Upload folder using huggingface_hub
Browse filesThis view is limited to 50 files because it contains too many changes. See raw diff
- cola__SAM_Muon__lr2e-05__bs32__ep3__rho0.002/checkpoint_final.pt +3 -0
- cola__SAM_Muon__lr2e-05__bs32__ep3__rho0.002/metrics_history.json +14 -0
- cola__SAM_Muon__lr2e-05__bs32__ep3__rho0.002/params.json +99 -0
- cola__SAM__lr2e-05__bs32__ep3__rho0.002/checkpoint_final.pt +3 -0
- cola__SAM__lr2e-05__bs32__ep3__rho0.002/metrics_history.json +14 -0
- cola__SAM__lr2e-05__bs32__ep3__rho0.002/params.json +99 -0
- cola__SGD__lr2e-05__bs32__ep3__rho_na/checkpoint_final.pt +3 -0
- cola__SGD__lr2e-05__bs32__ep3__rho_na/metrics_history.json +14 -0
- cola__SGD__lr2e-05__bs32__ep3__rho_na/params.json +98 -0
- mnli__SAM_Muon__lr2e-05__bs32__ep3__rho0.005/checkpoint_final.pt +3 -0
- mnli__SAM_Muon__lr2e-05__bs32__ep3__rho0.005/metrics_history.json +20 -0
- mnli__SAM_Muon__lr2e-05__bs32__ep3__rho0.005/params.json +99 -0
- mnli__SAM__lr2e-05__bs32__ep3__rho0.005/checkpoint_final.pt +3 -0
- mnli__SAM__lr2e-05__bs32__ep3__rho0.005/metrics_history.json +20 -0
- mnli__SAM__lr2e-05__bs32__ep3__rho0.005/params.json +99 -0
- mnli__SGD__lr2e-05__bs32__ep3__rho_na/checkpoint_final.pt +3 -0
- mnli__SGD__lr2e-05__bs32__ep3__rho_na/metrics_history.json +20 -0
- mnli__SGD__lr2e-05__bs32__ep3__rho_na/params.json +98 -0
- mrpc__SAM_Muon__lr2e-05__bs32__ep3__rho0.005/checkpoint_final.pt +3 -0
- mrpc__SAM_Muon__lr2e-05__bs32__ep3__rho0.005/metrics_history.json +17 -0
- mrpc__SAM_Muon__lr2e-05__bs32__ep3__rho0.005/params.json +99 -0
- mrpc__SAM__lr2e-05__bs32__ep3__rho0.005/checkpoint_final.pt +3 -0
- mrpc__SAM__lr2e-05__bs32__ep3__rho0.005/metrics_history.json +17 -0
- mrpc__SAM__lr2e-05__bs32__ep3__rho0.005/params.json +99 -0
- mrpc__SGD__lr2e-05__bs32__ep3__rho_na/checkpoint_final.pt +3 -0
- mrpc__SGD__lr2e-05__bs32__ep3__rho_na/metrics_history.json +17 -0
- mrpc__SGD__lr2e-05__bs32__ep3__rho_na/params.json +98 -0
- qnli__SAM_Muon__lr2e-05__bs32__ep3__rho0.005/checkpoint_final.pt +3 -0
- qnli__SAM_Muon__lr2e-05__bs32__ep3__rho0.005/metrics_history.json +14 -0
- qnli__SAM_Muon__lr2e-05__bs32__ep3__rho0.005/params.json +99 -0
- qnli__SAM__lr2e-05__bs32__ep3__rho0.005/checkpoint_final.pt +3 -0
- qnli__SAM__lr2e-05__bs32__ep3__rho0.005/metrics_history.json +14 -0
- qnli__SAM__lr2e-05__bs32__ep3__rho0.005/params.json +99 -0
- qnli__SGD__lr2e-05__bs32__ep3__rho_na/checkpoint_final.pt +3 -0
- qnli__SGD__lr2e-05__bs32__ep3__rho_na/metrics_history.json +14 -0
- qnli__SGD__lr2e-05__bs32__ep3__rho_na/params.json +98 -0
- qqp__SAM_Muon__lr2e-05__bs32__ep3__rho0.005/checkpoint_final.pt +3 -0
- qqp__SAM_Muon__lr2e-05__bs32__ep3__rho0.005/metrics_history.json +17 -0
- qqp__SAM_Muon__lr2e-05__bs32__ep3__rho0.005/params.json +99 -0
- qqp__SAM__lr2e-05__bs32__ep3__rho0.005/checkpoint_final.pt +3 -0
- qqp__SAM__lr2e-05__bs32__ep3__rho0.005/metrics_history.json +17 -0
- qqp__SAM__lr2e-05__bs32__ep3__rho0.005/params.json +99 -0
- qqp__SGD__lr2e-05__bs32__ep3__rho_na/checkpoint_final.pt +3 -0
- qqp__SGD__lr2e-05__bs32__ep3__rho_na/metrics_history.json +17 -0
- qqp__SGD__lr2e-05__bs32__ep3__rho_na/params.json +98 -0
- rte__SAM_Muon__lr2e-05__bs32__ep3__rho0.004/checkpoint_final.pt +3 -0
- rte__SAM_Muon__lr2e-05__bs32__ep3__rho0.004/metrics_history.json +14 -0
- rte__SAM_Muon__lr2e-05__bs32__ep3__rho0.004/params.json +99 -0
- rte__SAM__lr2e-05__bs32__ep3__rho0.004/checkpoint_final.pt +3 -0
- rte__SAM__lr2e-05__bs32__ep3__rho0.004/metrics_history.json +14 -0
cola__SAM_Muon__lr2e-05__bs32__ep3__rho0.002/checkpoint_final.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:0e64ecdc33e3260309257bef47efb37bad92e7e4a5eb5295375bc23fda7be1e2
|
| 3 |
+
size 438019989
|
cola__SAM_Muon__lr2e-05__bs32__ep3__rho0.002/metrics_history.json
ADDED
|
@@ -0,0 +1,14 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
[
|
| 2 |
+
{
|
| 3 |
+
"epoch": 1,
|
| 4 |
+
"mcc": 0.00
|
| 5 |
+
},
|
| 6 |
+
{
|
| 7 |
+
"epoch": 2,
|
| 8 |
+
"mcc": 0.00
|
| 9 |
+
},
|
| 10 |
+
{
|
| 11 |
+
"epoch": 3,
|
| 12 |
+
"mcc": 0.00
|
| 13 |
+
}
|
| 14 |
+
]
|
cola__SAM_Muon__lr2e-05__bs32__ep3__rho0.002/params.json
ADDED
|
@@ -0,0 +1,99 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"label_smoothing": 0.0,
|
| 3 |
+
"ESAMbeta": 0.5,
|
| 4 |
+
"ESAMgamma": 0.5,
|
| 5 |
+
"lookSAMk": 5,
|
| 6 |
+
"lookSAMalpha": 0.7,
|
| 7 |
+
"lookSAMLayerwise": false,
|
| 8 |
+
"optimizer": "SAM_Muon",
|
| 9 |
+
"weightDecay": 0.01,
|
| 10 |
+
"momentum": 0.9,
|
| 11 |
+
"rho": 0.002,
|
| 12 |
+
"samRho": 0.002,
|
| 13 |
+
"asamRho": 0.002,
|
| 14 |
+
"esamRho": 0.002,
|
| 15 |
+
"looksamRho": 0.002,
|
| 16 |
+
"friendlysamRho": 0.002,
|
| 17 |
+
"fishersamRho": 0.002,
|
| 18 |
+
"mongesamRho": 0.002,
|
| 19 |
+
"gsamRho": 0.002,
|
| 20 |
+
"bayesiansamRho": 0.002,
|
| 21 |
+
"msamRho": 0.002,
|
| 22 |
+
"adamwSamRho": 0.002,
|
| 23 |
+
"adamSamRho": 0.002,
|
| 24 |
+
"adamwMsamRho": 0.002,
|
| 25 |
+
"adamwFriendlysamRho": 0.002,
|
| 26 |
+
"sigma": 1.0,
|
| 27 |
+
"lmbda": 0.9,
|
| 28 |
+
"fisherEta": 1.0,
|
| 29 |
+
"gsamAlpha": 0.01,
|
| 30 |
+
"gsamAdaptive": false,
|
| 31 |
+
"bayesMSharpness": 8,
|
| 32 |
+
"bayesDelta": 10.0,
|
| 33 |
+
"bayesBeta1": 0.9,
|
| 34 |
+
"bayesBeta2": 0.999,
|
| 35 |
+
"bayesSInit": 1.0,
|
| 36 |
+
"bayesDamping": 1e-08,
|
| 37 |
+
"bayesEps": 1e-12,
|
| 38 |
+
"nesterov": false,
|
| 39 |
+
"grad_clip_norm": 1.0,
|
| 40 |
+
"rndSeed": true,
|
| 41 |
+
"deterministic": false,
|
| 42 |
+
"glue_task": "cola",
|
| 43 |
+
"glue_output_root": "/mnt/hungpv/projects/SAM_Variant/logs_nlp",
|
| 44 |
+
"widthFactor": 4,
|
| 45 |
+
"ViTPatchsize": 32,
|
| 46 |
+
"ViTSize": "S",
|
| 47 |
+
"pyramidAlpha": 270,
|
| 48 |
+
"pyramidBottleneck": true,
|
| 49 |
+
"model": "WRN",
|
| 50 |
+
"dropout": 0.0,
|
| 51 |
+
"BN": true,
|
| 52 |
+
"depth": 16,
|
| 53 |
+
"dataThreads": 4,
|
| 54 |
+
"dataDir": "/home/hungpv/.datasets",
|
| 55 |
+
"batchSize": 32,
|
| 56 |
+
"dataset": "CIFAR100",
|
| 57 |
+
"imageSize": 0,
|
| 58 |
+
"flip": false,
|
| 59 |
+
"crop": false,
|
| 60 |
+
"cut": false,
|
| 61 |
+
"cutoutProp": 0.5,
|
| 62 |
+
"randAugment": false,
|
| 63 |
+
"randAugment_magnitude": 10,
|
| 64 |
+
"mixup": false,
|
| 65 |
+
"mixupProp": 0.2,
|
| 66 |
+
"normalize": "std",
|
| 67 |
+
"truncate": true,
|
| 68 |
+
"verbose": true,
|
| 69 |
+
"metrics": [
|
| 70 |
+
"learningRate",
|
| 71 |
+
"loss",
|
| 72 |
+
"accuracy"
|
| 73 |
+
],
|
| 74 |
+
"logEach": 50,
|
| 75 |
+
"learningRate": 2e-05,
|
| 76 |
+
"lrScheduler": [
|
| 77 |
+
"const"
|
| 78 |
+
],
|
| 79 |
+
"LRScheduler_step_steps": 3.0,
|
| 80 |
+
"LRScheduler_step_gamma": 0.2,
|
| 81 |
+
"LRScheduler_exp_maxDecay": 0.01,
|
| 82 |
+
"LRScheduler_cos_periods": 0.5,
|
| 83 |
+
"LRScheduler_StartRamp_epochs": 1.0,
|
| 84 |
+
"LRScheduler_WRN_T0": 50.0,
|
| 85 |
+
"LRScheduler_WRN_Tmult": 1.2,
|
| 86 |
+
"rhoScheduler": [
|
| 87 |
+
"const"
|
| 88 |
+
],
|
| 89 |
+
"rhoScheduler_StartRamp_epochs": 8.0,
|
| 90 |
+
"rhoScheduler_StartJump_epochs": 8.0,
|
| 91 |
+
"saveCheckpoint": false,
|
| 92 |
+
"keepLastCheckpoint": false,
|
| 93 |
+
"checkpointsList": [],
|
| 94 |
+
"saveCheckpointInterval": 0,
|
| 95 |
+
"logDir": "/mnt/hungpv/projects/SAM_Variant/logs_nlp",
|
| 96 |
+
"logSubDir": "glue_bert",
|
| 97 |
+
"epochs": 3,
|
| 98 |
+
"contin": false
|
| 99 |
+
}
|
cola__SAM__lr2e-05__bs32__ep3__rho0.002/checkpoint_final.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:6966b8c6c5229ea00168ec0ef1a5087b243b5a9b0e5732387c5302cfcc16998b
|
| 3 |
+
size 438019989
|
cola__SAM__lr2e-05__bs32__ep3__rho0.002/metrics_history.json
ADDED
|
@@ -0,0 +1,14 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
[
|
| 2 |
+
{
|
| 3 |
+
"epoch": 1,
|
| 4 |
+
"mcc": 4.191
|
| 5 |
+
},
|
| 6 |
+
{
|
| 7 |
+
"epoch": 2,
|
| 8 |
+
"mcc": -2.07
|
| 9 |
+
},
|
| 10 |
+
{
|
| 11 |
+
"epoch": 3,
|
| 12 |
+
"mcc": -2.07
|
| 13 |
+
}
|
| 14 |
+
]
|
cola__SAM__lr2e-05__bs32__ep3__rho0.002/params.json
ADDED
|
@@ -0,0 +1,99 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"label_smoothing": 0.0,
|
| 3 |
+
"ESAMbeta": 0.5,
|
| 4 |
+
"ESAMgamma": 0.5,
|
| 5 |
+
"lookSAMk": 5,
|
| 6 |
+
"lookSAMalpha": 0.7,
|
| 7 |
+
"lookSAMLayerwise": false,
|
| 8 |
+
"optimizer": "SAM",
|
| 9 |
+
"weightDecay": 0.01,
|
| 10 |
+
"momentum": 0.9,
