "seconds": 711.3, "max_memory_gb": 17.34, "ce": 0.02596671186029198, "rank": 0.0} +{"step": 385, "steps": 500, "lr": 2e-05, "grad_norm": 0.9091566801071167, "loss_scale": 4.0, "retries": 0, "seconds": 713.1, "max_memory_gb": 17.34, "ce": 0.12713730382802169, "rank": 0.0} +{"step": 386, "steps": 500, "lr": 2e-05, "grad_norm": 0.994623064994812, "loss_scale": 4.0, "retries": 0, "seconds": 715.0, "max_memory_gb": 17.34, "ce": 0.05900792195188842, "rank": 0.0} +{"step": 387, "steps": 500, "lr": 2e-05, "grad_norm": 0.41496607661247253, "loss_scale": 4.0, "retries": 0, "seconds": 716.8, "max_memory_gb": 17.34, "ce": 0.049384599843051546, "rank": 0.0} +{"step": 388, "steps": 500, "lr": 2e-05, "grad_norm": 0.40551090240478516, "loss_scale": 4.0, "retries": 0, "seconds": 718.7, "max_memory_gb": 17.34, "ce": 0.03433827618709984, "rank": 0.0} +{"step": 389, "steps": 500, "lr": 2e-05, "grad_norm": 8.620865821838379, "loss_scale": 4.0, "retries": 0, "seconds": 720.5, "max_memory_gb": 17.34, "ce": 0.018575707940271968, "rank": 0.0} +{"step": 390, "steps": 500, "lr": 2e-05, "grad_norm": 93.87813568115234, "loss_scale": 4.0, "retries": 0, "seconds": 722.4, "max_memory_gb": 17.34, "ce": 0.9061178841129731, "rank": 0.0} +{"step": 391, "steps": 500, "lr": 2e-05, "grad_norm": 19.554357528686523, "loss_scale": 4.0, "retries": 0, "seconds": 724.2, "max_memory_gb": 17.34, "ce": 0.2848788694832365, "rank": 0.0} +{"step": 392, "steps": 500, "lr": 2e-05, "grad_norm": 0.5216012001037598, "loss_scale": 4.0, "retries": 0, "seconds": 726.1, "max_memory_gb": 17.34, "ce": 0.0037477636242329027, "rank": 0.10785400122404099} +{"step": 393, "steps": 500, "lr": 2e-05, "grad_norm": 3.023916482925415, "loss_scale": 4.0, "retries": 0, "seconds": 727.9, "max_memory_gb": 17.34, "ce": 0.16515289733615646, "rank": 0.0} +{"step": 394, "steps": 500, "lr": 2e-05, "grad_norm": 1.365289568901062, "loss_scale": 4.0, "retries": 0, "seconds": 729.7, "max_memory_gb": 17.34, "ce": 0.12159343736630035, "rank": 0.0} +{"step": 395, "steps": 500, "lr": 2e-05, "grad_norm": 0.3781697750091553, "loss_scale": 4.0, "retries": 0, "seconds": 731.5, "max_memory_gb": 17.34, "ce": 0.009448444874124107, "rank": 0.0} +{"step": 396, "steps": 500, "lr": 2e-05, "grad_norm": 2.1499898433685303, "loss_scale": 4.0, "retries": 0, "seconds": 733.4, "max_memory_gb": 17.34, "ce": 0.039801148221158655, "rank": 0.0} +{"step": 397, "steps": 500, "lr": 2e-05, "grad_norm": 0.5992816090583801, "loss_scale": 4.0, "retries": 0, "seconds": 735.2, "max_memory_gb": 17.34, "ce": 0.10293281776421281, "rank": 0.0} +{"step": 398, "steps": 500, "lr": 2e-05, "grad_norm": 1.2631739377975464, "loss_scale": 4.0, "retries": 0, "seconds": 737.0, "max_memory_gb": 17.34, "ce": 0.06958363836474746, "rank": 0.0} +{"step": 399, "steps": 500, "lr": 2e-05, "grad_norm": 2.3746249675750732, "loss_scale": 4.0, "retries": 0, "seconds": 738.9, "max_memory_gb": 17.34, "ce": 0.15882082160760547, "rank": 0.0} +{"step": 400, "steps": 500, "lr": 2e-05, "grad_norm": 1.4205327033996582, "loss_scale": 4.0, "retries": 0, "seconds": 740.7, "max_memory_gb": 17.34, "ce": 0.02272329750667268, "rank": 0.0} +{"step": 401, "steps": 500, "lr": 2e-05, "grad_norm": 3.67144775390625, "loss_scale": 4.0, "retries": 0, "seconds": 742.6, "max_memory_gb": 17.34, "ce": 0.2162819989343916, "rank": 0.0} +{"step": 402, "steps": 500, "lr": 2e-05, "grad_norm": 0.6838245391845703, "loss_scale": 4.0, "retries": 0, "seconds": 744.4, "max_memory_gb": 17.34, "ce": 0.09186075167053787, "rank": 0.09395226836204529} +{"step": 403, "steps": 500, "lr": 2e-05, "grad_norm": 0.1065824031829834, "loss_scale": 4.0, "retries": 0, "seconds": 746.2, "max_memory_gb": 17.34, "ce": 0.0037097245831319015, "rank": 0.0} +{"step": 404, "steps": 500, "lr": 2e-05, "grad_norm": 1.4582864046096802, "loss_scale": 4.0, "retries": 0, "seconds": 748.1, "max_memory_gb": 17.34, "ce": 0.14540024491589065, "rank": 0.0} +{"step": 405, "steps": 500, "lr": 2e-05, "grad_norm": 0.3469785749912262, "loss_scale": 4.0, "retries": 0, "seconds": 749.9, "max_memory_gb": 17.34, "ce": 0.014419630738302658, "rank": 0.0} +{"step": 406, "steps": 500, "lr": 2e-05, "grad_norm": 1.122660756111145, "loss_scale": 4.0, "retries": 0, "seconds": 751.7, "max_memory_gb": 17.34, "ce": 0.12758209003550292, "rank": 0.0} +{"step": 407, "steps": 500, "lr": 2e-05, "grad_norm": 3.228698253631592, "loss_scale": 4.0, "retries": 0, "seconds": 753.5, "max_memory_gb": 17.34, "ce": 0.7629116468251596, "rank": 0.2567048668861389} +{"step": 408, "steps": 500, "lr": 2e-05, "grad_norm": 2.0232629776000977, "loss_scale": 4.0, "retries": 0, "seconds": 755.4, "max_memory_gb": 17.34, "ce": 0.3394136133365464, "rank": 0.0} +{"step": 409, "steps": 500, "lr": 2e-05, "grad_norm": 0.8321226239204407, "loss_scale": 4.0, "retries": 0, "seconds": 757.2, "max_memory_gb": 17.34, "ce": 0.1437430781961666, "rank": 0.0} +{"step": 410, "steps": 500, "lr": 2e-05, "grad_norm": 2.5880069732666016, "loss_scale": 4.0, "retries": 0, "seconds": 759.0, "max_memory_gb": 17.34, "ce": 0.05478407653981776, "rank": 0.0} +{"step": 411, "steps": 500, "lr": 2e-05, "grad_norm": 7.654393196105957, "loss_scale": 4.0, "retries": 0, "seconds": 760.9, "max_memory_gb": 17.34, "ce": 0.38033139138997285, "rank": 0.0} +{"step": 412, "steps": 500, "lr": 2e-05, "grad_norm": 0.4442804455757141, "loss_scale": 4.0, "retries": 0, "seconds": 762.7, "max_memory_gb": 17.34, "ce": 0.14386495797862153, "rank": 0.12501409649848938} +{"step": 413, "steps": 500, "lr": 2e-05, "grad_norm": 0.6021381616592407, "loss_scale": 4.0, "retries": 0, "seconds": 764.5, "max_memory_gb": 17.34, "ce": 0.02759965309041945, "rank": 0.0} +{"step": 414, "steps": 500, "lr": 2e-05, "grad_norm": 2.439058303833008, "loss_scale": 4.0, "retries": 0, "seconds": 766.3, "max_memory_gb": 17.34, "ce": 0.18319109968570046, "rank": 0.0} +{"step": 415, "steps": 500, "lr": 2e-05, "grad_norm": 4.146321773529053, "loss_scale": 4.0, "retries": 0, "seconds": 768.1, "max_memory_gb": 17.34, "ce": 0.37016547873372474, "rank": 0.0} +{"step": 416, "steps": 500, "lr": 2e-05, "grad_norm": 2.072707414627075, "loss_scale": 4.0, "retries": 0, "seconds": 769.9, "max_memory_gb": 17.34, "ce": 0.015571963958791457, "rank": 0.0} +{"step": 417, "steps": 500, "lr": 2e-05, "grad_norm": 2.316887140274048, "loss_scale": 4.0, "retries": 0, "seconds": 771.8, "max_memory_gb": 17.34, "ce": 0.40388448274097755, "rank": 0.12773555517196655} +{"step": 418, "steps": 500, "lr": 2e-05, "grad_norm": 1.0927759408950806, "loss_scale": 4.0, "retries": 0, "seconds": 773.6, "max_memory_gb": 17.34, "ce": 0.05952636644815357, "rank": 0.0} +{"step": 419, "steps": 500, "lr": 2e-05, "grad_norm": 0.3958776295185089, "loss_scale": 4.0, "retries": 0, "seconds": 775.3, "max_memory_gb": 17.34, "ce": 0.06277707562730939, "rank": 0.0} +{"step": 420, "steps": 500, "lr": 2e-05, "grad_norm": 4.074398040771484, "loss_scale": 4.0, "retries": 0, "seconds": 777.2, "max_memory_gb": 17.34, "ce": 0.31505170743002964, "rank": 0.0} +{"step": 421, "steps": 500, "lr": 2e-05, "grad_norm": 0.3188856542110443, "loss_scale": 4.0, "retries": 0, "seconds": 779.0, "max_memory_gb": 17.34, "ce": 0.023096254852589482, "rank": 0.1218167394399643} +{"step": 422, "steps": 500, "lr": 2e-05, "grad_norm": 4.856051445007324, "loss_scale": 4.0, "retries": 0, "seconds": 780.8, "max_memory_gb": 17.34, "ce": 0.1811032938653625, "rank": 0.0} +{"step": 423, "steps": 500, "lr": 2e-05, "grad_norm": 3.3239688873291016, "loss_scale": 4.0, "retries": 0, "seconds": 782.6, "max_memory_gb": 17.34, "ce": 0.33101151282335195, "rank": 0.0} +{"step": 424, "steps": 500, "lr": 2e-05, "grad_norm": 0.5980757474899292, "loss_scale": 4.0, "retries": 0, "seconds": 784.5, "max_memory_gb": 17.34, "ce": 0.03143810033043337, "rank": 0.0} +{"step": 425, "steps": 500, "lr": 2e-05, "grad_norm": 0.8253774642944336, "loss_scale": 4.0, "retries": 0, "seconds": 786.3, "max_memory_gb": 17.34, "ce": 0.04253955126159781, "rank": 0.0} +{"step": 426, "steps": 500, "lr": 2e-05, "grad_norm": 1.5868449211120605, "loss_scale": 4.0, "retries": 0, "seconds": 788.2, "max_memory_gb": 17.34, "ce": 0.15538119587199617, "rank": 0.0} +{"step": 427, "steps": 500, "lr": 2e-05, "grad_norm": 0.6656688451766968, "loss_scale": 4.0, "retries": 0, "seconds": 790.1, "max_memory_gb": 17.34, "ce": 0.04288465254467155, "rank": 0.0} +{"step": 428, "steps": 500, "lr": 2e-05, "grad_norm": 2.561990976333618, "loss_scale": 4.0, "retries": 0, "seconds": 791.9, "max_memory_gb": 17.34, "ce": 0.05027310404102536, "rank": 0.0} +{"step": 429, "steps": 500, "lr": 2e-05, "grad_norm": 0.3197327256202698, "loss_scale": 4.0, "retries": 0, "seconds": 793.7, "max_memory_gb": 17.34, "ce": 0.006978986381454888, "rank": 0.0} +{"step": 430, "steps": 500, "lr": 2e-05, "grad_norm": 0.6917425990104675, "loss_scale": 4.0, "retries": 0, "seconds": 795.6, "max_memory_gb": 17.34, "ce": 0.08346220691373674, "rank": 0.0} +{"step": 431, "steps": 500, "lr": 2e-05, "grad_norm": 7.761377334594727, "loss_scale": 4.0, "retries": 0, "seconds": 797.5, "max_memory_gb": 17.34, "ce": 0.11454503195591315, "rank": 0.0} +{"step": 432, "steps": 500, "lr": 2e-05, "grad_norm": 0.7842202186584473, "loss_scale": 4.0, "retries": 0, "seconds": 799.3, "max_memory_gb": 17.34, "ce": 0.043962203903902264, "rank": 0.0} +{"step": 433, "steps": 500, "lr": 2e-05, "grad_norm": 41.963172912597656, "loss_scale": 4.0, "retries": 0, "seconds": 801.1, "max_memory_gb": 17.34, "ce": 0.1842884691659492, "rank": 0.0} +{"step": 434, "steps": 500, "lr": 2e-05, "grad_norm": 0.3364171087741852, "loss_scale": 4.0, "retries": 0, "seconds": 802.9, "max_memory_gb": 17.34, "ce": 0.003104721565705404, "rank": 0.0} +{"step": 435, "steps": 500, "lr": 2e-05, "grad_norm": 7.024860858917236, "loss_scale": 4.0, "retries": 0, "seconds": 804.8, "max_memory_gb": 17.34, "ce": 0.15944289791787014, "rank": 0.0} +{"step": 436, "steps": 500, "lr": 2e-05, "grad_norm": 0.9997687339782715, "loss_scale": 4.0, "retries": 0, "seconds": 806.6, "max_memory_gb": 17.34, "ce": 0.08051370709495131, "rank": 0.0} +{"step": 437, "steps": 500, "lr": 2e-05, "grad_norm": 19.01227569580078, "loss_scale": 4.0, "retries": 0, "seconds": 808.5, "max_memory_gb": 17.34, "ce": 0.3166374805518899, "rank": 0.12347191572189331} +{"step": 438, "steps": 500, "lr": 2e-05, "grad_norm": 0.7057813405990601, "loss_scale": 4.0, "retries": 0, "seconds": 810.3, "max_memory_gb": 17.34, "ce": 0.024912355141623266, "rank": 0.0} +{"step": 439, "steps": 500, "lr": 2e-05, "grad_norm": 1.5652024745941162, "loss_scale": 4.0, "retries": 0, "seconds": 812.1, "max_memory_gb": 17.34, "ce": 0.10929755596271207, "rank": 0.0} +{"step": 440, "steps": 500, "lr": 2e-05, "grad_norm": 0.5317223072052002, "loss_scale": 4.0, "retries": 0, "seconds": 814.0, "max_memory_gb": 17.34, "ce": 0.030648854376977397, "rank": 0.0} +{"step": 441, "steps": 500, "lr": 2e-05, "grad_norm": 11.373564720153809, "loss_scale": 4.0, "retries": 0, "seconds": 815.8, "max_memory_gb": 17.34, "ce": 0.2911179840289151, "rank": 0.0} +{"step": 442, "steps": 500, "lr": 2e-05, "grad_norm": 55.48529052734375, "loss_scale": 4.0, "retries": 0, "seconds": 817.6, "max_memory_gb": 17.34, "ce": 0.3120088058708461, "rank": 0.0} +{"step": 443, "steps": 500, "lr": 2e-05, "grad_norm": 0.7892128229141235, "loss_scale": 4.0, "retries": 0, "seconds": 819.4, "max_memory_gb": 17.34, "ce": 0.05191416407615179, "rank": 0.0} +{"step": 444, "steps": 500, "lr": 2e-05, "grad_norm": 0.7871862053871155, "loss_scale": 4.0, "retries": 0, "seconds": 821.3, "max_memory_gb": 17.34, "ce": 0.06304805837953609, "rank": 0.0} +{"step": 445, "steps": 500, "lr": 2e-05, "grad_norm": 2.2441117763519287, "loss_scale": 4.0, "retries": 0, "seconds": 823.1, "max_memory_gb": 17.34, "ce": 0.35402670257644786, "rank": 0.10522723942995071} +{"step": 446, "steps": 500, "lr": 2e-05, "grad_norm": 4.601688385009766, "loss_scale": 4.0, "retries": 0, "seconds": 824.9, "max_memory_gb": 17.34, "ce": 0.05334639875309222, "rank": 0.10438735783100128} +{"step": 447, "steps": 500, "lr": 2e-05, "grad_norm": 0.8564072251319885, "loss_scale": 4.0, "retries": 0, "seconds": 826.7, "max_memory_gb": 17.34, "ce": 0.13339206775208368, "rank": 0.0} +{"step": 448, "steps": 500, "lr": 2e-05, "grad_norm": 3.880969285964966, "loss_scale": 4.0, "retries": 0, "seconds": 828.5, "max_memory_gb": 17.34, "ce": 0.17765189440797258, "rank": 0.0} +{"step": 449, "steps": 500, "lr": 2e-05, "grad_norm": 0.5805427432060242, "loss_scale": 4.0, "retries": 0, "seconds": 830.3, "max_memory_gb": 17.34, "ce": 0.006921084057466942, "rank": 0.0} +{"step": 450, "steps": 500, "lr": 2e-05, "grad_norm": 4.290683269500732, "loss_scale": 4.0, "retries": 0, "seconds": 832.2, "max_memory_gb": 17.34, "ce": 0.04746835269997973, "rank": 0.0} +{"step": 451, "steps": 500, "lr": 2e-05, "grad_norm": 3.81318736076355, "loss_scale": 4.0, "retries": 0, "seconds": 834.0, "max_memory_gb": 17.34, "ce": 0.04717562434962019, "rank": 0.0} +{"step": 452, "steps": 500, "lr": 2e-05, "grad_norm": 5.543402194976807, "loss_scale": 4.0, "retries": 0, "seconds": 835.8, "max_memory_gb": 17.34, "ce": 0.18147414093164116, "rank": 0.0} +{"step": 453, "steps": 500, "lr": 2e-05, "grad_norm": 1.3408267498016357, "loss_scale": 4.0, "retries": 0, "seconds": 837.7, "max_memory_gb": 17.34, "ce": 0.034470662882995384, "rank": 0.0} +{"step": 454, "steps": 500, "lr": 2e-05, "grad_norm": 17.158750534057617, "loss_scale": 4.0, "retries": 0, "seconds": 839.5, "max_memory_gb": 17.34, "ce": 0.3688146025936021, "rank": 0.0} +{"step": 455, "steps": 500, "lr": 2e-05, "grad_norm": 5.678191661834717, "loss_scale": 4.0, "retries": 0, "seconds": 841.3, "max_memory_gb": 17.34, "ce": 0.1669801542782352, "rank": 0.0} +{"step": 456, "steps": 500, "lr": 2e-05, "grad_norm": 1.9306256771087646, "loss_scale": 4.0, "retries": 0, "seconds": 843.2, "max_memory_gb": 17.34, "ce": 0.00819802194064323, "rank": 0.0} +{"step": 457, "steps": 500, "lr": 2e-05, "grad_norm": 1.0588773488998413, "loss_scale": 4.0, "retries": 0, "seconds": 845.0, "max_memory_gb": 17.34, "ce": 0.027214252972953545, "rank": 0.10053952038288116} +{"step": 458, "steps": 500, "lr": 2e-05, "grad_norm": 8.309436798095703, "loss_scale": 4.0, "retries": 0, "seconds": 846.8, "max_memory_gb": 17.34, "ce": 0.2723228929489778, "rank": 0.0} +{"step": 459, "steps": 500, "lr": 2e-05, "grad_norm": 3.589798927307129, "loss_scale": 4.0, "retries": 0, "seconds": 848.7, "max_memory_gb": 17.34, "ce": 0.1472352467478686, "rank": 0.0} +{"step": 460, "steps": 500, "lr": 2e-05, "grad_norm": 0.26878657937049866, "loss_scale": 4.0, "retries": 0, "seconds": 850.6, "max_memory_gb": 17.34, "ce": 0.009476985623450673, "rank": 0.0} +{"step": 461, "steps": 500, "lr": 2e-05, "grad_norm": 2.716373920440674, "loss_scale": 4.0, "retries": 0, "seconds": 852.4, "max_memory_gb": 17.34, "ce": 0.2983079083405755, "rank": 0.0} +{"step": 462, "steps": 500, "lr": 2e-05, "grad_norm": 0.4147047698497772, "loss_scale": 4.0, "retries": 0, "seconds": 854.2, "max_memory_gb": 17.34, "ce": 0.014836394635267425, "rank": 0.0} +{"step": 463, "steps": 500, "lr": 2e-05, "grad_norm": 0.337294340133667, "loss_scale": 4.0, "retries": 0, "seconds": 856.0, "max_memory_gb": 17.34, "ce": 0.0101605681595629, "rank": 0.0} +{"step": 464, "steps": 500, "lr": 2e-05, "grad_norm": 0.2765999138355255, "loss_scale": 4.0, "retries": 0, "seconds": 857.9, "max_memory_gb": 17.34, "ce": 0.0024390628127548553, "rank": 0.0} +{"step": 465, "steps": 500, "lr": 2e-05, "grad_norm": 4.185530662536621, "loss_scale": 4.0, "retries": 0, "seconds": 859.8, "max_memory_gb": 17.34, "ce": 0.18193376112890292, "rank": 0.0} +{"step": 466, "steps": 500, "lr": 2e-05, "grad_norm": 3.40309476852417, "loss_scale": 4.0, "retries": 0, "seconds": 861.6, "max_memory_gb": 17.34, "ce": 0.11362397092329957, "rank": 0.0} +{"step": 467, "steps": 500, "lr": 2e-05, "grad_norm": 68.01908874511719, "loss_scale": 4.0, "retries": 0, "seconds": 863.4, "max_memory_gb": 17.34, "ce": 0.768818408885636, "rank": 0.0} +{"step": 468, "steps": 500, "lr": 2e-05, "grad_norm": 26.48828125, "loss_scale": 4.0, "retries": 0, "seconds": 865.3, "max_memory_gb": 17.34, "ce": 0.1723120135497993, "rank": 0.0} +{"step": 469, "steps": 500, "lr": 2e-05, "grad_norm": 97.63036346435547, "loss_scale": 4.0, "retries": 0, "seconds": 867.1, "max_memory_gb": 17.34, "ce": 1.1458019796314147, "rank": 0.0} +{"step": 470, "steps": 500, "lr": 2e-05, "grad_norm": 2.8553922176361084, "loss_scale": 4.0, "retries": 0, "seconds": 868.9, "max_memory_gb": 17.34, "ce": 0.12431578726182124, "rank": 0.0} +{"step": 471, "steps": 500, "lr": 2e-05, "grad_norm": 2.7178757190704346, "loss_scale": 4.0, "retries": 0, "seconds": 870.6, "max_memory_gb": 17.34, "ce": 0.2388116077891027, "rank": 0.0} +{"step": 472, "steps": 500, "lr": 2e-05, "grad_norm": 1.7286174297332764, "loss_scale": 4.0, "retries": 0, "seconds": 872.3, "max_memory_gb": 17.34, "ce": 0.006524579124743468, "rank": 0.0} +{"step": 473, "steps": 500, "lr": 2e-05, "grad_norm": 3.464618444442749, "loss_scale": 4.0, "retries": 0, "seconds": 874.0, "max_memory_gb": 17.34, "ce": 0.0355611567847518, "rank": 0.14619019627571106} +{"step": 474, "steps": 500, "lr": 2e-05, "grad_norm": 1.2432727813720703, "loss_scale": 4.0, "retries": 0, "seconds": 875.7, "max_memory_gb": 17.34, "ce": 0.09539469944684242, "rank": 0.0} +{"step": 475, "steps": 500, "lr": 2e-05, "grad_norm": 2.123135566711426, "loss_scale": 4.0, "retries": 0, "seconds": 877.4, "max_memory_gb": 17.34, "ce": 0.18949474462442595, "rank": 0.0} +{"step": 476, "steps": 500, "lr": 2e-05, "grad_norm": 0.1522102802991867, "loss_scale": 4.0, "retries": 0, "seconds": 879.1, "max_memory_gb": 17.34, "ce": 0.005044936334570593, "rank": 0.0} +{"step": 477, "steps": 500, "lr": 2e-05, "grad_norm": 3.7843637466430664, "loss_scale": 4.0, "retries": 0, "seconds": 880.8, "max_memory_gb": 17.34, "ce": 0.08477145088545512, "rank": 0.0} +{"step": 478, "steps": 500, "lr": 2e-05, "grad_norm": 1.3939625024795532, "loss_scale": 4.0, "retries": 0, "seconds": 882.5, "max_memory_gb": 17.34, "ce": 0.04953464338359481, "rank": 0.0} +{"step": 479, "steps": 500, "lr": 2e-05, "grad_norm": 5.954558849334717, "loss_scale": 4.0, "retries": 0, "seconds": 884.2, "max_memory_gb": 17.34, "ce": 0.3155568724978366, "rank": 0.05947369337081909} +{"step": 480, "steps": 500, "lr": 2e-05, "grad_norm": 0.678438663482666, "loss_scale": 4.0, "retries": 0, "seconds": 885.9, "max_memory_gb": 17.34, "ce": 0.11987719351918713, "rank": 0.0} +{"step": 481, "steps": 500, "lr": 2e-05, "grad_norm": 2.458392381668091, "loss_scale": 4.0, "retries": 0, "seconds": 887.7, "max_memory_gb": 17.34, "ce": 0.052368623758866306, "rank": 0.0} +{"step": 482, "steps": 500, "lr": 2e-05, "grad_norm": 9.250285148620605, "loss_scale": 4.0, "retries": 0, "seconds": 889.3, "max_memory_gb": 17.34, "ce": 0.11187134385545505, "rank": 0.0} +{"step": 483, "steps": 500, "lr": 2e-05, "grad_norm": 10.281015396118164, "loss_scale": 4.0, "retries": 0, "seconds": 891.0, "max_memory_gb": 17.34, "ce": 0.6257707646104791, "rank": 0.0} +{"step": 484, "steps": 500, "lr": 2e-05, "grad_norm": 2.286281108856201, "loss_scale": 4.0, "retries": 0, "seconds": 892.7, "max_memory_gb": 17.34, "ce": 0.4830640345298889, "rank": 0.0} +{"step": 485, "steps": 500, "lr": 2e-05, "grad_norm": 0.7361007332801819, "loss_scale": 4.0, "retries": 0, "seconds": 894.4, "max_memory_gb": 17.34, "ce": 0.03936191148841317, "rank": 0.0} +{"step": 486, "steps": 500, "lr": 2e-05, "grad_norm": 3.438387632369995, "loss_scale": 4.0, "retries": 0, "seconds": 896.1, "max_memory_gb": 17.34, "ce": 0.20501755702844093, "rank": 0.0} +{"step": 487, "steps": 500, "lr": 2e-05, "grad_norm": 20.312341690063477, "loss_scale": 4.0, "retries": 0, "seconds": 897.8, "max_memory_gb": 17.34, "ce": 0.2674434125924563, "rank": 0.0} +{"step": 488, "steps": 500, "lr": 2e-05, "grad_norm": 0.8764622807502747, "loss_scale": 4.0, "retries": 0, "seconds": 899.5, "max_memory_gb": 17.34, "ce": 0.0651863112771025, "rank": 0.0} +{"step": 489, "steps": 500, "lr": 2e-05, "grad_norm": 6.4064226150512695, "loss_scale": 4.0, "retries": 0, "seconds": 901.2, "max_memory_gb": 17.34, "ce": 0.26497039641481024, "rank": 0.0} +{"step": 490, "steps": 500, "lr": 2e-05, "grad_norm": 0.499362975358963, "loss_scale": 4.0, "retries": 0, "seconds": 903.0, "max_memory_gb": 17.34, "ce": 0.02291292746167528, "rank": 0.0} +{"step": 491, "steps": 500, "lr": 2e-05, "grad_norm": 0.6838342547416687, "loss_scale": 4.0, "retries": 0, "seconds": 904.8, "max_memory_gb": 17.34, "ce": 0.10434559711438851, "rank": 0.0} +{"step": 492, "steps": 500, "lr": 2e-05, "grad_norm": 0.3714657723903656, "loss_scale": 4.0, "retries": 0, "seconds": 906.7, "max_memory_gb": 17.34, "ce": 0.0057574651118557085, "rank": 0.0} +{"step": 493, "steps": 500, "lr": 2e-05, "grad_norm": 1.5839874744415283, "loss_scale": 4.0, "retries": 0, "seconds": 908.5, "max_memory_gb": 17.34, "ce": 0.048873500120862445, "rank": 0.0} +{"step": 494, "steps": 500, "lr": 2e-05, "grad_norm": 2.2061448097229004, "loss_scale": 4.0, "retries": 0, "seconds": 910.4, "max_memory_gb": 17.34, "ce": 0.09565279190064757, "rank": 0.0} +{"step": 495, "steps": 500, "lr": 2e-05, "grad_norm": 0.6851738691329956, "loss_scale": 4.0, "retries": 0, "seconds": 912.2, "max_memory_gb": 17.34, "ce": 0.024500371055182768, "rank": 0.0} +{"step": 496, "steps": 500, "lr": 2e-05, "grad_norm": 2.9791419506073, "loss_scale": 4.0, "retries": 0, "seconds": 914.0, "max_memory_gb": 17.34, "ce": 0.03250003759239917, "rank": 0.0} +{"step": 497, "steps": 500, "lr": 2e-05, "grad_norm": 2.781432867050171, "loss_scale": 4.0, "retries": 0, "seconds": 915.8, "max_memory_gb": 17.34, "ce": 0.15596125541742367, "rank": 0.0} +{"step": 498, "steps": 500, "lr": 2e-05, "grad_norm": 1.6822700500488281, "loss_scale": 4.0, "retries": 0, "seconds": 917.6, "max_memory_gb": 17.34, "ce": 0.114472691757328, "rank": 0.0} +{"step": 499, "steps": 500, "lr": 2e-05, "grad_norm": 3.096200704574585, "loss_scale": 4.0, "retries": 0, "seconds": 919.5, "max_memory_gb": 17.34, "ce": 0.18641262437586192, "rank": 0.0} +{"step": 500, "steps": 500, "lr": 2e-05, "grad_norm": 2.306658983230591, "loss_scale": 4.0, "retries": 0, "seconds": 921.3, "max_memory_gb": 17.34, "ce": 0.21509466995848925, "rank": 0.0} diff --git a/runs/single_llava/removal/L14_s13/adapter/README.md b/runs/single_llava/removal/L14_s13/adapter/README.md new file mode 100644 index 0000000000000000000000000000000000000000..ae587920ca55972d94e38f783837f71be14f1dda --- /dev/null +++ b/runs/single_llava/removal/L14_s13/adapter/README.md @@ -0,0 +1,207 @@ +--- +base_model: /data/OPSD/assets/models/llava-1.5-7b-hf +library_name: peft +pipeline_tag: text-generation +tags: +- base_model:adapter:/data/OPSD/assets/models/llava-1.5-7b-hf +- lora +- transformers +--- + +# Model Card for Model ID + + + + + +## Model Details + +### Model Description + + + + + +- **Developed by:** [More Information Needed] +- **Funded by [optional]:** [More Information Needed] +- **Shared by [optional]:** [More Information Needed] +- **Model type:** [More Information Needed] +- **Language(s) (NLP):** [More Information Needed] +- **License:** [More Information Needed] +- **Finetuned from model [optional]:** [More Information Needed] + +### Model Sources [optional] + + + +- **Repository:** [More Information Needed] +- **Paper [optional]:** [More Information Needed] +- **Demo [optional]:** [More Information Needed] + +## Uses + + + +### Direct Use + + + +[More Information Needed] + +### Downstream Use [optional] + + + +[More Information Needed] + +### Out-of-Scope Use + + + +[More Information Needed] + +## Bias, Risks, and Limitations + + + +[More Information Needed] + +### Recommendations + + + +Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. More information needed for further recommendations. + +## How to Get Started with the Model + +Use the code below to get started with the model. + +[More Information Needed] + +## Training Details + +### Training Data + + + +[More Information Needed] + +### Training Procedure + + + +#### Preprocessing [optional] + +[More Information Needed] + + +#### Training Hyperparameters + +- **Training regime:** [More Information Needed] + +#### Speeds, Sizes, Times [optional] + + + +[More Information Needed] + +## Evaluation + + + +### Testing Data, Factors & Metrics + +#### Testing Data + + + +[More Information Needed] + +#### Factors + + + +[More Information Needed] + +#### Metrics + + + +[More Information Needed] + +### Results + +[More Information Needed] + +#### Summary + + + +## Model Examination [optional] + + + +[More Information Needed] + +## Environmental Impact + + + +Carbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700). + +- **Hardware Type:** [More Information Needed] +- **Hours used:** [More Information Needed] +- **Cloud Provider:** [More Information Needed] +- **Compute Region:** [More Information Needed] +- **Carbon Emitted:** [More Information Needed] + +## Technical Specifications [optional] + +### Model Architecture and Objective + +[More Information Needed] + +### Compute Infrastructure + +[More Information Needed] + +#### Hardware + +[More Information Needed] + +#### Software + +[More Information Needed] + +## Citation [optional] + + + +**BibTeX:** + +[More Information Needed] + +**APA:** + +[More Information Needed] + +## Glossary [optional] + + + +[More Information Needed] + +## More Information [optional] + +[More Information Needed] + +## Model Card Authors [optional] + +[More Information Needed] + +## Model Card Contact + +[More Information Needed] +### Framework versions + +- PEFT 0.20.0 \ No newline at end of file diff --git a/runs/single_llava/removal/L14_s13/adapter/adapter_config.json b/runs/single_llava/removal/L14_s13/adapter/adapter_config.json new file mode 100644 index 0000000000000000000000000000000000000000..a089d45f59435e448888e0882f02a276b6e64e92 --- /dev/null +++ b/runs/single_llava/removal/L14_s13/adapter/adapter_config.json @@ -0,0 +1,50 @@ +{ + "alora_invocation_tokens": null, + "alpha_pattern": {}, + "arrow_config": null, + "auto_mapping": null, + "base_model_name_or_path": "/data/OPSD/assets/models/llava-1.5-7b-hf", + "bias": "none", + "corda_config": null, + "ensure_weight_tying": false, + "eva_config": null, + "exclude_modules": null, + "fan_in_fan_out": false, + "inference_mode": true, + "init_lora_weights": true, + "layer_replication": null, + "layers_pattern": null, + "layers_to_transform": null, + "loftq_config": {}, + "lora_alpha": 32, + "lora_bias": false, + "lora_dropout": 0.05, + "lora_ga_config": null, + "megatron_config": null, + "megatron_core": "megatron.core", + "modules_to_save": null, + "monteclora_config": null, + "peft_type": "LORA", + "peft_version": "0.20.0", + "qalora_group_size": 16, + "r": 16, + "rank_pattern": {}, + "revision": null, + "target_modules": [ + "model.language_model.layers.14.self_attn.q_proj", + "model.language_model.layers.14.mlp.gate_proj", + "model.language_model.layers.14.self_attn.k_proj", + "model.language_model.layers.14.self_attn.v_proj", + "model.language_model.layers.14.self_attn.o_proj", + "model.language_model.layers.14.mlp.up_proj", + "model.language_model.layers.14.mlp.down_proj" + ], + "target_parameters": null, + "task_type": "CAUSAL_LM", + "trainable_token_indices": null, + "use_bdlora": null, + "use_dora": false, + "use_qalora": false, + "use_rslora": false, + "velora_config": null +} \ No newline at end of file diff --git a/runs/single_llava/removal/L14_s13/config.json b/runs/single_llava/removal/L14_s13/config.json new file mode 100644 index 0000000000000000000000000000000000000000..4b3e59e431b776323984890855907754027b6ad3 --- /dev/null +++ b/runs/single_llava/removal/L14_s13/config.json @@ -0,0 +1,24 @@ +{ + "behavior": "removal", + "seed": 13, + "model": "/data/OPSD/assets/models/llava-1.5-7b-hf", + "backbone": "llava", + "train": "/data/OPSD/assets/v3/removal.jsonl", + "items": 1000, + "steps": 500, + "epochs": 2, + "steps_per_epoch": 250, + "groups_per_step": 4, + "lr": 2e-05, + "scheduler": "constant", + "warmup_steps": 0, + "final_fraction": 1.0, + "weight_decay": 0.0, + "clip": 1.0, + "initial_loss_scale": 4.0, + "adapter": { + "layer": 14, + "rank": 16, + "trainable_parameters": 1249280 + } +} \ No newline at end of file diff --git a/runs/single_llava/removal/L14_s13/steps.jsonl b/runs/single_llava/removal/L14_s13/steps.jsonl new file mode 100644 index 0000000000000000000000000000000000000000..e12260546ad52639ca53961aacb24ae656e1cb42 --- /dev/null +++ b/runs/single_llava/removal/L14_s13/steps.jsonl @@ -0,0 +1,500 @@ +{"step": 1, "steps": 500, "lr": 2e-05, "grad_norm": 0.2773131728172302, "loss_scale": 4.0, "retries": 0, "seconds": 2.7, "max_memory_gb": 16.48, "ce": 1.6287612388841808, "rank": 0.13243506848812103} +{"step": 2, "steps": 500, "lr": 2e-05, "grad_norm": 0.38780099153518677, "loss_scale": 4.0, "retries": 0, "seconds": 4.4, "max_memory_gb": 16.5, "ce": 2.6754561831476167, "rank": 0.032886385917663574} +{"step": 3, "steps": 500, "lr": 2e-05, "grad_norm": 0.8079825639724731, "loss_scale": 4.0, "retries": 0, "seconds": 6.1, "max_memory_gb": 16.5, "ce": 5.275363207794726, "rank": 0.1491079330444336} +{"step": 4, "steps": 500, "lr": 2e-05, "grad_norm": 0.561833381652832, "loss_scale": 4.0, "retries": 0, "seconds": 7.9, "max_memory_gb": 16.5, "ce": 3.7984784776344895, "rank": 0.12459981441497803} +{"step": 5, "steps": 500, "lr": 2e-05, "grad_norm": 0.09158334136009216, "loss_scale": 4.0, "retries": 0, "seconds": 9.6, "max_memory_gb": 16.5, "ce": 0.13952161301858723, "rank": 0.0} +{"step": 6, "steps": 500, "lr": 2e-05, "grad_norm": 0.24575451016426086, "loss_scale": 4.0, "retries": 0, "seconds": 11.3, "max_memory_gb": 16.5, "ce": 1.5110722952522337, "rank": 0.0} +{"step": 7, "steps": 500, "lr": 2e-05, "grad_norm": 0.6078246235847473, "loss_scale": 4.0, "retries": 0, "seconds": 13.0, "max_memory_gb": 16.5, "ce": 3.453049830859527, "rank": 0.020612239837646484} +{"step": 8, "steps": 500, "lr": 2e-05, "grad_norm": 0.7272250652313232, "loss_scale": 4.0, "retries": 0, "seconds": 14.8, "max_memory_gb": 16.5, "ce": 3.696386425755918, "rank": 0.20221960544586182} +{"step": 9, "steps": 500, "lr": 2e-05, "grad_norm": 0.5842805504798889, "loss_scale": 4.0, "retries": 0, "seconds": 16.5, "max_memory_gb": 16.5, "ce": 2.824977034702897, "rank": 0.038536906242370605} +{"step": 10, "steps": 500, "lr": 2e-05, "grad_norm": 0.3473511338233948, "loss_scale": 4.0, "retries": 0, "seconds": 18.2, "max_memory_gb": 16.5, "ce": 1.6116704912856221, "rank": 0.0} +{"step": 11, "steps": 500, "lr": 2e-05, "grad_norm": 0.6505975127220154, "loss_scale": 4.0, "retries": 0, "seconds": 20.0, "max_memory_gb": 16.5, "ce": 2.610559182241559, "rank": 0.21189281344413757} +{"step": 12, "steps": 500, "lr": 2e-05, "grad_norm": 0.425699919462204, "loss_scale": 4.0, "retries": 0, "seconds": 21.7, "max_memory_gb": 16.5, "ce": 1.501316661015153, "rank": 0.047830939292907715} +{"step": 13, "steps": 500, "lr": 2e-05, "grad_norm": 0.669762909412384, "loss_scale": 4.0, "retries": 0, "seconds": 23.4, "max_memory_gb": 16.5, "ce": 2.6493245291057974, "rank": 0.0} +{"step": 14, "steps": 500, "lr": 2e-05, "grad_norm": 1.0010169744491577, "loss_scale": 4.0, "retries": 0, "seconds": 25.1, "max_memory_gb": 16.5, "ce": 3.8378463461995125, "rank": 0.14062678813934326} +{"step": 15, "steps": 500, "lr": 2e-05, "grad_norm": 0.7254139184951782, "loss_scale": 4.0, "retries": 0, "seconds": 26.8, "max_memory_gb": 16.5, "ce": 2.5725772586883977, "rank": 0.16427326202392578} +{"step": 16, "steps": 500, "lr": 2e-05, "grad_norm": 0.4707626700401306, "loss_scale": 4.0, "retries": 0, "seconds": 28.5, "max_memory_gb": 16.5, "ce": 1.4005448944517411, "rank": 0.0} +{"step": 17, "steps": 500, "lr": 2e-05, "grad_norm": 0.5050439238548279, "loss_scale": 4.0, "retries": 0, "seconds": 30.3, "max_memory_gb": 16.5, "ce": 1.573844586731866, "rank": 0.02655208110809326} +{"step": 18, "steps": 500, "lr": 2e-05, "grad_norm": 1.0579851865768433, "loss_scale": 4.0, "retries": 0, "seconds": 32.0, "max_memory_gb": 16.51, "ce": 3.53701997641474, "rank": 0.18357586860656738} +{"step": 19, "steps": 500, "lr": 2e-05, "grad_norm": 1.6879777908325195, "loss_scale": 4.0, "retries": 0, "seconds": 33.7, "max_memory_gb": 16.51, "ce": 5.188868144527078, "rank": 0.294500470161438} +{"step": 20, "steps": 500, "lr": 2e-05, "grad_norm": 0.03222459554672241, "loss_scale": 4.0, "retries": 0, "seconds": 35.4, "max_memory_gb": 16.51, "ce": 0.06548178702360019, "rank": 0.10908487439155579} +{"step": 21, "steps": 500, "lr": 2e-05, "grad_norm": 1.0848547220230103, "loss_scale": 4.0, "retries": 0, "seconds": 37.2, "max_memory_gb": 16.51, "ce": 2.677734460332431, "rank": 0.14258348941802979} +{"step": 22, "steps": 500, "lr": 2e-05, "grad_norm": 1.1367403268814087, "loss_scale": 4.0, "retries": 0, "seconds": 38.9, "max_memory_gb": 16.51, "ce": 2.869890375994146, "rank": 0.17191720008850098} +{"step": 23, "steps": 500, "lr": 2e-05, "grad_norm": 1.0945954322814941, "loss_scale": 4.0, "retries": 0, "seconds": 40.7, "max_memory_gb": 16.51, "ce": 2.6721896089729853, "rank": 0.11023151874542236} +{"step": 24, "steps": 500, "lr": 2e-05, "grad_norm": 1.924021601676941, "loss_scale": 4.0, "retries": 0, "seconds": 42.4, "max_memory_gb": 16.51, "ce": 4.857375995256007, "rank": 0.2582143545150757} +{"step": 25, "steps": 500, "lr": 2e-05, "grad_norm": 1.9733988046646118, "loss_scale": 4.0, "retries": 0, "seconds": 44.2, "max_memory_gb": 16.51, "ce": 3.9738069851882756, "rank": 0.02242434024810791} +{"step": 26, "steps": 500, "lr": 2e-05, "grad_norm": 2.5145089626312256, "loss_scale": 4.0, "retries": 0, "seconds": 45.9, "max_memory_gb": 16.51, "ce": 4.95492900442332, "rank": 0.04176628589630127} +{"step": 27, "steps": 500, "lr": 2e-05, "grad_norm": 2.0632660388946533, "loss_scale": 4.0, "retries": 0, "seconds": 47.6, "max_memory_gb": 16.51, "ce": 4.026045020669699, "rank": 0.44386279582977295} +{"step": 28, "steps": 500, "lr": 2e-05, "grad_norm": 1.382419466972351, "loss_scale": 4.0, "retries": 0, "seconds": 49.4, "max_memory_gb": 16.51, "ce": 2.6389994425699115, "rank": 0.0} +{"step": 29, "steps": 500, "lr": 2e-05, "grad_norm": 1.4064018726348877, "loss_scale": 4.0, "retries": 0, "seconds": 51.1, "max_memory_gb": 16.51, "ce": 2.5446630514925346, "rank": 0.0} +{"step": 30, "steps": 500, "lr": 2e-05, "grad_norm": 1.7253360748291016, "loss_scale": 4.0, "retries": 0, "seconds": 52.8, "max_memory_gb": 16.51, "ce": 2.580828149570152, "rank": 0.0} +{"step": 31, "steps": 500, "lr": 2e-05, "grad_norm": 2.1312973499298096, "loss_scale": 4.0, "retries": 0, "seconds": 54.6, "max_memory_gb": 16.51, "ce": 2.6137311002239585, "rank": 0.11950147151947021} +{"step": 32, "steps": 500, "lr": 2e-05, "grad_norm": 3.5653297901153564, "loss_scale": 4.0, "retries": 0, "seconds": 56.3, "max_memory_gb": 16.51, "ce": 4.917109162546694, "rank": 0.19384706020355225} +{"step": 33, "steps": 500, "lr": 2e-05, "grad_norm": 1.561360478401184, "loss_scale": 4.0, "retries": 0, "seconds": 58.2, "max_memory_gb": 16.51, "ce": 2.3553810049779713, "rank": 0.00788898766040802} +{"step": 34, "steps": 500, "lr": 2e-05, "grad_norm": 1.9661388397216797, "loss_scale": 4.0, "retries": 0, "seconds": 60.1, "max_memory_gb": 16.51, "ce": 2.5203948294511065, "rank": 0.0} +{"step": 35, "steps": 500, "lr": 2e-05, "grad_norm": 2.601287841796875, "loss_scale": 4.0, "retries": 0, "seconds": 62.0, "max_memory_gb": 16.51, "ce": 2.4296678444370627, "rank": 0.07752025127410889} +{"step": 36, "steps": 500, "lr": 2e-05, "grad_norm": 2.0594093799591064, "loss_scale": 4.0, "retries": 0, "seconds": 63.9, "max_memory_gb": 16.51, "ce": 2.356866265181452, "rank": 0.0} +{"step": 37, "steps": 500, "lr": 2e-05, "grad_norm": 2.3054699897766113, "loss_scale": 4.0, "retries": 0, "seconds": 65.7, "max_memory_gb": 16.51, "ce": 2.3536547515541315, "rank": 0.0} +{"step": 38, "steps": 500, "lr": 2e-05, "grad_norm": 4.452315807342529, "loss_scale": 4.0, "retries": 0, "seconds": 67.5, "max_memory_gb": 16.51, "ce": 4.3165745958685875, "rank": 0.08819806575775146} +{"step": 39, "steps": 500, "lr": 2e-05, "grad_norm": 1.9289467334747314, "loss_scale": 4.0, "retries": 0, "seconds": 69.2, "max_memory_gb": 16.51, "ce": 1.2625408972380683, "rank": 0.12335080653429031} +{"step": 40, "steps": 500, "lr": 2e-05, "grad_norm": 2.786717414855957, "loss_scale": 4.0, "retries": 0, "seconds": 70.9, "max_memory_gb": 16.51, "ce": 2.346717283129692, "rank": 0.0} +{"step": 41, "steps": 500, "lr": 2e-05, "grad_norm": 1.5009773969650269, "loss_scale": 4.0, "retries": 0, "seconds": 72.6, "max_memory_gb": 16.51, "ce": 1.3642082959413528, "rank": 0.0} +{"step": 42, "steps": 500, "lr": 2e-05, "grad_norm": 3.817460775375366, "loss_scale": 4.0, "retries": 0, "seconds": 74.3, "max_memory_gb": 16.51, "ce": 1.9654100793413818, "rank": 0.0} +{"step": 43, "steps": 500, "lr": 2e-05, "grad_norm": 3.7624876499176025, "loss_scale": 4.0, "retries": 0, "seconds": 76.0, "max_memory_gb": 16.51, "ce": 2.0340397846885026, "rank": 0.0} +{"step": 44, "steps": 500, "lr": 2e-05, "grad_norm": 1.9157383441925049, "loss_scale": 4.0, "retries": 0, "seconds": 77.7, "max_memory_gb": 16.51, "ce": 1.2112032240838744, "rank": 0.0} +{"step": 45, "steps": 500, "lr": 2e-05, "grad_norm": 2.225297212600708, "loss_scale": 4.0, "retries": 0, "seconds": 79.4, "max_memory_gb": 16.51, "ce": 1.1799565837718546, "rank": 0.0} +{"step": 46, "steps": 500, "lr": 2e-05, "grad_norm": 3.971395969390869, "loss_scale": 4.0, "retries": 0, "seconds": 81.1, "max_memory_gb": 16.51, "ce": 1.8894886309280992, "rank": 0.0} +{"step": 47, "steps": 500, "lr": 2e-05, "grad_norm": 4.998388767242432, "loss_scale": 4.0, "retries": 0, "seconds": 82.8, "max_memory_gb": 16.51, "ce": 1.5021733650937676, "rank": 0.0} +{"step": 48, "steps": 500, "lr": 2e-05, "grad_norm": 9.327701568603516, "loss_scale": 4.0, "retries": 0, "seconds": 84.5, "max_memory_gb": 16.51, "ce": 2.690298884175718, "rank": 0.0} +{"step": 49, "steps": 500, "lr": 2e-05, "grad_norm": 9.168512344360352, "loss_scale": 4.0, "retries": 0, "seconds": 86.2, "max_memory_gb": 16.51, "ce": 2.512010058388114, "rank": 0.0} +{"step": 50, "steps": 500, "lr": 2e-05, "grad_norm": 8.609078407287598, "loss_scale": 4.0, "retries": 0, "seconds": 88.0, "max_memory_gb": 16.51, "ce": 2.568505237228237, "rank": 0.0} +{"step": 51, "steps": 500, "lr": 2e-05, "grad_norm": 3.0080299377441406, "loss_scale": 4.0, "retries": 0, "seconds": 89.7, "max_memory_gb": 16.51, "ce": 1.092145437025465, "rank": 0.0} +{"step": 52, "steps": 500, "lr": 2e-05, "grad_norm": 7.364981651306152, "loss_scale": 4.0, "retries": 0, "seconds": 91.4, "max_memory_gb": 16.51, "ce": 1.4970896784216166, "rank": 0.0} +{"step": 53, "steps": 500, "lr": 2e-05, "grad_norm": 11.068557739257812, "loss_scale": 4.0, "retries": 0, "seconds": 93.1, "max_memory_gb": 16.51, "ce": 2.4818566469475627, "rank": 0.0} +{"step": 54, "steps": 500, "lr": 2e-05, "grad_norm": 12.28930950164795, "loss_scale": 4.0, "retries": 0, "seconds": 94.8, "max_memory_gb": 16.51, "ce": 2.1913063935935497, "rank": 0.0} +{"step": 55, "steps": 500, "lr": 2e-05, "grad_norm": 8.472159385681152, "loss_scale": 4.0, "retries": 0, "seconds": 96.5, "max_memory_gb": 16.51, "ce": 1.2704018699005246, "rank": 0.0} +{"step": 56, "steps": 500, "lr": 2e-05, "grad_norm": 17.206621170043945, "loss_scale": 4.0, "retries": 0, "seconds": 98.3, "max_memory_gb": 16.51, "ce": 2.4525602916255593, "rank": 0.0} +{"step": 57, "steps": 500, "lr": 2e-05, "grad_norm": 4.292053699493408, "loss_scale": 4.0, "retries": 0, "seconds": 100.2, "max_memory_gb": 16.51, "ce": 0.705160659737885, "rank": 0.001152530312538147} +{"step": 58, "steps": 500, "lr": 2e-05, "grad_norm": 7.764725685119629, "loss_scale": 4.0, "retries": 0, "seconds": 102.1, "max_memory_gb": 16.51, "ce": 1.0115552445640787, "rank": 0.0} +{"step": 59, "steps": 500, "lr": 2e-05, "grad_norm": 4.498819828033447, "loss_scale": 4.0, "retries": 0, "seconds": 104.1, "max_memory_gb": 16.51, "ce": 0.85160246421583, "rank": 0.0} +{"step": 60, "steps": 500, "lr": 2e-05, "grad_norm": 8.706705093383789, "loss_scale": 4.0, "retries": 0, "seconds": 105.9, "max_memory_gb": 16.51, "ce": 1.0517126985359937, "rank": 0.0} +{"step": 61, "steps": 500, "lr": 2e-05, "grad_norm": 9.868825912475586, "loss_scale": 4.0, "retries": 0, "seconds": 107.7, "max_memory_gb": 16.51, "ce": 1.0123684024438262, "rank": 0.0} +{"step": 62, "steps": 500, "lr": 2e-05, "grad_norm": 3.745725393295288, "loss_scale": 4.0, "retries": 0, "seconds": 109.3, "max_memory_gb": 16.51, "ce": 0.48835796653293073, "rank": 0.0} +{"step": 63, "steps": 500, "lr": 2e-05, "grad_norm": 7.801331996917725, "loss_scale": 4.0, "retries": 0, "seconds": 111.0, "max_memory_gb": 16.51, "ce": 1.1904388777911663, "rank": 0.0} +{"step": 64, "steps": 500, "lr": 2e-05, "grad_norm": 5.800236225128174, "loss_scale": 4.0, "retries": 0, "seconds": 112.8, "max_memory_gb": 16.51, "ce": 0.8760731858201325, "rank": 0.0} +{"step": 65, "steps": 500, "lr": 2e-05, "grad_norm": 6.8485894203186035, "loss_scale": 4.0, "retries": 0, "seconds": 114.5, "max_memory_gb": 16.51, "ce": 1.1218517431989312, "rank": 0.0} +{"step": 66, "steps": 500, "lr": 2e-05, "grad_norm": 2.359586477279663, "loss_scale": 4.0, "retries": 0, "seconds": 116.2, "max_memory_gb": 16.51, "ce": 0.4697715804213658, "rank": 0.14880136400461197} +{"step": 67, "steps": 500, "lr": 2e-05, "grad_norm": 3.8260557651519775, "loss_scale": 4.0, "retries": 0, "seconds": 118.0, "max_memory_gb": 16.51, "ce": 0.6897895922884345, "rank": 0.0} +{"step": 68, "steps": 500, "lr": 2e-05, "grad_norm": 1.7582288980484009, "loss_scale": 4.0, "retries": 0, "seconds": 119.7, "max_memory_gb": 16.51, "ce": 0.30424341175239533, "rank": 0.0} +{"step": 69, "steps": 500, "lr": 2e-05, "grad_norm": 1.6746565103530884, "loss_scale": 4.0, "retries": 0, "seconds": 121.4, "max_memory_gb": 16.51, "ce": 0.46643277164548635, "rank": 0.0} +{"step": 70, "steps": 500, "lr": 2e-05, "grad_norm": 0.782952606678009, "loss_scale": 4.0, "retries": 0, "seconds": 123.1, "max_memory_gb": 16.51, "ce": 0.3950719400309026, "rank": 0.0242576003074646} +{"step": 71, "steps": 500, "lr": 2e-05, "grad_norm": 1.061442255973816, "loss_scale": 4.0, "retries": 0, "seconds": 124.9, "max_memory_gb": 16.51, "ce": 0.4943994196364656, "rank": 0.0} +{"step": 72, "steps": 500, "lr": 2e-05, "grad_norm": 1.191076636314392, "loss_scale": 4.0, "retries": 0, "seconds": 126.6, "max_memory_gb": 16.51, "ce": 0.6398005066439509, "rank": 0.0} +{"step": 73, "steps": 500, "lr": 2e-05, "grad_norm": 0.34374916553497314, "loss_scale": 4.0, "retries": 0, "seconds": 128.3, "max_memory_gb": 16.51, "ce": 0.24582816404290497, "rank": 0.0} +{"step": 74, "steps": 500, "lr": 2e-05, "grad_norm": 0.4015076458454132, "loss_scale": 4.0, "retries": 0, "seconds": 130.0, "max_memory_gb": 16.51, "ce": 0.4021689665969461, "rank": 0.0} +{"step": 75, "steps": 500, "lr": 2e-05, "grad_norm": 0.6856693625450134, "loss_scale": 4.0, "retries": 0, "seconds": 131.7, "max_memory_gb": 16.51, "ce": 0.7278734077699482, "rank": 0.13927894830703735} +{"step": 76, "steps": 500, "lr": 2e-05, "grad_norm": 0.45379090309143066, "loss_scale": 4.0, "retries": 0, "seconds": 133.4, "max_memory_gb": 16.51, "ce": 0.43831220967695117, "rank": 0.0} +{"step": 77, "steps": 500, "lr": 2e-05, "grad_norm": 0.724863588809967, "loss_scale": 4.0, "retries": 0, "seconds": 135.1, "max_memory_gb": 16.51, "ce": 0.6304335361346602, "rank": 0.0} +{"step": 78, "steps": 500, "lr": 2e-05, "grad_norm": 0.4055073857307434, "loss_scale": 4.0, "retries": 0, "seconds": 136.8, "max_memory_gb": 16.51, "ce": 0.3843190660700202, "rank": 0.0} +{"step": 79, "steps": 500, "lr": 2e-05, "grad_norm": 0.34542950987815857, "loss_scale": 4.0, "retries": 0, "seconds": 138.5, "max_memory_gb": 16.51, "ce": 0.26541026774793863, "rank": 0.01376575231552124} +{"step": 80, "steps": 500, "lr": 2e-05, "grad_norm": 0.36688241362571716, "loss_scale": 4.0, "retries": 0, "seconds": 140.2, "max_memory_gb": 16.51, "ce": 0.3180338906822726, "rank": 0.0} +{"step": 81, "steps": 500, "lr": 2e-05, "grad_norm": 0.6414570808410645, "loss_scale": 4.0, "retries": 0, "seconds": 141.8, "max_memory_gb": 16.51, "ce": 0.5594677207991481, "rank": 0.0} +{"step": 82, "steps": 500, "lr": 2e-05, "grad_norm": 0.49205806851387024, "loss_scale": 4.0, "retries": 0, "seconds": 143.6, "max_memory_gb": 16.51, "ce": 0.45988303795456886, "rank": 0.0} +{"step": 83, "steps": 500, "lr": 2e-05, "grad_norm": 0.5427339673042297, "loss_scale": 4.0, "retries": 0, "seconds": 145.3, "max_memory_gb": 16.51, "ce": 0.525485846796073, "rank": 0.0} +{"step": 84, "steps": 500, "lr": 2e-05, "grad_norm": 0.4408751428127289, "loss_scale": 4.0, "retries": 0, "seconds": 147.0, "max_memory_gb": 16.51, "ce": 0.2786808271193877, "rank": 0.0} +{"step": 85, "steps": 500, "lr": 2e-05, "grad_norm": 0.35335272550582886, "loss_scale": 4.0, "retries": 0, "seconds": 148.7, "max_memory_gb": 16.51, "ce": 0.11658529483247548, "rank": 0.0} +{"step": 86, "steps": 500, "lr": 2e-05, "grad_norm": 0.5887908339500427, "loss_scale": 4.0, "retries": 0, "seconds": 150.5, "max_memory_gb": 16.51, "ce": 0.5498388293199241, "rank": 0.0} +{"step": 87, "steps": 500, "lr": 2e-05, "grad_norm": 0.5490954518318176, "loss_scale": 4.0, "retries": 0, "seconds": 152.1, "max_memory_gb": 16.51, "ce": 0.467508532339707, "rank": 0.0} +{"step": 88, "steps": 500, "lr": 2e-05, "grad_norm": 0.5607101917266846, "loss_scale": 4.0, "retries": 0, "seconds": 153.8, "max_memory_gb": 16.51, "ce": 0.4358143199933693, "rank": 0.019914239645004272} +{"step": 89, "steps": 500, "lr": 2e-05, "grad_norm": 0.22471801936626434, "loss_scale": 4.0, "retries": 0, "seconds": 155.5, "max_memory_gb": 16.51, "ce": 0.15787353058112785, "rank": 0.1245766207575798} +{"step": 90, "steps": 500, "lr": 2e-05, "grad_norm": 0.44386956095695496, "loss_scale": 4.0, "retries": 0, "seconds": 157.3, "max_memory_gb": 16.51, "ce": 0.3742218401748687, "rank": 0.0} +{"step": 91, "steps": 500, "lr": 2e-05, "grad_norm": 0.4024732708930969, "loss_scale": 4.0, "retries": 0, "seconds": 159.0, "max_memory_gb": 16.51, "ce": 0.34527009841986, "rank": 0.0} +{"step": 92, "steps": 500, "lr": 2e-05, "grad_norm": 0.2637414038181305, "loss_scale": 4.0, "retries": 0, "seconds": 160.7, "max_memory_gb": 16.51, "ce": 0.2712959065102041, "rank": 0.0} +{"step": 93, "steps": 500, "lr": 2e-05, "grad_norm": 0.4192883372306824, "loss_scale": 4.0, "retries": 0, "seconds": 162.4, "max_memory_gb": 16.51, "ce": 0.31667009775992483, "rank": 0.0} +{"step": 94, "steps": 500, "lr": 2e-05, "grad_norm": 0.549897313117981, "loss_scale": 4.0, "retries": 0, "seconds": 164.1, "max_memory_gb": 16.51, "ce": 0.48045128025114536, "rank": 0.007378831505775452} +{"step": 95, "steps": 500, "lr": 2e-05, "grad_norm": 0.5159491300582886, "loss_scale": 4.0, "retries": 0, "seconds": 165.9, "max_memory_gb": 16.51, "ce": 0.44810212799347937, "rank": 0.0} +{"step": 96, "steps": 500, "lr": 2e-05, "grad_norm": 0.7639573812484741, "loss_scale": 4.0, "retries": 0, "seconds": 167.6, "max_memory_gb": 16.51, "ce": 0.5239384165033698, "rank": 0.019991695880889893} +{"step": 97, "steps": 500, "lr": 2e-05, "grad_norm": 0.5129684805870056, "loss_scale": 4.0, "retries": 0, "seconds": 169.3, "max_memory_gb": 16.51, "ce": 0.3025734964758158, "rank": 0.0} +{"step": 98, "steps": 500, "lr": 2e-05, "grad_norm": 0.3329416811466217, "loss_scale": 4.0, "retries": 0, "seconds": 171.0, "max_memory_gb": 16.51, "ce": 0.2827885348815471, "rank": 0.0} +{"step": 99, "steps": 500, "lr": 2e-05, "grad_norm": 0.6391260027885437, "loss_scale": 4.0, "retries": 0, "seconds": 172.7, "max_memory_gb": 16.51, "ce": 0.2081295215175487, "rank": 0.0} +{"step": 100, "steps": 500, "lr": 2e-05, "grad_norm": 0.5516203045845032, "loss_scale": 4.0, "retries": 0, "seconds": 174.5, "max_memory_gb": 16.51, "ce": 0.48690900951623917, "rank": 0.0} +{"step": 101, "steps": 500, "lr": 2e-05, "grad_norm": 0.4955575466156006, "loss_scale": 4.0, "retries": 0, "seconds": 176.2, "max_memory_gb": 16.51, "ce": 0.4575759880244732, "rank": 0.0} +{"step": 102, "steps": 500, "lr": 2e-05, "grad_norm": 0.9835334420204163, "loss_scale": 4.0, "retries": 0, "seconds": 178.0, "max_memory_gb": 16.51, "ce": 0.5735487532801926, "rank": 0.0} +{"step": 103, "steps": 500, "lr": 2e-05, "grad_norm": 0.3610716164112091, "loss_scale": 4.0, "retries": 0, "seconds": 179.7, "max_memory_gb": 16.51, "ce": 0.2894171304651536, "rank": 0.11806254088878632} +{"step": 104, "steps": 500, "lr": 2e-05, "grad_norm": 0.250998318195343, "loss_scale": 4.0, "retries": 0, "seconds": 181.3, "max_memory_gb": 16.51, "ce": 0.20176483754767105, "rank": 0.07648836076259613} +{"step": 105, "steps": 500, "lr": 2e-05, "grad_norm": 0.5861350297927856, "loss_scale": 4.0, "retries": 0, "seconds": 183.1, "max_memory_gb": 16.51, "ce": 0.4392059948295355, "rank": 0.10623020678758621} +{"step": 106, "steps": 500, "lr": 2e-05, "grad_norm": 0.5454708337783813, "loss_scale": 4.0, "retries": 0, "seconds": 184.8, "max_memory_gb": 16.51, "ce": 0.44385976158082485, "rank": 0.0854387953877449} +{"step": 107, "steps": 500, "lr": 2e-05, "grad_norm": 0.35173389315605164, "loss_scale": 4.0, "retries": 0, "seconds": 186.6, "max_memory_gb": 16.51, "ce": 0.25346365000586957, "rank": 0.0} +{"step": 108, "steps": 500, "lr": 2e-05, "grad_norm": 0.40584537386894226, "loss_scale": 4.0, "retries": 0, "seconds": 188.3, "max_memory_gb": 16.51, "ce": 0.2558133064303547, "rank": 0.0} +{"step": 109, "steps": 500, "lr": 2e-05, "grad_norm": 0.6293227672576904, "loss_scale": 4.0, "retries": 0, "seconds": 190.0, "max_memory_gb": 16.51, "ce": 0.5873052654787898, "rank": 0.0} +{"step": 110, "steps": 500, "lr": 2e-05, "grad_norm": 0.671319842338562, "loss_scale": 4.0, "retries": 0, "seconds": 191.8, "max_memory_gb": 16.51, "ce": 0.3641334194689989, "rank": 0.0} +{"step": 111, "steps": 500, "lr": 2e-05, "grad_norm": 0.40846046805381775, "loss_scale": 4.0, "retries": 0, "seconds": 193.5, "max_memory_gb": 16.51, "ce": 0.30393592338077724, "rank": 0.0} +{"step": 112, "steps": 500, "lr": 2e-05, "grad_norm": 0.4941970109939575, "loss_scale": 4.0, "retries": 0, "seconds": 195.2, "max_memory_gb": 16.51, "ce": 0.28341240249574184, "rank": 0.0} +{"step": 113, "steps": 500, "lr": 2e-05, "grad_norm": 0.4950173795223236, "loss_scale": 4.0, "retries": 0, "seconds": 197.0, "max_memory_gb": 16.51, "ce": 0.3556237993761897, "rank": 0.0} +{"step": 114, "steps": 500, "lr": 2e-05, "grad_norm": 0.9045910835266113, "loss_scale": 4.0, "retries": 0, "seconds": 198.7, "max_memory_gb": 16.51, "ce": 0.6499516637995839, "rank": 0.1223105937242508} +{"step": 115, "steps": 500, "lr": 2e-05, "grad_norm": 0.6629231572151184, "loss_scale": 4.0, "retries": 0, "seconds": 200.4, "max_memory_gb": 16.51, "ce": 0.4140552072785795, "rank": 0.08283315598964691} +{"step": 116, "steps": 500, "lr": 2e-05, "grad_norm": 0.2960026264190674, "loss_scale": 4.0, "retries": 0, "seconds": 202.2, "max_memory_gb": 16.51, "ce": 0.14686114597134292, "rank": 0.0} +{"step": 117, "steps": 500, "lr": 2e-05, "grad_norm": 0.9352782368659973, "loss_scale": 4.0, "retries": 0, "seconds": 203.9, "max_memory_gb": 16.51, "ce": 0.39848035061731935, "rank": 0.0} +{"step": 118, "steps": 500, "lr": 2e-05, "grad_norm": 0.3058614432811737, "loss_scale": 4.0, "retries": 0, "seconds": 205.6, "max_memory_gb": 16.51, "ce": 0.2754635721212253, "rank": 0.0} +{"step": 119, "steps": 500, "lr": 2e-05, "grad_norm": 0.3602238595485687, "loss_scale": 4.0, "retries": 0, "seconds": 207.4, "max_memory_gb": 16.51, "ce": 0.2448038721922785, "rank": 0.0} +{"step": 120, "steps": 500, "lr": 2e-05, "grad_norm": 0.8184032440185547, "loss_scale": 4.0, "retries": 0, "seconds": 209.1, "max_memory_gb": 16.51, "ce": 0.36172419250942767, "rank": 0.0} +{"step": 121, "steps": 500, "lr": 2e-05, "grad_norm": 0.6044161915779114, "loss_scale": 4.0, "retries": 0, "seconds": 210.9, "max_memory_gb": 16.51, "ce": 0.30629024043446407, "rank": 0.0} +{"step": 122, "steps": 500, "lr": 2e-05, "grad_norm": 0.35599496960639954, "loss_scale": 4.0, "retries": 0, "seconds": 212.6, "max_memory_gb": 16.51, "ce": 0.3280811350559816, "rank": 0.0} +{"step": 123, "steps": 500, "lr": 2e-05, "grad_norm": 0.9515331387519836, "loss_scale": 4.0, "retries": 0, "seconds": 214.3, "max_memory_gb": 16.51, "ce": 0.5150456152623519, "rank": 0.0} +{"step": 124, "steps": 500, "lr": 2e-05, "grad_norm": 0.7037315964698792, "loss_scale": 4.0, "retries": 0, "seconds": 216.1, "max_memory_gb": 16.51, "ce": 0.3791079795919359, "rank": 0.0} +{"step": 125, "steps": 500, "lr": 2e-05, "grad_norm": 0.38108915090560913, "loss_scale": 4.0, "retries": 0, "seconds": 217.8, "max_memory_gb": 16.51, "ce": 0.2633617885876447, "rank": 0.0} +{"step": 126, "steps": 500, "lr": 2e-05, "grad_norm": 0.5921306014060974, "loss_scale": 4.0, "retries": 0, "seconds": 219.5, "max_memory_gb": 16.51, "ce": 0.483893902041018, "rank": 0.0} +{"step": 127, "steps": 500, "lr": 2e-05, "grad_norm": 0.2557792663574219, "loss_scale": 4.0, "retries": 0, "seconds": 221.2, "max_memory_gb": 16.51, "ce": 0.14912502386141568, "rank": 0.0} +{"step": 128, "steps": 500, "lr": 2e-05, "grad_norm": 0.5982993841171265, "loss_scale": 4.0, "retries": 0, "seconds": 222.9, "max_memory_gb": 16.51, "ce": 0.34488543268525973, "rank": 0.0} +{"step": 129, "steps": 500, "lr": 2e-05, "grad_norm": 0.6596189737319946, "loss_scale": 4.0, "retries": 0, "seconds": 224.6, "max_memory_gb": 16.51, "ce": 0.439601878169924, "rank": 0.0} +{"step": 130, "steps": 500, "lr": 2e-05, "grad_norm": 0.9885342717170715, "loss_scale": 4.0, "retries": 0, "seconds": 226.3, "max_memory_gb": 16.51, "ce": 0.48800500156357884, "rank": 0.0} +{"step": 131, "steps": 500, "lr": 2e-05, "grad_norm": 0.8387547135353088, "loss_scale": 4.0, "retries": 0, "seconds": 228.0, "max_memory_gb": 16.51, "ce": 0.3243069760501385, "rank": 0.0} +{"step": 132, "steps": 500, "lr": 2e-05, "grad_norm": 1.1623399257659912, "loss_scale": 4.0, "retries": 0, "seconds": 229.8, "max_memory_gb": 16.51, "ce": 0.5141165700624697, "rank": 0.017257660627365112} +{"step": 133, "steps": 500, "lr": 2e-05, "grad_norm": 0.8015123009681702, "loss_scale": 4.0, "retries": 0, "seconds": 231.6, "max_memory_gb": 16.51, "ce": 0.41796410316601396, "rank": 0.0} +{"step": 134, "steps": 500, "lr": 2e-05, "grad_norm": 0.5350934267044067, "loss_scale": 4.0, "retries": 0, "seconds": 233.3, "max_memory_gb": 16.51, "ce": 0.15856465534307063, "rank": 0.0} +{"step": 135, "steps": 500, "lr": 2e-05, "grad_norm": 0.5420888066291809, "loss_scale": 4.0, "retries": 0, "seconds": 235.0, "max_memory_gb": 16.51, "ce": 0.3582624145783484, "rank": 0.0} +{"step": 136, "steps": 500, "lr": 2e-05, "grad_norm": 0.6922813653945923, "loss_scale": 4.0, "retries": 0, "seconds": 236.7, "max_memory_gb": 16.51, "ce": 0.3687397539615631, "rank": 0.0} +{"step": 137, "steps": 500, "lr": 2e-05, "grad_norm": 0.6455519199371338, "loss_scale": 4.0, "retries": 0, "seconds": 238.5, "max_memory_gb": 16.51, "ce": 0.30397528293542564, "rank": 0.0} +{"step": 138, "steps": 500, "lr": 2e-05, "grad_norm": 0.5036124587059021, "loss_scale": 4.0, "retries": 0, "seconds": 240.2, "max_memory_gb": 16.51, "ce": 0.301345644518733, "rank": 0.0} +{"step": 139, "steps": 500, "lr": 2e-05, "grad_norm": 0.8925439119338989, "loss_scale": 4.0, "retries": 0, "seconds": 242.0, "max_memory_gb": 16.51, "ce": 0.415852134115994, "rank": 0.0} +{"step": 140, "steps": 500, "lr": 2e-05, "grad_norm": 0.5041949152946472, "loss_scale": 4.0, "retries": 0, "seconds": 243.7, "max_memory_gb": 16.51, "ce": 0.22244380554184318, "rank": 0.0} +{"step": 141, "steps": 500, "lr": 2e-05, "grad_norm": 0.64067143201828, "loss_scale": 4.0, "retries": 0, "seconds": 245.5, "max_memory_gb": 16.51, "ce": 0.319744682405144, "rank": 0.0} +{"step": 142, "steps": 500, "lr": 2e-05, "grad_norm": 0.5342257022857666, "loss_scale": 4.0, "retries": 0, "seconds": 247.3, "max_memory_gb": 16.51, "ce": 0.2849706362467259, "rank": 0.0} +{"step": 143, "steps": 500, "lr": 2e-05, "grad_norm": 0.318604975938797, "loss_scale": 4.0, "retries": 0, "seconds": 249.0, "max_memory_gb": 16.51, "ce": 0.16601387699483894, "rank": 0.0} +{"step": 144, "steps": 500, "lr": 2e-05, "grad_norm": 0.8305810689926147, "loss_scale": 4.0, "retries": 0, "seconds": 250.8, "max_memory_gb": 16.51, "ce": 0.28333807119634, "rank": 0.0} +{"step": 145, "steps": 500, "lr": 2e-05, "grad_norm": 0.4319835603237152, "loss_scale": 4.0, "retries": 0, "seconds": 252.6, "max_memory_gb": 16.51, "ce": 0.16417883773101494, "rank": 0.0} +{"step": 146, "steps": 500, "lr": 2e-05, "grad_norm": 0.5191963911056519, "loss_scale": 4.0, "retries": 0, "seconds": 254.6, "max_memory_gb": 16.51, "ce": 0.24458532826974988, "rank": 0.0} +{"step": 147, "steps": 500, "lr": 2e-05, "grad_norm": 0.8891688585281372, "loss_scale": 4.0, "retries": 0, "seconds": 256.5, "max_memory_gb": 16.51, "ce": 0.46273571508936584, "rank": 0.0} +{"step": 148, "steps": 500, "lr": 2e-05, "grad_norm": 0.3463701605796814, "loss_scale": 4.0, "retries": 0, "seconds": 258.3, "max_memory_gb": 16.51, "ce": 0.12154646270209923, "rank": 0.0} +{"step": 149, "steps": 500, "lr": 2e-05, "grad_norm": 0.4953995943069458, "loss_scale": 4.0, "retries": 0, "seconds": 260.0, "max_memory_gb": 16.51, "ce": 0.1718846180010587, "rank": 0.12617728114128113} +{"step": 150, "steps": 500, "lr": 2e-05, "grad_norm": 0.9377866387367249, "loss_scale": 4.0, "retries": 0, "seconds": 261.7, "max_memory_gb": 16.51, "ce": 0.36342865135520697, "rank": 0.0} +{"step": 151, "steps": 500, "lr": 2e-05, "grad_norm": 0.7147558331489563, "loss_scale": 4.0, "retries": 0, "seconds": 263.5, "max_memory_gb": 16.51, "ce": 0.20395918702706695, "rank": 0.0} +{"step": 152, "steps": 500, "lr": 2e-05, "grad_norm": 0.9261727333068848, "loss_scale": 4.0, "retries": 0, "seconds": 265.2, "max_memory_gb": 16.51, "ce": 0.32676134863868356, "rank": 0.0} +{"step": 153, "steps": 500, "lr": 2e-05, "grad_norm": 0.7121407985687256, "loss_scale": 4.0, "retries": 0, "seconds": 266.9, "max_memory_gb": 16.51, "ce": 0.26661874289857224, "rank": 0.0} +{"step": 154, "steps": 500, "lr": 2e-05, "grad_norm": 1.4707399606704712, "loss_scale": 4.0, "retries": 0, "seconds": 268.7, "max_memory_gb": 16.51, "ce": 0.5014871158637106, "rank": 0.11775130033493042} +{"step": 155, "steps": 500, "lr": 2e-05, "grad_norm": 1.6585410833358765, "loss_scale": 4.0, "retries": 0, "seconds": 270.6, "max_memory_gb": 16.51, "ce": 0.5371859194128774, "rank": 0.0} +{"step": 156, "steps": 500, "lr": 2e-05, "grad_norm": 0.5394032001495361, "loss_scale": 4.0, "retries": 0, "seconds": 272.4, "max_memory_gb": 16.51, "ce": 0.21573727659415454, "rank": 0.0} +{"step": 157, "steps": 500, "lr": 2e-05, "grad_norm": 0.471638023853302, "loss_scale": 4.0, "retries": 0, "seconds": 274.1, "max_memory_gb": 16.51, "ce": 0.2012219496537, "rank": 0.0} +{"step": 158, "steps": 500, "lr": 2e-05, "grad_norm": 0.8407581448554993, "loss_scale": 4.0, "retries": 0, "seconds": 275.7, "max_memory_gb": 16.51, "ce": 0.36480378301348537, "rank": 0.0} +{"step": 159, "steps": 500, "lr": 2e-05, "grad_norm": 0.6093495488166809, "loss_scale": 4.0, "retries": 0, "seconds": 277.5, "max_memory_gb": 16.52, "ce": 0.17510619392851368, "rank": 0.0} +{"step": 160, "steps": 500, "lr": 2e-05, "grad_norm": 0.6452959179878235, "loss_scale": 4.0, "retries": 0, "seconds": 279.3, "max_memory_gb": 16.52, "ce": 0.22706766333431005, "rank": 0.0} +{"step": 161, "steps": 500, "lr": 2e-05, "grad_norm": 0.8121418952941895, "loss_scale": 4.0, "retries": 0, "seconds": 281.0, "max_memory_gb": 16.52, "ce": 0.2842519236728549, "rank": 0.0} +{"step": 162, "steps": 500, "lr": 2e-05, "grad_norm": 0.46525800228118896, "loss_scale": 4.0, "retries": 0, "seconds": 282.8, "max_memory_gb": 16.52, "ce": 0.1574747411068529, "rank": 0.0} +{"step": 163, "steps": 500, "lr": 2e-05, "grad_norm": 0.811498761177063, "loss_scale": 4.0, "retries": 0, "seconds": 284.6, "max_memory_gb": 16.52, "ce": 0.2127360925078392, "rank": 0.0} +{"step": 164, "steps": 500, "lr": 2e-05, "grad_norm": 1.0450338125228882, "loss_scale": 4.0, "retries": 0, "seconds": 286.6, "max_memory_gb": 16.52, "ce": 0.2773902867920697, "rank": 0.014478221535682678} +{"step": 165, "steps": 500, "lr": 2e-05, "grad_norm": 0.5183209776878357, "loss_scale": 4.0, "retries": 0, "seconds": 288.5, "max_memory_gb": 16.52, "ce": 0.15693336841650307, "rank": 0.0} +{"step": 166, "steps": 500, "lr": 2e-05, "grad_norm": 0.7745916247367859, "loss_scale": 4.0, "retries": 0, "seconds": 290.2, "max_memory_gb": 16.52, "ce": 0.22782274952623993, "rank": 0.0} +{"step": 167, "steps": 500, "lr": 2e-05, "grad_norm": 0.956184446811676, "loss_scale": 4.0, "retries": 0, "seconds": 292.1, "max_memory_gb": 16.52, "ce": 0.2491480801254511, "rank": 0.0} +{"step": 168, "steps": 500, "lr": 2e-05, "grad_norm": 0.5834096074104309, "loss_scale": 4.0, "retries": 0, "seconds": 293.8, "max_memory_gb": 16.52, "ce": 0.1450905802194029, "rank": 0.0} +{"step": 169, "steps": 500, "lr": 2e-05, "grad_norm": 1.7764451503753662, "loss_scale": 4.0, "retries": 0, "seconds": 295.5, "max_memory_gb": 16.52, "ce": 0.5973545943852514, "rank": 0.0} +{"step": 170, "steps": 500, "lr": 2e-05, "grad_norm": 0.5230827927589417, "loss_scale": 4.0, "retries": 0, "seconds": 297.2, "max_memory_gb": 16.52, "ce": 0.12952259683515877, "rank": 0.0} +{"step": 171, "steps": 500, "lr": 2e-05, "grad_norm": 0.8713659048080444, "loss_scale": 4.0, "retries": 0, "seconds": 299.0, "max_memory_gb": 16.52, "ce": 0.165591921802843, "rank": 0.0} +{"step": 172, "steps": 500, "lr": 2e-05, "grad_norm": 0.75847327709198, "loss_scale": 4.0, "retries": 0, "seconds": 300.8, "max_memory_gb": 16.52, "ce": 0.17682296899147332, "rank": 0.0} +{"step": 173, "steps": 500, "lr": 2e-05, "grad_norm": 1.4161992073059082, "loss_scale": 4.0, "retries": 0, "seconds": 302.4, "max_memory_gb": 16.52, "ce": 0.24348747706972063, "rank": 0.0} +{"step": 174, "steps": 500, "lr": 2e-05, "grad_norm": 0.9114918112754822, "loss_scale": 4.0, "retries": 0, "seconds": 304.1, "max_memory_gb": 16.52, "ce": 0.18496484769275412, "rank": 0.0} +{"step": 175, "steps": 500, "lr": 2e-05, "grad_norm": 0.6708497405052185, "loss_scale": 4.0, "retries": 0, "seconds": 305.9, "max_memory_gb": 16.52, "ce": 0.24493251065723598, "rank": 0.0} +{"step": 176, "steps": 500, "lr": 2e-05, "grad_norm": 0.6741058230400085, "loss_scale": 4.0, "retries": 0, "seconds": 307.6, "max_memory_gb": 16.52, "ce": 0.12123208516277373, "rank": 0.0} +{"step": 177, "steps": 500, "lr": 2e-05, "grad_norm": 1.047484278678894, "loss_scale": 4.0, "retries": 0, "seconds": 309.3, "max_memory_gb": 16.52, "ce": 0.2515194595907815, "rank": 0.1266445815563202} +{"step": 178, "steps": 500, "lr": 2e-05, "grad_norm": 0.44100093841552734, "loss_scale": 4.0, "retries": 0, "seconds": 311.0, "max_memory_gb": 16.52, "ce": 0.15582957229344174, "rank": 0.0} +{"step": 179, "steps": 500, "lr": 2e-05, "grad_norm": 0.8124174475669861, "loss_scale": 4.0, "retries": 0, "seconds": 312.7, "max_memory_gb": 16.52, "ce": 0.372924012131989, "rank": 0.0} +{"step": 180, "steps": 500, "lr": 2e-05, "grad_norm": 1.4462833404541016, "loss_scale": 4.0, "retries": 0, "seconds": 314.4, "max_memory_gb": 16.52, "ce": 0.343703645048663, "rank": 0.0} +{"step": 181, "steps": 500, "lr": 2e-05, "grad_norm": 0.7823835015296936, "loss_scale": 4.0, "retries": 0, "seconds": 316.1, "max_memory_gb": 16.52, "ce": 0.1247181196231395, "rank": 0.0} +{"step": 182, "steps": 500, "lr": 2e-05, "grad_norm": 1.7712788581848145, "loss_scale": 4.0, "retries": 0, "seconds": 317.8, "max_memory_gb": 16.52, "ce": 0.2308941253577359, "rank": 0.0} +{"step": 183, "steps": 500, "lr": 2e-05, "grad_norm": 0.3235873878002167, "loss_scale": 4.0, "retries": 0, "seconds": 319.6, "max_memory_gb": 16.52, "ce": 0.054597149312030524, "rank": 0.0} +{"step": 184, "steps": 500, "lr": 2e-05, "grad_norm": 0.6329445838928223, "loss_scale": 4.0, "retries": 0, "seconds": 321.4, "max_memory_gb": 16.52, "ce": 0.11777843712479807, "rank": 0.09558364748954773} +{"step": 185, "steps": 500, "lr": 2e-05, "grad_norm": 0.5916827321052551, "loss_scale": 4.0, "retries": 0, "seconds": 323.1, "max_memory_gb": 16.52, "ce": 0.11182663193903863, "rank": 0.0} +{"step": 186, "steps": 500, "lr": 2e-05, "grad_norm": 0.5338596701622009, "loss_scale": 4.0, "retries": 0, "seconds": 324.8, "max_memory_gb": 16.52, "ce": 0.07211123831802979, "rank": 0.0} +{"step": 187, "steps": 500, "lr": 2e-05, "grad_norm": 1.0953154563903809, "loss_scale": 4.0, "retries": 0, "seconds": 326.6, "max_memory_gb": 16.52, "ce": 0.17934932047501206, "rank": 0.1238347664475441} +{"step": 188, "steps": 500, "lr": 2e-05, "grad_norm": 0.27016139030456543, "loss_scale": 4.0, "retries": 0, "seconds": 328.3, "max_memory_gb": 16.52, "ce": 0.12224789868923835, "rank": 0.0} +{"step": 189, "steps": 500, "lr": 2e-05, "grad_norm": 0.9448840022087097, "loss_scale": 4.0, "retries": 0, "seconds": 330.0, "max_memory_gb": 16.52, "ce": 0.09533352172002196, "rank": 0.0} +{"step": 190, "steps": 500, "lr": 2e-05, "grad_norm": 1.08432137966156, "loss_scale": 4.0, "retries": 0, "seconds": 331.7, "max_memory_gb": 16.52, "ce": 0.11592851276509464, "rank": 0.0} +{"step": 191, "steps": 500, "lr": 2e-05, "grad_norm": 0.9678961038589478, "loss_scale": 4.0, "retries": 0, "seconds": 333.4, "max_memory_gb": 16.52, "ce": 0.28571961843408644, "rank": 0.0} +{"step": 192, "steps": 500, "lr": 2e-05, "grad_norm": 0.699103593826294, "loss_scale": 4.0, "retries": 0, "seconds": 335.1, "max_memory_gb": 16.52, "ce": 0.20166171138407663, "rank": 0.0} +{"step": 193, "steps": 500, "lr": 2e-05, "grad_norm": 0.8043795228004456, "loss_scale": 4.0, "retries": 0, "seconds": 336.9, "max_memory_gb": 16.52, "ce": 0.09636504147783853, "rank": 0.1253248155117035} +{"step": 194, "steps": 500, "lr": 2e-05, "grad_norm": 2.181041955947876, "loss_scale": 4.0, "retries": 0, "seconds": 338.6, "max_memory_gb": 16.52, "ce": 0.3404427021741867, "rank": 0.0} +{"step": 195, "steps": 500, "lr": 2e-05, "grad_norm": 1.1249092817306519, "loss_scale": 4.0, "retries": 0, "seconds": 340.3, "max_memory_gb": 16.52, "ce": 0.17934492230415344, "rank": 0.0} +{"step": 196, "steps": 500, "lr": 2e-05, "grad_norm": 0.49030008912086487, "loss_scale": 4.0, "retries": 0, "seconds": 342.1, "max_memory_gb": 16.52, "ce": 0.04954344022553414, "rank": 0.0} +{"step": 197, "steps": 500, "lr": 2e-05, "grad_norm": 0.6355521082878113, "loss_scale": 4.0, "retries": 0, "seconds": 343.8, "max_memory_gb": 16.52, "ce": 0.08689222822431475, "rank": 0.10293939709663391} +{"step": 198, "steps": 500, "lr": 2e-05, "grad_norm": 0.812598705291748, "loss_scale": 4.0, "retries": 0, "seconds": 345.5, "max_memory_gb": 16.52, "ce": 0.14113712403923273, "rank": 0.0} +{"step": 199, "steps": 500, "lr": 2e-05, "grad_norm": 1.5564353466033936, "loss_scale": 4.0, "retries": 0, "seconds": 347.2, "max_memory_gb": 16.52, "ce": 0.12319458369165659, "rank": 0.0} +{"step": 200, "steps": 500, "lr": 2e-05, "grad_norm": 0.6010298728942871, "loss_scale": 4.0, "retries": 0, "seconds": 348.9, "max_memory_gb": 