Oct 16 27 days ago 18:49 fix(xnnpack): preserve PLE dequantization and sensitive projection precision Oct 15 28 days ago 20:13 fix(vulkan): normalize fused dtypes and resize casts 19:03 Remove validation record and restore model card 19:00 fix(xnnpack): bound ring-cache prefill to 512 tokens Oct 14 29 days ago 17:45 docs: update model card, context lengths, and config specs Oct 13 30 days ago 19:24 fix(vulkan): use F.embedding for sliding window masks 18:40 fix(vulkan): reduce context length to 2048 to avoid device OOM 16:07 fix(vulkan): bind graph input start_pos directly in custom attention transforms 15:31 fix(vulkan): restore correct prefill attention masks and positions 14:44 fix(vulkan): pass scalar graph input input_pos to layers 14:12 fix(vulkan): re-export gemma_4_e2b_vulkan_8da4w with initialized mutable buffers Oct 12 13:28 update models 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 09:43 Use min/max dynamic shape bounds for forward inputs/outputs 08:39 Backfill constant metadata method values in config.json from .pte Jun 10 12:02 Delete e2b/mlx/tokenizer_config.json 12:02 Delete e2b/mlx/tokenizer.json 11:53 Delete e2b/xnnpack/tokenizer_config.json 11:53 Delete e2b/xnnpack/tokenizer.json 11:53 Delete e2b/vulkan/tokenizer_config.json 11:53 Delete e2b/vulkan/tokenizer.json 11:50 Upload folder using huggingface_hub Jun 9 10:44 Delete e2b/mlx/config.json 10:44 Rename e2b/mlx/tokenizer_config (1).json to e2b/mlx/tokenizer_config.json 10:44 Rename e2b/mlx/tokenizer (4).json to e2b/mlx/tokenizer.json 10:44 Upload 2 files 10:43 Rename e2b/mlx/gemma4_e2b_mlx_int4_labs_v5.pte to e2b/mlx/gemma4_e2b_mlx_int4.pte 10:43 Upload gemma4_e2b_mlx_int4_labs_v5.pte 10:26 Create mlx/config.json May 25 08:45 Upload folder using huggingface_hub May 22 13:13 Upload folder using huggingface_hub 
