{ "run_info": { "created_at": "2026-01-10T01:02:15+00:00", "total_time": 1158.007769990014, "experiment_name": "ln_tuning/llama-3.2-3B-default", "peft_branch": "main", "train_config": { "model_id": "meta-llama/Llama-3.2-3B", "dtype": "bfloat16", "max_seq_length": 768, "batch_size": 4, "batch_size_eval": 50, "max_steps": 5000, "eval_steps": 250, "compile": false, "query_template": "Question: {query} Think step by step.\nAnswer:", "seed": 0, "grad_norm_clip": 1.0, "optimizer_type": "AdamW", "optimizer_kwargs": { "lr": 0.0001, "weight_decay": 0.1 }, "lr_scheduler": "cosine", "use_amp": false, "autocast_adapter_dtype": true, "generation_kwargs": { "max_length": 800, "max_new_tokens": 300 }, "attn_implementation": null }, "peft_config": { "task_type": null, "peft_type": "LN_TUNING", "auto_mapping": null, "peft_version": "0.18.1.dev0@UNKNOWN", "base_model_name_or_path": "meta-llama/Llama-3.2-3B", "revision": null, "inference_mode": false, "target_modules": [ "norm", "input_layernorm", "post_attention_layernorm" ], "exclude_modules": null, "modules_to_save": null }, "error_msg": "" }, "train_info": { "accelerator_memory_reserved_avg": 13811999938, "accelerator_memory_max": 21179138048, "accelerator_memory_reserved_99th": 19146997760, "train_time": 933.4248713181005, "file_size": 358288, "num_trainable_params": 175104, "num_total_params": 3212924928, "status": "success", "metrics": [ { "step": 250, "valid accuracy": 0.0, "train loss": 1.3265766797065734, "train samples": 1000, "train time": 27.05523074005032, "eval time": 11.134952866006643, "tokens / sec": 7825.436864102909, "mem allocated avg": 6781727356.928, "mem reserved avg": 13917774413.824, "elapsed time": 59.58366395003395 }, { "step": 500, "valid accuracy": 0.0, "train loss": 1.3412033476829528, "train samples": 2000, "train time": 27.19167830166407, "eval time": 11.076495647954289, "tokens / sec": 7649.215237563, "mem allocated avg": 6774720333.824, "mem reserved avg": 13569009647.616, "elapsed time": 100.96434171200963 }, { "step": 750, "valid accuracy": 0.0, "train loss": 1.2789963104724884, "train samples": 3000, "train time": 27.539981623878703, "eval time": 11.047948513005394, "tokens / sec": 7785.081447334822, "mem allocated avg": 6784869052.416, "mem reserved avg": 13769388326.912, "elapsed time": 142.60399866703665 }, { "step": 1000, "valid accuracy": 0.0, "train loss": 1.2634205253124238, "train samples": 4000, "train time": 26.802510203735437, "eval time": 11.213655479019508, "tokens / sec": 7773.00329022781, "mem allocated avg": 6775999932.416, "mem reserved avg": 13779588874.24, "elapsed time": 183.79537835903466 }, { "step": 1250, "valid accuracy": 0.0, "train loss": 1.2484288651943207, "train samples": 5000, "train time": 26.859816205396783, "eval time": 11.04810167697724, "tokens / sec": 7763.939946770734, "mem allocated avg": 6776329017.344, "mem reserved avg": 13803437686.784, "elapsed time": 224.77404374204343 }, { "step": 1500, "valid accuracy": 0.0, "train loss": 1.231340999364853, "train samples": 6000, "train time": 27.544580540037714, "eval time": 11.118913470010739, "tokens / sec": 7599.716383254584, "mem allocated avg": 6778265929.728, "mem reserved avg": 13820894380.032, "elapsed time": 266.58437842602143 }, { "step": 1750, "valid accuracy": 0.0, "train loss": 1.2306929278373717, "train samples": 7000, "train time": 27.029298119421583, "eval time": 11.036927138979081, "tokens / sec": 7745.484143725154, "mem allocated avg": 6779571120.128, "mem reserved avg": 14093188595.712, "elapsed time": 307.76138371700654 }, { "step": 