16.52, "ce": 0.2016458244761452, "rank": 0.0} +{"step": 201, "steps": 500, "lr": 2e-05, "grad_norm": 0.6260170340538025, "loss_scale": 4.0, "retries": 0, "seconds": 350.7, "max_memory_gb": 16.52, "ce": 0.07799462170805782, "rank": 0.0} +{"step": 202, "steps": 500, "lr": 2e-05, "grad_norm": 0.20598456263542175, "loss_scale": 4.0, "retries": 0, "seconds": 352.4, "max_memory_gb": 16.52, "ce": 0.023583988309837878, "rank": 0.0} +{"step": 203, "steps": 500, "lr": 2e-05, "grad_norm": 0.8359333872795105, "loss_scale": 4.0, "retries": 0, "seconds": 354.1, "max_memory_gb": 16.52, "ce": 0.11583833815529943, "rank": 0.0} +{"step": 204, "steps": 500, "lr": 2e-05, "grad_norm": 0.7959834933280945, "loss_scale": 4.0, "retries": 0, "seconds": 355.8, "max_memory_gb": 16.52, "ce": 0.0767480821814388, "rank": 0.0} +{"step": 205, "steps": 500, "lr": 2e-05, "grad_norm": 0.8770341873168945, "loss_scale": 4.0, "retries": 0, "seconds": 357.5, "max_memory_gb": 16.52, "ce": 0.10275119464495219, "rank": 0.0} +{"step": 206, "steps": 500, "lr": 2e-05, "grad_norm": 0.9393287897109985, "loss_scale": 4.0, "retries": 0, "seconds": 359.2, "max_memory_gb": 16.52, "ce": 0.06862459162948653, "rank": 0.0} +{"step": 207, "steps": 500, "lr": 2e-05, "grad_norm": 1.6059761047363281, "loss_scale": 4.0, "retries": 0, "seconds": 360.9, "max_memory_gb": 16.52, "ce": 0.5251167925016489, "rank": 0.21321973204612732} +{"step": 208, "steps": 500, "lr": 2e-05, "grad_norm": 0.33486470580101013, "loss_scale": 4.0, "retries": 0, "seconds": 362.7, "max_memory_gb": 16.52, "ce": 0.023707221524091437, "rank": 0.0} +{"step": 209, "steps": 500, "lr": 2e-05, "grad_norm": 0.5135629773139954, "loss_scale": 4.0, "retries": 0, "seconds": 364.4, "max_memory_gb": 16.52, "ce": 0.14237585198134184, "rank": 0.12503458559513092} +{"step": 210, "steps": 500, "lr": 2e-05, "grad_norm": 1.3146709203720093, "loss_scale": 4.0, "retries": 0, "seconds": 366.2, "max_memory_gb": 16.52, "ce": 0.24937509794835933, "rank": 0.0} +{"step": 211, "steps": 500, "lr": 2e-05, "grad_norm": 0.5946349501609802, "loss_scale": 4.0, "retries": 0, "seconds": 367.9, "max_memory_gb": 16.52, "ce": 0.17859974305611104, "rank": 0.0} +{"step": 212, "steps": 500, "lr": 2e-05, "grad_norm": 1.4401344060897827, "loss_scale": 4.0, "retries": 0, "seconds": 369.7, "max_memory_gb": 16.52, "ce": 0.26539638394024223, "rank": 0.0} +{"step": 213, "steps": 500, "lr": 2e-05, "grad_norm": 0.9441565275192261, "loss_scale": 4.0, "retries": 0, "seconds": 371.4, "max_memory_gb": 16.52, "ce": 0.27788781223353, "rank": 0.0} +{"step": 214, "steps": 500, "lr": 2e-05, "grad_norm": 1.1345131397247314, "loss_scale": 4.0, "retries": 0, "seconds": 373.1, "max_memory_gb": 16.52, "ce": 0.25607772753573954, "rank": 0.0} +{"step": 215, "steps": 500, "lr": 2e-05, "grad_norm": 3.7763617038726807, "loss_scale": 4.0, "retries": 0, "seconds": 374.8, "max_memory_gb": 16.52, "ce": 0.5149070254992694, "rank": 0.0} +{"step": 216, "steps": 500, "lr": 2e-05, "grad_norm": 1.8736519813537598, "loss_scale": 4.0, "retries": 0, "seconds": 376.6, "max_memory_gb": 16.52, "ce": 0.1798430304042995, "rank": 0.0} +{"step": 217, "steps": 500, "lr": 2e-05, "grad_norm": 0.6919623613357544, "loss_scale": 4.0, "retries": 0, "seconds": 378.3, "max_memory_gb": 16.52, "ce": 0.05788006400689483, "rank": 0.0} +{"step": 218, "steps": 500, "lr": 2e-05, "grad_norm": 3.6477272510528564, "loss_scale": 4.0, "retries": 0, "seconds": 380.0, "max_memory_gb": 16.52, "ce": 0.49371363793034106, "rank": 0.0} +{"step": 219, "steps": 500, "lr": 2e-05, "grad_norm": 0.5812800526618958, "loss_scale": 4.0, "retries": 0, "seconds": 381.8, "max_memory_gb": 16.52, "ce": 0.06864391395356506, "rank": 0.0} +{"step": 220, "steps": 500, "lr": 2e-05, "grad_norm": 1.5325891971588135, "loss_scale": 4.0, "retries": 0, "seconds": 383.5, "max_memory_gb": 16.52, "ce": 0.10769143071956933, "rank": 0.0} +{"step": 221, "steps": 500, "lr": 2e-05, "grad_norm": 0.8638281226158142, "loss_scale": 4.0, "retries": 0, "seconds": 385.3, "max_memory_gb": 16.52, "ce": 0.08612395933596417, "rank": 0.0} +{"step": 222, "steps": 500, "lr": 2e-05, "grad_norm": 5.429413318634033, "loss_scale": 4.0, "retries": 0, "seconds": 387.0, "max_memory_gb": 16.52, "ce": 0.5089884174522012, "rank": 0.0} +{"step": 223, "steps": 500, "lr": 2e-05, "grad_norm": 0.7438832521438599, "loss_scale": 4.0, "retries": 0, "seconds": 388.8, "max_memory_gb": 16.52, "ce": 0.11130375176435336, "rank": 0.0} +{"step": 224, "steps": 500, "lr": 2e-05, "grad_norm": 0.5461704134941101, "loss_scale": 4.0, "retries": 0, "seconds": 390.5, "max_memory_gb": 16.52, "ce": 0.05783654534025118, "rank": 0.10803830623626709} +{"step": 225, "steps": 500, "lr": 2e-05, "grad_norm": 0.7098469734191895, "loss_scale": 4.0, "retries": 0, "seconds": 392.2, "max_memory_gb": 16.52, "ce": 0.045565738051664084, "rank": 0.0} +{"step": 226, "steps": 500, "lr": 2e-05, "grad_norm": 0.7050857543945312, "loss_scale": 4.0, "retries": 0, "seconds": 394.0, "max_memory_gb": 16.52, "ce": 0.06465215230127797, "rank": 0.0} +{"step": 227, "steps": 500, "lr": 2e-05, "grad_norm": 0.6898961663246155, "loss_scale": 4.0, "retries": 0, "seconds": 395.7, "max_memory_gb": 16.52, "ce": 0.020561575540341437, "rank": 0.0} +{"step": 228, "steps": 500, "lr": 2e-05, "grad_norm": 0.9947844743728638, "loss_scale": 4.0, "retries": 0, "seconds": 397.4, "max_memory_gb": 16.52, "ce": 0.0642223744362127, "rank": 0.0} +{"step": 229, "steps": 500, "lr": 2e-05, "grad_norm": 0.9397071003913879, "loss_scale": 4.0, "retries": 0, "seconds": 399.4, "max_memory_gb": 16.52, "ce": 0.14433958561858162, "rank": 0.0} +{"step": 230, "steps": 500, "lr": 2e-05, "grad_norm": 0.7716351747512817, "loss_scale": 4.0, "retries": 0, "seconds": 401.3, "max_memory_gb": 16.52, "ce": 0.18845499132294208, "rank": 0.0} +{"step": 231, "steps": 500, "lr": 2e-05, "grad_norm": 0.9384127259254456, "loss_scale": 4.0, "retries": 0, "seconds": 403.1, "max_memory_gb": 16.52, "ce": 0.19552903945441358, "rank": 0.0} +{"step": 232, "steps": 500, "lr": 2e-05, "grad_norm": 0.6160452961921692, "loss_scale": 4.0, "retries": 0, "seconds": 404.9, "max_memory_gb": 16.52, "ce": 0.028291515540331602, "rank": 0.0} +{"step": 233, "steps": 500, "lr": 2e-05, "grad_norm": 1.7232037782669067, "loss_scale": 4.0, "retries": 0, "seconds": 406.9, "max_memory_gb": 16.52, "ce": 0.11614172463305295, "rank": 0.0} +{"step": 234, "steps": 500, "lr": 2e-05, "grad_norm": 1.1939011812210083, "loss_scale": 4.0, "retries": 0, "seconds": 408.8, "max_memory_gb": 16.52, "ce": 0.1471165315597318, "rank": 0.0} +{"step": 235, "steps": 500, "lr": 2e-05, "grad_norm": 1.2150830030441284, "loss_scale": 4.0, "retries": 0, "seconds": 410.8, "max_memory_gb": 16.52, "ce": 0.31899869118933566, "rank": 0.0} +{"step": 236, "steps": 500, "lr": 2e-05, "grad_norm": 0.6987040042877197, "loss_scale": 4.0, "retries": 0, "seconds": 412.6, "max_memory_gb": 16.52, "ce": 0.0669794087880291, "rank": 0.0} +{"step": 237, "steps": 500, "lr": 2e-05, "grad_norm": 2.5353991985321045, "loss_scale": 4.0, "retries": 0, "seconds": 414.3, "max_memory_gb": 16.52, "ce": 0.18728455927339382, "rank": 0.0} +{"step": 238, "steps": 500, "lr": 2e-05, "grad_norm": 0.847722053527832, "loss_scale": 4.0, "retries": 0, "seconds": 416.0, "max_memory_gb": 16.52, "ce": 0.06603685556910932, "rank": 0.0} +{"step": 239, "steps": 500, "lr": 2e-05, "grad_norm": 0.5634291768074036, "loss_scale": 4.0, "retries": 0, "seconds": 417.7, "max_memory_gb": 16.52, "ce": 0.09454427467426285, "rank": 0.0} +{"step": 240, "steps": 500, "lr": 2e-05, "grad_norm": 1.424788236618042, "loss_scale": 4.0, "retries": 0, "seconds": 419.5, "max_memory_gb": 16.52, "ce": 0.09818637643184047, "rank": 0.0} +{"step": 241, "steps": 500, "lr": 2e-05, "grad_norm": 0.6929361820220947, "loss_scale": 4.0, "retries": 0, "seconds": 421.2, "max_memory_gb": 16.52, "ce": 0.11356413867906667, "rank": 0.0} +{"step": 242, "steps": 500, "lr": 2e-05, "grad_norm": 0.881844699382782, "loss_scale": 4.0, "retries": 0, "seconds": 422.9, "max_memory_gb": 16.52, "ce": 0.1088152902375441, "rank": 0.0} +{"step": 243, "steps": 500, "lr": 2e-05, "grad_norm": 6.631104946136475, "loss_scale": 4.0, "retries": 0, "seconds": 424.7, "max_memory_gb": 16.52, "ce": 0.2315881701942999, "rank": 0.0} +{"step": 244, "steps": 500, "lr": 2e-05, "grad_norm": 2.510012149810791, "loss_scale": 4.0, "retries": 0, "seconds": 426.4, "max_memory_gb": 16.52, "ce": 0.26076953299343586, "rank": 0.0} +{"step": 245, "steps": 500, "lr": 2e-05, "grad_norm": 1.4892362356185913, "loss_scale": 4.0, "retries": 0, "seconds": 428.2, "max_memory_gb": 16.52, "ce": 0.3120324162300676, "rank": 0.0} +{"step": 246, "steps": 500, "lr": 2e-05, "grad_norm": 0.10415041446685791, "loss_scale": 4.0, "retries": 0, "seconds": 429.9, "max_memory_gb": 16.52, "ce": 0.00786257311119698, "rank": 0.0} +{"step": 247, "steps": 500, "lr": 2e-05, "grad_norm": 20.573328018188477, "loss_scale": 4.0, "retries": 0, "seconds": 431.7, "max_memory_gb": 16.52, "ce": 0.3353769195819041, "rank": 0.0} +{"step": 248, "steps": 500, "lr": 2e-05, "grad_norm": 0.4975168704986572, "loss_scale": 4.0, "retries": 0, "seconds": 433.5, "max_memory_gb": 16.52, "ce": 0.04115091261337511, "rank": 0.0} +{"step": 249, "steps": 500, "lr": 2e-05, "grad_norm": 1.3289644718170166, "loss_scale": 4.0, "retries": 0, "seconds": 435.2, "max_memory_gb": 16.52, "ce": 0.032749955280451104, "rank": 0.0} +{"step": 250, "steps": 500, "lr": 2e-05, "grad_norm": 0.4006164073944092, "loss_scale": 4.0, "retries": 0, "seconds": 436.9, "max_memory_gb": 16.52, "ce": 0.024038857693085447, "rank": 0.0} +{"step": 251, "steps": 500, "lr": 2e-05, "grad_norm": 1.582850456237793, "loss_scale": 4.0, "retries": 0, "seconds": 438.7, "max_memory_gb": 16.52, "ce": 0.11858760882751085, "rank": 0.0} +{"step": 252, "steps": 500, "lr": 2e-05, "grad_norm": 0.8232820630073547, "loss_scale": 4.0, "retries": 0, "seconds": 440.4, "max_memory_gb": 16.52, "ce": 0.04621185520954896, "rank": 0.0} +{"step": 253, "steps": 500, "lr": 2e-05, "grad_norm": 1.291447639465332, "loss_scale": 4.0, "retries": 0, "seconds": 442.2, "max_memory_gb": 16.52, "ce": 0.1417883258254733, "rank": 0.0} +{"step": 254, "steps": 500, "lr": 2e-05, "grad_norm": 0.985244870185852, "loss_scale": 4.0, "retries": 0, "seconds": 443.9, "max_memory_gb": 16.52, "ce": 0.0997590157930972, "rank": 0.0} +{"step": 255, "steps": 500, "lr": 2e-05, "grad_norm": 0.7453593015670776, "loss_scale": 4.0, "retries": 0, "seconds": 445.6, "max_memory_gb": 16.52, "ce": 0.16026275523472577, "rank": 0.0} +{"step": 256, "steps": 500, "lr": 2e-05, "grad_norm": 0.7091312408447266, "loss_scale": 4.0, "retries": 0, "seconds": 447.4, "max_memory_gb": 16.52, "ce": 0.06067382384208031, "rank": 0.0} +{"step": 257, "steps": 500, "lr": 2e-05, "grad_norm": 0.9032131433486938, "loss_scale": 4.0, "retries": 0, "seconds": 449.1, "max_memory_gb": 16.52, "ce": 0.04606138229428325, "rank": 0.0} +{"step": 258, "steps": 500, "lr": 2e-05, "grad_norm": 0.7979499697685242, "loss_scale": 4.0, "retries": 0, "seconds": 450.8, "max_memory_gb": 16.52, "ce": 0.03550295054446906, "rank": 0.0} +{"step": 259, "steps": 500, "lr": 2e-05, "grad_norm": 0.35430610179901123, "loss_scale": 4.0, "retries": 0, "seconds": 452.6, "max_memory_gb": 16.52, "ce": 0.018009412437095307, "rank": 0.0} +{"step": 260, "steps": 500, "lr": 2e-05, "grad_norm": 0.5067445635795593, "loss_scale": 4.0, "retries": 0, "seconds": 454.3, "max_memory_gb": 16.52, "ce": 0.058450729338801466, "rank": 0.0} +{"step": 261, "steps": 500, "lr": 2e-05, "grad_norm": 0.6977230906486511, "loss_scale": 4.0, "retries": 0, "seconds": 456.1, "max_memory_gb": 16.52, "ce": 0.07592152014694875, "rank": 0.0} +{"step": 262, "steps": 500, "lr": 2e-05, "grad_norm": 1.186565637588501, "loss_scale": 4.0, "retries": 0, "seconds": 457.9, "max_memory_gb": 16.52, "ce": 0.12716466524580028, "rank": 0.0} +{"step": 263, "steps": 500, "lr": 2e-05, "grad_norm": 1.161161184310913, "loss_scale": 4.0, "retries": 0, "seconds": 459.6, "max_memory_gb": 16.52, "ce": 0.11874529526539845, "rank": 0.1247452050447464} +{"step": 264, "steps": 500, "lr": 2e-05, "grad_norm": 2.5724241733551025, "loss_scale": 4.0, "retries": 0, "seconds": 461.3, "max_memory_gb": 16.52, "ce": 0.20873516786377877, "rank": 0.0} +{"step": 265, "steps": 500, "lr": 2e-05, "grad_norm": 0.3602723181247711, "loss_scale": 4.0, "retries": 0, "seconds": 463.1, "max_memory_gb": 16.52, "ce": 0.050419270133716054, "rank": 0.0} +{"step": 266, "steps": 500, "lr": 2e-05, "grad_norm": 4.2419352531433105, "loss_scale": 4.0, "retries": 0, "seconds": 464.8, "max_memory_gb": 16.52, "ce": 0.5855143165681511, "rank": 0.0} +{"step": 267, "steps": 500, "lr": 2e-05, "grad_norm": 1.159033179283142, "loss_scale": 4.0, "retries": 0, "seconds": 466.6, "max_memory_gb": 16.52, "ce": 0.1387419406673871, "rank": 0.0} +{"step": 268, "steps": 500, "lr": 2e-05, "grad_norm": 0.4810091555118561, "loss_scale": 4.0, "retries": 0, "seconds": 468.3, "max_memory_gb": 16.52, "ce": 0.15824246776173823, "rank": 0.0} +{"step": 269, "steps": 500, "lr": 2e-05, "grad_norm": 4.528139114379883, "loss_scale": 4.0, "retries": 0, "seconds": 470.1, "max_memory_gb": 16.52, "ce": 0.47319877210247796, "rank": 0.0} +{"step": 270, "steps": 500, "lr": 2e-05, "grad_norm": 2.062316656112671, "loss_scale": 4.0, "retries": 0, "seconds": 471.9, "max_memory_gb": 16.52, "ce": 0.18833099064795533, "rank": 0.0} +{"step": 271, "steps": 500, "lr": 2e-05, "grad_norm": 1.0995675325393677, "loss_scale": 4.0, "retries": 0, "seconds": 473.6, "max_memory_gb": 16.52, "ce": 0.0456608656750177, "rank": 0.12235335260629654} +{"step": 272, "steps": 500, "lr": 2e-05, "grad_norm": 1.1602153778076172, "loss_scale": 4.0, "retries": 0, "seconds": 475.3, "max_memory_gb": 16.52, "ce": 0.09347128966692253, "rank": 0.0} +{"step": 273, "steps": 500, "lr": 2e-05, "grad_norm": 1.0605446100234985, "loss_scale": 4.0, "retries": 0, "seconds": 477.0, "max_memory_gb": 16.52, "ce": 0.07036326238085167, "rank": 0.1252024620771408} +{"step": 274, "steps": 500, "lr": 2e-05, "grad_norm": 0.9404903650283813, "loss_scale": 4.0, "retries": 0, "seconds": 478.8, "max_memory_gb": 16.52, "ce": 0.10038074169278843, "rank": 0.0} +{"step": 275, "steps": 500, "lr": 2e-05, "grad_norm": 2.2830474376678467, "loss_scale": 4.0, "retries": 0, "seconds": 480.5, "max_memory_gb": 16.52, "ce": 0.6973347246494086, "rank": 0.0} +{"step": 276, "steps": 500, "lr": 2e-05, "grad_norm": 0.8330244421958923, "loss_scale": 4.0, "retries": 0, "seconds": 482.3, "max_memory_gb": 16.52, "ce": 0.01580815387023904, "rank": 0.0} +{"step": 277, "steps": 500, "lr": 2e-05, "grad_norm": 0.4399951100349426, "loss_scale": 4.0, "retries": 0, "seconds": 484.0, "max_memory_gb": 16.52, "ce": 0.01588362193433568, "rank": 0.0} +{"step": 278, "steps": 500, "lr": 2e-05, "grad_norm": 1.7201502323150635, "loss_scale": 4.0, "retries": 0, "seconds": 485.8, "max_memory_gb": 16.52, "ce": 0.1297501686567557, "rank": 0.0} +{"step": 279, "steps": 500, "lr": 2e-05, "grad_norm": 0.8320791125297546, "loss_scale": 4.0, "retries": 0, "seconds": 487.6, "max_memory_gb": 16.52, "ce": 0.02817479645818821, "rank": 0.0} +{"step": 280, "steps": 500, "lr": 2e-05, "grad_norm": 0.7156739234924316, "loss_scale": 4.0, "retries": 0, "seconds": 489.3, "max_memory_gb": 16.52, "ce": 0.014930967949112528, "rank": 0.0} +{"step": 281, "steps": 500, "lr": 2e-05, "grad_norm": 1.4107086658477783, "loss_scale": 4.0, "retries": 0, "seconds": 491.1, "max_memory_gb": 16.52, "ce": 0.08000037576857721, "rank": 0.0} +{"step": 282, "steps": 500, "lr": 2e-05, "grad_norm": 0.7350911498069763, "loss_scale": 4.0, "retries": 0, "seconds": 492.8, "max_memory_gb": 16.52, "ce": 0.08234349390113493, "rank": 0.0} +{"step": 283, "steps": 500, "lr": 2e-05, "grad_norm": 0.9937864542007446, "loss_scale": 4.0, "retries": 0, "seconds": 494.6, "max_memory_gb": 16.52, "ce": 0.08460674362504506, "rank": 0.0} +{"step": 284, "steps": 500, "lr": 2e-05, "grad_norm": 1.11329984664917, "loss_scale": 4.0, "retries": 0, "seconds": 496.4, "max_memory_gb": 16.52, "ce": 0.020809728812309913, "rank": 0.0} +{"step": 285, "steps": 500, "lr": 2e-05, "grad_norm": 0.6359131932258606, "loss_scale": 4.0, "retries": 0, "seconds": 498.1, "max_memory_gb": 16.52, "ce": 0.062222371216194006, "rank": 0.0} +{"step": 286, "steps": 500, "lr": 2e-05, "grad_norm": 0.525428295135498, "loss_scale": 4.0, "retries": 0, "seconds": 499.9, "max_memory_gb": 16.52, "ce": 0.04141915380205319, "rank": 0.0} +{"step": 287, "steps": 500, "lr": 2e-05, "grad_norm": 1.9220211505889893, "loss_scale": 4.0, "retries": 0, "seconds": 501.7, "max_memory_gb": 16.52, "ce": 0.11817350537421589, "rank": 0.0} +{"step": 288, "steps": 500, "lr": 2e-05, "grad_norm": 0.058006465435028076, "loss_scale": 4.0, "retries": 0, "seconds": 503.5, "max_memory_gb": 16.52, "ce": 0.001983194169952185, "rank": 0.0} +{"step": 289, "steps": 500, "lr": 2e-05, "grad_norm": 1.268062949180603, "loss_scale": 4.0, "retries": 0, "seconds": 505.3, "max_memory_gb": 16.52, "ce": 0.057720275614883576, "rank": 0.0} +{"step": 290, "steps": 500, "lr": 2e-05, "grad_norm": 0.7070614695549011, "loss_scale": 4.0, "retries": 0, "seconds": 507.0, "max_memory_gb": 16.52, "ce": 0.07910647476091981, "rank": 0.0} +{"step": 291, "steps": 500, "lr": 2e-05, "grad_norm": 0.23642481863498688, "loss_scale": 4.0, "retries": 0, "seconds": 508.8, "max_memory_gb": 16.52, "ce": 0.0034733090724330395, "rank": 0.0} +{"step": 292, "steps": 500, "lr": 2e-05, "grad_norm": 1.9487135410308838, "loss_scale": 4.0, "retries": 0, "seconds": 510.5, "max_memory_gb": 16.52, "ce": 0.1232861605258222, "rank": 0.0} +{"step": 293, "steps": 500, "lr": 2e-05, "grad_norm": 0.3643014132976532, "loss_scale": 4.0, "retries": 0, "seconds": 512.2, "max_memory_gb": 16.52, "ce": 0.028187672216517967, "rank": 0.0} +{"step": 294, "steps": 500, "lr": 2e-05, "grad_norm": 22.86553192138672, "loss_scale": 4.0, "retries": 0, "seconds": 514.0, "max_memory_gb": 16.52, "ce": 0.38775378819264006, "rank": 0.0} +{"step": 295, "steps": 500, "lr": 2e-05, "grad_norm": 23.217697143554688, "loss_scale": 4.0, "retries": 0, "seconds": 515.7, "max_memory_gb": 16.52, "ce": 0.2461936204799713, "rank": 0.0} +{"step": 296, "steps": 500, "lr": 2e-05, "grad_norm": 6.205655097961426, "loss_scale": 4.0, "retries": 0, "seconds": 517.3, "max_memory_gb": 16.52, "ce": 0.25953143259084754, "rank": 0.0} +{"step": 297, "steps": 500, "lr": 2e-05, "grad_norm": 1.1381686925888062, "loss_scale": 4.0, "retries": 0, "seconds": 519.2, "max_memory_gb": 16.52, "ce": 0.10939136386514292, "rank": 0.0} +{"step": 298, "steps": 500, "lr": 2e-05, "grad_norm": 3.304192066192627, "loss_scale": 4.0, "retries": 0, "seconds": 521.1, "max_memory_gb": 16.52, "ce": 0.26547043682512594, "rank": 0.0} +{"step": 299, "steps": 500, "lr": 2e-05, "grad_norm": 0.6603720784187317, "loss_scale": 4.0, "retries": 0, "seconds": 523.1, "max_memory_gb": 16.52, "ce": 0.024135884511451877, "rank": 0.0} +{"step": 300, "steps": 500, "lr": 2e-05, "grad_norm": 1.4995759725570679, "loss_scale": 4.0, "retries": 0, "seconds": 524.9, "max_memory_gb": 16.52, "ce": 0.21851642625733803, "rank": 0.0} +{"step": 301, "steps": 500, "lr": 2e-05, "grad_norm": 1.7654310464859009, "loss_scale": 4.0, "retries": 0, "seconds": 526.6, "max_memory_gb": 16.52, "ce": 0.10290054061624687, "rank": 0.0} +{"step": 302, "steps": 500, "lr": 2e-05, "grad_norm": 0.3495897948741913, "loss_scale": 4.0, "retries": 0, "seconds": 528.4, "max_memory_gb": 16.52, "ce": 0.012628637909074314, "rank": 0.0} +{"step": 303, "steps": 500, "lr": 2e-05, "grad_norm": 9.010489463806152, "loss_scale": 4.0, "retries": 0, "seconds": 530.1, "max_memory_gb": 16.52, "ce": 0.052208236635124194, "rank": 0.0} +{"step": 304, "steps": 500, "lr": 2e-05, "grad_norm": 2.5733847618103027, "loss_scale": 4.0, "retries": 0, "seconds": 531.9, "max_memory_gb": 16.52, "ce": 0.043377852358389646, "rank": 0.0} +{"step": 305, "steps": 500, "lr": 2e-05, "grad_norm": 5.6469926834106445, "loss_scale": 4.0, "retries": 0, "seconds": 533.6, "max_memory_gb": 16.52, "ce": 0.12229630842193728, "rank": 0.0} +{"step": 306, "steps": 500, "lr": 2e-05, "grad_norm": 0.2888951599597931, "loss_scale": 4.0, "retries": 0, "seconds": 535.3, "max_memory_gb": 16.52, "ce": 0.007970433427544776, "rank": 0.0} +{"step": 307, "steps": 500, "lr": 2e-05, "grad_norm": 1.8394941091537476, "loss_scale": 4.0, "retries": 0, "seconds": 537.0, "max_memory_gb": 16.52, "ce": 0.16603412183758337, "rank": 0.0} +{"step": 308, "steps": 500, "lr": 2e-05, "grad_norm": 2.027074098587036, "loss_scale": 4.0, "retries": 0, "seconds": 538.8, "max_memory_gb": 16.52, "ce": 0.2219687669785344, "rank": 0.0} +{"step": 309, "steps": 500, "lr": 2e-05, "grad_norm": 1.865839958190918, "loss_scale": 4.0, "retries": 0, "seconds": 540.5, "max_memory_gb": 16.52, "ce": 0.15946455121593317, "rank": 0.0} +{"step": 310, "steps": 500, "lr": 2e-05, "grad_norm": 0.500627875328064, "loss_scale": 4.0, "retries": 0, "seconds": 542.4, "max_memory_gb": 16.52, "ce": 0.04185236343982979, "rank": 0.0} +{"step": 311, "steps": 500, "lr": 2e-05, "grad_norm": 0.9453162550926208, "loss_scale": 4.0, "retries": 0, "seconds": 544.2, "max_memory_gb": 16.52, "ce": 0.11618359840213088, "rank": 0.0} +{"step": 312, "steps": 500, "lr": 2e-05, "grad_norm": 0.15857845544815063, "loss_scale": 4.0, "retries": 0, "seconds": 546.3, "max_memory_gb": 16.52, "ce": 0.009339394116977928, "rank": 0.0} +{"step": 313, "steps": 500, "lr": 2e-05, "grad_norm": 2.408193588256836, "loss_scale": 4.0, "retries": 0, "seconds": 548.3, "max_memory_gb": 16.52, "ce": 0.16875426528531534, "rank": 0.0} +{"step": 314, "steps": 500, "lr": 2e-05, "grad_norm": 1.5419151782989502, "loss_scale": 4.0, "retries": 0, "seconds": 550.4, "max_memory_gb": 16.52, "ce": 0.22466037941194372, "rank": 0.0} +{"step": 315, "steps": 500, "lr": 2e-05, "grad_norm": 0.48646000027656555, "loss_scale": 4.0, "retries": 0, "seconds": 552.4, "max_memory_gb": 16.52, "ce": 0.03286783130988624, "rank": 0.12276541441679001} +{"step": 316, "steps": 500, "lr": 2e-05, "grad_norm": 1.6181811094284058, "loss_scale": 4.0, "retries": 0, "seconds": 554.3, "max_memory_gb": 16.52, "ce": 0.31978846501442604, "rank": 0.0} +{"step": 317, "steps": 500, "lr": 2e-05, "grad_norm": 0.8636467456817627, "loss_scale": 4.0, "retries": 0, "seconds": 556.3, "max_memory_gb": 16.52, "ce": 0.08560014886916179, "rank": 0.0} +{"step": 318, "steps": 500, "lr": 2e-05, "grad_norm": 10.180645942687988, "loss_scale": 4.0, "retries": 0, "seconds": 558.2, "max_memory_gb": 16.52, "ce": 0.42658456196659245, "rank": 0.0} +{"step": 319, "steps": 500, "lr": 2e-05, "grad_norm": 1.069067120552063, "loss_scale": 4.0, "retries": 0, "seconds": 560.2, "max_memory_gb": 16.52, "ce": 0.07362802906209254, "rank": 0.0} +{"step": 320, "steps": 500, "lr": 2e-05, "grad_norm": 0.7803412079811096, "loss_scale": 4.0, "retries": 0, "seconds": 562.2, "max_memory_gb": 16.52, "ce": 0.06317172913077229, "rank": 0.0} +{"step": 321, "steps": 500, "lr": 2e-05, "grad_norm": 1.7590808868408203, "loss_scale": 4.0, "retries": 0, "seconds": 564.2, "max_memory_gb": 16.52, "ce": 0.1292030292606796, "rank": 0.12561067938804626} +{"step": 322, "steps": 500, "lr": 2e-05, "grad_norm": 0.25045979022979736, "loss_scale": 4.0, "retries": 0, "seconds": 566.3, "max_memory_gb": 16.52, "ce": 0.011322116775772884, "rank": 0.0} +{"step": 323, "steps": 500, "lr": 2e-05, "grad_norm": 2.2313945293426514, "loss_scale": 4.0, "retries": 0, "seconds": 568.3, "max_memory_gb": 16.52, "ce": 0.05587373322850908, "rank": 0.0} +{"step": 324, "steps": 500, "lr": 2e-05, "grad_norm": 0.13160845637321472, "loss_scale": 4.0, "retries": 0, "seconds": 570.3, "max_memory_gb": 16.52, "ce": 0.005304116164552397, "rank": 0.0} +{"step": 325, "steps": 500, "lr": 2e-05, "grad_norm": 0.9234480857849121, "loss_scale": 4.0, "retries": 0, "seconds": 572.2, "max_memory_gb": 16.52, "ce": 0.13376168005925138, "rank": 0.0} +{"step": 326, "steps": 500, "lr": 2e-05, "grad_norm": 0.5284091830253601, "loss_scale": 4.0, "retries": 0, "seconds": 574.1, "max_memory_gb": 16.52, "ce": 0.027228479779296322, "rank": 0.08886386454105377} +{"step": 327, "steps": 500, "lr": 2e-05, "grad_norm": 1.059540867805481, "loss_scale": 4.0, "retries": 0, "seconds": 576.2, "max_memory_gb": 16.52, "ce": 0.04585597950335796, "rank": 0.0} +{"step": 328, "steps": 500, "lr": 2e-05, "grad_norm": 1.0543211698532104, "loss_scale": 4.0, "retries": 0, "seconds": 578.0, "max_memory_gb": 16.52, "ce": 0.1366742006939603, "rank": 0.0} +{"step": 329, "steps": 500, "lr": 2e-05, "grad_norm": 1.2402180433273315, "loss_scale": 4.0, "retries": 0, "seconds": 579.8, "max_memory_gb": 16.52, "ce": 0.1441284595948673, "rank": 0.0} +{"step": 330, "steps": 500, "lr": 2e-05, "grad_norm": 0.8207722306251526, "loss_scale": 4.0, "retries": 0, "seconds": 581.6, "max_memory_gb": 16.52, "ce": 0.11780202022055164, "rank": 0.0} +{"step": 331, "steps": 500, "lr": 2e-05, "grad_norm": 2.30643630027771, "loss_scale": 4.0, "retries": 0, "seconds": 583.6, "max_memory_gb": 16.52, "ce": 0.30094235662181745, "rank": 0.0} +{"step": 332, "steps": 500, "lr": 2e-05, "grad_norm": 1.2831742763519287, "loss_scale": 4.0, "retries": 0, "seconds": 585.6, "max_memory_gb": 16.52, "ce": 0.05991505997008062, "rank": 0.0} +{"step": 333, "steps": 500, "lr": 2e-05, "grad_norm": 1.3712835311889648, "loss_scale": 4.0, "retries": 0, "seconds": 587.4, "max_memory_gb": 16.52, "ce": 0.22557410244917264, "rank": 0.0} +{"step": 334, "steps": 500, "lr": 2e-05, "grad_norm": 0.9753001928329468, "loss_scale": 4.0, "retries": 0, "seconds": 589.4, "max_memory_gb": 16.52, "ce": 0.09312207530092564, "rank": 0.0} +{"step": 335, "steps": 500, "lr": 2e-05, "grad_norm": 3.8479435443878174, "loss_scale": 4.0, "retries": 0, "seconds": 591.3, "max_memory_gb": 16.52, "ce": 0.29473002998929587, "rank": 0.0} +{"step": 336, "steps": 500, "lr": 2e-05, "grad_norm": 0.363756388425827, "loss_scale": 4.0, "retries": 0, "seconds": 593.2, "max_memory_gb": 16.52, "ce": 0.012588142235472333, "rank": 0.0} +{"step": 337, "steps": 500, "lr": 2e-05, "grad_norm": 1.8887150287628174, "loss_scale": 4.0, "retries": 0, "seconds": 595.1, "max_memory_gb": 16.52, "ce": 0.12339337135927053, "rank": 0.0} +{"step": 338, "steps": 500, "lr": 2e-05, "grad_norm": 0.9881619811058044, "loss_scale": 4.0, "retries": 0, "seconds": 597.0, "max_memory_gb": 16.52, "ce": 0.18466690274908615, "rank": 0.0} +{"step": 339, "steps": 500, "lr": 2e-05, "grad_norm": 0.976628303527832, "loss_scale": 4.0, "retries": 0, "seconds": 598.8, "max_memory_gb": 16.52, "ce": 0.11680370488647895, "rank": 0.0} +{"step": 340, "steps": 500, "lr": 2e-05, "grad_norm": 1.1124714612960815, "loss_scale": 4.0, "retries": 0, "seconds": 600.5, "max_memory_gb": 16.52, "ce": 0.17157654281800205, "rank": 0.0} +{"step": 341, "steps": 500, "lr": 2e-05, "grad_norm": 7.27989387512207, "loss_scale": 4.0, "retries": 0, "seconds": 602.4, "max_memory_gb": 16.52, "ce": 0.3511634520982625, "rank": 0.0} +{"step": 342, "steps": 500, "lr": 2e-05, "grad_norm": 1.8644386529922485, "loss_scale": 4.0, "retries": 0, "seconds": 604.2, "max_memory_gb": 16.52, "ce": 0.5492040032713703, "rank": 0.0} +{"step": 343, "steps": 500, "lr": 2e-05, "grad_norm": 1.4480373859405518, "loss_scale": 4.0, "retries": 0, "seconds": 606.0, "max_memory_gb": 16.52, "ce": 0.062049927296357055, "rank": 0.0} +{"step": 344, "steps": 500, "lr": 2e-05, "grad_norm": 0.2564651072025299, "loss_scale": 4.0, "retries": 0, "seconds": 607.7, "max_memory_gb": 16.52, "ce": 0.013870715636585373, "rank": 0.0} +{"step": 345, "steps": 500, "lr": 2e-05, "grad_norm": 0.5292305946350098, "loss_scale": 4.0, "retries": 0, "seconds": 609.4, "max_memory_gb": 16.52, "ce": 0.16736914422108384, "rank": 0.0} +{"step": 346, "steps": 500, "lr": 2e-05, "grad_norm": 1.7114250659942627, "loss_scale": 4.0, "retries": 0, "seconds": 611.1, "max_memory_gb": 16.52, "ce": 0.20206927637082117, "rank": 0.0} +{"step": 347, "steps": 500, "lr": 2e-05, "grad_norm": 0.9866185188293457, "loss_scale": 4.0, "retries": 0, "seconds": 612.8, "max_memory_gb": 16.52, "ce": 0.12697909248163342, "rank": 0.0} +{"step": 348, "steps": 500, "lr": 2e-05, "grad_norm": 0.07798321545124054, "loss_scale": 4.0, "retries": 0, "seconds": 614.6, "max_memory_gb": 16.52, "ce": 0.004174570078248507, "rank": 0.0} +{"step": 349, "steps": 500, "lr": 2e-05, "grad_norm": 0.24183107912540436, "loss_scale": 4.0, "retries": 0, "seconds": 616.3, "max_memory_gb": 16.52, "ce": 0.0731559148262022, "rank": 0.0} +{"step": 350, "steps": 500, "lr": 2e-05, "grad_norm": 0.46176958084106445, "loss_scale": 4.0, "retries": 0, "seconds": 618.0, "max_memory_gb": 16.52, "ce": 0.05478414674325904, "rank": 0.0} +{"step": 351, "steps": 500, "lr": 2e-05, "grad_norm": 1.0412895679473877, "loss_scale": 4.0, "retries": 0, "seconds": 619.7, "max_memory_gb": 16.52, "ce": 0.043981057156088355, "rank": 0.0} +{"step": 352, "steps": 500, "lr": 2e-05, "grad_norm": 2.3954825401306152, "loss_scale": 4.0, "retries": 0, "seconds": 621.4, "max_memory_gb": 16.52, "ce": 0.2175825863432692, "rank": 0.0} +{"step": 353, "steps": 500, "lr": 2e-05, "grad_norm": 0.9876068830490112, "loss_scale": 4.0, "retries": 0, "seconds": 623.1, "max_memory_gb": 16.52, "ce": 0.09905110492582025, "rank": 0.0} +{"step": 354, "steps": 500, "lr": 2e-05, "grad_norm": 0.7280654907226562, "loss_scale": 4.0, "retries": 0, "seconds": 624.8, "max_memory_gb": 16.52, "ce": 0.13512164276562544, "rank": 0.0} +{"step": 355, "steps": 500, "lr": 2e-05, "grad_norm": 0.3721054494380951, "loss_scale": 4.0, "retries": 0, "seconds": 626.6, "max_memory_gb": 16.52, "ce": 0.06871308549125388, "rank": 0.0} +{"step": 356, "steps": 500, "lr": 2e-05, "grad_norm": 1.7462234497070312, "loss_scale": 4.0, "retries": 0, "seconds": 628.4, "max_memory_gb": 16.52, "ce": 0.07684234223597741, "rank": 0.0} +{"step": 357, "steps": 500, "lr": 2e-05, "grad_norm": 0.8926112651824951, "loss_scale": 4.0, "retries": 0, "seconds": 630.1, "max_memory_gb": 16.52, "ce": 0.06229114103462052, "rank": 0.0} +{"step": 358, "steps": 500, "lr": 2e-05, "grad_norm": 1.21731698513031, "loss_scale": 4.0, "retries": 0, "seconds": 631.8, "max_memory_gb": 16.52, "ce": 0.1752621475898195, "rank": 0.0} +{"step": 359, "steps": 500, "lr": 2e-05, "grad_norm": 0.8096113801002502, "loss_scale": 4.0, "retries": 0, "seconds": 633.6, "max_memory_gb": 16.52, "ce": 0.047264741069284355, "rank": 0.0} +{"step": 360, "steps": 500, "lr": 2e-05, "grad_norm": 1.2506651878356934, "loss_scale": 4.0, "retries": 0, "seconds": 635.3, "max_memory_gb": 16.52, "ce": 0.054198039666516706, "rank": 0.13241171836853027} +{"step": 361, "steps": 500, "lr": 2e-05, "grad_norm": 1.7898190021514893, "loss_scale": 4.0, "retries": 0, "seconds": 637.0, "max_memory_gb": 16.52, "ce": 0.13260929727039183, "rank": 0.0} +{"step": 362, "steps": 500, "lr": 2e-05, "grad_norm": 2.7369093894958496, "loss_scale": 4.0, "retries": 0, "seconds": 638.8, "max_memory_gb": 16.52, "ce": 0.1682320407780935, "rank": 0.0} +{"step": 363, "steps": 500, "lr": 2e-05, "grad_norm": 0.43502798676490784, "loss_scale": 4.0, "retries": 0, "seconds": 640.5, "max_memory_gb": 16.52, "ce": 0.031668355095462175, "rank": 0.0} +{"step": 364, "steps": 500, "lr": 2e-05, "grad_norm": 5.424763202667236, "loss_scale": 4.0, "retries": 0, "seconds": 642.2, "max_memory_gb": 16.52, "ce": 0.06127132427900506, "rank": 0.0} +{"step": 365, "steps": 500, "lr": 2e-05, "grad_norm": 0.9886475801467896, "loss_scale": 4.0, "retries": 0, "seconds": 644.0, "max_memory_gb": 16.52, "ce": 0.08039464622743253, "rank": 0.0} +{"step": 366, "steps": 500, "lr": 2e-05, "grad_norm": 1.9985132217407227, "loss_scale": 4.0, "retries": 0, "seconds": 645.7, "max_memory_gb": 16.52, "ce": 0.11735318914725212, "rank": 0.0} +{"step": 367, "steps": 500, "lr": 2e-05, "grad_norm": 0.7203109264373779, "loss_scale": 4.0, "retries": 0, "seconds": 647.4, "max_memory_gb": 16.52, "ce": 0.08027061777147537, "rank": 0.0} +{"step": 368, "steps": 500, "lr": 2e-05, "grad_norm": 0.5572752952575684, "loss_scale": 4.0, "retries": 0, "seconds": 649.1, "max_memory_gb": 16.52, "ce": 0.016352547112546745, "rank": 0.0} +{"step": 369, "steps": 500, "lr": 2e-05, "grad_norm": 0.4270314872264862, "loss_scale": 4.0, "retries": 0, "seconds": 650.8, "max_memory_gb": 16.52, "ce": 0.02361961123324363, "rank": 0.10173594951629639} +{"step": 370, "steps": 500, "lr": 2e-05, "grad_norm": 4.469074726104736, "loss_scale": 4.0, "retries": 0, "seconds": 652.5, "max_memory_gb": 16.52, "ce": 0.465076542602219, "rank": 0.0} +{"step": 371, "steps": 500, "lr": 2e-05, "grad_norm": 1.4400019645690918, "loss_scale": 4.0, "retries": 0, "seconds": 654.3, "max_memory_gb": 16.52, "ce": 0.045727235264166666, "rank": 0.0} +{"step": 372, "steps": 500, "lr": 2e-05, "grad_norm": 0.30771583318710327, "loss_scale": 4.0, "retries": 0, "seconds": 656.0, "max_memory_gb": 16.52, "ce": 0.019342878303177713, "rank": 0.0} +{"step": 373, "steps": 500, "lr": 2e-05, "grad_norm": 0.9596928358078003, "loss_scale": 4.0, "retries": 0, "seconds": 657.7, "max_memory_gb": 16.52, "ce": 0.08712627431759756, "rank": 0.0} +{"step": 374, "steps": 500, "lr": 2e-05, "grad_norm": 0.9808799028396606, "loss_scale": 4.0, "retries": 0, "seconds": 659.4, "max_memory_gb": 16.52, "ce": 0.12128202170515578, "rank": 0.0} +{"step": 375, "steps": 500, "lr": 2e-05, "grad_norm": 0.7353779673576355, "loss_scale": 4.0, "retries": 0, "seconds": 661.1, "max_memory_gb": 16.52, "ce": 0.1467368176281525, "rank": 0.0} +{"step": 376, "steps": 500, "lr": 2e-05, "grad_norm": 0.5550479888916016, "loss_scale": 4.0, "retries": 0, "seconds": 662.9, "max_memory_gb": 16.52, "ce": 0.03909980161097337, "rank": 0.0} +{"step": 377, "steps": 500, "lr": 2e-05, "grad_norm": 0.7084699273109436, "loss_scale": 4.0, "retries": 0, "seconds": 664.6, "max_memory_gb": 16.52, "ce": 0.08513851066163625, "rank": 0.0} +{"step": 378, "steps": 500, "lr": 2e-05, "grad_norm": 1.1231087446212769, "loss_scale": 4.0, "retries": 0, "seconds": 666.4, "max_memory_gb": 16.52, "ce": 0.06688638035438998, "rank": 0.0} +{"step": 379, "steps": 500, "lr": 2e-05, "grad_norm": 1.2627782821655273, "loss_scale": 4.0, "retries": 0, "seconds": 668.1, "max_memory_gb": 16.52, "ce": 0.10482693245012342, "rank": 0.0} +{"step": 380, "steps": 500, "lr": 2e-05, "grad_norm": 7.6371870040893555, "loss_scale": 4.0, "retries": 0, "seconds": 669.8, "max_memory_gb": 16.52, "ce": 0.22515678824993302, "rank": 0.0} +{"step": 381, "steps": 500, "lr": 2e-05, "grad_norm": 1.0345795154571533, "loss_scale": 4.0, "retries": 0, "seconds": 671.5, "max_memory_gb": 16.52, "ce": 0.21614456062616227, "rank": 0.0} +{"step": 382, "steps": 500, "lr": 2e-05, "grad_norm": 1.7024558782577515, "loss_scale": 4.0, "retries": 0, "seconds": 673.2, "max_memory_gb": 16.52, "ce": 0.3058068259488209, "rank": 0.0} +{"step": 383, "steps": 500, "lr": 2e-05, "grad_norm": 0.4653025269508362, "loss_scale": 4.0, "retries": 0, "seconds": 675.0, "max_memory_gb": 16.52, "ce": 0.07211448282578203, "rank": 0.0} +{"step": 384, "steps": 500, "lr": 2e-05, "grad_norm": 1.2549583911895752, "loss_scale": 4.0, "retries": 0, "seconds": 676.7, "max_memory_gb": 16.52, "ce": 0.04504821736009035, "rank": 0.0} +{"step": 385, "steps": 500, "lr": 2e-05, "grad_norm": 0.9845272898674011, "loss_scale": 4.0, "retries": 0, "seconds": 678.4, "max_memory_gb": 16.52, "ce": 0.13696212583272427, "rank": 0.0} +{"step": 386, "steps": 500, "lr": 2e-05, "grad_norm": 0.7609853148460388, "loss_scale": 4.0, "retries": 0, "seconds": 680.2, "max_memory_gb": 16.52, "ce": 0.052307698906588485, "rank": 0.0} +{"step": 387, "steps": 500, "lr": 2e-05, "grad_norm": 0.5194689631462097, "loss_scale": 4.0, "retries": 0, "seconds": 681.9, "max_memory_gb": 16.52, "ce": 0.04572510479511038, "rank": 0.0} +{"step": 388, "steps": 500, "lr": 2e-05, "grad_norm": 2.0003693103790283, "loss_scale": 4.0, "retries": 0, "seconds": 683.6, "max_memory_gb": 16.52, "ce": 0.0756023697249475, "rank": 0.0} +{"step": 389, "steps": 500, "lr": 2e-05, "grad_norm": 0.14844848215579987, "loss_scale": 4.0, "retries": 0, "seconds": 685.4, "max_memory_gb": 16.52, "ce": 0.00413438994291937, "rank": 0.0} +{"step": 390, "steps": 500, "lr": 2e-05, "grad_norm": 9.557419776916504, "loss_scale": 4.0, "retries": 0, "seconds": 687.1, "max_memory_gb": 16.52, "ce": 0.1021155726152756, "rank": 0.0} +{"step": 391, "steps": 500, "lr": 2e-05, "grad_norm": 2.0066311359405518, "loss_scale": 4.0, "retries": 0, "seconds": 688.8, "max_memory_gb": 16.52, "ce": 0.23073521595597413, "rank": 0.0} +{"step": 392, "steps": 500, "lr": 2e-05, "grad_norm": 0.19140957295894623, "loss_scale": 4.0, "retries": 0, "seconds": 690.6, "max_memory_gb": 16.52, "ce": 0.0009548369766889664, "rank": 0.08815581351518631} +{"step": 393, "steps": 500, "lr": 2e-05, "grad_norm": 1.1934648752212524, "loss_scale": 4.0, "retries": 0, "seconds": 692.3, "max_memory_gb": 16.52, "ce": 0.11345006615965758, "rank": 0.0} +{"step": 394, "steps": 500, "lr": 2e-05, "grad_norm": 0.978804886341095, "loss_scale": 4.0, "retries": 0, "seconds": 694.1, "max_memory_gb": 16.52, "ce": 0.09059870378268897, "rank": 0.0} +{"step": 395, "steps": 500, "lr": 2e-05, "grad_norm": 0.3115301728248596, "loss_scale": 4.0, "retries": 0, "seconds": 695.8, "max_memory_gb": 16.52, "ce": 0.009157484782008396, "rank": 0.0} +{"step": 396, "steps": 500, "lr": 2e-05, "grad_norm": 0.7048264145851135, "loss_scale": 4.0, "retries": 0, "seconds": 697.5, "max_memory_gb": 16.52, "ce": 0.026877014160163526, "rank": 0.0} +{"step": 397, "steps": 500, "lr": 2e-05, "grad_norm": 0.44518542289733887, "loss_scale": 4.0, "retries": 0, "seconds": 699.1, "max_memory_gb": 16.52, "ce": 0.09832353141746353, "rank": 0.0} +{"step": 398, "steps": 500, "lr": 2e-05, "grad_norm": 0.5595195293426514, "loss_scale": 4.0, "retries": 0, "seconds": 700.9, "max_memory_gb": 16.52, "ce": 0.05772007160840076, "rank": 0.0} +{"step": 399, "steps": 500, "lr": 2e-05, "grad_norm": 2.204803705215454, "loss_scale": 4.0, "retries": 0, "seconds": 702.9, "max_memory_gb": 16.52, "ce": 0.17932405657302297, "rank": 0.0} +{"step": 400, "steps": 500, "lr": 2e-05, "grad_norm": 0.5600865483283997, "loss_scale": 4.0, "retries": 0, "seconds": 704.8, "max_memory_gb": 16.52, "ce": 0.013135826820871443, "rank": 0.0} +{"step": 401, "steps": 500, "lr": 2e-05, "grad_norm": 2.5708436965942383, "loss_scale": 4.0, "retries": 0, "seconds": 706.7, "max_memory_gb": 16.52, "ce": 0.18241532924866988, "rank": 0.0} +{"step": 402, "steps": 500, "lr": 2e-05, "grad_norm": 0.7501639127731323, "loss_scale": 4.0, "retries": 0, "seconds": 708.6, "max_memory_gb": 16.52, "ce": 0.09517433730798075, "rank": 0.11792561411857605} +{"step": 403, "steps": 500, "lr": 2e-05, "grad_norm": 0.05668950453400612, "loss_scale": 4.0, "retries": 0, "seconds": 710.5, "max_memory_gb": 16.52, "ce": 0.0029143150750314817, "rank": 0.0} +{"step": 404, "steps": 500, "lr": 2e-05, "grad_norm": 1.5096426010131836, "loss_scale": 4.0, "retries": 0, "seconds": 712.5, "max_memory_gb": 16.52, "ce": 0.14159816920118828, "rank": 0.0} +{"step": 405, "steps": 500, "lr": 2e-05, "grad_norm": 0.3834083676338196, "loss_scale": 4.0, "retries": 0, "seconds": 714.4, "max_memory_gb": 16.52, "ce": 0.015047706196583022, "rank": 0.0} +{"step": 406, "steps": 500, "lr": 2e-05, "grad_norm": 0.8448371887207031, "loss_scale": 4.0, "retries": 0, "seconds": 716.2, "max_memory_gb": 16.52, "ce": 0.10608635779681208, "rank": 0.0} +{"step": 407, "steps": 500, "lr": 2e-05, "grad_norm": 2.30814528465271, "loss_scale": 4.0, "retries": 0, "seconds": 717.9, "max_memory_gb": 16.52, "ce": 0.7462588680800764, "rank": 0.20267164707183838} +{"step": 408, "steps": 500, "lr": 2e-05, "grad_norm": 1.9265966415405273, "loss_scale": 4.0, "retries": 0, "seconds": 719.6, "max_memory_gb": 16.52, "ce": 0.3547918751855832, "rank": 0.0} +{"step": 409, "steps": 500, "lr": 2e-05, "grad_norm": 0.4413246810436249, "loss_scale": 4.0, "retries": 0, "seconds": 721.4, "max_memory_gb": 16.52, "ce": 0.11893945236306536, "rank": 0.0} +{"step": 410, "steps": 500, "lr": 2e-05, "grad_norm": 0.31027382612228394, "loss_scale": 4.0, "retries": 0, "seconds": 723.1, "max_memory_gb": 16.52, "ce": 0.007446829109085229, "rank": 0.0} +{"step": 411, "steps": 500, "lr": 2e-05, "grad_norm": 6.019716739654541, "loss_scale": 4.0, "retries": 0, "seconds": 724.8, "max_memory_gb": 16.52, "ce": 0.25604984151914323, "rank": 0.0} +{"step": 412, "steps": 500, "lr": 2e-05, "grad_norm": 0.47691503167152405, "loss_scale": 4.0, "retries": 0, "seconds": 726.5, "max_memory_gb": 16.52, "ce": 0.13332893238657562, "rank": 0.12496024370193481} +{"step": 413, "steps": 500, "lr": 2e-05, "grad_norm": 0.5439453721046448, "loss_scale": 4.0, "retries": 0, "seconds": 728.2, "max_memory_gb": 16.52, "ce": 0.023697425356203894, "rank": 0.0} +{"step": 414, "steps": 500, "lr": 2e-05, "grad_norm": 2.414050579071045, "loss_scale": 4.0, "retries": 0, "seconds": 729.9, "max_memory_gb": 16.52, "ce": 0.20049165166528837, "rank": 0.0} +{"step": 415, "steps": 500, "lr": 2e-05, "grad_norm": 3.7176437377929688, "loss_scale": 4.0, "retries": 0, "seconds": 731.6, "max_memory_gb": 16.52, "ce": 0.3381141001332253, "rank": 0.0} +{"step": 416, "steps": 500, "lr": 2e-05, "grad_norm": 0.07537741214036942, "loss_scale": 4.0, "retries": 0, "seconds": 733.3, "max_memory_gb": 16.52, "ce": 0.004114057237075031, "rank": 0.0} +{"step": 417, "steps": 500, "lr": 2e-05, "grad_norm": 2.0089428424835205, "loss_scale": 4.0, "retries": 0, "seconds": 735.0, "max_memory_gb": 16.52, "ce": 0.3620541049226631, "rank": 0.12714849412441254} +{"step": 418, "steps": 500, "lr": 2e-05, "grad_norm": 0.6624183654785156, "loss_scale": 4.0, "retries": 0, "seconds": 736.7, "max_memory_gb": 16.52, "ce": 0.043429952378573944, "rank": 0.0} +{"step": 419, "steps": 500, "lr": 2e-05, "grad_norm": 0.4605428874492645, "loss_scale": 4.0, "retries": 0, "seconds": 738.4, "max_memory_gb": 16.52, "ce": 0.05836399505369627, "rank": 0.0} +{"step": 420, "steps": 500, "lr": 2e-05, "grad_norm": 3.3622426986694336, "loss_scale": 4.0, "retries": 0, "seconds": 740.1, "max_memory_gb": 16.52, "ce": 0.2727994295341887, "rank": 0.07338040322065353} +{"step": 421, "steps": 500, "lr": 2e-05, "grad_norm": 0.3512769639492035, "loss_scale": 4.0, "retries": 0, "seconds": 741.8, "max_memory_gb": 16.52, "ce": 0.022924976626200078, "rank": 0.11782291531562805} +{"step": 422, "steps": 500, "lr": 2e-05, "grad_norm": 2.3439018726348877, "loss_scale": 4.0, "retries": 0, "seconds": 743.6, "max_memory_gb": 16.52, "ce": 0.21917216873225698, "rank": 0.0} +{"step": 423, "steps": 500, "lr": 2e-05, "grad_norm": 5.967458724975586, "loss_scale": 4.0, "retries": 0, "seconds": 745.3, "max_memory_gb": 16.52, "ce": 0.440176995280126, "rank": 0.0} +{"step": 424, "steps": 500, "lr": 2e-05, "grad_norm": 0.8061440587043762, "loss_scale": 4.0, "retries": 0, "seconds": 747.1, "max_memory_gb": 16.52, "ce": 0.04295068199371599, "rank": 0.0} +{"step": 425, "steps": 500, "lr": 2e-05, "grad_norm": 0.878563404083252, "loss_scale": 4.0, "retries": 0, "seconds": 748.9, "max_memory_gb": 16.52, "ce": 0.05302099948494288, "rank": 0.0} +{"step": 426, "steps": 500, "lr": 2e-05, "grad_norm": 1.3252856731414795, "loss_scale": 4.0, "retries": 0, "seconds": 750.7, "max_memory_gb": 16.52, "ce": 0.14572230954581755, "rank": 0.0} +{"step": 427, "steps": 500, "lr": 2e-05, "grad_norm": 0.4101584255695343, "loss_scale": 4.0, "retries": 0, "seconds": 752.5, "max_memory_gb": 16.52, "ce": 0.04053709377103587, "rank": 0.0} +{"step": 428, "steps": 500, "lr": 2e-05, "grad_norm": 1.3316981792449951, "loss_scale": 4.0, "retries": 0, "seconds": 754.2, "max_memory_gb": 16.52, "ce": 0.04152867108996361, "rank": 0.0} +{"step": 429, "steps": 500, "lr": 2e-05, "grad_norm": 0.4665939509868622, "loss_scale": 4.0, "retries": 0, "seconds": 755.9, "max_memory_gb": 16.52, "ce": 0.017382046381044347, "rank": 0.0} +{"step": 430, "steps": 500, "lr": 2e-05, "grad_norm": 0.82137131690979, "loss_scale": 4.0, "retries": 0, "seconds": 757.7, "max_memory_gb": 16.52, "ce": 0.07737021075445227, "rank": 0.0} +{"step": 431, "steps": 500, "lr": 2e-05, "grad_norm": 7.303253173828125, "loss_scale": 4.0, "retries": 0, "seconds": 759.4, "max_memory_gb": 16.52, "ce": 0.509389283017299, "rank": 0.0} +{"step": 432, "steps": 500, "lr": 2e-05, "grad_norm": 0.7113655209541321, "loss_scale": 4.0, "retries": 0, "seconds": 761.2, "max_memory_gb": 16.52, "ce": 0.04052074174342124, "rank": 0.0} +{"step": 433, "steps": 500, "lr": 2e-05, "grad_norm": 1.8037166595458984, "loss_scale": 4.0, "retries": 0, "seconds": 762.9, "max_memory_gb": 16.52, "ce": 0.06260666433036022, "rank": 0.0} +{"step": 434, "steps": 500, "lr": 2e-05, "grad_norm": 0.17844361066818237, "loss_scale": 4.0, "retries": 0, "seconds": 764.6, "max_memory_gb": 16.52, "ce": 0.0039871671397122554, "rank": 0.0} +{"step": 435, "steps": 500, "lr": 2e-05, "grad_norm": 1.6653481721878052, "loss_scale": 4.0, "retries": 0, "seconds": 766.3, "max_memory_gb": 16.52, "ce": 0.15783320200625894, "rank": 0.0} +{"step": 436, "steps": 500, "lr": 2e-05, "grad_norm": 1.8718249797821045, "loss_scale": 4.0, "retries": 0, "seconds": 768.1, "max_memory_gb": 16.52, "ce": 0.11047638047102737, "rank": 0.0} +{"step": 437, "steps": 500, "lr": 2e-05, "grad_norm": 5.086726188659668, "loss_scale": 4.0, "retries": 0, "seconds": 769.8, "max_memory_gb": 16.52, "ce": 0.20129796294622793, "rank": 0.1155027523636818} +{"step": 438, "steps": 500, "lr": 2e-05, "grad_norm": 1.0685867071151733, "loss_scale": 4.0, "retries": 0, "seconds": 771.8, "max_memory_gb": 16.52, "ce": 0.06344812771203578, "rank": 0.0} +{"step": 439, "steps": 500, "lr": 2e-05, "grad_norm": 0.9330595135688782, "loss_scale": 4.0, "retries": 0, "seconds": 773.7, "max_memory_gb": 16.52, "ce": 0.10788267208044999, "rank": 0.0} +{"step": 440, "steps": 500, "lr": 2e-05, "grad_norm": 0.4671139419078827, "loss_scale": 4.0, "retries": 0, "seconds": 775.6, "max_memory_gb": 16.52, "ce": 0.0272819276442533, "rank": 0.0} +{"step": 441, "steps": 500, "lr": 2e-05, "grad_norm": 5.039419174194336, "loss_scale": 4.0, "retries": 0, "seconds": 777.6, "max_memory_gb": 16.52, "ce": 0.2281367562527521, "rank": 0.0} +{"step": 442, "steps": 500, "lr": 2e-05, "grad_norm": 1.0084294080734253, "loss_scale": 4.0, "retries": 0, "seconds": 779.4, "max_memory_gb": 16.52, "ce": 0.043672794515714486, "rank": 0.0} +{"step": 443, "steps": 500, "lr": 2e-05, "grad_norm": 0.6629226803779602, "loss_scale": 4.0, "retries": 0, "seconds": 781.4, "max_memory_gb": 16.52, "ce": 0.05046477224641421, "rank": 0.0} +{"step": 444, "steps": 500, "lr": 2e-05, "grad_norm": 0.41022828221321106, "loss_scale": 4.0, "retries": 0, "seconds": 783.2, "max_memory_gb": 16.52, "ce": 0.04183369903375933, "rank": 0.0} +{"step": 445, "steps": 500, "lr": 2e-05, "grad_norm": 2.1085760593414307, "loss_scale": 4.0, "retries": 0, "seconds": 785.2, "max_memory_gb": 16.52, "ce": 0.37794847897248474, "rank": 0.0900210291147232} +{"step": 446, "steps": 500, "lr": 2e-05, "grad_norm": 6.840194225311279, "loss_scale": 4.0, "retries": 0, "seconds": 787.1, "max_memory_gb": 16.52, "ce": 0.09922184862398353, "rank": 0.007966563105583191} +{"step": 447, "steps": 500, "lr": 2e-05, "grad_norm": 0.4864804446697235, "loss_scale": 4.0, "retries": 0, "seconds": 789.0, "max_memory_gb": 16.52, "ce": 0.12842909690516535, "rank": 0.0} +{"step": 448, "steps": 500, "lr": 2e-05, "grad_norm": 2.733489751815796, "loss_scale": 4.0, "retries": 0, "seconds": 790.9, "max_memory_gb": 16.52, "ce": 0.14286683374211862, "rank": 0.0} +{"step": 449, "steps": 500, "lr": 2e-05, "grad_norm": 1.212350606918335, "loss_scale": 4.0, "retries": 0, "seconds": 792.8, "max_memory_gb": 16.52, "ce": 0.015596612080116756, "rank": 0.0} +{"step": 450, "steps": 500, "lr": 2e-05, "grad_norm": 1.397266149520874, "loss_scale": 4.0, "retries": 0, "seconds": 794.8, "max_memory_gb": 16.52, "ce": 0.05601653800204076, "rank": 0.0} +{"step": 451, "steps": 500, "lr": 2e-05, "grad_norm": 1.1858645677566528, "loss_scale": 4.0, "retries": 0, "seconds": 796.7, "max_memory_gb": 16.52, "ce": 0.02388532839995605, "rank": 0.0} +{"step": 452, "steps": 500, "lr": 2e-05, "grad_norm": 2.177257537841797, "loss_scale": 4.0, "retries": 0, "seconds": 798.5, "max_memory_gb": 16.52, "ce": 0.18825229495587337, "rank": 0.0} +{"step": 453, "steps": 500, "lr": 2e-05, "grad_norm": 0.6394577026367188, "loss_scale": 4.0, "retries": 0, "seconds": 800.2, "max_memory_gb": 16.52, "ce": 0.027830750161228934, "rank": 0.0} +{"step": 454, "steps": 500, "lr": 2e-05, "grad_norm": 21.191116333007812, "loss_scale": 4.0, "retries": 0, "seconds": 802.0, "max_memory_gb": 16.52, "ce": 0.619664956007, "rank": 0.0} +{"step": 455, "steps": 500, "lr": 2e-05, "grad_norm": 2.4244813919067383, "loss_scale": 4.0, "retries": 0, "seconds": 803.7, "max_memory_gb": 16.52, "ce": 0.06301250284377602, "rank": 0.0} +{"step": 456, "steps": 500, "lr": 2e-05, "grad_norm": 0.11193536967039108, "loss_scale": 4.0, "retries": 0, "seconds": 805.5, "max_memory_gb": 16.52, "ce": 0.005567944807353342, "rank": 0.0} +{"step": 457, "steps": 500, "lr": 2e-05, "grad_norm": 0.7920624017715454, "loss_scale": 4.0, "retries": 0, "seconds": 807.5, "max_memory_gb": 16.52, "ce": 0.019706492158547917, "rank": 0.04783420264720917} +{"step": 458, "steps": 500, "lr": 2e-05, "grad_norm": 1.996948480606079, "loss_scale": 4.0, "retries": 0, "seconds": 809.4, "max_memory_gb": 16.52, "ce": 0.18687392638184974, "rank": 0.0} +{"step": 459, "steps": 500, "lr": 2e-05, "grad_norm": 3.2324557304382324, "loss_scale": 4.0, "retries": 0, "seconds": 811.4, "max_memory_gb": 16.52, "ce": 0.1308331047930551, "rank": 0.0} +{"step": 460, "steps": 500, "lr": 2e-05, "grad_norm": 0.11378471553325653, "loss_scale": 4.0, "retries": 0, "seconds": 813.3, "max_memory_gb": 16.52, "ce": 0.006997978922754555, "rank": 0.0} +{"step": 461, "steps": 500, "lr": 2e-05, "grad_norm": 1.820981502532959, "loss_scale": 4.0, "retries": 0, "seconds": 815.3, "max_memory_gb": 16.52, "ce": 0.2864447124838989, "rank": 0.0} +{"step": 462, "steps": 500, "lr": 2e-05, "grad_norm": 0.681878924369812, "loss_scale": 4.0, "retries": 0, "seconds": 817.3, "max_memory_gb": 16.52, "ce": 0.02497870097249688, "rank": 0.0} +{"step": 463, "steps": 500, "lr": 2e-05, "grad_norm": 0.3357759118080139, "loss_scale": 4.0, "retries": 0, "seconds": 819.1, "max_memory_gb": 16.52, "ce": 0.011122040933059907, "rank": 0.0} +{"step": 464, "steps": 500, "lr": 2e-05, "grad_norm": 0.03181097283959389, "loss_scale": 4.0, "retries": 0, "seconds": 820.9, "max_memory_gb": 16.52, "ce": 0.0015256335050253256, "rank": 0.0} +{"step": 465, "steps": 500, "lr": 2e-05, "grad_norm": 1.8802084922790527, "loss_scale": 4.0, "retries": 0, "seconds": 822.7, "max_memory_gb": 16.52, "ce": 0.16831098168813696, "rank": 0.0} +{"step": 466, "steps": 500, "lr": 2e-05, "grad_norm": 1.3956329822540283, "loss_scale": 4.0, "retries": 0, "seconds": 824.4, "max_memory_gb": 16.52, "ce": 0.07048868641186345, "rank": 0.0} +{"step": 467, "steps": 500, "lr": 2e-05, "grad_norm": 1.0717867612838745, "loss_scale": 4.0, "retries": 0, "seconds": 826.1, "max_memory_gb": 16.52, "ce": 0.028406039744368172, "rank": 0.0} +{"step": 468, "steps": 500, "lr": 2e-05, "grad_norm": 4.017094135284424, "loss_scale": 4.0, "retries": 0, "seconds": 827.8, "max_memory_gb": 16.52, "ce": 0.034876498390985944, "rank": 0.0} +{"step": 469, "steps": 500, "lr": 2e-05, "grad_norm": 1.9858306646347046, "loss_scale": 4.0, "retries": 0, "seconds": 829.6, "max_memory_gb": 16.52, "ce": 0.14107946856665876, "rank": 0.0} +{"step": 470, "steps": 500, "lr": 2e-05, "grad_norm": 1.5183229446411133, "loss_scale": 4.0, "retries": 0, "seconds": 831.3, "max_memory_gb": 16.52, "ce": 0.0895264404352929, "rank": 0.0} +{"step": 471, "steps": 500, "lr": 2e-05, "grad_norm": 1.9452062845230103, "loss_scale": 4.0, "retries": 0, "seconds": 833.0, "max_memory_gb": 16.52, "ce": 0.17625760948203606, "rank": 0.0} +{"step": 472, "steps": 500, "lr": 2e-05, "grad_norm": 1.305860161781311, "loss_scale": 4.0, "retries": 0, "seconds": 834.8, "max_memory_gb": 16.52, "ce": 0.05883448895974652, "rank": 0.0} +{"step": 473, "steps": 500, "lr": 2e-05, "grad_norm": 0.13068890571594238, "loss_scale": 4.0, "retries": 0, "seconds": 836.5, "max_memory_gb": 16.52, "ce": 0.005707339183572913, "rank": 0.12533695995807648} +{"step": 474, "steps": 500, "lr": 2e-05, "grad_norm": 0.9857247471809387, "loss_scale": 4.0, "retries": 0, "seconds": 838.2, "max_memory_gb": 16.52, "ce": 0.10336583139860522, "rank": 0.0} +{"step": 475, "steps": 500, "lr": 2e-05, "grad_norm": 1.8596112728118896, "loss_scale": 4.0, "retries": 0, "seconds": 839.9, "max_memory_gb": 16.52, "ce": 0.16987990101188188, "rank": 0.0} +{"step": 476, "steps": 500, "lr": 2e-05, "grad_norm": 0.12262877076864243, "loss_scale": 4.0, "retries": 0, "seconds": 841.7, "max_memory_gb": 16.52, "ce": 0.005217082174567622, "rank": 0.0} +{"step": 477, "steps": 500, "lr": 2e-05, "grad_norm": 1.0808348655700684, "loss_scale": 4.0, "retries": 0, "seconds": 843.4, "max_memory_gb": 16.52, "ce": 0.02826199782339245, "rank": 0.0} +{"step": 478, "steps": 500, "lr": 2e-05, "grad_norm": 1.0072513818740845, "loss_scale": 4.0, "retries": 0, "seconds": 845.1, "max_memory_gb": 16.52, "ce": 0.07168566937161813, "rank": 0.0} +{"step": 479, "steps": 500, "lr": 2e-05, "grad_norm": 3.4465761184692383, "loss_scale": 4.0, "retries": 0, "seconds": 846.9, "max_memory_gb": 16.52, "ce": 0.31235845180617616, "rank": 0.038096725940704346} +{"step": 480, "steps": 500, "lr": 2e-05, "grad_norm": 0.7310360670089722, "loss_scale": 4.0, "retries": 0, "seconds": 848.6, "max_memory_gb": 16.52, "ce": 0.10917540245645796, "rank": 0.0} +{"step": 481, "steps": 500, "lr": 2e-05, "grad_norm": 25.83781623840332, "loss_scale": 4.0, "retries": 0, "seconds": 850.3, "max_memory_gb": 16.52, "ce": 0.34983618242722514, "rank": 0.0} +{"step": 482, "steps": 500, "lr": 2e-05, "grad_norm": 1.4484992027282715, "loss_scale": 4.0, "retries": 0, "seconds": 852.0, "max_memory_gb": 16.52, "ce": 0.07471220633669873, "rank": 0.0} +{"step": 483, "steps": 500, "lr": 2e-05, "grad_norm": 5.007294178009033, "loss_scale": 4.0, "retries": 0, "seconds": 853.7, "max_memory_gb": 16.52, "ce": 0.620019138364114, "rank": 0.0} +{"step": 484, "steps": 500, "lr": 2e-05, "grad_norm": 1.8797458410263062, "loss_scale": 4.0, "retries": 0, "seconds": 855.4, "max_memory_gb": 16.52, "ce": 0.45980171003452597, "rank": 0.0} +{"step": 485, "steps": 500, "lr": 2e-05, "grad_norm": 1.2578699588775635, "loss_scale": 4.0, "retries": 0, "seconds": 857.2, "max_memory_gb": 16.52, "ce": 0.060725050148448645, "rank": 0.0} +{"step": 486, "steps": 500, "lr": 2e-05, "grad_norm": 2.277761697769165, "loss_scale": 4.0, "retries": 0, "seconds": 858.9, "max_memory_gb": 16.52, "ce": 0.20727613064264006, "rank": 0.0} +{"step": 487, "steps": 500, "lr": 2e-05, "grad_norm": 10.96161937713623, "loss_scale": 4.0, "retries": 0, "seconds": 860.5, "max_memory_gb": 16.52, "ce": 0.19913891430041986, "rank": 0.0} +{"step": 488, "steps": 500, "lr": 2e-05, "grad_norm": 0.46668940782546997, "loss_scale": 4.0, "retries": 0, "seconds": 862.3, "max_memory_gb": 16.52, "ce": 0.06684968924901113, "rank": 0.0} +{"step": 489, "steps": 500, "lr": 2e-05, "grad_norm": 3.0120649337768555, "loss_scale": 4.0, "retries": 0, "seconds": 864.0, "max_memory_gb": 16.52, "ce": 0.24836270867626808, "rank": 0.0} +{"step": 490, "steps": 500, "lr": 2e-05, "grad_norm": 0.46302342414855957, "loss_scale": 4.0, "retries": 0, "seconds": 865.7, "max_memory_gb": 16.52, "ce": 0.027257392795490887, "rank": 0.0} +{"step": 491, "steps": 500, "lr": 2e-05, "grad_norm": 0.7155588865280151, "loss_scale": 4.0, "retries": 0, "seconds": 867.4, "max_memory_gb": 16.52, "ce": 0.11668906279464863, "rank": 0.0} +{"step": 492, "steps": 500, "lr": 2e-05, "grad_norm": 0.13381944596767426, "loss_scale": 4.0, "retries": 0, "seconds": 869.1, "max_memory_gb": 16.52, "ce": 0.005028672421758529, "rank": 0.0} +{"step": 493, "steps": 500, "lr": 2e-05, "grad_norm": 0.9593689441680908, "loss_scale": 4.0, "retries": 0, "seconds": 870.9, "max_memory_gb": 16.52, "ce": 0.03947545890969195, "rank": 0.0} +{"step": 494, "steps": 500, "lr": 2e-05, "grad_norm": 1.073704719543457, "loss_scale": 4.0, "retries": 0, "seconds": 872.6, "max_memory_gb": 16.52, "ce": 0.07570963288117127, "rank": 0.0} +{"step": 495, "steps": 500, "lr": 2e-05, "grad_norm": 0.7578890323638916, "loss_scale": 4.0, "retries": 0, "seconds": 874.3, "max_memory_gb": 16.52, "ce": 0.03795197919225757, "rank": 0.0} +{"step": 496, "steps": 500, "lr": 2e-05, "grad_norm": 0.1653193086385727, "loss_scale": 4.0, "retries": 0, "seconds": 876.0, "max_memory_gb": 16.52, "ce": 0.006273273911574506, "rank": 0.0} +{"step": 497, "steps": 500, "lr": 2e-05, "grad_norm": 2.3222463130950928, "loss_scale": 4.0, "retries": 0, "seconds": 877.7, "max_memory_gb": 16.52, "ce": 0.18155368029010788, "rank": 0.0} +{"step": 498, "steps": 500, "lr": 2e-05, "grad_norm": 0.6340310573577881, "loss_scale": 4.0, "retries": 0, "seconds": 879.3, "max_memory_gb": 16.52, "ce": 0.09523273884224182, "rank": 0.0} +{"step": 499, "steps": 500, "lr": 2e-05, "grad_norm": 2.206364393234253, "loss_scale": 4.0, "retries": 0, "seconds": 881.1, "max_memory_gb": 16.52, "ce": 0.15939180089571892, "rank": 0.0} +{"step": 500, "steps": 500, "lr": 2e-05, "grad_norm": 2.0934839248657227, "loss_scale": 4.0, "retries": 0, "seconds": 882.8, "max_memory_gb": 16.52, "ce": 0.22262372373870676, "rank": 0.0} diff --git a/runs/single_llava/removal/L18_s13/adapter/README.md b/runs/single_llava/removal/L18_s13/adapter/README.md new file mode 100644 index 0000000000000000000000000000000000000000..ae587920ca55972d94e38f783837f71be14f1dda --- /dev/null +++ b/runs/single_llava/removal/L18_s13/adapter/README.md @@ -0,0 +1,207 @@ +--- +base_model: /data/OPSD/assets/models/llava-1.5-7b-hf +library_name: peft +pipeline_tag: text-generation +tags: +- base_model:adapter:/data/OPSD/assets/models/llava-1.5-7b-hf +- lora +- transformers +--- + +# Model Card for Model ID + + + + + +## Model Details + +### Model Description + + + + + +- **Developed by:** [More Information Needed] +- **Funded by [optional]:** [More Information Needed] +- **Shared by [optional]:** [More Information Needed] +- **Model type:** [More Information Needed] +- **Language(s) (NLP):** [More Information Needed] +- **License:** [More Information Needed] +- **Finetuned from model [optional]:** [More Information Needed] + +### Model Sources [optional] + + + +- **Repository:** [More Information Needed] +- **Paper [optional]:** [More Information Needed] +- **Demo [optional]:** [More Information Needed] + +## Uses + + + +### Direct Use + + + +[More Information Needed] + +### Downstream Use [optional] + + + +[More Information Needed] + +### Out-of-Scope Use + + + +[More Information Needed] + +## Bias, Risks, and Limitations + + + +[More Information Needed] + +### Recommendations + + + +Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. More information needed for further recommendations. + +## How to Get Started with the Model + +Use the code below to get started with the model. + +[More Information Needed] + +## Training Details + +### Training Data + + + +[More Information Needed] + +### Training Procedure + + + +#### Preprocessing [optional] + +[More Information Needed] + + +#### Training Hyperparameters + +- **Training regime:** [More Information Needed] + +#### Speeds, Sizes, Times [optional] + + + +[More Information Needed] + +## Evaluation + + + +### Testing Data, Factors & Metrics + +#### Testing Data + + + +[More Information Needed] + +#### Factors + + + +[More Information Needed] + +#### Metrics + + + +[More Information Needed] + +### Results + +[More Information Needed] + +#### Summary + + + +## Model Examination [optional] + + + +[More Information Needed] + +## Environmental Impact + + + +Carbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700). + +- **Hardware Type:** [More Information Needed] +- **Hours used:** [More Information Needed] +- **Cloud Provider:** [More Information Needed] +- **Compute Region:** [More Information Needed] +- **Carbon Emitted:** [More Information Needed] + +## Technical Specifications [optional] + +### Model Architecture and Objective + +[More Information Needed] + +### Compute Infrastructure + +[More Information Needed] + +#### Hardware + +[More Information Needed] + +#### Software + +[More Information Needed] + +## Citation [optional] + + + +**BibTeX:** + +[More Information Needed] + +**APA:** + +[More Information Needed] + +## Glossary [optional] + + + +[More Information Needed] + +## More Information [optional] + +[More Information Needed] + +## Model Card Authors [optional] + +[More Information Needed] + +## Model Card Contact + +[More Information Needed] +### Framework versions + +- PEFT 0.20.0 \ No newline at end of file diff --git a/runs/single_llava/removal/L18_s13/adapter/adapter_config.json b/runs/single_llava/removal/L18_s13/adapter/adapter_config.json new file mode 100644 index 0000000000000000000000000000000000000000..e84072a82c6a001c53cdbc7da2bd346f9c2d1be4 --- /dev/null +++ b/runs/single_llava/removal/L18_s13/adapter/adapter_config.json @@ -0,0 +1,50 @@ +{ + "alora_invocation_tokens": null, + "alpha_pattern": {}, + "arrow_config": null, + "auto_mapping": null, + "base_model_name_or_path": "/data/OPSD/assets/models/llava-1.5-7b-hf", + "bias": "none", + "corda_config": null, + "ensure_weight_tying": false, + "eva_config": null, + "exclude_modules": null, + "fan_in_fan_out": false, + "inference_mode": true, + "init_lora_weights": true, + "layer_replication": null, + "layers_pattern": null, + "layers_to_transform": null, + "loftq_config": {}, + "lora_alpha": 32, + "lora_bias": false, + "lora_dropout": 0.05, + "lora_ga_config": null, + "megatron_config": null, + "megatron_core": "megatron.core", + "modules_to_save": null, + "monteclora_config": null, + "peft_type": "LORA", + "peft_version": "0.20.0", + "qalora_group_size": 16, + "r": 16, + "rank_pattern": {}, + "revision": null, + "target_modules": [ + "model.language_model.layers.18.self_attn.q_proj", + "model.language_model.layers.18.self_attn.k_proj", + "model.language_model.layers.18.mlp.gate_proj", + "model.language_model.layers.18.mlp.up_proj", + "model.language_model.layers.18.mlp.down_proj", + "model.language_model.layers.18.self_attn.o_proj", + "model.language_model.layers.18.self_attn.v_proj" + ], + "target_parameters": null, + "task_type": "CAUSAL_LM", + "trainable_token_indices": null, + "use_bdlora": null, + "use_dora": false, + "use_qalora": false, + "use_rslora": false, + "velora_config": null +} \ No newline at end of file diff --git a/runs/single_llava/removal/L18_s13/config.json b/runs/single_llava/removal/L18_s13/config.json new file mode 100644 index 0000000000000000000000000000000000000000..9e3ead6b9f10399a1444a79104e7628e26b81d01 --- /dev/null +++ b/runs/single_llava/removal/L18_s13/config.json @@ -0,0 +1,24 @@ +{ + "behavior": "removal", + "seed": 13, + "model": "/data/OPSD/assets/models/llava-1.5-7b-hf", + "backbone": "llava", + "train": "/data/OPSD/assets/v3/removal.jsonl", + "items": 1000, + "steps": 500, + "epochs": 2, + "steps_per_epoch": 250, + "groups_per_step": 4, + "lr": 2e-05, + "scheduler": "constant", + "warmup_steps": 0, + "final_fraction": 1.0, + "weight_decay": 0.0, + "clip": 1.0, + "initial_loss_scale": 4.0, + "adapter": { + "layer": 18, + "rank": 16, + "trainable_parameters": 1249280 + } +} \ No newline at end of file diff --git a/runs/single_llava/removal/L18_s13/steps.jsonl b/runs/single_llava/removal/L18_s13/steps.jsonl new file mode 100644 index 0000000000000000000000000000000000000000..cc47a9cd97953579c60d49012ad658fc928b8b3e --- /dev/null +++ b/runs/single_llava/removal/L18_s13/steps.jsonl @@ -0,0 +1,500 @@ +{"step": 1, "steps": 500, "lr": 2e-05, "grad_norm": 0.17514729499816895, "loss_scale": 4.0, "retries": 0, "seconds": 3.2, "max_memory_gb": 15.94, "ce": 1.6287612388841808, "rank": 0.13243506848812103} +{"step": 2, "steps": 500, "lr": 2e-05, "grad_norm": 0.2862674593925476, "loss_scale": 4.0, "retries": 0, "seconds": 4.9, "max_memory_gb": 15.95, "ce": 2.673992426134646, "rank": 0.03306448459625244} +{"step": 3, "steps": 500, "lr": 2e-05, "grad_norm": 0.61821448802948, "loss_scale": 4.0, "retries": 0, "seconds": 6.6, "max_memory_gb": 15.95, "ce": 5.27529538795352, "rank": 0.15083348751068115} +{"step": 4, "steps": 500, "lr": 2e-05, "grad_norm": 0.44192081689834595, "loss_scale": 4.0, "retries": 0, "seconds": 8.3, "max_memory_gb": 15.95, "ce": 3.799318699282594, "rank": 0.12275028228759766} +{"step": 5, "steps": 500, "lr": 2e-05, "grad_norm": 0.051140014082193375, "loss_scale": 4.0, "retries": 0, "seconds": 10.0, "max_memory_gb": 15.95, "ce": 0.1394540760666132, "rank": 0.0} +{"step": 6, "steps": 500, "lr": 2e-05, "grad_norm": 0.19458319246768951, "loss_scale": 4.0, "retries": 0, "seconds": 11.7, "max_memory_gb": 15.95, "ce": 1.512481514364481, "rank": 0.0} +{"step": 7, "steps": 500, "lr": 2e-05, "grad_norm": 0.4248661696910858, "loss_scale": 4.0, "retries": 0, "seconds": 13.4, "max_memory_gb": 15.95, "ce": 3.45737760537304, "rank": 0.02034306526184082} +{"step": 8, "steps": 500, "lr": 2e-05, "grad_norm": 0.5240023136138916, "loss_scale": 4.0, "retries": 0, "seconds": 15.1, "max_memory_gb": 15.95, "ce": 3.701978189870715, "rank": 0.20740461349487305} +{"step": 9, "steps": 500, "lr": 2e-05, "grad_norm": 0.4370194971561432, "loss_scale": 4.0, "retries": 0, "seconds": 16.9, "max_memory_gb": 15.95, "ce": 2.8300669249147177, "rank": 0.04116940498352051} +{"step": 10, "steps": 500, "lr": 2e-05, "grad_norm": 0.25039321184158325, "loss_scale": 4.0, "retries": 0, "seconds": 18.5, "max_memory_gb": 15.95, "ce": 1.6130936993286014, "rank": 0.0} +{"step": 11, "steps": 500, "lr": 2e-05, "grad_norm": 0.45795291662216187, "loss_scale": 4.0, "retries": 0, "seconds": 20.2, "max_memory_gb": 15.95, "ce": 2.6193284299224615, "rank": 0.21385683119297028} +{"step": 12, "steps": 500, "lr": 2e-05, "grad_norm": 0.3030148446559906, "loss_scale": 4.0, "retries": 0, "seconds": 21.8, "max_memory_gb": 15.95, "ce": 1.5056753908284009, "rank": 0.04841279983520508} +{"step": 13, "steps": 500, "lr": 2e-05, "grad_norm": 0.5001691579818726, "loss_scale": 4.0, "retries": 0, "seconds": 23.5, "max_memory_gb": 15.95, "ce": 2.662230923771858, "rank": 0.0} +{"step": 14, "steps": 500, "lr": 2e-05, "grad_norm": 0.7554935812950134, "loss_scale": 4.0, "retries": 0, "seconds": 25.2, "max_memory_gb": 15.95, "ce": 3.8547039283439517, "rank": 0.1450408697128296} +{"step": 15, "steps": 500, "lr": 2e-05, "grad_norm": 0.5259355902671814, "loss_scale": 4.0, "retries": 0, "seconds": 26.8, "max_memory_gb": 15.95, "ce": 2.587008240807336, "rank": 0.17162013053894043} +{"step": 16, "steps": 500, "lr": 2e-05, "grad_norm": 0.30868566036224365, "loss_scale": 4.0, "retries": 0, "seconds": 28.6, "max_memory_gb": 15.95, "ce": 1.4081286783330142, "rank": 0.0} +{"step": 17, "steps": 500, "lr": 2e-05, "grad_norm": 0.3402215242385864, "loss_scale": 4.0, "retries": 