|
| 11 |
+
"rho": 0.002,
|
| 12 |
+
"samRho": 0.002,
|
| 13 |
+
"asamRho": 0.002,
|
| 14 |
+
"esamRho": 0.002,
|
| 15 |
+
"looksamRho": 0.002,
|
| 16 |
+
"friendlysamRho": 0.002,
|
| 17 |
+
"fishersamRho": 0.002,
|
| 18 |
+
"mongesamRho": 0.002,
|
| 19 |
+
"gsamRho": 0.002,
|
| 20 |
+
"bayesiansamRho": 0.002,
|
| 21 |
+
"msamRho": 0.002,
|
| 22 |
+
"adamwSamRho": 0.002,
|
| 23 |
+
"adamSamRho": 0.002,
|
| 24 |
+
"adamwMsamRho": 0.002,
|
| 25 |
+
"adamwFriendlysamRho": 0.002,
|
| 26 |
+
"sigma": 1.0,
|
| 27 |
+
"lmbda": 0.9,
|
| 28 |
+
"fisherEta": 1.0,
|
| 29 |
+
"gsamAlpha": 0.01,
|
| 30 |
+
"gsamAdaptive": false,
|
| 31 |
+
"bayesMSharpness": 8,
|
| 32 |
+
"bayesDelta": 10.0,
|
| 33 |
+
"bayesBeta1": 0.9,
|
| 34 |
+
"bayesBeta2": 0.999,
|
| 35 |
+
"bayesSInit": 1.0,
|
| 36 |
+
"bayesDamping": 1e-08,
|
| 37 |
+
"bayesEps": 1e-12,
|
| 38 |
+
"nesterov": false,
|
| 39 |
+
"grad_clip_norm": 1.0,
|
| 40 |
+
"rndSeed": true,
|
| 41 |
+
"deterministic": false,
|
| 42 |
+
"glue_task": "cola",
|
| 43 |
+
"glue_output_root": "/mnt/hungpv/projects/SAM_Variant/logs_nlp",
|
| 44 |
+
"widthFactor": 4,
|
| 45 |
+
"ViTPatchsize": 32,
|
| 46 |
+
"ViTSize": "S",
|
| 47 |
+
"pyramidAlpha": 270,
|
| 48 |
+
"pyramidBottleneck": true,
|
| 49 |
+
"model": "WRN",
|
| 50 |
+
"dropout": 0.0,
|
| 51 |
+
"BN": true,
|
| 52 |
+
"depth": 16,
|
| 53 |
+
"dataThreads": 4,
|
| 54 |
+
"dataDir": "/home/hungpv/.datasets",
|
| 55 |
+
"batchSize": 32,
|
| 56 |
+
"dataset": "CIFAR100",
|
| 57 |
+
"imageSize": 0,
|
| 58 |
+
"flip": false,
|
| 59 |
+
"crop": false,
|
| 60 |
+
"cut": false,
|
| 61 |
+
"cutoutProp": 0.5,
|
| 62 |
+
"randAugment": false,
|
| 63 |
+
"randAugment_magnitude": 10,
|
| 64 |
+
"mixup": false,
|
| 65 |
+
"mixupProp": 0.2,
|
| 66 |
+
"normalize": "std",
|
| 67 |
+
"truncate": true,
|
| 68 |
+
"verbose": true,
|
| 69 |
+
"metrics": [
|
| 70 |
+
"learningRate",
|
| 71 |
+
"loss",
|
| 72 |
+
"accuracy"
|
| 73 |
+
],
|
| 74 |
+
"logEach": 50,
|
| 75 |
+
"learningRate": 2e-05,
|
| 76 |
+
"lrScheduler": [
|
| 77 |
+
"const"
|
| 78 |
+
],
|
| 79 |
+
"LRScheduler_step_steps": 3.0,
|
| 80 |
+
"LRScheduler_step_gamma": 0.2,
|
| 81 |
+
"LRScheduler_exp_maxDecay": 0.01,
|
| 82 |
+
"LRScheduler_cos_periods": 0.5,
|
| 83 |
+
"LRScheduler_StartRamp_epochs": 1.0,
|
| 84 |
+
"LRScheduler_WRN_T0": 50.0,
|
| 85 |
+
"LRScheduler_WRN_Tmult": 1.2,
|
| 86 |
+
"rhoScheduler": [
|
| 87 |
+
"const"
|
| 88 |
+
],
|
| 89 |
+
"rhoScheduler_StartRamp_epochs": 8.0,
|
| 90 |
+
"rhoScheduler_StartJump_epochs": 8.0,
|
| 91 |
+
"saveCheckpoint": false,
|
| 92 |
+
"keepLastCheckpoint": false,
|
| 93 |
+
"checkpointsList": [],
|
| 94 |
+
"saveCheckpointInterval": 0,
|
| 95 |
+
"logDir": "/mnt/hungpv/projects/SAM_Variant/logs_nlp",
|
| 96 |
+
"logSubDir": "glue_bert",
|
| 97 |
+
"epochs": 3,
|
| 98 |
+
"contin": false
|
| 99 |
+
}
|
cola__SGD__lr2e-05__bs32__ep3__rho_na/checkpoint_final.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:14b8edf27a50b5e2e8210b03ee1b852ab99c864752ec2a0003a99a3060a18084
|
| 3 |
+
size 438019989
|
cola__SGD__lr2e-05__bs32__ep3__rho_na/metrics_history.json
ADDED
|
@@ -0,0 +1,14 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
[
|
| 2 |
+
{
|
| 3 |
+
"epoch": 1,
|
| 4 |
+
"mcc": 1.815
|
| 5 |
+
},
|
| 6 |
+
{
|
| 7 |
+
"epoch": 2,
|
| 8 |
+
"mcc": 1.815
|
| 9 |
+
},
|
| 10 |
+
{
|
| 11 |
+
"epoch": 3,
|
| 12 |
+
"mcc": 1.815
|
| 13 |
+
}
|
| 14 |
+
]
|
cola__SGD__lr2e-05__bs32__ep3__rho_na/params.json
ADDED
|
@@ -0,0 +1,98 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"label_smoothing": 0.0,
|
| 3 |
+
"ESAMbeta": 0.5,
|
| 4 |
+
"ESAMgamma": 0.5,
|
| 5 |
+
"lookSAMk": 5,
|
| 6 |
+
"lookSAMalpha": 0.7,
|
| 7 |
+
"lookSAMLayerwise": false,
|
| 8 |
+
"optimizer": "SGD",
|
| 9 |
+
"weightDecay": 0.01,
|
| 10 |
+
"momentum": 0.9,
|
| 11 |
+
"rho": 0.05,
|
| 12 |
+
"samRho": 0.05,
|
| 13 |
+
"asamRho": 0.05,
|
| 14 |
+
"esamRho": 0.05,
|
| 15 |
+
"looksamRho": 0.05,
|
| 16 |
+
"friendlysamRho": 0.05,
|
| 17 |
+
"fishersamRho": 0.05,
|
| 18 |
+
"mongesamRho": 0.05,
|
| 19 |
+
"gsamRho": 0.05,
|
| 20 |
+
"bayesiansamRho": 0.05,
|
| 21 |
+
"msamRho": 0.05,
|
| 22 |
+
"adamwSamRho": 0.05,
|
| 23 |
+
"adamwMsamRho": 0.05,
|
| 24 |
+
"adamwFriendlysamRho": 0.05,
|
| 25 |
+
"sigma": 1.0,
|
| 26 |
+
"lmbda": 0.9,
|
| 27 |
+
"fisherEta": 1.0,
|
| 28 |
+
"gsamAlpha": 0.01,
|
| 29 |
+
"gsamAdaptive": false,
|
| 30 |
+
"bayesMSharpness": 8,
|
| 31 |
+
"bayesDelta": 10.0,
|
| 32 |
+
"bayesBeta1": 0.9,
|
| 33 |
+
"bayesBeta2": 0.999,
|
| 34 |
+
"bayesSInit": 1.0,
|
| 35 |
+
"bayesDamping": 1e-08,
|
| 36 |
+
"bayesEps": 1e-12,
|
| 37 |
+
"nesterov": false,
|
| 38 |
+
"grad_clip_norm": 1.0,
|
| 39 |
+
"rndSeed": true,
|
| 40 |
+
"deterministic": false,
|
| 41 |
+
"glue_task": "cola",
|
| 42 |
+
"glue_output_root": "/mnt/hungpv/projects/SAM_Variant/logs_nlp",
|
| 43 |
+
"widthFactor": 4,
|
| 44 |
+
"ViTPatchsize": 32,
|
| 45 |
+
"ViTSize": "S",
|
| 46 |
+
"pyramidAlpha": 270,
|
| 47 |
+
"pyramidBottleneck": true,
|
| 48 |
+
"model": "WRN",
|
| 49 |
+
"dropout": 0.0,
|
| 50 |
+
"BN": true,
|
| 51 |
+
"depth": 16,
|
| 52 |
+
"dataThreads": 4,
|
| 53 |
+
"dataDir": "/home/hungpv/.datasets",
|
| 54 |
+
"batchSize": 32,
|
| 55 |
+
"dataset": "CIFAR100",
|
| 56 |
+
"imageSize": 0,
|
| 57 |
+
"flip": false,
|
| 58 |
+
"crop": false,
|
| 59 |
+
"cut": false,
|
| 60 |
+
"cutoutProp": 0.5,
|
| 61 |
+
"randAugment": false,
|
| 62 |
+
"randAugment_magnitude": 10,
|
| 63 |
+
"mixup": false,
|
| 64 |
+
"mixupProp": 0.2,
|
| 65 |
+
"normalize": "std",
|
| 66 |
+
"truncate": true,
|
| 67 |
+
"verbose": true,
|
| 68 |
+
"metrics": [
|
| 69 |
+
"learningRate",
|
| 70 |
+
"loss",
|
| 71 |
+
"accuracy"
|
| 72 |
+
],
|
| 73 |
+
"logEach": 50,
|
| 74 |
+
"learningRate": 2e-05,
|
| 75 |
+
"lrScheduler": [
|
| 76 |
+
"const"
|
| 77 |
+
],
|
| 78 |
+
"LRScheduler_step_steps": 3.0,
|
| 79 |
+
"LRScheduler_step_gamma": 0.2,
|
| 80 |
+
"LRScheduler_exp_maxDecay": 0.01,
|
| 81 |
+
"LRScheduler_cos_periods": 0.5,
|
| 82 |
+
"LRScheduler_StartRamp_epochs": 1.0,
|
| 83 |
+
"LRScheduler_WRN_T0": 50.0,
|
| 84 |
+
"LRScheduler_WRN_Tmult": 1.2,
|
| 85 |
+
"rhoScheduler": [
|
| 86 |
+
"const"
|
| 87 |
+
],
|
| 88 |
+
"rhoScheduler_StartRamp_epochs": 8.0,
|
| 89 |
+
"rhoScheduler_StartJump_epochs": 8.0,
|
| 90 |
+
"saveCheckpoint": false,
|
| 91 |
+
"keepLastCheckpoint": false,
|
| 92 |
+
"checkpointsList": [],
|
| 93 |
+
"saveCheckpointInterval": 0,
|
| 94 |
+
"logDir": "/mnt/hungpv/projects/SAM_Variant/logs_nlp",
|
| 95 |
+
"logSubDir": "glue_bert",
|
| 96 |
+
"epochs": 3,
|
| 97 |
+
"contin": false
|
| 98 |
+
}
|
mnli__SAM_Muon__lr2e-05__bs32__ep3__rho0.005/checkpoint_final.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:17e9f5f5a3d3849a252583b7fb6620bba91d177646f934c7152b694e113f3097
|
| 3 |
+
size 438023061
|
mnli__SAM_Muon__lr2e-05__bs32__ep3__rho0.005/metrics_history.json
ADDED
|
@@ -0,0 +1,20 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
[
|
| 2 |
+
{
|
| 3 |
+
"epoch": 1,
|
| 4 |
+
"matched_accuracy": 66.623,
|
| 5 |
+
"mismatched_accuracy": 68.033,
|
| 6 |
+
"mnli_accuracy_mean": 67.328
|
| 7 |
+
},
|
| 8 |
+
{
|
| 9 |
+
"epoch": 2,
|
| 10 |
+
"matched_accuracy": 68.813,
|
| 11 |
+
"mismatched_accuracy": 70.179,
|
| 12 |
+
"mnli_accuracy_mean": 69.496
|
| 13 |
+
},
|
| 14 |
+
{
|
| 15 |
+
"epoch": 3,
|
| 16 |
+
"matched_accuracy": 69.791,
|
| 17 |
+
"mismatched_accuracy": 70.962,
|
| 18 |
+
"mnli_accuracy_mean": 70.376
|
| 19 |
+
}
|
| 20 |
+
]
|
mnli__SAM_Muon__lr2e-05__bs32__ep3__rho0.005/params.json
ADDED
|
@@ -0,0 +1,99 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"label_smoothing": 0.0,
|
| 3 |
+
"ESAMbeta": 0.5,
|
| 4 |
+
"ESAMgamma": 0.5,
|
| 5 |
+
"lookSAMk": 5,
|
| 6 |
+
"lookSAMalpha": 0.7,
|
| 7 |
+
"lookSAMLayerwise": false,
|
| 8 |
+
"optimizer": "SAM_Muon",
|
| 9 |
+
"weightDecay": 0.01,
|
| 10 |
+
"momentum": 0.9,
|
| 11 |
+
"rho": 0.005,
|
| 12 |
+
"samRho": 0.005,
|
| 13 |
+
"asamRho": 0.005,
|
| 14 |
+
"esamRho": 0.005,
|
| 15 |
+
"looksamRho": 0.005,
|
| 16 |
+
"friendlysamRho": 0.005,
|
| 17 |
+
"fishersamRho": 0.005,
|
| 18 |
+
"mongesamRho": 0.005,
|
| 19 |
+
"gsamRho": 0.005,
|
| 20 |
+
"bayesiansamRho": 0.005,
|
| 21 |
+
"msamRho": 0.005,
|
| 22 |
+
"adamwSamRho": 0.005,
|
| 23 |
+
"adamSamRho": 0.005,
|
| 24 |
