{ "run_info": { "created_at": "2026-01-10T06:43:22+00:00", "total_time": 1838.2831721020048, "experiment_name": "prompt_tuning/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": "CAUSAL_LM", "peft_type": "PROMPT_TUNING", "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, "num_virtual_tokens": 200, "token_dim": 3072, "num_transformer_submodules": 1, "num_attention_heads": 24, "num_layers": 28, "modules_to_save": null, "prompt_tuning_init": "RANDOM", "prompt_tuning_init_text": null, "tokenizer_name_or_path": null, "tokenizer_kwargs": null }, "error_msg": "" }, "train_info": { "accelerator_memory_reserved_avg": 17625765642, "accelerator_memory_max": 24427626496, "accelerator_memory_reserved_99th": 22968008704, "train_time": 1505.838819740049, "file_size": 2457728, "num_trainable_params": 614400, "num_total_params": 3213364224, "status": "success", "metrics": [ { "step": 250, "valid accuracy": 0.0, "train loss": 3.462650680541992, "train samples": 1000, "train time": 48.139625351293944, "eval time": 16.530633710033726, "tokens / sec": 4398.019271130643, "mem allocated avg": 7084412407.808, "mem reserved avg": 17769940647.936, "elapsed time": 87.74912454502191 }, { "step": 500, "valid accuracy": 0.0, "train loss": 2.2589459805488588, "train samples": 2000, "train time": 48.43768710020231, "eval time": 16.48010197596159, "tokens / sec": 4294.073735802535, "mem allocated avg": 7076015853.568, "mem reserved avg": 17322089644.032, "elapsed time": 155.9924284190056 }, { "step": 750, "valid accuracy": 0.0, "train loss": 1.7582932782173157, "train samples": 3000, "train time": 48.9170593819581, "eval time": 16.461432596028317, "tokens / sec": 4382.94948038264, "mem allocated avg": 7086598868.992, "mem reserved avg": 17347867836.416, "elapsed time": 224.63146157900337 }, { "step": 1000, "valid accuracy": 0.0, "train loss": 1.6029385833740235, "train samples": 4000, "train time": 48.4513027088251, "eval time": 16.587105787999462, "tokens / sec": 4299.905025299824, "mem allocated avg": 7080125773.824, "mem reserved avg": 17638306611.2, "elapsed time": 293.0906825060374 }, { "step": 1250, "valid accuracy": 0.0, "train loss": 1.5049201951026916, "train samples": 5000, "train time": 47.84065633546561, "eval time": 16.44257097603986, "tokens / sec": 4359.012103381303, "mem allocated avg": 7076596897.792, "mem reserved avg": 17677858897.92, "elapsed time": 360.6927552940324 }, { "step": 1500, "valid accuracy": 0.0, "train loss": 1.4373888688087464, "train samples": 6000, "train time": 48.565593625826295, "eval time": 16.451919412997086, "tokens / sec": 4310.273680844737, "mem allocated avg": 7079910084.608, "mem reserved avg": 17610313826.304, "elapsed time": 429.08035932003986 }, { "step": 1750, "valid accuracy": 0.0, "train loss": 1.3829970264434814, "train samples": 7000, "train time": 47.46174641727703, "eval time": 16.44140770897502, "tokens / sec": 4411.026053685007, "mem allocated avg": 7081518635.008, "mem reserved avg": 17944507580.416, "elapsed time": 496.28030893800315 }, { "step": 2000, "valid accuracy": 0.0, "train loss": 1.333943516254425, "train samples": 8000, "train time": 47.37612354574958, "eval time": 16.461310581013095, "tokens / sec": 4383.980462213941, "mem allocated avg": 7076895195.136, "mem reserved avg": 17455510454.272, "elapsed time": 563.4433135589934 }, { "step": 2250, "valid accuracy": 0.0, "train loss": 1.2829430739879608, "train samples": 9000, "train time": 49.185617247305345, "eval time": 16.46588272700319, "tokens / sec": 4370.139321811113, "mem allocated avg": 7088224200.704, "mem reserved avg": 17821908074.496, "elapsed time": 632.4641255650204 }, { "step": 2500, "valid accuracy": 0.0, "train loss": 1.246446587562561, "train samples": 10000, "train time": 48.255648160818964, "eval time": 16.453739245946053, "tokens / sec": 4268.246471658303, "mem allocated avg": 7075521802.24, "mem reserved avg": 17499282210.816, "elapsed time": 700.5415165220038 }, { "step": 2750, "valid accuracy": 0.0, "train loss": 1.2048211748600006, "train samples": 11000, "train time": 48.75862841651542, "eval time": 16.46915528498357, "tokens / sec": 4345.507797923046, "mem allocated avg": 7083802673.152, "mem reserved avg": 18024199356.416, "elapsed time": 769.1840648030047 }, { "step": 3000, "valid accuracy": 0.06, "train loss": 1.1679602472782136, "train samples": 12000, "train time": 48.22104389790911, "eval time": 16.46740536397556, "tokens / sec": 4328.6288128045, "mem allocated avg": 7078835648.512, "mem reserved avg": 17764907483.136, "elapsed time": 837.2026147870347 }, { "step": 3250, "valid accuracy": 0.04, "train loss": 1.131537829875946, "train samples": 13000, "train time": 48.137974261830095, "eval time": 16.45086403901223, "tokens / sec": 4381.177297841325, "mem allocated avg": 7080606814.208, "mem reserved avg": 17735715127.296, "elapsed time": 905.0784538229927 }, { "step": 3500, "valid accuracy": 0.04, "train loss": 1.1094365301132203, "train samples": 14000, "train time": 48.85280603397405, "eval time": 16.464311491989065, "tokens / sec": 4293.509770024922, "mem allocated avg": 7081026895.872, "mem reserved avg": 17464184274.944, "elapsed time": 973.75803008402 }, { "step": 3750, "valid accuracy": 0.02, "train loss": 1.079718291759491, "train samples": 15000, "train time": 49.39224000740796, "eval time": 16.467753204982728, "tokens / sec": 4387.389597384091, "mem allocated avg": 7091974598.656, "mem reserved avg": 17988463886.336, "elapsed time": 1042.9676414930145 }, { "step": 4000, "valid accuracy": 0.02, "train loss": 1.0901593000888825, "train samples": 16000, "train time": 48.37513752095401, "eval time": 16.4682428810047, "tokens / sec": 4224.752847709724, "mem allocated avg": 7072683055.104, "mem reserved avg": 17487336833.024, "elapsed time": 1111.1566645530402 }, { "step": 4250, "valid accuracy": 0.04, "train loss": 1.0608918931484221, "train samples": 17000, "train time": 48.56447010586271, "eval time": 16.604359342018142, "tokens / sec": 4352.750056557934, "mem allocated avg": 7082847784.96, "mem reserved avg": 17545067233.28, "elapsed time": 1179.6303421389894 }, { "step": 4500, "valid accuracy": 0.06, "train loss": 1.068639324426651, "train samples": 18000, "train time": 48.767977151903324, "eval time": 16.501046893012244, "tokens / sec": 4261.36190460976, "mem allocated avg": 7077102155.776, "mem reserved avg": 17491698909.184, "elapsed time": 1248.2997376249987 }, { "step": 4750, "valid accuracy": 0.02, "train loss": 1.0587589330673217, "train samples": 19000, "train time": 48.74695514619816, "eval time": 16.455141162965447, "tokens / sec": 4306.710016458811, "mem allocated avg": 7080302989.312, "mem reserved avg": 17688931860.48, "elapsed time": 1316.8433299760218 }, { "step": 5000, "valid accuracy": 0.04, "train loss": 1.0654818024635315, "train samples": 20000, "train time": 48.002736268914305, "eval time": 16.48369751899736, "tokens / sec": 4338.91932395692, "mem allocated avg": 7078423576.576, "mem reserved avg": 17237222096.896, "elapsed time": 1384.6823478980223 }, { "step": 5000, "test accuracy": 0.053828658074298714, "train loss": 1.0654818024635315, "train samples": 20000, "train total tokens": 4198051, "forgetting": 2.5893402099609375 } ] }, "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" } }