0, "seconds": 30.2, "max_memory_gb": 15.95, "ce": 1.5857821304816753, "rank": 0.036470651626586914} +{"step": 18, "steps": 500, "lr": 2e-05, "grad_norm": 0.8164456486701965, "loss_scale": 4.0, "retries": 0, "seconds": 31.9, "max_memory_gb": 15.96, "ce": 3.5563633386045694, "rank": 0.19175612926483154} +{"step": 19, "steps": 500, "lr": 2e-05, "grad_norm": 1.3315315246582031, "loss_scale": 4.0, "retries": 0, "seconds": 33.7, "max_memory_gb": 15.96, "ce": 5.218875168822706, "rank": 0.3162040710449219} +{"step": 20, "steps": 500, "lr": 2e-05, "grad_norm": 0.02406836301088333, "loss_scale": 4.0, "retries": 0, "seconds": 35.4, "max_memory_gb": 15.96, "ce": 0.06519841676345095, "rank": 0.10886256396770477} +{"step": 21, "steps": 500, "lr": 2e-05, "grad_norm": 0.8390812873840332, "loss_scale": 4.0, "retries": 0, "seconds": 37.1, "max_memory_gb": 15.96, "ce": 2.69988375029061, "rank": 0.15734457969665527} +{"step": 22, "steps": 500, "lr": 2e-05, "grad_norm": 0.8100133538246155, "loss_scale": 4.0, "retries": 0, "seconds": 38.8, "max_memory_gb": 15.96, "ce": 2.898754104040563, "rank": 0.18484675884246826} +{"step": 23, "steps": 500, "lr": 2e-05, "grad_norm": 0.9014520049095154, "loss_scale": 4.0, "retries": 0, "seconds": 40.5, "max_memory_gb": 15.96, "ce": 2.6950565107399598, "rank": 0.12240207195281982} +{"step": 24, "steps": 500, "lr": 2e-05, "grad_norm": 1.395141363143921, "loss_scale": 4.0, "retries": 0, "seconds": 42.1, "max_memory_gb": 15.96, "ce": 4.9137179693207145, "rank": 0.25518977642059326} +{"step": 25, "steps": 500, "lr": 2e-05, "grad_norm": 1.3492658138275146, "loss_scale": 4.0, "retries": 0, "seconds": 43.8, "max_memory_gb": 15.96, "ce": 4.03355614002794, "rank": 0.0452350378036499} +{"step": 26, "steps": 500, "lr": 2e-05, "grad_norm": 1.7441171407699585, "loss_scale": 4.0, "retries": 0, "seconds": 45.6, "max_memory_gb": 15.96, "ce": 5.040237583220005, "rank": 0.06365966796875} +{"step": 27, "steps": 500, "lr": 2e-05, "grad_norm": 1.4939887523651123, "loss_scale": 4.0, "retries": 0, "seconds": 47.2, "max_memory_gb": 15.96, "ce": 4.0903165359050035, "rank": 0.4844210147857666} +{"step": 28, "steps": 500, "lr": 2e-05, "grad_norm": 1.0307607650756836, "loss_scale": 4.0, "retries": 0, "seconds": 48.9, "max_memory_gb": 15.96, "ce": 2.678448811173439, "rank": 0.0} +{"step": 29, "steps": 500, "lr": 2e-05, "grad_norm": 0.9670752882957458, "loss_scale": 4.0, "retries": 0, "seconds": 50.6, "max_memory_gb": 15.96, "ce": 2.5954738883301616, "rank": 0.0} +{"step": 30, "steps": 500, "lr": 2e-05, "grad_norm": 1.1445343494415283, "loss_scale": 4.0, "retries": 0, "seconds": 52.3, "max_memory_gb": 15.96, "ce": 2.653787260176614, "rank": 0.0} +{"step": 31, "steps": 500, "lr": 2e-05, "grad_norm": 1.2453333139419556, "loss_scale": 4.0, "retries": 0, "seconds": 54.0, "max_memory_gb": 15.96, "ce": 2.7085265861824155, "rank": 0.1496502161026001} +{"step": 32, "steps": 500, "lr": 2e-05, "grad_norm": 2.425424575805664, "loss_scale": 4.0, "retries": 0, "seconds": 55.6, "max_memory_gb": 15.96, "ce": 5.054229092784226, "rank": 0.23024976253509521} +{"step": 33, "steps": 500, "lr": 2e-05, "grad_norm": 1.05840265750885, "loss_scale": 4.0, "retries": 0, "seconds": 57.3, "max_memory_gb": 15.96, "ce": 2.418083977419883, "rank": 0.010293826460838318} +{"step": 34, "steps": 500, "lr": 2e-05, "grad_norm": 1.245829463005066, "loss_scale": 4.0, "retries": 0, "seconds": 59.0, "max_memory_gb": 15.96, "ce": 2.606510896002874, "rank": 0.0} +{"step": 35, "steps": 500, "lr": 2e-05, "grad_norm": 1.5450655221939087, "loss_scale": 4.0, "retries": 0, "seconds": 60.7, "max_memory_gb": 15.96, "ce": 2.5559125472791493, "rank": 0.11645936965942383} +{"step": 36, "steps": 500, "lr": 2e-05, "grad_norm": 1.2604268789291382, "loss_scale": 4.0, "retries": 0, "seconds": 62.4, "max_memory_gb": 15.96, "ce": 2.457556873559952, "rank": 0.0} +{"step": 37, "steps": 500, "lr": 2e-05, "grad_norm": 1.4643886089324951, "loss_scale": 4.0, "retries": 0, "seconds": 64.1, "max_memory_gb": 15.96, "ce": 2.4608832951635122, "rank": 0.0} +{"step": 38, "steps": 500, "lr": 2e-05, "grad_norm": 2.6918797492980957, "loss_scale": 4.0, "retries": 0, "seconds": 65.9, "max_memory_gb": 15.96, "ce": 4.525316638872027, "rank": 0.15605652332305908} +{"step": 39, "steps": 500, "lr": 2e-05, "grad_norm": 1.0538806915283203, "loss_scale": 4.0, "retries": 0, "seconds": 67.7, "max_memory_gb": 15.96, "ce": 1.344644763856195, "rank": 0.16749531030654907} +{"step": 40, "steps": 500, "lr": 2e-05, "grad_norm": 1.6574944257736206, "loss_scale": 4.0, "retries": 0, "seconds": 69.6, "max_memory_gb": 15.96, "ce": 2.4975889488123357, "rank": 0.0009582042694091797} +{"step": 41, "steps": 500, "lr": 2e-05, "grad_norm": 0.8115625977516174, "loss_scale": 4.0, "retries": 0, "seconds": 71.4, "max_memory_gb": 15.96, "ce": 1.437490620650351, "rank": 0.0} +{"step": 42, "steps": 500, "lr": 2e-05, "grad_norm": 1.8481190204620361, "loss_scale": 4.0, "retries": 0, "seconds": 73.0, "max_memory_gb": 15.96, "ce": 2.1941611487418413, "rank": 0.0} +{"step": 43, "steps": 500, "lr": 2e-05, "grad_norm": 1.8388537168502808, "loss_scale": 4.0, "retries": 0, "seconds": 74.6, "max_memory_gb": 15.96, "ce": 2.2747601743321866, "rank": 0.0} +{"step": 44, "steps": 500, "lr": 2e-05, "grad_norm": 0.9776744246482849, "loss_scale": 4.0, "retries": 0, "seconds": 76.1, "max_memory_gb": 15.96, "ce": 1.3193815458216704, "rank": 0.0} +{"step": 45, "steps": 500, "lr": 2e-05, "grad_norm": 0.9982602000236511, "loss_scale": 4.0, "retries": 0, "seconds": 77.7, "max_memory_gb": 15.96, "ce": 1.3176968391053379, "rank": 0.0} +{"step": 46, "steps": 500, "lr": 2e-05, "grad_norm": 2.0074524879455566, "loss_scale": 4.0, "retries": 0, "seconds": 79.4, "max_memory_gb": 15.96, "ce": 2.1319043338298798, "rank": 0.0} +{"step": 47, "steps": 500, "lr": 2e-05, "grad_norm": 1.7868187427520752, "loss_scale": 4.0, "retries": 0, "seconds": 81.0, "max_memory_gb": 15.96, "ce": 1.8821520400233567, "rank": 0.0} +{"step": 48, "steps": 500, "lr": 2e-05, "grad_norm": 4.086081027984619, "loss_scale": 4.0, "retries": 0, "seconds": 82.8, "max_memory_gb": 15.96, "ce": 3.291104869917035, "rank": 0.04070019721984863} +{"step": 49, "steps": 500, "lr": 2e-05, "grad_norm": 3.3105485439300537, "loss_scale": 4.0, "retries": 0, "seconds": 84.4, "max_memory_gb": 15.96, "ce": 3.1973318974487484, "rank": 0.0} +{"step": 50, "steps": 500, "lr": 2e-05, "grad_norm": 3.745044708251953, "loss_scale": 4.0, "retries": 0, "seconds": 86.0, "max_memory_gb": 15.96, "ce": 3.212412202381529, "rank": 0.0012898445129394531} +{"step": 51, "steps": 500, "lr": 2e-05, "grad_norm": 1.26856529712677, "loss_scale": 4.0, "retries": 0, "seconds": 87.6, "max_memory_gb": 15.96, "ce": 1.2831116906600073, "rank": 0.06527972221374512} +{"step": 52, "steps": 500, "lr": 2e-05, "grad_norm": 2.543783664703369, "loss_scale": 4.0, "retries": 0, "seconds": 89.2, "max_memory_gb": 15.96, "ce": 2.0988661805167794, "rank": 0.0} +{"step": 53, "steps": 500, "lr": 2e-05, "grad_norm": 3.765916347503662, "loss_scale": 4.0, "retries": 0, "seconds": 90.7, "max_memory_gb": 15.96, "ce": 3.2710904218256474, "rank": 0.0} +{"step": 54, "steps": 500, "lr": 2e-05, "grad_norm": 4.490849494934082, "loss_scale": 4.0, "retries": 0, "seconds": 92.3, "max_memory_gb": 15.96, "ce": 3.2461387757211924, "rank": 0.0} +{"step": 55, "steps": 500, "lr": 2e-05, "grad_norm": 2.954408884048462, "loss_scale": 4.0, "retries": 0, "seconds": 93.8, "max_memory_gb": 15.96, "ce": 2.0784441474825144, "rank": 0.0} +{"step": 56, "steps": 500, "lr": 2e-05, "grad_norm": 6.193009376525879, "loss_scale": 4.0, "retries": 0, "seconds": 95.5, "max_memory_gb": 15.96, "ce": 4.034362881444395, "rank": 0.0} +{"step": 57, "steps": 500, "lr": 2e-05, "grad_norm": 1.4850791692733765, "loss_scale": 4.0, "retries": 0, "seconds": 97.2, "max_memory_gb": 15.96, "ce": 1.0995133323594928, "rank": 0.004901111125946045} +{"step": 58, "steps": 500, "lr": 2e-05, "grad_norm": 3.701828956604004, "loss_scale": 4.0, "retries": 0, "seconds": 98.8, "max_memory_gb": 15.96, "ce": 1.9021552737103775, "rank": 0.0} +{"step": 59, "steps": 500, "lr": 2e-05, "grad_norm": 1.8252569437026978, "loss_scale": 4.0, "retries": 0, "seconds": 100.5, "max_memory_gb": 15.96, "ce": 1.2540565566159785, "rank": 0.0} +{"step": 60, "steps": 500, "lr": 2e-05, "grad_norm": 3.4458484649658203, "loss_scale": 4.0, "retries": 0, "seconds": 102.2, "max_memory_gb": 15.96, "ce": 1.9851570238824934, "rank": 0.0} +{"step": 61, "steps": 500, "lr": 2e-05, "grad_norm": 6.161250591278076, "loss_scale": 4.0, "retries": 0, "seconds": 104.0, "max_memory_gb": 15.96, "ce": 2.624469196423888, "rank": 0.0} +{"step": 62, "steps": 500, "lr": 2e-05, "grad_norm": 2.0736021995544434, "loss_scale": 4.0, "retries": 0, "seconds": 105.6, "max_memory_gb": 15.96, "ce": 1.0973512940108776, "rank": 0.0} +{"step": 63, "steps": 500, "lr": 2e-05, "grad_norm": 5.620537757873535, "loss_scale": 4.0, "retries": 0, "seconds": 107.5, "max_memory_gb": 15.96, "ce": 2.98740273877047, "rank": 0.210141122341156} +{"step": 64, "steps": 500, "lr": 2e-05, "grad_norm": 7.641270637512207, "loss_scale": 4.0, "retries": 0, "seconds": 109.4, "max_memory_gb": 15.96, "ce": 2.5342724344227463, "rank": 0.0} +{"step": 65, "steps": 500, "lr": 2e-05, "grad_norm": 6.808333396911621, "loss_scale": 4.0, "retries": 0, "seconds": 111.2, "max_memory_gb": 15.96, "ce": 2.737886601127684, "rank": 0.0} +{"step": 66, "steps": 500, "lr": 2e-05, "grad_norm": 1.4963219165802002, "loss_scale": 4.0, "retries": 0, "seconds": 113.1, "max_memory_gb": 15.96, "ce": 0.8102558691753075, "rank": 0.18886329233646393} +{"step": 67, "steps": 500, "lr": 2e-05, "grad_norm": 7.289936065673828, "loss_scale": 4.0, "retries": 0, "seconds": 114.9, "max_memory_gb": 15.96, "ce": 2.1842821650207043, "rank": 0.0} +{"step": 68, "steps": 500, "lr": 2e-05, "grad_norm": 3.0444180965423584, "loss_scale": 4.0, "retries": 0, "seconds": 116.7, "max_memory_gb": 15.96, "ce": 0.8122237918432802, "rank": 0.0} +{"step": 69, "steps": 500, "lr": 2e-05, "grad_norm": 5.437986373901367, "loss_scale": 4.0, "retries": 0, "seconds": 118.3, "max_memory_gb": 15.96, "ce": 1.4277255807537585, "rank": 0.0} +{"step": 70, "steps": 500, "lr": 2e-05, "grad_norm": 5.4329023361206055, "loss_scale": 4.0, "retries": 0, "seconds": 120.0, "max_memory_gb": 15.96, "ce": 1.3970146491192281, "rank": 0.0} +{"step": 71, "steps": 500, "lr": 2e-05, "grad_norm": 8.647075653076172, "loss_scale": 4.0, "retries": 0, "seconds": 121.6, "max_memory_gb": 15.96, "ce": 1.9952602150151506, "rank": 0.0} +{"step": 72, "steps": 500, "lr": 2e-05, "grad_norm": 10.871346473693848, "loss_scale": 4.0, "retries": 0, "seconds": 123.3, "max_memory_gb": 15.96, "ce": 2.2892056396231055, "rank": 0.06971943378448486} +{"step": 73, "steps": 500, "lr": 2e-05, "grad_norm": 0.18447698652744293, "loss_scale": 4.0, "retries": 0, "seconds": 125.0, "max_memory_gb": 15.96, "ce": 0.24672836100216955, "rank": 0.0} +{"step": 74, "steps": 500, "lr": 2e-05, "grad_norm": 6.215658187866211, "loss_scale": 4.0, "retries": 0, "seconds": 126.6, "max_memory_gb": 15.96, "ce": 1.1409474269021302, "rank": 0.0} +{"step": 75, "steps": 500, "lr": 2e-05, "grad_norm": 3.2710611820220947, "loss_scale": 4.0, "retries": 0, "seconds": 128.3, "max_memory_gb": 15.96, "ce": 1.0756148207001388, "rank": 0.15070918202400208} +{"step": 76, "steps": 500, "lr": 2e-05, "grad_norm": 6.0271172523498535, "loss_scale": 4.0, "retries": 0, "seconds": 130.0, "max_memory_gb": 15.96, "ce": 1.0974963805638254, "rank": 0.0} +{"step": 77, "steps": 500, "lr": 2e-05, "grad_norm": 8.953042030334473, "loss_scale": 4.0, "retries": 0, "seconds": 131.7, "max_memory_gb": 15.96, "ce": 1.5773826567456126, "rank": 0.0} +{"step": 78, "steps": 500, "lr": 2e-05, "grad_norm": 4.715014457702637, "loss_scale": 4.0, "retries": 0, "seconds": 133.3, "max_memory_gb": 15.96, "ce": 0.9185660500079393, "rank": 0.0} +{"step": 79, "steps": 500, "lr": 2e-05, "grad_norm": 2.755679130554199, "loss_scale": 4.0, "retries": 0, "seconds": 135.0, "max_memory_gb": 15.96, "ce": 0.4533432580064982, "rank": 0.0} +{"step": 80, "steps": 500, "lr": 2e-05, "grad_norm": 5.29343843460083, "loss_scale": 4.0, "retries": 0, "seconds": 136.6, "max_memory_gb": 15.96, "ce": 0.7126496067503467, "rank": 0.0} +{"step": 81, "steps": 500, "lr": 2e-05, "grad_norm": 7.922154426574707, "loss_scale": 4.0, "retries": 0, "seconds": 138.3, "max_memory_gb": 15.96, "ce": 1.0496599562466145, "rank": 0.0} +{"step": 82, "steps": 500, "lr": 2e-05, "grad_norm": 7.514275074005127, "loss_scale": 4.0, "retries": 0, "seconds": 140.0, "max_memory_gb": 15.96, "ce": 0.9001021762378514, "rank": 0.0} +{"step": 83, "steps": 500, "lr": 2e-05, "grad_norm": 6.354631423950195, "loss_scale": 4.0, "retries": 0, "seconds": 141.7, "max_memory_gb": 15.96, "ce": 0.9485203768126667, "rank": 0.0} +{"step": 84, "steps": 500, "lr": 2e-05, "grad_norm": 3.0088860988616943, "loss_scale": 4.0, "retries": 0, "seconds": 143.4, "max_memory_gb": 15.96, "ce": 0.45516041055088863, "rank": 0.0} +{"step": 85, "steps": 500, "lr": 2e-05, "grad_norm": 0.16879311203956604, "loss_scale": 4.0, "retries": 0, "seconds": 145.1, "max_memory_gb": 15.96, "ce": 0.12762363901128992, "rank": 0.0} +{"step": 86, "steps": 500, "lr": 2e-05, "grad_norm": 6.696173667907715, "loss_scale": 4.0, "retries": 0, "seconds": 146.8, "max_memory_gb": 15.96, "ce": 0.8822194971144199, "rank": 0.0} +{"step": 87, "steps": 500, "lr": 2e-05, "grad_norm": 4.720201015472412, "loss_scale": 4.0, "retries": 0, "seconds": 148.5, "max_memory_gb": 15.96, "ce": 0.659011437324807, "rank": 0.0} +{"step": 88, "steps": 500, "lr": 2e-05, "grad_norm": 1.9594836235046387, "loss_scale": 4.0, "retries": 0, "seconds": 150.2, "max_memory_gb": 15.96, "ce": 0.5056316034169868, "rank": 0.0} +{"step": 89, "steps": 500, "lr": 2e-05, "grad_norm": 0.09405481815338135, "loss_scale": 4.0, "retries": 0, "seconds": 151.9, "max_memory_gb": 15.96, "ce": 0.15710816701175645, "rank": 0.12443890422582626} +{"step": 90, "steps": 500, "lr": 2e-05, "grad_norm": 1.7489960193634033, "loss_scale": 4.0, "retries": 0, "seconds": 153.5, "max_memory_gb": 15.96, "ce": 0.44374167639762163, "rank": 0.0} +{"step": 91, "steps": 500, "lr": 2e-05, "grad_norm": 0.20122185349464417, "loss_scale": 4.0, "retries": 0, "seconds": 155.3, "max_memory_gb": 15.96, "ce": 0.3604784137569368, "rank": 0.0} +{"step": 92, "steps": 500, "lr": 2e-05, "grad_norm": 0.7387259006500244, "loss_scale": 4.0, "retries": 0, "seconds": 157.0, "max_memory_gb": 15.96, "ce": 0.2984305269783363, "rank": 0.0} +{"step": 93, "steps": 500, "lr": 2e-05, "grad_norm": 1.621791958808899, "loss_scale": 4.0, "retries": 0, "seconds": 158.6, "max_memory_gb": 15.96, "ce": 0.37485747097525746, "rank": 0.0} +{"step": 94, "steps": 500, "lr": 2e-05, "grad_norm": 1.6293352842330933, "loss_scale": 4.0, "retries": 0, "seconds": 160.3, "max_memory_gb": 15.96, "ce": 0.5342581868171692, "rank": 0.0} +{"step": 95, "steps": 500, "lr": 2e-05, "grad_norm": 0.917885422706604, "loss_scale": 4.0, "retries": 0, "seconds": 162.0, "max_memory_gb": 15.96, "ce": 0.4654629866126925, "rank": 0.0} +{"step": 96, "steps": 500, "lr": 2e-05, "grad_norm": 1.045771598815918, "loss_scale": 4.0, "retries": 0, "seconds": 163.7, "max_memory_gb": 15.96, "ce": 0.5266885310411453, "rank": 0.016808509826660156} +{"step": 97, "steps": 500, "lr": 2e-05, "grad_norm": 0.45543405413627625, "loss_scale": 4.0, "retries": 0, "seconds": 165.3, "max_memory_gb": 15.96, "ce": 0.3213323005475104, "rank": 0.0} +{"step": 98, "steps": 500, "lr": 2e-05, "grad_norm": 0.7582030296325684, "loss_scale": 4.0, "retries": 0, "seconds": 167.0, "max_memory_gb": 15.96, "ce": 0.3115436303196475, "rank": 0.0} +{"step": 99, "steps": 500, "lr": 2e-05, "grad_norm": 0.394423246383667, "loss_scale": 4.0, "retries": 0, "seconds": 168.7, "max_memory_gb": 15.96, "ce": 0.2378887350205332, "rank": 0.028127044439315796} +{"step": 100, "steps": 500, "lr": 2e-05, "grad_norm": 0.897810161113739, "loss_scale": 4.0, "retries": 0, "seconds": 170.4, "max_memory_gb": 15.96, "ce": 0.5078841270878911, "rank": 0.0} +{"step": 101, "steps": 500, "lr": 2e-05, "grad_norm": 0.7069432139396667, "loss_scale": 4.0, "retries": 0, "seconds": 172.1, "max_memory_gb": 15.96, "ce": 0.4663640297949314, "rank": 0.0} +{"step": 102, "steps": 500, "lr": 2e-05, "grad_norm": 1.0475635528564453, "loss_scale": 4.0, "retries": 0, "seconds": 173.8, "max_memory_gb": 15.96, "ce": 0.6058367257937789, "rank": 0.0} +{"step": 103, "steps": 500, "lr": 2e-05, "grad_norm": 0.4019366502761841, "loss_scale": 4.0, "retries": 0, "seconds": 175.5, "max_memory_gb": 15.96, "ce": 0.2988697903929278, "rank": 0.10911653935909271} +{"step": 104, "steps": 500, "lr": 2e-05, "grad_norm": 0.21588970720767975, "loss_scale": 4.0, "retries": 0, "seconds": 177.1, "max_memory_gb": 15.96, "ce": 0.21440112515119836, "rank": 0.0670439675450325} +{"step": 105, "steps": 500, "lr": 2e-05, "grad_norm": 0.5971521139144897, "loss_scale": 4.0, "retries": 0, "seconds": 178.9, "max_memory_gb": 15.96, "ce": 0.44450759957544506, "rank": 0.11578963696956635} +{"step": 106, "steps": 500, "lr": 2e-05, "grad_norm": 0.47862282395362854, "loss_scale": 4.0, "retries": 0, "seconds": 180.5, "max_memory_gb": 15.96, "ce": 0.4417112087830901, "rank": 0.08900170773267746} +{"step": 107, "steps": 500, "lr": 2e-05, "grad_norm": 0.24884136021137238, "loss_scale": 4.0, "retries": 0, "seconds": 182.2, "max_memory_gb": 15.96, "ce": 0.27062797895632684, "rank": 0.0} +{"step": 108, "steps": 500, "lr": 2e-05, "grad_norm": 0.304660439491272, "loss_scale": 4.0, "retries": 0, "seconds": 183.9, "max_memory_gb": 15.96, "ce": 0.2774441367946565, "rank": 0.0} +{"step": 109, "steps": 500, "lr": 2e-05, "grad_norm": 0.6278958916664124, "loss_scale": 4.0, "retries": 0, "seconds": 185.6, "max_memory_gb": 15.96, "ce": 0.5712803900241852, "rank": 0.0} +{"step": 110, "steps": 500, "lr": 2e-05, "grad_norm": 0.5240972638130188, "loss_scale": 4.0, "retries": 0, "seconds": 187.3, "max_memory_gb": 15.96, "ce": 0.39489789586514235, "rank": 0.05058474838733673} +{"step": 111, "steps": 500, "lr": 2e-05, "grad_norm": 0.3890696167945862, "loss_scale": 4.0, "retries": 0, "seconds": 189.0, "max_memory_gb": 15.96, "ce": 0.30864588799886405, "rank": 0.0} +{"step": 112, "steps": 500, "lr": 2e-05, "grad_norm": 0.4515511691570282, "loss_scale": 4.0, "retries": 0, "seconds": 190.6, "max_memory_gb": 15.96, "ce": 0.3030006014741957, "rank": 0.0} +{"step": 113, "steps": 500, "lr": 2e-05, "grad_norm": 0.5032243132591248, "loss_scale": 4.0, "retries": 0, "seconds": 192.3, "max_memory_gb": 15.96, "ce": 0.35540445195510983, "rank": 0.024021394550800323} +{"step": 114, "steps": 500, "lr": 2e-05, "grad_norm": 0.7087250351905823, "loss_scale": 4.0, "retries": 0, "seconds": 194.0, "max_memory_gb": 15.96, "ce": 0.631286523770541, "rank": 0.145700603723526} +{"step": 115, "steps": 500, "lr": 2e-05, "grad_norm": 0.47383734583854675, "loss_scale": 4.0, "retries": 0, "seconds": 195.7, "max_memory_gb": 15.96, "ce": 0.41370570636354387, "rank": 0.1345807909965515} +{"step": 116, "steps": 500, "lr": 2e-05, "grad_norm": 0.10887791961431503, "loss_scale": 4.0, "retries": 0, "seconds": 197.4, "max_memory_gb": 15.96, "ce": 0.149498337530531, "rank": 0.0} +{"step": 117, "steps": 500, "lr": 2e-05, "grad_norm": 0.5704694390296936, "loss_scale": 4.0, "retries": 0, "seconds": 199.1, "max_memory_gb": 15.96, "ce": 0.4314029214438051, "rank": 0.0} +{"step": 118, "steps": 500, "lr": 2e-05, "grad_norm": 0.2502656877040863, "loss_scale": 4.0, "retries": 0, "seconds": 200.8, "max_memory_gb": 15.96, "ce": 0.2785144897643477, "rank": 0.0} +{"step": 119, "steps": 500, "lr": 2e-05, "grad_norm": 0.31343770027160645, "loss_scale": 4.0, "retries": 0, "seconds": 202.4, "max_memory_gb": 15.96, "ce": 0.254558474291116, "rank": 0.0} +{"step": 120, "steps": 500, "lr": 2e-05, "grad_norm": 0.5622391104698181, "loss_scale": 4.0, "retries": 0, "seconds": 204.2, "max_memory_gb": 15.96, "ce": 0.3732811757363379, "rank": 0.0} +{"step": 121, "steps": 500, "lr": 2e-05, "grad_norm": 0.4287782907485962, "loss_scale": 4.0, "retries": 0, "seconds": 205.9, "max_memory_gb": 15.96, "ce": 0.3242546209366992, "rank": 0.0} +{"step": 122, "steps": 500, "lr": 2e-05, "grad_norm": 0.3982437551021576, "loss_scale": 4.0, "retries": 0, "seconds": 207.6, "max_memory_gb": 15.96, "ce": 0.33123051025904715, "rank": 0.0} +{"step": 123, "steps": 500, "lr": 2e-05, "grad_norm": 0.884381115436554, "loss_scale": 4.0, "retries": 0, "seconds": 209.2, "max_memory_gb": 15.96, "ce": 0.5169360332656652, "rank": 0.00393618643283844} +{"step": 124, "steps": 500, "lr": 2e-05, "grad_norm": 0.3784199357032776, "loss_scale": 4.0, "retries": 0, "seconds": 210.9, "max_memory_gb": 15.96, "ce": 0.3452472995268181, "rank": 0.060288965702056885} +{"step": 125, "steps": 500, "lr": 2e-05, "grad_norm": 0.2750062346458435, "loss_scale": 4.0, "retries": 0, "seconds": 212.5, "max_memory_gb": 15.96, "ce": 0.2666918314062059, "rank": 0.0} +{"step": 126, "steps": 500, "lr": 2e-05, "grad_norm": 0.5461562275886536, "loss_scale": 4.0, "retries": 0, "seconds": 214.2, "max_memory_gb": 15.96, "ce": 0.48760802997276187, "rank": 0.0} +{"step": 127, "steps": 500, "lr": 2e-05, "grad_norm": 0.2684479355812073, "loss_scale": 4.0, "retries": 0, "seconds": 215.9, "max_memory_gb": 15.96, "ce": 0.14463488693581894, "rank": 0.0} +{"step": 128, "steps": 500, "lr": 2e-05, "grad_norm": 0.5956563353538513, "loss_scale": 4.0, "retries": 0, "seconds": 217.6, "max_memory_gb": 15.96, "ce": 0.33419483981560916, "rank": 0.0} +{"step": 129, "steps": 500, "lr": 2e-05, "grad_norm": 0.8788784742355347, "loss_scale": 4.0, "retries": 0, "seconds": 219.2, "max_memory_gb": 15.96, "ce": 0.39887560717761517, "rank": 0.0} +{"step": 130, "steps": 500, "lr": 2e-05, "grad_norm": 0.7207000851631165, "loss_scale": 4.0, "retries": 0, "seconds": 220.9, "max_memory_gb": 15.96, "ce": 0.49155384488403797, "rank": 0.0} +{"step": 131, "steps": 500, "lr": 2e-05, "grad_norm": 0.5470963716506958, "loss_scale": 4.0, "retries": 0, "seconds": 222.6, "max_memory_gb": 15.96, "ce": 0.3109722346998751, "rank": 0.0} +{"step": 132, "steps": 500, "lr": 2e-05, "grad_norm": 1.2905384302139282, "loss_scale": 4.0, "retries": 0, "seconds": 224.3, "max_memory_gb": 15.96, "ce": 0.5005324988160282, "rank": 0.0977080762386322} +{"step": 133, "steps": 500, "lr": 2e-05, "grad_norm": 0.8318023085594177, "loss_scale": 4.0, "retries": 0, "seconds": 225.9, "max_memory_gb": 15.96, "ce": 0.3725145999342203, "rank": 0.057773083448410034} +{"step": 134, "steps": 500, "lr": 2e-05, "grad_norm": 0.29945018887519836, "loss_scale": 4.0, "retries": 0, "seconds": 227.6, "max_memory_gb": 15.96, "ce": 0.19824083452112973, "rank": 0.0} +{"step": 135, "steps": 500, "lr": 2e-05, "grad_norm": 0.6985977292060852, "loss_scale": 4.0, "retries": 0, "seconds": 229.3, "max_memory_gb": 15.96, "ce": 0.31619034381583333, "rank": 0.0} +{"step": 136, "steps": 500, "lr": 2e-05, "grad_norm": 0.5491570234298706, "loss_scale": 4.0, "retries": 0, "seconds": 230.8, "max_memory_gb": 15.96, "ce": 0.3298598430119455, "rank": 0.0} +{"step": 137, "steps": 500, "lr": 2e-05, "grad_norm": 0.3999730348587036, "loss_scale": 4.0, "retries": 0, "seconds": 232.5, "max_memory_gb": 15.96, "ce": 0.2552696840139106, "rank": 0.0} +{"step": 138, "steps": 500, "lr": 2e-05, "grad_norm": 0.5459385514259338, "loss_scale": 4.0, "retries": 0, "seconds": 234.2, "max_memory_gb": 15.96, "ce": 0.2624644157476723, "rank": 0.02167447656393051} +{"step": 139, "steps": 500, "lr": 2e-05, "grad_norm": 0.9750049710273743, "loss_scale": 4.0, "retries": 0, "seconds": 235.9, "max_memory_gb": 15.96, "ce": 0.37473359052091837, "rank": 0.0} +{"step": 140, "steps": 500, "lr": 2e-05, "grad_norm": 0.43113529682159424, "loss_scale": 4.0, "retries": 0, "seconds": 237.6, "max_memory_gb": 15.96, "ce": 0.18587266886606812, "rank": 0.0} +{"step": 141, "steps": 500, "lr": 2e-05, "grad_norm": 0.49804654717445374, "loss_scale": 4.0, "retries": 0, "seconds": 239.3, "max_memory_gb": 15.96, "ce": 0.27798636886291206, "rank": 0.0} +{"step": 142, "steps": 500, "lr": 2e-05, "grad_norm": 0.4249209463596344, "loss_scale": 4.0, "retries": 0, "seconds": 241.0, "max_memory_gb": 15.96, "ce": 0.25916941149625927, "rank": 0.0} +{"step": 143, "steps": 500, "lr": 2e-05, "grad_norm": 0.17640049755573273, "loss_scale": 4.0, "retries": 0, "seconds": 242.7, "max_memory_gb": 15.96, "ce": 0.17664318578317761, "rank": 0.0} +{"step": 144, "steps": 500, "lr": 2e-05, "grad_norm": 0.6112992167472839, "loss_scale": 4.0, "retries": 0, "seconds": 244.4, "max_memory_gb": 15.96, "ce": 0.28988287574611604, "rank": 0.0} +{"step": 145, "steps": 500, "lr": 2e-05, "grad_norm": 0.32401034235954285, "loss_scale": 4.0, "retries": 0, "seconds": 246.1, "max_memory_gb": 15.96, "ce": 0.13183436257531866, "rank": 0.0} +{"step": 146, "steps": 500, "lr": 2e-05, "grad_norm": 0.4089866876602173, "loss_scale": 4.0, "retries": 0, "seconds": 247.8, "max_memory_gb": 15.96, "ce": 0.2333207712508738, "rank": 0.0} +{"step": 147, "steps": 500, "lr": 2e-05, "grad_norm": 0.6182811260223389, "loss_scale": 4.0, "retries": 0, "seconds": 249.5, "max_memory_gb": 15.96, "ce": 0.4128992594778538, "rank": 0.0} +{"step": 148, "steps": 500, "lr": 2e-05, "grad_norm": 0.19490040838718414, "loss_scale": 4.0, "retries": 0, "seconds": 251.2, "max_memory_gb": 15.96, "ce": 0.0975335537805222, "rank": 0.0} +{"step": 149, "steps": 500, "lr": 2e-05, "grad_norm": 0.3798063397407532, "loss_scale": 4.0, "retries": 0, "seconds": 252.9, "max_memory_gb": 15.96, "ce": 0.1966710949782282, "rank": 0.1377527415752411} +{"step": 150, "steps": 500, "lr": 2e-05, "grad_norm": 0.6017696261405945, "loss_scale": 4.0, "retries": 0, "seconds": 254.6, "max_memory_gb": 15.96, "ce": 0.3104542079381645, "rank": 0.03273387253284454} +{"step": 151, "steps": 500, "lr": 2e-05, "grad_norm": 0.4433678388595581, "loss_scale": 4.0, "retries": 0, "seconds": 256.3, "max_memory_gb": 15.96, "ce": 0.1827006242237985, "rank": 0.0} +{"step": 152, "steps": 500, "lr": 2e-05, "grad_norm": 0.6473239660263062, "loss_scale": 4.0, "retries": 0, "seconds": 257.9, "max_memory_gb": 15.96, "ce": 0.26635082717984915, "rank": 0.0} +{"step": 153, "steps": 500, "lr": 2e-05, "grad_norm": 0.5952196717262268, "loss_scale": 4.0, "retries": 0, "seconds": 259.7, "max_memory_gb": 15.96, "ce": 0.20652625465299934, "rank": 0.0} +{"step": 154, "steps": 500, "lr": 2e-05, "grad_norm": 1.3057265281677246, "loss_scale": 4.0, "retries": 0, "seconds": 261.6, "max_memory_gb": 15.96, "ce": 0.4744767560623586, "rank": 0.08605802059173584} +{"step": 155, "steps": 500, "lr": 2e-05, "grad_norm": 1.9890602827072144, "loss_scale": 4.0, "retries": 0, "seconds": 263.4, "max_memory_gb": 15.96, "ce": 0.5252420625183731, "rank": 0.0} +{"step": 156, "steps": 500, "lr": 2e-05, "grad_norm": 0.43728581070899963, "loss_scale": 4.0, "retries": 0, "seconds": 265.2, "max_memory_gb": 15.96, "ce": 0.21184415381867439, "rank": 0.0} +{"step": 157, "steps": 500, "lr": 2e-05, "grad_norm": 0.31366485357284546, "loss_scale": 4.0, "retries": 0, "seconds": 266.8, "max_memory_gb": 15.96, "ce": 0.21666477026883513, "rank": 0.0} +{"step": 158, "steps": 500, "lr": 2e-05, "grad_norm": 0.6879229545593262, "loss_scale": 4.0, "retries": 0, "seconds": 268.5, "max_memory_gb": 15.96, "ce": 0.3783222553320229, "rank": 0.0} +{"step": 159, "steps": 500, "lr": 2e-05, "grad_norm": 0.3722938597202301, "loss_scale": 4.0, "retries": 0, "seconds": 270.2, "max_memory_gb": 15.97, "ce": 0.1718060034327209, "rank": 0.0} +{"step": 160, "steps": 500, "lr": 2e-05, "grad_norm": 0.47025373578071594, "loss_scale": 4.0, "retries": 0, "seconds": 271.9, "max_memory_gb": 15.97, "ce": 0.22168996930122375, "rank": 0.0} +{"step": 161, "steps": 500, "lr": 2e-05, "grad_norm": 0.4792565107345581, "loss_scale": 4.0, "retries": 0, "seconds": 273.6, "max_memory_gb": 15.97, "ce": 0.2702498340513557, "rank": 0.0} +{"step": 162, "steps": 500, "lr": 2e-05, "grad_norm": 0.24327681958675385, "loss_scale": 4.0, "retries": 0, "seconds": 275.3, "max_memory_gb": 15.97, "ce": 0.16093935910612345, "rank": 0.0} +{"step": 163, "steps": 500, "lr": 2e-05, "grad_norm": 0.7985435724258423, "loss_scale": 4.0, "retries": 0, "seconds": 277.0, "max_memory_gb": 15.97, "ce": 0.19735357887111604, "rank": 0.0} +{"step": 164, "steps": 500, "lr": 2e-05, "grad_norm": 0.712617814540863, "loss_scale": 4.0, "retries": 0, "seconds": 278.8, "max_memory_gb": 15.97, "ce": 0.29396508913487196, "rank": 0.06778010725975037} +{"step": 165, "steps": 500, "lr": 2e-05, "grad_norm": 0.40646645426750183, "loss_scale": 4.0, "retries": 0, "seconds": 280.5, "max_memory_gb": 15.97, "ce": 0.1855996218509972, "rank": 0.0} +{"step": 166, "steps": 500, "lr": 2e-05, "grad_norm": 0.6491143703460693, "loss_scale": 4.0, "retries": 0, "seconds": 282.2, "max_memory_gb": 15.97, "ce": 0.2558210985735059, "rank": 0.0} +{"step": 167, "steps": 500, "lr": 2e-05, "grad_norm": 0.5381629467010498, "loss_scale": 4.0, "retries": 0, "seconds": 284.0, "max_memory_gb": 15.97, "ce": 0.2540454906411469, "rank": 0.0} +{"step": 168, "steps": 500, "lr": 2e-05, "grad_norm": 0.6585780382156372, "loss_scale": 4.0, "retries": 0, "seconds": 285.6, "max_memory_gb": 15.97, "ce": 0.16223752684891224, "rank": 0.0} +{"step": 169, "steps": 500, "lr": 2e-05, "grad_norm": 2.0361406803131104, "loss_scale": 4.0, "retries": 0, "seconds": 287.3, "max_memory_gb": 15.97, "ce": 0.5812805206514895, "rank": 0.0476527214050293} +{"step": 170, "steps": 500, "lr": 2e-05, "grad_norm": 0.4425223469734192, "loss_scale": 4.0, "retries": 0, "seconds": 288.8, "max_memory_gb": 15.97, "ce": 0.1343854967271909, "rank": 0.0} +{"step": 171, "steps": 500, "lr": 2e-05, "grad_norm": 0.8533207178115845, "loss_scale": 4.0, "retries": 0, "seconds": 290.6, "max_memory_gb": 15.97, "ce": 0.19363722956040874, "rank": 0.0} +{"step": 172, "steps": 500, "lr": 2e-05, "grad_norm": 0.5993763208389282, "loss_scale": 4.0, "retries": 0, "seconds": 292.5, "max_memory_gb": 15.97, "ce": 0.17550403671339154, "rank": 0.0} +{"step": 173, "steps": 500, "lr": 2e-05, "grad_norm": 1.1846821308135986, "loss_scale": 4.0, "retries": 0, "seconds": 294.2, "max_memory_gb": 15.97, "ce": 0.24331332184374332, "rank": 0.012835368514060974} +{"step": 174, "steps": 500, "lr": 2e-05, "grad_norm": 0.6669949293136597, "loss_scale": 4.0, "retries": 0, "seconds": 296.0, "max_memory_gb": 15.97, "ce": 0.17341359652346, "rank": 0.0} +{"step": 175, "steps": 500, "lr": 2e-05, "grad_norm": 0.43940189480781555, "loss_scale": 4.0, "retries": 0, "seconds": 297.7, "max_memory_gb": 15.97, "ce": 0.2806274222675711, "rank": 0.0} +{"step": 176, "steps": 500, "lr": 2e-05, "grad_norm": 0.48848822712898254, "loss_scale": 4.0, "retries": 0, "seconds": 299.4, "max_memory_gb": 15.97, "ce": 0.11690023867413402, "rank": 0.0} +{"step": 177, "steps": 500, "lr": 2e-05, "grad_norm": 0.7745523452758789, "loss_scale": 4.0, "retries": 0, "seconds": 301.0, "max_memory_gb": 15.97, "ce": 0.25126402522437274, "rank": 0.12795332074165344} +{"step": 178, "steps": 500, "lr": 2e-05, "grad_norm": 0.4847721755504608, "loss_scale": 4.0, "retries": 0, "seconds": 302.7, "max_memory_gb": 15.97, "ce": 0.2196516200201586, "rank": 0.0} +{"step": 179, "steps": 500, "lr": 2e-05, "grad_norm": 0.34789586067199707, "loss_scale": 4.0, "retries": 0, "seconds": 304.3, "max_memory_gb": 15.97, "ce": 0.3943133303546347, "rank": 0.0} +{"step": 180, "steps": 500, "lr": 2e-05, "grad_norm": 1.0017377138137817, "loss_scale": 4.0, "retries": 0, "seconds": 306.0, "max_memory_gb": 15.97, "ce": 0.3230156502686441, "rank": 0.0} +{"step": 181, "steps": 500, "lr": 2e-05, "grad_norm": 0.7152063846588135, "loss_scale": 4.0, "retries": 0, "seconds": 307.7, "max_memory_gb": 15.97, "ce": 0.15631167241372168, "rank": 0.0} +{"step": 182, "steps": 500, "lr": 2e-05, "grad_norm": 0.5163620114326477, "loss_scale": 4.0, "retries": 0, "seconds": 309.3, "max_memory_gb": 15.97, "ce": 0.17821748385904357, "rank": 0.0} +{"step": 183, "steps": 500, "lr": 2e-05, "grad_norm": 0.2538190484046936, "loss_scale": 4.0, "retries": 0, "seconds": 311.0, "max_memory_gb": 15.97, "ce": 0.06293086457299069, "rank": 0.0} +{"step": 184, "steps": 500, "lr": 2e-05, "grad_norm": 0.48813503980636597, "loss_scale": 4.0, "retries": 0, "seconds": 312.7, "max_memory_gb": 15.97, "ce": 0.11499026417732239, "rank": 0.11012183874845505} +{"step": 185, "steps": 500, "lr": 2e-05, "grad_norm": 0.38124316930770874, "loss_scale": 4.0, "retries": 0, "seconds": 314.3, "max_memory_gb": 15.97, "ce": 0.13087488245218992, "rank": 0.0} +{"step": 186, "steps": 500, "lr": 2e-05, "grad_norm": 0.3875490725040436, "loss_scale": 4.0, "retries": 0, "seconds": 316.0, "max_memory_gb": 15.97, "ce": 0.08577954833162948, "rank": 0.0} +{"step": 187, "steps": 500, "lr": 2e-05, "grad_norm": 0.8748704195022583, "loss_scale": 4.0, "retries": 0, "seconds": 317.7, "max_memory_gb": 15.97, "ce": 0.22925309860147536, "rank": 0.1292063593864441} +{"step": 188, "steps": 500, "lr": 2e-05, "grad_norm": 0.1729312390089035, "loss_scale": 4.0, "retries": 0, "seconds": 319.2, "max_memory_gb": 15.97, "ce": 0.1488961864670273, "rank": 0.0} +{"step": 189, "steps": 500, "lr": 2e-05, "grad_norm": 0.4035317301750183, "loss_scale": 4.0, "retries": 0, "seconds": 320.7, "max_memory_gb": 15.97, "ce": 0.10945629212073982, "rank": 0.0} +{"step": 190, "steps": 500, "lr": 2e-05, "grad_norm": 0.9109958410263062, "loss_scale": 4.0, "retries": 0, "seconds": 322.3, "max_memory_gb": 15.97, "ce": 0.16213837987743318, "rank": 0.0} +{"step": 191, "steps": 500, "lr": 2e-05, "grad_norm": 0.3422265350818634, "loss_scale": 4.0, "retries": 0, "seconds": 323.9, "max_memory_gb": 15.97, "ce": 0.2716045998968184, "rank": 0.0} +{"step": 192, "steps": 500, "lr": 2e-05, "grad_norm": 0.3795287013053894, "loss_scale": 4.0, "retries": 0, "seconds": 325.5, "max_memory_gb": 15.97, "ce": 0.20894976169802248, "rank": 0.0} +{"step": 193, "steps": 500, "lr": 2e-05, "grad_norm": 0.5019509196281433, "loss_scale": 4.0, "retries": 0, "seconds": 327.3, "max_memory_gb": 15.97, "ce": 0.09933754685334861, "rank": 0.12553280591964722} +{"step": 194, "steps": 500, "lr": 2e-05, "grad_norm": 0.8636553287506104, "loss_scale": 4.0, "retries": 0, "seconds": 328.9, "max_memory_gb": 15.97, "ce": 0.3083056206814945, "rank": 0.0} +{"step": 195, "steps": 500, "lr": 2e-05, "grad_norm": 0.4763840138912201, "loss_scale": 4.0, "retries": 0, "seconds": 330.6, "max_memory_gb": 15.97, "ce": 0.1867138361558318, "rank": 0.0} +{"step": 196, "steps": 500, "lr": 2e-05, "grad_norm": 0.41596701741218567, "loss_scale": 4.0, "retries": 0, "seconds": 332.4, "max_memory_gb": 15.97, "ce": 0.08889344753697515, "rank": 0.0} +{"step": 197, "steps": 500, "lr": 2e-05, "grad_norm": 0.4218289256095886, "loss_scale": 4.0, "retries": 0, "seconds": 334.0, "max_memory_gb": 15.97, "ce": 0.153187045478262, "rank": 0.06978748738765717} +{"step": 198, "steps": 500, "lr": 2e-05, "grad_norm": 0.38573530316352844, "loss_scale": 4.0, "retries": 0, "seconds": 335.7, "max_memory_gb": 15.97, "ce": 0.2097815970191732, "rank": 0.0} +{"step": 199, "steps": 500, "lr": 2e-05, "grad_norm": 1.6297023296356201, "loss_scale": 4.0, "retries": 0, "seconds": 337.4, "max_memory_gb": 15.97, "ce": 0.17717516608536243, "rank": 0.0} +{"step": 200, "steps": 500, "lr": 2e-05, "grad_norm": 0.3465699255466461, "loss_scale": 4.0, "retries": 0, "seconds": 339.1, "max_memory_gb": 15.97, "ce": 0.24572350189555436, "rank": 0.0} +{"step": 201, "steps": 500, "lr": 2e-05, "grad_norm": 0.3720211088657379, "loss_scale": 4.0, "retries": 0, "seconds": 340.8, "max_memory_gb": 15.97, "ce": 0.13135400880128145, "rank": 0.0} +{"step": 202, "steps": 500, "lr": 2e-05, "grad_norm": 0.23384228348731995, "loss_scale": 4.0, "retries": 0, "seconds": 342.5, "max_memory_gb": 15.97, "ce": 0.03959836839931086, "rank": 0.03567443788051605} +{"step": 203, "steps": 500, "lr": 2e-05, "grad_norm": 0.6096465587615967, "loss_scale": 4.0, "retries": 0, "seconds": 344.2, "max_memory_gb": 15.97, "ce": 0.16087653988506645, "rank": 0.0} +{"step": 204, "steps": 500, "lr": 2e-05, "grad_norm": 1.059982180595398, "loss_scale": 4.0, "retries": 0, "seconds": 345.8, "max_memory_gb": 15.97, "ce": 0.15116012055659667, "rank": 0.0} +{"step": 205, "steps": 500, "lr": 2e-05, "grad_norm": 0.7675436735153198, "loss_scale": 4.0, "retries": 0, "seconds": 347.5, "max_memory_gb": 15.97, "ce": 0.15098264298285358, "rank": 0.0} +{"step": 206, "steps": 500, "lr": 2e-05, "grad_norm": 0.5644130110740662, "loss_scale": 4.0, "retries": 0, "seconds": 349.2, "max_memory_gb": 15.97, "ce": 0.09942266356665641, "rank": 0.0} +{"step": 207, "steps": 500, "lr": 2e-05, "grad_norm": 0.7352108955383301, "loss_scale": 4.0, "retries": 0, "seconds": 350.9, "max_memory_gb": 15.97, "ce": 0.47193575627170503, "rank": 0.19141173362731934} +{"step": 208, "steps": 500, "lr": 2e-05, "grad_norm": 0.6068315505981445, "loss_scale": 4.0, "retries": 0, "seconds": 352.6, "max_memory_gb": 15.97, "ce": 0.06231394785572775, "rank": 0.0} +{"step": 209, "steps": 500, "lr": 2e-05, "grad_norm": 0.40981727838516235, "loss_scale": 4.0, "retries": 0, "seconds": 354.3, "max_memory_gb": 15.97, "ce": 0.20632119209039956, "rank": 0.12632128596305847} +{"step": 210, "steps": 500, "lr": 2e-05, "grad_norm": 0.5700134634971619, "loss_scale": 4.0, "retries": 0, "seconds": 356.0, "max_memory_gb": 15.97, "ce": 0.25059142312966287, "rank": 0.03867757320404053} +{"step": 211, "steps": 500, "lr": 2e-05, "grad_norm": 0.41998621821403503, "loss_scale": 4.0, "retries": 0, "seconds": 357.7, "max_memory_gb": 15.97, "ce": 0.20632144785486162, "rank": 0.0} +{"step": 212, "steps": 500, "lr": 2e-05, "grad_norm": 1.224255919456482, "loss_scale": 4.0, "retries": 0, "seconds": 359.4, "max_memory_gb": 15.97, "ce": 0.3367761053959839, "rank": 0.0} +{"step": 213, "steps": 500, "lr": 2e-05, "grad_norm": 0.5037229657173157, "loss_scale": 4.0, "retries": 0, "seconds": 361.0, "max_memory_gb": 15.97, "ce": 0.2802413636818528, "rank": 0.0} +{"step": 214, "steps": 500, "lr": 2e-05, "grad_norm": 1.3010214567184448, "loss_scale": 4.0, "retries": 0, "seconds": 362.7, "max_memory_gb": 15.97, "ce": 0.30149270358379, "rank": 0.0} +{"step": 215, "steps": 500, "lr": 2e-05, "grad_norm": 1.4220404624938965, "loss_scale": 4.0, "retries": 0, "seconds": 364.4, "max_memory_gb": 15.97, "ce": 0.42901099065784365, "rank": 0.0} +{"step": 216, "steps": 500, "lr": 2e-05, "grad_norm": 1.7496871948242188, "loss_scale": 4.0, "retries": 0, "seconds": 366.1, "max_memory_gb": 15.97, "ce": 0.26989135704934597, "rank": 0.0} +{"step": 217, "steps": 500, "lr": 2e-05, "grad_norm": 0.2990466058254242, "loss_scale": 4.0, "retries": 0, "seconds": 367.8, "max_memory_gb": 15.97, "ce": 0.07025110017275438, "rank": 0.0} +{"step": 218, "steps": 500, "lr": 2e-05, "grad_norm": 1.0973963737487793, "loss_scale": 4.0, "retries": 0, "seconds": 369.4, "max_memory_gb": 15.97, "ce": 0.29585708043305203, "rank": 0.0} +{"step": 219, "steps": 500, "lr": 2e-05, "grad_norm": 0.5241538882255554, "loss_scale": 4.0, "retries": 0, "seconds": 371.1, "max_memory_gb": 15.97, "ce": 0.12682521803071722, "rank": 0.0} +{"step": 220, "steps": 500, "lr": 2e-05, "grad_norm": 1.8322601318359375, "loss_scale": 4.0, "retries": 0, "seconds": 372.8, "max_memory_gb": 15.97, "ce": 0.13129854726139456, "rank": 0.0} +{"step": 221, "steps": 500, "lr": 2e-05, "grad_norm": 1.4399718046188354, "loss_scale": 4.0, "retries": 0, "seconds": 374.5, "max_memory_gb": 15.97, "ce": 0.16851136507466435, "rank": 0.0} +{"step": 222, "steps": 500, "lr": 2e-05, "grad_norm": 4.826598167419434, "loss_scale": 4.0, "retries": 0, "seconds": 376.2, "max_memory_gb": 15.97, "ce": 0.5376496991375461, "rank": 0.0} +{"step": 223, "steps": 500, "lr": 2e-05, "grad_norm": 0.36664873361587524, "loss_scale": 4.0, "retries": 0, "seconds": 377.9, "max_memory_gb": 15.97, "ce": 0.18891223787795752, "rank": 0.0} +{"step": 224, "steps": 500, "lr": 2e-05, "grad_norm": 0.4085841476917267, "loss_scale": 4.0, "retries": 0, "seconds": 379.5, "max_memory_gb": 15.97, "ce": 0.0977970683015883, "rank": 0.09712017327547073} +{"step": 225, "steps": 500, "lr": 2e-05, "grad_norm": 0.4183683395385742, "loss_scale": 4.0, "retries": 0, "seconds": 381.2, "max_memory_gb": 15.97, "ce": 0.08668680302798748, "rank": 0.0} +{"step": 226, "steps": 500, "lr": 2e-05, "grad_norm": 0.414103627204895, "loss_scale": 4.0, "retries": 0, "seconds": 382.8, "max_memory_gb": 15.97, "ce": 0.10909225046634674, "rank": 0.0} +{"step": 227, "steps": 500, "lr": 2e-05, "grad_norm": 0.2739315629005432, "loss_scale": 4.0, "retries": 0, "seconds": 384.5, "max_memory_gb": 15.97, "ce": 0.03612093938863836, "rank": 0.0} +{"step": 228, "steps": 500, "lr": 2e-05, "grad_norm": 0.5309640169143677, "loss_scale": 4.0, "retries": 0, "seconds": 386.1, "max_memory_gb": 15.97, "ce": 0.09989316732389852, "rank": 0.0} +{"step": 229, "steps": 500, "lr": 2e-05, "grad_norm": 0.4681938886642456, "loss_scale": 4.0, "retries": 0, "seconds": 387.8, "max_memory_gb": 15.97, "ce": 0.20600416540401056, "rank": 0.0} +{"step": 230, "steps": 500, "lr": 2e-05, "grad_norm": 0.6049517393112183, "loss_scale": 4.0, "retries": 0, "seconds": 389.5, "max_memory_gb": 15.97, "ce": 0.22426383686251938, "rank": 0.0} +{"step": 231, "steps": 500, "lr": 2e-05, "grad_norm": 0.34920644760131836, "loss_scale": 4.0, "retries": 0, "seconds": 391.0, "max_memory_gb": 15.97, "ce": 0.20183468062896281, "rank": 0.0} +{"step": 232, "steps": 500, "lr": 2e-05, "grad_norm": 0.5258957743644714, "loss_scale": 4.0, "retries": 0, "seconds": 392.6, "max_memory_gb": 15.97, "ce": 0.061941527121234685, "rank": 0.0} +{"step": 233, "steps": 500, "lr": 2e-05, "grad_norm": 2.050236463546753, "loss_scale": 4.0, "retries": 0, "seconds": 394.3, "max_memory_gb": 15.97, "ce": 0.16299906314816326, "rank": 0.0} +{"step": 234, "steps": 500, "lr": 2e-05, "grad_norm": 0.5372351408004761, "loss_scale": 4.0, "retries": 0, "seconds": 395.9, "max_memory_gb": 15.97, "ce": 0.1905848536407575, "rank": 0.0} +{"step": 235, "steps": 500, "lr": 2e-05, "grad_norm": 0.5509185791015625, "loss_scale": 4.0, "retries": 0, "seconds": 397.6, "max_memory_gb": 15.97, "ce": 0.28665669925976545, "rank": 0.0} +{"step": 236, "steps": 500, "lr": 2e-05, "grad_norm": 0.3519577383995056, "loss_scale": 4.0, "retries": 0, "seconds": 399.3, "max_memory_gb": 15.97, "ce": 0.10549175774212927, "rank": 0.0} +{"step": 237, "steps": 500, "lr": 2e-05, "grad_norm": 1.706970453262329, "loss_scale": 4.0, "retries": 0, "seconds": 400.9, "max_memory_gb": 15.97, "ce": 0.20618129655485973, "rank": 0.0} +{"step": 238, "steps": 500, "lr": 2e-05, "grad_norm": 0.7094486951828003, "loss_scale": 4.0, "retries": 0, "seconds": 402.6, "max_memory_gb": 15.97, "ce": 0.09805290529038757, "rank": 0.0} +{"step": 239, "steps": 500, "lr": 2e-05, "grad_norm": 0.4672143757343292, "loss_scale": 4.0, "retries": 0, "seconds": 404.2, "max_memory_gb": 15.97, "ce": 0.17049046169267967, "rank": 0.0} +{"step": 240, "steps": 500, "lr": 2e-05, "grad_norm": 0.7489423155784607, "loss_scale": 4.0, "retries": 0, "seconds": 405.9, "max_memory_gb": 15.97, "ce": 0.0962223265087232, "rank": 0.0} +{"step": 241, "steps": 500, "lr": 2e-05, "grad_norm": 0.492169052362442, "loss_scale": 4.0, "retries": 0, "seconds": 407.7, "max_memory_gb": 15.97, "ce": 0.1692780003650114, "rank": 0.0} +{"step": 242, "steps": 500, "lr": 2e-05, "grad_norm": 0.5568295121192932, "loss_scale": 4.0, "retries": 0, "seconds": 409.5, "max_memory_gb": 15.97, "ce": 0.1288887948030606, "rank": 0.0} +{"step": 243, "steps": 500, "lr": 2e-05, "grad_norm": 0.6465861797332764, "loss_scale": 4.0, "retries": 0, "seconds": 411.2, "max_memory_gb": 15.97, "ce": 0.05624502996215597, "rank": 0.0} +{"step": 244, "steps": 500, "lr": 2e-05, "grad_norm": 0.9345635771751404, "loss_scale": 4.0, "retries": 0, "seconds": 413.0, "max_memory_gb": 15.97, "ce": 0.28083963016979396, "rank": 0.0} +{"step": 245, "steps": 500, "lr": 2e-05, "grad_norm": 0.48077327013015747, "loss_scale": 4.0, "retries": 0, "seconds": 414.9, "max_memory_gb": 15.97, "ce": 0.32833772752201185, "rank": 0.0} +{"step": 246, "steps": 500, "lr": 2e-05, "grad_norm": 0.27296867966651917, "loss_scale": 4.0, "retries": 0, "seconds": 416.8, "max_memory_gb": 15.97, "ce": 0.04486769950017333, "rank": 0.0} +{"step": 247, "steps": 500, "lr": 2e-05, "grad_norm": 0.5621668696403503, "loss_scale": 4.0, "retries": 0, "seconds": 418.7, "max_memory_gb": 15.97, "ce": 0.10886945715174079, "rank": 0.0} +{"step": 248, "steps": 500, "lr": 2e-05, "grad_norm": 0.5490620732307434, "loss_scale": 4.0, "retries": 0, "seconds": 420.4, "max_memory_gb": 15.97, "ce": 0.08582317008404061, "rank": 0.0} +{"step": 249, "steps": 500, "lr": 2e-05, "grad_norm": 0.5892318487167358, "loss_scale": 4.0, "retries": 0, "seconds": 422.1, "max_memory_gb": 15.97, "ce": 0.08024793383083306, "rank": 0.0} +{"step": 250, "steps": 500, "lr": 2e-05, "grad_norm": 0.5855017304420471, "loss_scale": 4.0, "retries": 0, "seconds": 423.8, "max_memory_gb": 15.97, "ce": 0.04973500088090077, "rank": 0.0} +{"step": 251, "steps": 500, "lr": 2e-05, "grad_norm": 0.3379223644733429, "loss_scale": 4.0, "retries": 0, "seconds": 425.4, "max_memory_gb": 15.97, "ce": 0.13316785628558137, "rank": 0.0} +{"step": 252, "steps": 500, "lr": 2e-05, "grad_norm": 0.7326961159706116, "loss_scale": 4.0, "retries": 0, "seconds": 427.1, "max_memory_gb": 15.97, "ce": 0.06764380994718522, "rank": 0.0} +{"step": 253, "steps": 500, "lr": 2e-05, "grad_norm": 0.42378348112106323, "loss_scale": 4.0, "retries": 0, "seconds": 428.7, "max_memory_gb": 15.97, "ce": 0.11252828370197676, "rank": 0.0} +{"step": 254, "steps": 500, "lr": 2e-05, "grad_norm": 1.1255155801773071, "loss_scale": 4.0, "retries": 0, "seconds": 430.4, "max_memory_gb": 15.97, "ce": 0.17277200747048482, "rank": 0.0} +{"step": 255, "steps": 500, "lr": 2e-05, "grad_norm": 0.42377349734306335, "loss_scale": 4.0, "retries": 0, "seconds": 432.0, "max_memory_gb": 15.97, "ce": 0.20492814644239843, "rank": 0.0} +{"step": 256, "steps": 500, "lr": 2e-05, "grad_norm": 0.2975444793701172, "loss_scale": 4.0, "retries": 0, "seconds": 433.7, "max_memory_gb": 15.97, "ce": 0.07671249716077, "rank": 0.0} +{"step": 257, "steps": 500, "lr": 2e-05, "grad_norm": 0.46148180961608887, "loss_scale": 4.0, "retries": 0, "seconds": 435.4, "max_memory_gb": 15.97, "ce": 0.07865941227646545, "rank": 0.0} +{"step": 258, "steps": 500, "lr": 2e-05, "grad_norm": 2.0414578914642334, "loss_scale": 4.0, "retries": 0, "seconds": 437.1, "max_memory_gb": 15.97, "ce": 0.10233918350422755, "rank": 0.0} +{"step": 259, "steps": 500, "lr": 2e-05, "grad_norm": 0.24305786192417145, "loss_scale": 4.0, "retries": 0, "seconds": 438.8, "max_memory_gb": 15.97, "ce": 0.037255028379149735, "rank": 0.0} +{"step": 260, "steps": 500, "lr": 2e-05, "grad_norm": 0.6768408417701721, "loss_scale": 4.0, "retries": 0, "seconds": 440.6, "max_memory_gb": 15.97, "ce": 0.11089557407831308, "rank": 0.0} +{"step": 261, "steps": 500, "lr": 2e-05, "grad_norm": 1.9178935289382935, "loss_scale": 4.0, "retries": 0, "seconds": 442.3, "max_memory_gb": 15.97, "ce": 0.16555006425187457, "rank": 0.0} +{"step": 262, "steps": 500, "lr": 2e-05, "grad_norm": 0.70828777551651, "loss_scale": 4.0, "retries": 0, "seconds": 444.0, "max_memory_gb": 15.97, "ce": 0.16596190276322886, "rank": 0.0} +{"step": 263, "steps": 500, "lr": 2e-05, "grad_norm": 0.9156332015991211, "loss_scale": 4.0, "retries": 0, "seconds": 445.7, "max_memory_gb": 15.97, "ce": 0.18622132696327753, "rank": 0.12502574920654297} +{"step": 264, "steps": 500, "lr": 2e-05, "grad_norm": 1.1926374435424805, "loss_scale": 4.0, "retries": 0, "seconds": 447.4, "max_memory_gb": 15.97, "ce": 0.14607847342267632, "rank": 0.0} +{"step": 265, "steps": 500, "lr": 2e-05, "grad_norm": 0.36426088213920593, "loss_scale": 4.0, "retries": 0, "seconds": 449.1, "max_memory_gb": 15.97, "ce": 0.0842398727836553, "rank": 0.0} +{"step": 266, "steps": 500, "lr": 2e-05, "grad_norm": 3.274188280105591, "loss_scale": 4.0, "retries": 0, "seconds": 450.8, "max_memory_gb": 15.97, "ce": 0.5178831785742659, "rank": 0.0} +{"step": 267, "steps": 500, "lr": 2e-05, "grad_norm": 0.30685436725616455, "loss_scale": 4.0, "retries": 0, "seconds": 452.5, "max_memory_gb": 15.97, "ce": 0.1315154225740116, "rank": 0.0} +{"step": 268, "steps": 500, "lr": 2e-05, "grad_norm": 0.35614892840385437, "loss_scale": 4.0, "retries": 0, "seconds": 454.2, "max_memory_gb": 15.97, "ce": 0.22066504578106105, "rank": 0.0} +{"step": 269, "steps": 500, "lr": 2e-05, "grad_norm": 3.8761558532714844, "loss_scale": 4.0, "retries": 0, "seconds": 455.8, "max_memory_gb": 15.97, "ce": 0.4470474958943669, "rank": 0.0} +{"step": 270, "steps": 500, "lr": 2e-05, "grad_norm": 0.7868456244468689, "loss_scale": 4.0, "retries": 0, "seconds": 457.5, "max_memory_gb": 15.97, "ce": 0.1484512201277539, "rank": 0.0} +{"step": 271, "steps": 500, "lr": 2e-05, "grad_norm": 0.43243369460105896, "loss_scale": 4.0, "retries": 0, "seconds": 459.1, "max_memory_gb": 15.97, "ce": 0.06540000590030104, "rank": 0.11922066658735275} +{"step": 272, "steps": 500, "lr": 2e-05, "grad_norm": 1.0573761463165283, "loss_scale": 4.0, "retries": 0, "seconds": 460.8, "max_memory_gb": 15.97, "ce": 0.10713469356414862, "rank": 0.0} +{"step": 273, "steps": 500, "lr": 2e-05, "grad_norm": 1.0488195419311523, "loss_scale": 4.0, "retries": 0, "seconds": 462.5, "max_memory_gb": 15.97, "ce": 0.07501634131767787, "rank": 0.1252851039171219} +{"step": 274, "steps": 500, "lr": 2e-05, "grad_norm": 0.5895031094551086, "loss_scale": 4.0, "retries": 0, "seconds": 464.2, "max_memory_gb": 15.97, "ce": 0.12144765979610384, "rank": 0.0} +{"step": 275, "steps": 500, "lr": 2e-05, "grad_norm": 1.4355741739273071, "loss_scale": 4.0, "retries": 0, "seconds": 465.9, "max_memory_gb": 15.97, "ce": 0.6159668123873416, "rank": 0.05688130855560303} +{"step": 276, "steps": 500, "lr": 2e-05, "grad_norm": 0.451901376247406, "loss_scale": 4.0, "retries": 0, "seconds": 467.6, "max_memory_gb": 15.97, "ce": 0.027779080322943628, "rank": 0.0} +{"step": 277, "steps": 500, "lr": 2e-05, "grad_norm": 0.4465535879135132, "loss_scale": 4.0, "retries": 0, "seconds": 469.3, "max_memory_gb": 15.97, "ce": 0.030171260179486126, "rank": 0.0} +{"step": 278, "steps": 500, "lr": 2e-05, "grad_norm": 1.3920278549194336, "loss_scale": 4.0, "retries": 0, "seconds": 471.0, "max_memory_gb": 15.97, "ce": 0.21291973792540375, "rank": 0.0338037833571434} +{"step": 279, "steps": 500, "lr": 2e-05, "grad_norm": 1.084174394607544, "loss_scale": 4.0, "retries": 0, "seconds": 472.6, "max_memory_gb": 15.97, "ce": 0.048444035433931276, "rank": 0.0} +{"step": 280, "steps": 500, "lr": 2e-05, "grad_norm": 1.3381404876708984, "loss_scale": 4.0, "retries": 0, "seconds": 474.3, "max_memory_gb": 15.97, "ce": 0.04483283541048877, "rank": 0.0} +{"step": 281, "steps": 500, "lr": 2e-05, "grad_norm": 1.5953463315963745, "loss_scale": 4.0, "retries": 0, "seconds": 476.1, "max_memory_gb": 15.97, "ce": 0.1322933626943268, "rank": 0.0} +{"step": 282, "steps": 500, "lr": 2e-05, "grad_norm": 0.3936297297477722, "loss_scale": 4.0, "retries": 0, "seconds": 477.8, "max_memory_gb": 15.97, "ce": 0.08838211544207297, "rank": 0.0} +{"step": 283, "steps": 500, "lr": 2e-05, "grad_norm": 0.9506015181541443, "loss_scale": 4.0, "retries": 0, "seconds": 479.5, "max_memory_gb": 15.97, "ce": 0.08776871873124037, "rank": 0.0} +{"step": 284, "steps": 500, "lr": 2e-05, "grad_norm": 0.8843268752098083, "loss_scale": 4.0, "retries": 0, "seconds": 481.2, "max_memory_gb": 15.97, "ce": 0.06742159265559167, "rank": 0.0} +{"step": 285, "steps": 500, "lr": 2e-05, "grad_norm": 0.6436477899551392, "loss_scale": 4.0, "retries": 0, "seconds": 482.9, "max_memory_gb": 15.97, "ce": 0.11056832119356841, "rank": 0.0} +{"step": 286, "steps": 500, "lr": 2e-05, "grad_norm": 0.3500247299671173, "loss_scale": 4.0, "retries": 0, "seconds": 484.5, "max_memory_gb": 15.97, "ce": 0.049915233088540845, "rank": 0.0} +{"step": 287, "steps": 500, "lr": 2e-05, "grad_norm": 0.6192502975463867, "loss_scale": 4.0, "retries": 0, "seconds": 486.1, "max_memory_gb": 15.97, "ce": 0.11072199267800897, "rank": 0.0} +{"step": 288, "steps": 500, "lr": 2e-05, "grad_norm": 0.12963369488716125, "loss_scale": 4.0, "retries": 0, "seconds": 487.8, "max_memory_gb": 15.97, "ce": 0.010703515137720387, "rank": 0.0} +{"step": 289, "steps": 500, "lr": 2e-05, "grad_norm": 0.5193353891372681, "loss_scale": 4.0, "retries": 0, "seconds": 489.5, "max_memory_gb": 15.97, "ce": 0.08238957372668665, "rank": 0.0} +{"step": 290, "steps": 500, "lr": 2e-05, "grad_norm": 0.5923224091529846, "loss_scale": 4.0, "retries": 0, "seconds": 491.2, "max_memory_gb": 15.97, "ce": 0.1199782418407267, "rank": 0.0} +{"step": 291, "steps": 500, "lr": 2e-05, "grad_norm": 0.398958295583725, "loss_scale": 4.0, "retries": 0, "seconds": 492.9, "max_memory_gb": 15.97, "ce": 0.020187301284749992, "rank": 0.0} +{"step": 292, "steps": 500, "lr": 2e-05, "grad_norm": 1.342308521270752, "loss_scale": 4.0, "retries": 0, "seconds": 494.7, "max_memory_gb": 15.97, "ce": 0.10879518017100054, "rank": 0.0} +{"step": 293, "steps": 500, "lr": 2e-05, "grad_norm": 0.40830153226852417, "loss_scale": 4.0, "retries": 0, "seconds": 496.4, "max_memory_gb": 15.97, "ce": 0.05325017204449978, "rank": 0.0} +{"step": 294, "steps": 500, "lr": 2e-05, "grad_norm": 1.299197793006897, "loss_scale": 4.0, "retries": 0, "seconds": 498.0, "max_memory_gb": 15.97, "ce": 0.07429180253529921, "rank": 0.0} +{"step": 295, "steps": 500, "lr": 2e-05, "grad_norm": 1.8615632057189941, "loss_scale": 4.0, "retries": 0, "seconds": 499.6, "max_memory_gb": 15.97, "ce": 0.1831771519209724, "rank": 0.0} +{"step": 296, "steps": 500, "lr": 2e-05, "grad_norm": 1.581964135169983, "loss_scale": 4.0, "retries": 0, "seconds": 501.4, "max_memory_gb": 15.97, "ce": 0.2702555555733852, "rank": 0.0} +{"step": 297, "steps": 500, "lr": 2e-05, "grad_norm": 0.5198065638542175, "loss_scale": 4.0, "retries": 0, "seconds": 502.9, "max_memory_gb": 15.97, "ce": 0.12712196898064576, "rank": 0.0} +{"step": 298, "steps": 500, "lr": 2e-05, "grad_norm": 3.084944486618042, "loss_scale": 4.0, "retries": 0, "seconds": 504.6, "max_memory_gb": 15.97, "ce": 0.3762478038261179, "rank": 0.0} +{"step": 299, "steps": 500, "lr": 2e-05, "grad_norm": 0.7107788324356079, "loss_scale": 4.0, "retries": 0, "seconds": 506.3, "max_memory_gb": 15.97, "ce": 0.040637604128278326, "rank": 0.0} +{"step": 300, "steps": 500, "lr": 2e-05, "grad_norm": 0.8871150016784668, "loss_scale": 4.0, "retries": 0, "seconds": 508.0, "max_memory_gb": 15.97, "ce": 0.2578498542570742, "rank": 0.0} +{"step": 301, "steps": 500, "lr": 2e-05, "grad_norm": 0.5210071206092834, "loss_scale": 4.0, "retries": 0, "seconds": 509.5, "max_memory_gb": 15.97, "ce": 0.1032947624407825, "rank": 0.0} +{"step": 302, "steps": 500, "lr": 2e-05, "grad_norm": 0.48845162987709045, "loss_scale": 4.0, "retries": 0, "seconds": 511.2, "max_memory_gb": 15.97, "ce": 0.01962350851681549, "rank": 0.0} +{"step": 303, "steps": 500, "lr": 2e-05, "grad_norm": 0.6080824732780457, "loss_scale": 4.0, "retries": 0, "seconds": 512.8, "max_memory_gb": 15.97, "ce": 0.02532466480624862, "rank": 0.0} +{"step": 304, "steps": 500, "lr": 2e-05, "grad_norm": 1.412037968635559, "loss_scale": 4.0, "retries": 0, "seconds": 514.5, "max_memory_gb": 15.97, "ce": 0.05044349182935548, "rank": 0.012337103486061096} +{"step": 305, "steps": 500, "lr": 2e-05, "grad_norm": 1.855017900466919, "loss_scale": 4.0, "retries": 0, "seconds": 516.2, "max_memory_gb": 15.97, "ce": 0.09799634621595033, "rank": 0.0} +{"step": 306, "steps": 500, "lr": 2e-05, "grad_norm": 0.5123005509376526, "loss_scale": 4.0, "retries": 0, "seconds": 517.9, "max_memory_gb": 15.97, "ce": 0.02374698868516134, "rank": 0.0} +{"step": 307, "steps": 500, "lr": 2e-05, "grad_norm": 0.7702963948249817, "loss_scale": 4.0, "retries": 0, "seconds": 519.5, "max_memory_gb": 15.97, "ce": 0.10939688631333411, "rank": 0.0} +{"step": 308, "steps": 500, "lr": 2e-05, "grad_norm": 1.6198979616165161, "loss_scale": 4.0, "retries": 0, "seconds": 521.3, "max_memory_gb": 15.97, "ce": 0.22259544520420604, "rank": 0.0} +{"step": 309, "steps": 500, "lr": 2e-05, "grad_norm": 0.5126925110816956, "loss_scale": 4.0, "retries": 0, "seconds": 522.9, "max_memory_gb": 15.97, "ce": 0.13980716883088462, "rank": 0.0} +{"step": 310, "steps": 500, "lr": 2e-05, "grad_norm": 0.6056327223777771, "loss_scale": 4.0, "retries": 0, "seconds": 524.6, "max_memory_gb": 15.97, "ce": 0.04076239946880378, "rank": 0.0} +{"step": 311, "steps": 500, "lr": 2e-05, "grad_norm": 0.49992650747299194, "loss_scale": 4.0, "retries": 0, "seconds": 526.3, "max_memory_gb": 15.97, "ce": 0.1066243256063899, "rank": 0.0} +{"step": 312, "steps": 500, "lr": 2e-05, "grad_norm": 0.18774406611919403, "loss_scale": 4.0, "retries": 0, "seconds": 528.2, "max_memory_gb": 15.97, "ce": 0.016631827922537923, "rank": 0.0} +{"step": 313, "steps": 500, "lr": 2e-05, "grad_norm": 1.3337814807891846, "loss_scale": 4.0, "retries": 0, "seconds": 530.0, "max_memory_gb": 15.97, "ce": 0.19207003092742525, "rank": 0.0} +{"step": 314, "steps": 500, "lr": 2e-05, "grad_norm": 0.9958330392837524, "loss_scale": 4.0, "retries": 0, "seconds": 531.9, "max_memory_gb": 15.97, "ce": 0.19078665121924132, "rank": 0.0} +{"step": 315, "steps": 500, "lr": 2e-05, "grad_norm": 0.533900797367096, "loss_scale": 4.0, "retries": 0, "seconds": 533.6, "max_memory_gb": 15.97, "ce": 0.07316511707722384, "rank": 0.12066259980201721} +{"step": 316, "steps": 500, "lr": 2e-05, "grad_norm": 1.6368519067764282, "loss_scale": 4.0, "retries": 0, "seconds": 535.3, "max_memory_gb": 15.97, "ce": 0.34431169024901465, "rank": 0.0} +{"step": 317, "steps": 500, "lr": 2e-05, "grad_norm": 0.623355507850647, "loss_scale": 4.0, "retries": 0, "seconds": 536.9, "max_memory_gb": 15.97, "ce": 0.10711427627757075, "rank": 0.0} +{"step": 318, "steps": 500, "lr": 2e-05, "grad_norm": 3.810015916824341, "loss_scale": 4.0, "retries": 0, "seconds": 538.6, "max_memory_gb": 15.97, "ce": 0.229798696393118, "rank": 0.0} +{"step": 319, "steps": 500, "lr": 2e-05, "grad_norm": 0.608708381652832, "loss_scale": 4.0, "retries": 0, "seconds": 540.2, "max_memory_gb": 15.97, "ce": 0.07027724824729376, "rank": 0.0} +{"step": 320, "steps": 500, "lr": 2e-05, "grad_norm": 0.8354688286781311, "loss_scale": 4.0, "retries": 0, "seconds": 541.9, "max_memory_gb": 15.97, "ce": 0.09718627467009355, "rank": 0.0} +{"step": 321, "steps": 500, "lr": 2e-05, "grad_norm": 0.8564211130142212, "loss_scale": 4.0, "retries": 0, "seconds": 543.5, "max_memory_gb": 15.97, "ce": 0.12661700070020743, "rank": 0.12683388590812683} +{"step": 322, "steps": 500, "lr": 2e-05, "grad_norm": 0.22467292845249176, "loss_scale": 4.0, "retries": 0, "seconds": 545.2, "max_memory_gb": 15.97, "ce": 0.015071751950017642, "rank": 0.0} +{"step": 323, "steps": 500, "lr": 2e-05, "grad_norm": 0.4847482144832611, "loss_scale": 4.0, "retries": 0, "seconds": 546.9, "max_memory_gb": 15.97, "ce": 0.041984356445027515, "rank": 0.0} +{"step": 324, "steps": 500, "lr": 2e-05, "grad_norm": 0.42536744475364685, "loss_scale": 4.0, "retries": 0, "seconds": 548.6, "max_memory_gb": 15.97, "ce": 0.025532282841595588, "rank": 0.0} +{"step": 325, "steps": 500, "lr": 2e-05, "grad_norm": 0.6857891082763672, "loss_scale": 4.0, "retries": 0, "seconds": 550.4, "max_memory_gb": 15.97, "ce": 0.18843497766647488, "rank": 0.0} +{"step": 326, "steps": 500, "lr": 2e-05, "grad_norm": 0.39978480339050293, "loss_scale": 4.0, "retries": 0, "seconds": 552.3, "max_memory_gb": 15.97, "ce": 0.04465716277627507, "rank": 0.09179866313934326} +{"step": 327, "steps": 500, "lr": 2e-05, "grad_norm": 1.6992218494415283, "loss_scale": 4.0, "retries": 0, "seconds": 554.3, "max_memory_gb": 15.97, "ce": 0.06981941903359257, "rank": 0.0} +{"step": 328, "steps": 500, "lr": 2e-05, "grad_norm": 0.6599891781806946, "loss_scale": 4.0, "retries": 0, "seconds": 556.2, "max_memory_gb": 15.97, "ce": 0.14344665037788218, "rank": 0.0} +{"step": 329, "steps": 500, "lr": 2e-05, "grad_norm": 0.6284317970275879, "loss_scale": 4.0, "retries": 0, "seconds": 558.0, "max_memory_gb": 15.97, "ce": 0.14335628219123464, "rank": 0.0} +{"step": 330, "steps": 500, "lr": 2e-05, "grad_norm": 0.4815906286239624, "loss_scale": 4.0, "retries": 0, "seconds": 559.9, "max_memory_gb": 15.97, "ce": 0.12294617199222557, "rank": 0.0} +{"step": 331, "steps": 500, "lr": 2e-05, "grad_norm": 1.6662267446517944, "loss_scale": 4.0, "retries": 0, "seconds": 561.8, "max_memory_gb": 15.97, "ce": 0.29482693543468486, "rank": 0.0} +{"step": 332, "steps": 500, "lr": 2e-05, "grad_norm": 1.1380407810211182, "loss_scale": 4.0, "retries": 0, "seconds": 563.6, "max_memory_gb": 15.97, "ce": 0.08153011800277454, "rank": 0.0} +{"step": 333, "steps": 500, "lr": 2e-05, "grad_norm": 0.829347550868988, "loss_scale": 4.0, "retries": 0, "seconds": 565.5, "max_memory_gb": 15.97, "ce": 0.25082850069884444, "rank": 0.0} +{"step": 334, "steps": 500, "lr": 2e-05, "grad_norm": 0.6745835542678833, "loss_scale": 4.0, "retries": 0, "seconds": 567.3, "max_memory_gb": 15.97, "ce": 0.12781246722806827, "rank": 0.0} +{"step": 335, "steps": 500, "lr": 2e-05, "grad_norm": 4.293911457061768, "loss_scale": 4.0, "retries": 0, "seconds": 569.2, "max_memory_gb": 15.97, "ce": 0.30360684628612944, "rank": 0.0} +{"step": 336, "steps": 500, "lr": 2e-05, "grad_norm": 1.3391929864883423, "loss_scale": 4.0, "retries": 0, "seconds": 571.0, "max_memory_gb": 15.97, "ce": 0.03765431975989486, "rank": 0.0} +{"step": 337, "steps": 500, "lr": 2e-05, "grad_norm": 2.154407024383545, "loss_scale": 4.0, "retries": 0, "seconds": 572.8, "max_memory_gb": 15.97, "ce": 0.1624146911490243, "rank": 0.0} +{"step": 338, "steps": 500, "lr": 2e-05, "grad_norm": 0.9832291007041931, "loss_scale": 4.0, "retries": 0, "seconds": 574.8, "max_memory_gb": 15.97, "ce": 0.19212356068601366, "rank": 0.0} +{"step": 339, "steps": 500, "lr": 2e-05, "grad_norm": 0.578428328037262, "loss_scale": 4.0, "retries": 0, "seconds": 576.7, "max_memory_gb": 15.97, "ce": 0.12563551935818396, "rank": 0.0} +{"step": 340, "steps": 500, "lr": 2e-05, "grad_norm": 0.6803988218307495, "loss_scale": 4.0, "retries": 0, "seconds": 578.6, "max_memory_gb": 15.97, "ce": 0.1602541999454843, "rank": 0.0} +{"step": 341, "steps": 500, "lr": 2e-05, "grad_norm": 4.395832061767578, "loss_scale": 4.0, "retries": 0, "seconds": 580.3, "max_memory_gb": 15.97, "ce": 0.24086382082896307, "rank": 0.0} +{"step": 342, "steps": 500, "lr": 2e-05, "grad_norm": 1.6587289571762085, "loss_scale": 4.0, "retries": 0, "seconds": 582.1, "max_memory_gb": 15.97, "ce": 0.5257019935525022, "rank": 0.0} +{"step": 343, "steps": 500, "lr": 2e-05, "grad_norm": 0.643463671207428, "loss_scale": 4.0, "retries": 0, "seconds": 583.9, "max_memory_gb": 15.97, "ce": 0.059366102039348334, "rank": 0.0} +{"step": 344, "steps": 500, "lr": 2e-05, "grad_norm": 0.16787850856781006, "loss_scale": 4.0, "retries": 0, "seconds": 585.7, "max_memory_gb": 15.97, "ce": 0.014376832186826505, "rank": 0.0} +{"step": 345, "steps": 500, "lr": 2e-05, "grad_norm": 0.29621243476867676, "loss_scale": 4.0, "retries": 0, "seconds": 587.3, "max_memory_gb": 15.97, "ce": 0.1684983065642882, "rank": 0.0} +{"step": 346, "steps": 500, "lr": 2e-05, "grad_norm": 0.9878698587417603, "loss_scale": 4.0, "retries": 0, "seconds": 589.0, "max_memory_gb": 15.97, "ce": 0.2038887129019713, "rank": 0.0} +{"step": 347, "steps": 500, "lr": 2e-05, "grad_norm": 0.5468770265579224, "loss_scale": 4.0, "retries": 0, "seconds": 590.8, "max_memory_gb": 15.97, "ce": 0.1319225068364176, "rank": 0.0} +{"step": 348, "steps": 500, "lr": 2e-05, "grad_norm": 0.39447537064552307, "loss_scale": 4.0, "retries": 0, "seconds": 592.7, "max_memory_gb": 15.97, "ce": 0.014387559607712319, "rank": 0.0} +{"step": 349, "steps": 500, "lr": 2e-05, "grad_norm": 0.29339858889579773, "loss_scale": 4.0, "retries": 0, "seconds": 594.6, "max_memory_gb": 15.97, "ce": 0.09214304563283804, "rank": 0.0} +{"step": 350, "steps": 500, "lr": 2e-05, "grad_norm": 0.5575858950614929, "loss_scale": 4.0, "retries": 0, "seconds": 596.4, "max_memory_gb": 15.97, "ce": 0.10854594940610696, "rank": 0.0} +{"step": 351, "steps": 500, "lr": 2e-05, "grad_norm": 1.2422202825546265, "loss_scale": 4.0, "retries": 0, "seconds": 598.2, "max_memory_gb": 15.97, "ce": 0.07123046114975295, "rank": 0.0} +{"step": 352, "steps": 500, "lr": 2e-05, "grad_norm": 1.6451493501663208, "loss_scale": 4.0, "retries": 0, "seconds": 600.0, "max_memory_gb": 15.97, "ce": 0.2667147779939114, "rank": 0.0} +{"step": 353, "steps": 500, "lr": 2e-05, "grad_norm": 0.632592499256134, "loss_scale": 4.0, "retries": 0, "seconds": 602.0, "max_memory_gb": 15.97, "ce": 0.11884234175158781, "rank": 0.0} +{"step": 354, "steps": 500, "lr": 2e-05, "grad_norm": 1.6011673212051392, "loss_scale": 4.0, "retries": 0, "seconds": 603.8, "max_memory_gb": 15.97, "ce": 0.18664024754980346, "rank": 0.0} +{"step": 355, "steps": 500, "lr": 2e-05, "grad_norm": 0.4802493155002594, "loss_scale": 4.0, "retries": 0, "seconds": 605.6, "max_memory_gb": 15.97, "ce": 0.08997845004341798, "rank": 0.0} +{"step": 356, "steps": 500, "lr": 2e-05, "grad_norm": 0.5777913928031921, "loss_scale": 4.0, "retries": 0, "seconds": 607.3, "max_memory_gb": 15.97, "ce": 0.05570735255605541, "rank": 0.0} +{"step": 357, "steps": 500, "lr": 2e-05, "grad_norm": 0.42687517404556274, "loss_scale": 4.0, "retries": 0, "seconds": 609.1, "max_memory_gb": 15.97, "ce": 0.06285772301998804, "rank": 0.0} +{"step": 358, "steps": 500, "lr": 2e-05, "grad_norm": 0.8812829852104187, "loss_scale": 4.0, "retries": 0, "seconds": 610.9, "max_memory_gb": 15.97, "ce": 0.21014295866552857, "rank": 0.0} +{"step": 359, "steps": 500, "lr": 2e-05, "grad_norm": 0.4913709759712219, "loss_scale": 4.0, "retries": 0, "seconds": 612.7, "max_memory_gb": 15.97, "ce": 0.06585929865286744, "rank": 0.0} +{"step": 360, "steps": 500, "lr": 2e-05, "grad_norm": 0.7316274046897888, "loss_scale": 4.0, "retries": 0, "seconds": 614.3, "max_memory_gb": 15.97, "ce": 0.060476309301520814, "rank": 0.13669830560684204} +{"step": 361, "steps": 500, "lr": 2e-05, "grad_norm": 1.3902076482772827, "loss_scale": 4.0, "retries": 0, "seconds": 616.0, "max_memory_gb": 15.97, "ce": 0.1405102485223324, "rank": 0.0} +{"step": 362, "steps": 500, "lr": 2e-05, "grad_norm": 3.2627320289611816, "loss_scale": 4.0, "retries": 0, "seconds": 617.7, "max_memory_gb": 15.97, "ce": 0.23734235675874515, "rank": 0.0} +{"step": 363, "steps": 500, "lr": 2e-05, "grad_norm": 0.338730126619339, "loss_scale": 4.0, "retries": 0, "seconds": 619.3, "max_memory_gb": 15.97, "ce": 0.05902281095586659, "rank": 0.0} +{"step": 364, "steps": 500, "lr": 2e-05, "grad_norm": 1.451307773590088, "loss_scale": 4.0, "retries": 0, "seconds": 621.0, "max_memory_gb": 15.97, "ce": 0.0622781705824309, "rank": 0.07407373934984207} +{"step": 365, "steps": 500, "lr": 2e-05, "grad_norm": 0.6763289570808411, "loss_scale": 4.0, "retries": 0, "seconds": 622.7, "max_memory_gb": 15.97, "ce": 0.10966243661823682, "rank": 0.0} +{"step": 