+
"adamwMsamRho": 0.005,
|
| 25 |
+
"adamwFriendlysamRho": 0.005,
|
| 26 |
+
"sigma": 1.0,
|
| 27 |
+
"lmbda": 0.9,
|
| 28 |
+
"fisherEta": 1.0,
|
| 29 |
+
"gsamAlpha": 0.01,
|
| 30 |
+
"gsamAdaptive": false,
|
| 31 |
+
"bayesMSharpness": 8,
|
| 32 |
+
"bayesDelta": 10.0,
|
| 33 |
+
"bayesBeta1": 0.9,
|
| 34 |
+
"bayesBeta2": 0.999,
|
| 35 |
+
"bayesSInit": 1.0,
|
| 36 |
+
"bayesDamping": 1e-08,
|
| 37 |
+
"bayesEps": 1e-12,
|
| 38 |
+
"nesterov": false,
|
| 39 |
+
"grad_clip_norm": 1.0,
|
| 40 |
+
"rndSeed": true,
|
| 41 |
+
"deterministic": false,
|
| 42 |
+
"glue_task": "mnli",
|
| 43 |
+
"glue_output_root": "/mnt/hungpv/projects/SAM_Variant/logs_nlp",
|
| 44 |
+
"widthFactor": 4,
|
| 45 |
+
"ViTPatchsize": 32,
|
| 46 |
+
"ViTSize": "S",
|
| 47 |
+
"pyramidAlpha": 270,
|
| 48 |
+
"pyramidBottleneck": true,
|
| 49 |
+
"model": "WRN",
|
| 50 |
+
"dropout": 0.0,
|
| 51 |
+
"BN": true,
|
| 52 |
+
"depth": 16,
|
| 53 |
+
"dataThreads": 4,
|
| 54 |
+
"dataDir": "/home/hungpv/.datasets",
|
| 55 |
+
"batchSize": 32,
|
| 56 |
+
"dataset": "CIFAR100",
|
| 57 |
+
"imageSize": 0,
|
| 58 |
+
"flip": false,
|
| 59 |
+
"crop": false,
|
| 60 |
+
"cut": false,
|
| 61 |
+
"cutoutProp": 0.5,
|
| 62 |
+
"randAugment": false,
|
| 63 |
+
"randAugment_magnitude": 10,
|
| 64 |
+
"mixup": false,
|
| 65 |
+
"mixupProp": 0.2,
|
| 66 |
+
"normalize": "std",
|
| 67 |
+
"truncate": true,
|
| 68 |
+
"verbose": true,
|
| 69 |
+
"metrics": [
|
| 70 |
+
"learningRate",
|
| 71 |
+
"loss",
|
| 72 |
+
"accuracy"
|
| 73 |
+
],
|
| 74 |
+
"logEach": 50,
|
| 75 |
+
"learningRate": 2e-05,
|
| 76 |
+
"lrScheduler": [
|
| 77 |
+
"const"
|
| 78 |
+
],
|
| 79 |
+
"LRScheduler_step_steps": 3.0,
|
| 80 |
+
"LRScheduler_step_gamma": 0.2,
|
| 81 |
+
"LRScheduler_exp_maxDecay": 0.01,
|
| 82 |
+
"LRScheduler_cos_periods": 0.5,
|
| 83 |
+
"LRScheduler_StartRamp_epochs": 1.0,
|
| 84 |
+
"LRScheduler_WRN_T0": 50.0,
|
| 85 |
+
"LRScheduler_WRN_Tmult": 1.2,
|
| 86 |
+
"rhoScheduler": [
|
| 87 |
+
"const"
|
| 88 |
+
],
|
| 89 |
+
"rhoScheduler_StartRamp_epochs": 8.0,
|
| 90 |
+
"rhoScheduler_StartJump_epochs": 8.0,
|
| 91 |
+
"saveCheckpoint": false,
|
| 92 |
+
"keepLastCheckpoint": false,
|
| 93 |
+
"checkpointsList": [],
|
| 94 |
+
"saveCheckpointInterval": 0,
|
| 95 |
+
"logDir": "/mnt/hungpv/projects/SAM_Variant/logs_nlp",
|
| 96 |
+
"logSubDir": "glue_bert",
|
| 97 |
+
"epochs": 3,
|
| 98 |
+
"contin": false
|
| 99 |
+
}
|
mnli__SAM__lr2e-05__bs32__ep3__rho0.005/checkpoint_final.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:e3745554ca0999b3436bc17b776218ff325eb06d836d50cf64f6e177bca40b2f
|
| 3 |
+
size 438023061
|
mnli__SAM__lr2e-05__bs32__ep3__rho0.005/metrics_history.json
ADDED
|
@@ -0,0 +1,20 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
[
|
| 2 |
+
{
|
| 3 |
+
"epoch": 1,
|
| 4 |
+
"matched_accuracy": 42.649,
|
| 5 |
+
"mismatched_accuracy": 43.379,
|
| 6 |
+
"mnli_accuracy_mean": 43.014
|
| 7 |
+
},
|
| 8 |
+
{
|
| 9 |
+
"epoch": 2,
|
| 10 |
+
"matched_accuracy": 47.845,
|
| 11 |
+
"mismatched_accuracy": 49.247,
|
| 12 |
+
"mnli_accuracy_mean": 48.546
|
| 13 |
+
},
|
| 14 |
+
{
|
| 15 |
+
"epoch": 3,
|
| 16 |
+
"matched_accuracy": 49.292,
|
| 17 |
+
"mismatched_accuracy": 50.417,
|
| 18 |
+
"mnli_accuracy_mean": 49.855
|
| 19 |
+
}
|
| 20 |
+
]
|
mnli__SAM__lr2e-05__bs32__ep3__rho0.005/params.json
ADDED
|
@@ -0,0 +1,99 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"label_smoothing": 0.0,
|
| 3 |
+
"ESAMbeta": 0.5,
|
| 4 |
+
"ESAMgamma": 0.5,
|
| 5 |
+
"lookSAMk": 5,
|
| 6 |
+
"lookSAMalpha": 0.7,
|
| 7 |
+
"lookSAMLayerwise": false,
|
| 8 |
+
"optimizer": "SAM",
|
| 9 |
+
"weightDecay": 0.01,
|
| 10 |
+
"momentum": 0.9,
|
| 11 |
+
"rho": 0.005,
|
| 12 |
+
"samRho": 0.005,
|
| 13 |
+
"asamRho": 0.005,
|
| 14 |
+
"esamRho": 0.005,
|
| 15 |
+
"looksamRho": 0.005,
|
| 16 |
+
"friendlysamRho": 0.005,
|
| 17 |
+
"fishersamRho": 0.005,
|
| 18 |
+
"mongesamRho": 0.005,
|
| 19 |
+
"gsamRho": 0.005,
|
| 20 |
+
"bayesiansamRho": 0.005,
|
| 21 |
+
"msamRho": 0.005,
|
| 22 |
+
"adamwSamRho": 0.005,
|
| 23 |
+
"adamSamRho": 0.005,
|
| 24 |
+
"adamwMsamRho": 0.005,
|
| 25 |
+
"adamwFriendlysamRho": 0.005,
|
| 26 |
+
"sigma": 1.0,
|
| 27 |
+
"lmbda": 0.9,
|
| 28 |
+
"fisherEta": 1.0,
|
| 29 |
+
"gsamAlpha": 0.01,
|
| 30 |
+
"gsamAdaptive": false,
|
| 31 |
+
"bayesMSharpness": 8,
|
| 32 |
+
"bayesDelta": 10.0,
|
| 33 |
+
"bayesBeta1": 0.9,
|
| 34 |
+
"bayesBeta2": 0.999,
|
| 35 |
+
"bayesSInit": 1.0,
|
| 36 |
+
"bayesDamping": 1e-08,
|
| 37 |
+
"bayesEps": 1e-12,
|
| 38 |
+
"nesterov": false,
|
| 39 |
+
"grad_clip_norm": 1.0,
|
| 40 |
+
"rndSeed": true,
|
| 41 |
+
"deterministic": false,
|
| 42 |
+
"glue_task": "mnli",
|
| 43 |
+
"glue_output_root": "/mnt/hungpv/projects/SAM_Variant/logs_nlp",
|
| 44 |
+
"widthFactor": 4,
|
| 45 |
+
"ViTPatchsize": 32,
|
| 46 |
+
"ViTSize": "S",
|
| 47 |
+
"pyramidAlpha": 270,
|
| 48 |
+
"pyramidBottleneck": true,
|
| 49 |
+
"model": "WRN",
|
| 50 |
+
"dropout": 0.0,
|
| 51 |
+
"BN": true,
|
| 52 |
+
"depth": 16,
|
| 53 |
+
"dataThreads": 4,
|
| 54 |
+
"dataDir": "/home/hungpv/.datasets",
|
| 55 |
+
"batchSize": 32,
|
| 56 |
+
"dataset": "CIFAR100",
|
| 57 |
+
"imageSize": 0,
|
| 58 |
+
"flip": false,
|
| 59 |
+
"crop": false,
|
| 60 |
+
"cut": false,
|
| 61 |
+
"cutoutProp": 0.5,
|
| 62 |
+
"randAugment": false,
|
| 63 |
+
"randAugment_magnitude": 10,
|
| 64 |
+
"mixup": false,
|
| 65 |
+
"mixupProp": 0.2,
|
| 66 |
+
"normalize": "std",
|
| 67 |
+
"truncate": true,
|
| 68 |
+
"verbose": true,
|
| 69 |
+
"metrics": [
|
| 70 |
+
"learningRate",
|
| 71 |
+
"loss",
|
| 72 |
+
"accuracy"
|
| 73 |
+
],
|
| 74 |
+
"logEach": 50,
|
| 75 |
+
"learningRate": 2e-05,
|
| 76 |
+
"lrScheduler": [
|
| 77 |
+
"const"
|
| 78 |
+
],
|
| 79 |
+
"LRScheduler_step_steps": 3.0,
|
| 80 |
+
"LRScheduler_step_gamma": 0.2,
|
| 81 |
+
"LRScheduler_exp_maxDecay": 0.01,
|
| 82 |
+
"LRScheduler_cos_periods": 0.5,
|
| 83 |
+
"LRScheduler_StartRamp_epochs": 1.0,
|
| 84 |
+
"LRScheduler_WRN_T0": 50.0,
|
| 85 |
+
"LRScheduler_WRN_Tmult": 1.2,
|
| 86 |
+
"rhoScheduler": [
|
| 87 |
+
"const"
|
| 88 |
+
],
|
| 89 |
+
"rhoScheduler_StartRamp_epochs": 8.0,
|
| 90 |
+
"rhoScheduler_StartJump_epochs": 8.0,
|
| 91 |
+
"saveCheckpoint": false,
|
| 92 |
+
"keepLastCheckpoint": false,
|
| 93 |
+
"checkpointsList": [],
|
| 94 |
+
"saveCheckpointInterval": 0,
|
| 95 |
+
"logDir": "/mnt/hungpv/projects/SAM_Variant/logs_nlp",
|
| 96 |
+
"logSubDir": "glue_bert",
|
| 97 |
+
"epochs": 3,
|
| 98 |
+
"contin": false
|
| 99 |
+
}
|
mnli__SGD__lr2e-05__bs32__ep3__rho_na/checkpoint_final.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:14e83e5cf89e24acadd580cce15adbdfece5de884d855088916bf85e93cdb4cc
|
| 3 |
+
size 438023061
|
mnli__SGD__lr2e-05__bs32__ep3__rho_na/metrics_history.json
ADDED
|
@@ -0,0 +1,20 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
[
|
| 2 |
+
{
|
| 3 |
+
"epoch": 1,
|
| 4 |
+
"matched_accuracy": 43.617,
|
| 5 |
+
"mismatched_accuracy": 45.362,
|
| 6 |
+
"mnli_accuracy_mean": 44.489
|
| 7 |
+
},
|
| 8 |
+
{
|
| 9 |
+
"epoch": 2,
|
| 10 |
+
"matched_accuracy": 48.548,
|
| 11 |
+
"mismatched_accuracy": 50.376,
|
| 12 |
+
"mnli_accuracy_mean": 49.462
|
| 13 |
+
},
|
| 14 |
+
{
|
| 15 |
+
"epoch": 3,
|
| 16 |
+
"matched_accuracy": 49.71,
|
| 17 |
+
"mismatched_accuracy": 51.221,
|
| 18 |
+
"mnli_accuracy_mean": 50.465
|
| 19 |
+
}
|
| 20 |
+
]
|
mnli__SGD__lr2e-05__bs32__ep3__rho_na/params.json
ADDED
|
@@ -0,0 +1,98 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"label_smoothing": 0.0,
|
| 3 |
+
"ESAMbeta": 0.5,
|
| 4 |
+
"ESAMgamma": 0.5,
|
| 5 |
+
"lookSAMk": 5,
|
| 6 |
+
"lookSAMalpha": 0.7,
|
| 7 |
+
"lookSAMLayerwise": false,
|
| 8 |
+
"optimizer": "SGD",
|
| 9 |
+
"weightDecay": 0.01,
|
| 10 |
+
"momentum": 0.9,
|
| 11 |
+
"rho": 0.05,
|
| 12 |
+
"samRho": 0.05,
|
| 13 |
+
"asamRho": 0.05,
|
| 14 |
+
"esamRho": 0.05,
|
| 15 |
+
"looksamRho": 0.05,
|
| 16 |
+
"friendlysamRho": 0.05,
|
| 17 |
+
"fishersamRho": 0.05,
|
| 18 |
+
"mongesamRho": 0.05,
|
| 19 |
+
"gsamRho": 0.05,
|
| 20 |
+
"bayesiansamRho": 0.05,
|
| 21 |
+
"msamRho": 0.05,
|
| 22 |
+
"adamwSamRho": 0.05,
|
| 23 |
+
"adamwMsamRho": 0.05,
|
| 24 |
+
"adamwFriendlysamRho": 0.05,
|
| 25 |
+
"sigma": 1.0,
|
| 26 |
+
"lmbda": 0.9,
|
| 27 |
+
"fisherEta": 1.0,
