{ "run_info": { "created_at": "2026-01-09T23:21:09+00:00", "total_time": 2347.218174619018, "experiment_name": "full-finetuning/llama-3.2-3B-lr_0.00001", "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": 1e-05 }, "lr_scheduler": "cosine", "use_amp": false, "autocast_adapter_dtype": true, "generation_kwargs": { "max_length": 800, "max_new_tokens": 300 }, "attn_implementation": null }, "peft_config": null, "error_msg": "" }, "train_info": { "accelerator_memory_reserved_avg": 34309618532, "accelerator_memory_max": 38549848064, "accelerator_memory_reserved_99th": 37237492285, "train_time": 2154.846133010811, "file_size": 6425499648, "num_trainable_params": 3212749824, "num_total_params": 3212749824, "status": "success", "metrics": [ { "step": 250, "valid accuracy": 0.3, "train loss": 1.0749022269248962, "train samples": 1000, "train time": 91.49652248725761, "eval time": 11.212762297014706, "tokens / sec": 2313.956795784073, "mem allocated avg": 26068575909.888, "mem reserved avg": 34185951576.064, "elapsed time": 124.61640950996662 }, { "step": 500, "valid accuracy": 0.4, "train loss": 0.7238605138063431, "train samples": 2000, "train time": 91.91453206090955, "eval time": 11.141172014991753, "tokens / sec": 2262.917466219234, "mem allocated avg": 26061334226.944, "mem reserved avg": 34147850518.528, "elapsed time": 230.16530759999296 }, { "step": 750, "valid accuracy": 0.42, "train loss": 0.6648618532419205, "train samples": 3000, "train time": 92.5216026121634, "eval time": 6.087067044980358, "tokens / sec": 2317.3074606017867, "mem allocated avg": 26070816337.92, "mem reserved avg": 34269409837.056, "elapsed time": 331.3244165930082 }, { "step": 1000, "valid accuracy": 0.4, "train loss": 0.6407386429309845, "train samples": 4000, "train time": 91.41802564635873, "eval time": 11.26410798204597, "tokens / sec": 2278.9378629322678, "mem allocated avg": 26063015718.912, "mem reserved avg": 34418341183.488, "elapsed time": 436.4462315859855 }, { "step": 1250, "valid accuracy": 0.46, "train loss": 0.6343499521017074, "train samples": 5000, "train time": 91.51783608621918, "eval time": 9.06717514997581, "tokens / sec": 2278.6596462304447, "mem allocated avg": 26061503963.136, "mem reserved avg": 34260576632.832, "elapsed time": 539.5919883140014 }, { "step": 1500, "valid accuracy": 0.58, "train loss": 0.624906349182129, "train samples": 6000, "train time": 91.87651025888044, "eval time": 11.171614417049568, "tokens / sec": 2278.395200363706, "mem allocated avg": 26064701278.208, "mem reserved avg": 34224690167.808, "elapsed time": 645.2011065029656 }, { "step": 1750, "valid accuracy": 0.54, "train loss": 0.6175477323532105, "train samples": 7000, "train time": 90.95103025593562, "eval time": 11.160093326005153, "tokens / sec": 2301.843084249583, "mem allocated avg": 26063144126.464, "mem reserved avg": 34362062012.416, "elapsed time": 749.8216157860006 }, { "step": 2000, "valid accuracy": 0.42, "train loss": 0.6184427078962326, "train samples": 8000, "train time": 90.84291155746905, "eval time": 11.153122653020546, "tokens / sec": 2286.3203792032505, "mem allocated avg": 26060605181.952, "mem reserved avg": 34304549715.968, "elapsed time": 854.322792426974 }, { "step": 2250, "valid accuracy": 0.46, "train loss": 0.6107840378284455, "train samples": 9000, "train time": 92.96169786009705, "eval time": 11.16127427702304, "tokens / sec": 2312.221107702729, "mem allocated avg": 26073147115.52, "mem reserved avg": 34312711831.552, "elapsed time": 960.9334694420104 }, { "step": 2500, "valid accuracy": 0.54, "train loss": 0.6089447345733643, "train