Oct 20 5 days ago 13:35 Publish schema-compliant artifact method signatures Bartosz Hanc
Oct 16 9 days ago 18:49 fix(xnnpack): preserve PLE dequantization and sensitive projection precision Oct 15 10 days ago 19:03 Remove validation record and restore model card Bartosz Hanc
19:00 fix(xnnpack): bound ring-cache prefill to 512 tokens Bartosz Hanc
11:35 fix(config): fix get_max_context_len typo from 4048 to 4096 Bartosz Hanc
Oct 13 12 days ago 21:03 Update Gemma 4 E2B XNNPACK multimodal model (v0.11.0) Bartosz Hanc
- Add get_model_schema method and companion schema
- Fix token_embedding output boundary dtypes to float32
- Fix audio_encoder to return both audio_embeds and audio_mask with 1-D num_valid_samples input
- Align audio chunking constants with RNE spec Sep 3 14:48 Stop restating quantized, and drop the unused default quantized duplicated precisions.json, which is authoritative for the
quantized/non-quantized partition. default expressed a per-repository
recommendation nothing consumed: the library resolves its own per-platform
default. Both are gone from the schema. Jun 22 08:39 Backfill constant metadata method values in config.json from .pte Jun 10 11:52 Delete e2b/xnnpack/tokenizer_config.json 11:52 Delete e2b/xnnpack/tokenizer.json May 25 13:03 Upload folder using huggingface_hub 08:47 Upload folder using huggingface_hub 
