{ "run_info": { "created_at": "2026-01-10T10:27:51+00:00", "total_time": 1263.101585368975, "experiment_name": "vera/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": 0.001 }, "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": "VERA", "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, "r": 256, "target_modules": [ "q_proj", "v_proj" ], "projection_prng_key": 0, "save_projection": true, "vera_dropout": 0.0, "d_initial": 0.1, "fan_in_fan_out": false, "bias": "none", "modules_to_save": null, "init_weights": true, "layers_to_transform": null, "layers_pattern": null }, "error_msg": "" }, "train_info": { "accelerator_memory_reserved_avg": 13999608640, "accelerator_memory_max": 21617442816, "accelerator_memory_reserved_99th": 19627245568, "train_time": 1015.8542694449425, "file_size": 6821968, "num_trainable_params": 129024, "num_total_params": 3212878848, "status": "success", "metrics": [ { "step": 250, "valid accuracy": 0.0, "train loss": 1.301754014492035, "train samples": 1000, "train time": 31.943585795292165, "eval time": 12.187114808009937, "tokens / sec": 6627.903371800015, "mem allocated avg": 6789185892.352, "mem reserved avg": 14110276190.208, "elapsed time": 66.61339723097626 }, { "step": 500, "valid accuracy": 0.28, "train loss": 1.0202444262504577, "train samples": 2000, "train time": 31.78171904978808, "eval time": 12.214804693008773, "tokens / sec": 6544.48551615986, "mem allocated avg": 6782031269.888, "mem reserved avg": 13716204552.192, "elapsed time": 113.8828261839808 }, { "step": 750, "valid accuracy": 0.32, "train loss": 0.8039965138435364, "train samples": 3000, "train time": 31.674388502840884, "eval time": 12.142891115974635, "tokens / sec": 6768.90731389401, "mem allocated avg": 6791776645.12, "mem reserved avg": 13949500129.28, "elapsed time": 160.88182786101243 }, { "step": 1000, "valid accuracy": 0.32, "train loss": 0.7543835270404816, "train samples": 4000, "train time": 31.252004174282774, "eval time": 12.184918653976638, "tokens / sec": 6666.324464766307, "mem allocated avg": 6783568500.736, "mem reserved avg": 14006072901.632, "elapsed time": 207.57197764096782 }, { "step": 1250, "valid accuracy": 0.42, "train loss": 0.7379774446487427, "train samples": 5000, "train time": 31.04683574225055, "eval time": 12.13903358601965, "tokens / sec": 6716.884185276503, "mem allocated avg": 6783591102.464, "mem reserved avg": 14041858703.36, "elapsed time": 253.9649221689906 }, { "step": 1500, "valid accuracy": 0.42, "train loss": 0.725366308093071, "train samples": 6000, "train time": 31.466947606997564, "eval time": 12.193409828003496, "tokens / sec": 6652.408826379123, "mem allocated avg": 6785737506.816, "mem reserved avg": 13954449408.0, "elapsed time": 300.88531206996413 }, { "step": 1750, "valid accuracy": 0.38, "train loss": 0.7149190576076507, "train samples": 7000, "train time": 32.161987620871514, "eval time": 12.206813755037729, "tokens / sec": 6509.392468770777, "mem allocated avg": 6786781550.592, "mem reserved avg": 14342623854.592, "elapsed time": 348.52885854797205 }, { "step": 2000, "valid accuracy": 0.42, "train loss": 0.7139762370586396, "train samples": 8000, "train time": 31.66324118396733, "eval time": 12.41653193399543, "tokens / sec": 6559.530617641469, "mem allocated avg": 6782663055.36, "mem reserved avg": 13990637862.912, "elapsed time": 395.93328011297854 }, { "step": 2250, "valid accuracy": 0.38, "train loss": 0.7068367917537689, "train samples": 9000, "train time": 32.87184925627662, "eval time": 12.149466416973155, "tokens / sec": 6538.968900843246, "mem allocated avg": 6793174691.84, "mem reserved avg": 14273040351.232, "elapsed time": 