jung-geun/PSO
1
0
Fork 0
mirror of https://github.com/jung-geun/PSO.git synced 2026-09-20 14:11:48 +09:00
Code Issues Packages Projects Releases Wiki Activity

feat: modernize PSO and add convergence research

Browse Source 
Migrate the package and examples to the tensor-native PyTorch implementation, add benchmark evidence, and add the guarded post-training convergence protocol with TensorBoard progress monitoring and hash-verified recovery.

Constraint: Preserve one-shot official-test sealing and auditable research artifacts

Rejected: Commit local .omc runs and downloaded datasets | multi-gigabyte runtime state is machine-local

Confidence: high

Scope-risk: broad

Not-tested: Production CUDA run on pieroot-server
This commit is contained in:
jung-geun
2026-09-07 22:03:25 +09:00
parent 9745bb7ad4
commit 813433000a
120 changed files with 198148 additions and 4745 deletions
Download Patch File Download Diff File Expand all files Collapse all files
benchmark_results/pso_v4_120p80_replication.csv
+16
View File
@@ -0,0 +1,16 @@
cohort,method,candidate_label,regimen,seed,n_particles,epochs,particle_epochs,train_loss,train_acc,test_loss,test_acc,test_mse,fit_time_sec,data_fingerprint,model_fingerprint,device,completed,error
baseline,adaptive_moment,am_b0.06_s0.5,fixed_epoch,71,120,80,9600,0.6516683101654053,0.7950000166893005,0.8584634065628052,0.7360000014305115,0.037673790007829666,11.962705624988303,dfe645918ece54c0,0777bd52fd76272d,mps,True,
baseline,adaptive_moment,am_b0.06_s0.5,fixed_epoch,72,120,80,9600,0.6982801556587219,0.7914999723434448,0.9025362133979797,0.7239999771118164,0.038956169039011,11.906837583053857,dfe645918ece54c0,6fcb6e473bdacbd2,mps,True,
baseline,adaptive_moment,am_b0.06_s0.5,fixed_epoch,73,120,80,9600,0.7527623772621155,0.7749999761581421,1.001193642616272,0.6919999718666077,0.04309915751218796,10.664651792030782,dfe645918ece54c0,2e6c351372592f10,mps,True,
baseline,adaptive_moment,am_b0.06_s0.5,fixed_epoch,74,120,80,9600,0.6858600974082947,0.784500002861023,0.8601324558258057,0.7350000143051147,0.03845023736357689,10.921957665821537,dfe645918ece54c0,4000fe3fb26ef207,mps,True,
baseline,adaptive_moment,am_b0.06_s0.5,fixed_epoch,75,120,80,9600,0.6326538324356079,0.8125,0.889915406703949,0.7300000190734863,0.03931796923279762,10.411899874918163,dfe645918ece54c0,0966039f5ef7af88,mps,True,
replay,adaptive_moment,am_b0.06_s0.5,fixed_epoch,71,120,80,9600,0.6516683101654053,0.7950000166893005,0.8584634065628052,0.7360000014305115,0.037673790007829666,11.090910458937287,dfe645918ece54c0,0777bd52fd76272d,mps,True,
replay,adaptive_moment,am_b0.06_s0.5,fixed_epoch,72,120,80,9600,0.6982801556587219,0.7914999723434448,0.9025362133979797,0.7239999771118164,0.038956169039011,10.668682500021532,dfe645918ece54c0,6fcb6e473bdacbd2,mps,True,
replay,adaptive_moment,am_b0.06_s0.5,fixed_epoch,73,120,80,9600,0.7527623772621155,0.7749999761581421,1.001193642616272,0.6919999718666077,0.04309915751218796,11.617381499847397,dfe645918ece54c0,2e6c351372592f10,mps,True,
replay,adaptive_moment,am_b0.06_s0.5,fixed_epoch,74,120,80,9600,0.6858600974082947,0.784500002861023,0.8601324558258057,0.7350000143051147,0.03845023736357689,12.46007908298634,dfe645918ece54c0,4000fe3fb26ef207,mps,True,
replay,adaptive_moment,am_b0.06_s0.5,fixed_epoch,75,120,80,9600,0.6326538324356079,0.8125,0.889915406703949,0.7300000190734863,0.03931796923279762,11.499297459144145,dfe645918ece54c0,0966039f5ef7af88,mps,True,
independent,adaptive_moment,am_b0.06_s0.5,fixed_epoch,81,120,80,9600,0.7000858783721924,0.781000018119812,0.9081020355224609,0.7250000238418579,0.039418451488018036,10.6923490408808,dfe645918ece54c0,b9e1b8cbb9345add,mps,True,
independent,adaptive_moment,am_b0.06_s0.5,fixed_epoch,82,120,80,9600,0.6899945139884949,0.796999990940094,0.9245963096618652,0.718999981880188,0.03965267166495323,11.110405791085213,dfe645918ece54c0,f1a0025f5b3b5b7c,mps,True,
independent,adaptive_moment,am_b0.06_s0.5,fixed_epoch,83,120,80,9600,0.6943607330322266,0.7929999828338623,0.8056192398071289,0.7609999775886536,0.034950967878103256,10.71645870897919,dfe645918ece54c0,9cb7fe904bf992ac,mps,True,
