{ "run_info": { "created_at": "2026-01-10T03:01:44+00:00", "total_time": 1173.714544863964, "experiment_name": "lora/llama-3.2-3B-rank32", "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": [ "v_proj", "q_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": false, "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": 14427567318, "accelerator_memory_max": 22286434304, "accelerator_memory_reserved_99th": 20122173440, "train_time": 958.3276851720293, "file_size": 36715216, "num_trainable_params": 9175040, "num_total_params": 3221924864, "status": "success", "metrics": [ { "step": 250, "valid accuracy": 0.34, "train loss": 0.9826480431556701, "train samples": 1000, "train time": 29.567459934856743, "eval time": 12.036156770016532, "tokens / sec": 7160.5406912349235, "mem allocated avg": 6927427098.624, "mem reserved avg": 14540435619.84, "elapsed time": 63.864961831015535 }, { "step": 500, "valid accuracy": 0.46, "train loss": 0.7163139179944992, "train samples": 2000, "train time": 29.46686314494582, "eval time": 12.008591033983976, "tokens / sec": 7058.606780670357, "mem allocated avg": 6920238368.768, "mem reserved avg": 14166437920.768, "elapsed time": 108.54651641799137 }, { "step": 750, "valid accuracy": 0.36, "train loss": 0.679160294175148, "train samples": 3000, "train time": 30.308068415150046, "eval time": 9.108271537988912, "tokens / sec": 7074.056883573211, "mem allocated avg": 6930467016.704, "mem reserved avg": 14381723156.48, "elapsed time": 151.1071816039621 }, { "step": 1000, "valid accuracy": 0.42, "train loss": 0.6593781963586808, "train samples": 4000, "train time": 30.217666860786267, "eval time": 12.128453835961409, "tokens / sec": 6894.50978991232, "mem allocated avg": 6921353275.392, "mem reserved avg": 14449670881.28, "elapsed time": 196.77825986297103 }, { "step": 1250, "valid accuracy": 0.42, "train loss": 0.6545941867828369, "train samples": 5000, "train time": 29.99424036074197, "eval time": 11.980647846998181, "tokens / sec": 6952.601482548144, "mem allocated avg": 6922116595.712, "mem reserved avg": 14466657812.48, "elapsed time": 241.93410539499018 }, { "step": 1500, "valid accuracy": 0.38, "train loss": 0.647283132314682, "train samples": 6000, "train time": 30.01751733769197, "eval time": 11.991809643048327, "tokens / sec": 6973.628020101123, "mem allocated avg": 6923839727.616, "mem reserved avg": 14383426043.904, "elapsed time": 287.2019039680017 }, { "step": 1750, "valid accuracy": 0.42, "train loss": 0.6378610955476761, "train samples": 7000, "train time": 30.147432959987782, "eval time": 8.299405578000005, "tokens / sec": 6944.372354285015, "mem allocated avg": 6925383284.736, "mem reserved avg": 14772699398.144, "elapsed time": 328.93911194999237 }, { "step": 2000, "valid accuracy": 0.36, "train loss": 0.6409116793870926, "train samples": 8000, "train time": 30.117346432991326, "eval time": 11.998839316016529, "tokens / sec": 6896.225086167764, "mem allocated avg": 6921043718.144, "mem reserved avg": 14419597721.6, "elapsed time": 374.35504649899667 }, { "step": 2250, "valid accuracy": 0.42, "train loss": 0.633124389886856, "train samples": 9000, "train time": 30.76248943991959, "eval time": 11.96315730799688, "tokens / sec": 6987.340878890907, "mem allocated avg": 6931688802.304, "mem reserved avg": 14696019132.416, "elapsed time": 420.3003967599943 }, { "step": 2500, "valid accuracy": 0.42, "train loss": 0.6295496643781662, "train samples": 10000, "train time": 29.52961014426546, "eval