{ "run_info": { "created_at": "2026-01-10T05:17:10+00:00", "total_time": 1145.4119831069838, "experiment_name": "miss/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": "MISS", "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": 64, "miss_dropout": 0.0, "mini_r": 1, "target_modules": [ "v_proj", "q_proj" ], "exclude_modules": null, "init_weights": true, "layers_to_transform": null, "layers_pattern": null, "bias": "none", "modules_to_save": null }, "error_msg": "" }, "train_info": { "accelerator_memory_reserved_avg": 13443749694, "accelerator_memory_max": 20275265536, "accelerator_memory_reserved_99th": 18498977792, "train_time": 930.4467204429675, "file_size": 29367496, "num_trainable_params": 7340032, "num_total_params": 3220089856, "status": "success", "metrics": [ { "step": 250, "valid accuracy": 0.4, "train loss": 0.8770495002269745, "train samples": 1000, "train time": 29.581142008537427, "eval time": 11.88444348500343, "tokens / sec": 7157.228748602595, "mem allocated avg": 6897714161.664, "mem reserved avg": 13519852404.736, "elapsed time": 63.32356589497067 }, { "step": 500, "valid accuracy": 0.42, "train loss": 0.6948347247838974, "train samples": 2000, "train time": 28.6214671752532, "eval time": 11.315816484973766, "tokens / sec": 7267.097760097966, "mem allocated avg": 6890609534.976, "mem reserved avg": 13203794821.12, "elapsed time": 105.93744991498534 }, { "step": 750, "valid accuracy": 0.4, "train loss": 0.6675079126358032, "train samples": 3000, "train time": 28.836697526625358, "eval time": 8.011410179024097, "tokens / sec": 7435.005336586838, "mem allocated avg": 6900360769.536, "mem reserved avg": 13417939206.144, "elapsed time": 145.43328489601845 }, { "step": 1000, "valid accuracy": 0.36, "train loss": 0.6478000595569611, "train samples": 4000, "train time": 28.465044133423362, "eval time": 7.135868438985199, "tokens / sec": 7319.0120142963, "mem allocated avg": 6891700822.016, "mem reserved avg": 13453288800.256, "elapsed time": 183.82717936398694 }, { "step": 1250, "valid accuracy": 0.34, "train loss": 0.644065090060234, "train samples": 5000, "train time": 28.294503198412713, "eval time": 6.2051020989893, "tokens / sec": 7370.265473037135, "mem allocated avg": 6892015517.696, "mem reserved avg": 13493218574.336, "elapsed time": 220.99409208400175 }, { "step": 1500, "valid accuracy": 0.5, "train loss": 0.6371824450492859, "train samples": 6000, "train time": 28.5445017926977, "eval time": 7.893026220030151, "tokens / sec": 7333.496360183501, "mem allocated avg": 6894076573.696, "mem reserved avg": 13404207054.848, "elapsed time": 260.1458303320105 }, { "step": 1750, "valid accuracy": 0.46, "train loss": 0.6281298241615295, "train samples": 7000, "train time": 28.726283402822446, "eval time": 11.832006459997501, "tokens / sec": 7287.925035907368, "mem allocated avg": 6895286140.928, "mem reserved avg": 13746437095.424, "elapsed time": 303.4682386519853 }, { "step": 2000, "valid accuracy": 0.4, "train loss": 0.6284154652357101, "train samples": 8000, "train time": 28.592419658671133, "eval time": 11.856868351984303, "tokens / sec": 7264.023208928129, "mem allocated avg": 6891065575.424, "mem reserved avg": 13428383023.104, "elapsed time": 346.6774228439899 }, { "step": 2250, "valid accuracy": 0.44, "train loss": 0.6165187916755677, "train samples": 9000, "train time": 29.799386566970497, "eval time": 11.83944596600486, "tokens / sec": 7213.168617311317, "mem allocated avg": 6901582643.2, "mem reserved avg": 13693672751.104, "elapsed time": 391.026491408993 }, { "step": 2500, "valid accuracy": 