Search results

Inspect accuracy discrepancy between .err & .out logs in multi-GPU linear probing

debugcode inspection/debuggingcode inspection/debugging

Status: Open.

#88 In kutayeroglu/ijepa;

· kutayeroglu opened on Apr 11, 2026

Implement semantic segmentation evaluation procedure

feature

Status: Open.

#87 In kutayeroglu/ijepa;

· kutayeroglu opened on Mar 31, 2026

Get ADE20k dataset on MN5

sys

Status: Open.

#86 In kutayeroglu/ijepa;

· kutayeroglu opened on Mar 31, 2026

Run semantic segmentation evaluation on multinoise and multiblock

feature

Status: Open.

#85 In kutayeroglu/ijepa;

· kutayeroglu opened on Mar 30, 2026

Remove linear scaling of hyperparameters use same evaluation setting to see whether performance improves

training

Status: Open.

#84 In kutayeroglu/ijepa;

· kutayeroglu opened on Mar 30, 2026

Linear probe pipeline is not fully reproducible across runs

feature

Status: Open.

#82 In kutayeroglu/ijepa;

· kutayeroglu opened on Mar 20, 2026

Copy log output of linear probe directly to linked pre-training run

nice to have

sys

Status: Open.

#80 In kutayeroglu/ijepa;

· kutayeroglu opened on Mar 16, 2026

Improve experiment tracking procedures

nice to have

sys

Status: Open.

#79 In kutayeroglu/ijepa;

· kutayeroglu opened on Mar 16, 2026

Multi-node training (2 nodes × 4 GPUs) for I-JEPA with SLURM + Singularity

feature

sys

Status: Open.

#73 In kutayeroglu/ijepa;

· kutayeroglu opened on Feb 28, 2026

setup wandb

feature

Status: Open.

#53 In kutayeroglu/ijepa;

· kutayeroglu opened on Jan 1, 2026

Check why there is no resize in train data preprocess

understandingfor understanding source codefor understanding source code

Status: Open.

#51 In kutayeroglu/ijepa;

· kutayeroglu opened on Dec 28, 2025

Why does training logs go under .err slurm outputs?

bugSomething isn't workingSomething isn't working

understandingfor understanding source codefor understanding source code

Status: Open.

#47 In kutayeroglu/ijepa;

· kutayeroglu opened on Dec 10, 2025