444.18720530200517 }, { "step": 2500, "valid accuracy": 0.4, "train loss": 0.7048668594360351, "train samples": 10000, "train time": 31.740153488412034, "eval time": 12.253064631018788, "tokens / sec": 6489.162066440422, "mem allocated avg": 6780151590.912, "mem reserved avg": 13770160078.848, "elapsed time": 491.48139633599203 }, { "step": 2750, "valid accuracy": 0.36, "train loss": 0.6994699697494506, "train samples": 11000, "train time": 31.831262842926662, "eval time": 12.147469694027677, "tokens / sec": 6656.380585512423, "mem allocated avg": 6789358209.024, "mem reserved avg": 14143369248.768, "elapsed time": 538.6766653700033 }, { "step": 3000, "valid accuracy": 0.38, "train loss": 0.6912990446090699, "train samples": 12000, "train time": 31.59112943575019, "eval time": 12.194770312984474, "tokens / sec": 6607.266144900441, "mem allocated avg": 6785734909.952, "mem reserved avg": 14090034479.104, "elapsed time": 585.7846582700149 }, { "step": 3250, "valid accuracy": 0.42, "train loss": 0.7005903688669205, "train samples": 13000, "train time": 32.01279983797576, "eval time": 12.153572589973919, "tokens / sec": 6588.021074926875, "mem allocated avg": 6787881736.192, "mem reserved avg": 13914804846.592, "elapsed time": 633.0662019159645 }, { "step": 3500, "valid accuracy": 0.4, "train loss": 0.6888112008571625, "train samples": 14000, "train time": 31.394535964296665, "eval time": 12.198143090994563, "tokens / sec": 6681.098909649039, "mem allocated avg": 6784882763.776, "mem reserved avg": 13954659123.2, "elapsed time": 680.0055505509954 }, { "step": 3750, "valid accuracy": 0.36, "train loss": 0.6853219895362854, "train samples": 15000, "train time": 32.20891197276069, "eval time": 12.149777695012745, "tokens / sec": 6728.044715800003, "mem allocated avg": 6795943520.256, "mem reserved avg": 14263871602.688, "elapsed time": 727.6112716209609 }, { "step": 4000, "valid accuracy": 0.38, "train loss": 0.7033915517330169, "train samples": 16000, "train time": 31.605115150858182, "eval time": 12.220193530025426, "tokens / sec": 6466.453263165871, "mem allocated avg": 6777292670.976, "mem reserved avg": 13970463260.672, "elapsed time": 774.7509838030091 }, { "step": 4250, "valid accuracy": 0.34, "train loss": 0.6845373644828796, "train samples": 17000, "train time": 31.9817577842623, "eval time": 12.144686578016263, "tokens / sec": 6609.6742219722855, "mem allocated avg": 6788761931.776, "mem reserved avg": 13960019443.712, "elapsed time": 821.9710977289942 }, { "step": 4500, "valid accuracy": 0.4, "train loss": 0.6940178046226502, "train samples": 18000, "train time": 31.950214379117824, "eval time": 12.20877941895742, "tokens / sec": 6504.4320997053055, "mem allocated avg": 6783152502.784, "mem reserved avg": 13920844644.352, "elapsed time": 869.4368790709996 }, { "step": 4750, "valid accuracy": 0.36, "train loss": 0.6864041843414307, "train samples": 19000, "train time": 31.99477967567509, "eval time": 12.149096956010908, "tokens / sec": 6561.664187974136, "mem allocated avg": 6785828032.512, "mem reserved avg": 13979548123.136, "elapsed time": 916.7890869269613 }, { "step": 5000, "valid accuracy": 0.38, "train loss": 0.6927255519628525, "train samples": 20000, "train time": 30.892451956111472, "eval time": 12.215222013997845, "tokens / sec": 6742.099989210984, "mem allocated avg": 6782438621.184, "mem reserved avg": 13639734001.664, "elapsed time": 963.0640164369834 }, { "step": 5000, "test accuracy": 0.37604245640636846, "train loss": 0.6927255519628525, "train samples": 20000, "train total tokens": 4198051, "forgetting": 0.1016998291015625 } ] }, "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" } }