Spaces:

DandyDonUnhinged
/

spock-body-lora-training

Paused

App Files Files Community

spock-body-lora-training

8.22 MB

Ctrl+K

Ctrl+K

3 contributors

History: 59 commits

DandyDonUnhinged's picture

DandyDonUnhinged

body LoRA: PUSH adapter to HF after training (preserves artifact past /tmp/ container exit)

3a82806 verified 8 days ago

ai-toolkit
fix: comprehensive substrate fork v2 + bump diffusers to >=0.34.0 (FLAX_WEIGHTS_NAME fix) 8 days ago
dataset
body LoRA: HF Trainer recipe (mirrors train_rafael_0.8b.py) 8 days ago
l2_audit
feat: vendored L2 pre-push audit script (5-invariant oracle for HF Spaces content) 8 days ago
.gitattributes

1.52 kB
initial: cloud_package code only (dataset excluded, push via xet separately) 14 days ago
.gitignore

100 Bytes
revert: dataset/ back in .gitignore (will ship via HF Dataset repo instead) 8 days ago
Dockerfile

2.38 kB
body LoRA: cache-bust (transformers 4.50.0 for qwen3_5) 8 days ago
README.md

3.38 kB
body LoRA: trigger fresh build (transformers 4.50.0 for qwen3_5) 8 days ago
comprehensive_fork.py

11.6 kB
fix: comprehensive substrate fork v2 + bump diffusers to >=0.34.0 (FLAX_WEIGHTS_NAME fix) 8 days ago
requirements.txt

265 Bytes
body LoRA: real change to requirements.txt (cache-bust with comment line) 8 days ago
run.py

3.37 kB
body LoRA: do NOT pass --output-dir (let train script use probed path) - real change 8 days ago
train_config.yaml

1.96 kB
fix: HF Spaces canonical user pattern (UID 1000); user-writable HOME; stable local_dir model paths 8 days ago
train_spock_0.8b.py

11.7 kB
body LoRA: PUSH adapter to HF after training (preserves artifact past /tmp/ container exit) 8 days ago