{ "run_info": { "created_at": "2026-01-10T00:21:54+00:00", "total_time": 1252.7159106170293, "experiment_name": "ia3/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": "IA3", "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": [ "k_proj", "down_proj", "v_proj" ], "exclude_modules": null, "feedforward_modules": [ "down_proj" ], "fan_in_fan_out": false, "modules_to_save": null, "init_ia3_weights": true }, "error_msg": "" }, "train_info": { "accelerator_memory_reserved_avg": 14786832524, "accelerator_memory_max": 23240638464, "accelerator_memory_reserved_99th": 20933771264, "train_time": 1015.700665248034, "file_size": 1157064, "num_trainable_params": 286720, "num_total_params": 3213036544, "status": "success", "metrics": [ { "step": 250, "valid accuracy": 0.0, "train loss": 1.3155594708919525, "train samples": 1000, "train time": 29.70507554640062, "eval time": 11.724285421019886, "tokens / sec": 7127.367835482717, "mem allocated avg": 6782600361.984, "mem reserved avg": 14885492621.312, "elapsed time": 63.95026855904143 }, { "step": 500, "valid accuracy": 0.0, "train loss": 1.2050744531154634, "train samples": 2000, "train time": 29.590323601732962, "eval time": 11.715296942042187, "tokens / sec": 7029.155976780827, "mem allocated avg": 6775425632.256, "mem reserved avg": 14480272523.264, "elapsed time": 108.9384753460181 }, { "step": 750, "valid accuracy": 0.12, "train loss": 1.0194114191532135, "train samples": 3000, "train time": 30.585519072716124, "eval time": 11.650421504979022, "tokens / sec": 7009.885936225842, "mem allocated avg": 6785668968.448, "mem reserved avg": 14756492607.488, "elapsed time": 154.82391193299554 }, { "step": 1000, "valid accuracy": 0.3, "train loss": 0.9196182670593261, "train samples": 4000, "train time": 29.49092253309209, "eval time": 11.840653097024187, "tokens / sec": 7064.411083316362, "mem allocated avg": 6777244168.192, "mem reserved avg": 14776985976.832, "elapsed time": 199.8590077710105 }, { "step": 1250, "valid accuracy": 0.3, "train loss": 0.8686861591339111, "train samples": 5000, "train time": 29.44121315999655, "eval time": 11.64985877397703, "tokens / sec": 7083.1999641696975, "mem allocated avg": 6777135923.2, "mem reserved avg": 14839564992.512, "elapsed time": 244.56394972099224 }, { "step": 1500, "valid accuracy": 0.3, "train loss": 0.8333434355258942, "train samples": 6000, "train time": 29.930749938997906, "eval time": 11.69840682001086, "tokens / sec": 6993.844137772663, "mem allocated avg": 6779035060.224, "mem reserved avg": 14722384527.36, "elapsed time": 289.89695524802664 }, { "step": 1750, "valid accuracy": 0.3, "train loss": 0.8170836737155914, "train samples": 7000, "train time": 30.3804316627793, "eval time": 11.654387208051048, "tokens / sec": 6891.113408914859, "mem allocated avg": 6779937654.784, "mem reserved avg": 15161679151.104, "elapsed time": 335.59190371102886 }, { "step": 2000, "valid accuracy": 0.3, "train loss": 0.8073079003095627, "train samples": 8000, "train time": 29.52141359890811, "eval time": 11.679582766955718, "tokens / sec": 7035.435457863099, "mem allocated avg": 6776053977.088, "mem reserved avg": 14763992023.04, "elapsed time": 380.5167595850071 }, { "step": 2250, "valid accuracy": 0.28, "train loss": 0.795267580151558, "train samples": 9000, "train time": 30.315992950985674, "eval time": 11.681108721997589, "tokens / sec": 7090.25102187891, "mem allocated avg": 6787896084.48, "mem reserved avg": 15088715038.72, "elapsed time": 426.1353191550006 }, { "step": 2500, "valid accuracy": 0.3, "train loss": 0.7891060090065003, "train