366, "steps": 500, "lr": 2e-05, "grad_norm": 1.9001729488372803, "loss_scale": 4.0, "retries": 0, "seconds": 624.4, "max_memory_gb": 15.97, "ce": 0.13454940255724068, "rank": 0.0} +{"step": 367, "steps": 500, "lr": 2e-05, "grad_norm": 0.5225584506988525, "loss_scale": 4.0, "retries": 0, "seconds": 626.1, "max_memory_gb": 15.97, "ce": 0.10005278316202748, "rank": 0.0} +{"step": 368, "steps": 500, "lr": 2e-05, "grad_norm": 0.6916511058807373, "loss_scale": 4.0, "retries": 0, "seconds": 627.7, "max_memory_gb": 15.97, "ce": 0.021951899190753466, "rank": 0.0} +{"step": 369, "steps": 500, "lr": 2e-05, "grad_norm": 0.38229233026504517, "loss_scale": 4.0, "retries": 0, "seconds": 629.4, "max_memory_gb": 15.97, "ce": 0.04222217800634098, "rank": 0.11222072690725327} +{"step": 370, "steps": 500, "lr": 2e-05, "grad_norm": 2.394906997680664, "loss_scale": 4.0, "retries": 0, "seconds": 631.0, "max_memory_gb": 15.97, "ce": 0.3713889201408165, "rank": 0.0} +{"step": 371, "steps": 500, "lr": 2e-05, "grad_norm": 1.3651514053344727, "loss_scale": 4.0, "retries": 0, "seconds": 632.7, "max_memory_gb": 15.97, "ce": 0.05432157796531101, "rank": 0.0} +{"step": 372, "steps": 500, "lr": 2e-05, "grad_norm": 0.3003334701061249, "loss_scale": 4.0, "retries": 0, "seconds": 634.3, "max_memory_gb": 15.97, "ce": 0.027796988482805318, "rank": 0.0} +{"step": 373, "steps": 500, "lr": 2e-05, "grad_norm": 0.5652337670326233, "loss_scale": 4.0, "retries": 0, "seconds": 636.0, "max_memory_gb": 15.97, "ce": 0.09257882392557804, "rank": 0.0} +{"step": 374, "steps": 500, "lr": 2e-05, "grad_norm": 0.6252080202102661, "loss_scale": 4.0, "retries": 0, "seconds": 637.7, "max_memory_gb": 15.97, "ce": 0.12578728570588282, "rank": 0.0} +{"step": 375, "steps": 500, "lr": 2e-05, "grad_norm": 0.39739683270454407, "loss_scale": 4.0, "retries": 0, "seconds": 639.2, "max_memory_gb": 15.97, "ce": 0.15395514083866146, "rank": 0.0} +{"step": 376, "steps": 500, "lr": 2e-05, "grad_norm": 0.5400328040122986, "loss_scale": 4.0, "retries": 0, "seconds": 640.8, "max_memory_gb": 15.97, "ce": 0.059474044350281474, "rank": 0.0} +{"step": 377, "steps": 500, "lr": 2e-05, "grad_norm": 0.51706463098526, "loss_scale": 4.0, "retries": 0, "seconds": 642.4, "max_memory_gb": 15.97, "ce": 0.09105409745097859, "rank": 0.0} +{"step": 378, "steps": 500, "lr": 2e-05, "grad_norm": 1.0843415260314941, "loss_scale": 4.0, "retries": 0, "seconds": 644.1, "max_memory_gb": 15.97, "ce": 0.08365566418069648, "rank": 0.0} +{"step": 379, "steps": 500, "lr": 2e-05, "grad_norm": 0.35980579257011414, "loss_scale": 4.0, "retries": 0, "seconds": 645.8, "max_memory_gb": 15.97, "ce": 0.10297915066621499, "rank": 0.0} +{"step": 380, "steps": 500, "lr": 2e-05, "grad_norm": 1.2426971197128296, "loss_scale": 4.0, "retries": 0, "seconds": 647.4, "max_memory_gb": 15.97, "ce": 0.06248813755246374, "rank": 0.0} +{"step": 381, "steps": 500, "lr": 2e-05, "grad_norm": 0.7912470698356628, "loss_scale": 4.0, "retries": 0, "seconds": 649.1, "max_memory_gb": 15.97, "ce": 0.2467686568015779, "rank": 0.0} +{"step": 382, "steps": 500, "lr": 2e-05, "grad_norm": 1.1025137901306152, "loss_scale": 4.0, "retries": 0, "seconds": 650.7, "max_memory_gb": 15.97, "ce": 0.2688007210745127, "rank": 0.0} +{"step": 383, "steps": 500, "lr": 2e-05, "grad_norm": 0.2621373236179352, "loss_scale": 4.0, "retries": 0, "seconds": 652.4, "max_memory_gb": 15.97, "ce": 0.0720755948113947, "rank": 0.0} +{"step": 384, "steps": 500, "lr": 2e-05, "grad_norm": 0.4384917616844177, "loss_scale": 4.0, "retries": 0, "seconds": 654.0, "max_memory_gb": 15.97, "ce": 0.027352494864317123, "rank": 0.0} +{"step": 385, "steps": 500, "lr": 2e-05, "grad_norm": 0.5650014281272888, "loss_scale": 4.0, "retries": 0, "seconds": 655.7, "max_memory_gb": 15.97, "ce": 0.17106326694556628, "rank": 0.0} +{"step": 386, "steps": 500, "lr": 2e-05, "grad_norm": 0.4915285110473633, "loss_scale": 4.0, "retries": 0, "seconds": 657.2, "max_memory_gb": 15.97, "ce": 0.057163077748555224, "rank": 0.0} +{"step": 387, "steps": 500, "lr": 2e-05, "grad_norm": 0.3391872048377991, "loss_scale": 4.0, "retries": 0, "seconds": 658.8, "max_memory_gb": 15.97, "ce": 0.051142767053534044, "rank": 0.0} +{"step": 388, "steps": 500, "lr": 2e-05, "grad_norm": 0.8413510918617249, "loss_scale": 4.0, "retries": 0, "seconds": 660.5, "max_memory_gb": 15.97, "ce": 0.05478223616228206, "rank": 0.0} +{"step": 389, "steps": 500, "lr": 2e-05, "grad_norm": 0.17180167138576508, "loss_scale": 4.0, "retries": 0, "seconds": 662.3, "max_memory_gb": 15.97, "ce": 0.006642046469096385, "rank": 0.0} +{"step": 390, "steps": 500, "lr": 2e-05, "grad_norm": 2.1708738803863525, "loss_scale": 4.0, "retries": 0, "seconds": 664.0, "max_memory_gb": 15.97, "ce": 0.06125239508764935, "rank": 0.0} +{"step": 391, "steps": 500, "lr": 2e-05, "grad_norm": 2.9544801712036133, "loss_scale": 4.0, "retries": 0, "seconds": 665.7, "max_memory_gb": 15.97, "ce": 0.2711659706410501, "rank": 0.0} +{"step": 392, "steps": 500, "lr": 2e-05, "grad_norm": 0.13671742379665375, "loss_scale": 4.0, "retries": 0, "seconds": 667.4, "max_memory_gb": 15.97, "ce": 0.002806858483381802, "rank": 0.06743408739566803} +{"step": 393, "steps": 500, "lr": 2e-05, "grad_norm": 1.5158791542053223, "loss_scale": 4.0, "retries": 0, "seconds": 668.9, "max_memory_gb": 15.97, "ce": 0.17871041715716274, "rank": 0.0} +{"step": 394, "steps": 500, "lr": 2e-05, "grad_norm": 0.9213796257972717, "loss_scale": 4.0, "retries": 0, "seconds": 670.5, "max_memory_gb": 15.97, "ce": 0.10279258481205034, "rank": 0.0} +{"step": 395, "steps": 500, "lr": 2e-05, "grad_norm": 0.47259560227394104, "loss_scale": 4.0, "retries": 0, "seconds": 672.1, "max_memory_gb": 15.97, "ce": 0.023782584735272394, "rank": 0.0} +{"step": 396, "steps": 500, "lr": 2e-05, "grad_norm": 1.223973274230957, "loss_scale": 4.0, "retries": 0, "seconds": 673.6, "max_memory_gb": 15.97, "ce": 0.03489736951632949, "rank": 0.0} +{"step": 397, "steps": 500, "lr": 2e-05, "grad_norm": 0.4176880717277527, "loss_scale": 4.0, "retries": 0, "seconds": 675.3, "max_memory_gb": 15.97, "ce": 0.10481115865695756, "rank": 0.0} +{"step": 398, "steps": 500, "lr": 2e-05, "grad_norm": 0.3813675045967102, "loss_scale": 4.0, "retries": 0, "seconds": 676.9, "max_memory_gb": 15.97, "ce": 0.058010182762700424, "rank": 0.0} +{"step": 399, "steps": 500, "lr": 2e-05, "grad_norm": 1.1864259243011475, "loss_scale": 4.0, "retries": 0, "seconds": 678.5, "max_memory_gb": 15.97, "ce": 0.15217977149950457, "rank": 0.0} +{"step": 400, "steps": 500, "lr": 2e-05, "grad_norm": 1.0613902807235718, "loss_scale": 4.0, "retries": 0, "seconds": 680.2, "max_memory_gb": 15.97, "ce": 0.02863753488054499, "rank": 0.0} +{"step": 401, "steps": 500, "lr": 2e-05, "grad_norm": 1.8588440418243408, "loss_scale": 4.0, "retries": 0, "seconds": 681.9, "max_memory_gb": 15.97, "ce": 0.19668248420202872, "rank": 0.0} +{"step": 402, "steps": 500, "lr": 2e-05, "grad_norm": 0.3465002775192261, "loss_scale": 4.0, "retries": 0, "seconds": 683.6, "max_memory_gb": 15.97, "ce": 0.09459597404384112, "rank": 0.09548970311880112} +{"step": 403, "steps": 500, "lr": 2e-05, "grad_norm": 0.09975554794073105, "loss_scale": 4.0, "retries": 0, "seconds": 685.2, "max_memory_gb": 15.97, "ce": 0.006144109556771582, "rank": 0.0} +{"step": 404, "steps": 500, "lr": 2e-05, "grad_norm": 0.48277634382247925, "loss_scale": 4.0, "retries": 0, "seconds": 686.8, "max_memory_gb": 15.97, "ce": 0.11497014562701224, "rank": 0.0} +{"step": 405, "steps": 500, "lr": 2e-05, "grad_norm": 0.4114937484264374, "loss_scale": 4.0, "retries": 0, "seconds": 688.5, "max_memory_gb": 15.97, "ce": 0.037007292772614164, "rank": 0.0} +{"step": 406, "steps": 500, "lr": 2e-05, "grad_norm": 0.6149478554725647, "loss_scale": 4.0, "retries": 0, "seconds": 690.2, "max_memory_gb": 15.97, "ce": 0.12056918042981124, "rank": 0.0} +{"step": 407, "steps": 500, "lr": 2e-05, "grad_norm": 1.7022582292556763, "loss_scale": 4.0, "retries": 0, "seconds": 691.8, "max_memory_gb": 15.97, "ce": 0.6687338616029592, "rank": 0.09280842542648315} +{"step": 408, "steps": 500, "lr": 2e-05, "grad_norm": 1.5335432291030884, "loss_scale": 4.0, "retries": 0, "seconds": 693.6, "max_memory_gb": 15.97, "ce": 0.32432946100198023, "rank": 0.05220872163772583} +{"step": 409, "steps": 500, "lr": 2e-05, "grad_norm": 0.3267282545566559, "loss_scale": 4.0, "retries": 0, "seconds": 695.2, "max_memory_gb": 15.97, "ce": 0.12486270256340504, "rank": 0.0} +{"step": 410, "steps": 500, "lr": 2e-05, "grad_norm": 1.3734492063522339, "loss_scale": 4.0, "retries": 0, "seconds": 696.9, "max_memory_gb": 15.97, "ce": 0.05006051709642634, "rank": 0.0} +{"step": 411, "steps": 500, "lr": 2e-05, "grad_norm": 5.909807205200195, "loss_scale": 4.0, "retries": 0, "seconds": 698.6, "max_memory_gb": 15.97, "ce": 0.2416662509949674, "rank": 0.0} +{"step": 412, "steps": 500, "lr": 2e-05, "grad_norm": 0.29736775159835815, "loss_scale": 4.0, "retries": 0, "seconds": 700.2, "max_memory_gb": 15.97, "ce": 0.13604049566674803, "rank": 0.12479938566684723} +{"step": 413, "steps": 500, "lr": 2e-05, "grad_norm": 0.5238308310508728, "loss_scale": 4.0, "retries": 0, "seconds": 701.9, "max_memory_gb": 15.97, "ce": 0.04246207795040391, "rank": 0.0} +{"step": 414, "steps": 500, "lr": 2e-05, "grad_norm": 1.3103562593460083, "loss_scale": 4.0, "retries": 0, "seconds": 703.6, "max_memory_gb": 15.97, "ce": 0.20907893284856982, "rank": 0.0} +{"step": 415, "steps": 500, "lr": 2e-05, "grad_norm": 2.77193546295166, "loss_scale": 4.0, "retries": 0, "seconds": 705.2, "max_memory_gb": 15.97, "ce": 0.35039180331477837, "rank": 0.0} +{"step": 416, "steps": 500, "lr": 2e-05, "grad_norm": 0.40025267004966736, "loss_scale": 4.0, "retries": 0, "seconds": 706.9, "max_memory_gb": 15.97, "ce": 0.019131008153635776, "rank": 0.0} +{"step": 417, "steps": 500, "lr": 2e-05, "grad_norm": 1.0699470043182373, "loss_scale": 4.0, "retries": 0, "seconds": 708.6, "max_memory_gb": 15.97, "ce": 0.33801461752227624, "rank": 0.12758122384548187} +{"step": 418, "steps": 500, "lr": 2e-05, "grad_norm": 0.3083900213241577, "loss_scale": 4.0, "retries": 0, "seconds": 710.4, "max_memory_gb": 15.97, "ce": 0.04259202183857269, "rank": 0.0} +{"step": 419, "steps": 500, "lr": 2e-05, "grad_norm": 0.30822107195854187, "loss_scale": 4.0, "retries": 0, "seconds": 712.2, "max_memory_gb": 15.97, "ce": 0.06103093516139779, "rank": 0.0} +{"step": 420, "steps": 500, "lr": 2e-05, "grad_norm": 1.9756428003311157, "loss_scale": 4.0, "retries": 0, "seconds": 714.1, "max_memory_gb": 15.97, "ce": 0.2800266056892724, "rank": 0.130604550242424} +{"step": 421, "steps": 500, "lr": 2e-05, "grad_norm": 0.31638669967651367, "loss_scale": 4.0, "retries": 0, "seconds": 715.9, "max_memory_gb": 15.97, "ce": 0.031499566066486295, "rank": 0.10946999490261078} +{"step": 422, "steps": 500, "lr": 2e-05, "grad_norm": 1.7142189741134644, "loss_scale": 4.0, "retries": 0, "seconds": 717.8, "max_memory_gb": 15.97, "ce": 0.22062860258483852, "rank": 0.0} +{"step": 423, "steps": 500, "lr": 2e-05, "grad_norm": 4.418954849243164, "loss_scale": 4.0, "retries": 0, "seconds": 719.7, "max_memory_gb": 15.97, "ce": 0.414777417860023, "rank": 0.0} +{"step": 424, "steps": 500, "lr": 2e-05, "grad_norm": 0.648068904876709, "loss_scale": 4.0, "retries": 0, "seconds": 721.5, "max_memory_gb": 15.97, "ce": 0.05856898005276889, "rank": 0.0} +{"step": 425, "steps": 500, "lr": 2e-05, "grad_norm": 0.8143153190612793, "loss_scale": 4.0, "retries": 0, "seconds": 723.2, "max_memory_gb": 15.97, "ce": 0.08042688858222391, "rank": 0.0} +{"step": 426, "steps": 500, "lr": 2e-05, "grad_norm": 1.0916856527328491, "loss_scale": 4.0, "retries": 0, "seconds": 724.9, "max_memory_gb": 15.97, "ce": 0.18848598497061175, "rank": 0.0} +{"step": 427, "steps": 500, "lr": 2e-05, "grad_norm": 0.2527453601360321, "loss_scale": 4.0, "retries": 0, "seconds": 726.5, "max_memory_gb": 15.97, "ce": 0.05309699427016312, "rank": 0.0} +{"step": 428, "steps": 500, "lr": 2e-05, "grad_norm": 1.3752715587615967, "loss_scale": 4.0, "retries": 0, "seconds": 728.3, "max_memory_gb": 15.97, "ce": 0.03763972394517623, "rank": 0.0} +{"step": 429, "steps": 500, "lr": 2e-05, "grad_norm": 0.30704009532928467, "loss_scale": 4.0, "retries": 0, "seconds": 730.0, "max_memory_gb": 15.97, "ce": 0.014585127766622463, "rank": 0.0} +{"step": 430, "steps": 500, "lr": 2e-05, "grad_norm": 0.6167378425598145, "loss_scale": 4.0, "retries": 0, "seconds": 731.6, "max_memory_gb": 15.97, "ce": 0.0845042927776376, "rank": 0.0} +{"step": 431, "steps": 500, "lr": 2e-05, "grad_norm": 7.205959796905518, "loss_scale": 4.0, "retries": 0, "seconds": 733.3, "max_memory_gb": 15.97, "ce": 0.3350212727291364, "rank": 0.0} +{"step": 432, "steps": 500, "lr": 2e-05, "grad_norm": 0.5603015422821045, "loss_scale": 4.0, "retries": 0, "seconds": 735.1, "max_memory_gb": 15.97, "ce": 0.056542628139141016, "rank": 0.0} +{"step": 433, "steps": 500, "lr": 2e-05, "grad_norm": 0.5307062864303589, "loss_scale": 4.0, "retries": 0, "seconds": 736.7, "max_memory_gb": 15.97, "ce": 0.06041403852213989, "rank": 0.0} +{"step": 434, "steps": 500, "lr": 2e-05, "grad_norm": 0.08416435867547989, "loss_scale": 4.0, "retries": 0, "seconds": 738.4, "max_memory_gb": 15.97, "ce": 0.005377546860472648, "rank": 0.0} +{"step": 435, "steps": 500, "lr": 2e-05, "grad_norm": 1.239395022392273, "loss_scale": 4.0, "retries": 0, "seconds": 740.0, "max_memory_gb": 15.97, "ce": 0.1549265822623056, "rank": 0.0} +{"step": 436, "steps": 500, "lr": 2e-05, "grad_norm": 1.158418893814087, "loss_scale": 4.0, "retries": 0, "seconds": 741.7, "max_memory_gb": 15.97, "ce": 0.12487518185662339, "rank": 0.0} +{"step": 437, "steps": 500, "lr": 2e-05, "grad_norm": 4.8477864265441895, "loss_scale": 4.0, "retries": 0, "seconds": 743.4, "max_memory_gb": 15.97, "ce": 0.19282669056792656, "rank": 0.11521273851394653} +{"step": 438, "steps": 500, "lr": 2e-05, "grad_norm": 0.6232281923294067, "loss_scale": 4.0, "retries": 0, "seconds": 745.1, "max_memory_gb": 15.97, "ce": 0.0568855196770528, "rank": 0.0} +{"step": 439, "steps": 500, "lr": 2e-05, "grad_norm": 1.0757039785385132, "loss_scale": 4.0, "retries": 0, "seconds": 746.7, "max_memory_gb": 15.97, "ce": 0.10625843310026539, "rank": 0.0} +{"step": 440, "steps": 500, "lr": 2e-05, "grad_norm": 0.39968717098236084, "loss_scale": 4.0, "retries": 0, "seconds": 748.4, "max_memory_gb": 15.97, "ce": 0.0363463450994459, "rank": 0.0} +{"step": 441, "steps": 500, "lr": 2e-05, "grad_norm": 2.0115294456481934, "loss_scale": 4.0, "retries": 0, "seconds": 750.0, "max_memory_gb": 15.97, "ce": 0.19006491729396657, "rank": 0.0} +{"step": 442, "steps": 500, "lr": 2e-05, "grad_norm": 0.571366012096405, "loss_scale": 4.0, "retries": 0, "seconds": 751.7, "max_memory_gb": 15.97, "ce": 0.052286519701283396, "rank": 0.0} +{"step": 443, "steps": 500, "lr": 2e-05, "grad_norm": 0.5579499006271362, "loss_scale": 4.0, "retries": 0, "seconds": 753.4, "max_memory_gb": 15.97, "ce": 0.049633031063422095, "rank": 0.0} +{"step": 444, "steps": 500, "lr": 2e-05, "grad_norm": 0.448842853307724, "loss_scale": 4.0, "retries": 0, "seconds": 755.0, "max_memory_gb": 15.97, "ce": 0.060094218549238576, "rank": 0.0} +{"step": 445, "steps": 500, "lr": 2e-05, "grad_norm": 2.1269445419311523, "loss_scale": 4.0, "retries": 0, "seconds": 756.7, "max_memory_gb": 15.97, "ce": 0.38824680874313344, "rank": 0.061687640845775604} +{"step": 446, "steps": 500, "lr": 2e-05, "grad_norm": 6.921499729156494, "loss_scale": 4.0, "retries": 0, "seconds": 758.3, "max_memory_gb": 15.97, "ce": 0.14041459921281785, "rank": 0.24112485349178314} +{"step": 447, "steps": 500, "lr": 2e-05, "grad_norm": 0.4897305965423584, "loss_scale": 4.0, "retries": 0, "seconds": 760.0, "max_memory_gb": 15.97, "ce": 0.134588811220965, "rank": 0.0} +{"step": 448, "steps": 500, "lr": 2e-05, "grad_norm": 1.8183066844940186, "loss_scale": 4.0, "retries": 0, "seconds": 761.6, "max_memory_gb": 15.97, "ce": 0.1591154515526796, "rank": 0.0} +{"step": 449, "steps": 500, "lr": 2e-05, "grad_norm": 0.8490037322044373, "loss_scale": 4.0, "retries": 0, "seconds": 763.3, "max_memory_gb": 15.97, "ce": 0.015272328768332954, "rank": 0.0} +{"step": 450, "steps": 500, "lr": 2e-05, "grad_norm": 0.9260918498039246, "loss_scale": 4.0, "retries": 0, "seconds": 765.0, "max_memory_gb": 15.97, "ce": 0.025034887466972577, "rank": 0.0} +{"step": 451, "steps": 500, "lr": 2e-05, "grad_norm": 0.5471746921539307, "loss_scale": 4.0, "retries": 0, "seconds": 766.7, "max_memory_gb": 15.97, "ce": 0.025109794491072535, "rank": 0.0} +{"step": 452, "steps": 500, "lr": 2e-05, "grad_norm": 1.3505334854125977, "loss_scale": 4.0, "retries": 0, "seconds": 768.3, "max_memory_gb": 15.97, "ce": 0.1530485722269077, "rank": 0.0} +{"step": 453, "steps": 500, "lr": 2e-05, "grad_norm": 1.37962806224823, "loss_scale": 4.0, "retries": 0, "seconds": 770.0, "max_memory_gb": 15.97, "ce": 0.0393801518903274, "rank": 0.0} +{"step": 454, "steps": 500, "lr": 2e-05, "grad_norm": 2.145359516143799, "loss_scale": 4.0, "retries": 0, "seconds": 771.7, "max_memory_gb": 15.97, "ce": 0.14025784132763874, "rank": 0.0} +{"step": 455, "steps": 500, "lr": 2e-05, "grad_norm": 2.0662503242492676, "loss_scale": 4.0, "retries": 0, "seconds": 773.4, "max_memory_gb": 15.97, "ce": 0.10372325343860211, "rank": 0.0} +{"step": 456, "steps": 500, "lr": 2e-05, "grad_norm": 0.21518278121948242, "loss_scale": 4.0, "retries": 0, "seconds": 775.0, "max_memory_gb": 15.97, "ce": 0.011128515983273246, "rank": 0.0} +{"step": 457, "steps": 500, "lr": 2e-05, "grad_norm": 0.6887024641036987, "loss_scale": 4.0, "retries": 0, "seconds": 776.7, "max_memory_gb": 15.97, "ce": 0.031242376746376976, "rank": 0.03115680068731308} +{"step": 458, "steps": 500, "lr": 2e-05, "grad_norm": 1.3209633827209473, "loss_scale": 4.0, "retries": 0, "seconds": 778.5, "max_memory_gb": 15.97, "ce": 0.16164232012488355, "rank": 0.0} +{"step": 459, "steps": 500, "lr": 2e-05, "grad_norm": 2.3655128479003906, "loss_scale": 4.0, "retries": 0, "seconds": 780.5, "max_memory_gb": 15.97, "ce": 0.13262370689153613, "rank": 0.0} +{"step": 460, "steps": 500, "lr": 2e-05, "grad_norm": 0.2057892382144928, "loss_scale": 4.0, "retries": 0, "seconds": 782.4, "max_memory_gb": 15.97, "ce": 0.01603190859168535, "rank": 0.0} +{"step": 461, "steps": 500, "lr": 2e-05, "grad_norm": 1.0158895254135132, "loss_scale": 4.0, "retries": 0, "seconds": 784.2, "max_memory_gb": 15.97, "ce": 0.24681616625639435, "rank": 0.0} +{"step": 462, "steps": 500, "lr": 2e-05, "grad_norm": 0.5620952844619751, "loss_scale": 4.0, "retries": 0, "seconds": 786.0, "max_memory_gb": 15.97, "ce": 0.026520658460867708, "rank": 0.0} +{"step": 463, "steps": 500, "lr": 2e-05, "grad_norm": 0.368592768907547, "loss_scale": 4.0, "retries": 0, "seconds": 787.8, "max_memory_gb": 15.97, "ce": 0.015833818387363863, "rank": 0.0} +{"step": 464, "steps": 500, "lr": 2e-05, "grad_norm": 0.2386801689863205, "loss_scale": 4.0, "retries": 0, "seconds": 789.7, "max_memory_gb": 15.97, "ce": 0.008071330474194838, "rank": 0.0} +{"step": 465, "steps": 500, "lr": 2e-05, "grad_norm": 1.989155888557434, "loss_scale": 4.0, "retries": 0, "seconds": 791.5, "max_memory_gb": 15.97, "ce": 0.22519498736983223, "rank": 0.0} +{"step": 466, "steps": 500, "lr": 2e-05, "grad_norm": 2.0305142402648926, "loss_scale": 4.0, "retries": 0, "seconds": 793.4, "max_memory_gb": 15.97, "ce": 0.07754540741098026, "rank": 0.0} +{"step": 467, "steps": 500, "lr": 2e-05, "grad_norm": 0.6666870713233948, "loss_scale": 4.0, "retries": 0, "seconds": 795.2, "max_memory_gb": 15.97, "ce": 0.031712407608210924, "rank": 0.0} +{"step": 468, "steps": 500, "lr": 2e-05, "grad_norm": 0.24881845712661743, "loss_scale": 4.0, "retries": 0, "seconds": 797.1, "max_memory_gb": 15.97, "ce": 0.010966245550207532, "rank": 0.0} +{"step": 469, "steps": 500, "lr": 2e-05, "grad_norm": 3.9729435443878174, "loss_scale": 4.0, "retries": 0, "seconds": 799.0, "max_memory_gb": 15.97, "ce": 0.21142291459500484, "rank": 0.0} +{"step": 470, "steps": 500, "lr": 2e-05, "grad_norm": 1.6864290237426758, "loss_scale": 4.0, "retries": 0, "seconds": 800.8, "max_memory_gb": 15.97, "ce": 0.07601076964692766, "rank": 0.0} +{"step": 471, "steps": 500, "lr": 2e-05, "grad_norm": 1.4188501834869385, "loss_scale": 4.0, "retries": 0, "seconds": 802.7, "max_memory_gb": 15.97, "ce": 0.16515340919886512, "rank": 0.0} +{"step": 472, "steps": 500, "lr": 2e-05, "grad_norm": 0.5611962676048279, "loss_scale": 4.0, "retries": 0, "seconds": 804.5, "max_memory_gb": 15.97, "ce": 0.03703426314268654, "rank": 0.0} +{"step": 473, "steps": 500, "lr": 2e-05, "grad_norm": 0.5474820733070374, "loss_scale": 4.0, "retries": 0, "seconds": 806.2, "max_memory_gb": 15.97, "ce": 0.015873899956204696, "rank": 0.12918207049369812} +{"step": 474, "steps": 500, "lr": 2e-05, "grad_norm": 0.6269225478172302, "loss_scale": 4.0, "retries": 0, "seconds": 807.9, "max_memory_gb": 15.97, "ce": 0.09532818406660226, "rank": 0.0} +{"step": 475, "steps": 500, "lr": 2e-05, "grad_norm": 1.7073307037353516, "loss_scale": 4.0, "retries": 0, "seconds": 809.6, "max_memory_gb": 15.97, "ce": 0.20003245383941248, "rank": 0.0} +{"step": 476, "steps": 500, "lr": 2e-05, "grad_norm": 0.23905174434185028, "loss_scale": 4.0, "retries": 0, "seconds": 811.2, "max_memory_gb": 15.97, "ce": 0.008331814612574817, "rank": 0.0} +{"step": 477, "steps": 500, "lr": 2e-05, "grad_norm": 2.4490292072296143, "loss_scale": 4.0, "retries": 0, "seconds": 812.9, "max_memory_gb": 15.97, "ce": 0.08437340462205611, "rank": 0.0} +{"step": 478, "steps": 500, "lr": 2e-05, "grad_norm": 0.5656754374504089, "loss_scale": 4.0, "retries": 0, "seconds": 814.7, "max_memory_gb": 15.97, "ce": 0.08167036667600769, "rank": 0.0} +{"step": 479, "steps": 500, "lr": 2e-05, "grad_norm": 4.906678199768066, "loss_scale": 4.0, "retries": 0, "seconds": 816.6, "max_memory_gb": 15.97, "ce": 0.3546010775062314, "rank": 0.03050088882446289} +{"step": 480, "steps": 500, "lr": 2e-05, "grad_norm": 0.758303165435791, "loss_scale": 4.0, "retries": 0, "seconds": 818.5, "max_memory_gb": 15.97, "ce": 0.13219996168572834, "rank": 0.0} +{"step": 481, "steps": 500, "lr": 2e-05, "grad_norm": 0.25821587443351746, "loss_scale": 4.0, "retries": 0, "seconds": 820.3, "max_memory_gb": 15.97, "ce": 0.01899185864112951, "rank": 0.0} +{"step": 482, "steps": 500, "lr": 2e-05, "grad_norm": 0.6761943101882935, "loss_scale": 4.0, "retries": 0, "seconds": 822.2, "max_memory_gb": 15.97, "ce": 0.070563455707088, "rank": 0.0} +{"step": 483, "steps": 500, "lr": 2e-05, "grad_norm": 2.6488571166992188, "loss_scale": 4.0, "retries": 0, "seconds": 824.1, "max_memory_gb": 15.97, "ce": 0.379923739125843, "rank": 0.0} +{"step": 484, "steps": 500, "lr": 2e-05, "grad_norm": 1.7335762977600098, "loss_scale": 4.0, "retries": 0, "seconds": 825.9, "max_memory_gb": 15.97, "ce": 0.4771190529518208, "rank": 0.0} +{"step": 485, "steps": 500, "lr": 2e-05, "grad_norm": 1.1553620100021362, "loss_scale": 4.0, "retries": 0, "seconds": 827.7, "max_memory_gb": 15.97, "ce": 0.07445207412001764, "rank": 0.0} +{"step": 486, "steps": 500, "lr": 2e-05, "grad_norm": 1.425718069076538, "loss_scale": 4.0, "retries": 0, "seconds": 829.3, "max_memory_gb": 15.97, "ce": 0.1886235445517741, "rank": 0.0} +{"step": 487, "steps": 500, "lr": 2e-05, "grad_norm": 4.624532222747803, "loss_scale": 4.0, "retries": 0, "seconds": 830.9, "max_memory_gb": 15.97, "ce": 0.12022047447317163, "rank": 0.0} +{"step": 488, "steps": 500, "lr": 2e-05, "grad_norm": 0.3950486481189728, "loss_scale": 4.0, "retries": 0, "seconds": 832.5, "max_memory_gb": 15.97, "ce": 0.0757453177438947, "rank": 0.0} +{"step": 489, "steps": 500, "lr": 2e-05, "grad_norm": 3.1665608882904053, "loss_scale": 4.0, "retries": 0, "seconds": 834.1, "max_memory_gb": 15.97, "ce": 0.18132776914535498, "rank": 0.0} +{"step": 490, "steps": 500, "lr": 2e-05, "grad_norm": 0.44466352462768555, "loss_scale": 4.0, "retries": 0, "seconds": 835.7, "max_memory_gb": 15.97, "ce": 0.03901745144503366, "rank": 0.0} +{"step": 491, "steps": 500, "lr": 2e-05, "grad_norm": 0.34754616022109985, "loss_scale": 4.0, "retries": 0, "seconds": 837.4, "max_memory_gb": 15.97, "ce": 0.11085473547427682, "rank": 0.0} +{"step": 492, "steps": 500, "lr": 2e-05, "grad_norm": 0.17865422368049622, "loss_scale": 4.0, "retries": 0, "seconds": 839.1, "max_memory_gb": 15.97, "ce": 0.007425154975862824, "rank": 0.0} +{"step": 493, "steps": 500, "lr": 2e-05, "grad_norm": 0.7625251412391663, "loss_scale": 4.0, "retries": 0, "seconds": 840.8, "max_memory_gb": 15.97, "ce": 0.052828669437076314, "rank": 0.0} +{"step": 494, "steps": 500, "lr": 2e-05, "grad_norm": 1.5818583965301514, "loss_scale": 4.0, "retries": 0, "seconds": 842.4, "max_memory_gb": 15.97, "ce": 0.10196522489604831, "rank": 0.0} +{"step": 495, "steps": 500, "lr": 2e-05, "grad_norm": 0.716194748878479, "loss_scale": 4.0, "retries": 0, "seconds": 844.0, "max_memory_gb": 15.97, "ce": 0.04443023502426513, "rank": 0.0} +{"step": 496, "steps": 500, "lr": 2e-05, "grad_norm": 0.5035225749015808, "loss_scale": 4.0, "retries": 0, "seconds": 845.7, "max_memory_gb": 15.97, "ce": 0.01211180979044002, "rank": 0.0} +{"step": 497, "steps": 500, "lr": 2e-05, "grad_norm": 1.5239559412002563, "loss_scale": 4.0, "retries": 0, "seconds": 847.3, "max_memory_gb": 15.97, "ce": 0.12369466133895912, "rank": 0.0} +{"step": 498, "steps": 500, "lr": 2e-05, "grad_norm": 0.8662014603614807, "loss_scale": 4.0, "retries": 0, "seconds": 848.9, "max_memory_gb": 15.97, "ce": 0.12098025064915419, "rank": 0.0} +{"step": 499, "steps": 500, "lr": 2e-05, "grad_norm": 1.2565878629684448, "loss_scale": 4.0, "retries": 0, "seconds": 850.6, "max_memory_gb": 15.97, "ce": 0.12674159874222823, "rank": 0.0} +{"step": 500, "steps": 500, "lr": 2e-05, "grad_norm": 1.5525610446929932, "loss_scale": 4.0, "retries": 0, "seconds": 852.3, "max_memory_gb": 15.97, "ce": 0.1998795309154957, "rank": 0.0} diff --git a/runs/single_llava/removal/L22_s13/adapter/README.md b/runs/single_llava/removal/L22_s13/adapter/README.md new file mode 100644 index 0000000000000000000000000000000000000000..ae587920ca55972d94e38f783837f71be14f1dda --- /dev/null +++ b/runs/single_llava/removal/L22_s13/adapter/README.md @@ -0,0 +1,207 @@ +--- +base_model: /data/OPSD/assets/models/llava-1.5-7b-hf +library_name: peft +pipeline_tag: text-generation +tags: +- base_model:adapter:/data/OPSD/assets/models/llava-1.5-7b-hf +- lora +- transformers +--- + +# Model Card for Model ID + + + + + +## Model Details + +### Model Description + + + + + +- **Developed by:** [More Information Needed] +- **Funded by [optional]:** [More Information Needed] +- **Shared by [optional]:** [More Information Needed] +- **Model type:** [More Information Needed] +- **Language(s) (NLP):** [More Information Needed] +- **License:** [More Information Needed] +- **Finetuned from model [optional]:** [More Information Needed] + +### Model Sources [optional] + + + +- **Repository:** [More Information Needed] +- **Paper [optional]:** [More Information Needed] +- **Demo [optional]:** [More Information Needed] + +## Uses + + + +### Direct Use + + + +[More Information Needed] + +### Downstream Use [optional] + + + +[More Information Needed] + +### Out-of-Scope Use + + + +[More Information Needed] + +## Bias, Risks, and Limitations + + + +[More Information Needed] + +### Recommendations + + + +Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. More information needed for further recommendations. + +## How to Get Started with the Model + +Use the code below to get started with the model. + +[More Information Needed] + +## Training Details + +### Training Data + + + +[More Information Needed] + +### Training Procedure + + + +#### Preprocessing [optional] + +[More Information Needed] + + +#### Training Hyperparameters + +- **Training regime:** [More Information Needed] + +#### Speeds, Sizes, Times [optional] + + + +[More Information Needed] + +## Evaluation + + + +### Testing Data, Factors & Metrics + +#### Testing Data + + + +[More Information Needed] + +#### Factors + + + +[More Information Needed] + +#### Metrics + + + +[More Information Needed] + +### Results + +[More Information Needed] + +#### Summary + + + +## Model Examination [optional] + + + +[More Information Needed] + +## Environmental Impact + + + +Carbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700). + +- **Hardware Type:** [More Information Needed] +- **Hours used:** [More Information Needed] +- **Cloud Provider:** [More Information Needed] +- **Compute Region:** [More Information Needed] +- **Carbon Emitted:** [More Information Needed] + +## Technical Specifications [optional] + +### Model Architecture and Objective + +[More Information Needed] + +### Compute Infrastructure + +[More Information Needed] + +#### Hardware + +[More Information Needed] + +#### Software + +[More Information Needed] + +## Citation [optional] + + + +**BibTeX:** + +[More Information Needed] + +**APA:** + +[More Information Needed] + +## Glossary [optional] + + + +[More Information Needed] + +## More Information [optional] + +[More Information Needed] + +## Model Card Authors [optional] + +[More Information Needed] + +## Model Card Contact + +[More Information Needed] +### Framework versions + +- PEFT 0.20.0 \ No newline at end of file diff --git a/runs/single_llava/removal/L22_s13/adapter/adapter_config.json b/runs/single_llava/removal/L22_s13/adapter/adapter_config.json new file mode 100644 index 0000000000000000000000000000000000000000..78923e9517b37ce5d18be38e0f83308f98d320b9 --- /dev/null +++ b/runs/single_llava/removal/L22_s13/adapter/adapter_config.json @@ -0,0 +1,50 @@ +{ + "alora_invocation_tokens": null, + "alpha_pattern": {}, + "arrow_config": null, + "auto_mapping": null, + "base_model_name_or_path": "/data/OPSD/assets/models/llava-1.5-7b-hf", + "bias": "none", + "corda_config": null, + "ensure_weight_tying": false, + "eva_config": null, + "exclude_modules": null, + "fan_in_fan_out": false, + "inference_mode": true, + "init_lora_weights": true, + "layer_replication": null, + "layers_pattern": null, + "layers_to_transform": null, + "loftq_config": {}, + "lora_alpha": 32, + "lora_bias": false, + "lora_dropout": 0.05, + "lora_ga_config": null, + "megatron_config": null, + "megatron_core": "megatron.core", + "modules_to_save": null, + "monteclora_config": null, + "peft_type": "LORA", + "peft_version": "0.20.0", + "qalora_group_size": 16, + "r": 16, + "rank_pattern": {}, + "revision": null, + "target_modules": [ + "model.language_model.layers.22.mlp.gate_proj", + "model.language_model.layers.22.mlp.up_proj", + "model.language_model.layers.22.self_attn.k_proj", + "model.language_model.layers.22.mlp.down_proj", + "model.language_model.layers.22.self_attn.o_proj", + "model.language_model.layers.22.self_attn.v_proj", + "model.language_model.layers.22.self_attn.q_proj" + ], + "target_parameters": null, + "task_type": "CAUSAL_LM", + "trainable_token_indices": null, + "use_bdlora": null, + "use_dora": false, + "use_qalora": false, + "use_rslora": false, + "velora_config": null +} \ No newline at end of file diff --git a/runs/single_llava/removal/L22_s13/config.json b/runs/single_llava/removal/L22_s13/config.json new file mode 100644 index 0000000000000000000000000000000000000000..d8acb05e9930fa98e195dbb9eb9c87b72a37e0b9 --- /dev/null +++ b/runs/single_llava/removal/L22_s13/config.json @@ -0,0 +1,24 @@ +{ + "behavior": "removal", + "seed": 13, + "model": "/data/OPSD/assets/models/llava-1.5-7b-hf", + "backbone": "llava", + "train": "/data/OPSD/assets/v3/removal.jsonl", + "items": 1000, + "steps": 500, + "epochs": 2, + "steps_per_epoch": 250, + "groups_per_step": 4, + "lr": 2e-05, + "scheduler": "constant", + "warmup_steps": 0, + "final_fraction": 1.0, + "weight_decay": 0.0, + "clip": 1.0, + "initial_loss_scale": 4.0, + "adapter": { + "layer": 22, + "rank": 16, + "trainable_parameters": 1249280 + } +} \ No newline at end of file diff --git a/runs/single_llava/removal/L22_s13/steps.jsonl b/runs/single_llava/removal/L22_s13/steps.jsonl new file mode 100644 index 0000000000000000000000000000000000000000..95450fcb10b6b7774967df9a31cf26c97fde07a6 --- /dev/null +++ b/runs/single_llava/removal/L22_s13/steps.jsonl @@ -0,0 +1,500 @@ +{"step": 1, "steps": 500, "lr": 2e-05, "grad_norm": 0.1188163086771965, "loss_scale": 4.0, "retries": 0, "seconds": 2.5, "max_memory_gb": 15.39, "ce": 1.6287612388841808, "rank": 0.13243506848812103} +{"step": 2, "steps": 500, "lr": 2e-05, "grad_norm": 0.2245241105556488, "loss_scale": 4.0, "retries": 0, "seconds": 4.2, "max_memory_gb": 15.4, "ce": 2.6746547917136922, "rank": 0.032057762145996094} +{"step": 3, "steps": 500, "lr": 2e-05, "grad_norm": 0.4415091872215271, "loss_scale": 4.0, "retries": 0, "seconds": 5.9, "max_memory_gb": 15.4, "ce": 5.278622932732105, "rank": 0.15012836456298828} +{"step": 4, "steps": 500, "lr": 2e-05, "grad_norm": 0.3573002517223358, "loss_scale": 4.0, "retries": 0, "seconds": 7.6, "max_memory_gb": 15.4, "ce": 3.802366811200045, "rank": 0.12172389030456543} +{"step": 5, "steps": 500, "lr": 2e-05, "grad_norm": 0.03301753103733063, "loss_scale": 4.0, "retries": 0, "seconds": 9.2, "max_memory_gb": 15.4, "ce": 0.1396686751395464, "rank": 0.0} +{"step": 6, "steps": 500, "lr": 2e-05, "grad_norm": 0.15980073809623718, "loss_scale": 4.0, "retries": 0, "seconds": 10.8, "max_memory_gb": 15.4, "ce": 1.5137383784167469, "rank": 0.0} +{"step": 7, "steps": 500, "lr": 2e-05, "grad_norm": 0.36476650834083557, "loss_scale": 4.0, "retries": 0, "seconds": 12.5, "max_memory_gb": 15.4, "ce": 3.462360544828698, "rank": 0.02062535285949707} +{"step": 8, "steps": 500, "lr": 2e-05, "grad_norm": 0.3832465410232544, "loss_scale": 4.0, "retries": 0, "seconds": 14.1, "max_memory_gb": 15.4, "ce": 3.7090398613363504, "rank": 0.20828473567962646} +{"step": 9, "steps": 500, "lr": 2e-05, "grad_norm": 0.34835121035575867, "loss_scale": 4.0, "retries": 0, "seconds": 15.7, "max_memory_gb": 15.4, "ce": 2.835477450862527, "rank": 0.042327165603637695} +{"step": 10, "steps": 500, "lr": 2e-05, "grad_norm": 0.17968608438968658, "loss_scale": 4.0, "retries": 0, "seconds": 17.4, "max_memory_gb": 15.4, "ce": 1.6168688284233212, "rank": 0.0} +{"step": 11, "steps": 500, "lr": 2e-05, "grad_norm": 0.3418327867984772, "loss_scale": 4.0, "retries": 0, "seconds": 19.1, "max_memory_gb": 15.4, "ce": 2.6260154442861676, "rank": 0.21723291277885437} +{"step": 12, "steps": 500, "lr": 2e-05, "grad_norm": 0.201075479388237, "loss_scale": 4.0, "retries": 0, "seconds": 20.8, "max_memory_gb": 15.4, "ce": 1.51110658980906, "rank": 0.05254507064819336} +{"step": 13, "steps": 500, "lr": 2e-05, "grad_norm": 0.38177958130836487, "loss_scale": 4.0, "retries": 0, "seconds": 22.4, "max_memory_gb": 15.4, "ce": 2.671349123585969, "rank": 0.0} +{"step": 14, "steps": 500, "lr": 2e-05, "grad_norm": 0.6178197860717773, "loss_scale": 4.0, "retries": 0, "seconds": 24.1, "max_memory_gb": 15.4, "ce": 3.8680640012025833, "rank": 0.1457306146621704} +{"step": 15, "steps": 500, "lr": 2e-05, "grad_norm": 0.397378534078598, "loss_scale": 4.0, "retries": 0, "seconds": 25.8, "max_memory_gb": 15.4, "ce": 2.5988172516808845, "rank": 0.1743776798248291} +{"step": 16, "steps": 500, "lr": 2e-05, "grad_norm": 0.23851829767227173, "loss_scale": 4.0, "retries": 0, "seconds": 27.4, "max_memory_gb": 15.4, "ce": 1.4154103437904269, "rank": 0.0} +{"step": 17, "steps": 500, "lr": 2e-05, "grad_norm": 0.26116061210632324, "loss_scale": 4.0, "retries": 0, "seconds": 29.1, "max_memory_gb": 15.4, "ce": 1.5949621115578339, "rank": 0.03611278533935547} +{"step": 18, "steps": 500, "lr": 2e-05, "grad_norm": 0.654715895652771, "loss_scale": 4.0, "retries": 0, "seconds": 30.8, "max_memory_gb": 15.41, "ce": 3.579666757956147, "rank": 0.1947445273399353} +{"step": 19, "steps": 500, "lr": 2e-05, "grad_norm": 0.9937361478805542, "loss_scale": 4.0, "retries": 0, "seconds": 32.5, "max_memory_gb": 15.41, "ce": 5.263905907049775, "rank": 0.32267165184020996} +{"step": 20, "steps": 500, "lr": 2e-05, "grad_norm": 0.01712033897638321, "loss_scale": 4.0, "retries": 0, "seconds": 34.1, "max_memory_gb": 15.41, "ce": 0.06552455294877291, "rank": 0.10890404880046844} +{"step": 21, "steps": 500, "lr": 2e-05, "grad_norm": 0.6121114492416382, "loss_scale": 4.0, "retries": 0, "seconds": 35.8, "max_memory_gb": 15.41, "ce": 2.731437813490629, "rank": 0.1698446273803711} +{"step": 22, "steps": 500, "lr": 2e-05, "grad_norm": 0.5855420231819153, "loss_scale": 4.0, "retries": 0, "seconds": 37.5, "max_memory_gb": 15.41, "ce": 2.9340506615117192, "rank": 0.18494069576263428} +{"step": 23, "steps": 500, "lr": 2e-05, "grad_norm": 0.6278239488601685, "loss_scale": 4.0, "retries": 0, "seconds": 39.2, "max_memory_gb": 15.41, "ce": 2.731867381138727, "rank": 0.12548673152923584} +{"step": 24, "steps": 500, "lr": 2e-05, "grad_norm": 1.04805588722229, "loss_scale": 4.0, "retries": 0, "seconds": 40.8, "max_memory_gb": 15.41, "ce": 4.976392411626875, "rank": 0.2548891305923462} +{"step": 25, "steps": 500, "lr": 2e-05, "grad_norm": 0.9704387187957764, "loss_scale": 4.0, "retries": 0, "seconds": 42.5, "max_memory_gb": 15.41, "ce": 4.0925448797643185, "rank": 0.05977463722229004} +{"step": 26, "steps": 500, "lr": 2e-05, "grad_norm": 1.179738998413086, "loss_scale": 4.0, "retries": 0, "seconds": 44.2, "max_memory_gb": 15.41, "ce": 5.129008608870208, "rank": 0.07420861721038818} +{"step": 27, "steps": 500, "lr": 2e-05, "grad_norm": 1.042980670928955, "loss_scale": 4.0, "retries": 0, "seconds": 45.9, "max_memory_gb": 15.41, "ce": 4.166838517412543, "rank": 0.5043240785598755} +{"step": 28, "steps": 500, "lr": 2e-05, "grad_norm": 0.6466562151908875, "loss_scale": 4.0, "retries": 0, "seconds": 47.6, "max_memory_gb": 15.41, "ce": 2.7388248685747385, "rank": 0.0} +{"step": 29, "steps": 500, "lr": 2e-05, "grad_norm": 0.7045882344245911, "loss_scale": 4.0, "retries": 0, "seconds": 49.2, "max_memory_gb": 15.41, "ce": 2.643367361393757, "rank": 0.0} +{"step": 30, "steps": 500, "lr": 2e-05, "grad_norm": 0.7727327942848206, "loss_scale": 4.0, "retries": 0, "seconds": 51.0, "max_memory_gb": 15.41, "ce": 2.7251760823419318, "rank": 0.0} +{"step": 31, "steps": 500, "lr": 2e-05, "grad_norm": 0.8159219622612, "loss_scale": 4.0, "retries": 0, "seconds": 52.6, "max_memory_gb": 15.41, "ce": 2.784306381829083, "rank": 0.1714339256286621} +{"step": 32, "steps": 500, "lr": 2e-05, "grad_norm": 1.719802737236023, "loss_scale": 4.0, "retries": 0, "seconds": 54.3, "max_memory_gb": 15.41, "ce": 5.204097617417574, "rank": 0.26600003242492676} +{"step": 33, "steps": 500, "lr": 2e-05, "grad_norm": 0.6995014548301697, "loss_scale": 4.0, "retries": 0, "seconds": 56.0, "max_memory_gb": 15.41, "ce": 2.486010086722672, "rank": 0.010972917079925537} +{"step": 34, "steps": 500, "lr": 2e-05, "grad_norm": 0.8578401803970337, "loss_scale": 4.0, "retries": 0, "seconds": 57.7, "max_memory_gb": 15.41, "ce": 2.685389059362933, "rank": 0.0} +{"step": 35, "steps": 500, "lr": 2e-05, "grad_norm": 0.9597033262252808, "loss_scale": 4.0, "retries": 0, "seconds": 59.4, "max_memory_gb": 15.41, "ce": 2.6719252476468682, "rank": 0.14590191841125488} +{"step": 36, "steps": 500, "lr": 2e-05, "grad_norm": 0.8938037753105164, "loss_scale": 4.0, "retries": 0, "seconds": 61.3, "max_memory_gb": 15.41, "ce": 2.542445518542081, "rank": 0.0} +{"step": 37, "steps": 500, "lr": 2e-05, "grad_norm": 0.9337952733039856, "loss_scale": 4.0, "retries": 0, "seconds": 63.1, "max_memory_gb": 15.41, "ce": 2.5787329506129026, "rank": 0.0} +{"step": 38, "steps": 500, "lr": 2e-05, "grad_norm": 1.7166106700897217, "loss_scale": 4.0, "retries": 0, "seconds": 64.9, "max_memory_gb": 15.41, "ce": 4.73053817730397, "rank": 0.17619407176971436} +{"step": 39, "steps": 500, "lr": 2e-05, "grad_norm": 0.6561536192893982, "loss_scale": 4.0, "retries": 0, "seconds": 66.6, "max_memory_gb": 15.41, "ce": 1.4226866896497086, "rank": 0.18945487588644028} +{"step": 40, "steps": 500, "lr": 2e-05, "grad_norm": 1.0320922136306763, "loss_scale": 4.0, "retries": 0, "seconds": 68.2, "max_memory_gb": 15.41, "ce": 2.623544469475746, "rank": 0.04142570495605469} +{"step": 41, "steps": 500, "lr": 2e-05, "grad_norm": 0.5735093951225281, "loss_scale": 4.0, "retries": 0, "seconds": 69.9, "max_memory_gb": 15.41, "ce": 1.4903672114014626, "rank": 0.0} +{"step": 42, "steps": 500, "lr": 2e-05, "grad_norm": 1.1252214908599854, "loss_scale": 4.0, "retries": 0, "seconds": 71.5, "max_memory_gb": 15.41, "ce": 2.3502783321309835, "rank": 0.0} +{"step": 43, "steps": 500, "lr": 2e-05, "grad_norm": 1.0519829988479614, "loss_scale": 4.0, "retries": 0, "seconds": 73.1, "max_memory_gb": 15.41, "ce": 2.4576372485607862, "rank": 0.0} +{"step": 44, "steps": 500, "lr": 2e-05, "grad_norm": 0.6275352835655212, "loss_scale": 4.0, "retries": 0, "seconds": 74.8, "max_memory_gb": 15.41, "ce": 1.4027524751145393, "rank": 0.0} +{"step": 45, "steps": 500, "lr": 2e-05, "grad_norm": 0.7043789625167847, "loss_scale": 4.0, "retries": 0, "seconds": 76.4, "max_memory_gb": 15.41, "ce": 1.3971568434499204, "rank": 0.0} +{"step": 46, "steps": 500, "lr": 2e-05, "grad_norm": 1.097424864768982, "loss_scale": 4.0, "retries": 0, "seconds": 78.1, "max_memory_gb": 15.41, "ce": 2.3216345505788922, "rank": 0.0} +{"step": 47, "steps": 500, "lr": 2e-05, "grad_norm": 1.038183331489563, "loss_scale": 4.0, "retries": 0, "seconds": 79.7, "max_memory_gb": 15.41, "ce": 2.0502275032922626, "rank": 0.0} +{"step": 48, "steps": 500, "lr": 2e-05, "grad_norm": 2.092421293258667, "loss_scale": 4.0, "retries": 0, "seconds": 81.4, "max_memory_gb": 15.41, "ce": 3.7269994881935418, "rank": 0.13990581035614014} +{"step": 49, "steps": 500, "lr": 2e-05, "grad_norm": 1.8646560907363892, "loss_scale": 4.0, "retries": 0, "seconds": 83.1, "max_memory_gb": 15.41, "ce": 3.5497722839936614, "rank": 0.04280722141265869} +{"step": 50, "steps": 500, "lr": 2e-05, "grad_norm": 1.9527397155761719, "loss_scale": 4.0, "retries": 0, "seconds": 84.7, "max_memory_gb": 15.41, "ce": 3.608518877066672, "rank": 0.06693136692047119} +{"step": 51, "steps": 500, "lr": 2e-05, "grad_norm": 0.8063995242118835, "loss_scale": 4.0, "retries": 0, "seconds": 86.4, "max_memory_gb": 15.41, "ce": 1.3879730546614155, "rank": 0.07425487041473389} +{"step": 52, "steps": 500, "lr": 2e-05, "grad_norm": 1.4341257810592651, "loss_scale": 4.0, "retries": 0, "seconds": 88.0, "max_memory_gb": 15.41, "ce": 2.357458452694118, "rank": 0.0} +{"step": 53, "steps": 500, "lr": 2e-05, "grad_norm": 2.0609469413757324, "loss_scale": 4.0, "retries": 0, "seconds": 89.6, "max_memory_gb": 15.41, "ce": 3.664292074739933, "rank": 0.03731882572174072} +{"step": 54, "steps": 500, "lr": 2e-05, "grad_norm": 2.1309404373168945, "loss_scale": 4.0, "retries": 0, "seconds": 91.3, "max_memory_gb": 15.41, "ce": 3.752021010965109, "rank": 0.0} +{"step": 55, "steps": 500, "lr": 2e-05, "grad_norm": 1.448896050453186, "loss_scale": 4.0, "retries": 0, "seconds": 92.9, "max_memory_gb": 15.41, "ce": 2.4425736790290102, "rank": 0.0} +{"step": 56, "steps": 500, "lr": 2e-05, "grad_norm": 2.9841349124908447, "loss_scale": 4.0, "retries": 0, "seconds": 94.6, "max_memory_gb": 15.41, "ce": 4.764348589815199, "rank": 0.036224961280822754} +{"step": 57, "steps": 500, "lr": 2e-05, "grad_norm": 0.8569468855857849, "loss_scale": 4.0, "retries": 0, "seconds": 96.2, "max_memory_gb": 15.41, "ce": 1.2699482175521553, "rank": 0.026875808835029602} +{"step": 58, "steps": 500, "lr": 2e-05, "grad_norm": 1.7093541622161865, "loss_scale": 4.0, "retries": 0, "seconds": 98.0, "max_memory_gb": 15.41, "ce": 2.3509651604108512, "rank": 0.0} +{"step": 59, "steps": 500, "lr": 2e-05, "grad_norm": 1.012056589126587, "loss_scale": 4.0, "retries": 0, "seconds": 99.8, "max_memory_gb": 15.41, "ce": 1.4097079026978463, "rank": 0.00548243522644043} +{"step": 60, "steps": 500, "lr": 2e-05, "grad_norm": 1.5213669538497925, "loss_scale": 4.0, "retries": 0, "seconds": 101.5, "max_memory_gb": 15.41, "ce": 2.38636794523336, "rank": 0.0} +{"step": 61, "steps": 500, "lr": 2e-05, "grad_norm": 2.5553905963897705, "loss_scale": 4.0, "retries": 0, "seconds": 103.4, "max_memory_gb": 15.41, "ce": 3.4553757030516863, "rank": 0.030336856842041016} +{"step": 62, "steps": 500, "lr": 2e-05, "grad_norm": 0.907383918762207, "loss_scale": 4.0, "retries": 0, "seconds": 105.2, "max_memory_gb": 15.41, "ce": 1.3393505002604797, "rank": 0.0} +{"step": 63, "steps": 500, "lr": 2e-05, "grad_norm": 2.7130227088928223, "loss_scale": 4.0, "retries": 0, "seconds": 107.0, "max_memory_gb": 15.41, "ce": 3.5690661852713674, "rank": 0.23939990997314453} +{"step": 64, "steps": 500, "lr": 2e-05, "grad_norm": 2.9121317863464355, "loss_scale": 4.0, "retries": 0, "seconds": 108.7, "max_memory_gb": 15.41, "ce": 3.530500807799399, "rank": 0.016869544982910156} +{"step": 65, "steps": 500, "lr": 2e-05, "grad_norm": 2.6412150859832764, "loss_scale": 4.0, "retries": 0, "seconds": 110.4, "max_memory_gb": 15.41, "ce": 3.6429016906768084, "rank": 0.0} +{"step": 66, "steps": 500, "lr": 2e-05, "grad_norm": 0.72947096824646, "loss_scale": 4.0, "retries": 0, "seconds": 112.0, "max_memory_gb": 15.41, "ce": 0.9786345121683553, "rank": 0.22546425461769104} +{"step": 67, "steps": 500, "lr": 2e-05, "grad_norm": 2.8906726837158203, "loss_scale": 4.0, "retries": 0, "seconds": 113.7, "max_memory_gb": 15.41, "ce": 3.31424162350595, "rank": 0.0} +{"step": 68, "steps": 500, "lr": 2e-05, "grad_norm": 1.023248314857483, "loss_scale": 4.0, "retries": 0, "seconds": 115.4, "max_memory_gb": 15.41, "ce": 1.180085372645408, "rank": 0.0} +{"step": 69, "steps": 500, "lr": 2e-05, "grad_norm": 1.8727707862854004, "loss_scale": 4.0, "retries": 0, "seconds": 117.0, "max_memory_gb": 15.41, "ce": 2.2028208854608238, "rank": 0.0} +{"step": 70, "steps": 500, "lr": 2e-05, "grad_norm": 2.0346052646636963, "loss_scale": 4.0, "retries": 0, "seconds": 118.7, "max_memory_gb": 15.41, "ce": 2.221672880463302, "rank": 0.08609950542449951} +{"step": 71, "steps": 500, "lr": 2e-05, "grad_norm": 3.4053685665130615, "loss_scale": 4.0, "retries": 0, "seconds": 120.3, "max_memory_gb": 15.41, "ce": 3.2211821572855115, "rank": 0.06640756130218506} +{"step": 72, "steps": 500, "lr": 2e-05, "grad_norm": 3.693303108215332, "loss_scale": 4.0, "retries": 0, "seconds": 122.0, "max_memory_gb": 15.41, "ce": 3.870567667298019, "rank": 0.07701277732849121} +{"step": 73, "steps": 500, "lr": 2e-05, "grad_norm": 0.08300641179084778, "loss_scale": 4.0, "retries": 0, "seconds": 123.6, "max_memory_gb": 15.41, "ce": 0.248453154345043, "rank": 0.0} +{"step": 74, "steps": 500, "lr": 2e-05, "grad_norm": 2.225940227508545, "loss_scale": 4.0, "retries": 0, "seconds": 125.3, "max_memory_gb": 15.41, "ce": 2.189079230418429, "rank": 0.1329871416091919} +{"step": 75, "steps": 500, "lr": 2e-05, "grad_norm": 1.4329689741134644, "loss_scale": 4.0, "retries": 0, "seconds": 127.0, "max_memory_gb": 15.41, "ce": 1.61582683166489, "rank": 0.14406859874725342} +{"step": 76, "steps": 500, "lr": 2e-05, "grad_norm": 2.201343297958374, "loss_scale": 4.0, "retries": 0, "seconds": 128.6, "max_memory_gb": 15.41, "ce": 2.1626194422133267, "rank": 0.0} +{"step": 77, "steps": 500, "lr": 2e-05, "grad_norm": 3.789870023727417, "loss_scale": 4.0, "retries": 0, "seconds": 130.2, "max_memory_gb": 15.41, "ce": 3.322392991743982, "rank": 0.07337677478790283} +{"step": 78, "steps": 500, "lr": 2e-05, "grad_norm": 2.1334850788116455, "loss_scale": 4.0, "retries": 0, "seconds": 131.8, "max_memory_gb": 15.41, "ce": 1.8100111708045006, "rank": 0.0} +{"step": 79, "steps": 500, "lr": 2e-05, "grad_norm": 1.3888665437698364, "loss_scale": 4.0, "retries": 0, "seconds": 133.4, "max_memory_gb": 15.41, "ce": 0.9716585548594594, "rank": 0.00720369815826416} +{"step": 80, "steps": 500, "lr": 2e-05, "grad_norm": 2.3967232704162598, "loss_scale": 4.0, "retries": 0, "seconds": 135.1, "max_memory_gb": 15.41, "ce": 2.0291966571239755, "rank": 0.0} +{"step": 81, "steps": 500, "lr": 2e-05, "grad_norm": 3.739349603652954, "loss_scale": 4.0, "retries": 0, "seconds": 136.7, "max_memory_gb": 15.41, "ce": 2.951178398914635, "rank": 0.04038989543914795} +{"step": 82, "steps": 500, "lr": 2e-05, "grad_norm": 4.228692054748535, "loss_scale": 4.0, "retries": 0, "seconds": 138.3, "max_memory_gb": 15.41, "ce": 2.8527348046191037, "rank": 0.0} +{"step": 83, "steps": 500, "lr": 2e-05, "grad_norm": 3.440953016281128, "loss_scale": 4.0, "retries": 0, "seconds": 140.0, "max_memory_gb": 15.41, "ce": 2.616775971953757, "rank": 0.008682727813720703} +{"step": 84, "steps": 500, "lr": 2e-05, "grad_norm": 1.4682583808898926, "loss_scale": 4.0, "retries": 0, "seconds": 141.7, "max_memory_gb": 15.41, "ce": 1.07790675672004, "rank": 0.0} +{"step": 85, "steps": 500, "lr": 2e-05, "grad_norm": 0.08604800701141357, "loss_scale": 4.0, "retries": 0, "seconds": 143.4, "max_memory_gb": 15.41, "ce": 0.12788128905231133, "rank": 0.0} +{"step": 86, "steps": 500, "lr": 2e-05, "grad_norm": 4.382697105407715, "loss_scale": 4.0, "retries": 0, "seconds": 145.1, "max_memory_gb": 15.41, "ce": 2.7961214995011687, "rank": 0.0} +{"step": 87, "steps": 500, "lr": 2e-05, "grad_norm": 4.305049419403076, "loss_scale": 4.0, "retries": 0, "seconds": 146.7, "max_memory_gb": 15.41, "ce": 2.762384250294417, "rank": 0.0} +{"step": 88, "steps": 500, "lr": 2e-05, "grad_norm": 3.645681142807007, "loss_scale": 4.0, "retries": 0, "seconds": 148.4, "max_memory_gb": 15.41, "ce": 1.8219383427640423, "rank": 0.050559818744659424} +{"step": 89, "steps": 500, "lr": 2e-05, "grad_norm": 0.04996484890580177, "loss_scale": 4.0, "retries": 0, "seconds": 150.0, "max_memory_gb": 15.41, "ce": 0.155673042871058, "rank": 0.12443676590919495} +{"step": 90, "steps": 500, "lr": 2e-05, "grad_norm": 2.778733968734741, "loss_scale": 4.0, "retries": 0, "seconds": 151.6, "max_memory_gb": 15.41, "ce": 1.5338652246864513, "rank": 0.0052430033683776855} +{"step": 91, "steps": 500, "lr": 2e-05, "grad_norm": 0.10668669641017914, "loss_scale": 4.0, "retries": 0, "seconds": 153.3, "max_memory_gb": 15.41, "ce": 0.36090481374412775, "rank": 0.0} +{"step": 92, "steps": 500, "lr": 2e-05, "grad_norm": 1.9869848489761353, "loss_scale": 4.0, "retries": 0, "seconds": 155.0, "max_memory_gb": 15.41, "ce": 0.9376953634782694, "rank": 0.0} +{"step": 93, "steps": 500, "lr": 2e-05, "grad_norm": 3.638580083847046, "loss_scale": 4.0, "retries": 0, "seconds": 156.6, "max_memory_gb": 15.41, "ce": 1.5129372188821435, "rank": 0.0} +{"step": 94, "steps": 500, "lr": 2e-05, "grad_norm": 5.392441272735596, "loss_scale": 4.0, "retries": 0, "seconds": 158.3, "max_memory_gb": 15.41, "ce": 2.6696898778900504, "rank": 0.10835123062133789} +{"step": 95, "steps": 500, "lr": 2e-05, "grad_norm": 3.4231221675872803, "loss_scale": 4.0, "retries": 0, "seconds": 160.0, "max_memory_gb": 15.41, "ce": 1.7807288873009384, "rank": 0.0} +{"step": 96, "steps": 500, "lr": 2e-05, "grad_norm": 5.874547481536865, "loss_scale": 4.0, "retries": 0, "seconds": 161.6, "max_memory_gb": 15.41, "ce": 2.549156245775521, "rank": 0.017686963081359863} +{"step": 97, "steps": 500, "lr": 2e-05, "grad_norm": 2.233212947845459, "loss_scale": 4.0, "retries": 0, "seconds": 163.3, "max_memory_gb": 15.41, "ce": 1.0250864918343723, "rank": 0.0} +{"step": 98, "steps": 500, "lr": 2e-05, "grad_norm": 3.7538037300109863, "loss_scale": 4.0, "retries": 0, "seconds": 165.0, "max_memory_gb": 15.41, "ce": 1.506006799521856, "rank": 0.0} +{"step": 99, "steps": 500, "lr": 2e-05, "grad_norm": 0.17869409918785095, "loss_scale": 4.0, "retries": 0, "seconds": 166.6, "max_memory_gb": 15.41, "ce": 0.23964203096693382, "rank": 0.024888381361961365} +{"step": 100, "steps": 500, "lr": 2e-05, "grad_norm": 5.312029838562012, "loss_scale": 4.0, "retries": 0, "seconds": 168.3, "max_memory_gb": 15.41, "ce": 2.258165649138391, "rank": 0.0} +{"step": 101, "steps": 500, "lr": 2e-05, "grad_norm": 5.769137859344482, "loss_scale": 4.0, "retries": 0, "seconds": 170.0, "max_memory_gb": 15.41, "ce": 2.2440680945292115, "rank": 0.0} +{"step": 102, "steps": 500, "lr": 2e-05, "grad_norm": 5.902729034423828, "loss_scale": 4.0, "retries": 0, "seconds": 171.7, "max_memory_gb": 15.41, "ce": 2.320426730439067, "rank": 0.0} +{"step": 103, "steps": 500, "lr": 2e-05, "grad_norm": 4.627213478088379, "loss_scale": 4.0, "retries": 0, "seconds": 173.4, "max_memory_gb": 15.41, "ce": 1.3482487631263211, "rank": 0.13389486074447632} +{"step": 104, "steps": 500, "lr": 2e-05, "grad_norm": 2.8403491973876953, "loss_scale": 4.0, "retries": 0, "seconds": 175.0, "max_memory_gb": 15.41, "ce": 0.7909226382616907, "rank": 0.09920205175876617} +{"step": 105, "steps": 500, "lr": 2e-05, "grad_norm": 6.239508628845215, "loss_scale": 4.0, "retries": 0, "seconds": 176.7, "max_memory_gb": 15.41, "ce": 1.9675726378336549, "rank": 0.16615158319473267} +{"step": 106, "steps": 500, "lr": 2e-05, "grad_norm": 6.7899346351623535, "loss_scale": 4.0, "retries": 0, "seconds": 178.3, "max_memory_gb": 15.41, "ce": 1.8054971732199192, "rank": 0.0006909966468811035} +{"step": 107, "steps": 500, "lr": 2e-05, "grad_norm": 2.497464656829834, "loss_scale": 4.0, "retries": 0, "seconds": 180.0, "max_memory_gb": 15.41, "ce": 0.7689718529582024, "rank": 0.0} +{"step": 108, "steps": 500, "lr": 2e-05, "grad_norm": 2.6667094230651855, "loss_scale": 4.0, "retries": 0, "seconds": 181.7, "max_memory_gb": 15.41, "ce": 0.7559917287435383, "rank": 0.0} +{"step": 109, "steps": 500, "lr": 2e-05, "grad_norm": 7.203331470489502, "loss_scale": 4.0, "retries": 0, "seconds": 183.4, "max_memory_gb": 15.41, "ce": 1.864592095836997, "rank": 0.0} +{"step": 110, "steps": 500, "lr": 2e-05, "grad_norm": 5.130098342895508, "loss_scale": 4.0, "retries": 0, "seconds": 185.1, "max_memory_gb": 15.41, "ce": 1.3084523319266737, "rank": 0.06626874208450317} +{"step": 111, "steps": 500, "lr": 2e-05, "grad_norm": 4.353065490722656, "loss_scale": 4.0, "retries": 0, "seconds": 186.8, "max_memory_gb": 15.41, "ce": 1.0632751872763038, "rank": 0.0} +{"step": 112, "steps": 500, "lr": 2e-05, "grad_norm": 4.895893573760986, "loss_scale": 4.0, "retries": 0, "seconds": 188.5, "max_memory_gb": 15.41, "ce": 1.132935119094327, "rank": 0.0} +{"step": 113, "steps": 500, "lr": 2e-05, "grad_norm": 5.756685256958008, "loss_scale": 4.0, "retries": 0, "seconds": 190.2, "max_memory_gb": 15.41, "ce": 1.131269367877394, "rank": 0.0} +{"step": 114, "steps": 500, "lr": 2e-05, "grad_norm": 7.837313175201416, "loss_scale": 4.0, "retries": 0, "seconds": 191.9, "max_memory_gb": 15.41, "ce": 1.740518406033516, "rank": 0.15077684819698334} +{"step": 115, "steps": 500, "lr": 2e-05, "grad_norm": 5.074172019958496, "loss_scale": 4.0, "retries": 0, "seconds": 193.6, "max_memory_gb": 15.41, "ce": 1.1404594937339425, "rank": 0.002610146999359131} +{"step": 116, "steps": 500, "lr": 2e-05, "grad_norm": 0.063358373939991, "loss_scale": 4.0, "retries": 0, "seconds": 195.2, "max_memory_gb": 15.41, "ce": 0.15024757327046245, "rank": 0.0} +{"step": 117, "steps": 500, "lr": 2e-05, "grad_norm": 4.41282844543457, "loss_scale": 4.0, "retries": 0, "seconds": 196.8, "max_memory_gb": 15.41, "ce": 0.9773825682932511, "rank": 0.0} +{"step": 118, "steps": 500, "lr": 2e-05, "grad_norm": 2.602283477783203, "loss_scale": 4.0, "retries": 0, "seconds": 198.5, "max_memory_gb": 15.41, "ce": 0.6280031055212021, "rank": 0.0} +{"step": 119, "steps": 500, "lr": 2e-05, "grad_norm": 2.307675361633301, "loss_scale": 4.0, "retries": 0, "seconds": 200.1, "max_memory_gb": 15.41, "ce": 0.5068373340182006, "rank": 0.0} +{"step": 120, "steps": 500, "lr": 2e-05, "grad_norm": 4.364255905151367, "loss_scale": 4.0, "retries": 0, "seconds": 201.8, "max_memory_gb": 15.41, "ce": 0.824184306897223, "rank": 0.0} +{"step": 121, "steps": 500, "lr": 2e-05, "grad_norm": 2.595562696456909, "loss_scale": 4.0, "retries": 0, "seconds": 203.5, "max_memory_gb": 15.41, "ce": 0.6467070878716186, "rank": 0.0} +{"step": 122, "steps": 500, "lr": 2e-05, "grad_norm": 3.7274210453033447, "loss_scale": 4.0, "retries": 0, "seconds": 205.1, "max_memory_gb": 15.41, "ce": 0.6878561383346096, "rank": 0.0} +{"step": 123, "steps": 500, "lr": 2e-05, "grad_norm": 2.3291072845458984, "loss_scale": 4.0, "retries": 0, "seconds": 206.8, "max_memory_gb": 15.41, "ce": 0.7357929374556988, "rank": 0.0044787973165512085} +{"step": 124, "steps": 500, "lr": 2e-05, "grad_norm": 1.624730110168457, "loss_scale": 4.0, "retries": 0, "seconds": 208.5, "max_memory_gb": 15.41, "ce": 0.4621287318877876, "rank": 0.0} +{"step": 125, "steps": 500, "lr": 2e-05, "grad_norm": 1.9003138542175293, "loss_scale": 4.0, "retries": 0, "seconds": 210.1, "max_memory_gb": 15.41, "ce": 0.40572934271767735, "rank": 0.0} +{"step": 126, "steps": 500, "lr": 2e-05, "grad_norm": 3.882577896118164, "loss_scale": 4.0, "retries": 0, "seconds": 211.8, "max_memory_gb": 15.41, "ce": 0.835729175247252, "rank": 0.0} +{"step": 127, "steps": 500, "lr": 2e-05, "grad_norm": 1.6909465789794922, "loss_scale": 4.0, "retries": 0, "seconds": 213.4, "max_memory_gb": 15.41, "ce": 0.2797331918263808, "rank": 0.0} +{"step": 128, "steps": 500, "lr": 2e-05, "grad_norm": 2.849113941192627, "loss_scale": 4.0, "retries": 0, "seconds": 215.0, "max_memory_gb": 15.41, "ce": 0.5451728962361813, "rank": 0.0} +{"step": 129, "steps": 500, "lr": 2e-05, "grad_norm": 4.746500492095947, "loss_scale": 4.0, "retries": 0, "seconds": 216.8, "max_memory_gb": 15.41, "ce": 0.7517901593819261, "rank": 0.0} +{"step": 130, "steps": 500, "lr": 2e-05, "grad_norm": 4.260730743408203, "loss_scale": 4.0, "retries": 0, "seconds": 218.4, "max_memory_gb": 15.41, "ce": 0.7506521134637296, "rank": 0.0} +{"step": 131, "steps": 500, "lr": 2e-05, "grad_norm": 1.4222334623336792, "loss_scale": 4.0, "retries": 0, "seconds": 220.1, "max_memory_gb": 15.41, "ce": 0.40414795046672225, "rank": 0.0} +{"step": 132, "steps": 500, "lr": 2e-05, "grad_norm": 2.9680278301239014, "loss_scale": 4.0, "retries": 0, "seconds": 221.7, "max_memory_gb": 15.41, "ce": 0.737835381529294, "rank": 0.0} +{"step": 133, "steps": 500, "lr": 2e-05, "grad_norm": 2.327216863632202, "loss_scale": 4.0, "retries": 0, "seconds": 223.3, "max_memory_gb": 15.41, "ce": 0.5712152668274939, "rank": 0.0} +{"step": 134, "steps": 500, "lr": 2e-05, "grad_norm": 0.1527462750673294, "loss_scale": 4.0, "retries": 0, "seconds": 225.0, "max_memory_gb": 15.41, "ce": 0.22100141050759703, "rank": 0.0} +{"step": 135, "steps": 500, "lr": 2e-05, "grad_norm": 1.4890352487564087, "loss_scale": 4.0, "retries": 0, "seconds": 226.6, "max_memory_gb": 15.41, "ce": 0.47677295468747616, "rank": 0.0} +{"step": 136, "steps": 500, "lr": 2e-05, "grad_norm": 0.7049969434738159, "loss_scale": 4.0, "retries": 0, "seconds": 228.3, "max_memory_gb": 15.41, "ce": 0.41806466737762094, "rank": 0.0} +{"step": 137, "steps": 500, "lr": 2e-05, "grad_norm": 1.1041412353515625, "loss_scale": 4.0, "retries": 0, "seconds": 229.9, "max_memory_gb": 15.41, "ce": 0.369103345554322, "rank": 0.0} +{"step": 138, "steps": 500, "lr": 2e-05, "grad_norm": 0.7865124940872192, "loss_scale": 4.0, "retries": 0, "seconds": 231.7, "max_memory_gb": 15.41, "ce": 0.39125247579067945, "rank": 0.0} +{"step": 139, "steps": 500, "lr": 2e-05, "grad_norm": 1.1506041288375854, "loss_scale": 4.0, "retries": 0, "seconds": 233.4, "max_memory_gb": 15.41, "ce": 0.5928973662666976, "rank": 0.0} +{"step": 140, "steps": 500, "lr": 2e-05, "grad_norm": 0.34637653827667236, "loss_scale": 4.0, "retries": 0, "seconds": 235.1, "max_memory_gb": 15.41, "ce": 0.2884523749817163, "rank": 0.0} +{"step": 141, "steps": 500, "lr": 2e-05, "grad_norm": 0.36378276348114014, "loss_scale": 4.0, "retries": 0, "seconds": 236.9, "max_memory_gb": 15.41, "ce": 0.36341897235251963, "rank": 0.0} +{"step": 142, "steps": 500, "lr": 2e-05, "grad_norm": 0.39313939213752747, "loss_scale": 4.0, "retries": 0, "seconds": 238.6, "max_memory_gb": 15.41, "ce": 0.3435284427832812, "rank": 0.0} +{"step": 143, "steps": 500, "lr": 2e-05, "grad_norm": 0.10179539769887924, "loss_scale": 4.0, "retries": 0, "seconds": 240.3, "max_memory_gb": 15.41, "ce": 0.18502697930671275, "rank": 0.0} +{"step": 144, "steps": 500, "lr": 2e-05, "grad_norm": 0.5407907962799072, "loss_scale": 4.0, "retries": 0, "seconds": 242.0, "max_memory_gb": 15.41, "ce": 0.4167277957312763, "rank": 0.0} +{"step": 145, "steps": 500, "lr": 2e-05, "grad_norm": 0.35146573185920715, "loss_scale": 4.0, "retries": 0, "seconds": 243.7, "max_memory_gb": 15.41, "ce": 0.22869837598409504, "rank": 0.0} +{"step": 146, "steps": 500, "lr": 2e-05, "grad_norm": 0.37374448776245117, "loss_scale": 4.0, "retries": 0, "seconds": 245.4, "max_memory_gb": 15.41, "ce": 0.34003034653142095, "rank": 0.0} +{"step": 147, "steps": 500, "lr": 2e-05, "grad_norm": 0.4672335982322693, "loss_scale": 4.0, "retries": 0, "seconds": 247.1, "max_memory_gb": 15.41, "ce": 0.4791820429963991, "rank": 0.0} +{"step": 148, "steps": 500, "lr": 2e-05, "grad_norm": 0.22426921129226685, "loss_scale": 4.0, "retries": 0, "seconds": 248.7, "max_memory_gb": 15.41, "ce": 0.1499633370549418, "rank": 0.0} +{"step": 149, "steps": 500, "lr": 2e-05, "grad_norm": 0.3118468225002289, "loss_scale": 4.0, "retries": 0, "seconds": 250.4, "max_memory_gb": 15.41, "ce": 0.2532719795126468, "rank": 0.13968157768249512} +{"step": 150, "steps": 500, "lr": 2e-05, "grad_norm": 0.5637852549552917, "loss_scale": 4.0, "retries": 0, "seconds": 252.1, "max_memory_gb": 15.41, "ce": 0.45827787509188056, "rank": 0.05580522119998932} +{"step": 151, "steps": 500, "lr": 2e-05, "grad_norm": 0.370054692029953, "loss_scale": 4.0, "retries": 0, "seconds": 254.0, "max_memory_gb": 15.41, "ce": 0.2601611487916671, "rank": 0.0} +{"step": 152, "steps": 500, "lr": 2e-05, "grad_norm": 0.6126483678817749, "loss_scale": 4.0, "retries": 0, "seconds": 255.8, "max_memory_gb": 15.41, "ce": 0.4060729034245014, "rank": 0.0} +{"step": 153, "steps": 500, "lr": 2e-05, "grad_norm": 0.5142039060592651, "loss_scale": 4.0, "retries": 0, "seconds": 257.5, "max_memory_gb": 15.41, "ce": 0.294939138693735, "rank": 0.0} +{"step": 154, "steps": 500, "lr": 2e-05, "grad_norm": 1.078436017036438, "loss_scale": 4.0, "retries": 0, "seconds": 259.2, "max_memory_gb": 15.41, "ce": 0.6175643550232053, "rank": 0.0883718878030777} +{"step": 155, "steps": 500, "lr": 2e-05, "grad_norm": 1.5774000883102417, "loss_scale": 4.0, "retries": 0, "seconds": 260.9, "max_memory_gb": 15.41, "ce": 0.5087812920100987, "rank": 0.0} +{"step": 156, "steps": 500, "lr": 2e-05, "grad_norm": 0.3730999529361725, "loss_scale": 4.0, "retries": 0, "seconds": 262.5, "max_memory_gb": 15.41, "ce": 0.24989765451755375, "rank": 0.0} +{"step": 157, "steps": 500, "lr": 2e-05, "grad_norm": 0.3783732056617737, "loss_scale": 4.0, "retries": 0, "seconds": 264.1, "max_memory_gb": 15.41, "ce": 0.29335475293919444, "rank": 0.0} +{"step": 158, "steps": 500, "lr": 2e-05, "grad_norm": 0.5274533629417419, "loss_scale": 4.0, "retries": 0, "seconds": 265.8, "max_memory_gb": 15.41, "ce": 0.42164447670802474, "rank": 0.0} +{"step": 159, "steps": 500, "lr": 2e-05, "grad_norm": 0.34341883659362793, "loss_scale": 4.0, "retries": 0, "seconds": 267.5, "max_memory_gb": 15.41, "ce": 0.2509446262847632, "rank": 0.0} +{"step": 160, "steps": 500, "lr": 2e-05, "grad_norm": 0.44797471165657043, "loss_scale": 4.0, "retries": 0, "seconds": 269.3, "max_memory_gb": 15.41, "ce": 0.3271383512765169, "rank": 0.03188150376081467} +{"step": 161, "steps": 500, "lr": 2e-05, "grad_norm": 0.2894546687602997, "loss_scale": 4.0, "retries": 0, "seconds": 271.1, "max_memory_gb": 15.41, "ce": 0.31416436133440584, "rank": 0.0} +{"step": 162, "steps": 500, "lr": 2e-05, "grad_norm": 0.21340860426425934, "loss_scale": 4.0, "retries": 0, "seconds": 272.7, "max_memory_gb": 15.41, "ce": 0.2084277248941362, "rank": 0.0} +{"step": 163, "steps": 500, "lr": 2e-05, "grad_norm": 0.5301190614700317, "loss_scale": 4.0, "retries": 0, "seconds": 274.3, "max_memory_gb": 15.41, "ce": 0.2909209995996207, "rank": 0.0} +{"step": 164, "steps": 500, "lr": 2e-05, "grad_norm": 0.5440644025802612, "loss_scale": 4.0, "retries": 0, "seconds": 276.0, "max_memory_gb": 15.41, "ce": 0.3833202123641968, "rank": 0.0714874193072319} +{"step": 165, "steps": 500, "lr": 2e-05, "grad_norm": 0.2393227070569992, "loss_scale": 4.0, "retries": 0, "seconds": 277.7, "max_memory_gb": 15.41, "ce": 0.25267245015129447, "rank": 0.0} +{"step": 166, "steps": 500, "lr": 2e-05, "grad_norm": 0.4231252074241638, "loss_scale": 4.0, "retries": 0, "seconds": 279.4, "max_memory_gb": 15.41, "ce": 0.3180917245335877, "rank": 0.0} +{"step": 167, "steps": 500, "lr": 2e-05, "grad_norm": 0.42119884490966797, "loss_scale": 4.0, "retries": 0, "seconds": 281.1, "max_memory_gb": 15.41, "ce": 0.3836486570071429, "rank": 0.0} +{"step": 168, "steps": 500, "lr": 2e-05, "grad_norm": 0.30241668224334717, "loss_scale": 4.0, "retries": 0, "seconds": 282.9, "max_memory_gb": 15.41, "ce": 0.22863867576234043, "rank": 0.0} +{"step": 169, "steps": 500, "lr": 2e-05, "grad_norm": 1.0385373830795288, "loss_scale": 4.0, "retries": 0, "seconds": 284.8, "max_memory_gb": 15.41, "ce": 0.6206776075996459, "rank": 0.12517499923706055} +{"step": 170, "steps": 500, "lr": 2e-05, "grad_norm": 0.3365715444087982, "loss_scale": 4.0, "retries": 0, "seconds": 286.5, "max_memory_gb": 15.41, "ce": 0.22614543803501874, "rank": 0.0} +{"step": 171, "steps": 500, "lr": 2e-05, "grad_norm": 0.39834535121917725, "loss_scale": 4.0, "retries": 0, "seconds": 288.3, "max_memory_gb": 15.41, "ce": 0.25574750191299245, "rank": 0.008796796202659607} +{"step": 172, "steps": 500, "lr": 2e-05, "grad_norm": 0.6312347054481506, "loss_scale": 4.0, "retries": 0, "seconds": 290.0, "max_memory_gb": 15.41, "ce": 0.33528248011134565, "rank": 0.0} +{"step": 173, "steps": 500, "lr": 2e-05, "grad_norm": 0.6231979131698608, "loss_scale": 4.0, "retries": 0, "seconds": 291.6, "max_memory_gb": 15.41, "ce": 0.33192324987612665, "rank": 0.045454345643520355} +{"step": 174, "steps": 500, "lr": 2e-05, "grad_norm": 0.6303345561027527, "loss_scale": 4.0, "retries": 0, "seconds": 293.3, "max_memory_gb": 15.41, "ce": 0.2603512442437932, "rank": 0.0} +{"step": 175, "steps": 500, "lr": 2e-05, "grad_norm": 0.30583223700523376, "loss_scale": 4.0, "retries": 0, "seconds": 294.9, "max_memory_gb": 15.41, "ce": 0.3697627419605851, "rank": 0.0} +{"step": 176, "steps": 500, "lr": 2e-05, "grad_norm": 0.49358007311820984, "loss_scale": 4.0, "retries": 0, "seconds": 296.5, "max_memory_gb": 15.41, "ce": 0.24895190307870507, "rank": 0.041258350014686584} +{"step": 177, "steps": 500, "lr": 2e-05, "grad_norm": 0.9185922145843506, "loss_scale": 4.0, "retries": 0, "seconds": 298.1, "max_memory_gb": 15.41, "ce": 0.35443282534834, "rank": 0.12903417646884918} +{"step": 178, "steps": 500, "lr": 2e-05, "grad_norm": 0.2750815749168396, "loss_scale": 4.0, "retries": 0, "seconds": 299.8, "max_memory_gb": 15.41, "ce": 0.29857139999512583, "rank": 0.0} +{"step": 179, "steps": 500, "lr": 2e-05, "grad_norm": 0.20184890925884247, "loss_scale": 4.0, "retries": 0, "seconds": 301.4, "max_memory_gb": 15.41, "ce": 0.43154055636841804, "rank": 