samples": 10000, "train time": 90.26249368820572, "eval time": 11.146873010031413, "tokens / sec": 2281.8669370189687, "mem allocated avg": 26057948864.512, "mem reserved avg": 34399810748.416, "elapsed time": 1064.9780696959933 }, { "step": 2750, "valid accuracy": 0.56, "train loss": 0.6021645392179489, "train samples": 11000, "train time": 91.39197911316296, "eval time": 11.157545586000197, "tokens / sec": 2318.376317659624, "mem allocated avg": 26068202784.768, "mem reserved avg": 34373529239.552, "elapsed time": 1170.0117728059995 }, { "step": 3000, "valid accuracy": 0.48, "train loss": 0.5950663160085679, "train samples": 12000, "train time": 91.94340823084349, "eval time": 6.359558377007488, "tokens / sec": 2270.2116879976475, "mem allocated avg": 26062483675.136, "mem reserved avg": 34305480851.456, "elapsed time": 1270.7611719949637 }, { "step": 3250, "valid accuracy": 0.5, "train loss": 0.6067559220790864, "train samples": 13000, "train time": 91.35254304017872, "eval time": 6.207320767978672, "tokens / sec": 2308.649469202422, "mem allocated avg": 26064313700.352, "mem reserved avg": 34178494103.552, "elapsed time": 1370.8038186989725 }, { "step": 3500, "valid accuracy": 0.46, "train loss": 0.5922972733974456, "train samples": 14000, "train time": 91.81345162418438, "eval time": 7.237666736007668, "tokens / sec": 2284.523632316533, "mem allocated avg": 26062809985.024, "mem reserved avg": 34401211645.952, "elapsed time": 1472.3757120459923 }, { "step": 3750, "valid accuracy": 0.5, "train loss": 0.5926397043466568, "train samples": 15000, "train time": 92.5030986299389, "eval time": 11.144441604032181, "tokens / sec": 2342.6566591777223, "mem allocated avg": 26074840416.256, "mem reserved avg": 34314205003.776, "elapsed time": 1578.5238473899662 }, { "step": 4000, "valid accuracy": 0.52, "train loss": 0.6050900548696518, "train samples": 16000, "train time": 90.88473943283316, "eval time": 11.175483218976296, "tokens / sec": 2248.7053522449546, "mem allocated avg": 26056039493.632, "mem reserved avg": 34580459421.696, "elapsed time": 1683.2547346790088 }, { "step": 4250, "valid accuracy": 0.52, "train loss": 0.5931364653110505, "train samples": 17000, "train time": 91.31777060282184, "eval time": 7.878185064007994, "tokens / sec": 2314.8725445720397, "mem allocated avg": 26067197499.392, "mem reserved avg": 34410673995.776, "elapsed time": 1785.080884763971 }, { "step": 4500, "valid accuracy": 0.48, "train loss": 0.6010294594764709, "train samples": 18000, "train time": 91.71872854570393, "eval time": 6.507075314002577, "tokens / sec": 2265.8185879282355, "mem allocated avg": 26062461708.288, "mem reserved avg": 34202108035.072, "elapsed time": 1885.7456577859703 }, { "step": 4750, "valid accuracy": 0.5, "train loss": 0.5937523410320282, "train samples": 19000, "train time": 91.9850392206572, "eval time": 6.368945363035891, "tokens / sec": 2282.31679606496, "mem allocated avg": 26063243161.6, "mem reserved avg": 34250082484.224, "elapsed time": 1986.7437554029748 }, { "step": 5000, "valid accuracy": 0.46, "train loss": 0.5987640290260315, "train samples": 20000, "train time": 90.72794272896135, "eval time": 11.154609176970553, "tokens / sec": 2295.654389764034, "mem allocated avg": 26061294045.184, "mem reserved avg": 34290171641.856, "elapsed time": 2091.1681218339945 }, { "step": 5000, "test accuracy": 0.49507202426080366, "train loss": 0.5987640290260315, "train samples": 20000, "train total tokens": 4198051, "forgetting": 0.5911064147949219 } ] }, "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" } }