|
| 28 |
+
"gsamAlpha": 0.01,
|
| 29 |
+
"gsamAdaptive": false,
|
| 30 |
+
"bayesMSharpness": 8,
|
| 31 |
+
"bayesDelta": 10.0,
|
| 32 |
+
"bayesBeta1": 0.9,
|
| 33 |
+
"bayesBeta2": 0.999,
|
| 34 |
+
"bayesSInit": 1.0,
|
| 35 |
+
"bayesDamping": 1e-08,
|
| 36 |
+
"bayesEps": 1e-12,
|
| 37 |
+
"nesterov": false,
|
| 38 |
+
"grad_clip_norm": 1.0,
|
| 39 |
+
"rndSeed": true,
|
| 40 |
+
"deterministic": false,
|
| 41 |
+
"glue_task": "mnli",
|
| 42 |
+
"glue_output_root": "/mnt/hungpv/projects/SAM_Variant/logs_nlp",
|
| 43 |
+
"widthFactor": 4,
|
| 44 |
+
"ViTPatchsize": 32,
|
| 45 |
+
"ViTSize": "S",
|
| 46 |
+
"pyramidAlpha": 270,
|
| 47 |
+
"pyramidBottleneck": true,
|
| 48 |
+
"model": "WRN",
|
| 49 |
+
"dropout": 0.0,
|
| 50 |
+
"BN": true,
|
| 51 |
+
"depth": 16,
|
| 52 |
+
"dataThreads": 4,
|
| 53 |
+
"dataDir": "/home/hungpv/.datasets",
|
| 54 |
+
"batchSize": 32,
|
| 55 |
+
"dataset": "CIFAR100",
|
| 56 |
+
"imageSize": 0,
|
| 57 |
+
"flip": false,
|
| 58 |
+
"crop": false,
|
| 59 |
+
"cut": false,
|
| 60 |
+
"cutoutProp": 0.5,
|
| 61 |
+
"randAugment": false,
|
| 62 |
+
"randAugment_magnitude": 10,
|
| 63 |
+
"mixup": false,
|
| 64 |
+
"mixupProp": 0.2,
|
| 65 |
+
"normalize": "std",
|
| 66 |
+
"truncate": true,
|
| 67 |
+
"verbose": true,
|
| 68 |
+
"metrics": [
|
| 69 |
+
"learningRate",
|
| 70 |
+
"loss",
|
| 71 |
+
"accuracy"
|
| 72 |
+
],
|
| 73 |
+
"logEach": 50,
|
| 74 |
+
"learningRate": 2e-05,
|
| 75 |
+
"lrScheduler": [
|
| 76 |
+
"const"
|
| 77 |
+
],
|
| 78 |
+
"LRScheduler_step_steps": 3.0,
|
| 79 |
+
"LRScheduler_step_gamma": 0.2,
|
| 80 |
+
"LRScheduler_exp_maxDecay": 0.01,
|
| 81 |
+
"LRScheduler_cos_periods": 0.5,
|
| 82 |
+
"LRScheduler_StartRamp_epochs": 1.0,
|
| 83 |
+
"LRScheduler_WRN_T0": 50.0,
|
| 84 |
+
"LRScheduler_WRN_Tmult": 1.2,
|
| 85 |
+
"rhoScheduler": [
|
| 86 |
+
"const"
|
| 87 |
+
],
|
| 88 |
+
"rhoScheduler_StartRamp_epochs": 8.0,
|
| 89 |
+
"rhoScheduler_StartJump_epochs": 8.0,
|
| 90 |
+
"saveCheckpoint": false,
|
| 91 |
+
"keepLastCheckpoint": false,
|
| 92 |
+
"checkpointsList": [],
|
| 93 |
+
"saveCheckpointInterval": 0,
|
| 94 |
+
"logDir": "/mnt/hungpv/projects/SAM_Variant/logs_nlp",
|
| 95 |
+
"logSubDir": "glue_bert",
|
| 96 |
+
"epochs": 3,
|
| 97 |
+
"contin": false
|
| 98 |
+
}
|
mrpc__SAM_Muon__lr2e-05__bs32__ep3__rho0.005/checkpoint_final.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:4616778e98e335762523a8dfba87e9d3ef970d806a2282cd1c69784a08436b6a
|
| 3 |
+
size 438019989
|
mrpc__SAM_Muon__lr2e-05__bs32__ep3__rho0.005/metrics_history.json
ADDED
|
@@ -0,0 +1,17 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
[
|
| 2 |
+
{
|
| 3 |
+
"epoch": 1,
|
| 4 |
+
"accuracy": 68.382,
|
| 5 |
+
"f1": 81.223
|
| 6 |
+
},
|
| 7 |
+
{
|
| 8 |
+
"epoch": 2,
|
| 9 |
+
"accuracy": 68.382,
|
| 10 |
+
"f1": 81.223
|
| 11 |
+
},
|
| 12 |
+
{
|
| 13 |
+
"epoch": 3,
|
| 14 |
+
"accuracy": 68.382,
|
| 15 |
+
"f1": 81.223
|
| 16 |
+
}
|
| 17 |
+
]
|
mrpc__SAM_Muon__lr2e-05__bs32__ep3__rho0.005/params.json
ADDED
|
@@ -0,0 +1,99 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"label_smoothing": 0.0,
|
| 3 |
+
"ESAMbeta": 0.5,
|
| 4 |
+
"ESAMgamma": 0.5,
|
| 5 |
+
"lookSAMk": 5,
|
| 6 |
+
"lookSAMalpha": 0.7,
|
| 7 |
+
"lookSAMLayerwise": false,
|
| 8 |
+
"optimizer": "SAM_Muon",
|
| 9 |
+
"weightDecay": 0.01,
|
| 10 |
+
"momentum": 0.9,
|
| 11 |
+
"rho": 0.005,
|
| 12 |
+
"samRho": 0.005,
|
| 13 |
+
"asamRho": 0.005,
|
| 14 |
+
"esamRho": 0.005,
|
| 15 |
+
"looksamRho": 0.005,
|
| 16 |
+
"friendlysamRho": 0.005,
|
| 17 |
+
"fishersamRho": 0.005,
|
| 18 |
+
"mongesamRho": 0.005,
|
| 19 |
+
"gsamRho": 0.005,
|
| 20 |
+
"bayesiansamRho": 0.005,
|
| 21 |
+
"msamRho": 0.005,
|
| 22 |
+
"adamwSamRho": 0.005,
|
| 23 |
+
"adamSamRho": 0.005,
|
| 24 |
+
"adamwMsamRho": 0.005,
|
| 25 |
+
"adamwFriendlysamRho": 0.005,
|
| 26 |
+
"sigma": 1.0,
|
| 27 |
+
"lmbda": 0.9,
|
| 28 |
+
"fisherEta": 1.0,
|
| 29 |
+
"gsamAlpha": 0.01,
|
| 30 |
+
"gsamAdaptive": false,
|
| 31 |
+
"bayesMSharpness": 8,
|
| 32 |
+
"bayesDelta": 10.0,
|
| 33 |
+
"bayesBeta1": 0.9,
|
| 34 |
+
"bayesBeta2": 0.999,
|
| 35 |
+
"bayesSInit": 1.0,
|
| 36 |
+
"bayesDamping": 1e-08,
|
| 37 |
+
"bayesEps": 1e-12,
|
| 38 |
+
"nesterov": false,
|
| 39 |
+
"grad_clip_norm": 1.0,
|
| 40 |
+
"rndSeed": true,
|
| 41 |
+
"deterministic": false,
|
| 42 |
+
"glue_task": "mrpc",
|
| 43 |
+
"glue_output_root": "/mnt/hungpv/projects/SAM_Variant/logs_nlp",
|
| 44 |
+
"widthFactor": 4,
|
| 45 |
+
"ViTPatchsize": 32,
|
| 46 |
+
"ViTSize": "S",
|
| 47 |
+
"pyramidAlpha": 270,
|
| 48 |
+
"pyramidBottleneck": true,
|
| 49 |
+
"model": "WRN",
|
| 50 |
+
"dropout": 0.0,
|
| 51 |
+
"BN": true,
|
| 52 |
+
"depth": 16,
|
| 53 |
+
"dataThreads": 4,
|
| 54 |
+
"dataDir": "/home/hungpv/.datasets",
|
| 55 |
+
"batchSize": 32,
|
| 56 |
+
"dataset": "CIFAR100",
|
| 57 |
+
"imageSize": 0,
|
| 58 |
+
"flip": false,
|
| 59 |
+
"crop": false,
|
| 60 |
+
"cut": false,
|
| 61 |
+
"cutoutProp": 0.5,
|
| 62 |
+
"randAugment": false,
|
| 63 |
+
"randAugment_magnitude": 10,
|
| 64 |
+
"mixup": false,
|
| 65 |
+
"mixupProp": 0.2,
|
| 66 |
+
"normalize": "std",
|
| 67 |
+
"truncate": true,
|
| 68 |
+
"verbose": true,
|
| 69 |
+
"metrics": [
|
| 70 |
+
"learningRate",
|
| 71 |
+
"loss",
|
| 72 |
+
"accuracy"
|
| 73 |
+
],
|
| 74 |
+
"logEach": 50,
|
| 75 |
+
"learningRate": 2e-05,
|
| 76 |
+
"lrScheduler": [
|
| 77 |
+
"const"
|
| 78 |
+
],
|
| 79 |
+
"LRScheduler_step_steps": 3.0,
|
| 80 |
+
"LRScheduler_step_gamma": 0.2,
|
| 81 |
+
"LRScheduler_exp_maxDecay": 0.01,
|
| 82 |
+
"LRScheduler_cos_periods": 0.5,
|
| 83 |
+
"LRScheduler_StartRamp_epochs": 1.0,
|
| 84 |
+
"LRScheduler_WRN_T0": 50.0,
|
| 85 |
+
"LRScheduler_WRN_Tmult": 1.2,
|
| 86 |
+
"rhoScheduler": [
|
| 87 |
+
"const"
|
| 88 |
+
],
|
| 89 |
+
"rhoScheduler_StartRamp_epochs": 8.0,
|
| 90 |
+
"rhoScheduler_StartJump_epochs": 8.0,
|
| 91 |
+
"saveCheckpoint": false,
|
| 92 |
+
"keepLastCheckpoint": false,
|
| 93 |
+
"checkpointsList": [],
|
| 94 |
+
"saveCheckpointInterval": 0,
|
| 95 |
+
"logDir": "/mnt/hungpv/projects/SAM_Variant/logs_nlp",
|
| 96 |
+
"logSubDir": "glue_bert",
|
| 97 |
+
"epochs": 3,
|
| 98 |
+
"contin": false
|
| 99 |
+
}
|
mrpc__SAM__lr2e-05__bs32__ep3__rho0.005/checkpoint_final.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:d29426cf54d9fe894382439ea33d4f823823c09f20426937ae04305780fbdf5c
|
| 3 |
+
size 438019989
|
mrpc__SAM__lr2e-05__bs32__ep3__rho0.005/metrics_history.json
ADDED
|
@@ -0,0 +1,17 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
[
|
| 2 |
+
{
|
| 3 |
+
"epoch": 1,
|
| 4 |
+
"accuracy": 68.137,
|
| 5 |
+
"f1": 81.05
|
| 6 |
+
},
|
| 7 |
+
{
|
| 8 |
+
"epoch": 2,
|
| 9 |
+
"accuracy": 68.382,
|
| 10 |
+
"f1": 81.223
|
| 11 |
+
},
|
| 12 |
+
{
|
| 13 |
+
"epoch": 3,
|
| 14 |
+
"accuracy": 68.382,
|
| 15 |
+
"f1": 81.223
|
| 16 |
+
}
|
| 17 |
+
]
|
mrpc__SAM__lr2e-05__bs32__ep3__rho0.005/params.json
ADDED
|
@@ -0,0 +1,99 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"label_smoothing": 0.0,
|
| 3 |
+
"ESAMbeta": 0.5,
|
| 4 |
+
"ESAMgamma": 0.5,
|
| 5 |
+
"lookSAMk": 5,
|
| 6 |
+
"lookSAMalpha": 0.7,
|
| 7 |
+
"lookSAMLayerwise": false,
|
| 8 |
+
"optimizer": "SAM",
|
| 9 |
+
"weightDecay": 0.01,
|
| 10 |
+
"momentum": 0.9,
|
| 11 |
+
"rho": 0.005,
|
| 12 |
+
"samRho": 0.005,
|
| 13 |
+
"asamRho": 0.005,
|
| 14 |
+
"esamRho": 0.005,
|
| 15 |
+
"looksamRho": 0.005,
|
| 16 |
+
"friendlysamRho": 0.005,
|
| 17 |
+
"fishersamRho": 0.005,
|
| 18 |
+
"mongesamRho": 0.005,
|
| 19 |
+
"gsamRho": 0.005,
|
| 20 |
+
"bayesiansamRho": 0.005,
|
| 21 |
+
"msamRho": 0.005,
|
| 22 |
+
"adamwSamRho": 0.005,
|
| 23 |
+
"adamSamRho": 0.005,
|
| 24 |
+
"adamwMsamRho": 0.005,
|
| 25 |
+
"adamwFriendlysamRho": 0.005,
|
| 26 |
+
"sigma": 1.0,
|
| 27 |
+
"lmbda": 0.9,
|
| 28 |
+
"fisherEta": 1.0,
|
| 29 |
+
"gsamAlpha": 0.01,
|
| 30 |
+
"gsamAdaptive": false,
|
| 31 |
+
"bayesMSharpness": 8,
|
| 32 |
+
"bayesDelta": 10.0,
|
| 33 |
+
"bayesBeta1": 0.9,
|
| 34 |
+
"bayesBeta2": 0.999,
|
| 35 |
+
"bayesSInit": 1.0,
|
| 36 |
+
"bayesDamping": 1e-08,
|
| 37 |
+
"bayesEps": 1e-12,
|
| 38 |
+
"nesterov": false,
|
| 39 |
+
"grad_clip_norm": 1.0,
|
| 40 |
+
"rndSeed": true,
|
| 41 |
+
"deterministic": false,
|
| 42 |
+
"glue_task": "mrpc",
|
| 43 |
+
