Ramzes / peft /method_comparison /MetaMathQA /results /vera--llama-3.2-3B-default.json
Bordoglor's picture
Upload folder using huggingface_hub
f472b08 verified
{
"run_info": {
"created_at": "2025-06-19T20:53:39+00:00",
"total_time": 2024.6820476150024,
"experiment_name": "vera/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
},
"peft_config": {
"task_type": null,
"peft_type": "VERA",
"auto_mapping": null,
"base_model_name_or_path": "meta-llama/Llama-3.2-3B",
"revision": null,
"inference_mode": false,
"r": 256,
"target_modules": [
"v_proj",
"q_proj"
],
"projection_prng_key": 0,
"save_projection": true,
"vera_dropout": 0.0,
"d_initial": 0.1,
"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": 11489715316,
"accelerator_memory_max": 21596471296,
"accelerator_memory_reserved_99th": 17291123097,
"train_time": 1819.9693055349999,
"file_size": 6821968,
"num_trainable_params": 129024,
"num_total_params": 3212878848,
"status": "success",
"metrics": [
{
"step": 250,
"valid accuracy": 0.0,
"train loss": 1.3017588877677917,
"train samples": 1000,
"train time": 32.843521857023006,
"eval time": 11.480974874997628,
"tokens / sec": 6446.294064372017,
"mem allocated avg": 6784826523.648,
"mem reserved avg": 11538438029.312,
"elapsed time": 95.45296428899746
},
{
"step": 500,
"valid accuracy": 0.28,
"train loss": 1.0202219936847687,
"train samples": 2000,
"train time": 32.35236015598639,
"eval time": 11.4980273259971,
"tokens / sec": 6429.051821788439,
"mem allocated avg": 6777359808.512,
"mem reserved avg": 11429948162.048,
"elapsed time": 183.95939499299857
},
{
"step": 750,
"valid accuracy": 0.38,
"train loss": 0.8040032889842987,
"train samples": 3000,
"train time": 32.52055500800634,
"eval time": 11.426841341002728,
"tokens / sec": 6592.784162115804,
"mem allocated avg": 6787965165.568,
"mem reserved avg": 11585061912.576,
"elapsed time": 272.8589564269969
},
{
"step": 1000,
"valid accuracy": 0.3,
"train loss": 0.7544035723209381,
"train samples": 4000,
"train time": 32.27830113501477,
"eval time": 11.54098314699877,
"tokens / sec": 6454.3669485133405,
"mem allocated avg": 6779215933.44,
"mem reserved avg": 11460172316.672,
"elapsed time": 361.1500098109973
},
{
"step": 1250,
"valid accuracy": 0.44,
"train loss": 0.7379197257757187,
"train samples": 5000,
"train time": 32.060909217962035,
"eval time": 11.406497389998549,
"tokens / sec": 6504.431879404317,
"mem allocated avg": 6779128844.288,
"mem reserved avg": 11454770053.12,
"elapsed time": 449.3482204989996
},
{
"step": 1500,
"valid accuracy": 0.4,
"train loss": 0.7252234178781509,
"train samples": 6000,
"train time": 31.98088176901365,
"eval time": 11.480169268001191,
"tokens / sec": 6545.504326988923,
"mem allocated avg": 6780286265.344,
"mem reserved avg": 11479667441.664,
"elapsed time": 537.3097453219998
},
{
"step": 1750,
"valid accuracy": 0.4,
"train loss": 0.7148357192277909,
"train samples": 7000,
"train time": 32.29452324002341,
"eval time": 11.44221062500219,
"tokens / sec": 6482.678144650271,
"mem allocated avg": 6782215264.256,
"mem reserved avg": 11493600919.552,
"elapsed time": 625.780868398997
},
{
"step": 2000,
"valid accuracy": 0.4,
"train loss": 0.7139411936998368,
"train samples": 8000,
"train time": 32.33002986999054,
"eval time": 11.472246884000924,
"tokens / sec": 6424.243987253105,
"mem allocated avg": 6778636718.08,
"mem reserved avg": 11439217573.888,
"elapsed time": 714.3076436519987
},
{
"step": 2250,
"valid accuracy": 0.38,
"train loss": 0.7067342863082886,
"train samples": 9000,
"train time": 32.69249906902769,
"eval time": 11.424881449998793,
"tokens / sec": 6574.841511692143,