2000, "valid accuracy": 0.0, "train loss": 1.2307198526859284, "train samples": 8000, "train time": 26.702352902852, "eval time": 11.090556368988473, "tokens / sec": 7778.1909615094855, "mem allocated avg": 6775307636.736, "mem reserved avg": 13810257625.088, "elapsed time": 348.7669416820281 }, { "step": 2250, "valid accuracy": 0.0, "train loss": 1.2245776391029357, "train samples": 9000, "train time": 27.69884210533928, "eval time": 11.05135836795671, "tokens / sec": 7760.179980901304, "mem allocated avg": 6785839908.864, "mem reserved avg": 14071948640.256, "elapsed time": 390.5765162970056 }, { "step": 2500, "valid accuracy": 0.0, "train loss": 1.2369521839618682, "train samples": 10000, "train time": 27.172044244071003, "eval time": 11.106051086972002, "tokens / sec": 7580.10689773341, "mem allocated avg": 6772880177.152, "mem reserved avg": 13577423421.44, "elapsed time": 431.9964616240468 }, { "step": 2750, "valid accuracy": 0.0, "train loss": 1.2313981459140777, "train samples": 11000, "train time": 27.588061350223143, "eval time": 11.063589482975658, "tokens / sec": 7680.1699586725845, "mem allocated avg": 6782132961.28, "mem reserved avg": 13968802316.288, "elapsed time": 473.7305173620116 }, { "step": 3000, "valid accuracy": 0.0, "train loss": 1.2322822377681732, "train samples": 12000, "train time": 27.48076619656058, "eval time": 11.10096250998322, "tokens / sec": 7595.53057971594, "mem allocated avg": 6777935728.64, "mem reserved avg": 13851705737.216, "elapsed time": 515.4748477530084 }, { "step": 3250, "valid accuracy": 0.0, "train loss": 1.231231040239334, "train samples": 13000, "train time": 27.788235317741055, "eval time": 11.049090766988229, "tokens / sec": 7589.578740372652, "mem allocated avg": 6780943562.752, "mem reserved avg": 13802355556.352, "elapsed time": 557.366414040036 }, { "step": 3500, "valid accuracy": 0.0, "train loss": 1.2321618583202363, "train samples": 14000, "train time": 27.022960201604292, "eval time": 11.094650793995243, "tokens / sec": 7761.917955515015, "mem allocated avg": 6777875570.688, "mem reserved avg": 13817966755.84, "elapsed time": 598.6570834580343 }, { "step": 3750, "valid accuracy": 0.0, "train loss": 1.215895681142807, "train samples": 15000, "train time": 27.39095462608384, "eval time": 11.043718303029891, "tokens / sec": 7911.4803758478065, "mem allocated avg": 6788866940.928, "mem reserved avg": 14072342904.832, "elapsed time": 640.2053242530092 }, { "step": 4000, "valid accuracy": 0.0, "train loss": 1.2472150621414184, "train samples": 16000, "train time": 26.27257864776766, "eval time": 11.109217509045266, "tokens / sec": 7778.947119732583, "mem allocated avg": 6769699028.992, "mem reserved avg": 13815643111.424, "elapsed time": 680.7284473860054 }, { "step": 4250, "valid accuracy": 0.0, "train loss": 1.21587810254097, "train samples": 17000, "train time": 26.962625220243353, "eval time": 11.047069397987798, "tokens / sec": 7840.074854480067, "mem allocated avg": 6782053310.464, "mem reserved avg": 13778800345.088, "elapsed time": 721.68619730504 }, { "step": 4500, "valid accuracy": 0.0, "train loss": 1.241484440803528, "train samples": 18000, "train time": 26.840785892680287, "eval time": 11.098841341969091, "tokens / sec": 7742.61978881452, "mem allocated avg": 6775733297.152, "mem reserved avg": 13672147582.976, "elapsed time": 762.8068605409935 }, { "step": 4750, "valid accuracy": 0.0, "train loss": 1.2253197972774506, "train samples": 19000, "train time": 27.470003902097233, "eval time": 11.055242774018552, "tokens / sec": 