"glue_output_root": "/mnt/hungpv/projects/SAM_Variant/logs_nlp",
|
| 44 |
+
"widthFactor": 4,
|
| 45 |
+
"ViTPatchsize": 32,
|
| 46 |
+
"ViTSize": "S",
|
| 47 |
+
"pyramidAlpha": 270,
|
| 48 |
+
"pyramidBottleneck": true,
|
| 49 |
+
"model": "WRN",
|
| 50 |
+
"dropout": 0.0,
|
| 51 |
+
"BN": true,
|
| 52 |
+
"depth": 16,
|
| 53 |
+
"dataThreads": 4,
|
| 54 |
+
"dataDir": "/home/hungpv/.datasets",
|
| 55 |
+
"batchSize": 32,
|
| 56 |
+
"dataset": "CIFAR100",
|
| 57 |
+
"imageSize": 0,
|
| 58 |
+
"flip": false,
|
| 59 |
+
"crop": false,
|
| 60 |
+
"cut": false,
|
| 61 |
+
"cutoutProp": 0.5,
|
| 62 |
+
"randAugment": false,
|
| 63 |
+
"randAugment_magnitude": 10,
|
| 64 |
+
"mixup": false,
|
| 65 |
+
"mixupProp": 0.2,
|
| 66 |
+
"normalize": "std",
|
| 67 |
+
"truncate": true,
|
| 68 |
+
"verbose": true,
|
| 69 |
+
"metrics": [
|
| 70 |
+
"learningRate",
|
| 71 |
+
"loss",
|
| 72 |
+
"accuracy"
|
| 73 |
+
],
|
| 74 |
+
"logEach": 50,
|
| 75 |
+
"learningRate": 2e-05,
|
| 76 |
+
"lrScheduler": [
|
| 77 |
+
"const"
|
| 78 |
+
],
|
| 79 |
+
"LRScheduler_step_steps": 3.0,
|
| 80 |
+
"LRScheduler_step_gamma": 0.2,
|
| 81 |
+
"LRScheduler_exp_maxDecay": 0.01,
|
| 82 |
+
"LRScheduler_cos_periods": 0.5,
|
| 83 |
+
"LRScheduler_StartRamp_epochs": 1.0,
|
| 84 |
+
"LRScheduler_WRN_T0": 50.0,
|
| 85 |
+
"LRScheduler_WRN_Tmult": 1.2,
|
| 86 |
+
"rhoScheduler": [
|
| 87 |
+
"const"
|
| 88 |
+
],
|
| 89 |
+
"rhoScheduler_StartRamp_epochs": 8.0,
|
| 90 |
+
"rhoScheduler_StartJump_epochs": 8.0,
|
| 91 |
+
"saveCheckpoint": false,
|
| 92 |
+
"keepLastCheckpoint": false,
|
| 93 |
+
"checkpointsList": [],
|
| 94 |
+
"saveCheckpointInterval": 0,
|
| 95 |
+
"logDir": "/mnt/hungpv/projects/SAM_Variant/logs_nlp",
|
| 96 |
+
"logSubDir": "glue_bert",
|
| 97 |
+
"epochs": 3,
|
| 98 |
+
"contin": false
|
| 99 |
+
}
|
mrpc__SGD__lr2e-05__bs32__ep3__rho_na/checkpoint_final.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:5871ddc9ae62ff178690465754810966419b021786556ef6b4ac362f8837b2d9
|
| 3 |
+
size 438019989
|
mrpc__SGD__lr2e-05__bs32__ep3__rho_na/metrics_history.json
ADDED
|
@@ -0,0 +1,17 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
[
|
| 2 |
+
{
|
| 3 |
+
"epoch": 1,
|
| 4 |
+
"accuracy": 68.382,
|
| 5 |
+
"f1": 81.223
|
| 6 |
+
},
|
| 7 |
+
{
|
| 8 |
+
"epoch": 2,
|
| 9 |
+
"accuracy": 68.382,
|
| 10 |
+
"f1": 81.223
|
| 11 |
+
},
|
| 12 |
+
{
|
| 13 |
+
"epoch": 3,
|
| 14 |
+
"accuracy": 68.382,
|
| 15 |
+
"f1": 81.223
|
| 16 |
+
}
|
| 17 |
+
]
|
mrpc__SGD__lr2e-05__bs32__ep3__rho_na/params.json
ADDED
|
@@ -0,0 +1,98 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"label_smoothing": 0.0,
|
| 3 |
+
"ESAMbeta": 0.5,
|
| 4 |
+
"ESAMgamma": 0.5,
|
| 5 |
+
"lookSAMk": 5,
|
| 6 |
+
"lookSAMalpha": 0.7,
|
| 7 |
+
"lookSAMLayerwise": false,
|
| 8 |
+
"optimizer": "SGD",
|
| 9 |
+
"weightDecay": 0.01,
|
| 10 |
+
"momentum": 0.9,
|
| 11 |
+
"rho": 0.05,
|
| 12 |
+
"samRho": 0.05,
|
| 13 |
+
"asamRho": 0.05,
|
| 14 |
+
"esamRho": 0.05,
|
| 15 |
+
"looksamRho": 0.05,
|
| 16 |
+
"friendlysamRho": 0.05,
|
| 17 |
+
"fishersamRho": 0.05,
|
| 18 |
+
"mongesamRho": 0.05,
|
| 19 |
+
"gsamRho": 0.05,
|
| 20 |
+
"bayesiansamRho": 0.05,
|
| 21 |
+
"msamRho": 0.05,
|
| 22 |
+
"adamwSamRho": 0.05,
|
| 23 |
+
"adamwMsamRho": 0.05,
|
| 24 |
+
"adamwFriendlysamRho": 0.05,
|
| 25 |
+
"sigma": 1.0,
|
| 26 |
+
"lmbda": 0.9,
|
| 27 |
+
"fisherEta": 1.0,
|
| 28 |
+
"gsamAlpha": 0.01,
|
| 29 |
+
"gsamAdaptive": false,
|
| 30 |
+
"bayesMSharpness": 8,
|
| 31 |
+
"bayesDelta": 10.0,
|
| 32 |
+
"bayesBeta1": 0.9,
|
| 33 |
+
"bayesBeta2": 0.999,
|
| 34 |
+
"bayesSInit": 1.0,
|
| 35 |
+
"bayesDamping": 1e-08,
|
| 36 |
+
"bayesEps": 1e-12,
|
| 37 |
+
"nesterov": false,
|
| 38 |
+
"grad_clip_norm": 1.0,
|
| 39 |
+
"rndSeed": true,
|
| 40 |
+
"deterministic": false,
|
| 41 |
+
"glue_task": "mrpc",
|
| 42 |
+
"glue_output_root": "/mnt/hungpv/projects/SAM_Variant/logs_nlp",
|
| 43 |
+
"widthFactor": 4,
|
| 44 |
+
"ViTPatchsize": 32,
|
| 45 |
+
"ViTSize": "S",
|
| 46 |
+
"pyramidAlpha": 270,
|
| 47 |
+
"pyramidBottleneck": true,
|
| 48 |
+
"model": "WRN",
|
| 49 |
+
"dropout": 0.0,
|
| 50 |
+
"BN": true,
|
| 51 |
+
"depth": 16,
|
| 52 |
+
"dataThreads": 4,
|
| 53 |
+
"dataDir": "/home/hungpv/.datasets",
|
| 54 |
+
"batchSize": 32,
|
| 55 |
+
"dataset": "CIFAR100",
|
| 56 |
+
"imageSize": 0,
|
| 57 |
+
"flip": false,
|
| 58 |
+
"crop": false,
|
| 59 |
+
"cut": false,
|
| 60 |
+
"cutoutProp": 0.5,
|
| 61 |
+
"randAugment": false,
|
| 62 |
+
"randAugment_magnitude": 10,
|
| 63 |
+
"mixup": false,
|
| 64 |
+
"mixupProp": 0.2,
|
| 65 |
+
"normalize": "std",
|
| 66 |
+
"truncate": true,
|
| 67 |
+
"verbose": true,
|
| 68 |
+
"metrics": [
|
| 69 |
+
"learningRate",
|
| 70 |
+
"loss",
|
| 71 |
+
"accuracy"
|
| 72 |
+
],
|
| 73 |
+
"logEach": 50,
|
| 74 |
+
"learningRate": 2e-05,
|
| 75 |
+
"lrScheduler": [
|
| 76 |
+
"const"
|
| 77 |
+
],
|
| 78 |
+
"LRScheduler_step_steps": 3.0,
|
| 79 |
+
"LRScheduler_step_gamma": 0.2,
|
| 80 |
+
"LRScheduler_exp_maxDecay": 0.01,
|
| 81 |
+
"LRScheduler_cos_periods": 0.5,
|
| 82 |
+
"LRScheduler_StartRamp_epochs": 1.0,
|
| 83 |
+
"LRScheduler_WRN_T0": 50.0,
|
| 84 |
+
"LRScheduler_WRN_Tmult": 1.2,
|
| 85 |
+
"rhoScheduler": [
|
| 86 |
+
"const"
|
| 87 |
+
],
|
| 88 |
+
"rhoScheduler_StartRamp_epochs": 8.0,
|
| 89 |
+
"rhoScheduler_StartJump_epochs": 8.0,
|
| 90 |
+
"saveCheckpoint": false,
|
| 91 |
+
"keepLastCheckpoint": false,
|
| 92 |
+
"checkpointsList": [],
|
| 93 |
+
"saveCheckpointInterval": 0,
|
| 94 |
+
"logDir": "/mnt/hungpv/projects/SAM_Variant/logs_nlp",
|
| 95 |
+
"logSubDir": "glue_bert",
|
| 96 |
+
"epochs": 3,
|
| 97 |
+
"contin": false
|
| 98 |
+
}
|
qnli__SAM_Muon__lr2e-05__bs32__ep3__rho0.005/checkpoint_final.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:2dde3d7cac1c35cf1f30f06708ee04c2128faeb81184036d34013276088f1b04
|
| 3 |
+
size 438019989
|
qnli__SAM_Muon__lr2e-05__bs32__ep3__rho0.005/metrics_history.json
ADDED
|
@@ -0,0 +1,14 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
[
|
| 2 |
+
{
|
| 3 |
+
"epoch": 1,
|
| 4 |
+
"accuracy": 79.791
|
| 5 |
+
},
|
| 6 |
+
{
|
| 7 |
+
"epoch": 2,
|
| 8 |
+
"accuracy": 80.432
|
| 9 |
+
},
|
| 10 |
+
{
|
| 11 |
+
"epoch": 3,
|
| 12 |
+
"accuracy": 80.981
|
| 13 |
+
}
|
| 14 |
+
]
|
qnli__SAM_Muon__lr2e-05__bs32__ep3__rho0.005/params.json
ADDED
|
@@ -0,0 +1,99 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"label_smoothing": 0.0,
|
| 3 |
+
"ESAMbeta": 0.5,
|
| 4 |
+
"ESAMgamma": 0.5,
|
| 5 |
+
"lookSAMk": 5,
|
| 6 |
+
"lookSAMalpha": 0.7,
|
| 7 |
+
"lookSAMLayerwise": false,
|
| 8 |
+
"optimizer": "SAM_Muon",
|
| 9 |
+
"weightDecay": 0.01,
|
| 10 |
+
"momentum": 0.9,
|
| 11 |
+
"rho": 0.005,
|
| 12 |
+
"samRho": 0.005,
|
| 13 |
+
"asamRho": 0.005,
|
| 14 |
+
"esamRho": 0.005,
|
| 15 |
+
"looksamRho": 0.005,
|
| 16 |
+
"friendlysamRho": 0.005,
|
| 17 |
+
"fishersamRho": 0.005,
|
| 18 |
+
"mongesamRho": 0.005,
|
| 19 |
+
"gsamRho": 0.005,
|
| 20 |
+
"bayesiansamRho": 0.005,
|
| 21 |
+
"msamRho": 0.005,
|
| 22 |
+
"adamwSamRho": 0.005,
|
| 23 |
+
"adamSamRho": 0.005,
|
| 24 |
+
"adamwMsamRho": 0.005,
|
| 25 |
+
"adamwFriendlysamRho": 0.005,
|
| 26 |
+
"sigma": 1.0,
|
| 27 |
+
"lmbda": 0.9,
|
| 28 |
+
"fisherEta": 1.0,
|
| 29 |
+
"gsamAlpha": 0.01,
|
| 30 |
+
"gsamAdaptive": false,
|
| 31 |
+
"bayesMSharpness": 8,
|
| 32 |
+
"bayesDelta": 10.0,
|
| 33 |
+
"bayesBeta1": 0.9,
|
| 34 |
+
"bayesBeta2": 0.999,
|
| 35 |
+
"bayesSInit": 1.0,
|
| 36 |
+
"bayesDamping": 1e-08,
|
| 37 |
+
"bayesEps": 1e-12,
|
| 38 |
+
"nesterov": false,
|
| 39 |
+
"grad_clip_norm": 1.0,
|
| 40 |
+
"rndSeed": true,
|
| 41 |
+
"deterministic": false,
|
| 42 |
+
"glue_task": "qnli",
|
| 43 |
+
"glue_output_root": "/mnt/hungpv/projects/SAM_Variant/logs_nlp",
|
| 44 |
+
"widthFactor": 4,
|
| 45 |
+
"ViTPatchsize": 32,
|
| 46 |
+
"ViTSize": "S",
|
| 47 |
+
"pyramidAlpha": 270,
|
| 48 |
+
"pyramidBottleneck": true,
|
| 49 |
+
"model": "WRN",
|
| 50 |
+
"dropout": 0.0,
|
| 51 |
+
"BN": true,
|
| 52 |
+
"depth": 16,
|
| 53 |
+
"dataThreads": 4,
|
| 54 |
+
"dataDir": "/home/hungpv/.datasets",
|
| 55 |
+
"batchSize": 32,
|
| 56 |
+
"dataset": "CIFAR100",
|
| 57 |
+
"imageSize": 0,
|
| 58 |
+
"flip": false,
|
| 59 |
+
"crop": false,
|
| 60 |
+
"cut": false,
|
| 61 |
+
"cutoutProp": 0.5,
|
| 62 |
+
"randAugment": false,
