{ "run_info": { "created_at": "2026-01-09T21:25:18+00:00", "total_time": 1366.5376071679639, "experiment_name": "c3a/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": 0.3, "weight_decay": 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": { "task_type": null, "peft_type": "C3A", "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, "block_size": 64, "target_modules": [ "v_proj", "q_proj" ], "bias": "none", "modules_to_save": null, "layers_to_transform": null, "layers_pattern": null, "block_size_pattern": {}, "init_weights": false }, "error_msg": "" }, "train_info": { "accelerator_memory_reserved_avg": 14377058251, "accelerator_memory_max": 22166896640, "accelerator_memory_reserved_99th": 20061356032, "train_time": 1153.291345822101, "file_size": 22027512, "num_trainable_params": 5505024, "num_total_params": 3218254848, "status": "success", "metrics": [ { "step": 250, "valid accuracy": 0.38, "train loss": 0.8909596583843231, "train samples": 1000, "train time": 37.80949870531913, "eval time": 11.432839135988615, "tokens / sec": 5599.624624756394, "mem allocated avg": 6869184182.272, "mem reserved avg": 14482872991.744, "elapsed time": 72.47078875597799 }, { "step": 500, "valid accuracy": 0.42, "train loss": 0.7013351047039031, "train samples": 2000, "train time": 37.42838709271746, "eval time": 8.331702384981327, "tokens / sec": 5557.1456895739475, "mem allocated avg": 6862181789.696, "mem reserved avg": 14091510874.112, "elapsed time": 121.3354702090146 }, { "step": 750, "valid accuracy": 0.46, "train loss": 0.674084732890129, "train samples": 3000, "train time": 38.329594842216466, "eval time": 9.41572032001568, "tokens / sec": 5593.615087312568, "mem allocated avg": 6872815554.56, "mem reserved avg": 14353973641.216, "elapsed time": 172.14090229501016 }, { "step": 1000, "valid accuracy": 0.44, "train loss": 0.6545811468362808, "train samples": 4000, "train time": 37.05278595397249, "eval time": 13.860594167024828, "tokens / sec": 5622.681119276645, "mem allocated avg": 6864588425.216, "mem reserved avg": 14391965646.848, "elapsed time": 226.19000285398215 }, { "step": 1250, "valid accuracy": 0.38, "train loss": 0.6509714391231537, "train samples": 5000, "train time": 36.79460979660507, "eval time": 8.696108918986283, "tokens / sec": 5667.62363163425, "mem allocated avg": 6863943450.624, "mem reserved avg": 14405840404.48, "elapsed time": 274.7266224689665 }, { "step": 1500, "valid accuracy": 0.52, "train loss": 0.6443188463449478, "train samples": 6000, "train time": 37.157257698476315, "eval time": 9.571486864006147, "tokens / sec": 5633.650408183485, "mem allocated avg": 6865824692.224, "mem reserved avg": 14339352297.472, "elapsed time": 324.5449085899745 }, { "step": 1750, "valid accuracy": 0.46, "train loss": 0.6351410290002822, "train samples": 7000, "train time": 37.588308651640546, "eval time": 13.673957422957756, "tokens / sec": 5569.6839658376775, "mem allocated avg": 6866604124.16, "mem reserved avg": 14722149646.336, "elapsed time": 378.88215018698247 }, { "step": 2000, "valid accuracy": 0.42, "train loss": 0.6367183820009231, "train samples": 8000, "train time": 37.49864329502452, "eval time": 9.148618383973371, "tokens / sec": 5538.760385700621, "mem allocated avg": 6862256517.12, "mem reserved avg": 14371757490.176, "elapsed time": 428.6968064419925 }, { "step": 2250, "valid accuracy": 0.4, "train loss": 0.6258963789939881, "train samples": 9000, "train time": 38.2491970359697, "eval time": 13.646375201002229, "tokens / sec": 5619.673526685071, "mem allocated avg": 6874371139.584, "mem reserved avg": 14650997473.28, "elapsed time": 483.6756669579772 }, { "step": 2500, "valid accuracy": 0.48, "train loss": 0.6206584950685501, "train samples": 10000, "train time": 37.62900273769628, "eval time": 13.673765863000881, "tokens / sec": 5473.623668311165, "mem allocated avg": 6859448377.344, "mem reserved avg": 14140047360.0, "elapsed time": 538.0801370120025 }, { "step": 2750, "valid accuracy": 0.5, "train loss": 0.6105827242136002, "train samples": 11000, "train time": 37.365518155740574, "eval time": 9.94396861596033, "tokens / sec": 5670.495431560023, "mem allocated avg": 6870005147.648, "mem reserved avg": 14540318179.328, "elapsed time": 588.4858504809672 }, { "step": 3000, "valid accuracy": 0.48, "train loss": 0.5981686524152756, "train samples": 12000, "train time": 36.890319129568525, "eval time": 9.116078838997055, "tokens / sec": 5658.151106442905, "mem allocated avg": 6865474818.048, "mem reserved avg": 14431375327.232, "elapsed time": 637.5967296099989 }, { "step": 3250, "valid accuracy": 0.58, "train loss": 0.6065964109897614, "train samples": 13000, "train time": 36.91873880266212, "eval time": 13.868325658957474, "tokens / sec": 5712.573257914012, "mem allocated avg": 6867266957.312, "mem reserved avg": 14327809572.864, "elapsed time": 691.4201774260146 }, { "step": 3500, "valid accuracy": 0.54, "train loss": 0.5876260681152343, "train samples": 14000, "train time": 37.31881154741859, "eval time": 13.723029766988475, "tokens / sec": 5620.4898093682405, "mem allocated avg": 6864456710.144, "mem reserved avg": 14320821862.4, "elapsed time": 745.5659840449807 }, { "step": 3750, "valid accuracy": 0.52, "train loss": 0.5839207136631012, "train samples": 15000, "train time": 38.572791439597495, "eval time": 8.601323624025099, "tokens / sec": 5618.027420684421, "mem allocated avg": 6877443866.624, "mem reserved avg": 14659335749.632, "elapsed time": 795.8486898440169 }, { "step": 4000, "valid accuracy": 0.56, "train loss": 0.5916516542434692, "train samples": 16000, "train time": 36.93544478120748, "eval time": 8.182845938019454, "tokens / sec": 5533.248650737345, "mem allocated avg": 6857094844.416, "mem reserved avg": 14323103563.776, "elapsed time": 844.0624020369723 }, { "step": 4250, "valid accuracy": 0.56, "train loss": 0.5781249686479568, "train samples": 17000, "train time": 37.122223554353695, "eval time": 9.616572241007816, "tokens / sec": 5694.405662163205, "mem allocated avg": 6869011593.216, "mem reserved avg": 14312148041.728, "elapsed time": 893.7695867909933 }, { "step": 4500, "valid accuracy": 0.56, "train loss": 0.5846204442977905, "train samples": 18000, "train time": 37.07799867092399, "eval time": 8.460158887959551, "tokens / sec": 5604.887195893012, "mem allocated avg": 6863795603.456, "mem reserved avg": 14295052058.624, "elapsed time": 942.4601713569718 }, { "step": 4750, "valid accuracy": 0.52, "train loss": 0.5748073060512543, "train samples": 19000, "train time": 37.64343991095666, "eval time": 8.636093506007455, "tokens / sec": 5577.040793737192, "mem allocated avg": 6865605898.24, "mem reserved avg": 14353235443.712, "elapsed time": 991.785048523976 }, { "step": 5000, "valid accuracy": 0.5, "train loss": 0.5811544748544692, "train samples": 20000, "train time": 37.4456382485223, "eval time": 8.431857075018343, "tokens / sec": 5562.1965532452705, "mem allocated avg": 6862668220.416, "mem reserved avg": 14027497406.464, "elapsed time": 1040.7068556279992 }, { "step": 5000, "test accuracy": 0.5026535253980288, "train loss": 0.5811544748544692, "train samples": 20000, "train total tokens": 4198051, "forgetting": 0.4295167922973633 } ] }, "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" } }