{ "run_info": { "created_at": "2026-01-10T03:21:22+00:00", "total_time": 1503.135829265986, "experiment_name": "lora/llama-3.2-3B-rank32-dora", "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": "CAUSAL_LM", "peft_type": "LORA", "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": 32, "target_modules": [ "q_proj", "v_proj" ], "exclude_modules": null, "lora_alpha": 64, "lora_dropout": 0.0, "fan_in_fan_out": false, "bias": "none", "use_rslora": false, "modules_to_save": null, "init_lora_weights": true, "layers_to_transform": null, "layers_pattern": null, "rank_pattern": {}, "alpha_pattern": {}, "megatron_config": null, "megatron_core": "megatron.core", "trainable_token_indices": null, "loftq_config": {}, "eva_config": null, "corda_config": null, "use_dora": true, "alora_invocation_tokens": null, "use_qalora": false, "qalora_group_size": 16, "layer_replication": null, "lora_bias": false, "target_parameters": null, "use_bdlora": null, "arrow_config": null, "ensure_weight_tying": false }, "error_msg": "" }, "train_info": { "accelerator_memory_reserved_avg": 15412689371, "accelerator_memory_max": 24547164160, "accelerator_memory_reserved_99th": 22162702336, "train_time": 1216.6712898398982, "file_size": 37181760, "num_trainable_params": 9289728, "num_total_params": 3222039552, "status": "success", "metrics": [ { "step": 250, "valid accuracy": 0.26, "train loss": 0.9799610931873322, "train samples": 1000, "train time": 35.417795918125194, "eval time": 17.352634231036063, "tokens / sec": 5977.757636003882, "mem allocated avg": 6925407713.28, "mem reserved avg": 15541154611.2, "elapsed time": 75.89598628797103 }, { "step": 500, "valid accuracy": 0.44, "train loss": 0.7162615118026734, "train samples": 2000, "train time": 35.72266378562199, "eval time": 17.35569141904125, "tokens / sec": 5822.494124408378, "mem allocated avg": 6918067554.304, "mem reserved avg": 15118301659.136, "elapsed time": 132.76655074994778 }, { "step": 750, "valid accuracy": 0.46, "train loss": 0.6789851016998291, "train samples": 3000, "train time": 36.546013267710805, "eval time": 11.214822898968123, "tokens / sec": 5866.60433873995, "mem allocated avg": 6928052195.328, "mem reserved avg": 15356437463.04, "elapsed time": 184.19187769596465 }, { "step": 1000, "valid accuracy": 0.4, "train loss": 0.6588249838352204, "train samples": 4000, "train time": 36.061139613215346, "eval time": 13.717305009020492, "tokens / sec": 5777.299393046663, "mem allocated avg": 6919732948.992, "mem reserved avg": 15453527212.032, "elapsed time": 237.84873885498382 }, { "step": 1250, "valid accuracy": 0.36, "train loss": 0.6543591912984849, "train samples": 5000, "train time": 35.93075305066304, "eval time": 17.338175220997073, "tokens / sec": 5803.886150283504, "mem allocated avg": 6919911497.728, "mem reserved avg": 15469197131.776, "elapsed time": 294.79154046898475 }, { "step": 1500, "valid accuracy": 0.4, "train loss": 0.6470256053209305, "train samples": 6000, "train time": 36.02197089209221, "eval time": 17.308061218995135, "tokens / sec": 5811.203407694546, "mem allocated avg": 6921449252.864, "mem reserved avg": 15339878350.848, "elapsed time": 351.9409700029646 }, { "step": 1750, "valid accuracy": 0.46, "train loss": 0.6376578369140625, "train samples": 7000, "train time": 36.077694381529, "eval time": 12.42961401498178, "tokens / sec": 5802.892994935542, "mem allocated avg": 6922675496.96, "mem reserved avg": 15837507354.624, "elapsed time": 404.2594750869903 }, { "step": 2000, "valid accuracy": 0.38, "train loss": 0.6405653357505798, "train samples": 8000, "train time": 36.009058863681275, "eval time": 11.470333072997164, "tokens / sec": 5767.881931773622, "mem allocated avg": 6919334260.736, "mem reserved avg": 15421482729.472, "elapsed time": 455.6199949949514 }, { "step": 2250, "valid accuracy": 0.46, "train loss": 0.6328598493337632, "train