independent,adaptive_moment,am_b0.06_s0.5,fixed_epoch,84,120,80,9600,0.7076351046562195,0.7885000109672546,0.8160682320594788,0.7429999709129333,0.0366184301674366,10.859140583081171,dfe645918ece54c0,b833ebd886fce382,mps,True,
independent,adaptive_moment,am_b0.06_s0.5,fixed_epoch,85,120,80,9600,0.6802449822425842,0.7875000238418579,0.8681835532188416,0.7319999933242798,0.038917701691389084,10.97266870806925,dfe645918ece54c0,e6bdf9e5d849521b,mps,True,
1 cohort method candidate_label regimen seed n_particles epochs particle_epochs train_loss train_acc test_loss test_acc test_mse fit_time_sec data_fingerprint model_fingerprint device completed error
2 baseline adaptive_moment am_b0.06_s0.5 fixed_epoch 71 120 80 9600 0.6516683101654053 0.7950000166893005 0.8584634065628052 0.7360000014305115 0.037673790007829666 11.962705624988303 dfe645918ece54c0 0777bd52fd76272d mps True
3 baseline adaptive_moment am_b0.06_s0.5 fixed_epoch 72 120 80 9600 0.6982801556587219 0.7914999723434448 0.9025362133979797 0.7239999771118164 0.038956169039011 11.906837583053857 dfe645918ece54c0 6fcb6e473bdacbd2 mps True
4 baseline adaptive_moment am_b0.06_s0.5 fixed_epoch 73 120 80 9600 0.7527623772621155 0.7749999761581421 1.001193642616272 0.6919999718666077 0.04309915751218796 10.664651792030782 dfe645918ece54c0 2e6c351372592f10 mps True
5 baseline adaptive_moment am_b0.06_s0.5 fixed_epoch 74 120 80 9600 0.6858600974082947 0.784500002861023 0.8601324558258057 0.7350000143051147 0.03845023736357689 10.921957665821537 dfe645918ece54c0 4000fe3fb26ef207 mps True
6 baseline adaptive_moment am_b0.06_s0.5 fixed_epoch 75 120 80 9600 0.6326538324356079 0.8125 0.889915406703949 0.7300000190734863 0.03931796923279762 10.411899874918163 dfe645918ece54c0 0966039f5ef7af88 mps True
7 replay adaptive_moment am_b0.06_s0.5 fixed_epoch 71 120 80 9600 0.6516683101654053 0.7950000166893005 0.8584634065628052 0.7360000014305115 0.037673790007829666 11.090910458937287 dfe645918ece54c0 0777bd52fd76272d mps True
8 replay adaptive_moment am_b0.06_s0.5 fixed_epoch 72 120 80 9600 0.6982801556587219 0.7914999723434448 0.9025362133979797 0.7239999771118164 0.038956169039011 10.668682500021532 dfe645918ece54c0 6fcb6e473bdacbd2 mps True
9 replay adaptive_moment am_b0.06_s0.5 fixed_epoch 73 120 80 9600 0.7527623772621155 0.7749999761581421 1.001193642616272 0.6919999718666077 0.04309915751218796 11.617381499847397 dfe645918ece54c0 2e6c351372592f10 mps True
10 replay adaptive_moment am_b0.06_s0.5 fixed_epoch 74 120 80 9600 0.6858600974082947 0.784500002861023 0.8601324558258057 0.7350000143051147 0.03845023736357689 12.46007908298634 dfe645918ece54c0 4000fe3fb26ef207 mps True
11 replay adaptive_moment am_b0.06_s0.5 fixed_epoch 75 120 80 9600 0.6326538324356079 0.8125 0.889915406703949 0.7300000190734863 0.03931796923279762 11.499297459144145 dfe645918ece54c0 0966039f5ef7af88 mps True
12 independent adaptive_moment am_b0.06_s0.5 fixed_epoch 81 120 80 9600 0.7000858783721924 0.781000018119812 0.9081020355224609 0.7250000238418579 0.039418451488018036 10.6923490408808 dfe645918ece54c0 b9e1b8cbb9345add mps True
13 independent adaptive_moment am_b0.06_s0.5 fixed_epoch 82 120 80 9600 0.6899945139884949 0.796999990940094 0.9245963096618652 0.718999981880188 0.03965267166495323 11.110405791085213 dfe645918ece54c0 f1a0025f5b3b5b7c mps True
14 independent adaptive_moment am_b0.06_s0.5 fixed_epoch 83 120 80 9600 0.6943607330322266 0.7929999828338623 0.8056192398071289 0.7609999775886536 0.034950967878103256 10.71645870897919 dfe645918ece54c0 9cb7fe904bf992ac mps True
15 independent adaptive_moment am_b0.06_s0.5 fixed_epoch 84 120 80 9600 0.7076351046562195 0.7885000109672546 0.8160682320594788 0.7429999709129333 0.0366184301674366 10.859140583081171 dfe645918ece54c0 b833ebd886fce382 mps True
16 independent adaptive_moment am_b0.06_s0.5 fixed_epoch 85 120 80 9600 0.6802449822425842 0.7875000238418579 0.8681835532188416 0.7319999933242798 0.038917701691389084 10.97266870806925 dfe645918ece54c0 e6bdf9e5d849521b mps True