0.0} +{"step": 180, "steps": 500, "lr": 2e-05, "grad_norm": 0.9524094462394714, "loss_scale": 4.0, "retries": 0, "seconds": 303.0, "max_memory_gb": 15.41, "ce": 0.4384877064730972, "rank": 0.0} +{"step": 181, "steps": 500, "lr": 2e-05, "grad_norm": 0.6039732098579407, "loss_scale": 4.0, "retries": 0, "seconds": 304.7, "max_memory_gb": 15.41, "ce": 0.2955335583537817, "rank": 0.02087312936782837} +{"step": 182, "steps": 500, "lr": 2e-05, "grad_norm": 0.2733018696308136, "loss_scale": 4.0, "retries": 0, "seconds": 306.3, "max_memory_gb": 15.41, "ce": 0.19886545708868653, "rank": 0.0} +{"step": 183, "steps": 500, "lr": 2e-05, "grad_norm": 0.23374265432357788, "loss_scale": 4.0, "retries": 0, "seconds": 308.0, "max_memory_gb": 15.41, "ce": 0.14032335591036826, "rank": 0.0} +{"step": 184, "steps": 500, "lr": 2e-05, "grad_norm": 0.3047653138637543, "loss_scale": 4.0, "retries": 0, "seconds": 309.6, "max_memory_gb": 15.41, "ce": 0.16984217090066522, "rank": 0.11470501124858856} +{"step": 185, "steps": 500, "lr": 2e-05, "grad_norm": 0.34112265706062317, "loss_scale": 4.0, "retries": 0, "seconds": 311.3, "max_memory_gb": 15.41, "ce": 0.23759983200579882, "rank": 0.0} +{"step": 186, "steps": 500, "lr": 2e-05, "grad_norm": 0.3259398341178894, "loss_scale": 4.0, "retries": 0, "seconds": 313.0, "max_memory_gb": 15.41, "ce": 0.16333442856557667, "rank": 0.06051751971244812} +{"step": 187, "steps": 500, "lr": 2e-05, "grad_norm": 0.7487143278121948, "loss_scale": 4.0, "retries": 0, "seconds": 314.7, "max_memory_gb": 15.41, "ce": 0.35850151162594557, "rank": 0.13657823204994202} +{"step": 188, "steps": 500, "lr": 2e-05, "grad_norm": 0.09787549078464508, "loss_scale": 4.0, "retries": 0, "seconds": 316.3, "max_memory_gb": 15.41, "ce": 0.1690816211339552, "rank": 0.0} +{"step": 189, "steps": 500, "lr": 2e-05, "grad_norm": 0.3049637973308563, "loss_scale": 4.0, "retries": 0, "seconds": 317.9, "max_memory_gb": 15.41, "ce": 0.1720415314193815, "rank": 0.019495196640491486} +{"step": 190, "steps": 500, "lr": 2e-05, "grad_norm": 0.6403149366378784, "loss_scale": 4.0, "retries": 0, "seconds": 319.6, "max_memory_gb": 15.41, "ce": 0.2790722013451159, "rank": 0.0} +{"step": 191, "steps": 500, "lr": 2e-05, "grad_norm": 0.1527053564786911, "loss_scale": 4.0, "retries": 0, "seconds": 321.2, "max_memory_gb": 15.41, "ce": 0.3085479447618127, "rank": 0.0} +{"step": 192, "steps": 500, "lr": 2e-05, "grad_norm": 0.27943113446235657, "loss_scale": 4.0, "retries": 0, "seconds": 322.9, "max_memory_gb": 15.41, "ce": 0.28356674255337566, "rank": 0.0} +{"step": 193, "steps": 500, "lr": 2e-05, "grad_norm": 0.4342039227485657, "loss_scale": 4.0, "retries": 0, "seconds": 324.5, "max_memory_gb": 15.41, "ce": 0.2023866978706792, "rank": 0.17866577953100204} +{"step": 194, "steps": 500, "lr": 2e-05, "grad_norm": 0.25133630633354187, "loss_scale": 4.0, "retries": 0, "seconds": 326.2, "max_memory_gb": 15.41, "ce": 0.3012288175523281, "rank": 0.0} +{"step": 195, "steps": 500, "lr": 2e-05, "grad_norm": 0.3703971803188324, "loss_scale": 4.0, "retries": 0, "seconds": 327.8, "max_memory_gb": 15.41, "ce": 0.2968172924593091, "rank": 0.0} +{"step": 196, "steps": 500, "lr": 2e-05, "grad_norm": 0.5816243886947632, "loss_scale": 4.0, "retries": 0, "seconds": 329.5, "max_memory_gb": 15.41, "ce": 0.2145005203783512, "rank": 0.04528851807117462} +{"step": 197, "steps": 500, "lr": 2e-05, "grad_norm": 0.3194056451320648, "loss_scale": 4.0, "retries": 0, "seconds": 331.2, "max_memory_gb": 15.41, "ce": 0.23373559536412358, "rank": 0.05382990092039108} +{"step": 198, "steps": 500, "lr": 2e-05, "grad_norm": 0.2922821044921875, "loss_scale": 4.0, "retries": 0, "seconds": 332.8, "max_memory_gb": 15.41, "ce": 0.2871594117023051, "rank": 0.0} +{"step": 199, "steps": 500, "lr": 2e-05, "grad_norm": 1.1229090690612793, "loss_scale": 4.0, "retries": 0, "seconds": 334.5, "max_memory_gb": 15.41, "ce": 0.2718808485660702, "rank": 0.0} +{"step": 200, "steps": 500, "lr": 2e-05, "grad_norm": 0.2235068380832672, "loss_scale": 4.0, "retries": 0, "seconds": 336.2, "max_memory_gb": 15.41, "ce": 0.31173964496701956, "rank": 0.06619846075773239} +{"step": 201, "steps": 500, "lr": 2e-05, "grad_norm": 0.2830781638622284, "loss_scale": 4.0, "retries": 0, "seconds": 337.9, "max_memory_gb": 15.41, "ce": 0.21184502402320504, "rank": 0.0} +{"step": 202, "steps": 500, "lr": 2e-05, "grad_norm": 0.1517258733510971, "loss_scale": 4.0, "retries": 0, "seconds": 339.5, "max_memory_gb": 15.41, "ce": 0.07477651664521545, "rank": 0.06043383479118347} +{"step": 203, "steps": 500, "lr": 2e-05, "grad_norm": 0.39256927371025085, "loss_scale": 4.0, "retries": 0, "seconds": 341.2, "max_memory_gb": 15.41, "ce": 0.24727981188334525, "rank": 0.007180526852607727} +{"step": 204, "steps": 500, "lr": 2e-05, "grad_norm": 0.8968574404716492, "loss_scale": 4.0, "retries": 0, "seconds": 342.8, "max_memory_gb": 15.41, "ce": 0.30827279458753765, "rank": 0.0} +{"step": 205, "steps": 500, "lr": 2e-05, "grad_norm": 0.4735858738422394, "loss_scale": 4.0, "retries": 0, "seconds": 344.4, "max_memory_gb": 15.41, "ce": 0.23142522078705952, "rank": 0.0} +{"step": 206, "steps": 500, "lr": 2e-05, "grad_norm": 0.5921293497085571, "loss_scale": 4.0, "retries": 0, "seconds": 346.1, "max_memory_gb": 15.41, "ce": 0.20201844669645652, "rank": 0.02699822187423706} +{"step": 207, "steps": 500, "lr": 2e-05, "grad_norm": 0.3807857930660248, "loss_scale": 4.0, "retries": 0, "seconds": 347.8, "max_memory_gb": 15.41, "ce": 0.4815384173998609, "rank": 0.22064703702926636} +{"step": 208, "steps": 500, "lr": 2e-05, "grad_norm": 0.3498803973197937, "loss_scale": 4.0, "retries": 0, "seconds": 349.4, "max_memory_gb": 15.41, "ce": 0.13498409849125892, "rank": 0.0} +{"step": 209, "steps": 500, "lr": 2e-05, "grad_norm": 0.3569404184818268, "loss_scale": 4.0, "retries": 0, "seconds": 351.1, "max_memory_gb": 15.41, "ce": 0.2896508644334972, "rank": 0.17352112382650375} +{"step": 210, "steps": 500, "lr": 2e-05, "grad_norm": 0.2663828432559967, "loss_scale": 4.0, "retries": 0, "seconds": 352.8, "max_memory_gb": 15.41, "ce": 0.2568686972372234, "rank": 0.05979180335998535} +{"step": 211, "steps": 500, "lr": 2e-05, "grad_norm": 0.5977458953857422, "loss_scale": 4.0, "retries": 0, "seconds": 354.4, "max_memory_gb": 15.41, "ce": 0.2993239478673786, "rank": 0.0} +{"step": 212, "steps": 500, "lr": 2e-05, "grad_norm": 0.7181785106658936, "loss_scale": 4.0, "retries": 0, "seconds": 356.1, "max_memory_gb": 15.41, "ce": 0.39103620569221675, "rank": 0.0} +{"step": 213, "steps": 500, "lr": 2e-05, "grad_norm": 0.28538650274276733, "loss_scale": 4.0, "retries": 0, "seconds": 357.8, "max_memory_gb": 15.41, "ce": 0.31677853176370263, "rank": 0.0} +{"step": 214, "steps": 500, "lr": 2e-05, "grad_norm": 0.9160215854644775, "loss_scale": 4.0, "retries": 0, "seconds": 359.4, "max_memory_gb": 15.41, "ce": 0.35812902366160415, "rank": 0.05198092758655548} +{"step": 215, "steps": 500, "lr": 2e-05, "grad_norm": 0.5163705945014954, "loss_scale": 4.0, "retries": 0, "seconds": 361.1, "max_memory_gb": 15.41, "ce": 0.37421798682771623, "rank": 0.02132074534893036} +{"step": 216, "steps": 500, "lr": 2e-05, "grad_norm": 0.7495042681694031, "loss_scale": 4.0, "retries": 0, "seconds": 362.7, "max_memory_gb": 15.41, "ce": 0.3177318120142445, "rank": 0.0} +{"step": 217, "steps": 500, "lr": 2e-05, "grad_norm": 0.25708645582199097, "loss_scale": 4.0, "retries": 0, "seconds": 364.4, "max_memory_gb": 15.41, "ce": 0.1313134152442217, "rank": 0.0} +{"step": 218, "steps": 500, "lr": 2e-05, "grad_norm": 0.5938242077827454, "loss_scale": 4.0, "retries": 0, "seconds": 366.1, "max_memory_gb": 15.41, "ce": 0.2964478626381606, "rank": 0.0} +{"step": 219, "steps": 500, "lr": 2e-05, "grad_norm": 0.4188278019428253, "loss_scale": 4.0, "retries": 0, "seconds": 367.8, "max_memory_gb": 15.41, "ce": 0.17551869270391762, "rank": 0.0} +{"step": 220, "steps": 500, "lr": 2e-05, "grad_norm": 0.7250410318374634, "loss_scale": 4.0, "retries": 0, "seconds": 369.5, "max_memory_gb": 15.41, "ce": 0.15370140253799036, "rank": 0.0} +{"step": 221, "steps": 500, "lr": 2e-05, "grad_norm": 0.8777823448181152, "loss_scale": 4.0, "retries": 0, "seconds": 371.1, "max_memory_gb": 15.41, "ce": 0.24905722425319254, "rank": 0.0} +{"step": 222, "steps": 500, "lr": 2e-05, "grad_norm": 2.199540138244629, "loss_scale": 4.0, "retries": 0, "seconds": 372.8, "max_memory_gb": 15.41, "ce": 0.414779543178156, "rank": 0.0} +{"step": 223, "steps": 500, "lr": 2e-05, "grad_norm": 0.17139790952205658, "loss_scale": 4.0, "retries": 0, "seconds": 374.4, "max_memory_gb": 15.41, "ce": 0.217024277895689, "rank": 0.0} +{"step": 224, "steps": 500, "lr": 2e-05, "grad_norm": 0.3442637026309967, "loss_scale": 4.0, "retries": 0, "seconds": 376.1, "max_memory_gb": 15.41, "ce": 0.17538526421412826, "rank": 0.10126971453428268} +{"step": 225, "steps": 500, "lr": 2e-05, "grad_norm": 0.3616185784339905, "loss_scale": 4.0, "retries": 0, "seconds": 377.8, "max_memory_gb": 15.41, "ce": 0.15779745997861028, "rank": 0.03924482315778732} +{"step": 226, "steps": 500, "lr": 2e-05, "grad_norm": 0.29574474692344666, "loss_scale": 4.0, "retries": 0, "seconds": 379.4, "max_memory_gb": 15.41, "ce": 0.1885832406114787, "rank": 0.0} +{"step": 227, "steps": 500, "lr": 2e-05, "grad_norm": 0.23830674588680267, "loss_scale": 4.0, "retries": 0, "seconds": 381.1, "max_memory_gb": 15.41, "ce": 0.0687622613622807, "rank": 0.0} +{"step": 228, "steps": 500, "lr": 2e-05, "grad_norm": 0.2335943877696991, "loss_scale": 4.0, "retries": 0, "seconds": 382.8, "max_memory_gb": 15.41, "ce": 0.14373225369490683, "rank": 0.0} +{"step": 229, "steps": 500, "lr": 2e-05, "grad_norm": 0.23072536289691925, "loss_scale": 4.0, "retries": 0, "seconds": 384.4, "max_memory_gb": 15.41, "ce": 0.23844176472630352, "rank": 0.0} +{"step": 230, "steps": 500, "lr": 2e-05, "grad_norm": 0.3941769003868103, "loss_scale": 4.0, "retries": 0, "seconds": 386.1, "max_memory_gb": 15.41, "ce": 0.2885300130583346, "rank": 0.0} +{"step": 231, "steps": 500, "lr": 2e-05, "grad_norm": 0.2110835611820221, "loss_scale": 4.0, "retries": 0, "seconds": 387.7, "max_memory_gb": 15.41, "ce": 0.2350008423672989, "rank": 0.0} +{"step": 232, "steps": 500, "lr": 2e-05, "grad_norm": 0.6715928912162781, "loss_scale": 4.0, "retries": 0, "seconds": 389.3, "max_memory_gb": 15.41, "ce": 0.1397254354087636, "rank": 0.024329297244548798} +{"step": 233, "steps": 500, "lr": 2e-05, "grad_norm": 1.3458781242370605, "loss_scale": 4.0, "retries": 0, "seconds": 391.0, "max_memory_gb": 15.41, "ce": 0.26935181324370205, "rank": 0.0} +{"step": 234, "steps": 500, "lr": 2e-05, "grad_norm": 0.2704213559627533, "loss_scale": 4.0, "retries": 0, "seconds": 392.6, "max_memory_gb": 15.41, "ce": 0.2387545086676255, "rank": 0.0} +{"step": 235, "steps": 500, "lr": 2e-05, "grad_norm": 0.22609099745750427, "loss_scale": 4.0, "retries": 0, "seconds": 394.3, "max_memory_gb": 15.41, "ce": 0.28643944626674056, "rank": 0.0} +{"step": 236, "steps": 500, "lr": 2e-05, "grad_norm": 0.31414467096328735, "loss_scale": 4.0, "retries": 0, "seconds": 396.0, "max_memory_gb": 15.41, "ce": 0.13705050817225128, "rank": 0.0} +{"step": 237, "steps": 500, "lr": 2e-05, "grad_norm": 1.0804020166397095, "loss_scale": 4.0, "retries": 0, "seconds": 397.6, "max_memory_gb": 15.41, "ce": 0.25328624539542943, "rank": 0.048661813139915466} +{"step": 238, "steps": 500, "lr": 2e-05, "grad_norm": 0.4909144639968872, "loss_scale": 4.0, "retries": 0, "seconds": 399.4, "max_memory_gb": 15.41, "ce": 0.14171206136234105, "rank": 0.0} +{"step": 239, "steps": 500, "lr": 2e-05, "grad_norm": 0.3320046067237854, "loss_scale": 4.0, "retries": 0, "seconds": 401.2, "max_memory_gb": 15.41, "ce": 0.24682555010076612, "rank": 0.0} +{"step": 240, "steps": 500, "lr": 2e-05, "grad_norm": 0.5669166445732117, "loss_scale": 4.0, "retries": 0, "seconds": 402.8, "max_memory_gb": 15.41, "ce": 0.16290791262872517, "rank": 0.0} +{"step": 241, "steps": 500, "lr": 2e-05, "grad_norm": 0.25729385018348694, "loss_scale": 4.0, "retries": 0, "seconds": 404.5, "max_memory_gb": 15.41, "ce": 0.22303922730498016, "rank": 0.0} +{"step": 242, "steps": 500, "lr": 2e-05, "grad_norm": 0.5117441415786743, "loss_scale": 4.0, "retries": 0, "seconds": 406.4, "max_memory_gb": 15.41, "ce": 0.17959844949655235, "rank": 0.0} +{"step": 243, "steps": 500, "lr": 2e-05, "grad_norm": 0.47752103209495544, "loss_scale": 4.0, "retries": 0, "seconds": 408.3, "max_memory_gb": 15.41, "ce": 0.10738610185217112, "rank": 0.0} +{"step": 244, "steps": 500, "lr": 2e-05, "grad_norm": 0.5136061310768127, "loss_scale": 4.0, "retries": 0, "seconds": 410.1, "max_memory_gb": 15.41, "ce": 0.2767411894747056, "rank": 0.00920093059539795} +{"step": 245, "steps": 500, "lr": 2e-05, "grad_norm": 0.27218905091285706, "loss_scale": 4.0, "retries": 0, "seconds": 412.0, "max_memory_gb": 15.41, "ce": 0.35976017941720784, "rank": 0.0} +{"step": 246, "steps": 500, "lr": 2e-05, "grad_norm": 0.23290064930915833, "loss_scale": 4.0, "retries": 0, "seconds": 413.7, "max_memory_gb": 15.41, "ce": 0.07418632617918774, "rank": 0.0} +{"step": 247, "steps": 500, "lr": 2e-05, "grad_norm": 0.46423080563545227, "loss_scale": 4.0, "retries": 0, "seconds": 415.3, "max_memory_gb": 15.41, "ce": 0.16399108641780913, "rank": 0.04143211990594864} +{"step": 248, "steps": 500, "lr": 2e-05, "grad_norm": 0.5359357595443726, "loss_scale": 4.0, "retries": 0, "seconds": 417.0, "max_memory_gb": 15.41, "ce": 0.14652238402049989, "rank": 0.0} +{"step": 249, "steps": 500, "lr": 2e-05, "grad_norm": 0.5295213460922241, "loss_scale": 4.0, "retries": 0, "seconds": 418.7, "max_memory_gb": 15.41, "ce": 0.13866815710207447, "rank": 0.0} +{"step": 250, "steps": 500, "lr": 2e-05, "grad_norm": 0.43126437067985535, "loss_scale": 4.0, "retries": 0, "seconds": 420.3, "max_memory_gb": 15.41, "ce": 0.10053847054950893, "rank": 0.0} +{"step": 251, "steps": 500, "lr": 2e-05, "grad_norm": 0.18775615096092224, "loss_scale": 4.0, "retries": 0, "seconds": 422.0, "max_memory_gb": 15.41, "ce": 0.17070353694725782, "rank": 0.0} +{"step": 252, "steps": 500, "lr": 2e-05, "grad_norm": 0.46919485926628113, "loss_scale": 4.0, "retries": 0, "seconds": 423.6, "max_memory_gb": 15.41, "ce": 0.10050468443660066, "rank": 0.0} +{"step": 253, "steps": 500, "lr": 2e-05, "grad_norm": 0.23311257362365723, "loss_scale": 4.0, "retries": 0, "seconds": 425.3, "max_memory_gb": 15.41, "ce": 0.1328122813720256, "rank": 0.0} +{"step": 254, "steps": 500, "lr": 2e-05, "grad_norm": 0.8004326224327087, "loss_scale": 4.0, "retries": 0, "seconds": 426.9, "max_memory_gb": 15.41, "ce": 0.24661616649245843, "rank": 0.0} +{"step": 255, "steps": 500, "lr": 2e-05, "grad_norm": 0.27050939202308655, "loss_scale": 4.0, "retries": 0, "seconds": 428.5, "max_memory_gb": 15.41, "ce": 0.2403837509918958, "rank": 0.0} +{"step": 256, "steps": 500, "lr": 2e-05, "grad_norm": 0.1893235445022583, "loss_scale": 4.0, "retries": 0, "seconds": 430.0, "max_memory_gb": 15.41, "ce": 0.11346700001740828, "rank": 0.0} +{"step": 257, "steps": 500, "lr": 2e-05, "grad_norm": 0.16417647898197174, "loss_scale": 4.0, "retries": 0, "seconds": 431.7, "max_memory_gb": 15.41, "ce": 0.09087051555979997, "rank": 0.0} +{"step": 258, "steps": 500, "lr": 2e-05, "grad_norm": 1.7115387916564941, "loss_scale": 4.0, "retries": 0, "seconds": 433.4, "max_memory_gb": 15.41, "ce": 0.15293809396098368, "rank": 0.0} +{"step": 259, "steps": 500, "lr": 2e-05, "grad_norm": 0.12235335260629654, "loss_scale": 4.0, "retries": 0, "seconds": 435.1, "max_memory_gb": 15.41, "ce": 0.0596649469807744, "rank": 0.0} +{"step": 260, "steps": 500, "lr": 2e-05, "grad_norm": 0.3686465919017792, "loss_scale": 4.0, "retries": 0, "seconds": 436.8, "max_memory_gb": 15.41, "ce": 0.15946041035931557, "rank": 0.0} +{"step": 261, "steps": 500, "lr": 2e-05, "grad_norm": 1.303917646408081, "loss_scale": 4.0, "retries": 0, "seconds": 438.4, "max_memory_gb": 15.41, "ce": 0.22212882331223227, "rank": 0.0} +{"step": 262, "steps": 500, "lr": 2e-05, "grad_norm": 0.3506051301956177, "loss_scale": 4.0, "retries": 0, "seconds": 440.1, "max_memory_gb": 15.41, "ce": 0.23246016312623397, "rank": 0.0} +{"step": 263, "steps": 500, "lr": 2e-05, "grad_norm": 0.8212213516235352, "loss_scale": 4.0, "retries": 0, "seconds": 441.8, "max_memory_gb": 15.41, "ce": 0.28377963340608403, "rank": 0.12218701094388962} +{"step": 264, "steps": 500, "lr": 2e-05, "grad_norm": 0.3764273226261139, "loss_scale": 4.0, "retries": 0, "seconds": 443.4, "max_memory_gb": 15.41, "ce": 0.17393035811255686, "rank": 0.0} +{"step": 265, "steps": 500, "lr": 2e-05, "grad_norm": 0.312697172164917, "loss_scale": 4.0, "retries": 0, "seconds": 445.1, "max_memory_gb": 15.41, "ce": 0.1334515237249434, "rank": 0.0} +{"step": 266, "steps": 500, "lr": 2e-05, "grad_norm": 2.1421778202056885, "loss_scale": 4.0, "retries": 0, "seconds": 446.7, "max_memory_gb": 15.41, "ce": 0.35144223453244194, "rank": 0.0} +{"step": 267, "steps": 500, "lr": 2e-05, "grad_norm": 0.1658608764410019, "loss_scale": 4.0, "retries": 0, "seconds": 448.4, "max_memory_gb": 15.41, "ce": 0.1545260176062584, "rank": 0.0} +{"step": 268, "steps": 500, "lr": 2e-05, "grad_norm": 0.23368845880031586, "loss_scale": 4.0, "retries": 0, "seconds": 450.0, "max_memory_gb": 15.41, "ce": 0.2678928403183818, "rank": 0.0} +{"step": 269, "steps": 500, "lr": 2e-05, "grad_norm": 2.392214775085449, "loss_scale": 4.0, "retries": 0, "seconds": 451.7, "max_memory_gb": 15.41, "ce": 0.49521343427477404, "rank": 0.17050358653068542} +{"step": 270, "steps": 500, "lr": 2e-05, "grad_norm": 0.2635067105293274, "loss_scale": 4.0, "retries": 0, "seconds": 453.4, "max_memory_gb": 15.41, "ce": 0.14795894373673946, "rank": 0.0} +{"step": 271, "steps": 500, "lr": 2e-05, "grad_norm": 0.3516501486301422, "loss_scale": 4.0, "retries": 0, "seconds": 455.0, "max_memory_gb": 15.41, "ce": 0.1145150066295173, "rank": 0.11830053478479385} +{"step": 272, "steps": 500, "lr": 2e-05, "grad_norm": 0.5731061697006226, "loss_scale": 4.0, "retries": 0, "seconds": 456.6, "max_memory_gb": 15.41, "ce": 0.14292726438725367, "rank": 0.0} +{"step": 273, "steps": 500, "lr": 2e-05, "grad_norm": 0.8428523540496826, "loss_scale": 4.0, "retries": 0, "seconds": 458.3, "max_memory_gb": 15.41, "ce": 0.11001890030456707, "rank": 0.1257031410932541} +{"step": 274, "steps": 500, "lr": 2e-05, "grad_norm": 0.5942422151565552, "loss_scale": 4.0, "retries": 0, "seconds": 460.0, "max_memory_gb": 15.41, "ce": 0.1708451448357664, "rank": 0.0} +{"step": 275, "steps": 500, "lr": 2e-05, "grad_norm": 0.6233859658241272, "loss_scale": 4.0, "retries": 0, "seconds": 461.7, "max_memory_gb": 15.41, "ce": 0.5500609492883086, "rank": 0.13062727451324463} +{"step": 276, "steps": 500, "lr": 2e-05, "grad_norm": 0.28463485836982727, "loss_scale": 4.0, "retries": 0, "seconds": 463.5, "max_memory_gb": 15.41, "ce": 0.061263410665560514, "rank": 0.0} +{"step": 277, "steps": 500, "lr": 2e-05, "grad_norm": 0.3129427134990692, "loss_scale": 4.0, "retries": 0, "seconds": 465.2, "max_memory_gb": 15.41, "ce": 0.05774325120728463, "rank": 0.0} +{"step": 278, "steps": 500, "lr": 2e-05, "grad_norm": 0.6861150860786438, "loss_scale": 4.0, "retries": 0, "seconds": 466.8, "max_memory_gb": 15.41, "ce": 0.20342645345954224, "rank": 0.05842204391956329} +{"step": 279, "steps": 500, "lr": 2e-05, "grad_norm": 0.701464831829071, "loss_scale": 4.0, "retries": 0, "seconds": 468.4, "max_memory_gb": 15.41, "ce": 0.09741317582665943, "rank": 0.0} +{"step": 280, "steps": 500, "lr": 2e-05, "grad_norm": 0.4369755983352661, "loss_scale": 4.0, "retries": 0, "seconds": 470.1, "max_memory_gb": 15.41, "ce": 0.07123308524023741, "rank": 0.0} +{"step": 281, "steps": 500, "lr": 2e-05, "grad_norm": 1.5182424783706665, "loss_scale": 4.0, "retries": 0, "seconds": 471.8, "max_memory_gb": 15.41, "ce": 0.2158806727384217, "rank": 0.0} +{"step": 282, "steps": 500, "lr": 2e-05, "grad_norm": 0.23200705647468567, "loss_scale": 4.0, "retries": 0, "seconds": 473.4, "max_memory_gb": 15.41, "ce": 0.1265255100443028, "rank": 0.0} +{"step": 283, "steps": 500, "lr": 2e-05, "grad_norm": 1.0068347454071045, "loss_scale": 4.0, "retries": 0, "seconds": 475.1, "max_memory_gb": 15.41, "ce": 0.14399586606305093, "rank": 0.0} +{"step": 284, "steps": 500, "lr": 2e-05, "grad_norm": 0.822928786277771, "loss_scale": 4.0, "retries": 0, "seconds": 476.8, "max_memory_gb": 15.41, "ce": 0.12207054300233722, "rank": 0.0} +{"step": 285, "steps": 500, "lr": 2e-05, "grad_norm": 0.42002588510513306, "loss_scale": 4.0, "retries": 0, "seconds": 478.5, "max_memory_gb": 15.41, "ce": 0.1503673359984532, "rank": 0.0} +{"step": 286, "steps": 500, "lr": 2e-05, "grad_norm": 0.2653347849845886, "loss_scale": 4.0, "retries": 0, "seconds": 480.1, "max_memory_gb": 15.41, "ce": 0.08160829907865264, "rank": 0.0} +{"step": 287, "steps": 500, "lr": 2e-05, "grad_norm": 0.36306869983673096, "loss_scale": 4.0, "retries": 0, "seconds": 481.8, "max_memory_gb": 15.41, "ce": 0.1598347794788424, "rank": 0.0} +{"step": 288, "steps": 500, "lr": 2e-05, "grad_norm": 0.23810017108917236, "loss_scale": 4.0, "retries": 0, "seconds": 483.5, "max_memory_gb": 15.41, "ce": 0.04190426829154603, "rank": 0.0} +{"step": 289, "steps": 500, "lr": 2e-05, "grad_norm": 0.29937300086021423, "loss_scale": 4.0, "retries": 0, "seconds": 485.2, "max_memory_gb": 15.41, "ce": 0.12873992892855313, "rank": 0.0} +{"step": 290, "steps": 500, "lr": 2e-05, "grad_norm": 0.3190149962902069, "loss_scale": 4.0, "retries": 0, "seconds": 486.9, "max_memory_gb": 15.41, "ce": 0.17706379509763792, "rank": 0.0} +{"step": 291, "steps": 500, "lr": 2e-05, "grad_norm": 0.29820162057876587, "loss_scale": 4.0, "retries": 0, "seconds": 488.5, "max_memory_gb": 15.41, "ce": 0.04906993225449696, "rank": 0.0} +{"step": 292, "steps": 500, "lr": 2e-05, "grad_norm": 0.7140703201293945, "loss_scale": 4.0, "retries": 0, "seconds": 490.2, "max_memory_gb": 15.41, "ce": 0.18076385679887608, "rank": 0.0} +{"step": 293, "steps": 500, "lr": 2e-05, "grad_norm": 0.40206611156463623, "loss_scale": 4.0, "retries": 0, "seconds": 491.8, "max_memory_gb": 15.41, "ce": 0.11309103039093316, "rank": 0.0} +{"step": 294, "steps": 500, "lr": 2e-05, "grad_norm": 0.7435580492019653, "loss_scale": 4.0, "retries": 0, "seconds": 493.4, "max_memory_gb": 15.41, "ce": 0.12338372724479996, "rank": 0.0} +{"step": 295, "steps": 500, "lr": 2e-05, "grad_norm": 1.2129465341567993, "loss_scale": 4.0, "retries": 0, "seconds": 495.0, "max_memory_gb": 15.41, "ce": 0.27859621919924393, "rank": 0.0} +{"step": 296, "steps": 500, "lr": 2e-05, "grad_norm": 0.7440263628959656, "loss_scale": 4.0, "retries": 0, "seconds": 496.6, "max_memory_gb": 15.41, "ce": 0.27319602109491825, "rank": 0.0} +{"step": 297, "steps": 500, "lr": 2e-05, "grad_norm": 0.16014276444911957, "loss_scale": 4.0, "retries": 0, "seconds": 498.3, "max_memory_gb": 15.41, "ce": 0.13298101897817105, "rank": 0.0} +{"step": 298, "steps": 500, "lr": 2e-05, "grad_norm": 2.2493703365325928, "loss_scale": 4.0, "retries": 0, "seconds": 499.9, "max_memory_gb": 15.41, "ce": 0.4480013243155554, "rank": 0.0} +{"step": 299, "steps": 500, "lr": 2e-05, "grad_norm": 0.5531160831451416, "loss_scale": 4.0, "retries": 0, "seconds": 501.6, "max_memory_gb": 15.41, "ce": 0.06268493857351132, "rank": 0.0} +{"step": 300, "steps": 500, "lr": 2e-05, "grad_norm": 0.26687389612197876, "loss_scale": 4.0, "retries": 0, "seconds": 503.3, "max_memory_gb": 15.41, "ce": 0.2757908501953352, "rank": 0.0} +{"step": 301, "steps": 500, "lr": 2e-05, "grad_norm": 0.2562396824359894, "loss_scale": 4.0, "retries": 0, "seconds": 504.9, "max_memory_gb": 15.41, "ce": 0.13614031751058064, "rank": 0.0} +{"step": 302, "steps": 500, "lr": 2e-05, "grad_norm": 0.7079049944877625, "loss_scale": 4.0, "retries": 0, "seconds": 506.6, "max_memory_gb": 15.41, "ce": 0.06291019503260031, "rank": 0.0} +{"step": 303, "steps": 500, "lr": 2e-05, "grad_norm": 0.4139624536037445, "loss_scale": 4.0, "retries": 0, "seconds": 508.2, "max_memory_gb": 15.41, "ce": 0.05261396124842577, "rank": 0.04332268238067627} +{"step": 304, "steps": 500, "lr": 2e-05, "grad_norm": 0.3020716607570648, "loss_scale": 4.0, "retries": 0, "seconds": 509.9, "max_memory_gb": 15.41, "ce": 0.07094430830329657, "rank": 0.06423194706439972} +{"step": 305, "steps": 500, "lr": 2e-05, "grad_norm": 0.8110799789428711, "loss_scale": 4.0, "retries": 0, "seconds": 511.5, "max_memory_gb": 15.41, "ce": 0.1289631957479287, "rank": 0.0} +{"step": 306, "steps": 500, "lr": 2e-05, "grad_norm": 0.6103535294532776, "loss_scale": 4.0, "retries": 0, "seconds": 513.2, "max_memory_gb": 15.41, "ce": 0.07005574268987402, "rank": 0.0} +{"step": 307, "steps": 500, "lr": 2e-05, "grad_norm": 0.2529720664024353, "loss_scale": 4.0, "retries": 0, "seconds": 514.8, "max_memory_gb": 15.41, "ce": 0.09993741779180709, "rank": 0.0} +{"step": 308, "steps": 500, "lr": 2e-05, "grad_norm": 0.7918395400047302, "loss_scale": 4.0, "retries": 0, "seconds": 516.6, "max_memory_gb": 15.41, "ce": 0.18551528801617678, "rank": 0.0} +{"step": 309, "steps": 500, "lr": 2e-05, "grad_norm": 0.2713327705860138, "loss_scale": 4.0, "retries": 0, "seconds": 518.3, "max_memory_gb": 15.41, "ce": 0.1812280242156703, "rank": 0.0} +{"step": 310, "steps": 500, "lr": 2e-05, "grad_norm": 0.7299507260322571, "loss_scale": 4.0, "retries": 0, "seconds": 520.1, "max_memory_gb": 15.41, "ce": 0.08075186354108155, "rank": 0.0} +{"step": 311, "steps": 500, "lr": 2e-05, "grad_norm": 0.1525389552116394, "loss_scale": 4.0, "retries": 0, "seconds": 521.9, "max_memory_gb": 15.41, "ce": 0.1095821054186672, "rank": 0.0} +{"step": 312, "steps": 500, "lr": 2e-05, "grad_norm": 0.2790054380893707, "loss_scale": 4.0, "retries": 0, "seconds": 523.8, "max_memory_gb": 15.41, "ce": 0.046780320844845846, "rank": 0.0} +{"step": 313, "steps": 500, "lr": 2e-05, "grad_norm": 0.7065131068229675, "loss_scale": 4.0, "retries": 0, "seconds": 525.4, "max_memory_gb": 15.41, "ce": 0.1782266878435621, "rank": 0.0} +{"step": 314, "steps": 500, "lr": 2e-05, "grad_norm": 0.37896499037742615, "loss_scale": 4.0, "retries": 0, "seconds": 527.1, "max_memory_gb": 15.41, "ce": 0.1793562442471739, "rank": 0.0} +{"step": 315, "steps": 500, "lr": 2e-05, "grad_norm": 0.238704115152359, "loss_scale": 4.0, "retries": 0, "seconds": 528.8, "max_memory_gb": 15.41, "ce": 0.09468943004321773, "rank": 0.1178816705942154} +{"step": 316, "steps": 500, "lr": 2e-05, "grad_norm": 1.1276050806045532, "loss_scale": 4.0, "retries": 0, "seconds": 530.5, "max_memory_gb": 15.41, "ce": 0.33473750471603125, "rank": 0.0} +{"step": 317, "steps": 500, "lr": 2e-05, "grad_norm": 0.4348389804363251, "loss_scale": 4.0, "retries": 0, "seconds": 532.1, "max_memory_gb": 15.41, "ce": 0.15557401691330597, "rank": 0.0} +{"step": 318, "steps": 500, "lr": 2e-05, "grad_norm": 1.385187029838562, "loss_scale": 4.0, "retries": 0, "seconds": 533.7, "max_memory_gb": 15.41, "ce": 0.22663907654350623, "rank": 0.0} +{"step": 319, "steps": 500, "lr": 2e-05, "grad_norm": 0.590305745601654, "loss_scale": 4.0, "retries": 0, "seconds": 535.4, "max_memory_gb": 15.41, "ce": 0.10923455539159477, "rank": 0.0} +{"step": 320, "steps": 500, "lr": 2e-05, "grad_norm": 0.948674738407135, "loss_scale": 4.0, "retries": 0, "seconds": 537.0, "max_memory_gb": 15.41, "ce": 0.1341648473025998, "rank": 0.0} +{"step": 321, "steps": 500, "lr": 2e-05, "grad_norm": 0.37227582931518555, "loss_scale": 4.0, "retries": 0, "seconds": 538.7, "max_memory_gb": 15.41, "ce": 0.1394579212355893, "rank": 0.12773214280605316} +{"step": 322, "steps": 500, "lr": 2e-05, "grad_norm": 0.32207196950912476, "loss_scale": 4.0, "retries": 0, "seconds": 540.4, "max_memory_gb": 15.41, "ce": 0.043214578210609034, "rank": 0.0} +{"step": 323, "steps": 500, "lr": 2e-05, "grad_norm": 0.25296834111213684, "loss_scale": 4.0, "retries": 0, "seconds": 542.3, "max_memory_gb": 15.41, "ce": 0.060870124463690445, "rank": 0.0} +{"step": 324, "steps": 500, "lr": 2e-05, "grad_norm": 0.3262556195259094, "loss_scale": 4.0, "retries": 0, "seconds": 544.2, "max_memory_gb": 15.41, "ce": 0.04951421199075412, "rank": 0.0} +{"step": 325, "steps": 500, "lr": 2e-05, "grad_norm": 0.27831873297691345, "loss_scale": 4.0, "retries": 0, "seconds": 546.2, "max_memory_gb": 15.41, "ce": 0.21282264207547996, "rank": 0.0} +{"step": 326, "steps": 500, "lr": 2e-05, "grad_norm": 0.2443840503692627, "loss_scale": 4.0, "retries": 0, "seconds": 548.0, "max_memory_gb": 15.41, "ce": 0.06768194344476797, "rank": 0.08521862328052521} +{"step": 327, "steps": 500, "lr": 2e-05, "grad_norm": 1.4789588451385498, "loss_scale": 4.0, "retries": 0, "seconds": 549.9, "max_memory_gb": 15.41, "ce": 0.144533563463483, "rank": 0.0} +{"step": 328, "steps": 500, "lr": 2e-05, "grad_norm": 0.30989906191825867, "loss_scale": 4.0, "retries": 0, "seconds": 551.7, "max_memory_gb": 15.41, "ce": 0.152752782087191, "rank": 0.0} +{"step": 329, "steps": 500, "lr": 2e-05, "grad_norm": 0.2875640094280243, "loss_scale": 4.0, "retries": 0, "seconds": 553.5, "max_memory_gb": 15.41, "ce": 0.15733200559043325, "rank": 0.0} +{"step": 330, "steps": 500, "lr": 2e-05, "grad_norm": 0.3902517855167389, "loss_scale": 4.0, "retries": 0, "seconds": 555.4, "max_memory_gb": 15.41, "ce": 0.15115339873591438, "rank": 0.0} +{"step": 331, "steps": 500, "lr": 2e-05, "grad_norm": 0.4763405919075012, "loss_scale": 4.0, "retries": 0, "seconds": 557.2, "max_memory_gb": 15.41, "ce": 0.2746426864468958, "rank": 0.0} +{"step": 332, "steps": 500, "lr": 2e-05, "grad_norm": 1.5224112272262573, "loss_scale": 4.0, "retries": 0, "seconds": 559.1, "max_memory_gb": 15.41, "ce": 0.18563253548927605, "rank": 0.0} +{"step": 333, "steps": 500, "lr": 2e-05, "grad_norm": 0.30763375759124756, "loss_scale": 4.0, "retries": 0, "seconds": 560.9, "max_memory_gb": 15.41, "ce": 0.26942897189292125, "rank": 0.0} +{"step": 334, "steps": 500, "lr": 2e-05, "grad_norm": 0.3205794095993042, "loss_scale": 4.0, "retries": 0, "seconds": 562.7, "max_memory_gb": 15.41, "ce": 0.143749749615381, "rank": 0.0} +{"step": 335, "steps": 500, "lr": 2e-05, "grad_norm": 2.3771169185638428, "loss_scale": 4.0, "retries": 0, "seconds": 564.7, "max_memory_gb": 15.41, "ce": 0.28572541891480796, "rank": 0.0} +{"step": 336, "steps": 500, "lr": 2e-05, "grad_norm": 0.8848977088928223, "loss_scale": 4.0, "retries": 0, "seconds": 566.6, "max_memory_gb": 15.41, "ce": 0.06379143928643316, "rank": 0.0} +{"step": 337, "steps": 500, "lr": 2e-05, "grad_norm": 1.9039424657821655, "loss_scale": 4.0, "retries": 0, "seconds": 568.5, "max_memory_gb": 15.41, "ce": 0.21840601330040954, "rank": 0.0} +{"step": 338, "steps": 500, "lr": 2e-05, "grad_norm": 0.6822766661643982, "loss_scale": 4.0, "retries": 0, "seconds": 570.4, "max_memory_gb": 15.41, "ce": 0.20663718433934264, "rank": 0.0} +{"step": 339, "steps": 500, "lr": 2e-05, "grad_norm": 0.23509575426578522, "loss_scale": 4.0, "retries": 0, "seconds": 572.2, "max_memory_gb": 15.41, "ce": 0.1407789056247566, "rank": 0.0} +{"step": 340, "steps": 500, "lr": 2e-05, "grad_norm": 0.3330762982368469, "loss_scale": 4.0, "retries": 0, "seconds": 574.1, "max_memory_gb": 15.41, "ce": 0.1850429346668534, "rank": 0.0} +{"step": 341, "steps": 500, "lr": 2e-05, "grad_norm": 1.650132656097412, "loss_scale": 4.0, "retries": 0, "seconds": 576.1, "max_memory_gb": 15.41, "ce": 0.20073372732440475, "rank": 0.0} +{"step": 342, "steps": 500, "lr": 2e-05, "grad_norm": 0.9212096333503723, "loss_scale": 4.0, "retries": 0, "seconds": 577.9, "max_memory_gb": 15.41, "ce": 0.45369564059365075, "rank": 0.0} +{"step": 343, "steps": 500, "lr": 2e-05, "grad_norm": 0.384927362203598, "loss_scale": 4.0, "retries": 0, "seconds": 579.5, "max_memory_gb": 15.41, "ce": 0.07752057208563201, "rank": 0.0} +{"step": 344, "steps": 500, "lr": 2e-05, "grad_norm": 0.23761571943759918, "loss_scale": 4.0, "retries": 0, "seconds": 581.2, "max_memory_gb": 15.41, "ce": 0.03714589404989965, "rank": 0.0} +{"step": 345, "steps": 500, "lr": 2e-05, "grad_norm": 0.16498436033725739, "loss_scale": 4.0, "retries": 0, "seconds": 582.9, "max_memory_gb": 15.41, "ce": 0.19187641609460115, "rank": 0.0} +{"step": 346, "steps": 500, "lr": 2e-05, "grad_norm": 0.4366649389266968, "loss_scale": 4.0, "retries": 0, "seconds": 584.9, "max_memory_gb": 15.41, "ce": 0.2257594161783345, "rank": 0.0} +{"step": 347, "steps": 500, "lr": 2e-05, "grad_norm": 0.8625035881996155, "loss_scale": 4.0, "retries": 0, "seconds": 586.8, "max_memory_gb": 15.41, "ce": 0.18887452405760996, "rank": 0.0} +{"step": 348, "steps": 500, "lr": 2e-05, "grad_norm": 0.3284270167350769, "loss_scale": 4.0, "retries": 0, "seconds": 588.7, "max_memory_gb": 15.41, "ce": 0.03499294593348168, "rank": 0.0} +{"step": 349, "steps": 500, "lr": 2e-05, "grad_norm": 0.21506693959236145, "loss_scale": 4.0, "retries": 0, "seconds": 590.7, "max_memory_gb": 15.41, "ce": 0.12310423445887864, "rank": 0.0} +{"step": 350, "steps": 500, "lr": 2e-05, "grad_norm": 0.3196265995502472, "loss_scale": 4.0, "retries": 0, "seconds": 592.6, "max_memory_gb": 15.41, "ce": 0.1505851208930835, "rank": 0.0} +{"step": 351, "steps": 500, "lr": 2e-05, "grad_norm": 0.5752577781677246, "loss_scale": 4.0, "retries": 0, "seconds": 594.4, "max_memory_gb": 15.41, "ce": 0.08134914917172864, "rank": 0.0} +{"step": 352, "steps": 500, "lr": 2e-05, "grad_norm": 0.9305179715156555, "loss_scale": 4.0, "retries": 0, "seconds": 596.3, "max_memory_gb": 15.41, "ce": 0.3007730600074865, "rank": 0.0} +{"step": 353, "steps": 500, "lr": 2e-05, "grad_norm": 0.3661205470561981, "loss_scale": 4.0, "retries": 0, "seconds": 598.0, "max_memory_gb": 15.41, "ce": 0.12471611619548639, "rank": 0.0} +{"step": 354, "steps": 500, "lr": 2e-05, "grad_norm": 1.3937934637069702, "loss_scale": 4.0, "retries": 0, "seconds": 599.6, "max_memory_gb": 15.41, "ce": 0.23336534667760134, "rank": 0.0} +{"step": 355, "steps": 500, "lr": 2e-05, "grad_norm": 0.8129013180732727, "loss_scale": 4.0, "retries": 0, "seconds": 601.4, "max_memory_gb": 15.41, "ce": 0.14078254911873955, "rank": 0.0} +{"step": 356, "steps": 500, "lr": 2e-05, "grad_norm": 0.31168901920318604, "loss_scale": 4.0, "retries": 0, "seconds": 603.2, "max_memory_gb": 15.41, "ce": 0.05941389514191542, "rank": 0.0} +{"step": 357, "steps": 500, "lr": 2e-05, "grad_norm": 0.24553881585597992, "loss_scale": 4.0, "retries": 0, "seconds": 605.0, "max_memory_gb": 15.41, "ce": 0.06602434636442922, "rank": 0.0} +{"step": 358, "steps": 500, "lr": 2e-05, "grad_norm": 0.37474703788757324, "loss_scale": 4.0, "retries": 0, "seconds": 606.5, "max_memory_gb": 15.41, "ce": 0.23139915570936864, "rank": 0.0} +{"step": 359, "steps": 500, "lr": 2e-05, "grad_norm": 0.2987470328807831, "loss_scale": 4.0, "retries": 0, "seconds": 608.2, "max_memory_gb": 15.41, "ce": 0.07663466489611892, "rank": 0.0} +{"step": 360, "steps": 500, "lr": 2e-05, "grad_norm": 0.4181215167045593, "loss_scale": 4.0, "retries": 0, "seconds": 609.9, "max_memory_gb": 15.41, "ce": 0.07509255103650503, "rank": 0.1396997570991516} +{"step": 361, "steps": 500, "lr": 2e-05, "grad_norm": 0.7504790425300598, "loss_scale": 4.0, "retries": 0, "seconds": 611.5, "max_memory_gb": 15.41, "ce": 0.1566872704133857, "rank": 0.0} +{"step": 362, "steps": 500, "lr": 2e-05, "grad_norm": 2.2099785804748535, "loss_scale": 4.0, "retries": 0, "seconds": 613.2, "max_memory_gb": 15.41, "ce": 0.27486451227014186, "rank": 0.0} +{"step": 363, "steps": 500, "lr": 2e-05, "grad_norm": 0.2944051921367645, "loss_scale": 4.0, "retries": 0, "seconds": 614.9, "max_memory_gb": 15.41, "ce": 0.09085206560848746, "rank": 0.0} +{"step": 364, "steps": 500, "lr": 2e-05, "grad_norm": 0.6899276375770569, "loss_scale": 4.0, "retries": 0, "seconds": 616.4, "max_memory_gb": 15.41, "ce": 0.07647991205885774, "rank": 0.11153775453567505} +{"step": 365, "steps": 500, "lr": 2e-05, "grad_norm": 0.3633418679237366, "loss_scale": 4.0, "retries": 0, "seconds": 618.1, "max_memory_gb": 15.41, "ce": 0.14435079961185693, "rank": 0.0} +{"step": 366, "steps": 500, "lr": 2e-05, "grad_norm": 0.9245025515556335, "loss_scale": 4.0, "retries": 0, "seconds": 619.8, "max_memory_gb": 15.41, "ce": 0.13681534658826422, "rank": 0.0} +{"step": 367, "steps": 500, "lr": 2e-05, "grad_norm": 0.28392457962036133, "loss_scale": 4.0, "retries": 0, "seconds": 621.4, "max_memory_gb": 15.41, "ce": 0.12520018216309836, "rank": 0.0} +{"step": 368, "steps": 500, "lr": 2e-05, "grad_norm": 0.750924825668335, "loss_scale": 4.0, "retries": 0, "seconds": 623.1, "max_memory_gb": 15.41, "ce": 0.04616284977237228, "rank": 0.0} +{"step": 369, "steps": 500, "lr": 2e-05, "grad_norm": 0.2662236988544464, "loss_scale": 4.0, "retries": 0, "seconds": 624.8, "max_memory_gb": 15.41, "ce": 0.06444431228737812, "rank": 0.1143178790807724} +{"step": 370, "steps": 500, "lr": 2e-05, "grad_norm": 0.7183130979537964, "loss_scale": 4.0, "retries": 0, "seconds": 626.5, "max_memory_gb": 15.41, "ce": 0.3303517433232628, "rank": 0.0} +{"step": 371, "steps": 500, "lr": 2e-05, "grad_norm": 1.2212188243865967, "loss_scale": 4.0, "retries": 0, "seconds": 628.3, "max_memory_gb": 15.41, "ce": 0.08601913484744728, "rank": 0.0} +{"step": 372, "steps": 500, "lr": 2e-05, "grad_norm": 0.20532824099063873, "loss_scale": 4.0, "retries": 0, "seconds": 630.0, "max_memory_gb": 15.41, "ce": 0.04340533952199621, "rank": 0.0} +{"step": 373, "steps": 500, "lr": 2e-05, "grad_norm": 0.26658111810684204, "loss_scale": 4.0, "retries": 0, "seconds": 631.7, "max_memory_gb": 15.41, "ce": 0.10321469872724265, "rank": 0.0} +{"step": 374, "steps": 500, "lr": 2e-05, "grad_norm": 0.3812369406223297, "loss_scale": 4.0, "retries": 0, "seconds": 633.4, "max_memory_gb": 15.41, "ce": 0.1390850728857913, "rank": 0.0} +{"step": 375, "steps": 500, "lr": 2e-05, "grad_norm": 0.3246048390865326, "loss_scale": 4.0, "retries": 0, "seconds": 635.1, "max_memory_gb": 15.41, "ce": 0.16842491865099873, "rank": 0.0} +{"step": 376, "steps": 500, "lr": 2e-05, "grad_norm": 0.3480057716369629, "loss_scale": 4.0, "retries": 0, "seconds": 636.8, "max_memory_gb": 15.41, "ce": 0.09363887573272223, "rank": 0.0} +{"step": 377, "steps": 500, "lr": 2e-05, "grad_norm": 0.244543194770813, "loss_scale": 4.0, "retries": 0, "seconds": 638.3, "max_memory_gb": 15.41, "ce": 0.09473076739232056, "rank": 0.0} +{"step": 378, "steps": 500, "lr": 2e-05, "grad_norm": 0.5137525200843811, "loss_scale": 4.0, "retries": 0, "seconds": 640.0, "max_memory_gb": 15.41, "ce": 0.0887544898178021, "rank": 0.0} +{"step": 379, "steps": 500, "lr": 2e-05, "grad_norm": 0.2964165210723877, "loss_scale": 4.0, "retries": 0, "seconds": 641.7, "max_memory_gb": 15.41, "ce": 0.13124183141917456, "rank": 0.0} +{"step": 380, "steps": 500, "lr": 2e-05, "grad_norm": 0.628996729850769, "loss_scale": 4.0, "retries": 0, "seconds": 643.3, "max_memory_gb": 15.41, "ce": 0.07779736772499746, "rank": 0.0} +{"step": 381, "steps": 500, "lr": 2e-05, "grad_norm": 0.4351450204849243, "loss_scale": 4.0, "retries": 0, "seconds": 645.0, "max_memory_gb": 15.41, "ce": 0.2793578083437751, "rank": 0.0} +{"step": 382, "steps": 500, "lr": 2e-05, "grad_norm": 0.8696860671043396, "loss_scale": 4.0, "retries": 0, "seconds": 646.6, "max_memory_gb": 15.41, "ce": 0.30531068888376467, "rank": 0.0} +{"step": 383, "steps": 500, "lr": 2e-05, "grad_norm": 0.2732827663421631, "loss_scale": 4.0, "retries": 0, "seconds": 648.3, "max_memory_gb": 15.41, "ce": 0.09931194063210569, "rank": 0.0} +{"step": 384, "steps": 500, "lr": 2e-05, "grad_norm": 0.20474675297737122, "loss_scale": 4.0, "retries": 0, "seconds": 649.9, "max_memory_gb": 15.41, "ce": 0.04524718894390389, "rank": 0.0} +{"step": 385, "steps": 500, "lr": 2e-05, "grad_norm": 0.23890003561973572, "loss_scale": 4.0, "retries": 0, "seconds": 651.6, "max_memory_gb": 15.41, "ce": 0.18422909203218296, "rank": 0.0} +{"step": 386, "steps": 500, "lr": 2e-05, "grad_norm": 0.23312467336654663, "loss_scale": 4.0, "retries": 0, "seconds": 653.3, "max_memory_gb": 15.41, "ce": 0.06321940754423849, "rank": 0.0} +{"step": 387, "steps": 500, "lr": 2e-05, "grad_norm": 0.2430400252342224, "loss_scale": 4.0, "retries": 0, "seconds": 654.9, "max_memory_gb": 15.41, "ce": 0.06784713291199296, "rank": 0.0} +{"step": 388, "steps": 500, "lr": 2e-05, "grad_norm": 0.3037465810775757, "loss_scale": 4.0, "retries": 0, "seconds": 656.5, "max_memory_gb": 15.41, "ce": 0.06868968068738468, "rank": 0.0} +{"step": 389, "steps": 500, "lr": 2e-05, "grad_norm": 0.3253004550933838, "loss_scale": 4.0, "retries": 0, "seconds": 658.2, "max_memory_gb": 15.41, "ce": 0.024998007913382025, "rank": 0.10722197592258453} +{"step": 390, "steps": 500, "lr": 2e-05, "grad_norm": 1.3429319858551025, "loss_scale": 4.0, "retries": 0, "seconds": 659.9, "max_memory_gb": 15.41, "ce": 0.08589231498262961, "rank": 0.0} +{"step": 391, "steps": 500, "lr": 2e-05, "grad_norm": 1.908511757850647, "loss_scale": 4.0, "retries": 0, "seconds": 661.6, "max_memory_gb": 15.41, "ce": 0.3048527172941249, "rank": 0.0} +{"step": 392, "steps": 500, "lr": 2e-05, "grad_norm": 0.16081824898719788, "loss_scale": 4.0, "retries": 0, "seconds": 663.3, "max_memory_gb": 15.41, "ce": 0.013103933451930061, "rank": 0.07471679151058197} +{"step": 393, "steps": 500, "lr": 2e-05, "grad_norm": 1.102465271949768, "loss_scale": 4.0, "retries": 0, "seconds": 664.9, "max_memory_gb": 15.41, "ce": 0.16887899948233098, "rank": 0.0} +{"step": 394, "steps": 500, "lr": 2e-05, "grad_norm": 0.26514238119125366, "loss_scale": 4.0, "retries": 0, "seconds": 666.6, "max_memory_gb": 15.41, "ce": 0.1046229437160946, "rank": 0.0} +{"step": 395, "steps": 500, "lr": 2e-05, "grad_norm": 0.366022527217865, "loss_scale": 4.0, "retries": 0, "seconds": 668.2, "max_memory_gb": 15.41, "ce": 0.04890149989660131, "rank": 0.0} +{"step": 396, "steps": 500, "lr": 2e-05, "grad_norm": 0.9737012386322021, "loss_scale": 4.0, "retries": 0, "seconds": 669.9, "max_memory_gb": 15.41, "ce": 0.06426454327629472, "rank": 0.0} +{"step": 397, "steps": 500, "lr": 2e-05, "grad_norm": 0.3054809868335724, "loss_scale": 4.0, "retries": 0, "seconds": 671.6, "max_memory_gb": 15.41, "ce": 0.13536362070590258, "rank": 0.0} +{"step": 398, "steps": 500, "lr": 2e-05, "grad_norm": 0.348933607339859, "loss_scale": 4.0, "retries": 0, "seconds": 673.3, "max_memory_gb": 15.41, "ce": 0.07683723834270495, "rank": 0.0} +{"step": 399, "steps": 500, "lr": 2e-05, "grad_norm": 0.5834789872169495, "loss_scale": 4.0, "retries": 0, "seconds": 674.9, "max_memory_gb": 15.41, "ce": 0.17288456666574348, "rank": 0.0} +{"step": 400, "steps": 500, "lr": 2e-05, "grad_norm": 2.1401309967041016, "loss_scale": 4.0, "retries": 0, "seconds": 676.4, "max_memory_gb": 15.41, "ce": 0.11003925038312445, "rank": 0.0} +{"step": 401, "steps": 500, "lr": 2e-05, "grad_norm": 2.8366806507110596, "loss_scale": 4.0, "retries": 0, "seconds": 678.2, "max_memory_gb": 15.41, "ce": 0.1927795511764998, "rank": 0.0} +{"step": 402, "steps": 500, "lr": 2e-05, "grad_norm": 0.3083009719848633, "loss_scale": 4.0, "retries": 0, "seconds": 679.9, "max_memory_gb": 15.41, "ce": 0.11924150502454722, "rank": 0.09296762198209763} +{"step": 403, "steps": 500, "lr": 2e-05, "grad_norm": 0.16005969047546387, "loss_scale": 4.0, "retries": 0, "seconds": 681.6, "max_memory_gb": 15.41, "ce": 0.018203947187430458, "rank": 0.0} +{"step": 404, "steps": 500, "lr": 2e-05, "grad_norm": 0.32858380675315857, "loss_scale": 4.0, "retries": 0, "seconds": 683.4, "max_memory_gb": 15.41, "ce": 0.14741705908090807, "rank": 0.0} +{"step": 405, "steps": 500, "lr": 2e-05, "grad_norm": 0.43704771995544434, "loss_scale": 4.0, "retries": 0, "seconds": 685.1, "max_memory_gb": 15.41, "ce": 0.054301379481330514, "rank": 0.0} +{"step": 406, "steps": 500, "lr": 2e-05, "grad_norm": 0.4093169867992401, "loss_scale": 4.0, "retries": 0, "seconds": 686.8, "max_memory_gb": 15.41, "ce": 0.14874357120424975, "rank": 0.0} +{"step": 407, "steps": 500, "lr": 2e-05, "grad_norm": 1.0559005737304688, "loss_scale": 4.0, "retries": 0, "seconds": 688.5, "max_memory_gb": 15.41, "ce": 0.629436885821633, "rank": 0.12790340185165405} +{"step": 408, "steps": 500, "lr": 2e-05, "grad_norm": 0.6072866320610046, "loss_scale": 4.0, "retries": 0, "seconds": 690.3, "max_memory_gb": 15.41, "ce": 0.2662196591554675, "rank": 0.03448981046676636} +{"step": 409, "steps": 500, "lr": 2e-05, "grad_norm": 0.2995006740093231, "loss_scale": 4.0, "retries": 0, "seconds": 691.9, "max_memory_gb": 15.41, "ce": 0.15138545233639888, "rank": 0.0} +{"step": 410, "steps": 500, "lr": 2e-05, "grad_norm": 1.1128500699996948, "loss_scale": 4.0, "retries": 0, "seconds": 693.6, "max_memory_gb": 15.41, "ce": 0.08170381268428173, "rank": 0.0} +{"step": 411, "steps": 500, "lr": 2e-05, "grad_norm": 3.6261470317840576, "loss_scale": 4.0, "retries": 0, "seconds": 695.2, "max_memory_gb": 15.41, "ce": 0.19939136544417124, "rank": 0.0} +{"step": 412, "steps": 500, "lr": 2e-05, "grad_norm": 0.222580224275589, "loss_scale": 4.0, "retries": 0, "seconds": 696.9, "max_memory_gb": 15.41, "ce": 0.1519411021581618, "rank": 0.12471968680620193} +{"step": 413, "steps": 500, "lr": 2e-05, "grad_norm": 0.36790212988853455, "loss_scale": 4.0, "retries": 0, "seconds": 698.5, "max_memory_gb": 15.41, "ce": 0.06304989519048831, "rank": 0.0} +{"step": 414, "steps": 500, "lr": 2e-05, "grad_norm": 0.5827904343605042, "loss_scale": 4.0, "retries": 0, "seconds": 700.3, "max_memory_gb": 15.41, "ce": 0.20440146484907018, "rank": 0.0} +{"step": 415, "steps": 500, "lr": 2e-05, "grad_norm": 1.5142279863357544, "loss_scale": 4.0, "retries": 0, "seconds": 702.1, "max_memory_gb": 15.41, "ce": 0.3278832014912041, "rank": 0.0} +{"step": 416, "steps": 500, "lr": 2e-05, "grad_norm": 0.2578657269477844, "loss_scale": 4.0, "retries": 0, "seconds": 704.1, "max_memory_gb": 15.41, "ce": 0.035237553987826686, "rank": 0.0} +{"step": 417, "steps": 500, "lr": 2e-05, "grad_norm": 0.4401823580265045, "loss_scale": 4.0, "retries": 0, "seconds": 706.0, "max_memory_gb": 15.41, "ce": 0.314677765883971, "rank": 0.12823650240898132} +{"step": 418, "steps": 500, "lr": 2e-05, "grad_norm": 0.24205359816551208, "loss_scale": 4.0, "retries": 0, "seconds": 707.9, "max_memory_gb": 15.41, "ce": 0.06166604950703913, "rank": 0.0} +{"step": 419, "steps": 500, "lr": 2e-05, "grad_norm": 0.2165592610836029, "loss_scale": 4.0, "retries": 0, "seconds": 709.7, "max_memory_gb": 15.41, "ce": 0.08509818566381, "rank": 0.0} +{"step": 420, "steps": 500, "lr": 2e-05, "grad_norm": 1.116714596748352, "loss_scale": 4.0, "retries": 0, "seconds": 711.5, "max_memory_gb": 15.41, "ce": 0.2869284882181091, "rank": 0.1273089200258255} +{"step": 421, "steps": 500, "lr": 2e-05, "grad_norm": 0.2526918053627014, "loss_scale": 4.0, "retries": 0, "seconds": 713.3, "max_memory_gb": 15.41, "ce": 0.05686267739656614, "rank": 0.10942870378494263} +{"step": 422, "steps": 500, "lr": 2e-05, "grad_norm": 1.255109429359436, "loss_scale": 4.0, "retries": 0, "seconds": 715.1, "max_memory_gb": 15.41, "ce": 0.22677773233226617, "rank": 0.0} +{"step": 423, "steps": 500, "lr": 2e-05, "grad_norm": 2.2263927459716797, "loss_scale": 4.0, "retries": 0, "seconds": 716.8, "max_memory_gb": 15.41, "ce": 0.3979328506720776, "rank": 0.0} +{"step": 424, "steps": 500, "lr": 2e-05, "grad_norm": 0.28895819187164307, "loss_scale": 4.0, "retries": 0, "seconds": 718.5, "max_memory_gb": 15.41, "ce": 0.06298359994252678, "rank": 0.0} +{"step": 425, "steps": 500, "lr": 2e-05, "grad_norm": 0.40544313192367554, "loss_scale": 4.0, "retries": 0, "seconds": 720.2, "max_memory_gb": 15.41, "ce": 0.08986947218909336, "rank": 0.0} +{"step": 426, "steps": 500, "lr": 2e-05, "grad_norm": 0.7084959745407104, "loss_scale": 4.0, "retries": 0, "seconds": 721.9, "max_memory_gb": 15.41, "ce": 0.20235706599487457, "rank": 0.0} +{"step": 427, "steps": 500, "lr": 2e-05, "grad_norm": 0.17241649329662323, "loss_scale": 4.0, "retries": 0, "seconds": 723.6, "max_memory_gb": 15.41, "ce": 0.05338631939594052, "rank": 0.0} +{"step": 428, "steps": 500, "lr": 2e-05, "grad_norm": 0.9392373561859131, "loss_scale": 4.0, "retries": 0, "seconds": 725.3, "max_memory_gb": 15.41, "ce": 0.05202159971668152, "rank": 0.0} +{"step": 429, "steps": 500, "lr": 2e-05, "grad_norm": 0.23856863379478455, "loss_scale": 4.0, "retries": 0, "seconds": 727.0, "max_memory_gb": 15.41, "ce": 0.02251080435962649, "rank": 0.0} +{"step": 430, "steps": 500, "lr": 2e-05, "grad_norm": 0.3878486454486847, "loss_scale": 4.0, "retries": 0, "seconds": 728.7, "max_memory_gb": 15.41, "ce": 0.09737898790626787, "rank": 0.0} +{"step": 431, "steps": 500, "lr": 2e-05, "grad_norm": 4.297839641571045, "loss_scale": 4.0, "retries": 0, "seconds": 730.4, "max_memory_gb": 15.41, "ce": 0.2717334824901627, "rank": 0.0} +{"step": 432, "steps": 500, "lr": 2e-05, "grad_norm": 0.47195643186569214, "loss_scale": 4.0, "retries": 0, "seconds": 732.1, "max_memory_gb": 15.41, "ce": 0.07147425341827329, "rank": 0.0} +{"step": 433, "steps": 500, "lr": 2e-05, "grad_norm": 0.38453954458236694, "loss_scale": 4.0, "retries": 0, "seconds": 733.8, "max_memory_gb": 15.41, "ce": 0.08408581560070161, "rank": 0.0} +{"step": 434, "steps": 500, "lr": 2e-05, "grad_norm": 0.11751726269721985, "loss_scale": 4.0, "retries": 0, "seconds": 735.5, "max_memory_gb": 15.41, "ce": 0.014728394511621445, "rank": 0.0} +{"step": 435, "steps": 500, "lr": 2e-05, "grad_norm": 0.591284453868866, "loss_scale": 4.0, "retries": 0, "seconds": 737.2, "max_memory_gb": 15.41, "ce": 0.15101601518836105, "rank": 0.0} +{"step": 436, "steps": 500, "lr": 2e-05, "grad_norm": 0.5204747915267944, "loss_scale": 4.0, "retries": 0, "seconds": 738.9, "max_memory_gb": 15.41, "ce": 0.11052319192094728, "rank": 0.0} +{"step": 437, "steps": 500, "lr": 2e-05, "grad_norm": 1.8074215650558472, "loss_scale": 4.0, "retries": 0, "seconds": 740.6, "max_memory_gb": 15.41, "ce": 0.13334343383758096, "rank": 0.09151853621006012} +{"step": 438, "steps": 500, "lr": 2e-05, "grad_norm": 0.42562592029571533, "loss_scale": 4.0, "retries": 0, "seconds": 742.2, "max_memory_gb": 15.41, "ce": 0.0723456107207312, "rank": 0.0} +{"step": 439, "steps": 500, "lr": 2e-05, "grad_norm": 0.9810224771499634, "loss_scale": 4.0, "retries": 0, "seconds": 743.9, "max_memory_gb": 15.41, "ce": 0.1352838943603274, "rank": 0.0} +{"step": 440, "steps": 500, "lr": 2e-05, "grad_norm": 0.26718446612358093, "loss_scale": 4.0, "retries": 0, "seconds": 745.6, "max_memory_gb": 15.41, "ce": 0.05871709735947661, "rank": 0.0} +{"step": 441, "steps": 500, "lr": 2e-05, "grad_norm": 0.8530780076980591, "loss_scale": 4.0, "retries": 0, "seconds": 747.1, "max_memory_gb": 15.41, "ce": 0.18985370862355921, "rank": 0.0} +{"step": 442, "steps": 500, "lr": 2e-05, "grad_norm": 0.5343230962753296, "loss_scale": 4.0, "retries": 0, "seconds": 748.7, "max_memory_gb": 15.41, "ce": 0.0714594270903035, "rank": 0.0} +{"step": 443, "steps": 500, "lr": 2e-05, "grad_norm": 0.37860041856765747, "loss_scale": 4.0, "retries": 0, "seconds": 750.5, "max_memory_gb": 15.41, "ce": 0.07611000974429771, "rank": 0.0} +{"step": 444, "steps": 500, "lr": 2e-05, "grad_norm": 0.3618663549423218, "loss_scale": 4.0, "retries": 0, "seconds": 752.2, "max_memory_gb": 15.41, "ce": 0.09231206670665415, "rank": 0.0} +{"step": 445, "steps": 500, "lr": 2e-05, "grad_norm": 0.923911452293396, "loss_scale": 4.0, "retries": 0, "seconds": 753.9, "max_memory_gb": 15.41, "ce": 0.36893121217144653, "rank": 0.06623844057321548} +{"step": 446, "steps": 500, "lr": 2e-05, "grad_norm": 5.185521125793457, "loss_scale": 4.0, "retries": 0, "seconds": 755.6, "max_memory_gb": 15.41, "ce": 0.16072525328490883, "rank": 0.2656661570072174} +{"step": 447, "steps": 500, "lr": 2e-05, "grad_norm": 0.5584444403648376, "loss_scale": 4.0, "retries": 0, "seconds": 757.3, "max_memory_gb": 15.41, "ce": 0.18771625716908602, "rank": 0.0} +{"step": 448, "steps": 500, "lr": 2e-05, "grad_norm": 1.214926838874817, "loss_scale": 4.0, "retries": 0, "seconds": 758.9, "max_memory_gb": 15.41, "ce": 0.2084394364246691, "rank": 0.0} +{"step": 449, "steps": 500, "lr": 2e-05, "grad_norm": 0.6156787276268005, "loss_scale": 4.0, "retries": 0, "seconds": 760.6, "max_memory_gb": 15.41, "ce": 0.02561543930278276, "rank": 0.0} +{"step": 450, "steps": 500, "lr": 2e-05, "grad_norm": 0.516779363155365, "loss_scale": 4.0, "retries": 0, "seconds": 762.3, "max_memory_gb": 15.41, "ce": 0.03097687703666452, "rank": 0.0} +{"step": 451, "steps": 500, "lr": 2e-05, "grad_norm": 0.6813018918037415, "loss_scale": 4.0, "retries": 0, "seconds": 764.0, "max_memory_gb": 15.41, "ce": 0.048308126359188464, "rank": 0.0} +{"step": 452, "steps": 500, "lr": 2e-05, "grad_norm": 1.1700809001922607, "loss_scale": 4.0, "retries": 0, "seconds": 765.7, "max_memory_gb": 15.41, "ce": 0.17264934400463972, "rank": 0.0} +{"step": 453, "steps": 500, "lr": 2e-05, "grad_norm": 0.632738471031189, "loss_scale": 4.0, "retries": 0, "seconds": 767.4, "max_memory_gb": 15.41, "ce": 0.0453070796065731, "rank": 0.0} +{"step": 454, "steps": 500, "lr": 2e-05, "grad_norm": 1.0550838708877563, "loss_scale": 4.0, "retries": 0, "seconds": 769.1, "max_memory_gb": 15.41, "ce": 0.14388912324648118, "rank": 0.0} +{"step": 455, "steps": 500, "lr": 2e-05, "grad_norm": 1.88508939743042, "loss_scale": 4.0, "retries": 0, "seconds": 770.9, "max_memory_gb": 15.41, "ce": 0.13406299310099712, "rank": 0.0} +{"step": 456, "steps": 500, "lr": 2e-05, "grad_norm": 0.25269439816474915, "loss_scale": 4.0, "retries": 0, "seconds": 772.8, "max_memory_gb": 15.41, "ce": 0.02467686455565854, "rank": 0.0} +{"step": 457, "steps": 500, "lr": 2e-05, "grad_norm": 0.45697611570358276, "loss_scale": 4.0, "retries": 0, "seconds": 774.7, "max_memory_gb": 15.41, "ce": 0.060094224085332826, "rank": 0.08186381310224533} +{"step": 458, "steps": 500, "lr": 2e-05, "grad_norm": 0.6103665828704834, "loss_scale": 4.0, "retries": 0, "seconds": 776.6, "max_memory_gb": 15.41, "ce": 0.1634914623318764, "rank": 0.0} +{"step": 459, "steps": 500, "lr": 2e-05, "grad_norm": 1.5493130683898926, "loss_scale": 4.0, "retries": 0, "seconds": 778.5, "max_memory_gb": 15.41, "ce": 0.141093160355922, "rank": 0.0} +{"step": 460, "steps": 500, "lr": 2e-05, "grad_norm": 0.3664776384830475, "loss_scale": 4.0, "retries": 0, "seconds": 780.4, "max_memory_gb": 15.41, "ce": 0.03135563785326667, "rank": 0.0} +{"step": 461, "steps": 500, "lr": 2e-05, "grad_norm": 0.40235742926597595, "loss_scale": 4.0, "retries": 0, "seconds": 782.3, "max_memory_gb": 15.41, "ce": 0.2351570306782378, "rank": 0.0} +{"step": 462, "steps": 500, "lr": 2e-05, "grad_norm": 0.17713195085525513, "loss_scale": 4.0, "retries": 0, "seconds": 784.2, "max_memory_gb": 15.41, "ce": 0.024017898351303302, "rank": 0.0} +{"step": 463, "steps": 500, "lr": 2e-05, "grad_norm": 0.17401111125946045, "loss_scale": 4.0, "retries": 0, "seconds": 786.0, "max_memory_gb": 15.41, "ce": 0.022391057553249993, "rank": 0.0} +{"step": 464, "steps": 500, "lr": 2e-05, "grad_norm": 0.1897556334733963, "loss_scale": 4.0, "retries": 0, "seconds": 788.0, "max_memory_gb": 15.41, "ce": 0.01779891460955696, "rank": 0.0} +{"step": 465, "steps": 500, "lr": 2e-05, "grad_norm": 1.2127454280853271, "loss_scale": 4.0, "retries": 0, "seconds": 789.9, "max_memory_gb": 15.41, "ce": 0.2150165729035507, "rank": 0.0} +{"step": 466, "steps": 500, "lr": 2e-05, "grad_norm": 0.7087408900260925, "loss_scale": 4.0, "retries": 0, "seconds": 791.9, "max_memory_gb": 15.41, "ce": 0.028809579056542134, "rank": 0.0} +{"step": 467, "steps": 500, "lr": 2e-05, "grad_norm": 0.28920978307724, "loss_scale": 4.0, "retries": 0, "seconds": 793.8, "max_memory_gb": 15.41, "ce": 0.04209253274893854, "rank": 0.0} +{"step": 468, "steps": 500, "lr": 2e-05, "grad_norm": 0.19762763381004333, "loss_scale": 4.0, "retries": 0, "seconds": 795.6, "max_memory_gb": 15.41, "ce": 0.021054198675756197, "rank": 0.0} +{"step": 469, "steps": 500, "lr": 2e-05, "grad_norm": 2.9288647174835205, "loss_scale": 4.0, "retries": 0, "seconds": 797.4, "max_memory_gb": 15.41, "ce": 0.17058984396680898, "rank": 0.0} +{"step": 470, "steps": 500, "lr": 2e-05, "grad_norm": 1.147855520248413, "loss_scale": 4.0, "retries": 0, "seconds": 799.0, "max_memory_gb": 15.41, "ce": 0.09548167620232562, "rank": 0.0} +{"step": 471, "steps": 500, "lr": 2e-05, "grad_norm": 0.7609191536903381, "loss_scale": 4.0, "retries": 0, "seconds": 800.8, "max_memory_gb": 15.41, "ce": 0.16375779410827818, "rank": 0.0} +{"step": 472, "steps": 500, "lr": 2e-05, "grad_norm": 0.4205853343009949, "loss_scale": 4.0, "retries": 0, "seconds": 802.4, "max_memory_gb": 15.41, "ce": 0.040380292201462, "rank": 0.0} +{"step": 473, "steps": 500, "lr": 2e-05, "grad_norm": 0.6043763756752014, "loss_scale": 4.0, "retries": 0, "seconds": 804.1, "max_memory_gb": 15.41, "ce": 0.029121519502950832, "rank": 0.1318056583404541} +{"step": 474, "steps": 500, "lr": 2e-05, "grad_norm": 0.45569509267807007, "loss_scale": 4.0, "retries": 0, "seconds": 805.9, "max_memory_gb": 15.41, "ce": 0.1016920401743846, "rank": 0.0} +{"step": 475, "steps": 500, "lr": 2e-05, "grad_norm": 0.8439630270004272, "loss_scale": 4.0, "retries": 0, "seconds": 807.8, "max_memory_gb": 15.41, "ce": 0.14984418598760385, "rank": 0.0} +{"step": 476, "steps": 500, "lr": 2e-05, "grad_norm": 0.20559827983379364, "loss_scale": 4.0, "retries": 0, "seconds": 809.6, "max_memory_gb": 15.41, "ce": 0.013229720911112963, "rank": 0.0} +{"step": 477, "steps": 500, "lr": 2e-05, "grad_norm": 1.7407149076461792, "loss_scale": 4.0, "retries": 0, "seconds": 811.5, "max_memory_gb": 15.41, "ce": 0.09037566963115751, "rank": 0.0} +{"step": 478, "steps": 500, "lr": 2e-05, "grad_norm": 0.4527525305747986, "loss_scale": 4.0, "retries": 0, "seconds": 813.3, "max_memory_gb": 15.41, "ce": 0.08087681361939758, "rank": 0.0} +{"step": 479, "steps": 500, "lr": 2e-05, "grad_norm": 3.7222023010253906, "loss_scale": 4.0, "retries": 0, "seconds": 815.2, "max_memory_gb": 15.41, "ce": 0.35762841607720475, "rank": 0.026607483625411987} +{"step": 480, "steps": 500, "lr": 2e-05, "grad_norm": 0.5173050761222839, "loss_scale": 4.0, "retries": 0, "seconds": 817.1, "max_memory_gb": 15.41, "ce": 0.16067100684176694, "rank": 0.0} +{"step": 481, "steps": 500, "lr": 2e-05, "grad_norm": 0.20424912869930267, "loss_scale": 4.0, "retries": 0, "seconds": 818.9, "max_memory_gb": 15.41, "ce": 0.035117736468237126, "rank": 0.0} +{"step": 482, "steps": 500, "lr": 2e-05, "grad_norm": 0.4452568590641022, "loss_scale": 4.0, "retries": 0, "seconds": 820.6, "max_memory_gb": 15.41, "ce": 0.09049005291308276, "rank": 0.0} +{"step": 483, "steps": 500, "lr": 2e-05, "grad_norm": 1.7937902212142944, "loss_scale": 4.0, "retries": 0, "seconds": 822.2, "max_memory_gb": 15.41, "ce": 0.3610633196553863, "rank": 0.0} +{"step": 484, "steps": 500, "lr": 2e-05, "grad_norm": 1.193367838859558, "loss_scale": 4.0, "retries": 0, "seconds": 823.9, "max_memory_gb": 15.41, "ce": 0.4308824211702813, "rank": 0.0} +{"step": 485, "steps": 500, "lr": 2e-05, "grad_norm": 0.6336275935173035, "loss_scale": 4.0, "retries": 0, "seconds": 825.6, "max_memory_gb": 15.41, "ce": 0.08777482675213832, "rank": 0.0} +{"step": 486, "steps": 500, "lr": 2e-05, "grad_norm": 0.7945519089698792, "loss_scale": 4.0, "retries": 0, "seconds": 827.3, "max_memory_gb": 15.41, "ce": 0.18548933053125438, "rank": 0.0} +{"step": 487, "steps": 500, "lr": 2e-05, "grad_norm": 4.132506370544434, "loss_scale": 4.0, "retries": 0, "seconds": 828.9, "max_memory_gb": 15.41, "ce": 0.13224947456455993, "rank": 0.0} +{"step": 488, "steps": 500, "lr": 2e-05, "grad_norm": 0.4128722548484802, "loss_scale": 4.0, "retries": 0, "seconds": 830.7, "max_memory_gb": 15.41, "ce": 0.09785693666890438, "rank": 0.0} +{"step": 489, "steps": 500, "lr": 2e-05, "grad_norm": 2.383992910385132, "loss_scale": 4.0, "retries": 0, "seconds": 832.3, "max_memory_gb": 15.41, "ce": 0.2092604732424661, "rank": 0.0} +{"step": 490, "steps": 500, "lr": 2e-05, "grad_norm": 0.3140162229537964, "loss_scale": 4.0, "retries": 0, "seconds": 833.9, "max_memory_gb": 15.41, "ce": 0.050432545140211005, "rank": 0.0} +{"step": 491, "steps": 500, "lr": 2e-05, "grad_norm": 0.18794241547584534, "loss_scale": 4.0, "retries": 0, "seconds": 835.6, "max_memory_gb": 15.41, "ce": 0.10456789930321975, "rank": 0.0} +{"step": 492, "steps": 500, "lr": 2e-05, "grad_norm": 0.48016634583473206, "loss_scale": 4.0, "retries": 0, "seconds": 837.3, "max_memory_gb": 15.41, "ce": 0.02776543913932983, "rank": 0.0} +{"step": 493, "steps": 500, "lr": 2e-05, "grad_norm": 0.9163577556610107, "loss_scale": 4.0, "retries": 0, "seconds": 839.0, "max_memory_gb": 15.41, "ce": 0.0941422572650481, "rank": 0.0} +{"step": 494, "steps": 500, "lr": 2e-05, "grad_norm": 1.1105012893676758, "loss_scale": 4.0, "retries": 0, "seconds": 840.7, "max_memory_gb": 15.41, "ce": 0.08076900670130271, "rank": 0.0} +{"step": 495, "steps": 500, "lr": 2e-05, "grad_norm": 0.5031759738922119, "loss_scale": 4.0, "retries": 0, "seconds": 842.4, "max_memory_gb": 15.41, "ce": 0.04725246255293314, "rank": 0.0} +{"step": 496, "steps": 500, "lr": 2e-05, "grad_norm": 1.4092607498168945, "loss_scale": 4.0, "retries": 0, "seconds": 844.1, "max_memory_gb": 15.41, "ce": 0.049368657026207075, "rank": 0.0} +{"step": 497, "steps": 500, "lr": 2e-05, "grad_norm": 1.1789860725402832, "loss_scale": 4.0, "retries": 0, "seconds": 845.8, "max_memory_gb": 15.41, "ce": 0.13941049451932486, "rank": 0.0} +{"step": 498, "steps": 500, "lr": 2e-05, "grad_norm": 0.6875226497650146, "loss_scale": 4.0, "retries": 0, "seconds": 847.5, "max_memory_gb": 15.41, "ce": 0.14377432171204418, "rank": 0.0} +{"step": 499, "steps": 500, "lr": 2e-05, "grad_norm": 0.7890111804008484, "loss_scale": 4.0, "retries": 0, "seconds": 849.2, "max_memory_gb": 15.41, "ce": 0.1359527165277541, "rank": 0.0} +{"step": 500, "steps": 500, "lr": 2e-05, "grad_norm": 0.9545114636421204, "loss_scale": 4.0, "retries": 0, "seconds": 850.9, "max_memory_gb": 15.41, "ce": 0.2214560605607403, "rank": 0.0} diff --git a/runs/single_llava/removal/L26_s13/adapter/README.md b/runs/single_llava/removal/L26_s13/adapter/README.md new file mode 100644 index 0000000000000000000000000000000000000000..ae587920ca55972d94e38f783837f71be14f1dda --- /dev/null +++ b/runs/single_llava/removal/L26_s13/adapter/README.md @@ -0,0 +1,207 @@ +--- +base_model: /data/OPSD/assets/models/llava-1.5-7b-hf +library_name: peft +pipeline_tag: text-generation +tags: +- base_model:adapter:/data/OPSD/assets/models/llava-1.5-7b-hf +- lora +- transformers +--- + +# Model Card for Model ID + + + + + +## Model Details + +### Model Description + + + + + +- **Developed by:** [More Information Needed] +- **Funded by [optional]:** [More Information Needed] +- **Shared by [optional]:** [More Information Needed] +- **Model type:** [More Information Needed] +- **Language(s) (NLP):** [More Information Needed] +- **License:** [More Information Needed] +- **Finetuned from model [optional]:** [More Information Needed] + +### Model Sources [optional] + + + +- **Repository:** [More Information Needed] +- **Paper [optional]:** [More Information Needed] +- **Demo [optional]:** [More Information Needed] + +## Uses + + + +### Direct Use + + + +[More Information Needed] + +### Downstream Use [optional] + + + +[More Information Needed] + +### Out-of-Scope Use + + + +[More Information Needed] + +## Bias, Risks, and Limitations + + + +[More Information Needed] + +### Recommendations + + + +Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. More information needed for further recommendations. + +## How to Get Started with the Model + +Use the code below to get started with the model. + +[More Information Needed] + +## Training Details + +### Training Data + + + +[More Information Needed] + +### Training Procedure + + + +#### Preprocessing [optional] + +[More Information Needed] + + +#### Training Hyperparameters + +- **Training regime:** [More Information Needed] + +#### Speeds, Sizes, Times [optional] + + + +[More Information Needed] + +## Evaluation + + + +### Testing Data, Factors & Metrics + +#### Testing Data + + + +[More Information Needed] + +#### Factors + + + +[More Information Needed] + +#### Metrics + + + +[More Information Needed] + +### Results + +[More Information Needed] + +#### Summary + + + +## Model Examination [optional] + + + +[More Information Needed] + +## Environmental Impact + + + +Carbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700). + +- **Hardware Type:** [More Information Needed] +- **Hours used:** [More Information Needed] +- **Cloud Provider:** [More Information Needed] +- **Compute Region:** [More Information Needed] +- **Carbon Emitted:** [More Information Needed] + +## Technical Specifications [optional] + +### Model Architecture and Objective + +[More Information Needed] + +### Compute Infrastructure + +[More Information Needed] + +#### Hardware + +[More Information Needed] + +#### Software + +[More Information Needed] + +## Citation [optional] + + + +**BibTeX:** + +[More Information Needed] + +**APA:** + +[More Information Needed] + +## Glossary [optional] + + + +[More Information Needed] + +## More Information [optional] + +[More Information Needed] + +## Model Card Authors [optional] + +[More Information Needed] + +## Model Card Contact + +[More Information Needed] +### Framework versions + +- PEFT 0.20.0 \ No newline at end of file diff --git a/runs/single_llava/removal/L26_s13/adapter/adapter_config.json b/runs/single_llava/removal/L26_s13/adapter/adapter_config.json new file mode 100644 index 0000000000000000000000000000000000000000..d66f66b7856ed023732edaed2964d713118f0ac0 --- /dev/null +++ b/runs/single_llava/removal/L26_s13/adapter/adapter_config.json @@ -0,0 +1,50 @@ +{ + "alora_invocation_tokens": null, + "alpha_pattern": {}, + "arrow_config": null, + "auto_mapping": null, + "base_model_name_or_path": "/data/OPSD/assets/models/llava-1.5-7b-hf", + "bias": "none", + "corda_config": null, + "ensure_weight_tying": false, + "eva_config": null, + "exclude_modules": null, + "fan_in_fan_out": false, + "inference_mode": true, + "init_lora_weights": true, + "layer_replication": null, + "layers_pattern": null, + "layers_to_transform": null, + "loftq_config": {}, + "lora_alpha": 32, + "lora_bias": false, + "lora_dropout": 0.05, + "lora_ga_config": null, + "megatron_config": null, + "megatron_core": "megatron.core", + "modules_to_save": null, + "monteclora_config": null, + "peft_type": "LORA", + "peft_version": "0.20.0", + "qalora_group_size": 16, + "r": 16, + "rank_pattern": {}, + "revision": null, + "target_modules": [ + "model.language_model.layers.26.self_attn.v_proj", + "model.language_model.layers.26.mlp.gate_proj", + "model.language_model.layers.26.self_attn.k_proj", + "model.language_model.layers.26.mlp.up_proj", + "model.language_model.layers.26.self_attn.q_proj", + "model.language_model.layers.26.self_attn.o_proj", + "model.language_model.layers.26.mlp.down_proj" + ], + "target_parameters": null, + "task_type": "CAUSAL_LM", + "trainable_token_indices": null, + "use_bdlora": null, + "use_dora": false, + "use_qalora": false, + "use_rslora": false, + "velora_config": null +} \ No newline at end of file diff --git