samples": 9000, "train time": 36.205427291512024, "eval time": 17.303009946015663, "tokens / sec": 5936.899964453458, "mem allocated avg": 6930608842.752, "mem reserved avg": 15744427360.256, "elapsed time": 512.8717058099573 }, { "step": 2500, "valid accuracy": 0.42, "train loss": 0.629319528579712, "train samples": 10000, "train time": 34.97963203344261, "eval time": 17.334575710003264, "tokens / sec": 5888.1980177230935, "mem allocated avg": 6915153993.728, "mem reserved avg": 15108914806.784, "elapsed time": 568.9176407279447 }, { "step": 2750, "valid accuracy": 0.34, "train loss": 0.6210904417037963, "train samples": 11000, "train time": 35.76373431546381, "eval time": 14.790576178987976, "tokens / sec": 5924.4652175034535, "mem allocated avg": 6926034331.648, "mem reserved avg": 15580086140.928, "elapsed time": 623.1745745909866 }, { "step": 3000, "valid accuracy": 0.42, "train loss": 0.6132080693244935, "train samples": 12000, "train time": 35.911868128867354, "eval time": 14.77954756503459, "tokens / sec": 5812.312499338176, "mem allocated avg": 6920991168.512, "mem reserved avg": 15530618519.552, "elapsed time": 677.7459287579986 }, { "step": 3250, "valid accuracy": 0.48, "train loss": 0.6221653089523316, "train samples": 13000, "train time": 35.66093143681064, "eval time": 17.444324638985563, "tokens / sec": 5914.063135835525, "mem allocated avg": 6923086110.72, "mem reserved avg": 15357922246.656, "elapsed time": 734.5032512069447 }, { "step": 3500, "valid accuracy": 0.5, "train loss": 0.6054375174045563, "train samples": 14000, "train time": 35.69942569779232, "eval time": 10.496662761026528, "tokens / sec": 5875.444657726555, "mem allocated avg": 6920997902.336, "mem reserved avg": 15300158291.968, "elapsed time": 784.4919563499861 }, { "step": 3750, "valid accuracy": 0.54, "train loss": 0.6029420564174652, "train samples": 15000, "train time": 36.43003757862607, "eval time": 12.152991220995318, "tokens / sec": 5948.47039430841, "mem allocated avg": 6933233727.488, "mem reserved avg": 15711619514.368, "elapsed time": 836.8070459629525 }, { "step": 4000, "valid accuracy": 0.46, "train loss": 0.616380462884903, "train samples": 16000, "train time": 35.82399892428657, "eval time": 10.243308471050113, "tokens / sec": 5704.918661703261, "mem allocated avg": 6913596911.616, "mem reserved avg": 15358928879.616, "elapsed time": 886.7407023639535 }, { "step": 4250, "valid accuracy": 0.5, "train loss": 0.6010915513038635, "train samples": 17000, "train time": 36.03616644133581, "eval time": 17.357441558968276, "tokens / sec": 5866.023522344573, "mem allocated avg": 6924514760.704, "mem reserved avg": 15282240225.28, "elapsed time": 943.718220997951 }, { "step": 4500, "valid accuracy": 0.48, "train loss": 0.6073519963026047, "train samples": 18000, "train time": 36.01844966417411, "eval time": 10.97284383297665, "tokens / sec": 5769.765271343896, "mem allocated avg": 6919068428.288, "mem reserved avg": 15324342648.832, "elapsed time": 994.5949146979838 }, { "step": 4750, "valid accuracy": 0.48, "train loss": 0.59943410551548, "train samples": 19000, "train time": 36.32360130321467, "eval time": 11.35271465400001, "tokens / sec": 5779.685726850554, "mem allocated avg": 6921889050.624, "mem reserved avg": 15411919716.352, "elapsed time": 1045.9306690069498 }, { "step": 5000, "valid accuracy": 0.5, "train loss": 0.6067105796337128, "train samples": 20000, "train time": 35.769867127528414, "eval time": 10.729407122998964, "tokens / sec": 5822.778129352014, "mem allocated avg": 6918440036.352, "mem reserved avg": 15005122560.0, "elapsed time": 1096.1181151779601 }, { "step": 5000, "test accuracy": 0.4799090219863533, "train loss": 0.6067105796337128, "train samples": 20000, "train total tokens": 4198051, "forgetting": 0.42719078063964844 } ] }, "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" } }