{ "run_info": { "created_at": "2025-06-20T08:46:44+00:00", "total_time": 2700.1305744579877, "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, "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, "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": 15297773830, "accelerator_memory_max": 24379392000, "accelerator_memory_reserved_99th": 20669781770, "train_time": 2379.557773831024, "file_size": 2457728, "num_trainable_params": 614400, "num_total_params": 3213364224, "status": "success", "metrics": [ { "step": 250, "valid accuracy": 0.0, "train loss": 3.462425223350525, "train samples": 1000, "train time": 46.206722402057494, "eval time": 15.901069569998072, "tokens / sec": 4581.9956273412845, "mem allocated avg": 7082871494.656, "mem reserved avg": 15331489742.848, "elapsed time": 119.40567356300016 }, { "step": 500, "valid accuracy": 0.0, "train loss": 2.259350722312927, "train samples": 2000, "train time": 45.66361523300293, "eval time": 15.856271529002697, "tokens / sec": 4554.939396249854, "mem allocated avg": 7075523266.56, "mem reserved avg": 15240674672.64, "elapsed time": 232.12755202699918 }, { "step": 750, "valid accuracy": 0.0, "train loss": 1.758247773170471, "train samples": 3000, "train time": 46.58154148896574, "eval time": 15.854417883005226, "tokens / sec": 4602.70298377282, "mem allocated avg": 7085465481.216, "mem reserved avg": 15376771448.832, "elapsed time": 346.0752758900053 }, { "step": 1000, "valid accuracy": 0.0, "train loss": 1.6028480381965637, "train samples": 4000, "train time": 45.41573346107907, "eval time": 15.861826895998092, "tokens / sec": 4587.30893729906, "mem allocated avg": 7077486481.408, "mem reserved avg": 15288170971.136, "elapsed time": 458.6240012299968 }, { "step": 1250, "valid accuracy": 0.0, "train loss": 1.5049157681465148, "train samples": 5000, "train time": 46.04039786210342, "eval time": 15.877354786993237, "tokens / sec": 4529.456948321703, "mem allocated avg": 7076584331.264, "mem reserved avg": 15265983102.976, "elapsed time": 571.9228152269934 }, { "step": 1500, "valid accuracy": 0.0, "train loss": 1.4375499501228333, "train samples": 6000, "train time": 45.70124057796784, "eval time": 15.84707298700232, "tokens / sec": 4580.4227052190045, "mem allocated avg": 7078481408.0, "mem reserved avg": 15279463596.032, "elapsed time": 684.8850296739984 }, { "step": 1750, "valid accuracy": 0.0, "train loss": 1.3827230257987977, "train samples": 7000, "train time": 44.976750778907444, "eval time": 15.845691901995451, "tokens / sec": 4654.7382008346485, "mem allocated avg": 7079360505.856, "mem reserved avg": 15298052751.36, "elapsed time": 796.8428356289951 }, { "step": 2000, "valid accuracy": 0.0, "train loss": 1.3338124525547028, "train samples": 8000, "train time": 45.10262611102371, "eval time": 15.857041016992298, "tokens / sec": 4604.964675199615, "mem allocated avg": 7075931449.344, "mem reserved avg": 15257242173.44, "elapsed time": 908.9726742479979 }, { "step": 2250, "valid accuracy": 0.0, "train loss": 1.2829065501689911, "train samples": 9000, "train time": 46.84363810600189, "eval time": 15.872781344005489, "tokens / sec": 4588.627371631486, "mem allocated avg": 7087554078.72, "mem reserved avg": 15416986435.584, "elapsed time": 1023.331907868007 }, { "step": 2500, "valid accuracy": 0.0, "train loss": 1.2462495183944702, "train samples": 10000, "train time": 45.55510413390584, "eval time": 15.84976143699896, "tokens / sec": 4521.271631705095, "mem allocated avg": 7072915062.784, "mem reserved avg": 15202909159.424, "elapsed time": 1136.1328145180014 }, { "step": 2750, "valid accuracy": 0.0, "train loss": 1.2045790712833404, "train samples": 11000, "train time": 45.34144312601711, "eval time": 15.8525270359969, "tokens / sec": 4673.009621928461, "mem allocated avg": 7083153442.816, "mem reserved avg": 15344005545.984, "elapsed time": 1248.7101804669946 }, { "step": 3000, "valid accuracy": 0.0, "train loss": 1.1678078708648683, "train samples": 12000, "train time": 45.599694666831056, "eval time": 15.870247816987103, "tokens / sec": 4577.464860786221, "mem allocated avg": 7077996111.872, "mem reserved avg": 15283892781.056, "elapsed