"mem allocated avg": 6789716504.576,
"mem reserved avg": 11617542602.752,
"elapsed time": 803.4051666009982
},
{
"step": 2500,
"valid accuracy": 0.4,
"train loss": 0.7048580280542374,
"train samples": 10000,
"train time": 31.796681229010574,
"eval time": 11.401134708998143,
"tokens / sec": 6477.625715607085,
"mem allocated avg": 6775192217.6,
"mem reserved avg": 11386755219.456,
"elapsed time": 890.7853266579987
},
{
"step": 2750,
"valid accuracy": 0.36,
"train loss": 0.6994425257444382,
"train samples": 11000,
"train time": 32.589996781029186,
"eval time": 6.453208308001194,
"tokens / sec": 6501.412118068606,
"mem allocated avg": 6785945655.296,
"mem reserved avg": 11552530890.752,
"elapsed time": 974.6122346880002
},
{
"step": 3000,
"valid accuracy": 0.4,
"train loss": 0.6912879683971405,
"train samples": 12000,
"train time": 32.34826778500428,
"eval time": 11.457833226999355,
"tokens / sec": 6452.617536966281,
"mem allocated avg": 6780318763.008,
"mem reserved avg": 11474030297.088,
"elapsed time": 1062.897270567999
},
{
"step": 3250,
"valid accuracy": 0.4,
"train loss": 0.700449936747551,
"train samples": 13000,
"train time": 32.51472582996939,
"eval time": 8.004199169998174,
"tokens / sec": 6486.322569744963,
"mem allocated avg": 6782387701.76,
"mem reserved avg": 11501452656.64,
"elapsed time": 1148.3985279560002
},
{
"step": 3500,
"valid accuracy": 0.36,
"train loss": 0.6886729755401612,
"train samples": 14000,
"train time": 32.572147220984334,
"eval time": 11.456443364000734,
"tokens / sec": 6439.550901479111,
"mem allocated avg": 6781381988.352,
"mem reserved avg": 11484943876.096,
"elapsed time": 1237.2252680229976
},
{
"step": 3750,
"valid accuracy": 0.38,
"train loss": 0.6851948540210724,
"train samples": 15000,
"train time": 32.8770313250061,
"eval time": 8.042231839001033,
"tokens / sec": 6591.318962402083,
"mem allocated avg": 6791807023.104,
"mem reserved avg": 11653781389.312,
"elapsed time": 1323.4750151669978
},
{
"step": 4000,
"valid accuracy": 0.36,
"train loss": 0.7032276903390884,
"train samples": 16000,
"train time": 31.65130396198947,
"eval time": 7.9955749260007,
"tokens / sec": 6457.016754994822,
"mem allocated avg": 6773653422.08,
"mem reserved avg": 11367989903.36,
"elapsed time": 1407.2714081800004
},
{
"step": 4250,
"valid accuracy": 0.36,
"train loss": 0.684476065993309,
"train samples": 17000,
"train time": 32.02934406197164,
"eval time": 8.007123895000404,
"tokens / sec": 6599.854170943876,
"mem allocated avg": 6784119472.128,
"mem reserved avg": 11519949537.28,
"elapsed time": 1492.0019941529972
},
{
"step": 4500,
"valid accuracy": 0.38,
"train loss": 0.6939880999326706,
"train samples": 18000,
"train time": 31.936327281997364,
"eval time": 9.855819755000994,
"tokens / sec": 6507.260467522446,
"mem allocated avg": 6777879162.88,
"mem reserved avg": 11436331892.736,
"elapsed time": 1578.2498042659972
},
{
"step": 4750,
"valid accuracy": 0.36,
"train loss": 0.68637368786335,
"train samples": 19000,
"train time": 32.33460194401778,
"eval time": 6.469711448000453,
"tokens / sec": 6492.704019164238,
"mem allocated avg": 6781104441.344,
"mem reserved avg": 11484004352.0,
"elapsed time": 1662.171022565999
},
{
"step": 5000,
"valid accuracy": 0.38,
"train loss": 0.6926896897554398,
"train samples": 20000,
"train time": 32.14674746405217,
"eval time": 8.441190715999255,
"tokens / sec": 6479.038049896257,
"mem allocated avg": 6777818853.376,
"mem reserved avg": 11434117300.224,
"elapsed time": 1747.4833575960001
},
{
"step": 5000,
"test accuracy": 0.3684609552691433,
"train loss": 0.6926896897554398,
"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"
}
}