time": 12.078024981019553, "tokens / sec": 6974.931229831966, "mem allocated avg": 6918595411.968, "mem reserved avg": 14198885056.512, "elapsed time": 465.105449871975 }, { "step": 2750, "valid accuracy": 0.4, "train loss": 0.6213833647966385, "train samples": 11000, "train time": 30.2069385854993, "eval time": 8.256086013978347, "tokens / sec": 7014.3155818416, "mem allocated avg": 6927818975.232, "mem reserved avg": 14563151970.304, "elapsed time": 506.76149754499784 }, { "step": 3000, "valid accuracy": 0.48, "train loss": 0.6133703490495682, "train samples": 12000, "train time": 30.294824214477558, "eval time": 12.005937494977843, "tokens / sec": 6889.988815325418, "mem allocated avg": 6924054609.92, "mem reserved avg": 14496630308.864, "elapsed time": 552.3840983010014 }, { "step": 3250, "valid accuracy": 0.44, "train loss": 0.6224808418750762, "train samples": 13000, "train time": 29.94148239895003, "eval time": 7.373289940995164, "tokens / sec": 7043.772822931297, "mem allocated avg": 6926655715.328, "mem reserved avg": 14371212230.656, "elapsed time": 592.8639876859961 }, { "step": 3500, "valid accuracy": 0.5, "train loss": 0.6059285970926285, "train samples": 14000, "train time": 30.129387284861878, "eval time": 11.973128407029435, "tokens / sec": 6961.6417359202715, "mem allocated avg": 6923676155.904, "mem reserved avg": 14364719448.064, "elapsed time": 638.2131494429777 }, { "step": 3750, "valid accuracy": 0.5, "train loss": 0.6033647983074188, "train samples": 15000, "train time": 30.893273340305313, "eval time": 7.364776138972957, "tokens / sec": 7014.569081524799, "mem allocated avg": 6934723543.04, "mem reserved avg": 14704281911.296, "elapsed time": 679.7110908210161 }, { "step": 4000, "valid accuracy": 0.42, "train loss": 0.6165854910612106, "train samples": 16000, "train time": 29.51253367715981, "eval time": 12.00497411499964, "tokens / sec": 6924.956096133736, "mem allocated avg": 6915841265.664, "mem reserved avg": 14385749688.32, "elapsed time": 724.5038471769658 }, { "step": 4250, "valid accuracy": 0.48, "train loss": 0.601499475479126, "train samples": 17000, "train time": 29.835361877689138, "eval time": 8.425500162993558, "tokens / sec": 7085.183041070353, "mem allocated avg": 6927651903.488, "mem reserved avg": 14371388391.424, "elapsed time": 765.8859701670008 }, { "step": 4500, "valid accuracy": 0.48, "train loss": 0.6076976944208146, "train samples": 18000, "train time": 29.980451503419317, "eval time": 11.977038996003103, "tokens / sec": 6931.7835315554885, "mem allocated avg": 6921458522.112, "mem reserved avg": 14316107464.704, "elapsed time": 811.13678931701 }, { "step": 4750, "valid accuracy": 0.54, "train loss": 0.5996475745439529, "train samples": 19000, "train time": 30.189162221679, "eval time": 7.3034177150111645, "tokens / sec": 6954.11811889373, "mem allocated avg": 6924435124.224, "mem reserved avg": 14412157026.304, "elapsed time": 851.8043771950179 }, { "step": 5000, "valid accuracy": 0.5, "train loss": 0.6068337166309357, "train samples": 20000, "train time": 29.628619944676757, "eval time": 12.001391948957462, "tokens / sec": 7029.689549796961, "mem allocated avg": 6920384235.52, "mem reserved avg": 14090395189.248, "elapsed time": 896.6036795430118 }, { "step": 5000, "test accuracy": 0.49052312357846856, "train loss": 0.6068337166309357, "train samples": 20000, "train total tokens": 4198051, "forgetting": 0.4156990051269531 } ] }, "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" } }