a/runs/single_llava/removal/L26_s13/config.json b/runs/single_llava/removal/L26_s13/config.json new file mode 100644 index 0000000000000000000000000000000000000000..61c55128547ae9976b7fbad251df9a9111f87320 --- /dev/null +++ b/runs/single_llava/removal/L26_s13/config.json @@ -0,0 +1,24 @@ +{ + "behavior": "removal", + "seed": 13, + "model": "/data/OPSD/assets/models/llava-1.5-7b-hf", + "backbone": "llava", + "train": "/data/OPSD/assets/v3/removal.jsonl", + "items": 1000, + "steps": 500, + "epochs": 2, + "steps_per_epoch": 250, + "groups_per_step": 4, + "lr": 2e-05, + "scheduler": "constant", + "warmup_steps": 0, + "final_fraction": 1.0, + "weight_decay": 0.0, + "clip": 1.0, + "initial_loss_scale": 4.0, + "adapter": { + "layer": 26, + "rank": 16, + "trainable_parameters": 1249280 + } +} \ No newline at end of file diff --git a/runs/single_llava/removal/L26_s13/steps.jsonl b/runs/single_llava/removal/L26_s13/steps.jsonl new file mode 100644 index 0000000000000000000000000000000000000000..dcd0073ac457cc77643aca563283224fa091f7f6 --- /dev/null +++ b/runs/single_llava/removal/L26_s13/steps.jsonl @@ -0,0 +1,500 @@ +{"step": 1, "steps": 500, "lr": 2e-05, "grad_norm": 0.10756666958332062, "loss_scale": 4.0, "retries": 0, "seconds": 2.5, "max_memory_gb": 14.84, "ce": 1.6287612388841808, "rank": 0.13243506848812103} +{"step": 2, "steps": 500, "lr": 2e-05, "grad_norm": 0.215862438082695, "loss_scale": 4.0, "retries": 0, "seconds": 4.0, "max_memory_gb": 14.85, "ce": 2.6742991826031357, "rank": 0.033866047859191895} +{"step": 3, "steps": 500, "lr": 2e-05, "grad_norm": 0.41505998373031616, "loss_scale": 4.0, "retries": 0, "seconds": 5.5, "max_memory_gb": 14.85, "ce": 5.2785763293504715, "rank": 0.15198862552642822} +{"step": 4, "steps": 500, "lr": 2e-05, "grad_norm": 0.3628225326538086, "loss_scale": 4.0, "retries": 0, "seconds": 7.0, "max_memory_gb": 14.85, "ce": 3.802344939438626, "rank": 0.12449777126312256} +{"step": 5, "steps": 500, "lr": 2e-05, "grad_norm": 0.02011784352362156, "loss_scale": 4.0, "retries": 0, "seconds": 8.6, "max_memory_gb": 14.85, "ce": 0.13960898481309414, "rank": 0.0} +{"step": 6, "steps": 500, "lr": 2e-05, "grad_norm": 0.1538223922252655, "loss_scale": 4.0, "retries": 0, "seconds": 10.1, "max_memory_gb": 14.85, "ce": 1.5146495876833797, "rank": 0.0} +{"step": 7, "steps": 500, "lr": 2e-05, "grad_norm": 0.3839884102344513, "loss_scale": 4.0, "retries": 0, "seconds": 11.6, "max_memory_gb": 14.85, "ce": 3.462684052065015, "rank": 0.022104740142822266} +{"step": 8, "steps": 500, "lr": 2e-05, "grad_norm": 0.37630435824394226, "loss_scale": 4.0, "retries": 0, "seconds": 13.2, "max_memory_gb": 14.85, "ce": 3.710068824701011, "rank": 0.2114924192428589} +{"step": 9, "steps": 500, "lr": 2e-05, "grad_norm": 0.2995079457759857, "loss_scale": 4.0, "retries": 0, "seconds": 14.7, "max_memory_gb": 14.85, "ce": 2.8379820426926017, "rank": 0.042589783668518066} +{"step": 10, "steps": 500, "lr": 2e-05, "grad_norm": 0.16266080737113953, "loss_scale": 4.0, "retries": 0, "seconds": 16.2, "max_memory_gb": 14.86, "ce": 1.6177564095705748, "rank": 0.0} +{"step": 11, "steps": 500, "lr": 2e-05, "grad_norm": 0.3793671429157257, "loss_scale": 4.0, "retries": 0, "seconds": 17.8, "max_memory_gb": 14.86, "ce": 2.627635861746967, "rank": 0.21805544197559357} +{"step": 12, "steps": 500, "lr": 2e-05, "grad_norm": 0.1886693686246872, "loss_scale": 4.0, "retries": 0, "seconds": 19.3, "max_memory_gb": 14.86, "ce": 1.5113694535102695, "rank": 0.052756547927856445} +{"step": 13, "steps": 500, "lr": 2e-05, "grad_norm": 0.3906996250152588, "loss_scale": 4.0, "retries": 0, "seconds": 20.9, "max_memory_gb": 14.86, "ce": 2.6724055183585733, "rank": 0.0} +{"step": 14, "steps": 500, "lr": 2e-05, "grad_norm": 0.5643200278282166, "loss_scale": 4.0, "retries": 0, "seconds": 22.4, "max_memory_gb": 14.86, "ce": 3.871150257997215, "rank": 0.14944350719451904} +{"step": 15, "steps": 500, "lr": 2e-05, "grad_norm": 0.3802323043346405, "loss_scale": 4.0, "retries": 0, "seconds": 24.0, "max_memory_gb": 14.86, "ce": 2.600851304421667, "rank": 0.17483019828796387} +{"step": 16, "steps": 500, "lr": 2e-05, "grad_norm": 0.24839197099208832, "loss_scale": 4.0, "retries": 0, "seconds": 25.5, "max_memory_gb": 14.86, "ce": 1.4164808336645365, "rank": 0.0} +{"step": 17, "steps": 500, "lr": 2e-05, "grad_norm": 0.22970616817474365, "loss_scale": 4.0, "retries": 0, "seconds": 27.1, "max_memory_gb": 14.86, "ce": 1.5975004351930693, "rank": 0.03858387470245361} +{"step": 18, "steps": 500, "lr": 2e-05, "grad_norm": 0.5905339121818542, "loss_scale": 4.0, "retries": 0, "seconds": 28.6, "max_memory_gb": 14.86, "ce": 3.5814368575811386, "rank": 0.19830739498138428} +{"step": 19, "steps": 500, "lr": 2e-05, "grad_norm": 0.9049792885780334, "loss_scale": 4.0, "retries": 0, "seconds": 30.1, "max_memory_gb": 14.86, "ce": 5.2678466103971004, "rank": 0.3302898406982422} +{"step": 20, "steps": 500, "lr": 2e-05, "grad_norm": 0.01686866022646427, "loss_scale": 4.0, "retries": 0, "seconds": 31.7, "max_memory_gb": 14.86, "ce": 0.06537132436642423, "rank": 0.10915207862854004} +{"step": 21, "steps": 500, "lr": 2e-05, "grad_norm": 0.5080054998397827, "loss_scale": 4.0, "retries": 0, "seconds": 33.2, "max_memory_gb": 14.86, "ce": 2.7352484384318814, "rank": 0.1749821901321411} +{"step": 22, "steps": 500, "lr": 2e-05, "grad_norm": 0.49501699209213257, "loss_scale": 4.0, "retries": 0, "seconds": 34.7, "max_memory_gb": 14.86, "ce": 2.94043624214828, "rank": 0.18942618370056152} +{"step": 23, "steps": 500, "lr": 2e-05, "grad_norm": 0.5590342879295349, "loss_scale": 4.0, "retries": 0, "seconds": 36.3, "max_memory_gb": 14.86, "ce": 2.7379486351273954, "rank": 0.13188529014587402} +{"step": 24, "steps": 500, "lr": 2e-05, "grad_norm": 0.9952799677848816, "loss_scale": 4.0, "retries": 0, "seconds": 37.7, "max_memory_gb": 14.86, "ce": 4.977634923532605, "rank": 0.2563284635543823} +{"step": 25, "steps": 500, "lr": 2e-05, "grad_norm": 0.9566236138343811, "loss_scale": 4.0, "retries": 0, "seconds": 39.2, "max_memory_gb": 14.86, "ce": 4.094995021820068, "rank": 0.06469690799713135} +{"step": 26, "steps": 500, "lr": 2e-05, "grad_norm": 1.0486664772033691, "loss_scale": 4.0, "retries": 0, "seconds": 40.8, "max_memory_gb": 14.86, "ce": 5.143932375125587, "rank": 0.08363008499145508} +{"step": 27, "steps": 500, "lr": 2e-05, "grad_norm": 0.9013077616691589, "loss_scale": 4.0, "retries": 0, "seconds": 42.3, "max_memory_gb": 14.86, "ce": 4.172728830948472, "rank": 0.517443060874939} +{"step": 28, "steps": 500, "lr": 2e-05, "grad_norm": 0.623151957988739, "loss_scale": 4.0, "retries": 0, "seconds": 43.9, "max_memory_gb": 14.86, "ce": 2.746110623702407, "rank": 0.0} +{"step": 29, "steps": 500, "lr": 2e-05, "grad_norm": 0.7492589354515076, "loss_scale": 4.0, "retries": 0, "seconds": 45.4, "max_memory_gb": 14.86, "ce": 2.647678079898469, "rank": 0.0} +{"step": 30, "steps": 500, "lr": 2e-05, "grad_norm": 0.7250482439994812, "loss_scale": 4.0, "retries": 0, "seconds": 46.9, "max_memory_gb": 14.86, "ce": 2.7318247138755396, "rank": 0.0013266801834106445} +{"step": 31, "steps": 500, "lr": 2e-05, "grad_norm": 0.8088438510894775, "loss_scale": 4.0, "retries": 0, "seconds": 48.4, "max_memory_gb": 14.86, "ce": 2.786921321414411, "rank": 0.1719217300415039} +{"step": 32, "steps": 500, "lr": 2e-05, "grad_norm": 1.440927505493164, "loss_scale": 4.0, "retries": 0, "seconds": 50.0, "max_memory_gb": 14.86, "ce": 5.216612748801708, "rank": 0.28054165840148926} +{"step": 33, "steps": 500, "lr": 2e-05, "grad_norm": 0.6720592975616455, "loss_scale": 4.0, "retries": 0, "seconds": 51.5, "max_memory_gb": 14.86, "ce": 2.48959695501253, "rank": 0.010600626468658447} +{"step": 34, "steps": 500, "lr": 2e-05, "grad_norm": 0.7935006022453308, "loss_scale": 4.0, "retries": 0, "seconds": 53.1, "max_memory_gb": 14.86, "ce": 2.6936979691963643, "rank": 0.0} +{"step": 35, "steps": 500, "lr": 2e-05, "grad_norm": 0.8250741958618164, "loss_scale": 4.0, "retries": 0, "seconds": 54.6, "max_memory_gb": 14.86, "ce": 2.6815667846240103, "rank": 0.1587369441986084} +{"step": 36, "steps": 500, "lr": 2e-05, "grad_norm": 0.824670672416687, "loss_scale": 4.0, "retries": 0, "seconds": 56.2, "max_memory_gb": 14.86, "ce": 2.546244430821389, "rank": 0.0} +{"step": 37, "steps": 500, "lr": 2e-05, "grad_norm": 0.8573230504989624, "loss_scale": 4.0, "retries": 0, "seconds": 57.7, "max_memory_gb": 14.86, "ce": 2.584338428452611, "rank": 0.01009511947631836} +{"step": 38, "steps": 500, "lr": 2e-05, "grad_norm": 1.588487982749939, "loss_scale": 4.0, "retries": 0, "seconds": 59.3, "max_memory_gb": 14.86, "ce": 4.742607395164669, "rank": 0.19677674770355225} +{"step": 39, "steps": 500, "lr": 2e-05, "grad_norm": 0.577329158782959, "loss_scale": 4.0, "retries": 0, "seconds": 60.8, "max_memory_gb": 14.86, "ce": 1.4310121937887743, "rank": 0.20343893766403198} +{"step": 40, "steps": 500, "lr": 2e-05, "grad_norm": 0.9996715188026428, "loss_scale": 4.0, "retries": 0, "seconds": 62.3, "max_memory_gb": 14.86, "ce": 2.6262209401465952, "rank": 0.05812668800354004} +{"step": 41, "steps": 500, "lr": 2e-05, "grad_norm": 0.6441724300384521, "loss_scale": 4.0, "retries": 0, "seconds": 63.9, "max_memory_gb": 14.86, "ce": 1.4849119847640395, "rank": 0.0} +{"step": 42, "steps": 500, "lr": 2e-05, "grad_norm": 1.0305942296981812, "loss_scale": 4.0, "retries": 0, "seconds": 65.4, "max_memory_gb": 14.86, "ce": 2.3657796373590827, "rank": 0.0} +{"step": 43, "steps": 500, "lr": 2e-05, "grad_norm": 1.1138360500335693, "loss_scale": 4.0, "retries": 0, "seconds": 66.9, "max_memory_gb": 14.86, "ce": 2.463446588953957, "rank": 0.002553224563598633} +{"step": 44, "steps": 500, "lr": 2e-05, "grad_norm": 0.6099492907524109, "loss_scale": 4.0, "retries": 0, "seconds": 68.5, "max_memory_gb": 14.86, "ce": 1.4001291561871767, "rank": 0.0} +{"step": 45, "steps": 500, "lr": 2e-05, "grad_norm": 0.6007959842681885, "loss_scale": 4.0, "retries": 0, "seconds": 70.0, "max_memory_gb": 14.86, "ce": 1.4084914033301175, "rank": 0.0} +{"step": 46, "steps": 500, "lr": 2e-05, "grad_norm": 1.0814623832702637, "loss_scale": 4.0, "retries": 0, "seconds": 71.5, "max_memory_gb": 14.86, "ce": 2.3271043384447694, "rank": 0.0} +{"step": 47, "steps": 500, "lr": 2e-05, "grad_norm": 1.138757348060608, "loss_scale": 4.0, "retries": 0, "seconds": 73.0, "max_memory_gb": 14.86, "ce": 2.048093256074935, "rank": 0.0} +{"step": 48, "steps": 500, "lr": 2e-05, "grad_norm": 1.8408559560775757, "loss_scale": 4.0, "retries": 0, "seconds": 74.6, "max_memory_gb": 14.86, "ce": 3.7544495034962893, "rank": 0.16629815101623535} +{"step": 49, "steps": 500, "lr": 2e-05, "grad_norm": 1.7083560228347778, "loss_scale": 4.0, "retries": 0, "seconds": 76.1, "max_memory_gb": 14.86, "ce": 3.551056528929621, "rank": 0.08376502990722656} +{"step": 50, "steps": 500, "lr": 2e-05, "grad_norm": 1.793739914894104, "loss_scale": 4.0, "retries": 0, "seconds": 77.7, "max_memory_gb": 14.86, "ce": 3.611616703448817, "rank": 0.08352923393249512} +{"step": 51, "steps": 500, "lr": 2e-05, "grad_norm": 0.6975740790367126, "loss_scale": 4.0, "retries": 0, "seconds": 79.2, "max_memory_gb": 14.86, "ce": 1.3908716177102178, "rank": 0.08999800682067871} +{"step": 52, "steps": 500, "lr": 2e-05, "grad_norm": 1.6252838373184204, "loss_scale": 4.0, "retries": 0, "seconds": 80.8, "max_memory_gb": 14.86, "ce": 2.355765457265079, "rank": 0.0} +{"step": 53, "steps": 500, "lr": 2e-05, "grad_norm": 1.8048720359802246, "loss_scale": 4.0, "retries": 0, "seconds": 82.3, "max_memory_gb": 14.86, "ce": 3.662549588829279, "rank": 0.05488014221191406} +{"step": 54, "steps": 500, "lr": 2e-05, "grad_norm": 1.9736961126327515, "loss_scale": 4.0, "retries": 0, "seconds": 83.8, "max_memory_gb": 14.86, "ce": 3.758350939489901, "rank": 0.0} +{"step": 55, "steps": 500, "lr": 2e-05, "grad_norm": 1.2080490589141846, "loss_scale": 4.0, "retries": 0, "seconds": 85.2, "max_memory_gb": 14.86, "ce": 2.4560192573117092, "rank": 0.0} +{"step": 56, "steps": 500, "lr": 2e-05, "grad_norm": 2.5120294094085693, "loss_scale": 4.0, "retries": 0, "seconds": 86.8, "max_memory_gb": 14.86, "ce": 4.772312194108963, "rank": 0.0423508882522583} +{"step": 57, "steps": 500, "lr": 2e-05, "grad_norm": 0.7240650653839111, "loss_scale": 4.0, "retries": 0, "seconds": 88.3, "max_memory_gb": 14.86, "ce": 1.2778861313126981, "rank": 0.05715687572956085} +{"step": 58, "steps": 500, "lr": 2e-05, "grad_norm": 1.568738579750061, "loss_scale": 4.0, "retries": 0, "seconds": 89.8, "max_memory_gb": 14.86, "ce": 2.365514399949461, "rank": 0.025583744049072266} +{"step": 59, "steps": 500, "lr": 2e-05, "grad_norm": 0.854161262512207, "loss_scale": 4.0, "retries": 0, "seconds": 91.4, "max_memory_gb": 14.86, "ce": 1.4154698771890253, "rank": 0.023983478546142578} +{"step": 60, "steps": 500, "lr": 2e-05, "grad_norm": 1.3060442209243774, "loss_scale": 4.0, "retries": 0, "seconds": 92.9, "max_memory_gb": 14.86, "ce": 2.38791615748778, "rank": 0.0} +{"step": 61, "steps": 500, "lr": 2e-05, "grad_norm": 2.444904088973999, "loss_scale": 4.0, "retries": 0, "seconds": 94.4, "max_memory_gb": 14.86, "ce": 3.475938935764134, "rank": 0.07128655910491943} +{"step": 62, "steps": 500, "lr": 2e-05, "grad_norm": 0.8325263857841492, "loss_scale": 4.0, "retries": 0, "seconds": 96.0, "max_memory_gb": 14.86, "ce": 1.330605679191649, "rank": 0.0} +{"step": 63, "steps": 500, "lr": 2e-05, "grad_norm": 2.280071258544922, "loss_scale": 4.0, "retries": 0, "seconds": 97.5, "max_memory_gb": 14.86, "ce": 3.567372713238001, "rank": 0.2671382427215576} +{"step": 64, "steps": 500, "lr": 2e-05, "grad_norm": 2.280738353729248, "loss_scale": 4.0, "retries": 0, "seconds": 99.1, "max_memory_gb": 14.86, "ce": 3.5507757000159472, "rank": 0.07843494415283203} +{"step": 65, "steps": 500, "lr": 2e-05, "grad_norm": 2.2623212337493896, "loss_scale": 4.0, "retries": 0, "seconds": 100.7, "max_memory_gb": 14.86, "ce": 3.635788256302476, "rank": 0.0} +{"step": 66, "steps": 500, "lr": 2e-05, "grad_norm": 0.7125515341758728, "loss_scale": 4.0, "retries": 0, "seconds": 102.2, "max_memory_gb": 14.86, "ce": 0.9581747823976912, "rank": 0.24443446844816208} +{"step": 67, "steps": 500, "lr": 2e-05, "grad_norm": 2.6543378829956055, "loss_scale": 4.0, "retries": 0, "seconds": 103.7, "max_memory_gb": 14.86, "ce": 3.3344023497775197, "rank": 0.0} +{"step": 68, "steps": 500, "lr": 2e-05, "grad_norm": 0.8927332758903503, "loss_scale": 4.0, "retries": 0, "seconds": 105.3, "max_memory_gb": 14.86, "ce": 1.1730469921603799, "rank": 0.0} +{"step": 69, "steps": 500, "lr": 2e-05, "grad_norm": 1.7031351327896118, "loss_scale": 4.0, "retries": 0, "seconds": 106.8, "max_memory_gb": 14.86, "ce": 2.2174145453609526, "rank": 0.0} +{"step": 70, "steps": 500, "lr": 2e-05, "grad_norm": 1.9917268753051758, "loss_scale": 4.0, "retries": 0, "seconds": 108.3, "max_memory_gb": 14.86, "ce": 2.1838361169211566, "rank": 0.10474932193756104} +{"step": 71, "steps": 500, "lr": 2e-05, "grad_norm": 2.730851888656616, "loss_scale": 4.0, "retries": 0, "seconds": 109.9, "max_memory_gb": 14.86, "ce": 3.2098681307397783, "rank": 0.09730219841003418} +{"step": 72, "steps": 500, "lr": 2e-05, "grad_norm": 3.313765287399292, "loss_scale": 4.0, "retries": 0, "seconds": 111.4, "max_memory_gb": 14.86, "ce": 3.8225396517664194, "rank": 0.10763514041900635} +{"step": 73, "steps": 500, "lr": 2e-05, "grad_norm": 0.05823316425085068, "loss_scale": 4.0, "retries": 0, "seconds": 112.9, "max_memory_gb": 14.86, "ce": 0.2470271933125332, "rank": 0.0} +{"step": 74, "steps": 500, "lr": 2e-05, "grad_norm": 2.0261058807373047, "loss_scale": 4.0, "retries": 0, "seconds": 114.4, "max_memory_gb": 14.86, "ce": 2.1709128071088344, "rank": 0.1636737585067749} +{"step": 75, "steps": 500, "lr": 2e-05, "grad_norm": 1.286962628364563, "loss_scale": 4.0, "retries": 0, "seconds": 116.0, "max_memory_gb": 14.86, "ce": 1.613030454609543, "rank": 0.1426037847995758} +{"step": 76, "steps": 500, "lr": 2e-05, "grad_norm": 2.00020432472229, "loss_scale": 4.0, "retries": 0, "seconds": 117.5, "max_memory_gb": 14.86, "ce": 2.135884389979765, "rank": 0.0} +{"step": 77, "steps": 500, "lr": 2e-05, "grad_norm": 2.9817490577697754, "loss_scale": 4.0, "retries": 0, "seconds": 119.1, "max_memory_gb": 14.86, "ce": 3.3178874626755714, "rank": 0.09743261337280273} +{"step": 78, "steps": 500, "lr": 2e-05, "grad_norm": 2.072152853012085, "loss_scale": 4.0, "retries": 0, "seconds": 120.6, "max_memory_gb": 14.86, "ce": 1.7684484142810106, "rank": 0.0} +{"step": 79, "steps": 500, "lr": 2e-05, "grad_norm": 1.3203305006027222, "loss_scale": 4.0, "retries": 0, "seconds": 122.0, "max_memory_gb": 14.86, "ce": 0.9327568751759827, "rank": 0.023498892784118652} +{"step": 80, "steps": 500, "lr": 2e-05, "grad_norm": 2.3591949939727783, "loss_scale": 4.0, "retries": 0, "seconds": 123.5, "max_memory_gb": 14.86, "ce": 1.9543347174767405, "rank": 0.0} +{"step": 81, "steps": 500, "lr": 2e-05, "grad_norm": 3.2768590450286865, "loss_scale": 4.0, "retries": 0, "seconds": 125.0, "max_memory_gb": 14.86, "ce": 2.8937159897759557, "rank": 0.041007399559020996} +{"step": 82, "steps": 500, "lr": 2e-05, "grad_norm": 3.3583385944366455, "loss_scale": 4.0, "retries": 0, "seconds": 126.5, "max_memory_gb": 14.86, "ce": 2.846518397331238, "rank": 0.0} +{"step": 83, "steps": 500, "lr": 2e-05, "grad_norm": 3.186763048171997, "loss_scale": 4.0, "retries": 0, "seconds": 128.1, "max_memory_gb": 14.86, "ce": 2.49028548959177, "rank": 0.0} +{"step": 84, "steps": 500, "lr": 2e-05, "grad_norm": 1.2541273832321167, "loss_scale": 4.0, "retries": 0, "seconds": 129.6, "max_memory_gb": 14.86, "ce": 1.0531013181316666, "rank": 0.0} +{"step": 85, "steps": 500, "lr": 2e-05, "grad_norm": 0.07403907179832458, "loss_scale": 4.0, "retries": 0, "seconds": 131.2, "max_memory_gb": 14.86, "ce": 0.12539021956035867, "rank": 0.0} +{"step": 86, "steps": 500, "lr": 2e-05, "grad_norm": 3.711878776550293, "loss_scale": 4.0, "retries": 0, "seconds": 132.7, "max_memory_gb": 14.86, "ce": 2.7614706861786544, "rank": 0.0} +{"step": 87, "steps": 500, "lr": 2e-05, "grad_norm": 3.7206757068634033, "loss_scale": 4.0, "retries": 0, "seconds": 134.2, "max_memory_gb": 14.86, "ce": 2.6639912193641067, "rank": 0.0} +{"step": 88, "steps": 500, "lr": 2e-05, "grad_norm": 2.665163040161133, "loss_scale": 4.0, "retries": 0, "seconds": 135.7, "max_memory_gb": 14.86, "ce": 1.8060604070778936, "rank": 0.2320346236228943} +{"step": 89, "steps": 500, "lr": 2e-05, "grad_norm": 0.05015658587217331, "loss_scale": 4.0, "retries": 0, "seconds": 137.2, "max_memory_gb": 14.86, "ce": 0.15367922274162993, "rank": 0.12437077611684799} +{"step": 90, "steps": 500, "lr": 2e-05, "grad_norm": 2.2807888984680176, "loss_scale": 4.0, "retries": 0, "seconds": 138.7, "max_memory_gb": 14.86, "ce": 1.505408946541138, "rank": 0.04854398965835571} +{"step": 91, "steps": 500, "lr": 2e-05, "grad_norm": 0.0852205827832222, "loss_scale": 4.0, "retries": 0, "seconds": 140.2, "max_memory_gb": 14.86, "ce": 0.35916032874956727, "rank": 0.0} +{"step": 92, "steps": 500, "lr": 2e-05, "grad_norm": 1.6496120691299438, "loss_scale": 4.0, "retries": 0, "seconds": 141.7, "max_memory_gb": 14.86, "ce": 0.9120647870004177, "rank": 0.0} +{"step": 93, "steps": 500, "lr": 2e-05, "grad_norm": 3.390451192855835, "loss_scale": 4.0, "retries": 0, "seconds": 143.3, "max_memory_gb": 14.86, "ce": 1.4636895874282345, "rank": 0.0} +{"step": 94, "steps": 500, "lr": 2e-05, "grad_norm": 4.588296413421631, "loss_scale": 4.0, "retries": 0, "seconds": 144.8, "max_memory_gb": 14.86, "ce": 2.4915722273290157, "rank": 0.1239248514175415} +{"step": 95, "steps": 500, "lr": 2e-05, "grad_norm": 3.018112897872925, "loss_scale": 4.0, "retries": 0, "seconds": 146.3, "max_memory_gb": 14.86, "ce": 1.7232624476309866, "rank": 0.015703022480010986} +{"step": 96, "steps": 500, "lr": 2e-05, "grad_norm": 4.785130977630615, "loss_scale": 4.0, "retries": 0, "seconds": 147.8, "max_memory_gb": 14.86, "ce": 2.4264698335900903, "rank": 0.07301834225654602} +{"step": 97, "steps": 500, "lr": 2e-05, "grad_norm": 1.7100117206573486, "loss_scale": 4.0, "retries": 0, "seconds": 149.4, "max_memory_gb": 14.86, "ce": 0.991112154442817, "rank": 0.0} +{"step": 98, "steps": 500, "lr": 2e-05, "grad_norm": 3.3391847610473633, "loss_scale": 4.0, "retries": 0, "seconds": 150.9, "max_memory_gb": 14.86, "ce": 1.3951883169356734, "rank": 0.006580948829650879} +{"step": 99, "steps": 500, "lr": 2e-05, "grad_norm": 0.12644092738628387, "loss_scale": 4.0, "retries": 0, "seconds": 152.4, "max_memory_gb": 14.86, "ce": 0.23533291433705017, "rank": 0.02517913281917572} +{"step": 100, "steps": 500, "lr": 2e-05, "grad_norm": 4.585390090942383, "loss_scale": 4.0, "retries": 0, "seconds": 154.0, "max_memory_gb": 14.86, "ce": 2.1175307435914874, "rank": 0.0} +{"step": 101, "steps": 500, "lr": 2e-05, "grad_norm": 5.041620254516602, "loss_scale": 4.0, "retries": 0, "seconds": 155.5, "max_memory_gb": 14.86, "ce": 1.9985856590792537, "rank": 0.0} +{"step": 102, "steps": 500, "lr": 2e-05, "grad_norm": 4.860674858093262, "loss_scale": 4.0, "retries": 0, "seconds": 157.0, "max_memory_gb": 14.86, "ce": 2.166177765466273, "rank": 0.0} +{"step": 103, "steps": 500, "lr": 2e-05, "grad_norm": 3.336453676223755, "loss_scale": 4.0, "retries": 0, "seconds": 158.5, "max_memory_gb": 14.86, "ce": 1.3423347302014008, "rank": 0.1840376853942871} +{"step": 104, "steps": 500, "lr": 2e-05, "grad_norm": 2.1631784439086914, "loss_scale": 4.0, "retries": 0, "seconds": 160.0, "max_memory_gb": 14.86, "ce": 0.70670779584907, "rank": 0.18311753869056702} +{"step": 105, "steps": 500, "lr": 2e-05, "grad_norm": 5.227514266967773, "loss_scale": 4.0, "retries": 0, "seconds": 161.6, "max_memory_gb": 14.86, "ce": 1.7482749512419105, "rank": 0.18305540084838867} +{"step": 106, "steps": 500, "lr": 2e-05, "grad_norm": 5.337345123291016, "loss_scale": 4.0, "retries": 0, "seconds": 163.1, "max_memory_gb": 14.86, "ce": 1.5822160369716585, "rank": 0.07710957527160645} +{"step": 107, "steps": 500, "lr": 2e-05, "grad_norm": 2.047161817550659, "loss_scale": 4.0, "retries": 0, "seconds": 164.7, "max_memory_gb": 14.86, "ce": 0.7011906142579392, "rank": 0.0} +{"step": 108, "steps": 500, "lr": 2e-05, "grad_norm": 1.7874109745025635, "loss_scale": 4.0, "retries": 0, "seconds": 166.2, "max_memory_gb": 14.86, "ce": 0.7032987684942782, "rank": 0.0} +{"step": 109, "steps": 500, "lr": 2e-05, "grad_norm": 5.568029880523682, "loss_scale": 4.0, "retries": 0, "seconds": 167.7, "max_memory_gb": 14.86, "ce": 1.7804207182489336, "rank": 0.07240831851959229} +{"step": 110, "steps": 500, "lr": 2e-05, "grad_norm": 4.005682945251465, "loss_scale": 4.0, "retries": 0, "seconds": 169.2, "max_memory_gb": 14.86, "ce": 1.117740364279598, "rank": 0.13620981574058533} +{"step": 111, "steps": 500, "lr": 2e-05, "grad_norm": 3.359123945236206, "loss_scale": 4.0, "retries": 0, "seconds": 170.8, "max_memory_gb": 14.86, "ce": 0.9469668774399906, "rank": 0.0} +{"step": 112, "steps": 500, "lr": 2e-05, "grad_norm": 3.871716260910034, "loss_scale": 4.0, "retries": 0, "seconds": 172.3, "max_memory_gb": 14.86, "ce": 0.849942879518494, "rank": 0.0013292431831359863} +{"step": 113, "steps": 500, "lr": 2e-05, "grad_norm": 3.9305644035339355, "loss_scale": 4.0, "retries": 0, "seconds": 173.8, "max_memory_gb": 14.86, "ce": 1.0112301181070507, "rank": 0.09487974643707275} +{"step": 114, "steps": 500, "lr": 2e-05, "grad_norm": 5.403469562530518, "loss_scale": 4.0, "retries": 0, "seconds": 175.4, "max_memory_gb": 14.86, "ce": 1.361425081267953, "rank": 0.15253865718841553} +{"step": 115, "steps": 500, "lr": 2e-05, "grad_norm": 3.8264999389648438, "loss_scale": 4.0, "retries": 0, "seconds": 176.9, "max_memory_gb": 14.86, "ce": 0.9818185660988092, "rank": 0.11471834778785706} +{"step": 116, "steps": 500, "lr": 2e-05, "grad_norm": 0.07211113721132278, "loss_scale": 4.0, "retries": 0, "seconds": 178.4, "max_memory_gb": 14.86, "ce": 0.14676611474715173, "rank": 0.0} +{"step": 117, "steps": 500, "lr": 2e-05, "grad_norm": 3.364441394805908, "loss_scale": 4.0, "retries": 0, "seconds": 179.9, "max_memory_gb": 14.86, "ce": 0.885951109114103, "rank": 0.0} +{"step": 118, "steps": 500, "lr": 2e-05, "grad_norm": 1.79848313331604, "loss_scale": 4.0, "retries": 0, "seconds": 181.5, "max_memory_gb": 14.86, "ce": 0.48974491306580603, "rank": 0.0} +{"step": 119, "steps": 500, "lr": 2e-05, "grad_norm": 1.6153504848480225, "loss_scale": 4.0, "retries": 0, "seconds": 183.0, "max_memory_gb": 14.86, "ce": 0.39827655255794525, "rank": 0.0} +{"step": 120, "steps": 500, "lr": 2e-05, "grad_norm": 2.874011278152466, "loss_scale": 4.0, "retries": 0, "seconds": 184.5, "max_memory_gb": 14.86, "ce": 0.6990938021335751, "rank": 0.0} +{"step": 121, "steps": 500, "lr": 2e-05, "grad_norm": 1.5468348264694214, "loss_scale": 4.0, "retries": 0, "seconds": 186.0, "max_memory_gb": 14.86, "ce": 0.4859065795317292, "rank": 0.0} +{"step": 122, "steps": 500, "lr": 2e-05, "grad_norm": 2.142238140106201, "loss_scale": 4.0, "retries": 0, "seconds": 187.6, "max_memory_gb": 14.86, "ce": 0.5559001239016652, "rank": 0.0} +{"step": 123, "steps": 500, "lr": 2e-05, "grad_norm": 1.61405611038208, "loss_scale": 4.0, "retries": 0, "seconds": 189.1, "max_memory_gb": 14.86, "ce": 0.6870696332771331, "rank": 0.0065369755029678345} +{"step": 124, "steps": 500, "lr": 2e-05, "grad_norm": 0.7627136707305908, "loss_scale": 4.0, "retries": 0, "seconds": 190.6, "max_memory_gb": 14.86, "ce": 0.3827961072092876, "rank": 0.006150931119918823} +{"step": 125, "steps": 500, "lr": 2e-05, "grad_norm": 1.1334669589996338, "loss_scale": 4.0, "retries": 0, "seconds": 192.1, "max_memory_gb": 14.86, "ce": 0.33981091249734163, "rank": 0.0} +{"step": 126, "steps": 500, "lr": 2e-05, "grad_norm": 2.4697322845458984, "loss_scale": 4.0, "retries": 0, "seconds": 193.6, "max_memory_gb": 14.86, "ce": 0.7032146183773875, "rank": 0.0} +{"step": 127, "steps": 500, "lr": 2e-05, "grad_norm": 1.0025697946548462, "loss_scale": 4.0, "retries": 0, "seconds": 195.1, "max_memory_gb": 14.86, "ce": 0.2084689427865669, "rank": 0.0} +{"step": 128, "steps": 500, "lr": 2e-05, "grad_norm": 1.288934588432312, "loss_scale": 4.0, "retries": 0, "seconds": 196.6, "max_memory_gb": 14.86, "ce": 0.40051144489552826, "rank": 0.0} +{"step": 129, "steps": 500, "lr": 2e-05, "grad_norm": 2.350344657897949, "loss_scale": 4.0, "retries": 0, "seconds": 198.1, "max_memory_gb": 14.86, "ce": 0.5347088547423482, "rank": 0.0} +{"step": 130, "steps": 500, "lr": 2e-05, "grad_norm": 1.2245934009552002, "loss_scale": 4.0, "retries": 0, "seconds": 199.6, "max_memory_gb": 14.86, "ce": 0.5458385306410491, "rank": 0.0} +{"step": 131, "steps": 500, "lr": 2e-05, "grad_norm": 0.7512431740760803, "loss_scale": 4.0, "retries": 0, "seconds": 201.1, "max_memory_gb": 14.86, "ce": 0.33926752861589193, "rank": 0.0} +{"step": 132, "steps": 500, "lr": 2e-05, "grad_norm": 1.8707367181777954, "loss_scale": 4.0, "retries": 0, "seconds": 202.6, "max_memory_gb": 14.86, "ce": 0.6008357591927052, "rank": 0.08792287111282349} +{"step": 133, "steps": 500, "lr": 2e-05, "grad_norm": 1.5877623558044434, "loss_scale": 4.0, "retries": 0, "seconds": 204.2, "max_memory_gb": 14.86, "ce": 0.48673644894734025, "rank": 0.06884750723838806} +{"step": 134, "steps": 500, "lr": 2e-05, "grad_norm": 0.12465206533670425, "loss_scale": 4.0, "retries": 0, "seconds": 205.7, "max_memory_gb": 14.86, "ce": 0.21503898315131664, "rank": 0.0} +{"step": 135, "steps": 500, "lr": 2e-05, "grad_norm": 1.0752133131027222, "loss_scale": 4.0, "retries": 0, "seconds": 207.2, "max_memory_gb": 14.86, "ce": 0.38822350883856416, "rank": 0.0} +{"step": 136, "steps": 500, "lr": 2e-05, "grad_norm": 0.6283225417137146, "loss_scale": 4.0, "retries": 0, "seconds": 208.7, "max_memory_gb": 14.86, "ce": 0.3566740232054144, "rank": 0.0} +{"step": 137, "steps": 500, "lr": 2e-05, "grad_norm": 0.5527922511100769, "loss_scale": 4.0, "retries": 0, "seconds": 210.2, "max_memory_gb": 14.86, "ce": 0.29065101221203804, "rank": 0.0} +{"step": 138, "steps": 500, "lr": 2e-05, "grad_norm": 0.7280495762825012, "loss_scale": 4.0, "retries": 0, "seconds": 211.8, "max_memory_gb": 14.86, "ce": 0.3189696604385972, "rank": 0.0} +{"step": 139, "steps": 500, "lr": 2e-05, "grad_norm": 1.151507019996643, "loss_scale": 4.0, "retries": 0, "seconds": 213.3, "max_memory_gb": 14.86, "ce": 0.45794670563191175, "rank": 0.0} +{"step": 140, "steps": 500, "lr": 2e-05, "grad_norm": 0.367715448141098, "loss_scale": 4.0, "retries": 0, "seconds": 214.9, "max_memory_gb": 14.86, "ce": 0.22349085472524166, "rank": 0.0} +{"step": 141, "steps": 500, "lr": 2e-05, "grad_norm": 0.43031221628189087, "loss_scale": 4.0, "retries": 0, "seconds": 216.3, "max_memory_gb": 14.86, "ce": 0.30026623932644725, "rank": 0.0} +{"step": 142, "steps": 500, "lr": 2e-05, "grad_norm": 0.5392041206359863, "loss_scale": 4.0, "retries": 0, "seconds": 217.8, "max_memory_gb": 14.86, "ce": 0.29070811392739415, "rank": 0.0} +{"step": 143, "steps": 500, "lr": 2e-05, "grad_norm": 0.09519799798727036, "loss_scale": 4.0, "retries": 0, "seconds": 219.4, "max_memory_gb": 14.86, "ce": 0.17972709226887673, "rank": 0.0} +{"step": 144, "steps": 500, "lr": 2e-05, "grad_norm": 0.7151612639427185, "loss_scale": 4.0, "retries": 0, "seconds": 220.9, "max_memory_gb": 14.86, "ce": 0.3518109943252057, "rank": 0.0} +{"step": 145, "steps": 500, "lr": 2e-05, "grad_norm": 0.36479273438453674, "loss_scale": 4.0, "retries": 0, "seconds": 222.5, "max_memory_gb": 14.86, "ce": 0.1749854536028579, "rank": 0.0} +{"step": 146, "steps": 500, "lr": 2e-05, "grad_norm": 0.38916829228401184, "loss_scale": 4.0, "retries": 0, "seconds": 224.0, "max_memory_gb": 14.86, "ce": 0.2718500290066004, "rank": 0.0} +{"step": 147, "steps": 500, "lr": 2e-05, "grad_norm": 0.5001197457313538, "loss_scale": 4.0, "retries": 0, "seconds": 225.6, "max_memory_gb": 14.86, "ce": 0.41522664576768875, "rank": 0.0} +{"step": 148, "steps": 500, "lr": 2e-05, "grad_norm": 0.24468941986560822, "loss_scale": 4.0, "retries": 0, "seconds": 227.0, "max_memory_gb": 14.86, "ce": 0.12010553595609963, "rank": 0.0} +{"step": 149, "steps": 500, "lr": 2e-05, "grad_norm": 0.2870957553386688, "loss_scale": 4.0, "retries": 0, "seconds": 228.3, "max_memory_gb": 14.86, "ce": 0.21802033646963537, "rank": 0.13959424197673798} +{"step": 150, "steps": 500, "lr": 2e-05, "grad_norm": 0.5414546132087708, "loss_scale": 4.0, "retries": 0, "seconds": 229.7, "max_memory_gb": 14.86, "ce": 0.38474735571071506, "rank": 0.060103580355644226} +{"step": 151, "steps": 500, "lr": 2e-05, "grad_norm": 0.3214384913444519, "loss_scale": 4.0, "retries": 0, "seconds": 231.2, "max_memory_gb": 14.86, "ce": 0.20749213983071968, "rank": 0.0} +{"step": 152, "steps": 500, "lr": 2e-05, "grad_norm": 0.5695504546165466, "loss_scale": 4.0, "retries": 0, "seconds": 232.8, "max_memory_gb": 14.86, "ce": 0.336033470928669, "rank": 0.0} +{"step": 153, "steps": 500, "lr": 2e-05, "grad_norm": 0.621250569820404, "loss_scale": 4.0, "retries": 0, "seconds": 234.3, "max_memory_gb": 14.86, "ce": 0.23055013245902956, "rank": 0.0} +{"step": 154, "steps": 500, "lr": 2e-05, "grad_norm": 1.2465628385543823, "loss_scale": 4.0, "retries": 0, "seconds": 235.8, "max_memory_gb": 14.86, "ce": 0.533220783341676, "rank": 0.0872555524110794} +{"step": 155, "steps": 500, "lr": 2e-05, "grad_norm": 2.1345181465148926, "loss_scale": 4.0, "retries": 0, "seconds": 237.3, "max_memory_gb": 14.86, "ce": 0.4763558936538175, "rank": 0.0} +{"step": 156, "steps": 500, "lr": 2e-05, "grad_norm": 0.5726127028465271, "loss_scale": 4.0, "retries": 0, "seconds": 238.9, "max_memory_gb": 14.86, "ce": 0.23432652896735817, "rank": 0.0} +{"step": 157, "steps": 500, "lr": 2e-05, "grad_norm": 0.3601604700088501, "loss_scale": 4.0, "retries": 0, "seconds": 240.4, "max_memory_gb": 14.86, "ce": 0.24938505818136036, "rank": 0.0} +{"step": 158, "steps": 500, "lr": 2e-05, "grad_norm": 0.6516664624214172, "loss_scale": 4.0, "retries": 0, "seconds": 241.9, "max_memory_gb": 14.86, "ce": 0.397909295046702, "rank": 0.0} +{"step": 159, "steps": 500, "lr": 2e-05, "grad_norm": 0.31796136498451233, "loss_scale": 4.0, "retries": 0, "seconds": 243.4, "max_memory_gb": 14.86, "ce": 0.2071294616907835, "rank": 0.0} +{"step": 160, "steps": 500, "lr": 2e-05, "grad_norm": 0.4198002815246582, "loss_scale": 4.0, "retries": 0, "seconds": 245.0, "max_memory_gb": 14.86, "ce": 0.2677426435984671, "rank": 0.023279085755348206} +{"step": 161, "steps": 500, "lr": 2e-05, "grad_norm": 0.3304927945137024, "loss_scale": 4.0, "retries": 0, "seconds": 246.5, "max_memory_gb": 14.86, "ce": 0.2869700387818739, "rank": 0.0} +{"step": 162, "steps": 500, "lr": 2e-05, "grad_norm": 0.1862247735261917, "loss_scale": 4.0, "retries": 0, "seconds": 248.0, "max_memory_gb": 14.86, "ce": 0.18185672955587506, "rank": 0.0} +{"step": 163, "steps": 500, "lr": 2e-05, "grad_norm": 0.6185584664344788, "loss_scale": 4.0, "retries": 0, "seconds": 249.5, "max_memory_gb": 14.86, "ce": 0.22674626088701189, "rank": 0.0} +{"step": 164, "steps": 500, "lr": 2e-05, "grad_norm": 0.7237277626991272, "loss_scale": 4.0, "retries": 0, "seconds": 251.0, "max_memory_gb": 14.86, "ce": 0.3405643696896732, "rank": 0.059923671185970306} +{"step": 165, "steps": 500, "lr": 2e-05, "grad_norm": 0.22605052590370178, "loss_scale": 4.0, "retries": 0, "seconds": 252.5, "max_memory_gb": 14.86, "ce": 0.22591970534995198, "rank": 0.0} +{"step": 166, "steps": 500, "lr": 2e-05, "grad_norm": 0.7484566569328308, "loss_scale": 4.0, "retries": 0, "seconds": 254.1, "max_memory_gb": 14.86, "ce": 0.301080709323287, "rank": 0.0} +{"step": 167, "steps": 500, "lr": 2e-05, "grad_norm": 0.4577244520187378, "loss_scale": 4.0, "retries": 0, "seconds": 255.7, "max_memory_gb": 14.86, "ce": 0.32925303489901125, "rank": 0.0} +{"step": 168, "steps": 500, "lr": 2e-05, "grad_norm": 0.3720490336418152, "loss_scale": 4.0, "retries": 0, "seconds": 257.2, "max_memory_gb": 14.86, "ce": 0.19152889749966562, "rank": 0.0} +{"step": 169, "steps": 500, "lr": 2e-05, "grad_norm": 1.2087517976760864, "loss_scale": 4.0, "retries": 0, "seconds": 258.7, "max_memory_gb": 14.86, "ce": 0.5547335790470243, "rank": 0.15142083913087845} +{"step": 170, "steps": 500, "lr": 2e-05, "grad_norm": 0.412857323884964, "loss_scale": 4.0, "retries": 0, "seconds": 260.1, "max_memory_gb": 14.86, "ce": 0.16468492639251053, "rank": 0.0} +{"step": 171, "steps": 500, "lr": 2e-05, "grad_norm": 0.522728681564331, "loss_scale": 4.0, "retries": 0, "seconds": 261.7, "max_memory_gb": 14.86, "ce": 0.21941131952917203, "rank": 0.010212019085884094} +{"step": 172, "steps": 500, "lr": 2e-05, "grad_norm": 0.8058488368988037, "loss_scale": 4.0, "retries": 0, "seconds": 263.2, "max_memory_gb": 14.86, "ce": 0.2556961129885167, "rank": 0.0} +{"step": 173, "steps": 500, "lr": 2e-05, "grad_norm": 0.696417510509491, "loss_scale": 4.0, "retries": 0, "seconds": 264.6, "max_memory_gb": 14.86, "ce": 0.2612144979648292, "rank": 0.06883211433887482} +{"step": 174, "steps": 500, "lr": 2e-05, "grad_norm": 0.9291673302650452, "loss_scale": 4.0, "retries": 0, "seconds": 266.1, "max_memory_gb": 14.86, "ce": 0.2220832679886371, "rank": 0.0} +{"step": 175, "steps": 500, "lr": 2e-05, "grad_norm": 0.35555416345596313, "loss_scale": 4.0, "retries": 0, "seconds": 267.6, "max_memory_gb": 14.86, "ce": 0.3204796062782407, "rank": 0.0} +{"step": 176, "steps": 500, "lr": 2e-05, "grad_norm": 0.5967445373535156, "loss_scale": 4.0, "retries": 0, "seconds": 269.2, "max_memory_gb": 14.86, "ce": 0.1750777547713369, "rank": 0.04563684016466141} +{"step": 177, "steps": 500, "lr": 2e-05, "grad_norm": 1.3675525188446045, "loss_scale": 4.0, "retries": 0, "seconds": 270.7, "max_memory_gb": 14.86, "ce": 0.31877171399537474, "rank": 0.12944702804088593} +{"step": 178, "steps": 500, "lr": 2e-05, "grad_norm": 0.31034523248672485, "loss_scale": 4.0, "retries": 0, "seconds": 272.2, "max_memory_gb": 14.86, "ce": 0.2762361348140985, "rank": 0.0} +{"step": 179, "steps": 500, "lr": 2e-05, "grad_norm": 0.2467346489429474, "loss_scale": 4.0, "retries": 0, "seconds": 273.7, "max_memory_gb": 14.86, "ce": 0.40572815167251974, "rank": 0.0} +{"step": 180, "steps": 500, "lr": 2e-05, "grad_norm": 1.228618860244751, "loss_scale": 4.0, "retries": 0, "seconds": 275.2, "max_memory_gb": 14.86, "ce": 0.3660956774838269, "rank": 0.0} +{"step": 181, "steps": 500, "lr": 2e-05, "grad_norm": 0.7383832335472107, "loss_scale": 4.0, "retries": 0, "seconds": 276.7, "max_memory_gb": 14.86, "ce": 0.20237557822838426, "rank": 0.009456261992454529} +{"step": 182, "steps": 500, "lr": 2e-05, "grad_norm": 0.29707619547843933, "loss_scale": 4.0, "retries": 0, "seconds": 278.2, "max_memory_gb": 14.86, "ce": 0.17564598459284753, "rank": 0.0} +{"step": 183, "steps": 500, "lr": 2e-05, "grad_norm": 0.2872770130634308, "loss_scale": 4.0, "retries": 0, "seconds": 279.7, "max_memory_gb": 14.86, "ce": 0.09912069141864777, "rank": 0.0} +{"step": 184, "steps": 500, "lr": 2e-05, "grad_norm": 0.4557539224624634, "loss_scale": 4.0, "retries": 0, "seconds": 281.3, "max_memory_gb": 14.86, "ce": 0.1343947744462639, "rank": 0.11474727839231491} +{"step": 185, "steps": 500, "lr": 2e-05, "grad_norm": 0.4211091697216034, "loss_scale": 4.0, "retries": 0, "seconds": 282.8, "max_memory_gb": 14.86, "ce": 0.17185741337016225, "rank": 0.0} +{"step": 186, "steps": 500, "lr": 2e-05, "grad_norm": 0.3805650472640991, "loss_scale": 4.0, "retries": 0, "seconds": 284.4, "max_memory_gb": 14.86, "ce": 0.12687086884398013, "rank": 0.08592560887336731} +{"step": 187, "steps": 500, "lr": 2e-05, "grad_norm": 1.0658000707626343, "loss_scale": 4.0, "retries": 0, "seconds": 285.9, "max_memory_gb": 14.86, "ce": 0.27635660278610885, "rank": 0.13437801599502563} +{"step": 188, "steps": 500, "lr": 2e-05, "grad_norm": 0.09612227976322174, "loss_scale": 4.0, "retries": 0, "seconds": 287.5, "max_memory_gb": 14.86, "ce": 0.1675527223269455, "rank": 0.0} +{"step": 189, "steps": 500, "lr": 2e-05, "grad_norm": 0.3840307295322418, "loss_scale": 4.0, "retries": 0, "seconds": 289.0, "max_memory_gb": 14.86, "ce": 0.12532776256557554, "rank": 0.007421717047691345} +{"step": 190, "steps": 500, "lr": 2e-05, "grad_norm": 0.7646045088768005, "loss_scale": 4.0, "retries": 0, "seconds": 290.5, "max_memory_gb": 14.86, "ce": 0.18702163384296, "rank": 0.0} +{"step": 191, "steps": 500, "lr": 2e-05, "grad_norm": 0.15806396305561066, "loss_scale": 4.0, "retries": 0, "seconds": 292.0, "max_memory_gb": 14.86, "ce": 0.2871752579230815, "rank": 0.0} +{"step": 192, "steps": 500, "lr": 2e-05, "grad_norm": 0.35122570395469666, "loss_scale": 4.0, "retries": 0, "seconds": 293.5, "max_memory_gb": 14.86, "ce": 0.23296134709380567, "rank": 0.0} +{"step": 193, "steps": 500, "lr": 2e-05, "grad_norm": 0.5092601776123047, "loss_scale": 4.0, "retries": 0, "seconds": 295.0, "max_memory_gb": 14.86, "ce": 0.13171718758530915, "rank": 0.18492314964532852} +{"step": 194, "steps": 500, "lr": 2e-05, "grad_norm": 0.2513817250728607, "loss_scale": 4.0, "retries": 0, "seconds": 296.6, "max_memory_gb": 14.86, "ce": 0.28054934018291533, "rank": 0.0} +{"step": 195, "steps": 500, "lr": 2e-05, "grad_norm": 0.534120500087738, "loss_scale": 4.0, "retries": 0, "seconds": 298.1, "max_memory_gb": 14.86, "ce": 0.2253839485347271, "rank": 0.0} +{"step": 196, "steps": 500, "lr": 2e-05, "grad_norm": 0.6169171333312988, "loss_scale": 4.0, "retries": 0, "seconds": 299.7, "max_memory_gb": 14.86, "ce": 0.1254890295676887, "rank": 0.05825459212064743} +{"step": 197, "steps": 500, "lr": 2e-05, "grad_norm": 0.35857540369033813, "loss_scale": 4.0, "retries": 0, "seconds": 301.2, "max_memory_gb": 14.86, "ce": 0.18340467079542577, "rank": 0.031764306128025055} +{"step": 198, "steps": 500, "lr": 2e-05, "grad_norm": 0.3187970519065857, "loss_scale": 4.0, "retries": 0, "seconds": 302.7, "max_memory_gb": 14.86, "ce": 0.23442452354356647, "rank": 0.0} +{"step": 199, "steps": 500, "lr": 2e-05, "grad_norm": 1.6007319688796997, "loss_scale": 4.0, "retries": 0, "seconds": 304.2, "max_memory_gb": 14.86, "ce": 0.23005359945818782, "rank": 0.0} +{"step": 200, "steps": 500, "lr": 2e-05, "grad_norm": 0.2069997638463974, "loss_scale": 4.0, "retries": 0, "seconds": 305.7, "max_memory_gb": 14.86, "ce": 0.28587235987652093, "rank": 0.07419119030237198} +{"step": 201, "steps": 500, "lr": 2e-05, "grad_norm": 0.2732866704463959, "loss_scale": 4.0, "retries": 0, "seconds": 307.3, "max_memory_gb": 14.86, "ce": 0.1665520155802369, "rank": 0.0} +{"step": 202, "steps": 500, "lr": 2e-05, "grad_norm": 0.16660498082637787, "loss_scale": 4.0, "retries": 0, "seconds": 308.7, "max_memory_gb": 14.86, "ce": 0.050724319415166974, "rank": 0.0635344609618187} +{"step": 203, "steps": 500, "lr": 2e-05, "grad_norm": 0.5539751052856445, "loss_scale": 4.0, "retries": 0, "seconds": 310.2, "max_memory_gb": 14.86, "ce": 0.19743821420706809, "rank": 0.011861294507980347} +{"step": 204, "steps": 500, "lr": 2e-05, "grad_norm": 1.2586473226547241, "loss_scale": 4.0, "retries": 0, "seconds": 311.7, "max_memory_gb": 14.86, "ce": 0.2114191377768293, "rank": 0.0} +{"step": 205, "steps": 500, "lr": 2e-05, "grad_norm": 0.6756076216697693, "loss_scale": 4.0, "retries": 0, "seconds": 313.3, "max_memory_gb": 14.86, "ce": 0.17533495876705274, "rank": 0.0} +{"step": 206, "steps": 500, "lr": 2e-05, "grad_norm": 0.6313722729682922, "loss_scale": 4.0, "retries": 0, "seconds": 314.8, "max_memory_gb": 14.86, "ce": 0.12212578777689487, "rank": 0.016032420098781586} +{"step": 207, "steps": 500, "lr": 2e-05, "grad_norm": 0.38579970598220825, "loss_scale": 4.0, "retries": 0, "seconds": 316.2, "max_memory_gb": 14.86, "ce": 0.4788633192656562, "rank": 0.23235958814620972} +{"step": 208, "steps": 500, "lr": 2e-05, "grad_norm": 0.3571237325668335, "loss_scale": 4.0, "retries": 0, "seconds": 317.7, "max_memory_gb": 14.86, "ce": 0.07887172454502434, "rank": 0.0} +{"step": 209, "steps": 500, "lr": 2e-05, "grad_norm": 0.39964720606803894, "loss_scale": 4.0, "retries": 0, "seconds": 319.3, "max_memory_gb": 14.86, "ce": 0.2386112060630694, "rank": 0.16700710356235504} +{"step": 210, "steps": 500, "lr": 2e-05, "grad_norm": 0.2519077956676483, "loss_scale": 4.0, "retries": 0, "seconds": 320.8, "max_memory_gb": 14.86, "ce": 0.22476157516939566, "rank": 0.06643104553222656} +{"step": 211, "steps": 500, "lr": 2e-05, "grad_norm": 0.6459388136863708, "loss_scale": 4.0, "retries": 0, "seconds": 322.3, "max_memory_gb": 14.86, "ce": 0.2433061702759005, "rank": 0.0} +{"step": 212, "steps": 500, "lr": 2e-05, "grad_norm": 0.9004353880882263, "loss_scale": 4.0, "retries": 0, "seconds": 323.8, "max_memory_gb": 14.86, "ce": 0.34525655227480456, "rank": 0.0} +{"step": 213, "steps": 500, "lr": 2e-05, "grad_norm": 0.286531001329422, "loss_scale": 4.0, "retries": 0, "seconds": 325.4, "max_memory_gb": 14.86, "ce": 0.27669477910967544, "rank": 0.0} +{"step": 214, "steps": 500, "lr": 2e-05, "grad_norm": 1.2654136419296265, "loss_scale": 4.0, "retries": 0, "seconds": 326.9, "max_memory_gb": 14.86, "ce": 0.3544409281166736, "rank": 0.06777454912662506} +{"step": 215, "steps": 500, "lr": 2e-05, "grad_norm": 0.29111412167549133, "loss_scale": 4.0, "retries": 0, "seconds": 328.5, "max_memory_gb": 14.86, "ce": 0.3655162525828928, "rank": 0.030658334493637085} +{"step": 216, "steps": 500, "lr": 2e-05, "grad_norm": 1.2241084575653076, "loss_scale": 4.0, "retries": 0, "seconds": 330.0, "max_memory_gb": 14.86, "ce": 0.32383173459675163, "rank": 0.0} +{"step": 217, "steps": 500, "lr": 2e-05, "grad_norm": 0.26140543818473816, "loss_scale": 4.0, "retries": 0, "seconds": 331.5, "max_memory_gb": 14.86, "ce": 0.09395118535030633, "rank": 0.0} +{"step": 218, "steps": 500, "lr": 2e-05, "grad_norm": 0.9023885130882263, "loss_scale": 4.0, "retries": 0, "seconds": 333.1, "max_memory_gb": 14.86, "ce": 0.28625839977758005, "rank": 0.0} +{"step": 219, "steps": 500, "lr": 2e-05, "grad_norm": 0.5774943828582764, "loss_scale": 4.0, "retries": 0, "seconds": 334.6, "max_memory_gb": 14.86, "ce": 0.1486441998858936, "rank": 0.0} +{"step": 220, "steps": 500, "lr": 2e-05, "grad_norm": 1.370826005935669, "loss_scale": 4.0, "retries": 0, "seconds": 336.2, "max_memory_gb": 14.86, "ce": 0.15513455733889714, "rank": 0.0} +{"step": 221, "steps": 500, "lr": 2e-05, "grad_norm": 0.8769212365150452, "loss_scale": 4.0, "retries": 0, "seconds": 337.7, "max_memory_gb": 14.86, "ce": 0.15828580397646874, "rank": 0.0} +{"step": 222, "steps": 500, "lr": 2e-05, "grad_norm": 3.4270052909851074, "loss_scale": 4.0, "retries": 0, "seconds": 339.2, "max_memory_gb": 14.86, "ce": 0.4054381090682, "rank": 0.0} +{"step": 223, "steps": 500, "lr": 2e-05, "grad_norm": 0.1675136238336563, "loss_scale": 4.0, "retries": 0, "seconds": 340.7, "max_memory_gb": 14.86, "ce": 0.2152662833686918, "rank": 0.0} +{"step": 224, "steps": 500, "lr": 2e-05, "grad_norm": 0.5500485301017761, "loss_scale": 4.0, "retries": 0, "seconds": 342.3, "max_memory_gb": 14.86, "ce": 0.1528583172475919, "rank": 0.10207395255565643} +{"step": 225, "steps": 500, "lr": 2e-05, "grad_norm": 0.4300213158130646, "loss_scale": 4.0, "retries": 0, "seconds": 343.8, "max_memory_gb": 14.86, "ce": 0.12279122782638296, "rank": 0.03710184991359711} +{"step": 226, "steps": 500, "lr": 2e-05, "grad_norm": 0.274973064661026, "loss_scale": 4.0, "retries": 0, "seconds": 345.3, "max_memory_gb": 14.86, "ce": 0.14965666661737487, "rank": 0.0} +{"step": 227, "steps": 500, "lr": 2e-05, "grad_norm": 0.33506086468696594, "loss_scale": 4.0, "retries": 0, "seconds": 346.8, "max_memory_gb": 14.86, "ce": 0.04525088082300499, "rank": 0.0} +{"step": 228, "steps": 500, "lr": 2e-05, "grad_norm": 0.19541071355342865, "loss_scale": 4.0, "retries": 0, "seconds": 348.3, "max_memory_gb": 14.86, "ce": 0.12673255504341796, "rank": 0.0} +{"step": 229, "steps": 500, "lr": 2e-05, "grad_norm": 0.208066925406456, "loss_scale": 4.0, "retries": 0, "seconds": 349.9, "max_memory_gb": 14.86, "ce": 0.23299154685810208, "rank": 0.0} +{"step": 230, "steps": 500, "lr": 2e-05, "grad_norm": 0.39320462942123413, "loss_scale": 4.0, "retries": 0, "seconds": 351.3, "max_memory_gb": 14.86, "ce": 0.26615455304272473, "rank": 0.0} +{"step": 231, "steps": 500, "lr": 2e-05, "grad_norm": 0.21235114336013794, "loss_scale": 4.0, "retries": 0, "seconds": 352.8, "max_memory_gb": 14.86, "ce": 0.21620049292687327, "rank": 0.0} +{"step": 232, "steps": 500, "lr": 2e-05, "grad_norm": 0.40995433926582336, "loss_scale": 4.0, "retries": 0, "seconds": 354.4, "max_memory_gb": 14.86, "ce": 0.07981183123774827, "rank": 0.0} +{"step": 233, "steps": 500, "lr": 2e-05, "grad_norm": 2.79189133644104, "loss_scale": 4.0, "retries": 0, "seconds": 356.0, "max_memory_gb": 14.86, "ce": 0.3216613298282027, "rank": 0.0} +{"step": 234, "steps": 500, "lr": 2e-05, "grad_norm": 0.22691063582897186, "loss_scale": 4.0, "retries": 0, "seconds": 357.5, "max_memory_gb": 14.86, "ce": 0.23057493340456858, "rank": 0.0} +{"step": 235, "steps": 500, "lr": 2e-05, "grad_norm": 0.20394566655158997, "loss_scale": 4.0, "retries": 0, "seconds": 359.0, "max_memory_gb": 14.86, "ce": 0.2726301712100394, "rank": 0.0} +{"step": 236, "steps": 500, "lr": 2e-05, "grad_norm": 0.30575090646743774, "loss_scale": 4.0, "retries": 0, "seconds": 360.5, "max_memory_gb": 14.86, "ce": 0.11684492306085303, "rank": 0.0} +{"step": 237, "steps": 500, "lr": 2e-05, "grad_norm": 2.434998035430908, "loss_scale": 4.0, "retries": 0, "seconds": 362.0, "max_memory_gb": 14.86, "ce": 0.2790311886637937, "rank": 0.08474910259246826} +{"step": 238, "steps": 500, "lr": 2e-05, "grad_norm": 0.5071592330932617, "loss_scale": 4.0, "retries": 0, "seconds": 363.5, "max_memory_gb": 14.86, "ce": 0.09692040312802419, "rank": 0.0} +{"step": 239, "steps": 500, "lr": 2e-05, "grad_norm": 0.31259456276893616, "loss_scale": 4.0, "retries": 0, "seconds": 365.0, "max_memory_gb": 14.86, "ce": 0.23141891101840883, "rank": 0.0} +{"step": 240, "steps": 500, "lr": 2e-05, "grad_norm": 0.6446253657341003, "loss_scale": 4.0, "retries": 0, "seconds": 366.5, "max_memory_gb": 14.86, "ce": 0.13425250176806003, "rank": 0.0} +{"step": 241, "steps": 500, "lr": 2e-05, "grad_norm": 0.2170916646718979, "loss_scale": 4.0, "retries": 0, "seconds": 368.0, "max_memory_gb": 14.86, "ce": 0.2279456416144967, "rank": 0.0} +{"step": 242, "steps": 500, "lr": 2e-05, "grad_norm": 0.55821293592453, "loss_scale": 4.0, "retries": 0, "seconds": 369.5, "max_memory_gb": 14.86, "ce": 0.16842446906957775, "rank": 0.0} +{"step": 243, "steps": 500, "lr": 2e-05, "grad_norm": 0.7901115417480469, "loss_scale": 4.0, "retries": 0, "seconds": 371.1, "max_memory_gb": 14.86, "ce": 0.09511954680783674, "rank": 0.0} +{"step": 244, "steps": 500, "lr": 2e-05, "grad_norm": 0.3805108070373535, "loss_scale": 4.0, "retries": 0, "seconds": 372.6, "max_memory_gb": 14.86, "ce": 0.2557186950580217, "rank": 0.008347302675247192} +{"step": 245, "steps": 500, "lr": 2e-05, "grad_norm": 0.2498190850019455, "loss_scale": 4.0, "retries": 0, "seconds": 374.2, "max_memory_gb": 14.86, "ce": 0.35855322319548577, "rank": 0.0} +{"step": 246, "steps": 500, "lr": 2e-05, "grad_norm": 0.2997020184993744, "loss_scale": 4.0, "retries": 0, "seconds": 375.7, "max_memory_gb": 14.86, "ce": 0.07132215448655188, "rank": 0.0} +{"step": 247, "steps": 500, "lr": 2e-05, "grad_norm": 0.38902416825294495, "loss_scale": 4.0, "retries": 0, "seconds": 377.2, "max_memory_gb": 14.86, "ce": 0.14876227593049407, "rank": 0.05705329775810242} +{"step": 248, "steps": 500, "lr": 2e-05, "grad_norm": 0.5199967622756958, "loss_scale": 4.0, "retries": 0, "seconds": 378.7, "max_memory_gb": 14.86, "ce": 0.11252708855317906, "rank": 0.0} +{"step": 249, "steps": 500, "lr": 2e-05, "grad_norm": 0.7001105546951294, "loss_scale": 4.0, "retries": 0, "seconds": 380.3, "max_memory_gb": 14.86, "ce": 0.11710400049923919, "rank": 0.0} +{"step": 250, "steps": 500, "lr": 2e-05, "grad_norm": 0.6916632056236267, "loss_scale": 4.0, "retries": 0, "seconds": 381.8, "max_memory_gb": 14.86, "ce": 0.1025118506513536, "rank": 0.0} +{"step": 251, "steps": 500, "lr": 2e-05, "grad_norm": 0.16833946108818054, "loss_scale": 4.0, "retries": 0, "seconds": 383.3, "max_memory_gb": 14.86, "ce": 0.17046270109131, "rank": 0.0} +{"step": 252, "steps": 500, "lr": 2e-05, "grad_norm": 0.5027092099189758, "loss_scale": 4.0, "retries": 0, "seconds": 384.8, "max_memory_gb": 14.86, "ce": 0.08314683177741244, "rank": 0.0} +{"step": 253, "steps": 500, "lr": 2e-05, "grad_norm": 0.19476152956485748, "loss_scale": 4.0, "retries": 0, "seconds": 386.2, "max_memory_gb": 14.86, "ce": 0.11960332226590253, "rank": 0.0} +{"step": 254, "steps": 500, "lr": 2e-05, "grad_norm": 0.8413715958595276, "loss_scale": 4.0, "retries": 0, "seconds": 387.6, "max_memory_gb": 14.86, "ce": 0.23421404545661062, "rank": 0.0} +{"step": 255, "steps": 500, "lr": 2e-05, "grad_norm": 0.26036307215690613, "loss_scale": 4.0, "retries": 0, "seconds": 389.1, "max_memory_gb": 14.86, "ce": 0.2453602731693536, "rank": 0.0} +{"step": 256, "steps": 500, "lr": 2e-05, "grad_norm": 0.28379398584365845, "loss_scale": 4.0, "retries": 0, "seconds": 390.6, "max_memory_gb": 14.86, "ce": 0.11539214663207531, "rank": 0.010636568069458008} +{"step": 257, "steps": 500, "lr": 2e-05, "grad_norm": 0.1408495157957077, "loss_scale": 4.0, "retries": 0, "seconds": 392.2, "max_memory_gb": 14.86, "ce": 0.08584088471252471, "rank": 0.0} +{"step": 258, "steps": 500, "lr": 2e-05, "grad_norm": 2.050231456756592, "loss_scale": 4.0, "retries": 0, "seconds": 393.7, "max_memory_gb": 14.86, "ce": 0.15893975854851305, "rank": 0.0} +{"step": 259, "steps": 500, "lr": 2e-05, "grad_norm": 0.10474443435668945, "loss_scale": 4.0, "retries": 0, "seconds": 395.2, "max_memory_gb": 14.86, "ce": 0.0650869378587231, "rank": 0.0} +{"step": 260, "steps": 500, "lr": 2e-05, "grad_norm": 0.3768208622932434, "loss_scale": 4.0, "retries": 0, "seconds": 396.7, "max_memory_gb": 14.86, "ce": 0.15189262523199432, "rank": 0.0} +{"step": 261, "steps": 500, "lr": 2e-05, "grad_norm": 1.5874340534210205, "loss_scale": 4.0, "retries": 0, "seconds": 398.1, "max_memory_gb": 14.86, "ce": 0.23085480227018707, "rank": 0.0} +{"step": 262, "steps": 500, "lr": 2e-05, "grad_norm": 0.296505331993103, "loss_scale": 4.0, "retries": 0, "seconds": 399.6, "max_memory_gb": 14.86, "ce": 0.22379212372470647, "rank": 0.0} +{"step": 263, "steps": 500, "lr": 2e-05, "grad_norm": 0.7261604070663452, "loss_scale": 4.0, "retries": 0, "seconds": 401.0, "max_memory_gb": 14.86, "ce": 0.2759626661427319, "rank": 0.11789123713970184} +{"step": 264, "steps": 500, "lr": 2e-05, "grad_norm": 0.2667108178138733, "loss_scale": 4.0, "retries": 0, "seconds": 402.5, "max_memory_gb": 14.86, "ce": 0.16754132288042456, "rank": 0.0} +{"step": 265, "steps": 500, "lr": 2e-05, "grad_norm": 0.2653596103191376, "loss_scale": 4.0, "retries": 0, "seconds": 404.0, "max_memory_gb": 14.86, "ce": 0.12691595588694327, "rank": 0.0} +{"step": 266, "steps": 500, "lr": 2e-05, "grad_norm": 2.442730188369751, "loss_scale": 4.0, "retries": 0, "seconds": 405.6, "max_memory_gb": 14.86, "ce": 0.3149790235911496, "rank": 0.0} +{"step": 267, "steps": 500, "lr": 2e-05, "grad_norm": 0.1617497354745865, "loss_scale": 4.0, "retries": 0, "seconds": 407.1, "max_memory_gb": 14.86, "ce": 0.1540773473388981, "rank": 0.0} +{"step": 268, "steps": 500, "lr": 2e-05, "grad_norm": 0.19350378215312958, "loss_scale": 4.0, "retries": 0, "seconds": 408.6, "max_memory_gb": 14.86, "ce": 0.2823014996247366, "rank": 0.0} +{"step": 269, "steps": 500, "lr": 2e-05, "grad_norm": 2.602482795715332, "loss_scale": 4.0, "retries": 0, "seconds": 410.1, "max_memory_gb": 14.86, "ce": 0.4801302098785527, "rank": 0.1165650486946106} +{"step": 270, "steps": 500, "lr": 2e-05, "grad_norm": 0.29218316078186035, "loss_scale": 4.0, "retries": 0, "seconds": 411.7, "max_memory_gb": 14.86, "ce": 0.13729739235714078, "rank": 0.0} +{"step": 271, "steps": 500, "lr": 2e-05, "grad_norm": 0.4634610712528229, "loss_scale": 4.0, "retries": 0, "seconds": 413.2, "max_memory_gb": 14.86, "ce": 0.10917155863717198, "rank": 0.11662637442350388} +{"step": 272, "steps": 500, "lr": 2e-05, "grad_norm": 0.6709378957748413, "loss_scale": 4.0, "retries": 0, "seconds": 414.7, "max_memory_gb": 14.86, "ce": 0.1471360516443383, "rank": 0.0} +{"step": 273, "steps": 500, "lr": 2e-05, "grad_norm": 0.8063362240791321, "loss_scale": 4.0, "retries": 0, "seconds": 416.2, "max_memory_gb": 14.86, "ce": 0.09378058012225665, "rank": 0.12579189240932465} +{"step": 274, "steps": 500, "lr": 2e-05, "grad_norm": 1.6710891723632812, "loss_scale": 4.0, "retries": 0, "seconds": 417.6, "max_memory_gb": 14.86, "ce": 0.21281878632726148, "rank": 0.004787415266036987} +{"step": 275, "steps": 500, "lr": 2e-05, "grad_norm": 0.5120457410812378, "loss_scale": 4.0, "retries": 0, "seconds": 419.3, "max_memory_gb": 14.86, "ce": 0.5264267796301283, "rank": 0.13003093004226685} +{"step": 276, "steps": 500, "lr": 2e-05, "grad_norm": 0.29490041732788086, "loss_scale": 4.0, "retries": 0, "seconds": 420.8, "max_memory_gb": 14.86, "ce": 0.053115470916964114, "rank": 0.0} +{"step": 277, "steps": 500, "lr": 2e-05, "grad_norm": 0.3192905783653259, "loss_scale": 4.0, "retries": 0, "seconds": 422.3, "max_memory_gb": 14.86, "ce": 0.04827511514304206, "rank": 0.0} +{"step": 278, "steps": 500, "lr": 2e-05, "grad_norm": 0.4813607633113861, "loss_scale": 4.0, "retries": 0, "seconds": 423.8, "max_memory_gb": 14.86, "ce": 0.18549709979561158, "rank": 0.06875220686197281} +{"step": 279, "steps": 500, "lr": 2e-05, "grad_norm": 1.2654110193252563, "loss_scale": 4.0, "retries": 0, "seconds": 425.3, "max_memory_gb": 14.86, "ce": 0.11852472924510948, "rank": 0.0} +{"step": 280, "steps": 500, "lr": 2e-05, "grad_norm": 0.39666396379470825, "loss_scale": 4.0, "retries": 0, "seconds": 426.8, "max_memory_gb": 14.86, "ce": 0.058239656063960865, "rank": 0.0} +{"step": 281, "steps": 500, "lr": 2e-05, "grad_norm": 1.8006511926651, "loss_scale": 4.0, "retries": 0, "seconds": 428.4, "max_memory_gb": 14.86, "ce": 0.21690209521329962, "rank": 0.0} +{"step": 282, "steps": 500, "lr": 2e-05, "grad_norm": 0.20235168933868408, "loss_scale": 4.0, "retries": 0, "seconds": 429.9, "max_memory_gb": 14.86, "ce": 0.12360912028816529, "rank": 0.0} +{"step": 283, "steps": 500, "lr": 2e-05, "grad_norm": 1.7305940389633179, "loss_scale": 4.0, "retries": 0, "seconds": 431.4, "max_memory_gb": 14.86, "ce": 0.1576323119807057, "rank": 0.010507404804229736} +{"step": 284, "steps": 500, "lr": 2e-05, "grad_norm": 1.4147614240646362, "loss_scale": 4.0, "retries": 0, "seconds": 432.9, "max_memory_gb": 14.86, "ce": 0.13780386558210012, "rank": 0.0} +{"step": 285, "steps": 500, "lr": 2e-05, "grad_norm": 0.5913348197937012, "loss_scale": 4.0, "retries": 0, "seconds": 434.5, "max_memory_gb": 14.86, "ce": 0.15519122139085084, "rank": 0.0} +{"step": 286, "steps": 500, "lr": 2e-05, "grad_norm": 0.29261791706085205, "loss_scale": 4.0, "retries": 0, "seconds": 436.0, "max_memory_gb": 14.86, "ce": 0.08213454677024856, "rank": 0.0} +{"step": 287, "steps": 500, "lr": 2e-05, "grad_norm": 0.3811994194984436, "loss_scale": 4.0, "retries": 0, "seconds": 437.6, "max_memory_gb": 14.86, "ce": 0.1649413769482635, "rank": 0.0} +{"step": 288, "steps": 500, "lr": 2e-05, "grad_norm": 0.9527761340141296, "loss_scale": 4.0, "retries": 0, "seconds": 439.0, "max_memory_gb": 14.86, "ce": 0.03734167829679791, "rank": 0.03354092687368393} +{"step": 289, "steps": 500, "lr": 2e-05, "grad_norm": 0.3529563546180725, "loss_scale": 4.0, "retries": 0, "seconds": 440.4, "max_memory_gb": 14.86, "ce": 0.13561370843672194, "rank": 0.0} +{"step": 290, "steps": 500, "lr": 2e-05, "grad_norm": 0.3061676621437073, "loss_scale": 4.0, "retries": 0, "seconds": 441.8, "max_memory_gb": 14.86, "ce": 0.18318728829035535, "rank": 0.0} +{"step": 291, "steps": 500, "lr": 2e-05, "grad_norm": 0.32329466938972473, "loss_scale": 4.0, "retries": 0, "seconds": 443.3, "max_memory_gb": 14.86, "ce": 0.04710102439275943, "rank": 0.0} +{"step": 292, "steps": 500, "lr": 2e-05, "grad_norm": 0.835892379283905, "loss_scale": 4.0, "retries": 0, "seconds": 444.9, "max_memory_gb": 14.86, "ce": 0.20399091494618915, "rank": 0.0} +{"step": 293, "steps": 500, "lr": 2e-05, "grad_norm": 0.4227380156517029, "loss_scale": 4.0, "retries": 0, "seconds": 446.4, "max_memory_gb": 14.86, "ce": 0.1190582895942498, "rank": 0.0} +{"step": 294, "steps": 500, "lr": 2e-05, "grad_norm": 0.9915773868560791, "loss_scale": 4.0, "retries": 0, "seconds": 447.7, "max_memory_gb": 14.86, "ce": 0.1201728964515496, "rank": 0.0} +{"step": 295, "steps": 500, "lr": 2e-05, "grad_norm": 0.7723273038864136, "loss_scale": 4.0, "retries": 0, "seconds": 449.0, "max_memory_gb": 14.86, "ce": 0.27056093409191817, "rank": 0.0} +{"step": 296, "steps": 500, "lr": 2e-05, "grad_norm": 0.7578871250152588, "loss_scale": 4.0, "retries": 0, "seconds": 450.3, "max_memory_gb": 14.86, "ce": 0.2741209827072453, "rank": 0.0} +{"step": 297, "steps": 500, "lr": 2e-05, "grad_norm": 0.14384311437606812, "loss_scale": 4.0, "retries": 0, "seconds": 451.5, "max_memory_gb": 14.86, "ce": 0.14070985012222081, "rank": 0.0} +{"step": 298, "steps": 500, "lr": 2e-05, "grad_norm": 2.559705972671509, "loss_scale": 4.0, "retries": 0, "seconds": 453.0, "max_memory_gb": 14.86, "ce": 0.4246192046557553, "rank": 0.0} +{"step": 299, "steps": 500, "lr": 2e-05, "grad_norm": 0.5582448244094849, "loss_scale": 4.0, "retries": 0, "seconds": 454.5, "max_memory_gb": 14.86, "ce": 0.054502224826137535, "rank": 0.0} +{"step": 300, "steps": 500, "lr": 2e-05, "grad_norm": 0.23297977447509766, "loss_scale": 4.0, "retries": 0, "seconds": 456.0, "max_memory_gb": 14.86, "ce": 0.2799136831454234, "rank": 0.0} +{"step": 301, "steps": 500, "lr": 2e-05, "grad_norm": 0.23483197391033173, "loss_scale": 4.0, "retries": 0, "seconds": 457.5, "max_memory_gb": 14.86, "ce": 0.14589896297547966, "rank": 0.0} +{"step": 302, "steps": 500, "lr": 2e-05, "grad_norm": 0.8218779563903809, "loss_scale": 4.0, "retries": 0, "seconds": 459.0, "max_memory_gb": 14.86, "ce": 0.0606834580830764, "rank": 0.0} +{"step": 303, "steps": 500, "lr": 2e-05, "grad_norm": 0.5751744508743286, "loss_scale": 4.0, "retries": 0, "seconds": 460.5, "max_memory_gb": 14.86, "ce": 0.054470144415972754, "rank": 0.029488958418369293} +{"step": 304, "steps": 500, "lr": 2e-05, "grad_norm": 0.32830220460891724, "loss_scale": 4.0, "retries": 0, "seconds": 462.1, "max_memory_gb": 14.86, "ce": 0.07093334700039122, "rank": 0.067705437541008} +{"step": 305, "steps": 500, "lr": 2e-05, "grad_norm": 0.9068417549133301, "loss_scale": 4.0, "retries": 0, "seconds": 463.7, "max_memory_gb": 14.86, "ce": 0.13636516359110828, "rank": 0.0} +{"step": 306, "steps": 500, "lr": 2e-05, "grad_norm": 0.8067301511764526, "loss_scale": 4.0, "retries": 0, "seconds": 465.2, "max_memory_gb": 14.86, "ce": 0.07469671661965549, "rank": 0.0} +{"step": 307, "steps": 500, "lr": 2e-05, "grad_norm": 0.28696003556251526, "loss_scale": 4.0, "retries": 0, "seconds": 466.7, "max_memory_gb": 14.86, "ce": 0.10916393555089599, "rank": 0.0} +{"step": 308, "steps": 500, "lr": 2e-05, "grad_norm": 0.8544347286224365, "loss_scale": 4.0, "retries": 0, "seconds": 468.3, "max_memory_gb": 14.86, "ce": 0.1985508153156843, "rank": 0.0} +{"step": 309, "steps": 500, "lr": 2e-05, "grad_norm": 0.24714066088199615, "loss_scale": 4.0, "retries": 0, "seconds": 469.8, "max_memory_gb": 14.86, "ce": 0.1924361424316885, "rank": 0.0} +{"step": 310, "steps": 500, "lr": 2e-05, "grad_norm": 1.0272401571273804, "loss_scale": 4.0, "retries": 0, "seconds": 471.3, "max_memory_gb": 14.86, "ce": 0.0907892393443035, "rank": 0.0} +{"step": 311, "steps": 500, "lr": 2e-05, "grad_norm": 0.1279400885105133, "loss_scale": 4.0, "retries": 0, "seconds": 472.8, "max_memory_gb": 14.86, "ce": 0.1156197430100292, "rank": 0.0} +{"step": 312, "steps": 500, "lr": 2e-05, "grad_norm": 0.27697235345840454, "loss_scale": 4.0, "retries": 0, "seconds": 474.3, "max_memory_gb": 14.86, "ce": 0.042305720009608194, "rank": 0.0} +{"step": 313, "steps": 500, "lr": 2e-05, "grad_norm": 0.6653028726577759, "loss_scale": 4.0, "retries": 0, "seconds": 475.9, "max_memory_gb": 14.86, "ce": 0.16761469519406091, "rank": 0.0} +{"step": 314, "steps": 500, "lr": 2e-05, "grad_norm": 0.341314435005188, "loss_scale": 4.0, "retries": 0, "seconds": 477.4, "max_memory_gb": 14.86, "ce": 0.1718403179984307, "rank": 0.0} +{"step": 315, "steps": 500, "lr": 2e-05, "grad_norm": 0.1775890439748764, "loss_scale": 4.0, "retries": 0, "seconds": 479.0, "max_memory_gb": 14.86, "ce": 0.09656102252483834, "rank": 0.11907492578029633} +{"step": 316, "steps": 500, "lr": 2e-05, "grad_norm": 1.2194252014160156, "loss_scale": 4.0, "retries": 0, "seconds": 480.5, "max_memory_gb": 14.86, "ce": 0.32375386671628803, "rank": 0.021536491811275482} +{"step": 317, "steps": 500, "lr": 2e-05, "grad_norm": 0.6850063800811768, "loss_scale": 4.0, "retries": 0, "seconds": 482.0, "max_memory_gb": 14.86, "ce": 0.180409059219528, "rank": 0.0} +{"step": 318, "steps": 500, "lr": 2e-05, "grad_norm": 1.8226613998413086, "loss_scale": 4.0, "retries": 0, "seconds": 483.5, "max_memory_gb": 14.86, "ce": 0.22853344900067896, "rank": 0.0} +{"step": 319, "steps": 500, "lr": 2e-05, "grad_norm": 0.7108386754989624, "loss_scale": 4.0, "retries": 0, "seconds": 485.0, "max_memory_gb": 14.86, "ce": 0.11526127302204259, "rank": 0.0} +{"step": 320, "steps": 500, "lr": 2e-05, "grad_norm": 0.8910152912139893, "loss_scale": 4.0, "retries": 0, "seconds": 486.5, "max_memory_gb": 14.86, "ce": 0.12236555256822612, "rank": 0.0} +{"step": 321, "steps": 500, "lr": 2e-05, "grad_norm": 0.25971826910972595, "loss_scale": 4.0, "retries": 0, "seconds": 488.0, "max_memory_gb": 14.86, "ce": 0.15160260509583168, "rank": 0.1284504234790802} +{"step": 322, "steps": 500, "lr": 2e-05, "grad_norm": 0.6010667681694031, "loss_scale": 4.0, "retries": 0, "seconds": 489.5, "max_memory_gb": 14.86, "ce": 0.04705124786414672, "rank": 0.0} +{"step": 323, "steps": 500, "lr": 2e-05, "grad_norm": 0.30763527750968933, "loss_scale": 4.0, "retries": 0, "seconds": 491.0, "max_memory_gb": 14.86, "ce": 0.06513376839575358, "rank": 0.0} +{"step": 324, "steps": 500, "lr": 2e-05, "grad_norm": 0.42668741941452026, "loss_scale": 4.0, "retries": 0, "seconds": 492.5, "max_memory_gb": 14.86, "ce": 0.05691839936480392, "rank": 0.0} +{"step": 325, "steps": 500, "lr": 2e-05, "grad_norm": 0.23157426714897156, "loss_scale": 4.0, "retries": 0, "seconds": 494.1, "max_memory_gb": 14.86, "ce": 0.22605076839681715, "rank": 0.0} +{"step": 326, "steps": 500, "lr": 2e-05, "grad_norm": 0.2183784693479538, "loss_scale": 4.0, "retries": 0, "seconds": 495.6, "max_memory_gb": 14.86, "ce": 0.06660314643522725, "rank": 0.08467523008584976} +{"step": 327, "steps": 500, "lr": 2e-05, "grad_norm": 2.3345742225646973, "loss_scale": 4.0, "retries": 0, "seconds": 497.1, "max_memory_gb": 14.86, "ce": 0.19522735601640306, "rank": 0.0} +{"step": 328, "steps": 500, "lr": 2e-05, "grad_norm": 0.221862331032753, "loss_scale": 4.0, "retries": 0, "seconds": 498.6, "max_memory_gb": 14.86, "ce": 0.15326350112445652, "rank": 0.0} +{"step": 329, "steps": 500, "lr": 2e-05, "grad_norm": 0.295354962348938, "loss_scale": 4.0, "retries": 0, "seconds": 500.2, "max_memory_gb": 14.86, "ce": 0.15221461217151955, "rank": 0.0} +{"step": 330, "steps": 500, "lr": 2e-05, "grad_norm": 0.3537248969078064, "loss_scale": 4.0, "retries": 0, "seconds": 501.7, "max_memory_gb": 14.86, "ce": 0.15218502335483208, "rank": 0.0} +{"step": 331, "steps": 500, "lr": 2e-05, "grad_norm": 0.6370604634284973, "loss_scale": 4.0, "retries": 0, "seconds": 503.1, "max_memory_gb": 14.86, "ce": 0.2741167164640501, "rank": 0.0} +{"step": 332, "steps": 500, "lr": 2e-05, "grad_norm": 1.5161330699920654, "loss_scale": 4.0, "retries": 0, "seconds": 504.7, "max_memory_gb": 14.86, "ce": 0.17284905354608782, "rank": 0.0} +{"step": 333, "steps": 500, "lr": 2e-05, "grad_norm": 0.25094765424728394, "loss_scale": 4.0, "retries": 0, "seconds": 506.2, "max_memory_gb": 14.86, "ce": 0.269997598050395, "rank": 0.0} +{"step": 334, "steps": 500, "lr": 2e-05, "grad_norm": 0.31917864084243774, "loss_scale": 4.0, "retries": 0, "seconds": 507.7, "max_memory_gb": 14.86, "ce": 0.14683441049419343, "rank": 0.0} +{"step": 335, "steps": 500, "lr": 2e-05, "grad_norm": 2.2425620555877686, "loss_scale": 4.0, "retries": 0, "seconds": 509.3, "max_memory_gb": 14.86, "ce": 0.2754543763148831, "rank": 0.0} +{"step": 336, "steps": 500, "lr": 2e-05, "grad_norm": 1.7277135848999023, "loss_scale": 4.0, "retries": 0, "seconds": 510.8, "max_memory_gb": 14.86, "ce": 0.09511414379812777, "rank": 0.0} +{"step": 337, "steps": 500, "lr": 2e-05, "grad_norm": 2.5036909580230713, "loss_scale": 4.0, "retries": 0, "seconds": 512.4, "max_memory_gb": 14.86, "ce": 0.2595329789619427, "rank": 0.0} +{"step": 338, "steps": 500, "lr": 2e-05, "grad_norm": 0.709621250629425, "loss_scale": 4.0, "retries": 0, "seconds": 513.8, "max_memory_gb": 14.86, "ce": 0.20596463940455578, "rank": 0.0} +{"step": 339, "steps": 500, "lr": 2e-05, "grad_norm": 0.2119237780570984, "loss_scale": 