7642.481622799185, "mem allocated avg": 6778600820.736, "mem reserved avg": 13782675881.984, "elapsed time": 804.4074719570344 }, { "step": 5000, "valid accuracy": 0.0, "train loss": 1.2366444897651672, "train samples": 20000, "train time": 26.569939231791068, "eval time": 11.095670820039231, "tokens / sec": 7838.93399917121, "mem allocated avg": 6775143475.2, "mem reserved avg": 13464646975.488, "elapsed time": 845.1059082580032 }, { "step": 5000, "test accuracy": 0.0, "train loss": 1.2366444897651672, "train samples": 20000, "train total tokens": 4198051, "forgetting": -0.032855987548828125 } ] }, "meta_info": { "model_info": { "sha": "13afe5124825b4f3751f836b40dafda64c1ed062", "created_at": "2024-09-18T15:23:48+00:00" }, "dataset_info": { "metamath": { "sha": "aa4f34d3d2d3231299b5b03d9b3e5a20da45aa18", "created_at": "2023-09-21T17:22:46+00:00" }, "gsm8k": { "sha": "cc7b047b6e5bb11b4f1af84efc572db110a51b3c", "created_at": "2022-04-12T10:22:10+00:00" } }, "package_info": { "transformers-version": "4.57.1", "transformers-commit-hash": null, "peft-version": "0.18.1.dev0", "peft-commit-hash": "8be1a16f5e06ca5e197d2af74bdfc5b3c8072d26", "datasets-version": "4.2.0", "datasets-commit-hash": null, "bitsandbytes-version": "0.46.0", "bitsandbytes-commit-hash": null, "torch-version": "2.9.0+cu128", "torch-commit-hash": null }, "system_info": { "system": "Linux", "release": "6.14.0-1016-aws", "version": "#16~24.04.1-Ubuntu SMP Tue Oct 14 02:15:09 UTC 2025", "machine": "x86_64", "processor": "x86_64", "accelerator": "NVIDIA L40S" }, "pytorch_info": "PyTorch built with:\n - GCC 13.3\n - C++ Version: 201703\n - Intel(R) oneAPI Math Kernel Library Version 2024.2-Product Build 20240605 for Intel(R) 64 architecture applications\n - Intel(R) MKL-DNN v3.7.1 (Git Hash 8d263e693366ef8db40acc569cc7d8edf644556d)\n - OpenMP 201511 (a.k.a. OpenMP 4.5)\n - LAPACK is enabled (usually provided by MKL)\n - NNPACK is enabled\n - CPU capability usage: AVX2\n - CUDA Runtime 12.8\n - NVCC architecture flags: -gencode;arch=compute_70,code=sm_70;-gencode;arch=compute_75,code=sm_75;-gencode;arch=compute_80,code=sm_80;-gencode;arch=compute_86,code=sm_86;-gencode;arch=compute_90,code=sm_90;-gencode;arch=compute_100,code=sm_100;-gencode;arch=compute_120,code=sm_120\n - CuDNN 90.7.1\n - Built with CuDNN 90.8\n - Magma 2.6.1\n - Build settings: BLAS_INFO=mkl, BUILD_TYPE=Release, COMMIT_SHA=0fabc3ba44823f257e70ce397d989c8de5e362c1, CUDA_VERSION=12.8, CUDNN_VERSION=9.8.0, CXX_COMPILER=/opt/rh/gcc-toolset-13/root/usr/bin/c++, CXX_FLAGS= -fvisibility-inlines-hidden -DUSE_PTHREADPOOL -DNDEBUG -DUSE_KINETO -DLIBKINETO_NOROCTRACER -DLIBKINETO_NOXPUPTI=ON -DUSE_FBGEMM -DUSE_PYTORCH_QNNPACK -DUSE_XNNPACK -DSYMBOLICATE_MOBILE_DEBUG_HANDLE -O2 -fPIC -DC10_NODEPRECATED -Wall -Wextra -Werror=return-type -Werror=non-virtual-dtor -Werror=range-loop-construct -Werror=bool-operation -Wnarrowing -Wno-missing-field-initializers -Wno-unknown-pragmas -Wno-unused-parameter -Wno-strict-overflow -Wno-strict-aliasing -Wno-stringop-overflow -Wsuggest-override -Wno-psabi -Wno-error=old-style-cast -faligned-new -Wno-maybe-uninitialized -fno-math-errno -fno-trapping-math -Werror=format -Wno-dangling-reference -Wno-error=dangling-reference -Wno-stringop-overflow, LAPACK_INFO=mkl, PERF_WITH_AVX=1, PERF_WITH_AVX2=1, TORCH_VERSION=2.9.0, USE_CUDA=ON, USE_CUDNN=ON, USE_CUSPARSELT=1, USE_GFLAGS=OFF, USE_GLOG=OFF, USE_GLOO=ON, USE_MKL=ON, USE_MKLDNN=ON, USE_MPI=OFF, USE_NCCL=1, USE_NNPACK=ON, USE_OPENMP=ON, USE_ROCM=OFF, USE_ROCM_KERNEL_ASSERT=OFF, USE_XCCL=OFF, USE_XPU=OFF, \n" } }