Skip to content

Actions: dieterich-lab/LLM-PPI-Extraction

Actions

All workflows

Actions

Loading...
Loading

Showing runs from all workflows
44 workflow runs
44 workflow runs

Filter by Workflow

⌥ + click/return to exclude

Filter by Event

⌥ + click/return to exclude

Filter by Status

⌥ + click/return to exclude

Filter by Branch

⌥ + click/return to exclude

Filter by Actor

⌥ + click/return to exclude
style: formatting CI #43: Commit c7f49c5 pushed by phiwi
17s master
master
17s
docs: add 70B HF link, no longer 'coming soon' CI #42: Commit 1327fcb pushed by phiwi
18s master
master
18s
fix: push to HF before GGUF save, wrap GGUF in try/except CI #41: Commit 0e5587f pushed by phiwi
18s master
master
18s
Add standalone GGUF conversion script for Llama 3.3 70B (bypasses Uns… CI #40: Commit 09ad428 pushed by phiwi
18s master
master
18s
chore: comment out non-baseline runs in 70B regu matrix (direct_norma… CI #39: Commit 3dde3af pushed by phiwi
16s master
master
16s
fix: disable checkpoint saves to avoid Unsloth/TRL pickle error CI #38: Commit 88fe47a pushed by phiwi
20s master
master
20s
docs: document fine-tuned models in README with training details and … CI #37: Commit a35006d pushed by phiwi
19s master
master
19s
fix: truncate doc text to 4000 chars for DynEx embedding (mxbai-embed… CI #36: Commit 6b4e1a7 pushed by phiwi
16s master
master
16s
feat: add Ollama import & matrix scripts for fine-tuned Llama 3.3 70B CI #35: Commit 407c995 pushed by phiwi
23s master
master
23s
fix: reduce 70B finetuning to 1 GPU — Unsloth doesn't auto-DP CI #34: Commit ad5e2ff pushed by phiwi
18s master
master
18s
fix: disable HF XET transfer for 70B model download CI #33: Commit 1a2bc8a pushed by phiwi
17s master
master
17s
chore: relax unsloth pin to >= for Llama 3.3 70B support CI #32: Commit ccf8f00 pushed by phiwi
18s master
master
18s
fix: update finetune_llama33.sh with proper paths, env, 2xH100, 400G RAM CI #31: Commit 5eaf394 pushed by phiwi
17s master
master
17s
chore: move matrix runs to gpu-g3-1 (Turing, 24GB) - inference only, … CI #30: Commit c2b29d2 pushed by phiwi
18s master
master
18s
fix: delete existing Ollama model before re-import CI #29: Commit 0c66b4e pushed by phiwi
19s master
master
19s
feat: add vanilla Llama 3.1 8B full-corpus matrix for baseline compar… CI #28: Commit 01e2076 pushed by phiwi
16s master
master
16s
feat: add RegulaTome matrix script for fine-tuned Llama 3.1 8B (llama… CI #27: Commit f7fa157 pushed by phiwi
17s master
master
17s
fix: use Unsloth-generated Modelfile from GGUF dir for Ollama import CI #26: Commit 493db0a pushed by phiwi
16s master
master
16s
feat: add Ollama import script for fine-tuned Llama 3.1 8B RegulaTome… CI #25: Commit 1448019 pushed by phiwi
16s master
master
16s
fix: revert to gpu-g4-1 (Ampere) - Turing GPU incompatible with Trito… CI #24: Commit b27676d pushed by phiwi
15s master
master
15s
chore: move finetune_llama31 to gpu-g3-1 (Turing, 24GB) - only 7.6GB … CI #23: Commit 802732f pushed by phiwi
20s master
master
20s
fix: remove cd ../ from finetune scripts (run from SLURM_SUBMIT_DIR) CI #22: Commit 9c08837 pushed by phiwi
16s master
master
16s
fix: correct .env path in SLURM scripts (SLURM_SUBMIT_DIR/.env) CI #21: Commit 20a9773 pushed by phiwi
17s master
master
17s
fix: use SLURM_SUBMIT_DIR for .env sourcing in all SLURM scripts CI #20: Commit 8028fa9 pushed by phiwi
19s master
master
19s
fix: use absolute path for SLURM output in finetune_llama31.sh CI #19: Commit c2938a9 pushed by phiwi
18s master
master
18s