samples": 10000, "train time": 29.285075834020972, "eval time": 11.683325350983068, "tokens / sec": 7033.17284091594, "mem allocated avg": 6772744847.36, "mem reserved avg": 14541761019.904, "elapsed time": 470.7990039800061 }, { "step": 2750, "valid accuracy": 0.3, "train loss": 0.7823175694942475, "train samples": 11000, "train time": 30.584478964097798, "eval time": 11.64602530299453, "tokens / sec": 6927.729592801655, "mem allocated avg": 6783104911.36, "mem reserved avg": 14943508234.24, "elapsed time": 516.6453109350405 }, { "step": 3000, "valid accuracy": 0.26, "train loss": 0.7710221862792969, "train samples": 12000, "train time": 29.543955663975794, "eval time": 11.677449156995863, "tokens / sec": 7065.099960683823, "mem allocated avg": 6778547337.216, "mem reserved avg": 14865569677.312, "elapsed time": 561.5630381150404 }, { "step": 3250, "valid accuracy": 0.3, "train loss": 0.7756292035579682, "train samples": 13000, "train time": 29.600664581230376, "eval time": 11.676535067963414, "tokens / sec": 7124.873815628153, "mem allocated avg": 6780569153.536, "mem reserved avg": 14715774304.256, "elapsed time": 606.3470957170357 }, { "step": 3500, "valid accuracy": 0.32, "train loss": 0.7658540964126587, "train samples": 14000, "train time": 29.628978170803748, "eval time": 11.7645476309699, "tokens / sec": 7079.218148896091, "mem allocated avg": 6778699649.024, "mem reserved avg": 14727702904.832, "elapsed time": 651.4321867520339 }, { "step": 3750, "valid accuracy": 0.36, "train loss": 0.7585025093555451, "train samples": 15000, "train time": 30.480809888918884, "eval time": 11.652502680022735, "tokens / sec": 7109.489570314241, "mem allocated avg": 6790036742.144, "mem reserved avg": 15097195921.408, "elapsed time": 697.2096698190435 }, { "step": 4000, "valid accuracy": 0.34, "train loss": 0.7766863429546356, "train samples": 16000, "train time": 29.984013571229298, "eval time": 11.786360332975164, "tokens / sec": 6816.065484845664, "mem allocated avg": 6770827464.704, "mem reserved avg": 14748657647.616, "elapsed time": 742.6605575150461 }, { "step": 4250, "valid accuracy": 0.32, "train loss": 0.7541669148206711, "train samples": 17000, "train time": 30.13022418995388, "eval time": 11.657879816018976, "tokens / sec": 7015.845573113327, "mem allocated avg": 6782400247.808, "mem reserved avg": 14721461780.48, "elapsed time": 787.9822572360281 }, { "step": 4500, "valid accuracy": 0.3, "train loss": 0.7659742934703827, "train samples": 18000, "train time": 30.020536605035886, "eval time": 11.697008132992778, "tokens / sec": 6922.527826006246, "mem allocated avg": 6776485750.784, "mem reserved avg": 14703325609.984, "elapsed time": 833.4761402930017 }, { "step": 4750, "valid accuracy": 0.32, "train loss": 0.7579846435785293, "train samples": 19000, "train time": 30.37128037889488, "eval time": 11.668571799993515, "tokens / sec": 6912.4184881546, "mem allocated avg": 6779605635.072, "mem reserved avg": 14800801234.944, "elapsed time": 879.1311652210425 }, { "step": 5000, "valid accuracy": 0.34, "train loss": 0.7657218432426453, "train samples": 20000, "train time": 29.584271008439828, "eval time": 11.69853776000673, "tokens / sec": 7040.227556750738, "mem allocated avg": 6776251918.336, "mem reserved avg": 14395312701.44, "elapsed time": 924.0385756980395 }, { "step": 5000, "test accuracy": 0.35178165276724793, "train loss": 0.7657218432426453, "train samples": 20000, "train total tokens": 4198051, "forgetting": 0.1822957992553711 } ] }, "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" } }