|
| 63 |
+
"randAugment_magnitude": 10,
|
| 64 |
+
"mixup": false,
|
| 65 |
+
"mixupProp": 0.2,
|
| 66 |
+
"normalize": "std",
|
| 67 |
+
"truncate": true,
|
| 68 |
+
"verbose": true,
|
| 69 |
+
"metrics": [
|
| 70 |
+
"learningRate",
|
| 71 |
+
"loss",
|
| 72 |
+
"accuracy"
|
| 73 |
+
],
|
| 74 |
+
"logEach": 50,
|
| 75 |
+
"learningRate": 2e-05,
|
| 76 |
+
"lrScheduler": [
|
| 77 |
+
"const"
|
| 78 |
+
],
|
| 79 |
+
"LRScheduler_step_steps": 3.0,
|
| 80 |
+
"LRScheduler_step_gamma": 0.2,
|
| 81 |
+
"LRScheduler_exp_maxDecay": 0.01,
|
| 82 |
+
"LRScheduler_cos_periods": 0.5,
|
| 83 |
+
"LRScheduler_StartRamp_epochs": 1.0,
|
| 84 |
+
"LRScheduler_WRN_T0": 50.0,
|
| 85 |
+
"LRScheduler_WRN_Tmult": 1.2,
|
| 86 |
+
"rhoScheduler": [
|
| 87 |
+
"const"
|
| 88 |
+
],
|
| 89 |
+
"rhoScheduler_StartRamp_epochs": 8.0,
|
| 90 |
+
"rhoScheduler_StartJump_epochs": 8.0,
|
| 91 |
+
"saveCheckpoint": false,
|
| 92 |
+
"keepLastCheckpoint": false,
|
| 93 |
+
"checkpointsList": [],
|
| 94 |
+
"saveCheckpointInterval": 0,
|
| 95 |
+
"logDir": "/mnt/hungpv/projects/SAM_Variant/logs_nlp",
|
| 96 |
+
"logSubDir": "glue_bert",
|
| 97 |
+
"epochs": 3,
|
| 98 |
+
"contin": false
|
| 99 |
+
}
|
qnli__SAM__lr2e-05__bs32__ep3__rho0.005/checkpoint_final.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:02b876fcd8afd89aa7c65133ada0d2ab06acb3ec0414f9be91bdb0d673292c80
|
| 3 |
+
size 438019989
|
qnli__SAM__lr2e-05__bs32__ep3__rho0.005/metrics_history.json
ADDED
|
@@ -0,0 +1,14 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
[
|
| 2 |
+
{
|
| 3 |
+
"epoch": 1,
|
| 4 |
+
"accuracy": 69.028
|
| 5 |
+
},
|
| 6 |
+
{
|
| 7 |
+
"epoch": 2,
|
| 8 |
+
"accuracy": 70.913
|
| 9 |
+
},
|
| 10 |
+
{
|
| 11 |
+
"epoch": 3,
|
| 12 |
+
"accuracy": 71.572
|
| 13 |
+
}
|
| 14 |
+
]
|
qnli__SAM__lr2e-05__bs32__ep3__rho0.005/params.json
ADDED
|
@@ -0,0 +1,99 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"label_smoothing": 0.0,
|
| 3 |
+
"ESAMbeta": 0.5,
|
| 4 |
+
"ESAMgamma": 0.5,
|
| 5 |
+
"lookSAMk": 5,
|
| 6 |
+
"lookSAMalpha": 0.7,
|
| 7 |
+
"lookSAMLayerwise": false,
|
| 8 |
+
"optimizer": "SAM",
|
| 9 |
+
"weightDecay": 0.01,
|
| 10 |
+
"momentum": 0.9,
|
| 11 |
+
"rho": 0.005,
|
| 12 |
+
"samRho": 0.005,
|
| 13 |
+
"asamRho": 0.005,
|
| 14 |
+
"esamRho": 0.005,
|
| 15 |
+
"looksamRho": 0.005,
|
| 16 |
+
"friendlysamRho": 0.005,
|
| 17 |
+
"fishersamRho": 0.005,
|
| 18 |
+
"mongesamRho": 0.005,
|
| 19 |
+
"gsamRho": 0.005,
|
| 20 |
+
"bayesiansamRho": 0.005,
|
| 21 |
+
"msamRho": 0.005,
|
| 22 |
+
"adamwSamRho": 0.005,
|
| 23 |
+
"adamSamRho": 0.005,
|
| 24 |
+
"adamwMsamRho": 0.005,
|
| 25 |
+
"adamwFriendlysamRho": 0.005,
|
| 26 |
+
"sigma": 1.0,
|
| 27 |
+
"lmbda": 0.9,
|
| 28 |
+
"fisherEta": 1.0,
|
| 29 |
+
"gsamAlpha": 0.01,
|
| 30 |
+
"gsamAdaptive": false,
|
| 31 |
+
"bayesMSharpness": 8,
|
| 32 |
+
"bayesDelta": 10.0,
|
| 33 |
+
"bayesBeta1": 0.9,
|
| 34 |
+
"bayesBeta2": 0.999,
|
| 35 |
+
"bayesSInit": 1.0,
|
| 36 |
+
"bayesDamping": 1e-08,
|
| 37 |
+
"bayesEps": 1e-12,
|
| 38 |
+
"nesterov": false,
|
| 39 |
+
"grad_clip_norm": 1.0,
|
| 40 |
+
"rndSeed": true,
|
| 41 |
+
"deterministic": false,
|
| 42 |
+
"glue_task": "qnli",
|
| 43 |
+
"glue_output_root": "/mnt/hungpv/projects/SAM_Variant/logs_nlp",
|
| 44 |
+
"widthFactor": 4,
|
| 45 |
+
"ViTPatchsize": 32,
|
| 46 |
+
"ViTSize": "S",
|
| 47 |
+
"pyramidAlpha": 270,
|
| 48 |
+
"pyramidBottleneck": true,
|
| 49 |
+
"model": "WRN",
|
| 50 |
+
"dropout": 0.0,
|
| 51 |
+
"BN": true,
|
| 52 |
+
"depth": 16,
|
| 53 |
+
"dataThreads": 4,
|
| 54 |
+
"dataDir": "/home/hungpv/.datasets",
|
| 55 |
+
"batchSize": 32,
|
| 56 |
+
"dataset": "CIFAR100",
|
| 57 |
+
"imageSize": 0,
|
| 58 |
+
"flip": false,
|
| 59 |
+
"crop": false,
|
| 60 |
+
"cut": false,
|
| 61 |
+
"cutoutProp": 0.5,
|
| 62 |
+
"randAugment": false,
|
| 63 |
+
"randAugment_magnitude": 10,
|
| 64 |
+
"mixup": false,
|
| 65 |
+
"mixupProp": 0.2,
|
| 66 |
+
"normalize": "std",
|
| 67 |
+
"truncate": true,
|
| 68 |
+
"verbose": true,
|
| 69 |
+
"metrics": [
|
| 70 |
+
"learningRate",
|
| 71 |
+
"loss",
|
| 72 |
+
"accuracy"
|
| 73 |
+
],
|
| 74 |
+
"logEach": 50,
|
| 75 |
+
"learningRate": 2e-05,
|
| 76 |
+
"lrScheduler": [
|
| 77 |
+
"const"
|
| 78 |
+
],
|
| 79 |
+
"LRScheduler_step_steps": 3.0,
|
| 80 |
+
"LRScheduler_step_gamma": 0.2,
|
| 81 |
+
"LRScheduler_exp_maxDecay": 0.01,
|
| 82 |
+
"LRScheduler_cos_periods": 0.5,
|
| 83 |
+
"LRScheduler_StartRamp_epochs": 1.0,
|
| 84 |
+
"LRScheduler_WRN_T0": 50.0,
|
| 85 |
+
"LRScheduler_WRN_Tmult": 1.2,
|
| 86 |
+
"rhoScheduler": [
|
| 87 |
+
"const"
|
| 88 |
+
],
|
| 89 |
+
"rhoScheduler_StartRamp_epochs": 8.0,
|
| 90 |
+
"rhoScheduler_StartJump_epochs": 8.0,
|
| 91 |
+
"saveCheckpoint": false,
|
| 92 |
+
"keepLastCheckpoint": false,
|
| 93 |
+
"checkpointsList": [],
|
| 94 |
+
"saveCheckpointInterval": 0,
|
| 95 |
+
"logDir": "/mnt/hungpv/projects/SAM_Variant/logs_nlp",
|
| 96 |
+
"logSubDir": "glue_bert",
|
| 97 |
+
"epochs": 3,
|
| 98 |
+
"contin": false
|
| 99 |
+
}
|
qnli__SGD__lr2e-05__bs32__ep3__rho_na/checkpoint_final.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:00e43002ce2e69bcfeb4f6f098ee458c3cc8aa41ebd0987fff38c7d0fe251e7c
|
| 3 |
+
size 438019989
|
qnli__SGD__lr2e-05__bs32__ep3__rho_na/metrics_history.json
ADDED
|
@@ -0,0 +1,14 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
[
|
| 2 |
+
{
|
| 3 |
+
"epoch": 1,
|
| 4 |
+
"accuracy": 68.9
|
| 5 |
+
},
|
| 6 |
+
{
|
| 7 |
+
"epoch": 2,
|
| 8 |
+
"accuracy": 71.682
|
| 9 |
+
},
|
| 10 |
+
{
|
| 11 |
+
"epoch": 3,
|
| 12 |
+
"accuracy": 72.305
|
| 13 |
+
}
|
| 14 |
+
]
|
qnli__SGD__lr2e-05__bs32__ep3__rho_na/params.json
ADDED
|
@@ -0,0 +1,98 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"label_smoothing": 0.0,
|
| 3 |
+
"ESAMbeta": 0.5,
|
| 4 |
+
"ESAMgamma": 0.5,
|
| 5 |
+
"lookSAMk": 5,
|
| 6 |
+
"lookSAMalpha": 0.7,
|
| 7 |
+
"lookSAMLayerwise": false,
|
| 8 |
+
"optimizer": "SGD",
|
| 9 |
+
"weightDecay": 0.01,
|
| 10 |
+
"momentum": 0.9,
|
| 11 |
+
"rho": 0.05,
|
| 12 |
+
"samRho": 0.05,
|
| 13 |
+
"asamRho": 0.05,
|
| 14 |
+
"esamRho": 0.05,
|
| 15 |
+
"looksamRho": 0.05,
|
| 16 |
+
"friendlysamRho": 0.05,
|
| 17 |
+
"fishersamRho": 0.05,
|
| 18 |
+
"mongesamRho": 0.05,
|
| 19 |
+
"gsamRho": 0.05,
|
| 20 |
+
"bayesiansamRho": 0.05,
|
| 21 |
+
"msamRho": 0.05,
|
| 22 |
+
"adamwSamRho": 0.05,
|
| 23 |
+
"adamwMsamRho": 0.05,
|
| 24 |
+
"adamwFriendlysamRho": 0.05,
|
| 25 |
+
"sigma": 1.0,
|
| 26 |
+
"lmbda": 0.9,
|
| 27 |
+
"fisherEta": 1.0,
|
| 28 |
+
"gsamAlpha": 0.01,
|
| 29 |
+
"gsamAdaptive": false,
|
| 30 |
+
"bayesMSharpness": 8,
|
| 31 |
+
"bayesDelta": 10.0,
|
| 32 |
+
"bayesBeta1": 0.9,
|
| 33 |
+
"bayesBeta2": 0.999,
|
| 34 |
+
"bayesSInit": 1.0,
|
| 35 |
+
"bayesDamping": 1e-08,
|
| 36 |
+
"bayesEps": 1e-12,
|
| 37 |
+
"nesterov": false,
|
| 38 |
+
"grad_clip_norm": 1.0,
|
| 39 |
+
"rndSeed": true,
|
| 40 |
+
"deterministic": false,
|
| 41 |
+
"glue_task": "qnli",
|
| 42 |
+
"glue_output_root": "/mnt/hungpv/projects/SAM_Variant/logs_nlp",
|
| 43 |
+
"widthFactor": 4,
|
| 44 |
+
"ViTPatchsize": 32,
|
| 45 |
+
"ViTSize": "S",
|
| 46 |
+
"pyramidAlpha": 270,
|
| 47 |
+
"pyramidBottleneck": true,
|
| 48 |
+
"model": "WRN",
|
| 49 |
+
"dropout": 0.0,
|
| 50 |
+
"BN": true,
|
| 51 |
+
"depth": 16,
|
| 52 |
+
"dataThreads": 4,
|
| 53 |
+
"dataDir": "/home/hungpv/.datasets",
|
| 54 |
+
"batchSize": 32,
|
| 55 |
+
"dataset": "CIFAR100",
|
| 56 |
+
"imageSize": 0,
|
| 57 |
+
"flip": false,
|
| 58 |
+
"crop": false,
|
| 59 |
+
"cut": false,
|
| 60 |
+
"cutoutProp": 0.5,
|
| 61 |
+
"randAugment": false,
|
| 62 |
+
"randAugment_magnitude": 10,
|
| 63 |
+
"mixup": false,
|
| 64 |
+
"mixupProp": 0.2,
|
| 65 |
+
"normalize": "std",
|
| 66 |
+
"truncate": true,
|
| 67 |
+
"verbose": true,
|
| 68 |
+
"metrics": [
|
| 69 |
+
"learningRate",
|
| 70 |
+
"loss",
|
| 71 |
+
"accuracy"
|
| 72 |
+
],
|
| 73 |
+
"logEach": 50,
|
| 74 |
+
"learningRate": 2e-05,
|
| 75 |
+
"lrScheduler": [
|
| 76 |
+
"const"
|
| 77 |
+
],
|
| 78 |
+
"LRScheduler_step_steps": 3.0,
|
| 79 |
+
"LRScheduler_step_gamma": 0.2,
|
| 80 |
+
"LRScheduler_exp_maxDecay": 0.01,
|
| 81 |
+
"LRScheduler_cos_periods": 0.5,
|
| 82 |
+
"LRScheduler_StartRamp_epochs": 1.0,
|
| 83 |
+
"LRScheduler_WRN_T0": 50.0,
|
| 84 |
+