4.0, "retries": 0, "seconds": 515.2, "max_memory_gb": 14.86, "ce": 0.1359192597301444, "rank": 0.0} +{"step": 340, "steps": 500, "lr": 2e-05, "grad_norm": 0.4334361255168915, "loss_scale": 4.0, "retries": 0, "seconds": 516.7, "max_memory_gb": 14.86, "ce": 0.20323643516167067, "rank": 0.0} +{"step": 341, "steps": 500, "lr": 2e-05, "grad_norm": 1.7211623191833496, "loss_scale": 4.0, "retries": 0, "seconds": 518.3, "max_memory_gb": 14.86, "ce": 0.22671525547048077, "rank": 0.0} +{"step": 342, "steps": 500, "lr": 2e-05, "grad_norm": 0.9472600221633911, "loss_scale": 4.0, "retries": 0, "seconds": 519.8, "max_memory_gb": 14.86, "ce": 0.43439558589307126, "rank": 0.07074082642793655} +{"step": 343, "steps": 500, "lr": 2e-05, "grad_norm": 0.5126144289970398, "loss_scale": 4.0, "retries": 0, "seconds": 521.3, "max_memory_gb": 14.86, "ce": 0.08569456997793168, "rank": 0.0} +{"step": 344, "steps": 500, "lr": 2e-05, "grad_norm": 0.2966688275337219, "loss_scale": 4.0, "retries": 0, "seconds": 522.8, "max_memory_gb": 14.86, "ce": 0.04514035297324881, "rank": 0.0} +{"step": 345, "steps": 500, "lr": 2e-05, "grad_norm": 0.140877828001976, "loss_scale": 4.0, "retries": 0, "seconds": 524.4, "max_memory_gb": 14.86, "ce": 0.19996521502616815, "rank": 0.0} +{"step": 346, "steps": 500, "lr": 2e-05, "grad_norm": 0.389212965965271, "loss_scale": 4.0, "retries": 0, "seconds": 525.9, "max_memory_gb": 14.86, "ce": 0.25001683985465206, "rank": 0.0} +{"step": 347, "steps": 500, "lr": 2e-05, "grad_norm": 0.9894531965255737, "loss_scale": 4.0, "retries": 0, "seconds": 527.4, "max_memory_gb": 14.86, "ce": 0.1962942982208915, "rank": 0.0} +{"step": 348, "steps": 500, "lr": 2e-05, "grad_norm": 0.30749866366386414, "loss_scale": 4.0, "retries": 0, "seconds": 528.9, "max_memory_gb": 14.86, "ce": 0.03170159391447669, "rank": 0.0} +{"step": 349, "steps": 500, "lr": 2e-05, "grad_norm": 0.23754218220710754, "loss_scale": 4.0, "retries": 0, "seconds": 530.4, "max_memory_gb": 14.86, "ce": 0.1262346436851658, "rank": 0.0} +{"step": 350, "steps": 500, "lr": 2e-05, "grad_norm": 0.22773407399654388, "loss_scale": 4.0, "retries": 0, "seconds": 531.9, "max_memory_gb": 14.86, "ce": 0.15743443928658962, "rank": 0.0} +{"step": 351, "steps": 500, "lr": 2e-05, "grad_norm": 0.6481227278709412, "loss_scale": 4.0, "retries": 0, "seconds": 533.5, "max_memory_gb": 14.86, "ce": 0.09396021444990765, "rank": 0.0} +{"step": 352, "steps": 500, "lr": 2e-05, "grad_norm": 0.9003402590751648, "loss_scale": 4.0, "retries": 0, "seconds": 535.0, "max_memory_gb": 14.86, "ce": 0.34186692400544416, "rank": 0.0} +{"step": 353, "steps": 500, "lr": 2e-05, "grad_norm": 0.2953346073627472, "loss_scale": 4.0, "retries": 0, "seconds": 536.5, "max_memory_gb": 14.86, "ce": 0.12151285224535968, "rank": 0.0} +{"step": 354, "steps": 500, "lr": 2e-05, "grad_norm": 1.2142163515090942, "loss_scale": 4.0, "retries": 0, "seconds": 538.0, "max_memory_gb": 14.86, "ce": 0.21795769277377985, "rank": 0.0} +{"step": 355, "steps": 500, "lr": 2e-05, "grad_norm": 1.358716607093811, "loss_scale": 4.0, "retries": 0, "seconds": 539.5, "max_memory_gb": 14.86, "ce": 0.16310766417882405, "rank": 0.0} +{"step": 356, "steps": 500, "lr": 2e-05, "grad_norm": 0.23669107258319855, "loss_scale": 4.0, "retries": 0, "seconds": 541.0, "max_memory_gb": 14.86, "ce": 0.06508365609624889, "rank": 0.0} +{"step": 357, "steps": 500, "lr": 2e-05, "grad_norm": 0.2669762074947357, "loss_scale": 4.0, "retries": 0, "seconds": 542.6, "max_memory_gb": 14.86, "ce": 0.07081041336641647, "rank": 0.0} +{"step": 358, "steps": 500, "lr": 2e-05, "grad_norm": 0.36328011751174927, "loss_scale": 4.0, "retries": 0, "seconds": 544.1, "max_memory_gb": 14.86, "ce": 0.2333712682375335, "rank": 0.0} +{"step": 359, "steps": 500, "lr": 2e-05, "grad_norm": 0.41577526926994324, "loss_scale": 4.0, "retries": 0, "seconds": 545.6, "max_memory_gb": 14.86, "ce": 0.09568429596401984, "rank": 0.0} +{"step": 360, "steps": 500, "lr": 2e-05, "grad_norm": 0.2633199989795685, "loss_scale": 4.0, "retries": 0, "seconds": 547.2, "max_memory_gb": 14.86, "ce": 0.08327112432743888, "rank": 0.14060625433921814} +{"step": 361, "steps": 500, "lr": 2e-05, "grad_norm": 0.7596959471702576, "loss_scale": 4.0, "retries": 0, "seconds": 548.7, "max_memory_gb": 14.86, "ce": 0.16702758784231264, "rank": 0.0} +{"step": 362, "steps": 500, "lr": 2e-05, "grad_norm": 2.257070541381836, "loss_scale": 4.0, "retries": 0, "seconds": 550.2, "max_memory_gb": 14.86, "ce": 0.3019120954195387, "rank": 0.0} +{"step": 363, "steps": 500, "lr": 2e-05, "grad_norm": 0.2106960415840149, "loss_scale": 4.0, "retries": 0, "seconds": 551.7, "max_memory_gb": 14.86, "ce": 0.09517624048748985, "rank": 0.0} +{"step": 364, "steps": 500, "lr": 2e-05, "grad_norm": 0.7023381590843201, "loss_scale": 4.0, "retries": 0, "seconds": 553.2, "max_memory_gb": 14.86, "ce": 0.08994124584569363, "rank": 0.1191125363111496} +{"step": 365, "steps": 500, "lr": 2e-05, "grad_norm": 0.30240121483802795, "loss_scale": 4.0, "retries": 0, "seconds": 554.8, "max_memory_gb": 14.86, "ce": 0.15958786642659106, "rank": 0.0} +{"step": 366, "steps": 500, "lr": 2e-05, "grad_norm": 0.9917829632759094, "loss_scale": 4.0, "retries": 0, "seconds": 556.3, "max_memory_gb": 14.86, "ce": 0.1323838818425429, "rank": 0.0} +{"step": 367, "steps": 500, "lr": 2e-05, "grad_norm": 0.2389511615037918, "loss_scale": 4.0, "retries": 0, "seconds": 557.8, "max_memory_gb": 14.86, "ce": 0.12320379503944423, "rank": 0.0} +{"step": 368, "steps": 500, "lr": 2e-05, "grad_norm": 0.5901649594306946, "loss_scale": 4.0, "retries": 0, "seconds": 559.3, "max_memory_gb": 14.86, "ce": 0.0406287937075831, "rank": 0.0} +{"step": 369, "steps": 500, "lr": 2e-05, "grad_norm": 0.2682086229324341, "loss_scale": 4.0, "retries": 0, "seconds": 560.8, "max_memory_gb": 14.86, "ce": 0.07141554667032324, "rank": 0.11239451169967651} +{"step": 370, "steps": 500, "lr": 2e-05, "grad_norm": 0.5488507151603699, "loss_scale": 4.0, "retries": 0, "seconds": 562.4, "max_memory_gb": 14.86, "ce": 0.3529568804151495, "rank": 0.0} +{"step": 371, "steps": 500, "lr": 2e-05, "grad_norm": 1.7133833169937134, "loss_scale": 4.0, "retries": 0, "seconds": 563.9, "max_memory_gb": 14.86, "ce": 0.10348464977141703, "rank": 0.0} +{"step": 372, "steps": 500, "lr": 2e-05, "grad_norm": 0.19523534178733826, "loss_scale": 4.0, "retries": 0, "seconds": 565.4, "max_memory_gb": 14.86, "ce": 0.0502407206222415, "rank": 0.0} +{"step": 373, "steps": 500, "lr": 2e-05, "grad_norm": 0.2559072971343994, "loss_scale": 4.0, "retries": 0, "seconds": 566.9, "max_memory_gb": 14.86, "ce": 0.12441838358063251, "rank": 0.0} +{"step": 374, "steps": 500, "lr": 2e-05, "grad_norm": 0.30603358149528503, "loss_scale": 4.0, "retries": 0, "seconds": 568.5, "max_memory_gb": 14.86, "ce": 0.16100554042350268, "rank": 0.0} +{"step": 375, "steps": 500, "lr": 2e-05, "grad_norm": 0.35894185304641724, "loss_scale": 4.0, "retries": 0, "seconds": 570.0, "max_memory_gb": 14.86, "ce": 0.17916017837706022, "rank": 0.0} +{"step": 376, "steps": 500, "lr": 2e-05, "grad_norm": 0.3565751314163208, "loss_scale": 4.0, "retries": 0, "seconds": 571.5, "max_memory_gb": 14.86, "ce": 0.10605626891265274, "rank": 0.0} +{"step": 377, "steps": 500, "lr": 2e-05, "grad_norm": 0.213127002120018, "loss_scale": 4.0, "retries": 0, "seconds": 573.0, "max_memory_gb": 14.86, "ce": 0.11018262274592416, "rank": 0.0} +{"step": 378, "steps": 500, "lr": 2e-05, "grad_norm": 0.5765653252601624, "loss_scale": 4.0, "retries": 0, "seconds": 574.6, "max_memory_gb": 14.86, "ce": 0.09174699685536325, "rank": 0.0} +{"step": 379, "steps": 500, "lr": 2e-05, "grad_norm": 0.26464375853538513, "loss_scale": 4.0, "retries": 0, "seconds": 576.1, "max_memory_gb": 14.86, "ce": 0.1396160125732422, "rank": 0.0} +{"step": 380, "steps": 500, "lr": 2e-05, "grad_norm": 1.2512613534927368, "loss_scale": 4.0, "retries": 0, "seconds": 577.6, "max_memory_gb": 14.86, "ce": 0.11892836702463683, "rank": 0.0} +{"step": 381, "steps": 500, "lr": 2e-05, "grad_norm": 0.3878404498100281, "loss_scale": 4.0, "retries": 0, "seconds": 579.1, "max_memory_gb": 14.86, "ce": 0.2851192104135407, "rank": 0.0} +{"step": 382, "steps": 500, "lr": 2e-05, "grad_norm": 0.9195445775985718, "loss_scale": 4.0, "retries": 0, "seconds": 580.6, "max_memory_gb": 14.86, "ce": 0.3170773073579767, "rank": 0.0} +{"step": 383, "steps": 500, "lr": 2e-05, "grad_norm": 0.23797349631786346, "loss_scale": 4.0, "retries": 0, "seconds": 582.2, "max_memory_gb": 14.86, "ce": 0.09678275312762707, "rank": 0.0} +{"step": 384, "steps": 500, "lr": 2e-05, "grad_norm": 0.24713045358657837, "loss_scale": 4.0, "retries": 0, "seconds": 583.7, "max_memory_gb": 14.86, "ce": 0.04310632059787167, "rank": 0.0} +{"step": 385, "steps": 500, "lr": 2e-05, "grad_norm": 0.1771048903465271, "loss_scale": 4.0, "retries": 0, "seconds": 585.2, "max_memory_gb": 14.86, "ce": 0.19761428842321038, "rank": 0.0} +{"step": 386, "steps": 500, "lr": 2e-05, "grad_norm": 0.20470696687698364, "loss_scale": 4.0, "retries": 0, "seconds": 586.7, "max_memory_gb": 14.86, "ce": 0.07205868675373495, "rank": 0.0} +{"step": 387, "steps": 500, "lr": 2e-05, "grad_norm": 0.23071254789829254, "loss_scale": 4.0, "retries": 0, "seconds": 588.3, "max_memory_gb": 14.86, "ce": 0.0826626889029285, "rank": 0.0} +{"step": 388, "steps": 500, "lr": 2e-05, "grad_norm": 0.3736353814601898, "loss_scale": 4.0, "retries": 0, "seconds": 589.8, "max_memory_gb": 14.86, "ce": 0.07307245009360486, "rank": 0.0} +{"step": 389, "steps": 500, "lr": 2e-05, "grad_norm": 0.31941837072372437, "loss_scale": 4.0, "retries": 0, "seconds": 591.4, "max_memory_gb": 14.86, "ce": 0.02779477048534318, "rank": 0.10841327905654907} +{"step": 390, "steps": 500, "lr": 2e-05, "grad_norm": 2.006118059158325, "loss_scale": 4.0, "retries": 0, "seconds": 592.9, "max_memory_gb": 14.86, "ce": 0.1184057813952677, "rank": 0.0} +{"step": 391, "steps": 500, "lr": 2e-05, "grad_norm": 1.8074510097503662, "loss_scale": 4.0, "retries": 0, "seconds": 594.5, "max_memory_gb": 14.86, "ce": 0.3205256847286364, "rank": 0.0} +{"step": 392, "steps": 500, "lr": 2e-05, "grad_norm": 0.1555093228816986, "loss_scale": 4.0, "retries": 0, "seconds": 596.0, "max_memory_gb": 14.86, "ce": 0.015122140408493578, "rank": 0.07586521655321121} +{"step": 393, "steps": 500, "lr": 2e-05, "grad_norm": 1.1106414794921875, "loss_scale": 4.0, "retries": 0, "seconds": 597.5, "max_memory_gb": 14.86, "ce": 0.14520230118432664, "rank": 0.0} +{"step": 394, "steps": 500, "lr": 2e-05, "grad_norm": 0.32478490471839905, "loss_scale": 4.0, "retries": 0, "seconds": 599.0, "max_memory_gb": 14.86, "ce": 0.11921219200303312, "rank": 0.0} +{"step": 395, "steps": 500, "lr": 2e-05, "grad_norm": 0.2486598938703537, "loss_scale": 4.0, "retries": 0, "seconds": 600.5, "max_memory_gb": 14.86, "ce": 0.06027097575861262, "rank": 0.0} +{"step": 396, "steps": 500, "lr": 2e-05, "grad_norm": 0.8789635300636292, "loss_scale": 4.0, "retries": 0, "seconds": 602.0, "max_memory_gb": 14.86, "ce": 0.06594508099806262, "rank": 0.0} +{"step": 397, "steps": 500, "lr": 2e-05, "grad_norm": 0.30596646666526794, "loss_scale": 4.0, "retries": 0, "seconds": 603.5, "max_memory_gb": 14.86, "ce": 0.14915136052877642, "rank": 0.0} +{"step": 398, "steps": 500, "lr": 2e-05, "grad_norm": 0.6175671815872192, "loss_scale": 4.0, "retries": 0, "seconds": 605.0, "max_memory_gb": 14.86, "ce": 0.09823051612329436, "rank": 0.0} +{"step": 399, "steps": 500, "lr": 2e-05, "grad_norm": 0.4811587333679199, "loss_scale": 4.0, "retries": 0, "seconds": 606.6, "max_memory_gb": 14.86, "ce": 0.16623248386895284, "rank": 0.0} +{"step": 400, "steps": 500, "lr": 2e-05, "grad_norm": 3.5297303199768066, "loss_scale": 4.0, "retries": 0, "seconds": 608.1, "max_memory_gb": 14.86, "ce": 0.20570554462756263, "rank": 0.0} +{"step": 401, "steps": 500, "lr": 2e-05, "grad_norm": 2.8393373489379883, "loss_scale": 4.0, "retries": 0, "seconds": 609.6, "max_memory_gb": 14.86, "ce": 0.22457744801613444, "rank": 0.0} +{"step": 402, "steps": 500, "lr": 2e-05, "grad_norm": 0.2919120192527771, "loss_scale": 4.0, "retries": 0, "seconds": 611.1, "max_memory_gb": 14.86, "ce": 0.14753677406406496, "rank": 0.10047321021556854} +{"step": 403, "steps": 500, "lr": 2e-05, "grad_norm": 0.20706984400749207, "loss_scale": 4.0, "retries": 0, "seconds": 612.6, "max_memory_gb": 14.86, "ce": 0.023348941875156015, "rank": 0.0} +{"step": 404, "steps": 500, "lr": 2e-05, "grad_norm": 0.30920135974884033, "loss_scale": 4.0, "retries": 0, "seconds": 614.2, "max_memory_gb": 14.86, "ce": 0.15441909755463712, "rank": 0.0} +{"step": 405, "steps": 500, "lr": 2e-05, "grad_norm": 0.5086883306503296, "loss_scale": 4.0, "retries": 0, "seconds": 615.7, "max_memory_gb": 14.86, "ce": 0.06730970926582813, "rank": 0.0} +{"step": 406, "steps": 500, "lr": 2e-05, "grad_norm": 0.3967989683151245, "loss_scale": 4.0, "retries": 0, "seconds": 617.2, "max_memory_gb": 14.86, "ce": 0.16340456093894318, "rank": 0.0} +{"step": 407, "steps": 500, "lr": 2e-05, "grad_norm": 0.973584771156311, "loss_scale": 4.0, "retries": 0, "seconds": 618.7, "max_memory_gb": 14.86, "ce": 0.6401271389913745, "rank": 0.24722003936767578} +{"step": 408, "steps": 500, "lr": 2e-05, "grad_norm": 0.47329583764076233, "loss_scale": 4.0, "retries": 0, "seconds": 620.2, "max_memory_gb": 14.86, "ce": 0.27381254549254663, "rank": 0.10114783048629761} +{"step": 409, "steps": 500, "lr": 2e-05, "grad_norm": 0.34986409544944763, "loss_scale": 4.0, "retries": 0, "seconds": 621.8, "max_memory_gb": 14.86, "ce": 0.1782873708725674, "rank": 0.0} +{"step": 410, "steps": 500, "lr": 2e-05, "grad_norm": 1.0108457803726196, "loss_scale": 4.0, "retries": 0, "seconds": 623.2, "max_memory_gb": 14.86, "ce": 0.07495489603752503, "rank": 0.0} +{"step": 411, "steps": 500, "lr": 2e-05, "grad_norm": 3.3001129627227783, "loss_scale": 4.0, "retries": 0, "seconds": 624.8, "max_memory_gb": 14.86, "ce": 0.22332446956716012, "rank": 0.0} +{"step": 412, "steps": 500, "lr": 2e-05, "grad_norm": 0.22720560431480408, "loss_scale": 4.0, "retries": 0, "seconds": 626.3, "max_memory_gb": 14.86, "ce": 0.1644214120169636, "rank": 0.12467983365058899} +{"step": 413, "steps": 500, "lr": 2e-05, "grad_norm": 0.2889711558818817, "loss_scale": 4.0, "retries": 0, "seconds": 627.8, "max_memory_gb": 14.86, "ce": 0.07617469046090264, "rank": 0.0} +{"step": 414, "steps": 500, "lr": 2e-05, "grad_norm": 0.5434660911560059, "loss_scale": 4.0, "retries": 0, "seconds": 629.3, "max_memory_gb": 14.86, "ce": 0.20347445186052937, "rank": 0.0} +{"step": 415, "steps": 500, "lr": 2e-05, "grad_norm": 1.050667643547058, "loss_scale": 4.0, "retries": 0, "seconds": 630.8, "max_memory_gb": 14.86, "ce": 0.3247346755961189, "rank": 0.0} +{"step": 416, "steps": 500, "lr": 2e-05, "grad_norm": 0.26501163840293884, "loss_scale": 4.0, "retries": 0, "seconds": 632.3, "max_memory_gb": 14.86, "ce": 0.041567617474356666, "rank": 0.0} +{"step": 417, "steps": 500, "lr": 2e-05, "grad_norm": 0.33474817872047424, "loss_scale": 4.0, "retries": 0, "seconds": 633.8, "max_memory_gb": 14.86, "ce": 0.325468851136975, "rank": 0.12823586165905} +{"step": 418, "steps": 500, "lr": 2e-05, "grad_norm": 0.20057569444179535, "loss_scale": 4.0, "retries": 0, "seconds": 635.3, "max_memory_gb": 14.86, "ce": 0.0717551824345719, "rank": 0.0} +{"step": 419, "steps": 500, "lr": 2e-05, "grad_norm": 0.2846822738647461, "loss_scale": 4.0, "retries": 0, "seconds": 636.7, "max_memory_gb": 14.86, "ce": 0.09181714613805525, "rank": 0.039925724267959595} +{"step": 420, "steps": 500, "lr": 2e-05, "grad_norm": 1.0317175388336182, "loss_scale": 4.0, "retries": 0, "seconds": 638.3, "max_memory_gb": 14.86, "ce": 0.2803615152297425, "rank": 0.12357094883918762} +{"step": 421, "steps": 500, "lr": 2e-05, "grad_norm": 0.25258487462997437, "loss_scale": 4.0, "retries": 0, "seconds": 639.8, "max_memory_gb": 14.86, "ce": 0.07372818034491502, "rank": 0.10375085473060608} +{"step": 422, "steps": 500, "lr": 2e-05, "grad_norm": 1.8926714658737183, "loss_scale": 4.0, "retries": 0, "seconds": 641.3, "max_memory_gb": 14.86, "ce": 0.2602699960989412, "rank": 0.0} +{"step": 423, "steps": 500, "lr": 2e-05, "grad_norm": 1.2261290550231934, "loss_scale": 4.0, "retries": 0, "seconds": 642.8, "max_memory_gb": 14.86, "ce": 0.33397702025831677, "rank": 0.0} +{"step": 424, "steps": 500, "lr": 2e-05, "grad_norm": 0.3011046051979065, "loss_scale": 4.0, "retries": 0, "seconds": 644.4, "max_memory_gb": 14.86, "ce": 0.06638687550730538, "rank": 0.0} +{"step": 425, "steps": 500, "lr": 2e-05, "grad_norm": 0.4053942859172821, "loss_scale": 4.0, "retries": 0, "seconds": 645.9, "max_memory_gb": 14.86, "ce": 0.10154504096135497, "rank": 0.0} +{"step": 426, "steps": 500, "lr": 2e-05, "grad_norm": 0.6590350270271301, "loss_scale": 4.0, "retries": 0, "seconds": 647.5, "max_memory_gb": 14.86, "ce": 0.2497845190227963, "rank": 0.0} +{"step": 427, "steps": 500, "lr": 2e-05, "grad_norm": 0.1876148134469986, "loss_scale": 4.0, "retries": 0, "seconds": 649.0, "max_memory_gb": 14.86, "ce": 0.06311412532522809, "rank": 0.0} +{"step": 428, "steps": 500, "lr": 2e-05, "grad_norm": 0.9473937153816223, "loss_scale": 4.0, "retries": 0, "seconds": 650.5, "max_memory_gb": 14.86, "ce": 0.06479583112741238, "rank": 0.0} +{"step": 429, "steps": 500, "lr": 2e-05, "grad_norm": 0.4091663658618927, "loss_scale": 4.0, "retries": 0, "seconds": 652.0, "max_memory_gb": 14.86, "ce": 0.032240699511021376, "rank": 0.0} +{"step": 430, "steps": 500, "lr": 2e-05, "grad_norm": 0.38378506898880005, "loss_scale": 4.0, "retries": 0, "seconds": 653.6, "max_memory_gb": 14.86, "ce": 0.12917500705225393, "rank": 0.0} +{"step": 431, "steps": 500, "lr": 2e-05, "grad_norm": 3.533674716949463, "loss_scale": 4.0, "retries": 0, "seconds": 655.1, "max_memory_gb": 14.86, "ce": 0.20349645932219573, "rank": 0.0} +{"step": 432, "steps": 500, "lr": 2e-05, "grad_norm": 0.40590447187423706, "loss_scale": 4.0, "retries": 0, "seconds": 656.6, "max_memory_gb": 14.86, "ce": 0.08514527640363667, "rank": 0.0} +{"step": 433, "steps": 500, "lr": 2e-05, "grad_norm": 0.5488014221191406, "loss_scale": 4.0, "retries": 0, "seconds": 658.1, "max_memory_gb": 14.86, "ce": 0.10059309501957614, "rank": 0.0} +{"step": 434, "steps": 500, "lr": 2e-05, "grad_norm": 0.1024596244096756, "loss_scale": 4.0, "retries": 0, "seconds": 659.7, "max_memory_gb": 14.86, "ce": 0.01956763633643277, "rank": 0.0} +{"step": 435, "steps": 500, "lr": 2e-05, "grad_norm": 0.5581048727035522, "loss_scale": 4.0, "retries": 0, "seconds": 661.2, "max_memory_gb": 14.86, "ce": 0.1475496480416041, "rank": 0.0} +{"step": 436, "steps": 500, "lr": 2e-05, "grad_norm": 0.44137659668922424, "loss_scale": 4.0, "retries": 0, "seconds": 662.7, "max_memory_gb": 14.86, "ce": 0.12339381924903137, "rank": 0.0} +{"step": 437, "steps": 500, "lr": 2e-05, "grad_norm": 2.1770005226135254, "loss_scale": 4.0, "retries": 0, "seconds": 664.3, "max_memory_gb": 14.86, "ce": 0.16416870919420035, "rank": 0.053614936769008636} +{"step": 438, "steps": 500, "lr": 2e-05, "grad_norm": 0.5007715821266174, "loss_scale": 4.0, "retries": 0, "seconds": 665.8, "max_memory_gb": 14.86, "ce": 0.09066380865988322, "rank": 0.0} +{"step": 439, "steps": 500, "lr": 2e-05, "grad_norm": 1.5225456953048706, "loss_scale": 4.0, "retries": 0, "seconds": 667.2, "max_memory_gb": 14.86, "ce": 0.18081222604450886, "rank": 0.0} +{"step": 440, "steps": 500, "lr": 2e-05, "grad_norm": 0.21873459219932556, "loss_scale": 4.0, "retries": 0, "seconds": 668.8, "max_memory_gb": 14.86, "ce": 0.0739322483423166, "rank": 0.0} +{"step": 441, "steps": 500, "lr": 2e-05, "grad_norm": 0.9387655854225159, "loss_scale": 4.0, "retries": 0, "seconds": 670.3, "max_memory_gb": 14.86, "ce": 0.19672246899426682, "rank": 0.0} +{"step": 442, "steps": 500, "lr": 2e-05, "grad_norm": 0.9927720427513123, "loss_scale": 4.0, "retries": 0, "seconds": 671.8, "max_memory_gb": 14.86, "ce": 0.09479838224069681, "rank": 0.0} +{"step": 443, "steps": 500, "lr": 2e-05, "grad_norm": 0.3433990776538849, "loss_scale": 4.0, "retries": 0, "seconds": 673.3, "max_memory_gb": 14.86, "ce": 0.09208067148574628, "rank": 0.0} +{"step": 444, "steps": 500, "lr": 2e-05, "grad_norm": 0.3495372533798218, "loss_scale": 4.0, "retries": 0, "seconds": 674.9, "max_memory_gb": 14.86, "ce": 0.10747909695055569, "rank": 0.0} +{"step": 445, "steps": 500, "lr": 2e-05, "grad_norm": 0.6147944927215576, "loss_scale": 4.0, "retries": 0, "seconds": 676.4, "max_memory_gb": 14.86, "ce": 0.3627107336287736, "rank": 0.06690370291471481} +{"step": 446, "steps": 500, "lr": 2e-05, "grad_norm": 7.365808010101318, "loss_scale": 4.0, "retries": 0, "seconds": 677.8, "max_memory_gb": 14.86, "ce": 0.2858131465036422, "rank": 0.412751704454422} +{"step": 447, "steps": 500, "lr": 2e-05, "grad_norm": 0.6470518112182617, "loss_scale": 4.0, "retries": 0, "seconds": 679.4, "max_memory_gb": 14.86, "ce": 0.2216004266811069, "rank": 0.0} +{"step": 448, "steps": 500, "lr": 2e-05, "grad_norm": 1.376718282699585, "loss_scale": 4.0, "retries": 0, "seconds": 680.8, "max_memory_gb": 14.86, "ce": 0.2570934948453214, "rank": 0.0} +{"step": 449, "steps": 500, "lr": 2e-05, "grad_norm": 0.7392958998680115, "loss_scale": 4.0, "retries": 0, "seconds": 682.3, "max_memory_gb": 14.86, "ce": 0.044011469872202724, "rank": 0.0} +{"step": 450, "steps": 500, "lr": 2e-05, "grad_norm": 0.7558397054672241, "loss_scale": 4.0, "retries": 0, "seconds": 683.8, "max_memory_gb": 14.86, "ce": 0.046706588720553555, "rank": 0.0} +{"step": 451, "steps": 500, "lr": 2e-05, "grad_norm": 0.6314703226089478, "loss_scale": 4.0, "retries": 0, "seconds": 685.3, "max_memory_gb": 14.86, "ce": 0.051497241234756075, "rank": 0.0} +{"step": 452, "steps": 500, "lr": 2e-05, "grad_norm": 1.1486175060272217, "loss_scale": 4.0, "retries": 0, "seconds": 686.8, "max_memory_gb": 14.86, "ce": 0.1991403439460555, "rank": 0.0} +{"step": 453, "steps": 500, "lr": 2e-05, "grad_norm": 0.30516520142555237, "loss_scale": 4.0, "retries": 0, "seconds": 688.4, "max_memory_gb": 14.86, "ce": 0.050865532393800095, "rank": 0.0} +{"step": 454, "steps": 500, "lr": 2e-05, "grad_norm": 1.0683082342147827, "loss_scale": 4.0, "retries": 0, "seconds": 689.8, "max_memory_gb": 14.86, "ce": 0.1542590136614308, "rank": 0.0} +{"step": 455, "steps": 500, "lr": 2e-05, "grad_norm": 2.129390001296997, "loss_scale": 4.0, "retries": 0, "seconds": 691.3, "max_memory_gb": 14.86, "ce": 0.1616406577959424, "rank": 0.0} +{"step": 456, "steps": 500, "lr": 2e-05, "grad_norm": 0.24530917406082153, "loss_scale": 4.0, "retries": 0, "seconds": 692.8, "max_memory_gb": 14.86, "ce": 0.03360755138237437, "rank": 0.0} +{"step": 457, "steps": 500, "lr": 2e-05, "grad_norm": 0.3451824188232422, "loss_scale": 4.0, "retries": 0, "seconds": 694.3, "max_memory_gb": 14.86, "ce": 0.08196000580210239, "rank": 0.11202680319547653} +{"step": 458, "steps": 500, "lr": 2e-05, "grad_norm": 1.100015640258789, "loss_scale": 4.0, "retries": 0, "seconds": 695.7, "max_memory_gb": 14.86, "ce": 0.17950967748765834, "rank": 0.0} +{"step": 459, "steps": 500, "lr": 2e-05, "grad_norm": 0.7727519869804382, "loss_scale": 4.0, "retries": 0, "seconds": 697.2, "max_memory_gb": 14.86, "ce": 0.15270535497620585, "rank": 0.0} +{"step": 460, "steps": 500, "lr": 2e-05, "grad_norm": 0.282589852809906, "loss_scale": 4.0, "retries": 0, "seconds": 698.8, "max_memory_gb": 14.86, "ce": 0.039076551256584935, "rank": 0.0} +{"step": 461, "steps": 500, "lr": 2e-05, "grad_norm": 0.3610619306564331, "loss_scale": 4.0, "retries": 0, "seconds": 700.3, "max_memory_gb": 14.86, "ce": 0.24127421532102744, "rank": 0.0} +{"step": 462, "steps": 500, "lr": 2e-05, "grad_norm": 0.2230282425880432, "loss_scale": 4.0, "retries": 0, "seconds": 701.9, "max_memory_gb": 14.86, "ce": 0.03426290981587954, "rank": 0.0} +{"step": 463, "steps": 500, "lr": 2e-05, "grad_norm": 0.17006777226924896, "loss_scale": 4.0, "retries": 0, "seconds": 703.4, "max_memory_gb": 14.86, "ce": 0.03093408772838302, "rank": 0.0} +{"step": 464, "steps": 500, "lr": 2e-05, "grad_norm": 0.32932159304618835, "loss_scale": 4.0, "retries": 0, "seconds": 705.0, "max_memory_gb": 14.86, "ce": 0.029613485701702302, "rank": 0.0} +{"step": 465, "steps": 500, "lr": 2e-05, "grad_norm": 1.125122308731079, "loss_scale": 4.0, "retries": 0, "seconds": 706.4, "max_memory_gb": 14.86, "ce": 0.2415767408838292, "rank": 0.0} +{"step": 466, "steps": 500, "lr": 2e-05, "grad_norm": 0.7528068423271179, "loss_scale": 4.0, "retries": 0, "seconds": 707.9, "max_memory_gb": 14.86, "ce": 0.04073346967197722, "rank": 0.0} +{"step": 467, "steps": 500, "lr": 2e-05, "grad_norm": 0.28459057211875916, "loss_scale": 4.0, "retries": 0, "seconds": 709.4, "max_memory_gb": 14.86, "ce": 0.04489497829308675, "rank": 0.0} +{"step": 468, "steps": 500, "lr": 2e-05, "grad_norm": 0.1647816300392151, "loss_scale": 4.0, "retries": 0, "seconds": 710.9, "max_memory_gb": 14.86, "ce": 0.02655730953119928, "rank": 0.0} +{"step": 469, "steps": 500, "lr": 2e-05, "grad_norm": 3.1050846576690674, "loss_scale": 4.0, "retries": 0, "seconds": 712.4, "max_memory_gb": 14.86, "ce": 0.1535624236403237, "rank": 0.0} +{"step": 470, "steps": 500, "lr": 2e-05, "grad_norm": 1.0564345121383667, "loss_scale": 4.0, "retries": 0, "seconds": 713.9, "max_memory_gb": 14.86, "ce": 0.09775144318700768, "rank": 0.0} +{"step": 471, "steps": 500, "lr": 2e-05, "grad_norm": 0.7570781707763672, "loss_scale": 4.0, "retries": 0, "seconds": 715.5, "max_memory_gb": 14.86, "ce": 0.17502098809200106, "rank": 0.0} +{"step": 472, "steps": 500, "lr": 2e-05, "grad_norm": 0.3227836787700653, "loss_scale": 4.0, "retries": 0, "seconds": 717.0, "max_memory_gb": 14.86, "ce": 0.043790492188236385, "rank": 0.0} +{"step": 473, "steps": 500, "lr": 2e-05, "grad_norm": 1.4171332120895386, "loss_scale": 4.0, "retries": 0, "seconds": 718.6, "max_memory_gb": 14.86, "ce": 0.051908970997828874, "rank": 0.1444176584482193} +{"step": 474, "steps": 500, "lr": 2e-05, "grad_norm": 0.40337586402893066, "loss_scale": 4.0, "retries": 0, "seconds": 720.1, "max_memory_gb": 14.86, "ce": 0.11813107934722211, "rank": 0.0} +{"step": 475, "steps": 500, "lr": 2e-05, "grad_norm": 0.536162257194519, "loss_scale": 4.0, "retries": 0, "seconds": 721.6, "max_memory_gb": 14.86, "ce": 0.13543360386393033, "rank": 0.0} +{"step": 476, "steps": 500, "lr": 2e-05, "grad_norm": 1.0088893175125122, "loss_scale": 4.0, "retries": 0, "seconds": 723.2, "max_memory_gb": 14.86, "ce": 0.03002886653302994, "rank": 0.028964556753635406} +{"step": 477, "steps": 500, "lr": 2e-05, "grad_norm": 1.9337702989578247, "loss_scale": 4.0, "retries": 0, "seconds": 724.7, "max_memory_gb": 14.86, "ce": 0.12301350930101762, "rank": 0.0} +{"step": 478, "steps": 500, "lr": 2e-05, "grad_norm": 0.3988134562969208, "loss_scale": 4.0, "retries": 0, "seconds": 726.2, "max_memory_gb": 14.86, "ce": 0.08290737758943578, "rank": 0.0} +{"step": 479, "steps": 500, "lr": 2e-05, "grad_norm": 3.161228656768799, "loss_scale": 4.0, "retries": 0, "seconds": 727.8, "max_memory_gb": 14.86, "ce": 0.34429145928879734, "rank": 0.03305056691169739} +{"step": 480, "steps": 500, "lr": 2e-05, "grad_norm": 0.5683278441429138, "loss_scale": 4.0, "retries": 0, "seconds": 729.3, "max_memory_gb": 14.86, "ce": 0.16774714822895476, "rank": 0.0} +{"step": 481, "steps": 500, "lr": 2e-05, "grad_norm": 0.2118980884552002, "loss_scale": 4.0, "retries": 0, "seconds": 730.9, "max_memory_gb": 14.86, "ce": 0.04044624795278651, "rank": 0.0} +{"step": 482, "steps": 500, "lr": 2e-05, "grad_norm": 0.5162537693977356, "loss_scale": 4.0, "retries": 0, "seconds": 732.4, "max_memory_gb": 14.86, "ce": 0.10879052896052599, "rank": 0.0} +{"step": 483, "steps": 500, "lr": 2e-05, "grad_norm": 1.78123140335083, "loss_scale": 4.0, "retries": 0, "seconds": 733.9, "max_memory_gb": 14.86, "ce": 0.3569594700284142, "rank": 0.0} +{"step": 484, "steps": 500, "lr": 2e-05, "grad_norm": 0.7785648107528687, "loss_scale": 4.0, "retries": 0, "seconds": 735.4, "max_memory_gb": 14.86, "ce": 0.36489504793098604, "rank": 0.0} +{"step": 485, "steps": 500, "lr": 2e-05, "grad_norm": 0.5644932985305786, "loss_scale": 4.0, "retries": 0, "seconds": 736.9, "max_memory_gb": 14.86, "ce": 0.1107434175646631, "rank": 0.0} +{"step": 486, "steps": 500, "lr": 2e-05, "grad_norm": 0.6938545107841492, "loss_scale": 4.0, "retries": 0, "seconds": 738.4, "max_memory_gb": 14.86, "ce": 0.18010784960733872, "rank": 0.0} +{"step": 487, "steps": 500, "lr": 2e-05, "grad_norm": 1.8762190341949463, "loss_scale": 4.0, "retries": 0, "seconds": 739.9, "max_memory_gb": 14.86, "ce": 0.09729146439713077, "rank": 0.0} +{"step": 488, "steps": 500, "lr": 2e-05, "grad_norm": 0.5228891372680664, "loss_scale": 4.0, "retries": 0, "seconds": 741.5, "max_memory_gb": 14.86, "ce": 0.10991237662528874, "rank": 0.0} +{"step": 489, "steps": 500, "lr": 2e-05, "grad_norm": 2.8064374923706055, "loss_scale": 4.0, "retries": 0, "seconds": 743.0, "max_memory_gb": 14.86, "ce": 0.21689657888418878, "rank": 0.0} +{"step": 490, "steps": 500, "lr": 2e-05, "grad_norm": 0.31675300002098083, "loss_scale": 4.0, "retries": 0, "seconds": 744.5, "max_memory_gb": 14.86, "ce": 0.06308054764667759, "rank": 0.0} +{"step": 491, "steps": 500, "lr": 2e-05, "grad_norm": 0.14810094237327576, "loss_scale": 4.0, "retries": 0, "seconds": 746.0, "max_memory_gb": 14.86, "ce": 0.10393750925140921, "rank": 0.0} +{"step": 492, "steps": 500, "lr": 2e-05, "grad_norm": 0.39946427941322327, "loss_scale": 4.0, "retries": 0, "seconds": 747.5, "max_memory_gb": 14.86, "ce": 0.023904081594082527, "rank": 0.0} +{"step": 493, "steps": 500, "lr": 2e-05, "grad_norm": 1.907543659210205, "loss_scale": 4.0, "retries": 0, "seconds": 749.1, "max_memory_gb": 14.86, "ce": 0.1421461514546536, "rank": 0.0} +{"step": 494, "steps": 500, "lr": 2e-05, "grad_norm": 1.1573381423950195, "loss_scale": 4.0, "retries": 0, "seconds": 750.6, "max_memory_gb": 14.86, "ce": 0.10247185584739782, "rank": 0.0} +{"step": 495, "steps": 500, "lr": 2e-05, "grad_norm": 0.4643551707267761, "loss_scale": 4.0, "retries": 0, "seconds": 752.0, "max_memory_gb": 14.86, "ce": 0.05140879246573604, "rank": 0.0} +{"step": 496, "steps": 500, "lr": 2e-05, "grad_norm": 1.7730262279510498, "loss_scale": 4.0, "retries": 0, "seconds": 753.5, "max_memory_gb": 14.86, "ce": 0.06901383613876533, "rank": 0.0} +{"step": 497, "steps": 500, "lr": 2e-05, "grad_norm": 0.6354601383209229, "loss_scale": 4.0, "retries": 0, "seconds": 755.0, "max_memory_gb": 14.86, "ce": 0.1282054765852081, "rank": 0.0} +{"step": 498, "steps": 500, "lr": 2e-05, "grad_norm": 0.5679410696029663, "loss_scale": 4.0, "retries": 0, "seconds": 756.5, "max_memory_gb": 14.86, "ce": 0.15584366345137823, "rank": 0.0} +{"step": 499, "steps": 500, "lr": 2e-05, "grad_norm": 0.8142927885055542, "loss_scale": 4.0, "retries": 0, "seconds": 758.1, "max_memory_gb": 14.86, "ce": 0.1414801555329177, "rank": 0.0} +{"step": 500, "steps": 500, "lr": 2e-05, "grad_norm": 0.822006106376648, "loss_scale": 4.0, "retries": 0, "seconds": 759.6, "max_memory_gb": 14.86, "ce": 0.2217175872028747, "rank": 0.0} diff --git a/runs/single_llava/removal/L30_s13/adapter/README.md b/runs/single_llava/removal/L30_s13/adapter/README.md new file mode 100644 index 0000000000000000000000000000000000000000..ae587920ca55972d94e38f783837f71be14f1dda --- /dev/null +++ b/runs/single_llava/removal/L30_s13/adapter/README.md @@ -0,0 +1,207 @@ +--- +base_model: /data/OPSD/assets/models/llava-1.5-7b-hf +library_name: peft +pipeline_tag: text-generation +tags: +- base_model:adapter:/data/OPSD/assets/models/llava-1.5-7b-hf +- lora +- transformers +--- + +# Model Card for Model ID + + + + + +## Model Details + +### Model Description + + + + + +- **Developed by:** [More Information Needed] +- **Funded by [optional]:** [More Information Needed] +- **Shared by [optional]:** [More Information Needed] +- **Model type:** [More Information Needed] +- **Language(s) (NLP):** [More Information Needed] +- **License:** [More Information Needed] +- **Finetuned from model [optional]:** [More Information Needed] + +### Model Sources [optional] + + + +- **Repository:** [More Information Needed] +- **Paper [optional]:** [More Information Needed] +- **Demo [optional]:** [More Information Needed] + +## Uses + + + +### Direct Use + + + +[More Information Needed] + +### Downstream Use [optional] + + + +[More Information Needed] + +### Out-of-Scope Use + + + +[More Information Needed] + +## Bias, Risks, and Limitations + + + +[More Information Needed] + +### Recommendations + + + +Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. More information needed for further recommendations. + +## How to Get Started with the Model + +Use the code below to get started with the model. + +[More Information Needed] + +## Training Details + +### Training Data + + + +[More Information Needed] + +### Training Procedure + + + +#### Preprocessing [optional] + +[More Information Needed] + + +#### Training Hyperparameters + +- **Training regime:** [More Information Needed] + +#### Speeds, Sizes, Times [optional] + + + +[More Information Needed] + +## Evaluation + + + +### Testing Data, Factors & Metrics + +#### Testing Data + + + +[More Information Needed] + +#### Factors + + + +[More Information Needed] + +#### Metrics + + + +[More Information Needed] + +### Results + +[More Information Needed] + +#### Summary + + + +## Model Examination [optional] + + + +[More Information Needed] + +## Environmental Impact + + + +Carbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700). + +- **Hardware Type:** [More Information Needed] +- **Hours used:** [More Information Needed] +- **Cloud Provider:** [More Information Needed] +- **Compute Region:** [More Information Needed] +- **Carbon Emitted:** [More Information Needed] + +## Technical Specifications [optional] + +### Model Architecture and Objective + +[More Information Needed] + +### Compute Infrastructure + +[More Information Needed] + +#### Hardware + +[More Information Needed] + +#### Software + +[More Information Needed] + +## Citation [optional] + + + +**BibTeX:** + +[More Information Needed] + +**APA:** + +[More Information Needed] + +## Glossary [optional] + + + +[More Information Needed] + +## More Information [optional] + +[More Information Needed] + +## Model Card Authors [optional] + +[More Information Needed] + +## Model Card Contact + +[More Information Needed] +### Framework versions + +- PEFT 0.20.0 \ No newline at end of file diff --git a/runs/single_llava/removal/L30_s13/adapter/adapter_config.json b/runs/single_llava/removal/L30_s13/adapter/adapter_config.json new file mode 100644 index 0000000000000000000000000000000000000000..72e4b3c69dde07812a116f1a02a870015e6a45ea --- /dev/null +++ b/runs/single_llava/removal/L30_s13/adapter/adapter_config.json @@ -0,0 +1,50 @@ +{ + "alora_invocation_tokens": null, + "alpha_pattern": {}, + "arrow_config": null, + "auto_mapping": null, + "base_model_name_or_path": "/data/OPSD/assets/models/llava-1.5-7b-hf", + "bias": "none", + "corda_config": null, + "ensure_weight_tying": false, + "eva_config": null, + "exclude_modules": null, + "fan_in_fan_out": false, + "inference_mode": true, + "init_lora_weights": true, + "layer_replication": null, + "layers_pattern": null, + "layers_to_transform": null, + "loftq_config": {}, + "lora_alpha": 32, + "lora_bias": false, + "lora_dropout": 0.05, + "lora_ga_config": null, + "megatron_config": null, + "megatron_core": "megatron.core", + "modules_to_save": null, + "monteclora_config": null, + "peft_type": "LORA", + "peft_version": "0.20.0", + "qalora_group_size": 16, + "r": 16, + "rank_pattern": {}, + "revision": null, + "target_modules": [ + "model.language_model.layers.30.mlp.up_proj", + "model.language_model.layers.30.self_attn.o_proj", + "model.language_model.layers.30.self_attn.v_proj", + "model.language_model.layers.30.mlp.down_proj", + "model.language_model.layers.30.self_attn.q_proj", + "model.language_model.layers.30.mlp.gate_proj", + "model.language_model.layers.30.self_attn.k_proj" + ], + "target_parameters": null, + "task_type": "CAUSAL_LM", + "trainable_token_indices": null, + "use_bdlora": null, + "use_dora": false, + "use_qalora": false, + "use_rslora": false, + "velora_config": null +} \ No newline at end of file diff --git a/runs/single_llava/removal/L30_s13/config.json b/runs/single_llava/removal/L30_s13/config.json new file mode 100644 index 0000000000000000000000000000000000000000..804a9dd0916bec0d42b8bea37d9183d702b0b9d0 --- /dev/null +++ b/runs/single_llava/removal/L30_s13/config.json @@ -0,0 +1,24 @@ +{ + "behavior": "removal", + "seed": 13, + "model": "/data/OPSD/assets/models/llava-1.5-7b-hf", + "backbone": "llava", + "train": "/data/OPSD/assets/v3/removal.jsonl", + "items": 1000, + "steps": 500, + "epochs": 2, + "steps_per_epoch": 250, + "groups_per_step": 4, + "lr": 2e-05, + "scheduler": "constant", + "warmup_steps": 0, + "final_fraction": 1.0, + "weight_decay": 0.0, + "clip": 1.0, + "initial_loss_scale": 4.0, + "adapter": { + "layer": 30, + "rank": 16, + "trainable_parameters": 1249280 + } +} \ No newline at end of file diff --git a/runs/single_llava/removal/L30_s13/steps.jsonl b/runs/single_llava/removal/L30_s13/steps.jsonl new file mode 100644 index 0000000000000000000000000000000000000000..284de6e032accc048c96c7bfab7e27ac00635ef7 --- /dev/null +++ b/runs/single_llava/removal/L30_s13/steps.jsonl @@ -0,0 +1,500 @@ +{"step": 1, "steps": 500, "lr": 2e-05, "grad_norm": 0.1919374018907547, "loss_scale": 4.0, "retries": 0, "seconds": 2.3, "max_memory_gb": 14.29, "ce": 1.6287612388841808, "rank": 0.13243506848812103} +{"step": 2, "steps": 500, "lr": 2e-05, "grad_norm": 0.35401585698127747, "loss_scale": 4.0, "retries": 0, "seconds": 3.6, "max_memory_gb": 14.31, "ce": 2.675456267199479, "rank": 0.033625006675720215} +{"step": 3, "steps": 500, "lr": 2e-05, "grad_norm": 0.847952663898468, "loss_scale": 4.0, "retries": 0, "seconds": 5.0, "max_memory_gb": 14.31, "ce": 5.278313662856817, "rank": 0.1505674123764038} +{"step": 4, "steps": 500, "lr": 2e-05, "grad_norm": 0.5781099796295166, "loss_scale": 4.0, "retries": 0, "seconds": 6.4, "max_memory_gb": 14.31, "ce": 3.8030319492099807, "rank": 0.12352776527404785} +{"step": 5, "steps": 500, "lr": 2e-05, "grad_norm": 0.031579893082380295, "loss_scale": 4.0, "retries": 0, "seconds": 7.6, "max_memory_gb": 14.31, "ce": 0.13957958458922803, "rank": 0.0} +{"step": 6, "steps": 500, "lr": 2e-05, "grad_norm": 0.2847771942615509, "loss_scale": 4.0, "retries": 0, "seconds": 8.9, "max_memory_gb": 14.31, "ce": 1.5140266995877028, "rank": 0.0} +{"step": 7, "steps": 500, "lr": 2e-05, "grad_norm": 0.6207830309867859, "loss_scale": 4.0, "retries": 0, "seconds": 10.3, "max_memory_gb": 14.31, "ce": 3.462694162270054, "rank": 0.02046060562133789} +{"step": 8, "steps": 500, "lr": 2e-05, "grad_norm": 0.8217577934265137, "loss_scale": 4.0, "retries": 0, "seconds": 11.7, "max_memory_gb": 14.31, "ce": 3.7079845648258924, "rank": 0.2099754810333252} +{"step": 9, "steps": 500, "lr": 2e-05, "grad_norm": 0.5095857381820679, "loss_scale": 4.0, "retries": 0, "seconds": 13.1, "max_memory_gb": 14.31, "ce": 2.836483597755432, "rank": 0.04407358169555664} +{"step": 10, "steps": 500, "lr": 2e-05, "grad_norm": 0.31661465764045715, "loss_scale": 4.0, "retries": 0, "seconds": 14.5, "max_memory_gb": 14.31, "ce": 1.6167410146445036, "rank": 0.0} +{"step": 11, "steps": 500, "lr": 2e-05, "grad_norm": 0.6944901347160339, "loss_scale": 4.0, "retries": 0, "seconds": 15.8, "max_memory_gb": 14.31, "ce": 2.625632087700069, "rank": 0.21860462427139282} +{"step": 12, "steps": 500, "lr": 2e-05, "grad_norm": 0.3647211790084839, "loss_scale": 4.0, "retries": 0, "seconds": 17.2, "max_memory_gb": 14.31, "ce": 1.5099505204707384, "rank": 0.052081942558288574} +{"step": 13, "steps": 500, "lr": 2e-05, "grad_norm": 0.6702440977096558, "loss_scale": 4.0, "retries": 0, "seconds": 18.5, "max_memory_gb": 14.31, "ce": 2.671206133789383, "rank": 0.0} +{"step": 14, "steps": 500, "lr": 2e-05, "grad_norm": 0.9867542386054993, "loss_scale": 4.0, "retries": 0, "seconds": 19.9, "max_memory_gb": 14.31, "ce": 3.865676431916654, "rank": 0.1474834680557251} +{"step": 15, "steps": 500, "lr": 2e-05, "grad_norm": 0.7263439297676086, "loss_scale": 4.0, "retries": 0, "seconds": 21.2, "max_memory_gb": 14.31, "ce": 2.5970110403141007, "rank": 0.17317676544189453} +{"step": 16, "steps": 500, "lr": 2e-05, "grad_norm": 0.43567439913749695, "loss_scale": 4.0, "retries": 0, "seconds": 22.5, "max_memory_gb": 14.31, "ce": 1.414091745391488, "rank": 0.0} +{"step": 17, "steps": 500, "lr": 2e-05, "grad_norm": 0.4414915442466736, "loss_scale": 4.0, "retries": 0, "seconds": 23.9, "max_memory_gb": 14.31, "ce": 1.5953994005685672, "rank": 0.03785908222198486} +{"step": 18, "steps": 500, "lr": 2e-05, "grad_norm": 1.2847321033477783, "loss_scale": 4.0, "retries": 0, "seconds": 25.2, "max_memory_gb": 14.31, "ce": 3.5728610632941127, "rank": 0.19652318954467773} +{"step": 19, "steps": 500, "lr": 2e-05, "grad_norm": 1.9365154504776, "loss_scale": 4.0, "retries": 0, "seconds": 26.6, "max_memory_gb": 14.31, "ce": 5.247206646949053, "rank": 0.32729458808898926} +{"step": 20, "steps": 500, "lr": 2e-05, "grad_norm": 0.019099319353699684, "loss_scale": 4.0, "retries": 0, "seconds": 28.0, "max_memory_gb": 14.31, "ce": 0.06544453476089984, "rank": 0.10915315896272659} +{"step": 21, "steps": 500, "lr": 2e-05, "grad_norm": 1.00411057472229, "loss_scale": 4.0, "retries": 0, "seconds": 29.3, "max_memory_gb": 14.31, "ce": 2.7227964010089636, "rank": 0.17668533325195312} +{"step": 22, "steps": 500, "lr": 2e-05, "grad_norm": 1.003851056098938, "loss_scale": 4.0, "retries": 0, "seconds": 30.7, "max_memory_gb": 14.31, "ce": 2.9301546281203628, "rank": 0.18962585926055908} +{"step": 23, "steps": 500, "lr": 2e-05, "grad_norm": 1.134323000907898, "loss_scale": 4.0, "retries": 0, "seconds": 32.0, "max_memory_gb": 14.31, "ce": 2.7193700712523423, "rank": 0.12515473365783691} +{"step": 24, "steps": 500, "lr": 2e-05, "grad_norm": 1.966476559638977, "loss_scale": 4.0, "retries": 0, "seconds": 33.4, "max_memory_gb": 14.31, "ce": 4.95299653057009, "rank": 0.2552229166030884} +{"step": 25, "steps": 500, "lr": 2e-05, "grad_norm": 1.881178379058838, "loss_scale": 4.0, "retries": 0, "seconds": 34.8, "max_memory_gb": 14.31, "ce": 4.073667999356985, "rank": 0.06383180618286133} +{"step": 26, "steps": 500, "lr": 2e-05, "grad_norm": 2.155491352081299, "loss_scale": 4.0, "retries": 0, "seconds": 36.1, "max_memory_gb": 14.31, "ce": 5.112923946231604, "rank": 0.07656586170196533} +{"step": 27, "steps": 500, "lr": 2e-05, "grad_norm": 1.693433403968811, "loss_scale": 4.0, "retries": 0, "seconds": 37.5, "max_memory_gb": 14.31, "ce": 4.152062482200563, "rank": 0.5148054361343384} +{"step": 28, "steps": 500, "lr": 2e-05, "grad_norm": 1.1003565788269043, "loss_scale": 4.0, "retries": 0, "seconds": 38.9, "max_memory_gb": 14.31, "ce": 2.7294855192303658, "rank": 0.0} +{"step": 29, "steps": 500, "lr": 2e-05, "grad_norm": 1.1086714267730713, "loss_scale": 4.0, "retries": 0, "seconds": 40.2, "max_memory_gb": 14.31, "ce": 2.6307738864561543, "rank": 0.0} +{"step": 30, "steps": 500, "lr": 2e-05, "grad_norm": 1.4200576543807983, "loss_scale": 4.0, "retries": 0, "seconds": 41.6, "max_memory_gb": 14.31, "ce": 2.7123742644907907, "rank": 0.0009807348251342773} +{"step": 31, "steps": 500, "lr": 2e-05, "grad_norm": 1.4574388265609741, "loss_scale": 4.0, "retries": 0, "seconds": 42.9, "max_memory_gb": 14.31, "ce": 2.7590548945590854, "rank": 0.1700146198272705} +{"step": 32, "steps": 500, "lr": 2e-05, "grad_norm": 2.7253990173339844, "loss_scale": 4.0, "retries": 0, "seconds": 44.3, "max_memory_gb": 14.31, "ce": 5.168157995678484, "rank": 0.27642500400543213} +{"step": 33, "steps": 500, "lr": 2e-05, "grad_norm": 1.3941540718078613, "loss_scale": 4.0, "retries": 0, "seconds": 45.7, "max_memory_gb": 14.31, "ce": 2.4729026085697114, "rank": 0.011544600129127502} +{"step": 34, "steps": 500, "lr": 2e-05, "grad_norm": 1.5069210529327393, "loss_scale": 4.0, "retries": 0, "seconds": 47.0, "max_memory_gb": 14.31, "ce": 2.6635346645489335, "rank": 0.0} +{"step": 35, "steps": 500, "lr": 2e-05, "grad_norm": 1.558336853981018, "loss_scale": 4.0, "retries": 0, "seconds": 48.4, "max_memory_gb": 14.31, "ce": 2.6530719762668014, "rank": 0.15487384796142578} +{"step": 36, "steps": 500, "lr": 2e-05, "grad_norm": 1.474217414855957, "loss_scale": 4.0, "retries": 0, "seconds": 49.8, "max_memory_gb": 14.31, "ce": 2.520920040551573, "rank": 0.0} +{"step": 37, "steps": 500, "lr": 2e-05, "grad_norm": 1.5959211587905884, "loss_scale": 4.0, "retries": 0, "seconds": 51.1, "max_memory_gb": 14.31, "ce": 2.5409758714959025, "rank": 0.0008981227874755859} +{"step": 38, "steps": 500, "lr": 2e-05, "grad_norm": 2.956392526626587, "loss_scale": 4.0, "retries": 0, "seconds": 52.5, "max_memory_gb": 14.31, "ce": 4.687647576443851, "rank": 0.18939507007598877} +{"step": 39, "steps": 500, "lr": 2e-05, "grad_norm": 1.0362932682037354, "loss_scale": 4.0, "retries": 0, "seconds": 53.8, "max_memory_gb": 14.31, "ce": 1.4136308385059237, "rank": 0.19987370073795319} +{"step": 40, "steps": 500, "lr": 2e-05, "grad_norm": 1.8836342096328735, "loss_scale": 4.0, "retries": 0, "seconds": 55.1, "max_memory_gb": 14.31, "ce": 2.581737110391259, "rank": 0.056325554847717285} +{"step": 41, "steps": 500, "lr": 2e-05, "grad_norm": 1.1571238040924072, "loss_scale": 4.0, "retries": 0, "seconds": 56.5, "max_memory_gb": 14.31, "ce": 1.4606882482767105, "rank": 0.0} +{"step": 42, "steps": 500, "lr": 2e-05, "grad_norm": 1.6048260927200317, "loss_scale": 4.0, "retries": 0, "seconds": 57.8, "max_memory_gb": 14.31, "ce": 2.324425932718441, "rank": 0.0} +{"step": 43, "steps": 500, "lr": 2e-05, "grad_norm": 1.8937959671020508, "loss_scale": 4.0, "retries": 0, "seconds": 59.2, "max_memory_gb": 14.31, "ce": 2.4064048405271024, "rank": 0.001911759376525879} +{"step": 44, "steps": 500, "lr": 2e-05, "grad_norm": 1.170283317565918, "loss_scale": 4.0, "retries": 0, "seconds": 60.5, "max_memory_gb": 14.31, "ce": 1.3791909744613804, "rank": 0.0} +{"step": 45, "steps": 500, "lr": 2e-05, "grad_norm": 0.9147579669952393, "loss_scale": 4.0, "retries": 0, "seconds": 61.9, "max_memory_gb": 14.31, "ce": 1.3839610782451928, "rank": 0.0} +{"step": 46, "steps": 500, "lr": 2e-05, "grad_norm": 2.0050065517425537, "loss_scale": 4.0, "retries": 0, "seconds": 63.2, "max_memory_gb": 14.31, "ce": 2.297932301647961, "rank": 0.0} +{"step": 47, "steps": 500, "lr": 2e-05, "grad_norm": 2.1766345500946045, "loss_scale": 4.0, "retries": 0, "seconds": 64.6, "max_memory_gb": 14.31, "ce": 2.0022591419983655, "rank": 0.0} +{"step": 48, "steps": 500, "lr": 2e-05, "grad_norm": 3.3576440811157227, "loss_scale": 4.0, "retries": 0, "seconds": 66.0, "max_memory_gb": 14.31, "ce": 3.673652924131602, "rank": 0.1552942991256714} +{"step": 49, "steps": 500, "lr": 2e-05, "grad_norm": 3.1804282665252686, "loss_scale": 4.0, "retries": 0, "seconds": 67.3, "max_memory_gb": 14.31, "ce": 3.4361499208025634, "rank": 0.06885433197021484} +{"step": 50, "steps": 500, "lr": 2e-05, "grad_norm": 3.271462917327881, "loss_scale": 4.0, "retries": 0, "seconds": 68.7, "max_memory_gb": 14.31, "ce": 3.51061950949952, "rank": 0.057526588439941406} +{"step": 51, "steps": 500, "lr": 2e-05, "grad_norm": 1.3777588605880737, "loss_scale": 4.0, "retries": 0, "seconds": 70.0, "max_memory_gb": 14.31, "ce": 1.3493488191161305, "rank": 0.07979404926300049} +{"step": 52, "steps": 500, "lr": 2e-05, "grad_norm": 2.8119893074035645, "loss_scale": 4.0, "retries": 0, "seconds": 71.3, "max_memory_gb": 14.31, "ce": 2.2723999116569757, "rank": 0.0} +{"step": 53, "steps": 500, "lr": 2e-05, "grad_norm": 3.1664912700653076, "loss_scale": 4.0, "retries": 0, "seconds": 72.6, "max_memory_gb": 14.31, "ce": 3.559533921070397, "rank": 0.04917478561401367} +{"step": 54, "steps": 500, "lr": 2e-05, "grad_norm": 3.4419949054718018, "loss_scale": 4.0, "retries": 0, "seconds": 74.0, "max_memory_gb": 14.31, "ce": 3.6832037828862667, "rank": 0.0} +{"step": 55, "steps": 500, "lr": 2e-05, "grad_norm": 2.1818506717681885, "loss_scale": 4.0, "retries": 0, "seconds": 75.3, "max_memory_gb": 14.31, "ce": 2.3930055055534467, "rank": 0.0} +{"step": 56, "steps": 500, "lr": 2e-05, "grad_norm": 4.57388973236084, "loss_scale": 4.0, "retries": 0, "seconds": 76.7, "max_memory_gb": 14.31, "ce": 4.636048750020564, "rank": 0.043846726417541504} +{"step": 57, "steps": 500, "lr": 2e-05, "grad_norm": 1.1142300367355347, "loss_scale": 4.0, "retries": 0, "seconds": 78.0, "max_memory_gb": 14.31, "ce": 1.2400669199414551, "rank": 0.04264768958091736} +{"step": 58, "steps": 500, "lr": 2e-05, "grad_norm": 2.6720917224884033, "loss_scale": 4.0, "retries": 0, "seconds": 79.4, "max_memory_gb": 14.31, "ce": 2.29901250556577, "rank": 0.0264894962310791} +{"step": 59, "steps": 500, "lr": 2e-05, "grad_norm": 1.3121986389160156, "loss_scale": 4.0, "retries": 0, "seconds": 80.7, "max_memory_gb": 14.31, "ce": 1.3790817093104124, "rank": 0.03269350528717041} +{"step": 60, "steps": 500, "lr": 2e-05, "grad_norm": 2.3573453426361084, "loss_scale": 4.0, "retries": 0, "seconds": 82.1, "max_memory_gb": 14.31, "ce": 2.3203381432686, "rank": 0.0} +{"step": 61, "steps": 500, "lr": 2e-05, "grad_norm": 3.9964067935943604, "loss_scale": 4.0, "retries": 0, "seconds": 83.4, "max_memory_gb": 14.31, "ce": 3.3115389961749315, "rank": 0.0661691427230835} +{"step": 62, "steps": 500, "lr": 2e-05, "grad_norm": 1.419227957725525, "loss_scale": 4.0, "retries": 0, "seconds": 84.8, "max_memory_gb": 14.31, "ce": 1.3012341097928584, "rank": 0.0} +{"step": 63, "steps": 500, "lr": 2e-05, "grad_norm": 3.8364999294281006, "loss_scale": 4.0, "retries": 0, "seconds": 86.1, "max_memory_gb": 14.31, "ce": 3.4022557933349162, "rank": 0.2538588047027588} +{"step": 64, "steps": 500, "lr": 2e-05, "grad_norm": 3.9919590950012207, "loss_scale": 4.0, "retries": 0, "seconds": 87.5, "max_memory_gb": 14.31, "ce": 3.407562547829002, "rank": 0.04550516605377197} +{"step": 65, "steps": 500, "lr": 2e-05, "grad_norm": 3.749711036682129, "loss_scale": 4.0, "retries": 0, "seconds": 88.8, "max_memory_gb": 14.31, "ce": 3.5524015398696065, "rank": 0.0} +{"step": 66, "steps": 500, "lr": 2e-05, "grad_norm": 1.3723782300949097, "loss_scale": 4.0, "retries": 0, "seconds": 90.2, "max_memory_gb": 14.31, "ce": 0.9459767605294473, "rank": 0.2529519200325012} +{"step": 67, "steps": 500, "lr": 2e-05, "grad_norm": 4.331387042999268, "loss_scale": 4.0, "retries": 0, "seconds": 91.6, "max_memory_gb": 14.31, "ce": 3.233394594863057, "rank": 0.0} +{"step": 68, "steps": 500, "lr": 2e-05, "grad_norm": 1.4460819959640503, "loss_scale": 4.0, "retries": 0, "seconds": 93.0, "max_memory_gb": 14.31, "ce": 1.138377872412093, "rank": 0.0} +{"step": 69, "steps": 500, "lr": 2e-05, "grad_norm": 2.5917036533355713, "loss_scale": 4.0, "retries": 0, "seconds": 94.3, "max_memory_gb": 14.31, "ce": 2.1407433738932014, "rank": 0.0} +{"step": 70, "steps": 500, "lr": 2e-05, "grad_norm": 3.0721895694732666, "loss_scale": 4.0, "retries": 0, "seconds": 95.6, "max_memory_gb": 14.31, "ce": 2.094460680615157, "rank": 0.09379947185516357} +{"step": 71, "steps": 500, "lr": 2e-05, "grad_norm": 4.537412643432617, "loss_scale": 4.0, "retries": 0, "seconds": 96.9, "max_memory_gb": 14.31, "ce": 3.054632214247249, "rank": 0.10216116905212402} +{"step": 72, "steps": 500, "lr": 2e-05, "grad_norm": 5.171882152557373, "loss_scale": 4.0, "retries": 0, "seconds": 98.3, "max_memory_gb": 14.31, "ce": 3.701306008733809, "rank": 0.06732034683227539} +{"step": 73, "steps": 500, "lr": 2e-05, "grad_norm": 0.09700620919466019, "loss_scale": 4.0, "retries": 0, "seconds": 99.6, "max_memory_gb": 14.31, "ce": 0.24982540763448924, "rank": 0.0} +{"step": 74, "steps": 500, "lr": 2e-05, "grad_norm": 3.086634635925293, "loss_scale": 4.0, "retries": 0, "seconds": 100.9, "max_memory_gb": 14.31, "ce": 2.0855714278295636, "rank": 0.1676180362701416} +{"step": 75, "steps": 500, "lr": 2e-05, "grad_norm": 1.950239896774292, "loss_scale": 4.0, "retries": 0, "seconds": 102.3, "max_memory_gb": 14.31, "ce": 1.5789746609516442, "rank": 0.1417863667011261} +{"step": 76, "steps": 500, "lr": 2e-05, "grad_norm": 2.9704537391662598, "loss_scale": 4.0, "retries": 0, "seconds": 103.6, "max_memory_gb": 14.31, "ce": 2.0528353748377413, "rank": 0.0} +{"step": 77, "steps": 500, "lr": 2e-05, "grad_norm": 4.667126178741455, "loss_scale": 4.0, "retries": 0, "seconds": 104.9, "max_memory_gb": 14.31, "ce": 3.173520459793508, "rank": 0.09807455539703369} +{"step": 78, "steps": 500, "lr": 2e-05, "grad_norm": 3.231656789779663, "loss_scale": 4.0, "retries": 0, "seconds": 106.2, "max_memory_gb": 14.31, "ce": 1.7530763102695346, "rank": 0.0} +{"step": 79, "steps": 500, "lr": 2e-05, "grad_norm": 2.005286931991577, "loss_scale": 4.0, "retries": 0, "seconds": 107.5, "max_memory_gb": 14.31, "ce": 0.9072676589712501, "rank": 0.02075093984603882} +{"step": 80, "steps": 500, "lr": 2e-05, "grad_norm": 3.384627103805542, "loss_scale": 4.0, "retries": 0, "seconds": 108.8, "max_memory_gb": 14.31, "ce": 1.8996085416292772, "rank": 0.0} +{"step": 81, "steps": 500, "lr": 2e-05, "grad_norm": 4.932344436645508, "loss_scale": 4.0, "retries": 0, "seconds": 110.2, "max_memory_gb": 14.31, "ce": 2.8146961154416203, "rank": 0.02146315574645996} +{"step": 82, "steps": 500, "lr": 2e-05, "grad_norm": 4.981822967529297, "loss_scale": 4.0, "retries": 0, "seconds": 111.5, "max_memory_gb": 14.31, "ce": 2.725998869165778, "rank": 0.0038707852363586426} +{"step": 83, "steps": 500, "lr": 2e-05, "grad_norm": 4.7275285720825195, "loss_scale": 4.0, "retries": 0, "seconds": 112.9, "max_memory_gb": 14.31, "ce": 2.4433098066365346, "rank": 0.02610301971435547} +{"step": 84, "steps": 500, "lr": 2e-05, "grad_norm": 1.6490408182144165, "loss_scale": 4.0, "retries": 0, "seconds": 114.3, "max_memory_gb": 14.31, "ce": 1.0381455202004872, "rank": 0.0} +{"step": 85, "steps": 500, "lr": 2e-05, "grad_norm": 0.08381516486406326, "loss_scale": 4.0, "retries": 0, "seconds": 115.6, "max_memory_gb": 14.31, "ce": 0.12963290145853534, "rank": 0.0} +{"step": 86, "steps": 500, "lr": 2e-05, "grad_norm": 4.795912265777588, "loss_scale": 4.0, "retries": 0, "seconds": 117.0, "max_memory_gb": 14.31, "ce": 2.6204302958212793, "rank": 0.0} +{"step": 87, "steps": 500, "lr": 2e-05, "grad_norm": 4.9938225746154785, "loss_scale": 4.0, "retries": 0, "seconds": 118.3, "max_memory_gb": 14.31, "ce": 2.590469185030088, "rank": 0.0} +{"step": 88, "steps": 500, "lr": 2e-05, "grad_norm": 3.5783169269561768, "loss_scale": 4.0, "retries": 0, "seconds": 119.6, "max_memory_gb": 14.31, "ce": 1.7401738140033558, "rank": 0.13016754388809204} +{"step": 89, "steps": 500, "lr": 2e-05, "grad_norm": 0.05859130620956421, "loss_scale": 4.0, "retries": 0, "seconds": 121.0, "max_memory_gb": 14.31, "ce": 0.1574817833607085, "rank": 0.12445016205310822} +{"step": 90, "steps": 500, "lr": 2e-05, "grad_norm": 2.9462242126464844, "loss_scale": 4.0, "retries": 0, "seconds": 122.3, "max_memory_gb": 14.31, "ce": 1.526573105249554, "rank": 0.044087886810302734} +{"step": 91, "steps": 500, "lr": 2e-05, "grad_norm": 0.1546722799539566, "loss_scale": 4.0, "retries": 0, "seconds": 123.7, "max_memory_gb": 14.31, "ce": 0.3614959791302681, "rank": 0.0} +{"step": 92, "steps": 500, "lr": 2e-05, "grad_norm": 2.002265214920044, "loss_scale": 4.0, "retries": 0, "seconds": 125.1, "max_memory_gb": 14.31, "ce": 0.9359601612668484, "rank": 0.0} +{"step": 93, "steps": 500, "lr": 2e-05, "grad_norm": 4.5854573249816895, "loss_scale": 4.0, "retries": 0, "seconds": 126.4, "max_memory_gb": 14.31, "ce": 1.4318201631540433, "rank": 0.0} +{"step": 94, "steps": 500, "lr": 2e-05, "grad_norm": 5.637260913848877, "loss_scale": 4.0, "retries": 0, "seconds": 127.8, "max_memory_gb": 14.31, "ce": 2.479329627007246, "rank": 0.09149223566055298} +{"step": 95, "steps": 500, "lr": 2e-05, "grad_norm": 3.5820491313934326, "loss_scale": 4.0, "retries": 0, "seconds": 129.2, "max_memory_gb": 14.31, "ce": 1.705605800729245, "rank": 0.0} +{"step": 96, "steps": 500, "lr": 2e-05, "grad_norm": 5.604701042175293, "loss_scale": 4.0, "retries": 0, "seconds": 130.5, "max_memory_gb": 14.31, "ce": 2.456278449855745, "rank": 0.03800041973590851} +{"step": 97, "steps": 500, "lr": 2e-05, "grad_norm": 1.953590750694275, "loss_scale": 4.0, "retries": 0, "seconds": 131.9, "max_memory_gb": 14.31, "ce": 1.0668957815505564, "rank": 0.0} +{"step": 98, "steps": 500, "lr": 2e-05, "grad_norm": 3.8440306186676025, "loss_scale": 4.0, "retries": 0, "seconds": 133.2, "max_memory_gb": 14.31, "ce": 1.4851482714293525, "rank": 0.00023686885833740234} +{"step": 99, "steps": 500, "lr": 2e-05, "grad_norm": 0.1552009880542755, "loss_scale": 4.0, "retries": 0, "seconds": 134.6, "max_memory_gb": 14.31, "ce": 0.24177075875923038, "rank": 0.026371553540229797} +{"step": 100, "steps": 500, "lr": 2e-05, "grad_norm": 5.118680000305176, "loss_scale": 4.0, "retries": 0, "seconds": 135.9, "max_memory_gb": 14.31, "ce": 2.242949071340263, "rank": 0.0} +{"step": 101, "steps": 500, "lr": 2e-05, "grad_norm": 5.516894817352295, "loss_scale": 4.0, "retries": 0, "seconds": 137.2, "max_memory_gb": 14.31, "ce": 2.1224722201004624, "rank": 0.0} +{"step": 102, "steps": 500, "lr": 2e-05, "grad_norm": 5.291504859924316, "loss_scale": 4.0, "retries": 0, "seconds": 138.5, "max_memory_gb": 14.31, "ce": 2.327276927884668, "rank": 0.0} +{"step": 103, "steps": 500, "lr": 2e-05, "grad_norm": 3.7329277992248535, "loss_scale": 4.0, "retries": 0, "seconds": 139.9, "max_memory_gb": 14.31, "ce": 1.389047902659513, "rank": 0.1100618839263916} +{"step": 104, "steps": 500, "lr": 2e-05, "grad_norm": 2.267913341522217, "loss_scale": 4.0, "retries": 0, "seconds": 141.1, "max_memory_gb": 14.31, "ce": 0.7620585324475542, "rank": 0.1551290899515152} +{"step": 105, "steps": 500, "lr": 2e-05, "grad_norm": 5.764068603515625, "loss_scale": 4.0, "retries": 0, "seconds": 142.5, "max_memory_gb": 14.31, "ce": 2.018013084307313, "rank": 0.12434715032577515} +{"step": 106, "steps": 500, "lr": 2e-05, "grad_norm": 5.865462779998779, "loss_scale": 4.0, "retries": 0, "seconds": 143.8, "max_memory_gb": 14.31, "ce": 1.8142284443601966, "rank": 0.05965691804885864} +{"step": 107, "steps": 500, "lr": 2e-05, "grad_norm": 1.9442158937454224, "loss_scale": 4.0, "retries": 0, "seconds": 145.2, "max_memory_gb": 14.31, "ce": 0.7914399146102369, "rank": 0.0} +{"step": 108, "steps": 500, "lr": 2e-05, "grad_norm": 2.017535448074341, "loss_scale": 4.0, "retries": 0, "seconds": 146.6, "max_memory_gb": 14.31, "ce": 0.803606124361977, "rank": 0.0} +{"step": 109, "steps": 500, "lr": 2e-05, "grad_norm": 6.000257968902588, "loss_scale": 4.0, "retries": 0, "seconds": 147.9, "max_memory_gb": 14.31, "ce": 2.028072520159185, "rank": 0.017304301261901855} +{"step": 110, "steps": 500, "lr": 2e-05, "grad_norm": 3.955190658569336, "loss_scale": 4.0, "retries": 0, "seconds": 149.3, "max_memory_gb": 14.31, "ce": 1.3096599243581295, "rank": 0.09335505962371826} +{"step": 111, "steps": 500, "lr": 2e-05, "grad_norm": 3.7003371715545654, "loss_scale": 4.0, "retries": 0, "seconds": 150.7, "max_memory_gb": 14.31, "ce": 1.2005671507213265, "rank": 0.0} +{"step": 112, "steps": 500, "lr": 2e-05, "grad_norm": 4.115142345428467, "loss_scale": 4.0, "retries": 0, "seconds": 152.0, "max_memory_gb": 14.31, "ce": 1.1536603369750082, "rank": 0.020064711570739746} +{"step": 113, "steps": 500, "lr": 2e-05, "grad_norm": 4.429904937744141, "loss_scale": 4.0, "retries": 0, "seconds": 153.4, "max_memory_gb": 14.31, "ce": 1.2967917555943131, "rank": 0.0926048755645752} +{"step": 114, "steps": 500, "lr": 2e-05, "grad_norm": 6.217803955078125, "loss_scale": 4.0, "retries": 0, "seconds": 154.8, "max_memory_gb": 14.31, "ce": 1.7954589324072003, "rank": 0.18377822637557983} +{"step": 115, "steps": 500, "lr": 2e-05, "grad_norm": 4.086572647094727, "loss_scale": 4.0, "retries": 0, "seconds": 156.1, "max_memory_gb": 14.31, "ce": 1.2105585848912597, "rank": 0.05789032578468323} +{"step": 116, "steps": 500, "lr": 2e-05, "grad_norm": 0.11020702868700027, "loss_scale": 4.0, "retries": 0, "seconds": 157.4, "max_memory_gb": 14.31, "ce": 0.1529178066411987, "rank": 0.0} +{"step": 117, "steps": 500, "lr": 2e-05, "grad_norm": 3.950124740600586, "loss_scale": 4.0, "retries": 0, "seconds": 158.8, "max_memory_gb": 14.31, "ce": 1.2136256974190474, "rank": 0.0} +{"step": 118, "steps": 500, "lr": 2e-05, "grad_norm": 2.203831195831299, "loss_scale": 4.0, "retries": 0, "seconds": 160.1, "max_memory_gb": 14.31, "ce": 0.6558849012944847, "rank": 0.0} +{"step": 119, "steps": 500, "lr": 2e-05, "grad_norm": 2.335526943206787, "loss_scale": 4.0, "retries": 0, "seconds": 161.5, "max_memory_gb": 14.31, "ce": 0.5249930475838482, "rank": 0.0} +{"step": 120, "steps": 500, "lr": 2e-05, "grad_norm": 3.74162220954895, "loss_scale": 4.0, "retries": 0, "seconds": 162.9, "max_memory_gb": 14.31, "ce": 0.9996319960337132, "rank": 0.0} +{"step": 121, "steps": 500, "lr": 2e-05, "grad_norm": 2.440502405166626, "loss_scale": 4.0, "retries": 0, "seconds": 164.2, "max_memory_gb": 14.31, "ce": 0.700706870527938, "rank": 0.007779538631439209} +{"step": 122, "steps": 500, "lr": 2e-05, "grad_norm": 3.311145782470703, "loss_scale": 4.0, "retries": 0, "seconds": 165.6, "max_memory_gb": 14.31, "ce": 0.8880284586921334, "rank": 0.0} +{"step": 123, "steps": 500, "lr": 2e-05, "grad_norm": 2.097191095352173, "loss_scale": 4.0, "retries": 0, "seconds": 166.9, "max_memory_gb": 14.31, "ce": 0.8535577941220254, "rank": 0.004590138792991638} +{"step": 124, "steps": 500, "lr": 2e-05, "grad_norm": 1.9120830297470093, "loss_scale": 4.0, "retries": 0, "seconds": 168.2, "max_memory_gb": 14.31, "ce": 0.4895694594597444, "rank": 0.0} +{"step": 125, "steps": 500, "lr": 2e-05, "grad_norm": 1.9118561744689941, "loss_scale": 4.0, "retries": 0, "seconds": 169.5, "max_memory_gb": 14.31, "ce": 0.5215689809992909, "rank": 0.0} +{"step": 126, "steps": 500, "lr": 2e-05, "grad_norm": 3.751549243927002, "loss_scale": 4.0, "retries": 0, "seconds": 170.9, "max_memory_gb": 14.31, "ce": 1.0806160122156143, "rank": 0.0} +{"step": 127, "steps": 500, "lr": 2e-05, "grad_norm": 1.7474470138549805, "loss_scale": 4.0, "retries": 0, "seconds": 172.2, "max_memory_gb": 14.31, "ce": 0.3279006983502768, "rank": 0.0} +{"step": 128, "steps": 500, "lr": 2e-05, "grad_norm": 3.546689510345459, "loss_scale": 4.0, "retries": 0, "seconds": 173.6, "max_memory_gb": 14.31, "ce": 0.7186421588994563, "rank": 0.0} +{"step": 129, "steps": 500, "lr": 2e-05, "grad_norm": 6.446577072143555, "loss_scale": 4.0, "retries": 0, "seconds": 174.9, "max_memory_gb": 14.31, "ce": 1.2323222868144512, "rank": 0.0} +{"step": 130, "steps": 500, "lr": 2e-05, "grad_norm": 3.5455572605133057, "loss_scale": 4.0, "retries": 0, "seconds": 176.3, "max_memory_gb": 14.31, "ce": 0.887803616002202, "rank": 0.0} +{"step": 131, "steps": 500, "lr": 2e-05, "grad_norm": 1.960054636001587, "loss_scale": 4.0, "retries": 0, "seconds": 177.6, "max_memory_gb": 14.31, "ce": 0.496754031162709, "rank": 0.0} +{"step": 132, "steps": 500, "lr": 2e-05, "grad_norm": 4.356414318084717, "loss_scale": 4.0, "retries": 0, "seconds": 179.0, "max_memory_gb": 14.31, "ce": 1.0330867152661085, "rank": 0.036717742681503296} +{"step": 133, "steps": 500, "lr": 2e-05, "grad_norm": 4.619503021240234, "loss_scale": 4.0, "retries": 0, "seconds": 180.4, "max_memory_gb": 14.31, "ce": 0.9804302449338138, "rank": 0.06555649638175964} +{"step": 134, "steps": 500, "lr": 2e-05, "grad_norm": 0.16171608865261078, "loss_scale": 4.0, "retries": 0, "seconds": 181.7, "max_memory_gb": 14.31, "ce": 0.221445006551221, "rank": 0.0} +{"step": 135, "steps": 500, "lr": 2e-05, "grad_norm": 3.9975719451904297, "loss_scale": 4.0, "retries": 0, "seconds": 183.0, "max_memory_gb": 14.31, "ce": 0.8252845886163414, "rank": 0.0} +{"step": 136, "steps": 500, "lr": 2e-05, "grad_norm": 3.561971664428711, "loss_scale": 4.0, "retries": 0, "seconds": 184.3, "max_memory_gb": 14.31, "ce": 0.6160265428479761, "rank": 0.0} +{"step": 137, "steps": 500, "lr": 2e-05, "grad_norm": 2.0251669883728027, "loss_scale": 4.0, "retries": 0, "seconds": 185.7, "max_memory_gb": 14.31, "ce": 0.4526874851435423, "rank": 0.0} +{"step": 138, "steps": 500, "lr": 2e-05, "grad_norm": 2.1880805492401123, "loss_scale": 4.0, "retries": 0, "seconds": 187.1, "max_memory_gb": 14.31, "ce": 0.5462467954494059, "rank": 0.0} +{"step": 139, "steps": 500, "lr": 2e-05, "grad_norm": 3.022434949874878, "loss_scale": 4.0, "retries": 0, "seconds": 188.5, "max_memory_gb": 14.31, "ce": 0.8314622375182807, "rank": 0.0} +{"step": 140, "steps": 500, "lr": 2e-05, "grad_norm": 1.3125886917114258, "loss_scale": 4.0, "retries": 0, "seconds": 189.8, "max_memory_gb": 14.31, "ce": 0.3600932857953012, "rank": 0.0} +{"step": 141, "steps": 500, "lr": 2e-05, "grad_norm": 1.300040364265442, "loss_scale": 4.0, "retries": 0, "seconds": 191.2, "max_memory_gb": 14.31, "ce": 0.4488518126308918, "rank": 0.0} +{"step": 142, "steps": 500, "lr": 2e-05, "grad_norm": 1.742870569229126, "loss_scale": 4.0, "retries": 0, "seconds": 192.4, "max_memory_gb": 14.31, "ce": 0.44727825187146664, "rank": 0.0} +{"step": 143, "steps": 500, "lr": 2e-05, "grad_norm": 0.1453661024570465, "loss_scale": 4.0, "retries": 0, "seconds": 193.7, "max_memory_gb": 14.31, "ce": 0.1910882566589862, "rank": 0.0} +{"step": 144, "steps": 500, "lr": 2e-05, "grad_norm": 2.680030107498169, "loss_scale": 4.0, "retries": 0, "seconds": 195.0, "max_memory_gb": 14.31, "ce": 0.6867644747253507, "rank": 0.0} +{"step": 145, "steps": 500, "lr": 2e-05, "grad_norm": 1.5584689378738403, "loss_scale": 4.0, "retries": 0, "seconds": 196.4, "max_memory_gb": 14.31, "ce": 0.35113837313838303, "rank": 0.0} +{"step": 146, "steps": 500, "lr": 2e-05, "grad_norm": 0.970578134059906, "loss_scale": 4.0, "retries": 0, "seconds": 197.7, "max_memory_gb": 14.31, "ce": 0.3991182981990278, "rank": 0.0} +{"step": 147, "steps": 500, "lr": 2e-05, "grad_norm": 1.1611567735671997, "loss_scale": 4.0, "retries": 0, "seconds": 199.1, "max_memory_gb": 14.31, "ce": 0.5520564699545503, "rank": 0.0} +{"step": 148, "steps": 500, "lr": 2e-05, "grad_norm": 0.5796814560890198, "loss_scale": 4.0, "retries": 0, "seconds": 200.5, "max_memory_gb": 14.31, "ce": 0.1927857703412883, "rank": 0.0} +{"step": 149, "steps": 500, "lr": 2e-05, "grad_norm": 1.1084696054458618, "loss_scale": 4.0, "retries": 0, "seconds": 201.8, "max_memory_gb": 14.31, "ce": 0.3017567249480635, "rank": 0.1377851516008377} +{"step": 150, "steps": 500, "lr": 2e-05, "grad_norm": 1.279615044593811, "loss_scale": 4.0, "retries": 0, "seconds": 203.2, "max_memory_gb": 14.31, "ce": 0.5765105420723557, "rank": 0.056415900588035583} +{"step": 151, "steps": 500, "lr": 2e-05, "grad_norm": 0.718294084072113, "loss_scale": 4.0, "retries": 0, "seconds": 204.5, "max_memory_gb": 14.31, "ce": 0.2859271151246503, "rank": 0.0} +{"step": 152, "steps": 500, "lr": 2e-05, "grad_norm": 0.8122401833534241, "loss_scale": 4.0, "retries": 0, "seconds": 205.8, "max_memory_gb": 