{ "run_info": { "created_at": "2026-04-14T19:04:28+00:00", "total_time": 1280.5890959829994, "experiment_name": "tinylora/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": 0.001 }, "lr_scheduler": "cosine", "use_amp": false, "autocast_adapter_dtype": true, "generation_kwargs": { "max_length": 800, "max_new_tokens": 300 }, "attn_implementation": null, "init_kv_cache_prefix": null }, "peft_config": { "task_type": null, "peft_type": "TINYLORA", "auto_mapping": null, "peft_version": "0.19.0", "base_model_name_or_path": "meta-llama/Llama-3.2-3B", "revision": null, "inference_mode": false, "r": 2, "u": 64, "weight_tying": 0.0, "projection_seed": 42, "save_projection": true, "init_v_bound": 0.02, "target_modules": [ "v_proj", "q_proj" ], "tinylora_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": 13315370437, "accelerator_memory_max": 20126367744, "accelerator_memory_reserved_99th": 18343788544, "train_time": 1005.6939271469892, "file_size": 2391912, "num_trainable_params": 3584, "num_total_params": 3212753408, "status": "success", "metrics": [ { "step": 250, "valid accuracy": 0.0, "train loss": 1.2781277766227723, "train samples": 1000, "train time": 29.62463812986971, "eval time": 12.268791488997522, "tokens / sec": 7146.720208761961, "mem allocated avg": 6782683836.416, "mem reserved avg": 13404014116.864, "elapsed time": 63.092878730007214 }, { "step": 500, "valid accuracy": 0.0, "train loss": 1.2406480107307434, "train samples": 2000, "train time": 29.568441232899204, "eval time": 12.359398319997126, "tokens / sec": 7034.357961642402, "mem allocated avg": 6775698290.688, "mem reserved avg": 13079584702.464, "elapsed time": 107.82620730499912 }, { "step": 750, "valid accuracy": 0.02, "train loss": 1.1951167840957642, "train samples": 3000, "train time": 29.82903349219123, "eval time": 12.353063451009803, "tokens / sec": 7187.661647036831, "mem allocated avg": 6785483120.64, "mem reserved avg": 13268521320.448, "elapsed time": 152.80012364900904 }, { "step": 1000, "valid accuracy": 0.02, "train loss": 1.196952908039093, "train samples": 4000, "train time": 29.21191224195354, "eval time": 12.419155650000903, "tokens / sec": 7131.885043143193, "mem allocated avg": 6776640491.52, "mem reserved avg": 13314843213.824, "elapsed time": 197.2873525449977 }, { "step": 1250, "valid accuracy": 0.0, "train loss": 1.1898907599449158, "train samples": 5000, "train time": 29.00116592599079, "eval time": 12.263450048994855, "tokens / sec": 7190.676420809297, "mem allocated avg": 6777077430.272, "mem reserved avg": 13365250359.296, "elapsed time": 241.3598588379973 }, { "step": 1500, "valid accuracy": 0.0, "train loss": 1.1793672075271606, "train samples": 6000, "train time": 29.21511149003345, "eval time": 12.238495423996937, "tokens / sec": 7165.1617715514085, "mem allocated avg": 6779007174.656, "mem reserved avg": 13277757177.856, "elapsed time": 285.6396420480014 }, { "step": 1750, "valid accuracy": 0.02, "train loss": 1.1763010482788085, "train samples": 7000, "train time": 29.251842216108344, "eval time": 12.174482814996736, "tokens / sec": 7156.98513800655, "mem allocated avg": 6780389603.328, "mem reserved avg": 13619811057.664, "elapsed time": 329.89888020399667 }, { "step": 2000, "valid accuracy": 0.02, "train loss": 1.1813358347415923, "train samples": 8000, "train time": 29.418175176848308, "eval time": 12.335072386005777, "tokens / sec": 7060.125203260529, "mem allocated avg": 6776134410.24, "mem reserved avg": 13301312389.12, "elapsed time": 374.5338296889968 }, { "step": 2250, "valid accuracy": 0.0, "train loss": 1.172850294113159, "train samples": 9000, "train time": 29.904061377135804, "eval time": 12.285082250004052, "tokens / sec": 7187.919971443946, "mem allocated avg": 6786565449.728, "mem reserved avg": 13564848898.048, "elapsed time": 419.51835140399635 }, { "step": 2500, "valid accuracy": 0.02, "train loss": 1.1793990023136138, "train samples": 10000, "train time": 29.20591944696207, "eval