software-mansion/react-native-executorch-distiluse-base-multilingual-cased-v2

Declare the int8 Core ML variant

Add int8 Core ML variant (embedding table quantized)

Core ML: accept a sequence length of 1

Apply the model card standard

Apply the model card standard

Stop restating quantized, and drop the unused default

Correct the published metadata

Add the Vulkan fp16 export

xnnpack: 8da4w is the default; fp32 stays as the accuracy reference

MLX int8; re-export xnnpack fp32 (stale schema); fix duplicate default

Core ML: pin compute_unit=CPU_ONLY, drop the fp32 variant

Fix Core ML artifacts: the published builds failed to load

Rebuild mlx export with fork toolchain (ExecuTorch 1.4.1)

Add backend config (ExecuTorch 1.4.1)

Add export (ExecuTorch 1.4.1)

Update .pte files for v0.10.0

Re-export text embeddings with get_dynamic_dims_forward dynamic-shape contract

Re-export xnnpack variants with get_dynamic_dims (variable-length input bounds)

fix(v0.9.0): snake_case rename of two dashed files

Add stub root config.json for HF download counter

Add spec-compliant config.json files

Add spec-compliant config.json files

Remove old-layout metadata orphaned by MODEL_SPEC.md restructure

Restructure to MODEL_SPEC.md convention

Add CoreML fp32+fp16 variants; update README with full variant matrix

Add XNNPACK 8da4w quantized variant

Initial upload: distiluse-base-multilingual-cased-v2 XNNPACK fp32 for RNE v0.9.0

initial commit