"LRScheduler_WRN_Tmult": 1.2,
|
| 85 |
+
"rhoScheduler": [
|
| 86 |
+
"const"
|
| 87 |
+
],
|
| 88 |
+
"rhoScheduler_StartRamp_epochs": 8.0,
|
| 89 |
+
"rhoScheduler_StartJump_epochs": 8.0,
|
| 90 |
+
"saveCheckpoint": false,
|
| 91 |
+
"keepLastCheckpoint": false,
|
| 92 |
+
"checkpointsList": [],
|
| 93 |
+
"saveCheckpointInterval": 0,
|
| 94 |
+
"logDir": "/mnt/hungpv/projects/SAM_Variant/logs_nlp",
|
| 95 |
+
"logSubDir": "glue_bert",
|
| 96 |
+
"epochs": 3,
|
| 97 |
+
"contin": false
|
| 98 |
+
}
|
qqp__SAM_Muon__lr2e-05__bs32__ep3__rho0.005/checkpoint_final.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:584b8028b067f59aa6297402ac078381058e98ca23acf6d4810a03dbcbd10795
|
| 3 |
+
size 438019989
|
qqp__SAM_Muon__lr2e-05__bs32__ep3__rho0.005/metrics_history.json
ADDED
|
@@ -0,0 +1,17 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
[
|
| 2 |
+
{
|
| 3 |
+
"epoch": 1,
|
| 4 |
+
"accuracy": 80.759,
|
| 5 |
+
"f1": 74.723
|
| 6 |
+
},
|
| 7 |
+
{
|
| 8 |
+
"epoch": 2,
|
| 9 |
+
"accuracy": 81.959,
|
| 10 |
+
"f1": 76.102
|
| 11 |
+
},
|
| 12 |
+
{
|
| 13 |
+
"epoch": 3,
|
| 14 |
+
"accuracy": 82.29,
|
| 15 |
+
"f1": 76.452
|
| 16 |
+
}
|
| 17 |
+
]
|
qqp__SAM_Muon__lr2e-05__bs32__ep3__rho0.005/params.json
ADDED
|
@@ -0,0 +1,99 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"label_smoothing": 0.0,
|
| 3 |
+
"ESAMbeta": 0.5,
|
| 4 |
+
"ESAMgamma": 0.5,
|
| 5 |
+
"lookSAMk": 5,
|
| 6 |
+
"lookSAMalpha": 0.7,
|
| 7 |
+
"lookSAMLayerwise": false,
|
| 8 |
+
"optimizer": "SAM_Muon",
|
| 9 |
+
"weightDecay": 0.01,
|
| 10 |
+
"momentum": 0.9,
|
| 11 |
+
"rho": 0.005,
|
| 12 |
+
"samRho": 0.005,
|
| 13 |
+
"asamRho": 0.005,
|
| 14 |
+
"esamRho": 0.005,
|
| 15 |
+
"looksamRho": 0.005,
|
| 16 |
+
"friendlysamRho": 0.005,
|
| 17 |
+
"fishersamRho": 0.005,
|
| 18 |
+
"mongesamRho": 0.005,
|
| 19 |
+
"gsamRho": 0.005,
|
| 20 |
+
"bayesiansamRho": 0.005,
|
| 21 |
+
"msamRho": 0.005,
|
| 22 |
+
"adamwSamRho": 0.005,
|
| 23 |
+
"adamSamRho": 0.005,
|
| 24 |
+
"adamwMsamRho": 0.005,
|
| 25 |
+
"adamwFriendlysamRho": 0.005,
|
| 26 |
+
"sigma": 1.0,
|
| 27 |
+
"lmbda": 0.9,
|
| 28 |
+
"fisherEta": 1.0,
|
| 29 |
+
"gsamAlpha": 0.01,
|
| 30 |
+
"gsamAdaptive": false,
|
| 31 |
+
"bayesMSharpness": 8,
|
| 32 |
+
"bayesDelta": 10.0,
|
| 33 |
+
"bayesBeta1": 0.9,
|
| 34 |
+
"bayesBeta2": 0.999,
|
| 35 |
+
"bayesSInit": 1.0,
|
| 36 |
+
"bayesDamping": 1e-08,
|
| 37 |
+
"bayesEps": 1e-12,
|
| 38 |
+
"nesterov": false,
|
| 39 |
+
"grad_clip_norm": 1.0,
|
| 40 |
+
"rndSeed": true,
|
| 41 |
+
"deterministic": false,
|
| 42 |
+
"glue_task": "qqp",
|
| 43 |
+
"glue_output_root": "/mnt/hungpv/projects/SAM_Variant/logs_nlp",
|
| 44 |
+
"widthFactor": 4,
|
| 45 |
+
"ViTPatchsize": 32,
|
| 46 |
+
"ViTSize": "S",
|
| 47 |
+
"pyramidAlpha": 270,
|
| 48 |
+
"pyramidBottleneck": true,
|
| 49 |
+
"model": "WRN",
|
| 50 |
+
"dropout": 0.0,
|
| 51 |
+
"BN": true,
|
| 52 |
+
"depth": 16,
|
| 53 |
+
"dataThreads": 4,
|
| 54 |
+
"dataDir": "/home/hungpv/.datasets",
|
| 55 |
+
"batchSize": 32,
|
| 56 |
+
"dataset": "CIFAR100",
|
| 57 |
+
"imageSize": 0,
|
| 58 |
+
"flip": false,
|
| 59 |
+
"crop": false,
|
| 60 |
+
"cut": false,
|
| 61 |
+
"cutoutProp": 0.5,
|
| 62 |
+
"randAugment": false,
|
| 63 |
+
"randAugment_magnitude": 10,
|
| 64 |
+
"mixup": false,
|
| 65 |
+
"mixupProp": 0.2,
|
| 66 |
+
"normalize": "std",
|
| 67 |
+
"truncate": true,
|
| 68 |
+
"verbose": true,
|
| 69 |
+
"metrics": [
|
| 70 |
+
"learningRate",
|
| 71 |
+
"loss",
|
| 72 |
+
"accuracy"
|
| 73 |
+
],
|
| 74 |
+
"logEach": 50,
|
| 75 |
+
"learningRate": 2e-05,
|
| 76 |
+
"lrScheduler": [
|
| 77 |
+
"const"
|
| 78 |
+
],
|
| 79 |
+
"LRScheduler_step_steps": 3.0,
|
| 80 |
+
"LRScheduler_step_gamma": 0.2,
|
| 81 |
+
"LRScheduler_exp_maxDecay": 0.01,
|
| 82 |
+
"LRScheduler_cos_periods": 0.5,
|
| 83 |
+
"LRScheduler_StartRamp_epochs": 1.0,
|
| 84 |
+
"LRScheduler_WRN_T0": 50.0,
|
| 85 |
+
"LRScheduler_WRN_Tmult": 1.2,
|
| 86 |
+
"rhoScheduler": [
|
| 87 |
+
"const"
|
| 88 |
+
],
|
| 89 |
+
"rhoScheduler_StartRamp_epochs": 8.0,
|
| 90 |
+
"rhoScheduler_StartJump_epochs": 8.0,
|
| 91 |
+
"saveCheckpoint": false,
|
| 92 |
+
"keepLastCheckpoint": false,
|
| 93 |
+
"checkpointsList": [],
|
| 94 |
+
"saveCheckpointInterval": 0,
|
| 95 |
+
"logDir": "/mnt/hungpv/projects/SAM_Variant/logs_nlp",
|
| 96 |
+
"logSubDir": "glue_bert",
|
| 97 |
+
"epochs": 3,
|
| 98 |
+
"contin": false
|
| 99 |
+
}
|
qqp__SAM__lr2e-05__bs32__ep3__rho0.005/checkpoint_final.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:af8ad46e1ee33dabe90a14bc6d48709d462bb05d408e5130fa3c9a94697b7ef7
|
| 3 |
+
size 438019989
|
qqp__SAM__lr2e-05__bs32__ep3__rho0.005/metrics_history.json
ADDED
|
@@ -0,0 +1,17 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
[
|
| 2 |
+
{
|
| 3 |
+
"epoch": 1,
|
| 4 |
+
"accuracy": 70.994,
|
| 5 |
+
"f1": 66.906
|
| 6 |
+
},
|
| 7 |
+
{
|
| 8 |
+
"epoch": 2,
|
| 9 |
+
"accuracy": 73.26,
|
| 10 |
+
"f1": 70.156
|
| 11 |
+
},
|
| 12 |
+
{
|
| 13 |
+
"epoch": 3,
|
| 14 |
+
"accuracy": 73.876,
|
| 15 |
+
"f1": 70.638
|
| 16 |
+
}
|
| 17 |
+
]
|
qqp__SAM__lr2e-05__bs32__ep3__rho0.005/params.json
ADDED
|
@@ -0,0 +1,99 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"label_smoothing": 0.0,
|
| 3 |
+
"ESAMbeta": 0.5,
|
| 4 |
+
"ESAMgamma": 0.5,
|
| 5 |
+
"lookSAMk": 5,
|
| 6 |
+
"lookSAMalpha": 0.7,
|
| 7 |
+
"lookSAMLayerwise": false,
|
| 8 |
+
"optimizer": "SAM",
|
| 9 |
+
"weightDecay": 0.01,
|
| 10 |
+
"momentum": 0.9,
|
| 11 |
+
"rho": 0.005,
|
| 12 |
+
"samRho": 0.005,
|
| 13 |
+
"asamRho": 0.005,
|
| 14 |
+
"esamRho": 0.005,
|
| 15 |
+
"looksamRho": 0.005,
|
| 16 |
+
"friendlysamRho": 0.005,
|
| 17 |
+
"fishersamRho": 0.005,
|
| 18 |
+
"mongesamRho": 0.005,
|
| 19 |
+
"gsamRho": 0.005,
|
| 20 |
+
"bayesiansamRho": 0.005,
|
| 21 |
+
"msamRho": 0.005,
|
| 22 |
+
"adamwSamRho": 0.005,
|
| 23 |
+
"adamSamRho": 0.005,
|
| 24 |
+
"adamwMsamRho": 0.005,
|
| 25 |
+
"adamwFriendlysamRho": 0.005,
|
| 26 |
+
"sigma": 1.0,
|
| 27 |
+
"lmbda": 0.9,
|
| 28 |
+
"fisherEta": 1.0,
|
| 29 |
+
"gsamAlpha": 0.01,
|
| 30 |
+
"gsamAdaptive": false,
|
| 31 |
+
"bayesMSharpness": 8,
|
| 32 |
+
"bayesDelta": 10.0,
|
| 33 |
+
"bayesBeta1": 0.9,
|
| 34 |
+
"bayesBeta2": 0.999,
|
| 35 |
+
"bayesSInit": 1.0,
|
| 36 |
+
"bayesDamping": 1e-08,
|
| 37 |
+
"bayesEps": 1e-12,
|
| 38 |
+
"nesterov": false,
|
| 39 |
+
"grad_clip_norm": 1.0,
|
| 40 |
+
"rndSeed": true,
|
| 41 |
+
"deterministic": false,
|
| 42 |
+
"glue_task": "qqp",
|
| 43 |
+
"glue_output_root": "/mnt/hungpv/projects/SAM_Variant/logs_nlp",
|
| 44 |
+
"widthFactor": 4,
|
| 45 |
+
"ViTPatchsize": 32,
|
| 46 |
+
"ViTSize": "S",
|
| 47 |
+
"pyramidAlpha": 270,
|
| 48 |
+
"pyramidBottleneck": true,
|
| 49 |
+
"model": "WRN",
|
| 50 |
+
"dropout": 0.0,
|
| 51 |
+
"BN": true,
|
| 52 |
+
"depth": 16,
|
| 53 |
+
"dataThreads": 4,
|
| 54 |
+
"dataDir": "/home/hungpv/.datasets",
|
| 55 |
+
"batchSize": 32,
|
| 56 |
+
"dataset": "CIFAR100",
|
| 57 |
+
"imageSize": 0,
|
| 58 |
+
"flip": false,
|
| 59 |
+
"crop": false,
|
| 60 |
+
"cut": false,
|
| 61 |
+
"cutoutProp": 0.5,
|
| 62 |
+
"randAugment": false,
|
| 63 |
+
"randAugment_magnitude": 10,
|
| 64 |
+
"mixup": false,
|
| 65 |
+
"mixupProp": 0.2,
|
| 66 |
+
"normalize": "std",
|
| 67 |
+
"truncate": true,
|
| 68 |
+
"verbose": true,
|
| 69 |
+
"metrics": [
|
| 70 |
+
"learningRate",
|
| 71 |
+
"loss",
|
| 72 |
+
"accuracy"
|
| 73 |
+
],
|
| 74 |
+
"logEach": 50,
|
| 75 |
+
"learningRate": 2e-05,
|
| 76 |
+
"lrScheduler": [
|
| 77 |
+
"const"
|
| 78 |
+
],
|
| 79 |
+
"LRScheduler_step_steps": 3.0,
|
| 80 |
+
"LRScheduler_step_gamma": 0.2,
|
| 81 |
+
"LRScheduler_exp_maxDecay": 0.01,
|
| 82 |
+
"LRScheduler_cos_periods": 0.5,
|
| 83 |
+
"LRScheduler_StartRamp_epochs": 1.0,
|
| 84 |
+
"LRScheduler_WRN_T0": 50.0,
|
| 85 |
+
"LRScheduler_WRN_Tmult": 1.2,
|
| 86 |
+
"rhoScheduler": [
|
| 87 |
+
"const"
|
| 88 |
+
],
|
| 89 |
+
"rhoScheduler_StartRamp_epochs": 8.0,
|
| 90 |
+
"rhoScheduler_StartJump_epochs": 8.0,
|
| 91 |
+
"saveCheckpoint": false,
|
| 92 |
+
"keepLastCheckpoint": false,
|
| 93 |
+
"checkpointsList": [],
|
| 94 |
+
"saveCheckpointInterval": 0,
|
| 95 |
+
"logDir": "/mnt/hungpv/projects/SAM_Variant/logs_nlp",
|
| 96 |
+
"logSubDir": "glue_bert",
|
| 97 |
+
"epochs": 3,
|
| 98 |
+
