{ "run_info": { "created_at": "2025-10-23T20:53:30+00:00", "total_time": 3265.5213168810005, "experiment_name": "waveft/llama-3.2-3B-n_frequency-5000", "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": "WAVEFT", "auto_mapping": null, "peft_version": "0.17.2.dev0@UNKNOWN", "base_model_name_or_path": "meta-llama/Llama-3.2-3B", "revision": null, "inference_mode": false, "n_frequency": 5000, "scaling": 25.0, "wavelet_family": "db1", "use_idwt": true, "random_loc_seed": 777, "fan_in_fan_out": false, "target_modules": [ "q_proj", "v_proj" ], "exclude_modules": null, "bias": "none", "modules_to_save": null, "layers_to_transform": null, "layers_pattern": null, "n_frequency_pattern": {}, "proportional_parameters": false, "init_weights": true }, "error_msg": "" }, "train_info": { "accelerator_memory_reserved_avg": 14582950762, "accelerator_memory_max": 24914165760, "accelerator_memory_reserved_99th": 20564693483, "train_time": 2783.132204494017, "file_size": 1127304, "num_trainable_params": 280000, "num_total_params": 3213029824, "status": "success", "metrics": [ { "step": 250, "valid accuracy": 0.34, "train loss": 1.057445770263672, "train samples": 1000, "train time": 66.99218972800008, "eval time": 29.364740246994188, "tokens / sec": 3160.3534809000257, "mem allocated avg": 6784647395.328, "mem reserved avg": 14636350963.712, "elapsed time": 149.00784370000474 }, { "step": 500, "valid accuracy": 0.34, "train loss": 0.7512865424156189, "train samples": 2000, "train time": 65.21943290103809, "eval time": 29.0264903579955, "tokens / sec": 3189.156831762169, "mem allocated avg": 6776712450.048, "mem reserved avg": 14526627971.072, "elapsed time": 283.5782854230056 }, { "step": 750, "valid accuracy": 0.34, "train loss": 0.6967935096025467, "train samples": 3000, "train time": 66.11754105806904, "eval time": 22.38963912300096, "tokens / sec": 3242.724949672555, "mem allocated avg": 6786416594.944, "mem reserved avg": 14687840239.616, "elapsed time": 412.45540837700537 }, { "step": 1000, "valid accuracy": 0.36, "train loss": 0.6762894586324691, "train samples": 4000, "train time": 65.32631866908923, "eval time": 22.478863092997926, "tokens / sec": 3189.1587379250154, "mem allocated avg": 6778948349.952, "mem reserved avg": 14555082129.408, "elapsed time": 540.8314235460057 }, { "step": 1250, "valid accuracy": 0.38, "train loss": 0.6724052220582962, "train samples": 5000, "train time": 65.4501353219166, "eval time": 29.181654504995095, "tokens / sec": 3186.2118997051034, "mem allocated avg": 6778097270.784, "mem reserved avg": 14548866170.88, "elapsed time": 675.7093939700062 }, { "step": 1500, "valid accuracy": 0.48, "train loss": 0.6683271112442016, "train samples": 6000, "train time": 65.66567935897183, "eval time": 19.425667663999775, "tokens / sec": 3187.8296553616533, "mem allocated avg": 6779162421.248, "mem reserved avg": 14573251854.336, "elapsed time": 801.0151559639999 }, { "step": 1750, "valid accuracy": 0.54, "train loss": 0.6589902213811875, "train samples": 7000, "train time": 65.90766002406599, "eval time": 24.991644790003193, "tokens / sec": 3176.48965118098, "mem allocated avg": 6781186451.456, "mem reserved avg": 14585163677.696, "elapsed time": 932.4019877410028 }, { "step": 2000, "valid accuracy": 0.42, "train loss": 0.6641829339265823, "train samples": 8000, "train time": 65.52569843604579, "eval time": 29.105937477994303, "tokens / sec": 3169.6876944045835, "mem allocated avg": 6777191518.208, "mem reserved avg": 14533355634.688, "elapsed time": 1067.1376723650028 }, { "step": 2250, "valid accuracy": 0.4, "train loss": 0.6568171486854554, "train samples": 9000, "train time": 66.60324803898402, "eval time": 18.03150882799673, "tokens / sec": 3227.290054595945, "mem allocated avg": 6789178621.952, "mem reserved avg": 14712368529.408, "elapsed time": 1192.7977026480003 }, { "step": 2500, "valid accuracy": 0.42, "train loss": 0.6552880892753601, "train samples": 10000, "train time": 