Hugging Face's logo

GetBeholder
/
Beholder-q4f16

Text Generation
MLC-LLM
beholder
state-extraction
roleplay
webllm
mlc
q4f16
webgpu
Model card Files
xet
 Community
Beholder-q4f16 / params_shard_4.bin

Commit History

  1. gc4b (v11 beta) q4f16: new shards + wasm + version.json bump v10->v11 (cache-bust)

    GetBeholder
    Verified signature
    b48e216

  1. v10 (cross-attribution v2): gc3 q4f16 clean weight-swap — new params + tensor-cache + version v9->v10 (wasm/config unchanged)

    GetBeholder
    Verified signature
    6d74380

  1. v9 (persona-sink fix): gc2 q4f16 clean weight-swap — new params + tensor-cache + version v8->v9 (wasm/config/tokenizer unchanged)

    GetBeholder
    Verified signature
    c026ac1

  2. gc1 -> v8 (grounding-clean): weight-swap + version bump v7->v8; worn P +5pts on gold, removal 38/38, drift Q4 |Δ|<=0.04

    GetBeholder
    Verified signature
    9678536

  1. v7 cross-attribution build (54->85%); weights+config+version, wasm unchanged

    GetBeholder
    Verified signature
    df3e496

  1. tkaug (v0.6, removal-capable): q4f16 bundle, version 0.6.0-tkaug

    GetBeholder
    Verified signature
    48ffbd8

  1. rc-cpc2-pad (0.8B, worn 0.67 @ max-new 2048; padding+bronze recipe) — overwrite 2026-07-05

    GetBeholder
    Verified signature
    323c9e7

  1. v5: MIN short-prompt build (q4f16) + clean README

    GetBeholder
    Verified signature
    2d91503

  1. Add throwaway Qwen3.5 q4f16 WebLLM bundle (weights + config + tokenizer + WebGPU .wasm)

    GetBeholder
    Verified signature
    6037ec9