14.31, "ce": 0.46768896002322435, "rank": 0.0} +{"step": 153, "steps": 500, "lr": 2e-05, "grad_norm": 1.2730308771133423, "loss_scale": 4.0, "retries": 0, "seconds": 207.2, "max_memory_gb": 14.31, "ce": 0.39021767722442746, "rank": 0.0} +{"step": 154, "steps": 500, "lr": 2e-05, "grad_norm": 1.6756914854049683, "loss_scale": 4.0, "retries": 0, "seconds": 208.5, "max_memory_gb": 14.31, "ce": 0.7070410656742752, "rank": 0.07279562950134277} +{"step": 155, "steps": 500, "lr": 2e-05, "grad_norm": 2.418445110321045, "loss_scale": 4.0, "retries": 0, "seconds": 209.9, "max_memory_gb": 14.31, "ce": 0.6480851251399145, "rank": 0.04715266823768616} +{"step": 156, "steps": 500, "lr": 2e-05, "grad_norm": 0.5174362063407898, "loss_scale": 4.0, "retries": 0, "seconds": 211.3, "max_memory_gb": 14.31, "ce": 0.2514820205979049, "rank": 0.0} +{"step": 157, "steps": 500, "lr": 2e-05, "grad_norm": 0.7977359890937805, "loss_scale": 4.0, "retries": 0, "seconds": 212.6, "max_memory_gb": 14.31, "ce": 0.3347759759053588, "rank": 0.0} +{"step": 158, "steps": 500, "lr": 2e-05, "grad_norm": 1.2469720840454102, "loss_scale": 4.0, "retries": 0, "seconds": 214.0, "max_memory_gb": 14.31, "ce": 0.4541999853681773, "rank": 0.0} +{"step": 159, "steps": 500, "lr": 2e-05, "grad_norm": 0.6260176301002502, "loss_scale": 4.0, "retries": 0, "seconds": 215.4, "max_memory_gb": 14.31, "ce": 0.295444926712662, "rank": 0.0} +{"step": 160, "steps": 500, "lr": 2e-05, "grad_norm": 0.687295138835907, "loss_scale": 4.0, "retries": 0, "seconds": 216.7, "max_memory_gb": 14.31, "ce": 0.3901960044167936, "rank": 0.005753517150878906} +{"step": 161, "steps": 500, "lr": 2e-05, "grad_norm": 0.7493671178817749, "loss_scale": 4.0, "retries": 0, "seconds": 218.0, "max_memory_gb": 14.31, "ce": 0.3552045605611056, "rank": 0.0} +{"step": 162, "steps": 500, "lr": 2e-05, "grad_norm": 0.3408021032810211, "loss_scale": 4.0, "retries": 0, "seconds": 219.3, "max_memory_gb": 14.31, "ce": 0.23029677895829082, "rank": 0.0} +{"step": 163, "steps": 500, "lr": 2e-05, "grad_norm": 1.033517599105835, "loss_scale": 4.0, "retries": 0, "seconds": 220.6, "max_memory_gb": 14.31, "ce": 0.36038328520953655, "rank": 0.0} +{"step": 164, "steps": 500, "lr": 2e-05, "grad_norm": 1.445068359375, "loss_scale": 4.0, "retries": 0, "seconds": 222.0, "max_memory_gb": 14.31, "ce": 0.46834943210706115, "rank": 0.08379284292459488} +{"step": 165, "steps": 500, "lr": 2e-05, "grad_norm": 0.3978257179260254, "loss_scale": 4.0, "retries": 0, "seconds": 223.4, "max_memory_gb": 14.31, "ce": 0.27231571497395635, "rank": 0.0} +{"step": 166, "steps": 500, "lr": 2e-05, "grad_norm": 0.8252670168876648, "loss_scale": 4.0, "retries": 0, "seconds": 224.8, "max_memory_gb": 14.31, "ce": 0.36135111656039953, "rank": 0.0} +{"step": 167, "steps": 500, "lr": 2e-05, "grad_norm": 0.8695201873779297, "loss_scale": 4.0, "retries": 0, "seconds": 226.1, "max_memory_gb": 14.31, "ce": 0.4500443828292191, "rank": 0.0} +{"step": 168, "steps": 500, "lr": 2e-05, "grad_norm": 0.5380493402481079, "loss_scale": 4.0, "retries": 0, "seconds": 227.5, "max_memory_gb": 14.31, "ce": 0.24247626680880785, "rank": 0.0} +{"step": 169, "steps": 500, "lr": 2e-05, "grad_norm": 1.4333596229553223, "loss_scale": 4.0, "retries": 0, "seconds": 228.8, "max_memory_gb": 14.31, "ce": 0.5863015931099653, "rank": 0.05684284865856171} +{"step": 170, "steps": 500, "lr": 2e-05, "grad_norm": 0.669135570526123, "loss_scale": 4.0, "retries": 0, "seconds": 230.2, "max_memory_gb": 14.31, "ce": 0.244993464788422, "rank": 0.0} +{"step": 171, "steps": 500, "lr": 2e-05, "grad_norm": 0.8515205979347229, "loss_scale": 4.0, "retries": 0, "seconds": 231.5, "max_memory_gb": 14.31, "ce": 0.26505275070667267, "rank": 0.008173778653144836} +{"step": 172, "steps": 500, "lr": 2e-05, "grad_norm": 1.3444254398345947, "loss_scale": 4.0, "retries": 0, "seconds": 232.9, "max_memory_gb": 14.31, "ce": 0.3892852580174804, "rank": 0.0} +{"step": 173, "steps": 500, "lr": 2e-05, "grad_norm": 1.4218113422393799, "loss_scale": 4.0, "retries": 0, "seconds": 234.2, "max_memory_gb": 14.31, "ce": 0.35926265362650156, "rank": 0.011152461171150208} +{"step": 174, "steps": 500, "lr": 2e-05, "grad_norm": 1.0803910493850708, "loss_scale": 4.0, "retries": 0, "seconds": 235.5, "max_memory_gb": 14.31, "ce": 0.277830608189106, "rank": 0.0} +{"step": 175, "steps": 500, "lr": 2e-05, "grad_norm": 0.6255413293838501, "loss_scale": 4.0, "retries": 0, "seconds": 236.9, "max_memory_gb": 14.31, "ce": 0.3865073248744011, "rank": 0.0} +{"step": 176, "steps": 500, "lr": 2e-05, "grad_norm": 0.8208978176116943, "loss_scale": 4.0, "retries": 0, "seconds": 238.2, "max_memory_gb": 14.31, "ce": 0.2800054117105901, "rank": 0.03075443208217621} +{"step": 177, "steps": 500, "lr": 2e-05, "grad_norm": 1.667219877243042, "loss_scale": 4.0, "retries": 0, "seconds": 239.5, "max_memory_gb": 14.31, "ce": 0.385996175580658, "rank": 0.12933090329170227} +{"step": 178, "steps": 500, "lr": 2e-05, "grad_norm": 0.4278872609138489, "loss_scale": 4.0, "retries": 0, "seconds": 240.9, "max_memory_gb": 14.31, "ce": 0.3065151165938005, "rank": 0.0} +{"step": 179, "steps": 500, "lr": 2e-05, "grad_norm": 0.38249462842941284, "loss_scale": 4.0, "retries": 0, "seconds": 242.2, "max_memory_gb": 14.31, "ce": 0.432241590693593, "rank": 0.0} +{"step": 180, "steps": 500, "lr": 2e-05, "grad_norm": 1.576795220375061, "loss_scale": 4.0, "retries": 0, "seconds": 243.5, "max_memory_gb": 14.31, "ce": 0.43971875170245767, "rank": 0.0} +{"step": 181, "steps": 500, "lr": 2e-05, "grad_norm": 1.1197195053100586, "loss_scale": 4.0, "retries": 0, "seconds": 244.9, "max_memory_gb": 14.31, "ce": 0.3183818166144192, "rank": 0.012834832072257996} +{"step": 182, "steps": 500, "lr": 2e-05, "grad_norm": 0.7797282934188843, "loss_scale": 4.0, "retries": 0, "seconds": 246.2, "max_memory_gb": 14.31, "ce": 0.21650095004588366, "rank": 0.0} +{"step": 183, "steps": 500, "lr": 2e-05, "grad_norm": 0.39043059945106506, "loss_scale": 4.0, "retries": 0, "seconds": 247.6, "max_memory_gb": 14.31, "ce": 0.14221475046360865, "rank": 0.0} +{"step": 184, "steps": 500, "lr": 2e-05, "grad_norm": 0.9635753035545349, "loss_scale": 4.0, "retries": 0, "seconds": 249.0, "max_memory_gb": 14.31, "ce": 0.20083741791313514, "rank": 0.11281335353851318} +{"step": 185, "steps": 500, "lr": 2e-05, "grad_norm": 0.5983079075813293, "loss_scale": 4.0, "retries": 0, "seconds": 250.3, "max_memory_gb": 14.31, "ce": 0.2365844319574535, "rank": 0.0} +{"step": 186, "steps": 500, "lr": 2e-05, "grad_norm": 0.7189469933509827, "loss_scale": 4.0, "retries": 0, "seconds": 251.7, "max_memory_gb": 14.31, "ce": 0.17962293315213174, "rank": 0.06402278691530228} +{"step": 187, "steps": 500, "lr": 2e-05, "grad_norm": 1.2870844602584839, "loss_scale": 4.0, "retries": 0, "seconds": 253.1, "max_memory_gb": 14.31, "ce": 0.3450417025014758, "rank": 0.13481858372688293} +{"step": 188, "steps": 500, "lr": 2e-05, "grad_norm": 0.171517014503479, "loss_scale": 4.0, "retries": 0, "seconds": 254.4, "max_memory_gb": 14.31, "ce": 0.16679974235012196, "rank": 0.0} +{"step": 189, "steps": 500, "lr": 2e-05, "grad_norm": 0.4175013303756714, "loss_scale": 4.0, "retries": 0, "seconds": 255.7, "max_memory_gb": 14.31, "ce": 0.1638365664985031, "rank": 0.03446713835000992} +{"step": 190, "steps": 500, "lr": 2e-05, "grad_norm": 1.061958909034729, "loss_scale": 4.0, "retries": 0, "seconds": 257.0, "max_memory_gb": 14.31, "ce": 0.2774536937940866, "rank": 0.0} +{"step": 191, "steps": 500, "lr": 2e-05, "grad_norm": 0.22333164513111115, "loss_scale": 4.0, "retries": 0, "seconds": 258.3, "max_memory_gb": 14.31, "ce": 0.30305184540338814, "rank": 0.0} +{"step": 192, "steps": 500, "lr": 2e-05, "grad_norm": 0.40773627161979675, "loss_scale": 4.0, "retries": 0, "seconds": 259.7, "max_memory_gb": 14.31, "ce": 0.27211158093996346, "rank": 0.0} +{"step": 193, "steps": 500, "lr": 2e-05, "grad_norm": 0.7913013696670532, "loss_scale": 4.0, "retries": 0, "seconds": 261.1, "max_memory_gb": 14.31, "ce": 0.21931611327454448, "rank": 0.1760639324784279} +{"step": 194, "steps": 500, "lr": 2e-05, "grad_norm": 0.49478065967559814, "loss_scale": 4.0, "retries": 0, "seconds": 262.4, "max_memory_gb": 14.31, "ce": 0.2925527517218143, "rank": 0.0} +{"step": 195, "steps": 500, "lr": 2e-05, "grad_norm": 0.6843462586402893, "loss_scale": 4.0, "retries": 0, "seconds": 263.8, "max_memory_gb": 14.31, "ce": 0.2853546920232475, "rank": 0.0} +{"step": 196, "steps": 500, "lr": 2e-05, "grad_norm": 0.872390627861023, "loss_scale": 4.0, "retries": 0, "seconds": 265.2, "max_memory_gb": 14.31, "ce": 0.193030736874789, "rank": 0.05618879199028015} +{"step": 197, "steps": 500, "lr": 2e-05, "grad_norm": 0.5182547569274902, "loss_scale": 4.0, "retries": 0, "seconds": 266.5, "max_memory_gb": 14.31, "ce": 0.23607793380506337, "rank": 0.048557065427303314} +{"step": 198, "steps": 500, "lr": 2e-05, "grad_norm": 0.7014018893241882, "loss_scale": 4.0, "retries": 0, "seconds": 267.9, "max_memory_gb": 14.31, "ce": 0.29451305512338877, "rank": 0.0} +{"step": 199, "steps": 500, "lr": 2e-05, "grad_norm": 1.1936366558074951, "loss_scale": 4.0, "retries": 0, "seconds": 269.2, "max_memory_gb": 14.31, "ce": 0.23883565212599933, "rank": 0.0} +{"step": 200, "steps": 500, "lr": 2e-05, "grad_norm": 0.2989684045314789, "loss_scale": 4.0, "retries": 0, "seconds": 270.5, "max_memory_gb": 14.31, "ce": 0.3052917786408216, "rank": 0.07917051762342453} +{"step": 201, "steps": 500, "lr": 2e-05, "grad_norm": 0.3810092508792877, "loss_scale": 4.0, "retries": 0, "seconds": 271.9, "max_memory_gb": 14.31, "ce": 0.20352897490374744, "rank": 0.0} +{"step": 202, "steps": 500, "lr": 2e-05, "grad_norm": 0.2897825241088867, "loss_scale": 4.0, "retries": 0, "seconds": 273.2, "max_memory_gb": 14.31, "ce": 0.06532279669772834, "rank": 0.04956212639808655} +{"step": 203, "steps": 500, "lr": 2e-05, "grad_norm": 0.6981496810913086, "loss_scale": 4.0, "retries": 0, "seconds": 274.6, "max_memory_gb": 14.31, "ce": 0.25207531987689435, "rank": 0.060769446194171906} +{"step": 204, "steps": 500, "lr": 2e-05, "grad_norm": 1.5827608108520508, "loss_scale": 4.0, "retries": 0, "seconds": 275.9, "max_memory_gb": 14.31, "ce": 0.28748156619258225, "rank": 0.0} +{"step": 205, "steps": 500, "lr": 2e-05, "grad_norm": 1.0252058506011963, "loss_scale": 4.0, "retries": 0, "seconds": 277.3, "max_memory_gb": 14.31, "ce": 0.22832191246561706, "rank": 0.0} +{"step": 206, "steps": 500, "lr": 2e-05, "grad_norm": 0.9514808058738708, "loss_scale": 4.0, "retries": 0, "seconds": 278.6, "max_memory_gb": 14.31, "ce": 0.18586642295122147, "rank": 0.07398857176303864} +{"step": 207, "steps": 500, "lr": 2e-05, "grad_norm": 0.5115421414375305, "loss_scale": 4.0, "retries": 0, "seconds": 280.0, "max_memory_gb": 14.31, "ce": 0.48236715397797525, "rank": 0.20784300565719604} +{"step": 208, "steps": 500, "lr": 2e-05, "grad_norm": 0.762638509273529, "loss_scale": 4.0, "retries": 0, "seconds": 281.3, "max_memory_gb": 14.31, "ce": 0.13485924602719024, "rank": 0.0} +{"step": 209, "steps": 500, "lr": 2e-05, "grad_norm": 0.5830050706863403, "loss_scale": 4.0, "retries": 0, "seconds": 282.7, "max_memory_gb": 14.31, "ce": 0.2609410497825593, "rank": 0.17208455502986908} +{"step": 210, "steps": 500, "lr": 2e-05, "grad_norm": 0.40270641446113586, "loss_scale": 4.0, "retries": 0, "seconds": 284.0, "max_memory_gb": 14.31, "ce": 0.2449359151069075, "rank": 0.05394464731216431} +{"step": 211, "steps": 500, "lr": 2e-05, "grad_norm": 0.992042601108551, "loss_scale": 4.0, "retries": 0, "seconds": 285.4, "max_memory_gb": 14.31, "ce": 0.2764683156274259, "rank": 0.0} +{"step": 212, "steps": 500, "lr": 2e-05, "grad_norm": 1.1550465822219849, "loss_scale": 4.0, "retries": 0, "seconds": 286.7, "max_memory_gb": 14.31, "ce": 0.3466115817427635, "rank": 0.0} +{"step": 213, "steps": 500, "lr": 2e-05, "grad_norm": 0.5445658564567566, "loss_scale": 4.0, "retries": 0, "seconds": 288.1, "max_memory_gb": 14.31, "ce": 0.3002529088407755, "rank": 0.0} +{"step": 214, "steps": 500, "lr": 2e-05, "grad_norm": 1.910706877708435, "loss_scale": 4.0, "retries": 0, "seconds": 289.5, "max_memory_gb": 14.31, "ce": 0.36931806104257703, "rank": 0.044232532382011414} +{"step": 215, "steps": 500, "lr": 2e-05, "grad_norm": 0.3421724736690521, "loss_scale": 4.0, "retries": 0, "seconds": 290.8, "max_memory_gb": 14.31, "ce": 0.3545764544978738, "rank": 0.018055245280265808} +{"step": 216, "steps": 500, "lr": 2e-05, "grad_norm": 1.03714919090271, "loss_scale": 4.0, "retries": 0, "seconds": 292.1, "max_memory_gb": 14.31, "ce": 0.3293033846421167, "rank": 0.0} +{"step": 217, "steps": 500, "lr": 2e-05, "grad_norm": 0.617283821105957, "loss_scale": 4.0, "retries": 0, "seconds": 293.4, "max_memory_gb": 14.31, "ce": 0.12114482920151204, "rank": 0.0} +{"step": 218, "steps": 500, "lr": 2e-05, "grad_norm": 0.789384126663208, "loss_scale": 4.0, "retries": 0, "seconds": 294.7, "max_memory_gb": 14.31, "ce": 0.2556650541955605, "rank": 0.0} +{"step": 219, "steps": 500, "lr": 2e-05, "grad_norm": 0.6186544895172119, "loss_scale": 4.0, "retries": 0, "seconds": 296.0, "max_memory_gb": 14.31, "ce": 0.14564170443918556, "rank": 0.0} +{"step": 220, "steps": 500, "lr": 2e-05, "grad_norm": 1.310683012008667, "loss_scale": 4.0, "retries": 0, "seconds": 297.4, "max_memory_gb": 14.31, "ce": 0.1542649178300053, "rank": 0.0} +{"step": 221, "steps": 500, "lr": 2e-05, "grad_norm": 1.299271583557129, "loss_scale": 4.0, "retries": 0, "seconds": 298.7, "max_memory_gb": 14.31, "ce": 0.19439383770804852, "rank": 0.0} +{"step": 222, "steps": 500, "lr": 2e-05, "grad_norm": 3.621764659881592, "loss_scale": 4.0, "retries": 0, "seconds": 300.1, "max_memory_gb": 14.31, "ce": 0.3897507607471198, "rank": 0.0} +{"step": 223, "steps": 500, "lr": 2e-05, "grad_norm": 0.2685911953449249, "loss_scale": 4.0, "retries": 0, "seconds": 301.4, "max_memory_gb": 14.31, "ce": 0.2100366281811148, "rank": 0.0} +{"step": 224, "steps": 500, "lr": 2e-05, "grad_norm": 0.6932056546211243, "loss_scale": 4.0, "retries": 0, "seconds": 302.8, "max_memory_gb": 14.31, "ce": 0.16774254047777504, "rank": 0.102693110704422} +{"step": 225, "steps": 500, "lr": 2e-05, "grad_norm": 0.6753541231155396, "loss_scale": 4.0, "retries": 0, "seconds": 304.1, "max_memory_gb": 14.31, "ce": 0.13835913536604494, "rank": 0.01221630722284317} +{"step": 226, "steps": 500, "lr": 2e-05, "grad_norm": 0.47494542598724365, "loss_scale": 4.0, "retries": 0, "seconds": 305.5, "max_memory_gb": 14.31, "ce": 0.15577729512006044, "rank": 0.0} +{"step": 227, "steps": 500, "lr": 2e-05, "grad_norm": 0.5753265619277954, "loss_scale": 4.0, "retries": 0, "seconds": 306.8, "max_memory_gb": 14.31, "ce": 0.04626765369903296, "rank": 0.0} +{"step": 228, "steps": 500, "lr": 2e-05, "grad_norm": 0.4069153368473053, "loss_scale": 4.0, "retries": 0, "seconds": 308.1, "max_memory_gb": 14.31, "ce": 0.14172808069270104, "rank": 0.0} +{"step": 229, "steps": 500, "lr": 2e-05, "grad_norm": 0.3487948179244995, "loss_scale": 4.0, "retries": 0, "seconds": 309.5, "max_memory_gb": 14.31, "ce": 0.22107816208153963, "rank": 0.0} +{"step": 230, "steps": 500, "lr": 2e-05, "grad_norm": 0.518728494644165, "loss_scale": 4.0, "retries": 0, "seconds": 310.8, "max_memory_gb": 14.31, "ce": 0.25673877325607464, "rank": 0.0} +{"step": 231, "steps": 500, "lr": 2e-05, "grad_norm": 0.4119829535484314, "loss_scale": 4.0, "retries": 0, "seconds": 312.1, "max_memory_gb": 14.31, "ce": 0.20971059007570148, "rank": 0.0} +{"step": 232, "steps": 500, "lr": 2e-05, "grad_norm": 0.7463088631629944, "loss_scale": 4.0, "retries": 0, "seconds": 313.5, "max_memory_gb": 14.31, "ce": 0.09438014694023877, "rank": 0.0} +{"step": 233, "steps": 500, "lr": 2e-05, "grad_norm": 2.2267048358917236, "loss_scale": 4.0, "retries": 0, "seconds": 314.8, "max_memory_gb": 14.31, "ce": 0.24755096348235384, "rank": 0.0} +{"step": 234, "steps": 500, "lr": 2e-05, "grad_norm": 0.3355656862258911, "loss_scale": 4.0, "retries": 0, "seconds": 316.1, "max_memory_gb": 14.31, "ce": 0.21902234031585976, "rank": 0.0} +{"step": 235, "steps": 500, "lr": 2e-05, "grad_norm": 0.3744480311870575, "loss_scale": 4.0, "retries": 0, "seconds": 317.5, "max_memory_gb": 14.31, "ce": 0.2700875822920352, "rank": 0.0} +{"step": 236, "steps": 500, "lr": 2e-05, "grad_norm": 0.3956650495529175, "loss_scale": 4.0, "retries": 0, "seconds": 318.8, "max_memory_gb": 14.31, "ce": 0.10633395472541451, "rank": 0.0} +{"step": 237, "steps": 500, "lr": 2e-05, "grad_norm": 2.880812406539917, "loss_scale": 4.0, "retries": 0, "seconds": 320.2, "max_memory_gb": 14.31, "ce": 0.27570988214574754, "rank": 0.05610215663909912} +{"step": 238, "steps": 500, "lr": 2e-05, "grad_norm": 0.6946873664855957, "loss_scale": 4.0, "retries": 0, "seconds": 321.5, "max_memory_gb": 14.31, "ce": 0.07558144052745774, "rank": 0.0} +{"step": 239, "steps": 500, "lr": 2e-05, "grad_norm": 0.6777441501617432, "loss_scale": 4.0, "retries": 0, "seconds": 322.8, "max_memory_gb": 14.31, "ce": 0.23216825307463296, "rank": 0.0} +{"step": 240, "steps": 500, "lr": 2e-05, "grad_norm": 0.8823796510696411, "loss_scale": 4.0, "retries": 0, "seconds": 324.1, "max_memory_gb": 14.31, "ce": 0.13985700492048636, "rank": 0.0} +{"step": 241, "steps": 500, "lr": 2e-05, "grad_norm": 0.27130815386772156, "loss_scale": 4.0, "retries": 0, "seconds": 325.5, "max_memory_gb": 14.31, "ce": 0.21617391402833164, "rank": 0.0} +{"step": 242, "steps": 500, "lr": 2e-05, "grad_norm": 1.2030686140060425, "loss_scale": 4.0, "retries": 0, "seconds": 326.9, "max_memory_gb": 14.31, "ce": 0.18411773478146642, "rank": 0.0} +{"step": 243, "steps": 500, "lr": 2e-05, "grad_norm": 0.8142539262771606, "loss_scale": 4.0, "retries": 0, "seconds": 328.2, "max_memory_gb": 14.31, "ce": 0.07726289481797721, "rank": 0.0} +{"step": 244, "steps": 500, "lr": 2e-05, "grad_norm": 0.8595879673957825, "loss_scale": 4.0, "retries": 0, "seconds": 329.6, "max_memory_gb": 14.31, "ce": 0.2581324346610927, "rank": 0.0} +{"step": 245, "steps": 500, "lr": 2e-05, "grad_norm": 0.4419519007205963, "loss_scale": 4.0, "retries": 0, "seconds": 330.9, "max_memory_gb": 14.31, "ce": 0.33829302377125714, "rank": 0.0} +{"step": 246, "steps": 500, "lr": 2e-05, "grad_norm": 0.5578590035438538, "loss_scale": 4.0, "retries": 0, "seconds": 332.3, "max_memory_gb": 14.31, "ce": 0.05792351292620879, "rank": 0.0} +{"step": 247, "steps": 500, "lr": 2e-05, "grad_norm": 0.9839972257614136, "loss_scale": 4.0, "retries": 0, "seconds": 333.6, "max_memory_gb": 14.31, "ce": 0.15161054019699804, "rank": 0.0} +{"step": 248, "steps": 500, "lr": 2e-05, "grad_norm": 0.7620556950569153, "loss_scale": 4.0, "retries": 0, "seconds": 335.0, "max_memory_gb": 14.31, "ce": 0.10883134655887261, "rank": 0.0} +{"step": 249, "steps": 500, "lr": 2e-05, "grad_norm": 1.6216646432876587, "loss_scale": 4.0, "retries": 0, "seconds": 336.3, "max_memory_gb": 14.31, "ce": 0.14728004774951842, "rank": 0.0} +{"step": 250, "steps": 500, "lr": 2e-05, "grad_norm": 1.0079646110534668, "loss_scale": 4.0, "retries": 0, "seconds": 337.6, "max_memory_gb": 14.31, "ce": 0.09834381905966438, "rank": 0.0} +{"step": 251, "steps": 500, "lr": 2e-05, "grad_norm": 0.2812977135181427, "loss_scale": 4.0, "retries": 0, "seconds": 339.0, "max_memory_gb": 14.31, "ce": 0.15149819252837915, "rank": 0.0} +{"step": 252, "steps": 500, "lr": 2e-05, "grad_norm": 0.6526393890380859, "loss_scale": 4.0, "retries": 0, "seconds": 340.4, "max_memory_gb": 14.31, "ce": 0.06166560419660527, "rank": 0.0} +{"step": 253, "steps": 500, "lr": 2e-05, "grad_norm": 0.34456461668014526, "loss_scale": 4.0, "retries": 0, "seconds": 341.7, "max_memory_gb": 14.31, "ce": 0.10693705585435964, "rank": 0.0} +{"step": 254, "steps": 500, "lr": 2e-05, "grad_norm": 2.330268144607544, "loss_scale": 4.0, "retries": 0, "seconds": 343.0, "max_memory_gb": 14.31, "ce": 0.25496548067894764, "rank": 0.0} +{"step": 255, "steps": 500, "lr": 2e-05, "grad_norm": 0.40634098649024963, "loss_scale": 4.0, "retries": 0, "seconds": 344.4, "max_memory_gb": 14.31, "ce": 0.22262032655999064, "rank": 0.0} +{"step": 256, "steps": 500, "lr": 2e-05, "grad_norm": 0.42011046409606934, "loss_scale": 4.0, "retries": 0, "seconds": 345.7, "max_memory_gb": 14.31, "ce": 0.10269516112748533, "rank": 0.0} +{"step": 257, "steps": 500, "lr": 2e-05, "grad_norm": 0.23835909366607666, "loss_scale": 4.0, "retries": 0, "seconds": 347.0, "max_memory_gb": 14.31, "ce": 0.0717176393372938, "rank": 0.0} +{"step": 258, "steps": 500, "lr": 2e-05, "grad_norm": 1.932548999786377, "loss_scale": 4.0, "retries": 0, "seconds": 348.4, "max_memory_gb": 14.31, "ce": 0.11343757080612704, "rank": 0.0} +{"step": 259, "steps": 500, "lr": 2e-05, "grad_norm": 0.15655474364757538, "loss_scale": 4.0, "retries": 0, "seconds": 349.8, "max_memory_gb": 14.31, "ce": 0.05746387201361358, "rank": 0.0} +{"step": 260, "steps": 500, "lr": 2e-05, "grad_norm": 0.4674360752105713, "loss_scale": 4.0, "retries": 0, "seconds": 351.1, "max_memory_gb": 14.31, "ce": 0.14618461098871194, "rank": 0.0} +{"step": 261, "steps": 500, "lr": 2e-05, "grad_norm": 1.4904335737228394, "loss_scale": 4.0, "retries": 0, "seconds": 352.5, "max_memory_gb": 14.31, "ce": 0.20786820838111453, "rank": 0.0} +{"step": 262, "steps": 500, "lr": 2e-05, "grad_norm": 0.5455617904663086, "loss_scale": 4.0, "retries": 0, "seconds": 353.8, "max_memory_gb": 14.31, "ce": 0.21511118925991468, "rank": 0.0} +{"step": 263, "steps": 500, "lr": 2e-05, "grad_norm": 1.024728536605835, "loss_scale": 4.0, "retries": 0, "seconds": 355.2, "max_memory_gb": 14.31, "ce": 0.26231146400095895, "rank": 0.12283022701740265} +{"step": 264, "steps": 500, "lr": 2e-05, "grad_norm": 0.4685496985912323, "loss_scale": 4.0, "retries": 0, "seconds": 356.5, "max_memory_gb": 14.31, "ce": 0.18030010159418453, "rank": 0.0} +{"step": 265, "steps": 500, "lr": 2e-05, "grad_norm": 0.39763522148132324, "loss_scale": 4.0, "retries": 0, "seconds": 357.9, "max_memory_gb": 14.31, "ce": 0.1178972378547769, "rank": 0.0} +{"step": 266, "steps": 500, "lr": 2e-05, "grad_norm": 2.7673327922821045, "loss_scale": 4.0, "retries": 0, "seconds": 359.2, "max_memory_gb": 14.31, "ce": 0.25665191088046413, "rank": 0.0} +{"step": 267, "steps": 500, "lr": 2e-05, "grad_norm": 0.2805120646953583, "loss_scale": 4.0, "retries": 0, "seconds": 360.6, "max_memory_gb": 14.31, "ce": 0.14085254834208172, "rank": 0.0} +{"step": 268, "steps": 500, "lr": 2e-05, "grad_norm": 0.3198009729385376, "loss_scale": 4.0, "retries": 0, "seconds": 361.9, "max_memory_gb": 14.31, "ce": 0.2537109318654984, "rank": 0.0} +{"step": 269, "steps": 500, "lr": 2e-05, "grad_norm": 2.6546173095703125, "loss_scale": 4.0, "retries": 0, "seconds": 363.2, "max_memory_gb": 14.31, "ce": 0.3785477529745549, "rank": 0.08274926245212555} +{"step": 270, "steps": 500, "lr": 2e-05, "grad_norm": 0.5632020831108093, "loss_scale": 4.0, "retries": 0, "seconds": 364.6, "max_memory_gb": 14.31, "ce": 0.12512108116061427, "rank": 0.0} +{"step": 271, "steps": 500, "lr": 2e-05, "grad_norm": 0.9917564392089844, "loss_scale": 4.0, "retries": 0, "seconds": 365.9, "max_memory_gb": 14.31, "ce": 0.1061270473874174, "rank": 0.1129993200302124} +{"step": 272, "steps": 500, "lr": 2e-05, "grad_norm": 0.9499532580375671, "loss_scale": 4.0, "retries": 0, "seconds": 367.1, "max_memory_gb": 14.31, "ce": 0.11103368969634175, "rank": 0.0} +{"step": 273, "steps": 500, "lr": 2e-05, "grad_norm": 0.9266161322593689, "loss_scale": 4.0, "retries": 0, "seconds": 368.3, "max_memory_gb": 14.31, "ce": 0.08758464688435197, "rank": 0.12543843686580658} +{"step": 274, "steps": 500, "lr": 2e-05, "grad_norm": 1.780539870262146, "loss_scale": 4.0, "retries": 0, "seconds": 369.6, "max_memory_gb": 14.31, "ce": 0.18415509673650376, "rank": 0.0007685571908950806} +{"step": 275, "steps": 500, "lr": 2e-05, "grad_norm": 0.794818103313446, "loss_scale": 4.0, "retries": 0, "seconds": 371.0, "max_memory_gb": 14.31, "ce": 0.5483726098609623, "rank": 0.06870460510253906} +{"step": 276, "steps": 500, "lr": 2e-05, "grad_norm": 0.4773421883583069, "loss_scale": 4.0, "retries": 0, "seconds": 372.3, "max_memory_gb": 14.31, "ce": 0.0414157590130344, "rank": 0.0} +{"step": 277, "steps": 500, "lr": 2e-05, "grad_norm": 0.41725537180900574, "loss_scale": 4.0, "retries": 0, "seconds": 373.7, "max_memory_gb": 14.31, "ce": 0.036377748983795755, "rank": 0.0} +{"step": 278, "steps": 500, "lr": 2e-05, "grad_norm": 0.6008747220039368, "loss_scale": 4.0, "retries": 0, "seconds": 375.1, "max_memory_gb": 14.31, "ce": 0.1722097728488734, "rank": 0.05048400163650513} +{"step": 279, "steps": 500, "lr": 2e-05, "grad_norm": 1.6356250047683716, "loss_scale": 4.0, "retries": 0, "seconds": 376.4, "max_memory_gb": 14.31, "ce": 0.1035466457833536, "rank": 0.0} +{"step": 280, "steps": 500, "lr": 2e-05, "grad_norm": 0.8908403515815735, "loss_scale": 4.0, "retries": 0, "seconds": 377.7, "max_memory_gb": 14.31, "ce": 0.053224825773213524, "rank": 0.0} +{"step": 281, "steps": 500, "lr": 2e-05, "grad_norm": 2.922806739807129, "loss_scale": 4.0, "retries": 0, "seconds": 379.0, "max_memory_gb": 14.31, "ce": 0.23856354258896317, "rank": 0.0} +{"step": 282, "steps": 500, "lr": 2e-05, "grad_norm": 0.24877384305000305, "loss_scale": 4.0, "retries": 0, "seconds": 380.4, "max_memory_gb": 14.31, "ce": 0.11695107194100274, "rank": 0.0} +{"step": 283, "steps": 500, "lr": 2e-05, "grad_norm": 2.291901111602783, "loss_scale": 4.0, "retries": 0, "seconds": 381.7, "max_memory_gb": 14.31, "ce": 0.16903246652509551, "rank": 0.0} +{"step": 284, "steps": 500, "lr": 2e-05, "grad_norm": 1.9386802911758423, "loss_scale": 4.0, "retries": 0, "seconds": 383.1, "max_memory_gb": 14.31, "ce": 0.13103016243258025, "rank": 0.0} +{"step": 285, "steps": 500, "lr": 2e-05, "grad_norm": 0.9153313636779785, "loss_scale": 4.0, "retries": 0, "seconds": 384.3, "max_memory_gb": 14.31, "ce": 0.12411079336015973, "rank": 0.0} +{"step": 286, "steps": 500, "lr": 2e-05, "grad_norm": 0.48776933550834656, "loss_scale": 4.0, "retries": 0, "seconds": 385.5, "max_memory_gb": 14.31, "ce": 0.06875783930445323, "rank": 0.0} +{"step": 287, "steps": 500, "lr": 2e-05, "grad_norm": 0.49366873502731323, "loss_scale": 4.0, "retries": 0, "seconds": 386.8, "max_memory_gb": 14.31, "ce": 0.16836800386954565, "rank": 0.0} +{"step": 288, "steps": 500, "lr": 2e-05, "grad_norm": 0.23645371198654175, "loss_scale": 4.0, "retries": 0, "seconds": 388.2, "max_memory_gb": 14.31, "ce": 0.017638751785852946, "rank": 0.0} +{"step": 289, "steps": 500, "lr": 2e-05, "grad_norm": 0.5620402097702026, "loss_scale": 4.0, "retries": 0, "seconds": 389.5, "max_memory_gb": 14.31, "ce": 0.13043476054735947, "rank": 0.0} +{"step": 290, "steps": 500, "lr": 2e-05, "grad_norm": 0.5491151213645935, "loss_scale": 4.0, "retries": 0, "seconds": 390.9, "max_memory_gb": 14.31, "ce": 0.16177222461556084, "rank": 0.0} +{"step": 291, "steps": 500, "lr": 2e-05, "grad_norm": 0.2616140842437744, "loss_scale": 4.0, "retries": 0, "seconds": 392.2, "max_memory_gb": 14.31, "ce": 0.02965520314319292, "rank": 0.0} +{"step": 292, "steps": 500, "lr": 2e-05, "grad_norm": 1.1874812841415405, "loss_scale": 4.0, "retries": 0, "seconds": 393.5, "max_memory_gb": 14.31, "ce": 0.16594857021118514, "rank": 0.0} +{"step": 293, "steps": 500, "lr": 2e-05, "grad_norm": 0.5274538397789001, "loss_scale": 4.0, "retries": 0, "seconds": 394.9, "max_memory_gb": 14.31, "ce": 0.08849290963553358, "rank": 0.0} +{"step": 294, "steps": 500, "lr": 2e-05, "grad_norm": 0.9624010324478149, "loss_scale": 4.0, "retries": 0, "seconds": 396.2, "max_memory_gb": 14.31, "ce": 0.0767305403132923, "rank": 0.0} +{"step": 295, "steps": 500, "lr": 2e-05, "grad_norm": 0.6571291089057922, "loss_scale": 4.0, "retries": 0, "seconds": 397.5, "max_memory_gb": 14.31, "ce": 0.24581431398837594, "rank": 0.0} +{"step": 296, "steps": 500, "lr": 2e-05, "grad_norm": 1.1313680410385132, "loss_scale": 4.0, "retries": 0, "seconds": 398.8, "max_memory_gb": 14.31, "ce": 0.24490806521498598, "rank": 0.0} +{"step": 297, "steps": 500, "lr": 2e-05, "grad_norm": 0.2538204491138458, "loss_scale": 4.0, "retries": 0, "seconds": 400.2, "max_memory_gb": 14.31, "ce": 0.11744543723762035, "rank": 0.0} +{"step": 298, "steps": 500, "lr": 2e-05, "grad_norm": 3.590569257736206, "loss_scale": 4.0, "retries": 0, "seconds": 401.5, "max_memory_gb": 14.31, "ce": 0.44062258434132673, "rank": 0.0} +{"step": 299, "steps": 500, "lr": 2e-05, "grad_norm": 0.6147991418838501, "loss_scale": 4.0, "retries": 0, "seconds": 402.9, "max_memory_gb": 14.31, "ce": 0.033945454662898555, "rank": 0.0} +{"step": 300, "steps": 500, "lr": 2e-05, "grad_norm": 0.4180974066257477, "loss_scale": 4.0, "retries": 0, "seconds": 404.2, "max_memory_gb": 14.31, "ce": 0.23898733855457976, "rank": 0.0} +{"step": 301, "steps": 500, "lr": 2e-05, "grad_norm": 0.3541077971458435, "loss_scale": 4.0, "retries": 0, "seconds": 405.5, "max_memory_gb": 14.31, "ce": 0.11693835463665891, "rank": 0.0} +{"step": 302, "steps": 500, "lr": 2e-05, "grad_norm": 0.8950376510620117, "loss_scale": 4.0, "retries": 0, "seconds": 406.9, "max_memory_gb": 14.31, "ce": 0.04257085095741786, "rank": 0.0} +{"step": 303, "steps": 500, "lr": 2e-05, "grad_norm": 0.6451747417449951, "loss_scale": 4.0, "retries": 0, "seconds": 408.2, "max_memory_gb": 14.31, "ce": 0.04090563110366929, "rank": 0.0} +{"step": 304, "steps": 500, "lr": 2e-05, "grad_norm": 0.616689920425415, "loss_scale": 4.0, "retries": 0, "seconds": 409.6, "max_memory_gb": 14.31, "ce": 0.05861289137828862, "rank": 0.047298409044742584} +{"step": 305, "steps": 500, "lr": 2e-05, "grad_norm": 1.259875774383545, "loss_scale": 4.0, "retries": 0, "seconds": 410.9, "max_memory_gb": 14.31, "ce": 0.11449157610331895, "rank": 0.0} +{"step": 306, "steps": 500, "lr": 2e-05, "grad_norm": 1.1157441139221191, "loss_scale": 4.0, "retries": 0, "seconds": 412.2, "max_memory_gb": 14.31, "ce": 0.05059912133583566, "rank": 0.0} +{"step": 307, "steps": 500, "lr": 2e-05, "grad_norm": 0.4124423861503601, "loss_scale": 4.0, "retries": 0, "seconds": 413.6, "max_memory_gb": 14.31, "ce": 0.09633893156205886, "rank": 0.0} +{"step": 308, "steps": 500, "lr": 2e-05, "grad_norm": 1.039819359779358, "loss_scale": 4.0, "retries": 0, "seconds": 415.0, "max_memory_gb": 14.31, "ce": 0.18117104283737717, "rank": 0.0} +{"step": 309, "steps": 500, "lr": 2e-05, "grad_norm": 0.45384764671325684, "loss_scale": 4.0, "retries": 0, "seconds": 416.3, "max_memory_gb": 14.31, "ce": 0.14850477701838827, "rank": 0.0} +{"step": 310, "steps": 500, "lr": 2e-05, "grad_norm": 1.0591998100280762, "loss_scale": 4.0, "retries": 0, "seconds": 417.6, "max_memory_gb": 14.31, "ce": 0.05337646698899334, "rank": 0.0} +{"step": 311, "steps": 500, "lr": 2e-05, "grad_norm": 0.2773599624633789, "loss_scale": 4.0, "retries": 0, "seconds": 418.9, "max_memory_gb": 14.31, "ce": 0.10165158288145903, "rank": 0.0} +{"step": 312, "steps": 500, "lr": 2e-05, "grad_norm": 0.2647876441478729, "loss_scale": 4.0, "retries": 0, "seconds": 420.3, "max_memory_gb": 14.31, "ce": 0.0198251802357845, "rank": 0.0} +{"step": 313, "steps": 500, "lr": 2e-05, "grad_norm": 0.9009084701538086, "loss_scale": 4.0, "retries": 0, "seconds": 421.6, "max_memory_gb": 14.31, "ce": 0.15886789999785833, "rank": 0.0} +{"step": 314, "steps": 500, "lr": 2e-05, "grad_norm": 0.6590065956115723, "loss_scale": 4.0, "retries": 0, "seconds": 422.9, "max_memory_gb": 14.31, "ce": 0.14171573292696849, "rank": 0.0} +{"step": 315, "steps": 500, "lr": 2e-05, "grad_norm": 0.922882616519928, "loss_scale": 4.0, "retries": 0, "seconds": 424.3, "max_memory_gb": 14.31, "ce": 0.08907469570840476, "rank": 0.12249543517827988} +{"step": 316, "steps": 500, "lr": 2e-05, "grad_norm": 2.1459267139434814, "loss_scale": 4.0, "retries": 0, "seconds": 425.7, "max_memory_gb": 14.31, "ce": 0.2848965385637712, "rank": 0.0} +{"step": 317, "steps": 500, "lr": 2e-05, "grad_norm": 0.9708598852157593, "loss_scale": 4.0, "retries": 0, "seconds": 427.0, "max_memory_gb": 14.31, "ce": 0.1651386067533167, "rank": 0.0} +{"step": 318, "steps": 500, "lr": 2e-05, "grad_norm": 4.917360782623291, "loss_scale": 4.0, "retries": 0, "seconds": 428.3, "max_memory_gb": 14.31, "ce": 0.25220418395474553, "rank": 0.0} +{"step": 319, "steps": 500, "lr": 2e-05, "grad_norm": 0.7626944184303284, "loss_scale": 4.0, "retries": 0, "seconds": 429.6, "max_memory_gb": 14.31, "ce": 0.10092794709998998, "rank": 0.0} +{"step": 320, "steps": 500, "lr": 2e-05, "grad_norm": 1.2936540842056274, "loss_scale": 4.0, "retries": 0, "seconds": 430.9, "max_memory_gb": 14.31, "ce": 0.11178657178970752, "rank": 0.0} +{"step": 321, "steps": 500, "lr": 2e-05, "grad_norm": 0.4495420753955841, "loss_scale": 4.0, "retries": 0, "seconds": 432.3, "max_memory_gb": 14.31, "ce": 0.16246189485536888, "rank": 0.12674082815647125} +{"step": 322, "steps": 500, "lr": 2e-05, "grad_norm": 0.5654058456420898, "loss_scale": 4.0, "retries": 0, "seconds": 433.6, "max_memory_gb": 14.31, "ce": 0.029249677569168853, "rank": 0.0} +{"step": 323, "steps": 500, "lr": 2e-05, "grad_norm": 0.5303423404693604, "loss_scale": 4.0, "retries": 0, "seconds": 435.0, "max_memory_gb": 14.31, "ce": 0.0663137189403642, "rank": 0.0} +{"step": 324, "steps": 500, "lr": 2e-05, "grad_norm": 0.677825391292572, "loss_scale": 4.0, "retries": 0, "seconds": 436.3, "max_memory_gb": 14.31, "ce": 0.04050792745692888, "rank": 0.0} +{"step": 325, "steps": 500, "lr": 2e-05, "grad_norm": 0.4814232587814331, "loss_scale": 4.0, "retries": 0, "seconds": 437.5, "max_memory_gb": 14.31, "ce": 0.2071490597736556, "rank": 0.0} +{"step": 326, "steps": 500, "lr": 2e-05, "grad_norm": 0.3661375045776367, "loss_scale": 4.0, "retries": 0, "seconds": 438.6, "max_memory_gb": 14.31, "ce": 0.05076407126762206, "rank": 0.08993908017873764} +{"step": 327, "steps": 500, "lr": 2e-05, "grad_norm": 3.8468496799468994, "loss_scale": 4.0, "retries": 0, "seconds": 440.0, "max_memory_gb": 14.31, "ce": 0.16456236752856057, "rank": 0.0} +{"step": 328, "steps": 500, "lr": 2e-05, "grad_norm": 0.4198140501976013, "loss_scale": 4.0, "retries": 0, "seconds": 441.2, "max_memory_gb": 14.31, "ce": 0.14067414629971609, "rank": 0.0} +{"step": 329, "steps": 500, "lr": 2e-05, "grad_norm": 0.39303481578826904, "loss_scale": 4.0, "retries": 0, "seconds": 442.3, "max_memory_gb": 14.31, "ce": 0.1189041536708828, "rank": 0.0} +{"step": 330, "steps": 500, "lr": 2e-05, "grad_norm": 0.9309187531471252, "loss_scale": 4.0, "retries": 0, "seconds": 443.4, "max_memory_gb": 14.31, "ce": 0.15380285841820296, "rank": 0.0} +{"step": 331, "steps": 500, "lr": 2e-05, "grad_norm": 0.7309045195579529, "loss_scale": 4.0, "retries": 0, "seconds": 444.5, "max_memory_gb": 14.31, "ce": 0.2577498850005213, "rank": 0.0} +{"step": 332, "steps": 500, "lr": 2e-05, "grad_norm": 2.5459327697753906, "loss_scale": 4.0, "retries": 0, "seconds": 445.7, "max_memory_gb": 14.31, "ce": 0.1588758410362061, "rank": 0.0} +{"step": 333, "steps": 500, "lr": 2e-05, "grad_norm": 0.49149060249328613, "loss_scale": 4.0, "retries": 0, "seconds": 446.9, "max_memory_gb": 14.31, "ce": 0.24128308062790893, "rank": 0.0} +{"step": 334, "steps": 500, "lr": 2e-05, "grad_norm": 0.5186738967895508, "loss_scale": 4.0, "retries": 0, "seconds": 448.2, "max_memory_gb": 14.31, "ce": 0.13217034556510043, "rank": 0.0} +{"step": 335, "steps": 500, "lr": 2e-05, "grad_norm": 2.64040207862854, "loss_scale": 4.0, "retries": 0, "seconds": 449.5, "max_memory_gb": 14.31, "ce": 0.23973772382669267, "rank": 0.0} +{"step": 336, "steps": 500, "lr": 2e-05, "grad_norm": 1.4032193422317505, "loss_scale": 4.0, "retries": 0, "seconds": 450.8, "max_memory_gb": 14.31, "ce": 0.058458984618482646, "rank": 0.0} +{"step": 337, "steps": 500, "lr": 2e-05, "grad_norm": 3.21622371673584, "loss_scale": 4.0, "retries": 0, "seconds": 452.2, "max_memory_gb": 14.31, "ce": 0.2326060956584115, "rank": 0.0} +{"step": 338, "steps": 500, "lr": 2e-05, "grad_norm": 0.8348211646080017, "loss_scale": 4.0, "retries": 0, "seconds": 453.6, "max_memory_gb": 14.31, "ce": 0.18558105992269702, "rank": 0.0} +{"step": 339, "steps": 500, "lr": 2e-05, "grad_norm": 0.4245525002479553, "loss_scale": 4.0, "retries": 0, "seconds": 455.0, "max_memory_gb": 14.31, "ce": 0.11828829550358932, "rank": 0.0} +{"step": 340, "steps": 500, "lr": 2e-05, "grad_norm": 0.5203116536140442, "loss_scale": 4.0, "retries": 0, "seconds": 456.3, "max_memory_gb": 14.31, "ce": 0.16561535837536212, "rank": 0.0} +{"step": 341, "steps": 500, "lr": 2e-05, "grad_norm": 2.5680644512176514, "loss_scale": 4.0, "retries": 0, "seconds": 457.7, "max_memory_gb": 14.31, "ce": 0.22128377325861948, "rank": 0.0} +{"step": 342, "steps": 500, "lr": 2e-05, "grad_norm": 1.5325204133987427, "loss_scale": 4.0, "retries": 0, "seconds": 459.0, "max_memory_gb": 14.31, "ce": 0.4345596979765105, "rank": 0.0} +{"step": 343, "steps": 500, "lr": 2e-05, "grad_norm": 0.5057061314582825, "loss_scale": 4.0, "retries": 0, "seconds": 460.4, "max_memory_gb": 14.31, "ce": 0.06430654750874965, "rank": 0.0} +{"step": 344, "steps": 500, "lr": 2e-05, "grad_norm": 0.3915494382381439, "loss_scale": 4.0, "retries": 0, "seconds": 461.7, "max_memory_gb": 14.31, "ce": 0.025191967564751394, "rank": 0.0} +{"step": 345, "steps": 500, "lr": 2e-05, "grad_norm": 0.2410583645105362, "loss_scale": 4.0, "retries": 0, "seconds": 463.0, "max_memory_gb": 14.31, "ce": 0.1668302545331244, "rank": 0.0} +{"step": 346, "steps": 500, "lr": 2e-05, "grad_norm": 0.6446455121040344, "loss_scale": 4.0, "retries": 0, "seconds": 464.4, "max_memory_gb": 14.31, "ce": 0.23171162718790583, "rank": 0.0} +{"step": 347, "steps": 500, "lr": 2e-05, "grad_norm": 1.2902030944824219, "loss_scale": 4.0, "retries": 0, "seconds": 465.7, "max_memory_gb": 14.31, "ce": 0.15828796328423778, "rank": 0.0} +{"step": 348, "steps": 500, "lr": 2e-05, "grad_norm": 0.491731196641922, "loss_scale": 4.0, "retries": 0, "seconds": 467.1, "max_memory_gb": 14.31, "ce": 0.016079788454590016, "rank": 0.0} +{"step": 349, "steps": 500, "lr": 2e-05, "grad_norm": 0.38214799761772156, "loss_scale": 4.0, "retries": 0, "seconds": 468.4, "max_memory_gb": 14.31, "ce": 0.08745190135960001, "rank": 0.0} +{"step": 350, "steps": 500, "lr": 2e-05, "grad_norm": 0.3254747986793518, "loss_scale": 4.0, "retries": 0, "seconds": 469.8, "max_memory_gb": 14.31, "ce": 0.11636632935551461, "rank": 0.0} +{"step": 351, "steps": 500, "lr": 2e-05, "grad_norm": 0.7953536510467529, "loss_scale": 4.0, "retries": 0, "seconds": 471.2, "max_memory_gb": 14.31, "ce": 0.06094081447372446, "rank": 0.0} +{"step": 352, "steps": 500, "lr": 2e-05, "grad_norm": 1.9465550184249878, "loss_scale": 4.0, "retries": 0, "seconds": 472.5, "max_memory_gb": 14.31, "ce": 0.35308726287985337, "rank": 0.0} +{"step": 353, "steps": 500, "lr": 2e-05, "grad_norm": 0.4392346441745758, "loss_scale": 4.0, "retries": 0, "seconds": 473.8, "max_memory_gb": 14.31, "ce": 0.12755055567504314, "rank": 0.0} +{"step": 354, "steps": 500, "lr": 2e-05, "grad_norm": 2.2667510509490967, "loss_scale": 4.0, "retries": 0, "seconds": 475.1, "max_memory_gb": 14.31, "ce": 0.22394820491899736, "rank": 0.0} +{"step": 355, "steps": 500, "lr": 2e-05, "grad_norm": 0.7372404932975769, "loss_scale": 4.0, "retries": 0, "seconds": 476.5, "max_memory_gb": 14.31, "ce": 0.101548279153576, "rank": 0.0} +{"step": 356, "steps": 500, "lr": 2e-05, "grad_norm": 0.33905965089797974, "loss_scale": 4.0, "retries": 0, "seconds": 477.9, "max_memory_gb": 14.31, "ce": 0.055804815301598865, "rank": 0.0} +{"step": 357, "steps": 500, "lr": 2e-05, "grad_norm": 0.33883821964263916, "loss_scale": 4.0, "retries": 0, "seconds": 479.2, "max_memory_gb": 14.31, "ce": 0.06974780657401425, "rank": 0.0} +{"step": 358, "steps": 500, "lr": 2e-05, "grad_norm": 0.5117934942245483, "loss_scale": 4.0, "retries": 0, "seconds": 480.4, "max_memory_gb": 14.31, "ce": 0.2038461602987809, "rank": 0.0} +{"step": 359, "steps": 500, "lr": 2e-05, "grad_norm": 0.4888851046562195, "loss_scale": 4.0, "retries": 0, "seconds": 481.8, "max_memory_gb": 14.31, "ce": 0.06685984412615653, "rank": 0.0} +{"step": 360, "steps": 500, "lr": 2e-05, "grad_norm": 0.348107248544693, "loss_scale": 4.0, "retries": 0, "seconds": 483.1, "max_memory_gb": 14.31, "ce": 0.06063837434339803, "rank": 0.13374409079551697} +{"step": 361, "steps": 500, "lr": 2e-05, "grad_norm": 0.9978609681129456, "loss_scale": 4.0, "retries": 0, "seconds": 484.4, "max_memory_gb": 14.31, "ce": 0.13939191092504188, "rank": 0.0} +{"step": 362, "steps": 500, "lr": 2e-05, "grad_norm": 5.310637474060059, "loss_scale": 4.0, "retries": 0, "seconds": 485.7, "max_memory_gb": 14.31, "ce": 0.3220842751670716, "rank": 0.0} +{"step": 363, "steps": 500, "lr": 2e-05, "grad_norm": 0.43254098296165466, "loss_scale": 4.0, "retries": 0, "seconds": 487.0, "max_memory_gb": 14.31, "ce": 0.0721258300691261, "rank": 0.0} +{"step": 364, "steps": 500, "lr": 2e-05, "grad_norm": 1.12973153591156, "loss_scale": 4.0, "retries": 0, "seconds": 488.3, "max_memory_gb": 14.31, "ce": 0.06507615227928909, "rank": 0.10581519454717636} +{"step": 365, "steps": 500, "lr": 2e-05, "grad_norm": 0.48625969886779785, "loss_scale": 4.0, "retries": 0, "seconds": 489.7, "max_memory_gb": 14.31, "ce": 0.11896763803088106, "rank": 0.0} +{"step": 366, "steps": 500, "lr": 2e-05, "grad_norm": 1.2505637407302856, "loss_scale": 4.0, "retries": 0, "seconds": 491.1, "max_memory_gb": 14.31, "ce": 0.11422889485220367, "rank": 0.0} +{"step": 367, "steps": 500, "lr": 2e-05, "grad_norm": 0.481975793838501, "loss_scale": 4.0, "retries": 0, "seconds": 492.4, "max_memory_gb": 14.31, "ce": 0.08042607902825694, "rank": 0.0} +{"step": 368, "steps": 500, "lr": 2e-05, "grad_norm": 0.9643419981002808, "loss_scale": 4.0, "retries": 0, "seconds": 493.7, "max_memory_gb": 14.31, "ce": 0.02749173291522311, "rank": 0.0} +{"step": 369, "steps": 500, "lr": 2e-05, "grad_norm": 0.335451602935791, "loss_scale": 4.0, "retries": 0, "seconds": 495.0, "max_memory_gb": 14.31, "ce": 0.046601061592809856, "rank": 0.11287662386894226} +{"step": 370, "steps": 500, "lr": 2e-05, "grad_norm": 1.1438850164413452, "loss_scale": 4.0, "retries": 0, "seconds": 496.4, "max_memory_gb": 14.31, "ce": 0.4181423705085763, "rank": 0.0} +{"step": 371, "steps": 500, "lr": 2e-05, "grad_norm": 2.4018051624298096, "loss_scale": 4.0, "retries": 0, "seconds": 497.7, "max_memory_gb": 14.31, "ce": 0.07451699754346919, "rank": 0.0} +{"step": 372, "steps": 500, "lr": 2e-05, "grad_norm": 0.2825275957584381, "loss_scale": 4.0, "retries": 0, "seconds": 499.0, "max_memory_gb": 14.31, "ce": 0.02939665040685213, "rank": 0.0} +{"step": 373, "steps": 500, "lr": 2e-05, "grad_norm": 0.42219269275665283, "loss_scale": 4.0, "retries": 0, "seconds": 500.4, "max_memory_gb": 14.31, "ce": 0.1050521582874353, "rank": 0.0} +{"step": 374, "steps": 500, "lr": 2e-05, "grad_norm": 0.500118613243103, "loss_scale": 4.0, "retries": 0, "seconds": 501.7, "max_memory_gb": 14.31, "ce": 0.14379386793007143, "rank": 0.0} +{"step": 375, "steps": 500, "lr": 2e-05, "grad_norm": 0.52887362241745, "loss_scale": 4.0, "retries": 0, "seconds": 503.1, "max_memory_gb": 14.31, "ce": 0.1639543635501468, "rank": 0.0} +{"step": 376, "steps": 500, "lr": 2e-05, "grad_norm": 0.4613296091556549, "loss_scale": 4.0, "retries": 0, "seconds": 504.3, "max_memory_gb": 14.31, "ce": 0.1028701320383334, "rank": 0.0} +{"step": 377, "steps": 500, "lr": 2e-05, "grad_norm": 0.3051157593727112, "loss_scale": 4.0, "retries": 0, "seconds": 505.6, "max_memory_gb": 14.31, "ce": 0.10107270010121283, "rank": 0.0} +{"step": 378, "steps": 500, "lr": 2e-05, "grad_norm": 0.7998569011688232, "loss_scale": 4.0, "retries": 0, "seconds": 506.9, "max_memory_gb": 14.31, "ce": 0.08544427667675336, "rank": 0.0} +{"step": 379, "steps": 500, "lr": 2e-05, "grad_norm": 0.35093310475349426, "loss_scale": 4.0, "retries": 0, "seconds": 508.3, "max_memory_gb": 14.31, "ce": 0.12571270345142693, "rank": 0.0} +{"step": 380, "steps": 500, "lr": 2e-05, "grad_norm": 1.4252156019210815, "loss_scale": 4.0, "retries": 0, "seconds": 509.6, "max_memory_gb": 14.31, "ce": 0.08834830069827149, "rank": 0.0} +{"step": 381, "steps": 500, "lr": 2e-05, "grad_norm": 0.6548739671707153, "loss_scale": 4.0, "retries": 0, "seconds": 510.9, "max_memory_gb": 14.31, "ce": 0.23925841272102844, "rank": 0.0} +{"step": 382, "steps": 500, "lr": 2e-05, "grad_norm": 0.9280415177345276, "loss_scale": 4.0, "retries": 0, "seconds": 512.2, "max_memory_gb": 14.31, "ce": 0.29398463886354875, "rank": 0.0} +{"step": 383, "steps": 500, "lr": 2e-05, "grad_norm": 0.31021663546562195, "loss_scale": 4.0, "retries": 0, "seconds": 513.5, "max_memory_gb": 14.31, "ce": 0.06824147330917185, "rank": 0.0} +{"step": 384, "steps": 500, "lr": 2e-05, "grad_norm": 0.3100737929344177, "loss_scale": 4.0, "retries": 0, "seconds": 514.9, "max_memory_gb": 14.31, "ce": 0.03935915944020962, "rank": 0.0} +{"step": 385, "steps": 500, "lr": 2e-05, "grad_norm": 0.27447909116744995, "loss_scale": 4.0, "retries": 0, "seconds": 516.3, "max_memory_gb": 14.31, "ce": 0.18752836921339622, "rank": 0.0} +{"step": 386, "steps": 500, "lr": 2e-05, "grad_norm": 0.3861882984638214, "loss_scale": 4.0, "retries": 0, "seconds": 517.6, "max_memory_gb": 14.31, "ce": 0.07137097405211534, "rank": 0.0} +{"step": 387, "steps": 500, "lr": 2e-05, "grad_norm": 0.41375845670700073, "loss_scale": 4.0, "retries": 0, "seconds": 519.0, "max_memory_gb": 14.31, "ce": 0.05826420359881013, "rank": 0.0} +{"step": 388, "steps": 500, "lr": 2e-05, "grad_norm": 0.5743300318717957, "loss_scale": 4.0, "retries": 0, "seconds": 520.3, "max_memory_gb": 14.31, "ce": 0.061653435210246244, "rank": 0.0} +{"step": 389, "steps": 500, "lr": 2e-05, "grad_norm": 0.799689531326294, "loss_scale": 4.0, "retries": 0, "seconds": 521.6, "max_memory_gb": 14.31, "ce": 0.020330044115326018, "rank": 0.10869129747152328} +{"step": 390, "steps": 500, "lr": 2e-05, "grad_norm": 5.2318220138549805, "loss_scale": 4.0, "retries": 0, "seconds": 522.9, "max_memory_gb": 14.31, "ce": 0.1631223237955055, "rank": 0.0} +{"step": 391, "steps": 500, "lr": 2e-05, "grad_norm": 3.7140450477600098, "loss_scale": 4.0, "retries": 0, "seconds": 524.3, "max_memory_gb": 14.31, "ce": 0.34012554657783767, "rank": 0.0} +{"step": 392, "steps": 500, "lr": 2e-05, "grad_norm": 0.3573310077190399, "loss_scale": 4.0, "retries": 0, "seconds": 525.6, "max_memory_gb": 14.31, "ce": 0.00873209228120686, "rank": 0.05697225034236908} +{"step": 393, "steps": 500, "lr": 2e-05, "grad_norm": 1.5892261266708374, "loss_scale": 4.0, "retries": 0, "seconds": 527.0, "max_memory_gb": 14.31, "ce": 0.11586651865763997, "rank": 0.0} +{"step": 394, "steps": 500, "lr": 2e-05, "grad_norm": 0.7231579422950745, "loss_scale": 4.0, "retries": 0, "seconds": 528.3, "max_memory_gb": 14.31, "ce": 0.10733471984713105, "rank": 0.0} +{"step": 395, "steps": 500, "lr": 2e-05, "grad_norm": 0.24991226196289062, "loss_scale": 4.0, "retries": 0, "seconds": 529.7, "max_memory_gb": 14.31, "ce": 0.033523920028528664, "rank": 0.0} +{"step": 396, "steps": 500, "lr": 2e-05, "grad_norm": 1.5363305807113647, "loss_scale": 4.0, "retries": 0, "seconds": 531.0, "max_memory_gb": 14.31, "ce": 0.04080619519481843, "rank": 0.0} +{"step": 397, "steps": 500, "lr": 2e-05, "grad_norm": 0.6193453073501587, "loss_scale": 4.0, "retries": 0, "seconds": 532.3, "max_memory_gb": 14.31, "ce": 0.13135220130425296, "rank": 0.0} +{"step": 398, "steps": 500, "lr": 2e-05, "grad_norm": 1.140594482421875, "loss_scale": 4.0, "retries": 0, "seconds": 533.7, "max_memory_gb": 14.31, "ce": 0.07306825288651453, "rank": 0.0} +{"step": 399, "steps": 500, "lr": 2e-05, "grad_norm": 0.8687056303024292, "loss_scale": 4.0, "retries": 0, "seconds": 535.0, "max_memory_gb": 14.31, "ce": 0.15806927811900096, "rank": 0.0} +{"step": 400, "steps": 500, "lr": 2e-05, "grad_norm": 3.2787015438079834, "loss_scale": 4.0, "retries": 0, "seconds": 536.4, "max_memory_gb": 14.31, "ce": 0.10652668639704643, "rank": 0.0} +{"step": 401, "steps": 500, "lr": 2e-05, "grad_norm": 2.914947509765625, "loss_scale": 4.0, "retries": 0, "seconds": 537.8, "max_memory_gb": 14.31, "ce": 0.20276531356103078, "rank": 0.0} +{"step": 402, "steps": 500, "lr": 2e-05, "grad_norm": 0.49273592233657837, "loss_scale": 4.0, "retries": 0, "seconds": 539.1, "max_memory_gb": 14.31, "ce": 0.121565768506116, "rank": 0.10780774056911469} +{"step": 403, "steps": 500, "lr": 2e-05, "grad_norm": 0.304914265871048, "loss_scale": 4.0, "retries": 0, "seconds": 540.4, "max_memory_gb": 14.31, "ce": 0.013886724307667464, "rank": 0.0} +{"step": 404, "steps": 500, "lr": 2e-05, "grad_norm": 0.595025897026062, "loss_scale": 4.0, "retries": 0, "seconds": 541.8, "max_memory_gb": 14.31, "ce": 0.12229933155686012, "rank": 0.0} +{"step": 405, "steps": 500, "lr": 2e-05, "grad_norm": 0.6564701795578003, "loss_scale": 4.0, "retries": 0, "seconds": 543.1, "max_memory_gb": 14.31, "ce": 0.0564923278070637, "rank": 0.0} +{"step": 406, "steps": 500, "lr": 2e-05, "grad_norm": 0.5460146069526672, "loss_scale": 4.0, "retries": 0, "seconds": 544.5, "max_memory_gb": 14.31, "ce": 0.14182741285549127, "rank": 0.0} +{"step": 407, "steps": 500, "lr": 2e-05, "grad_norm": 1.7641563415527344, "loss_scale": 4.0, "retries": 0, "seconds": 545.8, "max_memory_gb": 14.31, "ce": 0.7075795033360919, "rank": 0.12465488910675049} +{"step": 408, "steps": 500, "lr": 2e-05, "grad_norm": 0.6788272261619568, "loss_scale": 4.0, "retries": 0, "seconds": 547.2, "max_memory_gb": 14.31, "ce": 0.32115332430112176, "rank": 0.027390122413635254} +{"step": 409, "steps": 500, "lr": 2e-05, "grad_norm": 0.6610842347145081, "loss_scale": 4.0, "retries": 0, "seconds": 548.5, "max_memory_gb": 14.31, "ce": 0.15311965761793545, "rank": 0.0} +{"step": 410, "steps": 500, "lr": 2e-05, "grad_norm": 1.6245427131652832, "loss_scale": 4.0, "retries": 0, "seconds": 549.8, "max_memory_gb": 14.31, "ce": 0.06253041687705263, "rank": 0.0} +{"step": 411, "steps": 500, "lr": 2e-05, "grad_norm": 3.553316354751587, "loss_scale": 4.0, "retries": 0, "seconds": 551.2, "max_memory_gb": 14.31, "ce": 0.17820354956347728, "rank": 0.0} +{"step": 412, "steps": 500, "lr": 2e-05, "grad_norm": 0.4423668086528778, "loss_scale": 4.0, "retries": 0, "seconds": 552.5, "max_memory_gb": 14.31, "ce": 0.14218393499868398, "rank": 0.12485898286104202} +{"step": 413, "steps": 500, "lr": 2e-05, "grad_norm": 0.3626824617385864, "loss_scale": 4.0, "retries": 0, "seconds": 553.8, "max_memory_gb": 14.31, "ce": 0.06542103469837457, "rank": 0.0} +{"step": 414, "steps": 500, "lr": 2e-05, "grad_norm": 0.8555253148078918, "loss_scale": 4.0, "retries": 0, "seconds": 555.2, "max_memory_gb": 14.31, "ce": 0.17008214640372898, "rank": 0.0} +{"step": 415, "steps": 500, "lr": 2e-05, "grad_norm": 1.5838768482208252, "loss_scale": 4.0, "retries": 0, "seconds": 556.5, "max_memory_gb": 14.31, "ce": 0.3272358742215147, "rank": 0.0} +{"step": 416, "steps": 500, "lr": 2e-05, "grad_norm": 0.3629970848560333, "loss_scale": 4.0, "retries": 0, "seconds": 557.8, "max_memory_gb": 14.31, "ce": 0.029526978793001035, "rank": 0.0} +{"step": 417, "steps": 500, "lr": 2e-05, "grad_norm": 0.6679482460021973, "loss_scale": 4.0, "retries": 0, "seconds": 559.2, "max_memory_gb": 14.31, "ce": 0.3137528078950709, "rank": 0.12705036997795105} +{"step": 418, "steps": 500, "lr": 2e-05, "grad_norm": 0.23810526728630066, "loss_scale": 4.0, "retries": 0, "seconds": 560.6, "max_memory_gb": 14.31, "ce": 0.057489417580654845, "rank": 0.0} +{"step": 419, "steps": 500, "lr": 2e-05, "grad_norm": 0.3384636640548706, "loss_scale": 4.0, "retries": 0, "seconds": 561.8, "max_memory_gb": 14.31, "ce": 0.06239250668659224, "rank": 0.0} +{"step": 420, "steps": 500, "lr": 2e-05, "grad_norm": 1.5177278518676758, "loss_scale": 4.0, "retries": 0, "seconds": 563.2, "max_memory_gb": 14.31, "ce": 0.27360284921815037, "rank": 0.12224842607975006} +{"step": 421, "steps": 500, "lr": 2e-05, "grad_norm": 0.4308681786060333, "loss_scale": 4.0, "retries": 0, "seconds": 564.5, "max_memory_gb": 14.31, "ce": 0.06570040940277977, "rank": 0.09572216868400574} +{"step": 422, "steps": 500, "lr": 2e-05, "grad_norm": 2.976267099380493, "loss_scale": 4.0, "retries": 0, "seconds": 565.8, "max_memory_gb": 14.31, "ce": 0.2588291771662625, "rank": 0.0} +{"step": 423, "steps": 500, "lr": 2e-05, "grad_norm": 2.266702890396118, "loss_scale": 4.0, "retries": 0, "seconds": 567.2, "max_memory_gb": 14.31, "ce": 0.3056430322540109, "rank": 0.0} +{"step": 424, "steps": 500, "lr": 2e-05, "grad_norm": 0.518843412399292, "loss_scale": 4.0, "retries": 0, "seconds": 568.5, "max_memory_gb": 14.31, "ce": 0.07740016250318149, "rank": 0.0} +{"step": 425, "steps": 500, "lr": 2e-05, "grad_norm": 0.5542714595794678, "loss_scale": 4.0, "retries": 0, "seconds": 569.9, "max_memory_gb": 14.31, "ce": 0.11022032045912056, "rank": 0.0} +{"step": 426, "steps": 500, "lr": 2e-05, "grad_norm": 0.7162755131721497, "loss_scale": 4.0, "retries": 0, "seconds": 571.3, "max_memory_gb": 14.31, "ce": 0.2352472617494641, "rank": 0.0} +{"step": 427, "steps": 500, "lr": 2e-05, "grad_norm": 0.26113253831863403, "loss_scale": 4.0, "retries": 0, "seconds": 572.6, "max_memory_gb": 14.31, "ce": 0.05088500728561485, "rank": 0.0} +{"step": 428, "steps": 500, "lr": 2e-05, "grad_norm": 1.9293489456176758, "loss_scale": 4.0, "retries": 0, "seconds": 573.9, "max_memory_gb": 14.31, "ce": 0.07042501958221692, "rank": 0.0} +{"step": 429, "steps": 500, "lr": 2e-05, "grad_norm": 0.6607030034065247, "loss_scale": 4.0, "retries": 0, "seconds": 575.2, "max_memory_gb": 14.31, "ce": 0.023340416435530642, "rank": 0.0} +{"step": 430, "steps": 500, "lr": 2e-05, "grad_norm": 0.5320636630058289, "loss_scale": 4.0, "retries": 0, "seconds": 576.6, "max_memory_gb": 14.31, "ce": 0.11685411594589823, "rank": 0.0} +{"step": 431, "steps": 500, "lr": 2e-05, "grad_norm": 5.472825050354004, "loss_scale": 4.0, "retries": 0, "seconds": 578.0, "max_memory_gb": 14.31, "ce": 0.2542213397273372, "rank": 0.0} +{"step": 432, "steps": 500, "lr": 2e-05, "grad_norm": 0.5160524845123291, "loss_scale": 4.0, "retries": 0, "seconds": 579.3, "max_memory_gb": 14.31, "ce": 0.07316004942549625, "rank": 0.0} +{"step": 433, "steps": 500, "lr": 2e-05, "grad_norm": 0.818148136138916, "loss_scale": 4.0, "retries": 0, "seconds": 580.6, "max_memory_gb": 14.31, "ce": 0.08985995304828975, "rank": 0.0} +{"step": 434, "steps": 500, "lr": 2e-05, "grad_norm": 0.06892770528793335, "loss_scale": 4.0, "retries": 0, "seconds": 582.0, "max_memory_gb": 14.31, "ce": 0.007725826631030941, "rank": 0.0} +{"step": 435, "steps": 500, "lr": 2e-05, "grad_norm": 0.7242857813835144, "loss_scale": 4.0, "retries": 0, "seconds": 583.3, "max_memory_gb": 14.31, "ce": 0.13396304908383172, "rank": 0.0} +{"step": 436, "steps": 500, "lr": 2e-05, "grad_norm": 0.6825770139694214, "loss_scale": 4.0, "retries": 0, "seconds": 584.7, "max_memory_gb": 14.31, "ce": 0.14423368027291872, "rank": 0.0} +{"step": 437, "steps": 500, "lr": 2e-05, "grad_norm": 2.249507427215576, "loss_scale": 4.0, "retries": 0, "seconds": 586.0, "max_memory_gb": 14.31, "ce": 0.12997763425664743, "rank": 0.08194338530302048} +{"step": 438, "steps": 500, "lr": 2e-05, "grad_norm": 0.6880298256874084, "loss_scale": 4.0, "retries": 0, "seconds": 587.3, "max_memory_gb": 14.31, "ce": 0.09597323483649234, "rank": 0.0} +{"step": 439, "steps": 500, "lr": 2e-05, "grad_norm": 2.942896604537964, "loss_scale": 4.0, "retries": 0, "seconds": 588.7, "max_memory_gb": 14.31, "ce": 0.16685622720342508, "rank": 0.0} +{"step": 440, "steps": 500, "lr": 2e-05, "grad_norm": 0.26327189803123474, "loss_scale": 4.0, "retries": 0, "seconds": 590.1, "max_memory_gb": 14.31, "ce": 0.06057870179211022, "rank": 0.0} +{"step": 441, "steps": 500, "lr": 2e-05, "grad_norm": 1.310860514640808, "loss_scale": 4.0, "retries": 0, "seconds": 591.4, "max_memory_gb": 14.31, "ce": 0.17336103740672115, "rank": 0.0} +{"step": 442, "steps": 500, "lr": 2e-05, "grad_norm": 1.7978962659835815, "loss_scale": 4.0, "retries": 0, "seconds": 592.7, "max_memory_gb": 14.31, "ce": 0.09045626555598574, "rank": 0.0} +{"step": 443, "steps": 500, "lr": 2e-05, "grad_norm": 0.4452090561389923, "loss_scale": 4.0, "retries": 0, "seconds": 594.1, "max_memory_gb": 14.31, "ce": 0.08674739119851438, "rank": 0.0} +{"step": 444, "steps": 500, "lr": 2e-05, "grad_norm": 0.5207098126411438, "loss_scale": 4.0, "retries": 0, "seconds": 595.4, "max_memory_gb": 14.31, "ce": 0.08630627186175843, "rank": 0.0} +{"step": 445, "steps": 500, "lr": 2e-05, "grad_norm": 0.8044292330741882, "loss_scale": 4.0, "retries": 0, "seconds": 596.7, "max_memory_gb": 14.31, "ce": 0.3468537427306728, "rank": 0.06619074195623398} +{"step": 446, "steps": 500, "lr": 2e-05, "grad_norm": 10.69421100616455, "loss_scale": 4.0, "retries": 0, "seconds": 598.0, "max_memory_gb": 14.31, "ce": 0.35785341414884897, "rank": 0.4656437933444977} +{"step": 447, "steps": 500, "lr": 2e-05, "grad_norm": 0.7565954327583313, "loss_scale": 4.0, "retries": 0, "seconds": 599.3, "max_memory_gb": 14.31, "ce": 0.22849800278163457, "rank": 0.0} +{"step": 448, "steps": 500, "lr": 2e-05, "grad_norm": 2.008481740951538, "loss_scale": 4.0, "retries": 0, "seconds": 600.6, "max_memory_gb": 14.31, "ce": 0.29796834952867357, "rank": 0.0} +{"step": 449, "steps": 500, "lr": 2e-05, "grad_norm": 0.9356789588928223, "loss_scale": 4.0, "retries": 0, "seconds": 602.0, "max_memory_gb": 14.31, "ce": 0.02763635256269481, "rank": 0.0} +{"step": 450, "steps": 500, "lr": 2e-05, "grad_norm": 0.9881916642189026, "loss_scale": 4.0, "retries": 0, "seconds": 603.3, "max_memory_gb": 14.31, "ce": 0.041783950961871597, "rank": 0.0} +{"step": 451, "steps": 500, "lr": 2e-05, "grad_norm": 0.44091325998306274, "loss_scale": 4.0, "retries": 0, "seconds": 604.7, "max_memory_gb": 14.31, "ce": 0.03858639158352162, "rank": 0.0} +{"step": 452, "steps": 500, "lr": 2e-05, "grad_norm": 1.2310024499893188, "loss_scale": 4.0, "retries": 0, "seconds": 606.0, "max_memory_gb": 14.31, "ce": 0.17629267272968718, "rank": 0.0} +{"step": 453, "steps": 500, "lr": 2e-05, "grad_norm": 0.2965754568576813, "loss_scale": 4.0, "retries": 0, "seconds": 607.3, "max_memory_gb": 14.31, "ce": 0.03139853668471915, "rank": 0.0} +{"step": 454, "steps": 500, "lr": 2e-05, "grad_norm": 2.496615409851074, "loss_scale": 4.0, "retries": 0, "seconds": 608.6, "max_memory_gb": 14.31, "ce": 0.1711352834227, "rank": 0.0} +{"step": 455, "steps": 500, "lr": 2e-05, "grad_norm": 2.514329671859741, "loss_scale": 4.0, "retries": 0, "seconds": 610.0, "max_memory_gb": 14.31, "ce": 0.1416302940888272, "rank": 0.0} +{"step": 456, "steps": 500, "lr": 2e-05, "grad_norm": 0.16995543241500854, "loss_scale": 4.0, "retries": 0, "seconds": 611.3, "max_memory_gb": 14.31, "ce": 0.015881476385402493, "rank": 0.0} +{"step": 457, "steps": 500, "lr": 2e-05, "grad_norm": 0.4979867935180664, "loss_scale": 4.0, "retries": 0, "seconds": 612.7, "max_memory_gb": 14.31, "ce": 0.07606405206570344, "rank": 0.10370507836341858} +{"step": 458, "steps": 500, "lr": 2e-05, "grad_norm": 1.1307523250579834, "loss_scale": 4.0, "retries": 0, "seconds": 614.0, "max_memory_gb": 14.31, "ce": 0.14138779534278, "rank": 0.0} +{"step": 459, "steps": 500, "lr": 2e-05, "grad_norm": 0.9400743842124939, "loss_scale": 4.0, "retries": 0, "seconds": 615.3, "max_memory_gb": 14.31, "ce": 0.15738597209565341, "rank": 0.0} +{"step": 460, "steps": 500, "lr": 2e-05, "grad_norm": 0.3224273920059204, "loss_scale": 4.0, "retries": 0, "seconds": 616.7, "max_memory_gb": 14.31, "ce": 0.019048996949095454, "rank": 0.0} +{"step": 461, "steps": 500, "lr": 2e-05, "grad_norm": 0.5768384337425232, "loss_scale": 4.0, "retries": 0, "seconds": 618.0, "max_memory_gb": 14.31, "ce": 0.24155348941076227, "rank": 0.0} +{"step": 462, "steps": 500, "lr": 2e-05, "grad_norm": 0.26105430722236633, "loss_scale": 4.0, "retries": 0, "seconds": 619.4, "max_memory_gb": 14.31, "ce": 0.02757948591533932, "rank": 0.0} +{"step": 463, "steps": 500, "lr": 2e-05, "grad_norm": 0.15885639190673828, "loss_scale": 4.0, "retries": 0, "seconds": 620.7, "max_memory_gb": 14.31, "ce": 0.018529984254200826, "rank": 0.0} +{"step": 464, "steps": 500, "lr": 2e-05, "grad_norm": 0.44328945875167847, "loss_scale": 4.0, "retries": 0, "seconds": 622.1, "max_memory_gb": 14.31, "ce": 0.024723587166590733, "rank": 0.0} +{"step": 465, "steps": 500, "lr": 2e-05, "grad_norm": 1.5284970998764038, "loss_scale": 4.0, "retries": 0, "seconds": 623.4, "max_memory_gb": 14.31, "ce": 0.25795640282103705, "rank": 0.0} +{"step": 466, "steps": 500, "lr": 2e-05, "grad_norm": 0.8787432909011841, "loss_scale": 4.0, "retries": 0, "seconds": 624.8, "max_memory_gb": 14.31, "ce": 0.03196864045185066, "rank": 0.0} +{"step": 467, "steps": 500, "lr": 2e-05, "grad_norm": 0.3147842586040497, "loss_scale": 4.0, "retries": 0, "seconds": 626.1, "max_memory_gb": 14.31, "ce": 0.04171903129918064, "rank": 0.0} +{"step": 468, "steps": 500, "lr": 2e-05, "grad_norm": 0.20043884217739105, "loss_scale": 4.0, "retries": 0, "seconds": 627.4, "max_memory_gb": 14.31, "ce": 0.018261236830767302, "rank": 0.0} +{"step": 469, "steps": 500, "lr": 2e-05, "grad_norm": 3.837975025177002, "loss_scale": 4.0, "retries": 0, "seconds": 628.8, "max_memory_gb": 14.31, "ce": 0.13593106759799412, "rank": 0.0} +{"step": 470, "steps": 500, "lr": 2e-05, "grad_norm": 1.5386004447937012, "loss_scale": 4.0, "retries": 0, "seconds": 630.1, "max_memory_gb": 14.31, "ce": 0.09167932743366691, "rank": 0.0} +{"step": 471, "steps": 500, "lr": 2e-05, "grad_norm": 0.7994281649589539, "loss_scale": 4.0, "retries": 0, "seconds": 631.4, "max_memory_gb": 14.31, "ce": 0.1379868271842497, "rank": 0.0} +{"step": 472, "steps": 500, "lr": 2e-05, "grad_norm": 0.46097129583358765, "loss_scale": 4.0, "retries": 0, "seconds": 632.8, "max_memory_gb": 14.31, "ce": 0.036712284195346, "rank": 0.0} +{"step": 473, "steps": 500, "lr": 2e-05, "grad_norm": 0.7564078569412231, "loss_scale": 4.0, "retries": 0, "seconds": 634.2, "max_memory_gb": 14.31, "ce": 0.019517596550940652, "rank": 0.13153740763664246} +{"step": 474, "steps": 500, "lr": 2e-05, "grad_norm": 0.4266612231731415, "loss_scale": 4.0, "retries": 0, "seconds": 635.5, "max_memory_gb": 14.31, "ce": 0.09679604618759186, "rank": 0.0} +{"step": 475, "steps": 500, "lr": 2e-05, "grad_norm": 0.7517076134681702, "loss_scale": 4.0, "retries": 0, "seconds": 636.8, "max_memory_gb": 14.31, "ce": 0.12365267218319786, "rank": 0.0} +{"step": 476, "steps": 500, "lr": 2e-05, "grad_norm": 0.2881389260292053, "loss_scale": 4.0, "retries": 0, "seconds": 638.2, "max_memory_gb": 14.31, "ce": 0.012415387510372966, "rank": 0.0} +{"step": 477, "steps": 500, "lr": 2e-05, "grad_norm": 3.152191400527954, "loss_scale": 4.0, "retries": 0, "seconds": 639.5, "max_memory_gb": 14.31, "ce": 0.10871306856370211, "rank": 0.0} +{"step": 478, "steps": 500, "lr": 2e-05, "grad_norm": 0.3199784457683563, "loss_scale": 4.0, "retries": 0, "seconds": 640.8, "max_memory_gb": 14.31, "ce": 0.07690691344760126, "rank": 0.0} +{"step": 479, "steps": 500, "lr": 2e-05, "grad_norm": 5.051732063293457, "loss_scale": 4.0, "retries": 0, "seconds": 642.2, "max_memory_gb": 14.31, "ce": 0.38682363813222764, "rank": 0.0} +{"step": 480, "steps": 500, "lr": 2e-05, "grad_norm": 0.8147878050804138, "loss_scale": 4.0, "retries": 0, "seconds": 643.5, "max_memory_gb": 14.31, "ce": 0.16141483851924932, "rank": 0.0} +{"step": 481, "steps": 500, "lr": 2e-05, "grad_norm": 0.22401976585388184, "loss_scale": 4.0, "retries": 0, "seconds": 644.9, "max_memory_gb": 14.31, "ce": 0.03297034473143867, "rank": 0.0} +{"step": 482, "steps": 500, "lr": 2e-05, "grad_norm": 0.7776992321014404, "loss_scale": 4.0, "retries": 0, "seconds": 646.2, "max_memory_gb": 14.31, "ce": 0.08541737533596461, "rank": 0.0} +{"step": 483, "steps": 500, "lr": 2e-05, "grad_norm": 2.509174346923828, "loss_scale": 4.0, "retries": 0, "seconds": 647.5, "max_memory_gb": 14.31, "ce": 0.42974775958464306, "rank": 0.0} +{"step": 484, "steps": 500, "lr": 2e-05, "grad_norm": 1.0616765022277832, "loss_scale": 4.0, "retries": 0, "seconds": 648.8, "max_memory_gb": 14.31, "ce": 0.3665171457860197, "rank": 0.0} +{"step": 485, "steps": 500, "lr": 2e-05, "grad_norm": 0.6698402166366577, "loss_scale": 4.0, "retries": 0, "seconds": 650.2, "max_memory_gb": 14.31, "ce": 0.13960694886554847, "rank": 0.0} +{"step": 486, "steps": 500, "lr": 2e-05, "grad_norm": 1.0354299545288086, "loss_scale": 4.0, "retries": 0, "seconds": 651.4, "max_memory_gb": 14.31, "ce": 0.1755242246636044, "rank": 0.0} +{"step": 487, "steps": 500, "lr": 2e-05, "grad_norm": 1.9542572498321533, "loss_scale": 4.0, "retries": 0, "seconds": 652.6, "max_memory_gb": 14.31, "ce": 0.07365612015564693, "rank": 0.0} +{"step": 488, "steps": 500, "lr": 2e-05, "grad_norm": 0.6443522572517395, "loss_scale": 4.0, "retries": 0, "seconds": 653.8, "max_memory_gb": 14.31, "ce": 0.11125380691828468, "rank": 0.0} +{"step": 489, "steps": 500, "lr": 2e-05, "grad_norm": 4.890945911407471, "loss_scale": 4.0, "retries": 0, "seconds": 655.1, "max_memory_gb": 14.31, "ce": 0.26297477160005656, "rank": 0.0} +{"step": 490, "steps": 500, "lr": 2e-05, "grad_norm": 0.30534088611602783, "loss_scale": 4.0, "retries": 0, "seconds": 656.4, "max_memory_gb": 14.31, "ce": 0.047952394993444614, "rank": 0.0} +{"step": 491, "steps": 500, "lr": 2e-05, "grad_norm": 0.12197822332382202, "loss_scale": 4.0, "retries": 0, "seconds": 657.7, "max_memory_gb": 14.31, "ce": 0.10296174674476788, "rank": 0.0} +{"step": 492, "steps": 500, "lr": 2e-05, "grad_norm": 0.4002465009689331, "loss_scale": 4.0, "retries": 0, "seconds": 659.1, "max_memory_gb": 14.31, "ce": 0.018618271496961825, "rank": 0.0} +{"step": 493, "steps": 500, "lr": 2e-05, "grad_norm": 0.9406058192253113, "loss_scale": 4.0, "retries": 0, "seconds": 660.4, "max_memory_gb": 14.31, "ce": 0.11227174693249253, "rank": 0.0} +{"step": 494, "steps": 500, "lr": 2e-05, "grad_norm": 1.341115117073059, "loss_scale": 4.0, "retries": 0, "seconds": 661.8, "max_memory_gb": 14.31, "ce": 0.06674526272399817, "rank": 0.0} +{"step": 495, "steps": 500, "lr": 2e-05, "grad_norm": 0.6506909132003784, "loss_scale": 4.0, "retries": 0, "seconds": 663.1, "max_memory_gb": 14.31, "ce": 0.060273617284110514, "rank": 0.0} +{"step": 496, "steps": 500, "lr": 2e-05, "grad_norm": 1.4723135232925415, "loss_scale": 4.0, "retries": 0, "seconds": 664.5, "max_memory_gb": 14.31, "ce": 0.038337987041813903, "rank": 0.0} +{"step": 497, "steps": 500, "lr": 2e-05, "grad_norm": 0.7033885717391968, "loss_scale": 4.0, "retries": 0, "seconds": 665.8, "max_memory_gb": 14.31, "ce": 0.13668003341172152, "rank": 0.0} +{"step": 498, "steps": 500, "lr": 2e-05, "grad_norm": 0.7333466410636902, "loss_scale": 4.0, "retries": 0, "seconds": 667.1, "max_memory_gb": 14.31, "ce": 0.13333067632538587, "rank": 0.0} +{"step": 499, "steps": 500, "lr": 2e-05, "grad_norm": 0.7754046320915222, "loss_scale": 4.0, "retries": 0, "seconds": 668.5, "max_memory_gb": 14.31, "ce": 0.10632666449691897, "rank": 0.0} +{"step": 500, "steps": 500, "lr": 2e-05, "grad_norm": 0.8464720249176025, "loss_scale": 4.0, "retries": 0, "seconds": 669.8, "max_memory_gb": 14.31, "ce": 0.19651953382617648, "rank": 0.0}