{ "run_info": { "created_at": "2026-01-09T17:49:47+00:00", "total_time": 10059.440659368003, "experiment_name": "boft/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": "BOFT", "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, "boft_block_size": 4, "boft_block_num": 0, "boft_n_butterfly_factor": 1, "target_modules": [ "v_proj", "q_proj" ], "exclude_modules": null, "boft_dropout": 0.0, "fan_in_fan_out": false, "bias": "none", "modules_to_save": null, "init_weights": true, "layers_to_transform": null, "layers_pattern": null }, "error_msg": "" }, "train_info": { "accelerator_memory_reserved_avg": 17188165936, "accelerator_memory_max": 24410849280, "accelerator_memory_reserved_99th": 22345154560, "train_time": 7190.453845723998, "file_size": 3225360, "num_trainable_params": 802816, "num_total_params": 3213552640, "status": "success", "metrics": [ { "step": 250, "valid accuracy": 0.0, "train loss": 1.2914534993171691, "train samples": 1000, "train time": 168.16101133066695, "eval time": 142.30593162897276, "tokens / sec": 1259.025491846513, "mem allocated avg": 6795307169.792, "mem reserved avg": 17261339344.896, "elapsed time": 344.37489816197194 }, { "step": 500, "valid accuracy": 0.12, "train loss": 1.065816521883011, "train samples": 2000, "train time": 169.6081039829878, "eval time": 143.82841559901135, "tokens / sec": 1226.3270157236266, "mem allocated avg": 6787530438.656, "mem reserved avg": 16960280592.384, "elapsed time": 662.5959406199981 }, { "step": 750, "valid accuracy": 0.42, "train loss": 0.8761289038658142, "train samples": 3000, "train time": 170.50346154812723, "eval time": 143.55063587299082, "tokens / sec": 1257.4583416271698, "mem allocated avg": 6797964732.416, "mem reserved avg": 17119546703.872, "elapsed time": 981.3382467220072 }, { "step": 1000, "valid accuracy": 0.44, "train loss": 0.8187025270462036, "train samples": 4000, "train time": 170.603982186527, "eval time": 143.48508089897223, "tokens / sec": 1221.1672748190563, "mem allocated avg": 6789360863.232, "mem reserved avg": 17173468676.096, "elapsed time": 1300.2320007149829 }, { "step": 1250, "valid accuracy": 0.38, "train loss": 0.7968714184761048, "train samples": 5000, "train time": 170.08946738566738, "eval time": 143.52085828600684, "tokens / sec": 1226.0488741913275, "mem allocated avg": 6788928741.376, "mem reserved avg": 17237641527.296, "elapsed time": 1618.561195612012 }, { "step": 1500, "valid accuracy": 0.42, "train loss": 0.7768479048013687, "train samples": 6000, "train time": 171.03791188867763, "eval time": 142.4387985090143, "tokens / sec": 1223.8865505809376, "mem allocated avg": 6790983270.4, "mem reserved avg": 17204464582.656, "elapsed time": 1936.8438452039845 }, { "step": 1750, "valid accuracy": 0.34, "train loss": 0.7638826340436935, "train samples": 7000, "train time": 169.5759706167737, "eval time": 142.81106396700488, "tokens / sec": 1234.5793996551745, "mem allocated avg": 6792170039.296, "mem reserved avg": 17521872732.16, "elapsed time": 2253.9779722079984 }, { "step": 2000, "valid accuracy": 0.3, "train loss": 0.7574314765930176, "train samples": 8000, "train time": 169.80084089771844, "eval time": 142.94605548703112, "tokens / sec": 1223.1741545090943, "mem allocated avg": 6788512305.152, "mem reserved avg": 17193743941.632, "elapsed time": 2571.564919157012 }, { "step": 2250, "valid accuracy": 0.4, "train loss": 0.747995015501976, "train samples": 9000, "train time": 169.89184381318046, "eval time": 142.61838847299805, "tokens / sec": 1265.2049396577568, "mem allocated avg": 6799572060.16, "mem reserved avg": 17437013573.632, "elapsed time": 2888.81089515402 }, { "step": 2500, "valid accuracy": 0.3, "train loss": 0.7451818664073944, "train samples": 10000, "train time": 169.05202275590273, "eval time": 142.17101406800793, "tokens / sec": 1218.3645995020095, "mem allocated avg": 6785213417.472, "mem reserved avg": 17006149500.928, "elapsed time": 3204.817744602973 }, { "step": 2750, "valid accuracy": 0.26, "train loss": 0.7368416948318481, "train samples": 11000, "train time": 169.79861720540794, "eval time": 142.90656082698843, "tokens / sec": 1247.837017092338, "mem allocated avg": 6795389593.6, "mem reserved avg": 17333305212.928, "elapsed time": 3522.2434823570075 }, { "step": 3000, "valid accuracy": 0.42, "train loss": 0.7283624488115311, "train samples": 12000, "train time": 169.40914075018372, "eval time": 142.4628674259875, "tokens / sec": 1232.1117920537804, "mem allocated avg": 6790662739.968, "mem reserved avg": 17229924007.936, "elapsed time": 3838.9266921749804 }, { "step": 3250, "valid accuracy": 0.4, "train loss": 0.7360379147529602, "train samples": 13000, "train time": 169.2894103731378, "eval time": 142.70766545902006, "tokens / sec": 1245.8014918661738, "mem allocated avg": 6792828252.16, "mem reserved avg": 17116644245.504, "elapsed time": 4155.603474072996 }, { "step": 3500, "valid accuracy": 0.38, "train loss": 0.7245079389810563, "train samples": 14000, "train time": 169.62816236459184, "eval time": 142.80769011599477, "tokens / sec": 1236.528162989657, "mem allocated avg": 6790204116.992, "mem reserved avg": 17182729699.328, "elapsed time": 4472.8496758749825 }, { "step": 3750, "valid accuracy": 0.32, "train loss": 0.7198032699823379, "train samples": 15000, "train time": 169.99743940570625, "eval time": 143.21545482298825, "tokens / sec": 1274.742729993885, "mem allocated avg": 6802659770.368, "mem reserved avg": 17454201831.424, "elapsed time": 4790.820539581007 }, { "step": 4000, "valid accuracy": 0.36, "train loss": 0.7386080088615418, "train samples": 16000, "train time": 169.4405590199167, "eval time": 142.85203308903147, "tokens / sec": 1206.1633954829977, "mem allocated avg": 6782729986.048, "mem reserved avg": 17107366445.056, "elapsed time": 5107.90697863797 }, { "step": 4250, "valid accuracy": 0.4, "train loss": 0.7167673094272613, "train samples": 17000, "train time": 169.4217613734072, "eval time": 142.6639407540206, "tokens / sec": 1247.7086667402577, "mem allocated avg": 6793673689.088, "mem reserved avg": 17128346353.664, "elapsed time": 5424.600868158974 }, { "step": 4500, "valid accuracy": 0.36, "train loss": 0.7279775400161743, "train samples": 18000, "train time": 169.55897833802737, "eval time": 142.51284795097308, "tokens / sec": 1225.6384299845254, "mem allocated avg": 6789383653.376, "mem reserved avg": 17117466329.088, "elapsed time": 5741.48882275197 }, { "step": 4750, "valid accuracy": 0.4, "train loss": 0.7207833558321, "train samples": 19000, "train time": 169.6458221563953, "eval time": 142.71257099701324, "tokens / sec": 1237.513528664789, "mem allocated avg": 6791604719.616, "mem reserved avg": 17139536756.736, "elapsed time": 6058.54882254597 }, { "step": 5000, "valid accuracy": 0.36, "train loss": 0.7267992464303971, "train samples": 20000, "train time": 169.51394822495058, "eval time": 142.66391144797672, "tokens / sec": 1228.6894511099795, "mem allocated avg": 6788474263.552, "mem reserved avg": 16838276677.632, "elapsed time": 6375.395246869011 }, { "step": 5000, "test accuracy": 0.36239575435936316, "train loss": 0.7267992464303971, "train samples": 20000, "train total tokens": 4198051, "forgetting": 0.3084230422973633 } ] }, "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" } }