time": 12.41826296299405, "tokens / sec": 7052.234748987647, "mem allocated avg": 6773510297.6, "mem reserved avg": 13106612797.44, "elapsed time": 463.9779940340086 }, { "step": 2750, "valid accuracy": 0.0, "train loss": 1.180105079650879, "train samples": 11000, "train time": 29.320615836040815, "eval time": 12.377129669010174, "tokens / sec": 7226.348900201356, "mem allocated avg": 6782631518.208, "mem reserved avg": 13437618880.512, "elapsed time": 508.437297217999 }, { "step": 3000, "valid accuracy": 0.0, "train loss": 1.1744389948844909, "train samples": 12000, "train time": 29.53985286586976, "eval time": 12.229440000999602, "tokens / sec": 7066.081234316744, "mem allocated avg": 6778503475.2, "mem reserved avg": 13380526014.464, "elapsed time": 553.0742953360023 }, { "step": 3250, "valid accuracy": 0.02, "train loss": 1.172437875032425, "train samples": 13000, "train time": 29.548683459026506, "eval time": 12.297136758003035, "tokens / sec": 7137.40767139235, "mem allocated avg": 6781349326.848, "mem reserved avg": 13244496347.136, "elapsed time": 597.6882013260038 }, { "step": 3500, "valid accuracy": 0.0, "train loss": 1.1740667581558228, "train samples": 14000, "train time": 29.95824750296015, "eval time": 12.355832200992154, "tokens / sec": 7001.410879567464, "mem allocated avg": 6778354403.328, "mem reserved avg": 13265174265.856, "elapsed time": 642.876650516002 }, { "step": 3750, "valid accuracy": 0.0, "train loss": 1.1586440045833588, "train samples": 15000, "train time": 30.502618752914714, "eval time": 12.421693063995917, "tokens / sec": 7104.40640377124, "mem allocated avg": 6789483493.376, "mem reserved avg": 13573707268.096, "elapsed time": 688.6765086410014 }, { "step": 4000, "valid accuracy": 0.06, "train loss": 1.1851642508506774, "train samples": 16000, "train time": 29.493367298011435, "eval time": 12.399896686998545, "tokens / sec": 6929.456305715885, "mem allocated avg": 6770658635.776, "mem reserved avg": 13280558972.928, "elapsed time": 733.4192178939993 }, { "step": 4250, "valid accuracy": 0.0, "train loss": 1.157331362247467, "train samples": 17000, "train time": 29.721664497148595, "eval time": 12.352380377997179, "tokens / sec": 7112.286730114998, "mem allocated avg": 6782410606.592, "mem reserved avg": 13262255030.272, "elapsed time": 778.2037502510066 }, { "step": 4500, "valid accuracy": 0.0, "train loss": 1.1738232126235961, "train samples": 18000, "train time": 29.415511626095395, "eval time": 12.387085970010958, "tokens / sec": 7064.911963511059, "mem allocated avg": 6776753987.584, "mem reserved avg": 13233884758.016, "elapsed time": 822.8698514100106 }, { "step": 4750, "valid accuracy": 0.0, "train loss": 1.1605148227214814, "train samples": 19000, "train time": 30.141289916908136, "eval time": 12.342347552010324, "tokens / sec": 6965.163089527634, "mem allocated avg": 6779153461.248, "mem reserved avg": 13314037907.456, "elapsed time": 868.1438604020077 }, { "step": 5000, "valid accuracy": 0.02, "train loss": 1.1730116956233978, "train samples": 20000, "train time": 29.35209196503274, "eval time": 12.416217063000659, "tokens / sec": 7095.91671517399, "mem allocated avg": 6776144590.848, "mem reserved avg": 13012593278.976, "elapsed time": 912.6687434610067 }, { "step": 5000, "test accuracy": 0.009855951478392721, "train loss": 1.1730116956233978, "train samples": 20000, "train total tokens": 4198051, "forgetting": -0.7508955001831055 } ] }, "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": "740312add88f781978c0658806c59bc2815b9866", "created_at": "2022-04-12T10:22:10+00:00" } }, "package_info": { "transformers-version": "5.5.4", "transformers-commit-hash": null, "peft-version": "0.19.0", "peft-commit-hash": "6d5a6f4f2f902dbf13d21d2661d57c3c05df1dae", "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": "6.17.0-1009-aws", "version": "#9~24.04.2-Ubuntu SMP Fri Mar 6 23:50:29 UTC 2026", "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" } }