64.93178476598405, "eval time": 29.03553620800085, "tokens / sec": 3172.0520349519234, "mem allocated avg": 6774276726.784, "mem reserved avg": 14475549736.96, "elapsed time": 1327.1052960740053 }, { "step": 2750, "valid accuracy": 0.42, "train loss": 0.6487538056373596, "train samples": 11000, "train time": 66.10884880107187, "eval time": 20.587333617004333, "tokens / sec": 3205.0323646925253, "mem allocated avg": 6784980387.84, "mem reserved avg": 14651936997.376, "elapsed time": 1454.7400554460037 }, { "step": 3000, "valid accuracy": 0.4, "train loss": 0.6414109219312668, "train samples": 12000, "train time": 65.51606800403533, "eval time": 17.688484279002296, "tokens / sec": 3185.951269040499, "mem allocated avg": 6780026255.36, "mem reserved avg": 14562925477.888, "elapsed time": 1578.3467425210038 }, { "step": 3250, "valid accuracy": 0.4, "train loss": 0.6511869001388549, "train samples": 13000, "train time": 65.77464669098845, "eval time": 22.121913303002657, "tokens / sec": 3206.4178313388766, "mem allocated avg": 6781515575.296, "mem reserved avg": 14588720447.488, "elapsed time": 1706.5400248380029 }, { "step": 3500, "valid accuracy": 0.46, "train loss": 0.637642817735672, "train samples": 14000, "train time": 65.76092355793662, "eval time": 29.041672920000565, "tokens / sec": 3189.584158063204, "mem allocated avg": 6779834134.528, "mem reserved avg": 14574015217.664, "elapsed time": 1842.0509184040056 }, { "step": 3750, "valid accuracy": 0.42, "train loss": 0.6350828701257706, "train samples": 15000, "train time": 66.55924862711254, "eval time": 20.789683652998065, "tokens / sec": 3255.790960232193, "mem allocated avg": 6791231805.44, "mem reserved avg": 14752080199.68, "elapsed time": 1970.2113445850046 }, { "step": 4000, "valid accuracy": 0.38, "train loss": 0.65046697640419, "train samples": 16000, "train time": 65.04778776894818, "eval time": 19.624021877003543, "tokens / sec": 3141.890093571505, "mem allocated avg": 6772911845.376, "mem reserved avg": 14460534128.64, "elapsed time": 2095.6094995790045 }, { "step": 4250, "valid accuracy": 0.42, "train loss": 0.6331748945713043, "train samples": 17000, "train time": 65.85189565200562, "eval time": 23.701296111001284, "tokens / sec": 3210.067043735313, "mem allocated avg": 6782308450.304, "mem reserved avg": 14607057944.576, "elapsed time": 2225.6684009890014 }, { "step": 4500, "valid accuracy": 0.4, "train loss": 0.641278461933136, "train samples": 18000, "train time": 65.11867782095214, "eval time": 23.630613847002678, "tokens / sec": 3191.3731505944966, "mem allocated avg": 6778411657.216, "mem reserved avg": 14525831053.312, "elapsed time": 2354.2676229330027 }, { "step": 4750, "valid accuracy": 0.4, "train loss": 0.6345745379924774, "train samples": 19000, "train time": 65.44978067000193, "eval time": 23.75194463099615, "tokens / sec": 3207.634889695251, "mem allocated avg": 6780527521.792, "mem reserved avg": 14582739369.984, "elapsed time": 2484.0077965070013 }, { "step": 5000, "valid accuracy": 0.44, "train loss": 0.6398445825576782, "train samples": 20000, "train time": 65.40377733100468, "eval time": 19.663959343997703, "tokens / sec": 3184.5255503502062, "mem allocated avg": 6777134090.24, "mem reserved avg": 14518717513.728, "elapsed time": 2609.523235476001 }, { "step": 5000, "test accuracy": 0.4162244124336619, "train loss": 0.6398445825576782, "train samples": 20000, "train total tokens": 4198051 } ] }, "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": "e53f048856ff4f594e959d75785d2c2d37b678ee", "created_at": "2022-04-12T10:22:10+00:00" } }, "package_info": { "transformers-version": "4.57.1", "transformers-commit-hash": null, "peft-version": "0.17.2.dev0", "peft-commit-hash": "a18ba67f242ab2eb74cdabab76ea2fd836b5cd83", "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": "6.14.0-1014-aws", "version": "#14~24.04.1-Ubuntu SMP Tue Sep 23 14:51:14 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" } }