"contin": false
|
| 99 |
+
}
|
qqp__SGD__lr2e-05__bs32__ep3__rho_na/checkpoint_final.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ae328d07cd59416db72a8b1d3fdb1658350b4f6a5292ae668d141ab4b1e11332
|
| 3 |
+
size 438019989
|
qqp__SGD__lr2e-05__bs32__ep3__rho_na/metrics_history.json
ADDED
|
@@ -0,0 +1,17 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
[
|
| 2 |
+
{
|
| 3 |
+
"epoch": 1,
|
| 4 |
+
"accuracy": 70.952,
|
| 5 |
+
"f1": 65.703
|
| 6 |
+
},
|
| 7 |
+
{
|
| 8 |
+
"epoch": 2,
|
| 9 |
+
"accuracy": 73.208,
|
| 10 |
+
"f1": 70.291
|
| 11 |
+
},
|
| 12 |
+
{
|
| 13 |
+
"epoch": 3,
|
| 14 |
+
"accuracy": 73.93,
|
| 15 |
+
"f1": 70.908
|
| 16 |
+
}
|
| 17 |
+
]
|
qqp__SGD__lr2e-05__bs32__ep3__rho_na/params.json
ADDED
|
@@ -0,0 +1,98 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"label_smoothing": 0.0,
|
| 3 |
+
"ESAMbeta": 0.5,
|
| 4 |
+
"ESAMgamma": 0.5,
|
| 5 |
+
"lookSAMk": 5,
|
| 6 |
+
"lookSAMalpha": 0.7,
|
| 7 |
+
"lookSAMLayerwise": false,
|
| 8 |
+
"optimizer": "SGD",
|
| 9 |
+
"weightDecay": 0.01,
|
| 10 |
+
"momentum": 0.9,
|
| 11 |
+
"rho": 0.05,
|
| 12 |
+
"samRho": 0.05,
|
| 13 |
+
"asamRho": 0.05,
|
| 14 |
+
"esamRho": 0.05,
|
| 15 |
+
"looksamRho": 0.05,
|
| 16 |
+
"friendlysamRho": 0.05,
|
| 17 |
+
"fishersamRho": 0.05,
|
| 18 |
+
"mongesamRho": 0.05,
|
| 19 |
+
"gsamRho": 0.05,
|
| 20 |
+
"bayesiansamRho": 0.05,
|
| 21 |
+
"msamRho": 0.05,
|
| 22 |
+
"adamwSamRho": 0.05,
|
| 23 |
+
"adamwMsamRho": 0.05,
|
| 24 |
+
"adamwFriendlysamRho": 0.05,
|
| 25 |
+
"sigma": 1.0,
|
| 26 |
+
"lmbda": 0.9,
|
| 27 |
+
"fisherEta": 1.0,
|
| 28 |
+
"gsamAlpha": 0.01,
|
| 29 |
+
"gsamAdaptive": false,
|
| 30 |
+
"bayesMSharpness": 8,
|
| 31 |
+
"bayesDelta": 10.0,
|
| 32 |
+
"bayesBeta1": 0.9,
|
| 33 |
+
"bayesBeta2": 0.999,
|
| 34 |
+
"bayesSInit": 1.0,
|
| 35 |
+
"bayesDamping": 1e-08,
|
| 36 |
+
"bayesEps": 1e-12,
|
| 37 |
+
"nesterov": false,
|
| 38 |
+
"grad_clip_norm": 1.0,
|
| 39 |
+
"rndSeed": true,
|
| 40 |
+
"deterministic": false,
|
| 41 |
+
"glue_task": "qqp",
|
| 42 |
+
"glue_output_root": "/mnt/hungpv/projects/SAM_Variant/logs_nlp",
|
| 43 |
+
"widthFactor": 4,
|
| 44 |
+
"ViTPatchsize": 32,
|
| 45 |
+
"ViTSize": "S",
|
| 46 |
+
"pyramidAlpha": 270,
|
| 47 |
+
"pyramidBottleneck": true,
|
| 48 |
+
"model": "WRN",
|
| 49 |
+
"dropout": 0.0,
|
| 50 |
+
"BN": true,
|
| 51 |
+
"depth": 16,
|
| 52 |
+
"dataThreads": 4,
|
| 53 |
+
"dataDir": "/home/hungpv/.datasets",
|
| 54 |
+
"batchSize": 32,
|
| 55 |
+
"dataset": "CIFAR100",
|
| 56 |
+
"imageSize": 0,
|
| 57 |
+
"flip": false,
|
| 58 |
+
"crop": false,
|
| 59 |
+
"cut": false,
|
| 60 |
+
"cutoutProp": 0.5,
|
| 61 |
+
"randAugment": false,
|
| 62 |
+
"randAugment_magnitude": 10,
|
| 63 |
+
"mixup": false,
|
| 64 |
+
"mixupProp": 0.2,
|
| 65 |
+
"normalize": "std",
|
| 66 |
+
"truncate": true,
|
| 67 |
+
"verbose": true,
|
| 68 |
+
"metrics": [
|
| 69 |
+
"learningRate",
|
| 70 |
+
"loss",
|
| 71 |
+
"accuracy"
|
| 72 |
+
],
|
| 73 |
+
"logEach": 50,
|
| 74 |
+
"learningRate": 2e-05,
|
| 75 |
+
"lrScheduler": [
|
| 76 |
+
"const"
|
| 77 |
+
],
|
| 78 |
+
"LRScheduler_step_steps": 3.0,
|
| 79 |
+
"LRScheduler_step_gamma": 0.2,
|
| 80 |
+
"LRScheduler_exp_maxDecay": 0.01,
|
| 81 |
+
"LRScheduler_cos_periods": 0.5,
|
| 82 |
+
"LRScheduler_StartRamp_epochs": 1.0,
|
| 83 |
+
"LRScheduler_WRN_T0": 50.0,
|
| 84 |
+
"LRScheduler_WRN_Tmult": 1.2,
|
| 85 |
+
"rhoScheduler": [
|
| 86 |
+
"const"
|
| 87 |
+
],
|
| 88 |
+
"rhoScheduler_StartRamp_epochs": 8.0,
|
| 89 |
+
"rhoScheduler_StartJump_epochs": 8.0,
|
| 90 |
+
"saveCheckpoint": false,
|
| 91 |
+
"keepLastCheckpoint": false,
|
| 92 |
+
"checkpointsList": [],
|
| 93 |
+
"saveCheckpointInterval": 0,
|
| 94 |
+
"logDir": "/mnt/hungpv/projects/SAM_Variant/logs_nlp",
|
| 95 |
+
"logSubDir": "glue_bert",
|
| 96 |
+
"epochs": 3,
|
| 97 |
+
"contin": false
|
| 98 |
+
}
|
rte__SAM_Muon__lr2e-05__bs32__ep3__rho0.004/checkpoint_final.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:944f38c4f637c231bf5280993d262e8930e575a8334701d12305983e5bee8914
|
| 3 |
+
size 438019989
|
rte__SAM_Muon__lr2e-05__bs32__ep3__rho0.004/metrics_history.json
ADDED
|
@@ -0,0 +1,14 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
[
|
| 2 |
+
{
|
| 3 |
+
"epoch": 1,
|
| 4 |
+
"accuracy": 47.292
|
| 5 |
+
},
|
| 6 |
+
{
|
| 7 |
+
"epoch": 2,
|
| 8 |
+
"accuracy": 47.292
|
| 9 |
+
},
|
| 10 |
+
{
|
| 11 |
+
"epoch": 3,
|
| 12 |
+
"accuracy": 47.292
|
| 13 |
+
}
|
| 14 |
+
]
|
rte__SAM_Muon__lr2e-05__bs32__ep3__rho0.004/params.json
ADDED
|
@@ -0,0 +1,99 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"label_smoothing": 0.0,
|
| 3 |
+
"ESAMbeta": 0.5,
|
| 4 |
+
"ESAMgamma": 0.5,
|
| 5 |
+
"lookSAMk": 5,
|
| 6 |
+
"lookSAMalpha": 0.7,
|
| 7 |
+
"lookSAMLayerwise": false,
|
| 8 |
+
"optimizer": "SAM_Muon",
|
| 9 |
+
"weightDecay": 0.01,
|
| 10 |
+
"momentum": 0.9,
|
| 11 |
+
"rho": 0.004,
|
| 12 |
+
"samRho": 0.004,
|
| 13 |
+
"asamRho": 0.004,
|
| 14 |
+
"esamRho": 0.004,
|
| 15 |
+
"looksamRho": 0.004,
|
| 16 |
+
"friendlysamRho": 0.004,
|
| 17 |
+
"fishersamRho": 0.004,
|
| 18 |
+
"mongesamRho": 0.004,
|
| 19 |
+
"gsamRho": 0.004,
|
| 20 |
+
"bayesiansamRho": 0.004,
|
| 21 |
+
"msamRho": 0.004,
|
| 22 |
+
"adamwSamRho": 0.004,
|
| 23 |
+
"adamSamRho": 0.004,
|
| 24 |
+
"adamwMsamRho": 0.004,
|
| 25 |
+
"adamwFriendlysamRho": 0.004,
|
| 26 |
+
"sigma": 1.0,
|
| 27 |
+
"lmbda": 0.9,
|
| 28 |
+
"fisherEta": 1.0,
|
| 29 |
+
"gsamAlpha": 0.01,
|
| 30 |
+
"gsamAdaptive": false,
|
| 31 |
+
"bayesMSharpness": 8,
|
| 32 |
+
"bayesDelta": 10.0,
|
| 33 |
+
"bayesBeta1": 0.9,
|
| 34 |
+
"bayesBeta2": 0.999,
|
| 35 |
+
"bayesSInit": 1.0,
|
| 36 |
+
"bayesDamping": 1e-08,
|
| 37 |
+
"bayesEps": 1e-12,
|
| 38 |
+
"nesterov": false,
|
| 39 |
+
"grad_clip_norm": 1.0,
|
| 40 |
+
"rndSeed": true,
|
| 41 |
+
"deterministic": false,
|
| 42 |
+
"glue_task": "rte",
|
| 43 |
+
"glue_output_root": "/mnt/hungpv/projects/SAM_Variant/logs_nlp",
|
| 44 |
+
"widthFactor": 4,
|
| 45 |
+
"ViTPatchsize": 32,
|
| 46 |
+
"ViTSize": "S",
|
| 47 |
+
"pyramidAlpha": 270,
|
| 48 |
+
"pyramidBottleneck": true,
|
| 49 |
+
"model": "WRN",
|
| 50 |
+
"dropout": 0.0,
|
| 51 |
+
"BN": true,
|
| 52 |
+
"depth": 16,
|
| 53 |
+
"dataThreads": 4,
|
| 54 |
+
"dataDir": "/home/hungpv/.datasets",
|
| 55 |
+
"batchSize": 32,
|
| 56 |
+
"dataset": "CIFAR100",
|
| 57 |
+
"imageSize": 0,
|
| 58 |
+
"flip": false,
|
| 59 |
+
"crop": false,
|
| 60 |
+
"cut": false,
|
| 61 |
+
"cutoutProp": 0.5,
|
| 62 |
+
"randAugment": false,
|
| 63 |
+
"randAugment_magnitude": 10,
|
| 64 |
+
"mixup": false,
|
| 65 |
+
"mixupProp": 0.2,
|
| 66 |
+
"normalize": "std",
|
| 67 |
+
"truncate": true,
|
| 68 |
+
"verbose": true,
|
| 69 |
+
"metrics": [
|
| 70 |
+
"learningRate",
|
| 71 |
+
"loss",
|
| 72 |
+
"accuracy"
|
| 73 |
+
],
|
| 74 |
+
"logEach": 50,
|
| 75 |
+
"learningRate": 2e-05,
|
| 76 |
+
"lrScheduler": [
|
| 77 |
+
"const"
|
| 78 |
+
],
|
| 79 |
+
"LRScheduler_step_steps": 3.0,
|
| 80 |
+
"LRScheduler_step_gamma": 0.2,
|
| 81 |
+
"LRScheduler_exp_maxDecay": 0.01,
|
| 82 |
+
"LRScheduler_cos_periods": 0.5,
|
| 83 |
+
"LRScheduler_StartRamp_epochs": 1.0,
|
| 84 |
+
"LRScheduler_WRN_T0": 50.0,
|
| 85 |
+
"LRScheduler_WRN_Tmult": 1.2,
|
| 86 |
+
"rhoScheduler": [
|
| 87 |
+
"const"
|
| 88 |
+
],
|
| 89 |
+
"rhoScheduler_StartRamp_epochs": 8.0,
|
| 90 |
+
"rhoScheduler_StartJump_epochs": 8.0,
|
| 91 |
+
"saveCheckpoint": false,
|
| 92 |
+
"keepLastCheckpoint": false,
|
| 93 |
+
"checkpointsList": [],
|
| 94 |
+
"saveCheckpointInterval": 0,
|
| 95 |
+
"logDir": "/mnt/hungpv/projects/SAM_Variant/logs_nlp",
|
| 96 |
+
"logSubDir": "glue_bert",
|
| 97 |
+
"epochs": 3,
|
| 98 |
+
"contin": false
|
| 99 |
+
}
|
rte__SAM__lr2e-05__bs32__ep3__rho0.004/checkpoint_final.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:36aa572affaf64f610267f60042ec71f861250fedf711fbd6c93cbbfc907d466
|
| 3 |
+
size 438019989
|
rte__SAM__lr2e-05__bs32__ep3__rho0.004/metrics_history.json
ADDED
|
@@ -0,0 +1,14 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
[
|
| 2 |
+
{
|
| 3 |
+
"epoch": 1,
|
| 4 |
+
"accuracy": 47.292
|
| 5 |
+
},
|
| 6 |
+
{
|
| 7 |
+
"epoch": 2,
|
| 8 |
+
"accuracy": 47.292
|
| 9 |
+
},
|
| 10 |
+
{
|
| 11 |
+
"epoch": 3,
|
| 12 |
+
"accuracy": 47.292
|
| 13 |
+
}
|
| 14 |
+
]
|