FGIR-ViT checkpoint zoo

One deduplicated folder holding every kept checkpoint from the FGIR-ViT family of experiments: the ILA inter-layer adapter and SAW contrastive-warmup paper (ila_*), the attention-aggregation adapter (AAA) paper (aaa_*), the medical PETL and IDRiD studies (medical_*), and the full-fine-tune fine-grained benchmark (ft_*). It backs the fgirvit family of the fgir_zoo library and the HuggingFace mirror ERISLab/FGIR-ViT.

1051 checkpoints, one seed per configuration, last epoch only.

Layout

Files are grouped by experiment line, flat inside each group, named <dataset>_<model>_<method>_<serial>[_<variant>].pth. The serial is the wandb serial, and the name ends in it (or in the variant, where one is present) because only the last epoch is kept, so "last" is implied and never written into the name. <variant> is the ifc contrastive-head tag (jsd, dscsc) that disambiguates configurations sharing a serial, on the ila_contrastive and aaa_viz runs.

Group	Origin	wandb project	Serials	Backbone	Files
ila_main_224	ILA/SAW paper	ILA	1	vit_b16	206
ila_main_448	ILA/SAW paper	ILA	3	vit_b16	185
ila_ablations	ILA/SAW paper	ILA	4–18	vit_b16	33
ila_saw	ILA/SAW paper	ILA	90–93, 210–260	vit_b16	36
ila_saw_deit	ILA/SAW paper	ILA	71, 81	pedeit_base_patch16_224.fb_in1k	20
ila_saw_no_ila	ILA/SAW paper	ILA	270–272	vit_b16	21
ila_swin	ILA/SAW paper	ILA	80, 81, 161	peswin_base_patch4_window7_224.ms_in22k	31
ila_dinov2	ILA/SAW paper	ILA	27, 70, 71	pevit_base_patch14_dinov2.lvd142m	9
ila_foundation	ILA/SAW paper	ila-backbones-baselines	1–3, 100/200/300	MAE, DINOv2, CLIP pevit	12
ila_equal_compute	ILA/SAW paper	ILA	150	vit_b16	10
ila_contrastive	ILA/SAW paper	ILA	410–436	vit_b16	105
aaa_main_224	AAA paper	ILA	401	vit_b16	13
aaa_main_448	AAA paper	ILA	403	vit_b16	15
aaa_main_672	AAA paper	AAA	501	vit_b16	40
aaa_viz	AAA paper	ILA	420, 421	vit_b16	12
medical_idrid	AAA medical	AAA	541	vit_b16	6
medical_petl	AAA medical	AAA	1	vit_b16	55
ft_224	FGIR benchmark	FGIRFT	1	vit_b16	111
ft_448	FGIR benchmark	FGIRFT	3	vit_b16	131

The wandb project shown is the current one. Some older configs carry a legacy project_name in their embedded config (MLA for the ILA main tables, PEFGIR for part of the 224 benchmark); grouping here follows the experiment line, not that field.

aaa_main_672 holds the four adapter arms (attention-aggregation, ILA++ conv, plain adapter, frozen) across ten datasets. ila_foundation holds the frozen MAE/DINOv2/CLIP baselines (launched as serials 100/200/300, renumbered to 29). aaa_viz holds the attention-visualisation runs, which may be rerun.

Contents of a checkpoint

Each file is a dict with config (the run's full argument namespace), model (the state dict), accuracy (top-1 on the test set) and epoch. Load it with torch.load(path, weights_only=False); the embedded config drives model construction.

Loading by name

Install the standalone library and load any checkpoint by its <group>/<stem> name — no fgir_vit install needed:

pip install git+https://github.com/arkel23/fgir-zoo.git
python -c "import fgir_zoo; m = fgir_zoo.load('fgirvit/aaa_main_448/soylocal_vit_b16_aaa_dso_cls_adapter_fz_403')"

fgir_zoo.list_models(family='fgirvit') lists the names; the row set is manifest.csv.

Where the rest went

Deprecated files (best-epoch copies, learning-rate searches, extra seeds, reproduction and smoke runs, one corrupt file) are in /mnt/hdd10tb/edwin/ckpt_cleanup/deprecated/consolidation_20260920/, grouped by reason. The obsolete first-pass checkpoint tree and the redundant gather folders are in /mnt/hdd10tb/edwin/deprecated/, pending deletion. /mnt/hdd10tb/edwin/results_ila/ keeps only the ILA/SAW paper's analysis outputs (per-class accuracies, CKA, transferability feature dumps). The build scripts and the file inventory that produced this folder are in _build/.