-
Prune QHexRT HNPU bundle to runtime-minimum artifacts
Remove 5 non-runtime/sensitive files from the QHexRT HNPU bundle.
Strip non-runtime _comment metadata from 2 manifest(s) without changing runtime fields.
Audit artifacts: QHexRT/hf_hnpu_audit cleanup plan, manifest reference check, and multi-agent review.
-
Add chat template block (runtime-owned chat templating)
-
Scrub internal engineering notes from manifest _comment
-
Upload v81/experts_i8.bin with huggingface_hub
-
Upload v81/special_tokens_map.json with huggingface_hub
-
Upload v81/tokenizer_config.json with huggingface_hub
-
Upload v81/tokenizer.json with huggingface_hub
-
Upload v81/embed_f16.bin with huggingface_hub
-
Upload v81/experts_scale.f32 with huggingface_hub
-
Upload v81/lmhead_ar1.bin with huggingface_hub
-
Upload v81/ffn2.bin with huggingface_hub
-
Upload v81/a_hi.bin with huggingface_hub
-
Upload v81/a_lo.bin with huggingface_hub
-
Upload v81/phimoe.json with huggingface_hub