0.5, "train loss": 0.6125160049200058, "train samples": 10000, "train time": 28.65632294554962, "eval time": 8.679996104969177, "tokens / sec": 7187.488792311613, "mem allocated avg": 6888456951.808, "mem reserved avg": 13231275900.928, "elapsed time": 431.072143708996 }, { "step": 2750, "valid accuracy": 0.46, "train loss": 0.6009856187105179, "train samples": 11000, "train time": 28.90840060118353, "eval time": 11.810368279984687, "tokens / sec": 7329.391996571593, "mem allocated avg": 6897535533.056, "mem reserved avg": 13570335047.68, "elapsed time": 474.47477067 }, { "step": 3000, "valid accuracy": 0.44, "train loss": 0.5904879401922226, "train samples": 12000, "train time": 28.64282536803512, "eval time": 11.851109263021499, "tokens / sec": 7287.374667756766, "mem allocated avg": 6893574338.56, "mem reserved avg": 13508544561.152, "elapsed time": 517.7344847109634 }, { "step": 3250, "valid accuracy": 0.52, "train loss": 0.5999962613582611, "train samples": 13000, "train time": 28.622812291316222, "eval time": 11.942176382988691, "tokens / sec": 7368.283656179534, "mem allocated avg": 6896400535.552, "mem reserved avg": 13368513527.808, "elapsed time": 560.9433922549943 }, { "step": 3500, "valid accuracy": 0.54, "train loss": 0.581193753361702, "train samples": 14000, "train time": 28.940197857038584, "eval time": 8.767924475017935, "tokens / sec": 7247.704422621507, "mem allocated avg": 6893331273.728, "mem reserved avg": 13393645797.376, "elapsed time": 601.3920295889839 }, { "step": 3750, "valid accuracy": 0.52, "train loss": 0.5779475719928742, "train samples": 15000, "train time": 30.048772321257275, "eval time": 11.848424528026953, "tokens / sec": 7211.708940491347, "mem allocated avg": 6904537260.032, "mem reserved avg": 13701608374.272, "elapsed time": 646.0380702709663 }, { "step": 4000, "valid accuracy": 0.5, "train loss": 0.5872079440355301, "train samples": 16000, "train time": 28.764328929479234, "eval time": 11.874075043015182, "tokens / sec": 7105.084930055418, "mem allocated avg": 6885427052.544, "mem reserved avg": 13412377559.04, "elapsed time": 689.4377979059936 }, { "step": 4250, "valid accuracy": 0.5, "train loss": 0.5736851218938828, "train samples": 17000, "train time": 29.44270641100593, "eval time": 11.804472424963024, "tokens / sec": 7179.672855107539, "mem allocated avg": 6897404037.12, "mem reserved avg": 13391523479.552, "elapsed time": 733.3348887860193 }, { "step": 4500, "valid accuracy": 0.48, "train loss": 0.5799340176582336, "train samples": 18000, "train time": 28.752423408732284, "eval time": 11.81368973600911, "tokens / sec": 7227.842921125195, "mem allocated avg": 6891705860.096, "mem reserved avg": 13364461830.144, "elapsed time": 776.6694126090151 }, { "step": 4750, "valid accuracy": 0.46, "train loss": 0.569597121834755, "train samples": 19000, "train time": 28.624554165056907, "eval time": 11.791828219022136, "tokens / sec": 7334.2277678609435, "mem allocated avg": 6894110355.456, "mem reserved avg": 13432342446.08, "elapsed time": 819.7075462730136 }, { "step": 5000, "valid accuracy": 0.5, "train loss": 0.5774972836971283, "train samples": 20000, "train time": 28.596162587520666, "eval time": 11.84556078800233, "tokens / sec": 7283.49474732995, "mem allocated avg": 6891173048.32, "mem reserved avg": 13139571638.272, "elapsed time": 862.8139132329961 }, { "step": 5000, "test accuracy": 0.5072024260803639, "train loss": 0.5774972836971283, "train samples": 20000, "train total tokens": 4198051, "forgetting": 0.2503843307495117 } ] }, "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" } }