Spaces:

eliezeravihail
/

torchdocs-agent

Running

App Files Files Community

torchdocs-agent / eval

5.69 MB

Ctrl+K

Ctrl+K

4 contributors

History: 54 commits

github-actions[bot]

eval: agentic results from Actions run [skip ci]

f9534ae 8 days ago

results
eval: agentic results from Actions run [skip ci] 8 days ago
__init__.py

0 Bytes
Scaffold repo structure per M0 (pyproject, ruff, pytest, pre-commit) 16 days ago
agentic_v1.jsonl

3.77 kB
Add the v1 eval dataset: 100 valid + 100 invalid + 20 agentic 10 days ago
checks.py

4.66 kB
Codebase quality pass: kill silent-truncation hacks, fix latent bugs, cover them (#46) 11 days ago
diagnose_retrieval.py

5.03 kB
retrieval: remove the cross-encoder reranker (ablation showed no earned keep) (#100) 8 days ago
docs_inventory.jsonl

3.83 MB
eval: docs inventory + index manifest from Actions [skip ci] 9 days ago
hallucinations.md

5.7 kB
Document M2 grounded before/after: 100% grounded_api_rate, fabrications eliminated 12 days ago
index_manifest.jsonl

1.51 MB
eval: docs inventory + index manifest from Actions [skip ci] 9 days ago
invalid_v1.jsonl

12.5 kB
Add the v1 eval dataset: 100 valid + 100 invalid + 20 agentic 10 days ago
questions_v0.jsonl

1.85 kB
M1: Answer schema, LLM wrapper with retry+repair, static checks, 15-question eval 13 days ago
questions_v1.jsonl

15.5 kB
Add the v1 eval dataset: 100 valid + 100 invalid + 20 agentic 10 days ago
retrieval_v0.jsonl

813 Bytes
Add a retrieval benchmark: recall@k + MRR over the v0 questions 11 days ago
run_agentic.py

4.51 kB
Make the agentic eval completable: bounded sample + 120m timeout 10 days ago
run_judge.py

10.5 kB
perf: hydrate retrieved sections concurrently — cut the dominant answer latency (#96) 8 days ago
run_retrieval.py

4.84 kB
retrieval: remove the cross-encoder reranker (ablation showed no earned keep) (#100) 8 days ago
run_v0.py

4.66 kB
Codebase quality pass: kill silent-truncation hacks, fix latent bugs, cover them (#46) 11 days ago