time": 1361.5449211609957 }, { "step": 3250, "valid accuracy": 0.04, "train loss": 1.1313301923274994, "train samples": 13000, "train time": 45.95094640579191, "eval time": 15.868188906999421, "tokens / sec": 4589.698722144641, "mem allocated avg": 7079686449.152, "mem reserved avg": 15301248811.008, "elapsed time": 1474.734694629995 }, { "step": 3500, "valid accuracy": 0.06, "train loss": 1.1092858843803406, "train samples": 14000, "train time": 45.96525488591578, "eval time": 15.86030059499899, "tokens / sec": 4563.229346178814, "mem allocated avg": 7078805225.472, "mem reserved avg": 15302347718.656, "elapsed time": 1588.1363447299955 }, { "step": 3750, "valid accuracy": 0.06, "train loss": 1.079538120508194, "train samples": 15000, "train time": 46.46510764303093, "eval time": 15.86466599200503, "tokens / sec": 4663.779145091515, "mem allocated avg": 7089610215.424, "mem reserved avg": 15446287843.328, "elapsed time": 1702.2553167559963 }, { "step": 4000, "valid accuracy": 0.04, "train loss": 1.0899075508117675, "train samples": 16000, "train time": 45.08557640206709, "eval time": 15.860410296008922, "tokens / sec": 4533.001822521445, "mem allocated avg": 7071494891.52, "mem reserved avg": 15189319614.464, "elapsed time": 1814.3939928110049 }, { "step": 4250, "valid accuracy": 0.04, "train loss": 1.0607522547245025, "train samples": 17000, "train time": 46.2303190480452, "eval time": 15.875090683999588, "tokens / sec": 4572.518735601033, "mem allocated avg": 7082239875.072, "mem reserved avg": 15329283538.944, "elapsed time": 1928.1608909490024 }, { "step": 4500, "valid accuracy": 0.04, "train loss": 1.068591582775116, "train samples": 18000, "train time": 45.96484722109744, "eval time": 15.854171614992083, "tokens / sec": 4521.237697155087, "mem allocated avg": 7076175783.936, "mem reserved avg": 15251420479.488, "elapsed time": 2041.5032397750037 }, { "step": 4750, "valid accuracy": 0.06, "train loss": 1.0587167317867279, "train samples": 19000, "train time": 45.48911916205543, "eval time": 15.858397545001935, "tokens / sec": 4615.147619194169, "mem allocated avg": 7079419088.896, "mem reserved avg": 15298539290.624, "elapsed time": 2154.3035376479966 }, { "step": 5000, "valid accuracy": 0.02, "train loss": 1.0654937489032745, "train samples": 20000, "train time": 45.758550852071494, "eval time": 15.85034008299408, "tokens / sec": 4551.7175723796145, "mem allocated avg": 7075618770.944, "mem reserved avg": 15251386925.056, "elapsed time": 2267.4055672899995 }, { "step": 5000, "test accuracy": 0.050037907505686124, "train loss": 1.0654937489032745, "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": "4.52.4", "transformers-commit-hash": null, "peft-version": "0.15.2.dev0", "peft-commit-hash": "5fe7f8f8abe914d313fc3751f2ea92de7718fbaf", "datasets-version": "3.6.0", "datasets-commit-hash": null, "bitsandbytes-version": "0.46.0", "bitsandbytes-commit-hash": null, "torch-version": "2.7.1+cu126", "torch-commit-hash": null }, "system_info": { "system": "Linux", "release": "6.8.0-1029-aws", "version": "#31-Ubuntu SMP Wed Apr 23 18:42:41 UTC 2025", "machine": "x86_64", "processor": "x86_64", "accelerator": "NVIDIA L40S" }, "pytorch_info": "PyTorch built with:\n - GCC 11.2\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 12.6\n - NVCC architecture flags: -gencode;arch=compute_50,code=sm_50;-gencode;arch=compute_60,code=sm_60;-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\n - CuDNN 90.7.1 (built against CUDA 12.8)\n - Built with CuDNN 90.5.1\n - Magma 2.6.1\n - Build settings: BLAS_INFO=mkl, BUILD_TYPE=Release, COMMIT_SHA=e2d141dbde55c2a4370fac5165b0561b6af4798b, CUDA_VERSION=12.6, CUDNN_VERSION=9.5.1, CXX_COMPILER=/opt/rh/gcc-toolset-11/root/usr/bin/c++, CXX_FLAGS= -D_GLIBCXX_USE_CXX11_ABI=1 -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 -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 -fdiagnostics-color=always -faligned-new -Wno-maybe-uninitialized -fno-math-errno -fno-trapping-math -Werror=format -Wno-stringop-overflow, LAPACK_INFO=mkl, PERF_WITH_AVX=1, PERF_WITH_AVX2=1, TORCH_VERSION=2.7.1, 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, \n" } }