Commit Graph
4 Commits
Author SHA1 Message Date
Alexandre Teixeira b88154ecf9 Merge pull request #6730 from dxrpl/fix/hwfit-hf-param-count
fix(hwfit): keep the MoE active size for HF collection models (#6729)
2026-10-30 19:43:34 +00:00
dxrplandClaude Opus 5.5 6123b70b5c fix(hwfit): keep non-LLM HF collection models out of the LLM catalog (#6731)
Tracked HF collections also contain time-series forecasters, image
encoders and image/video generators. They were imported as "general"
LLMs and offered for vllm serve. Skip their pipeline tags; embedding,
reranker and RL-tagged LLM repos stay.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-10-15 01:15:08 +02:00
dxrplandClaude Opus 5.5 bfa514f6e8 fix(hwfit): keep the MoE active size for HF collection models (#6729)
_entry_from_collection_item parsed the repo name (the only source of the
"-A3B" active size) only when HF's numParameters was missing, which is
almost never. Every MoE model from the HF collection catalog was scored
as dense, with speed estimates 3-4x too low.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-10-15 01:11:46 +02:00
pewdiepie-archdaemon a1a14bd5c9 Checkpoint Odysseus local update 2026-07-07 00:50:07 +00:00