| Commit message (Collapse) | Author | Age | Files | Lines | |
|---|---|---|---|---|---|
| * | kernels: drop asr_ prefix naming | Onur Gündüz | 2026-06-14 | 1 | -1745/+0 |
| | | |||||
| * | kernels: fix OpenBLAS thread contention and numerical instability | Onur Gündüz | 2026-06-14 | 1 | -2/+30 |
| | | |||||
| * | kernels: parallelize large rmsnorm rows | Onur Gündüz | 2026-06-14 | 1 | -5/+48 |
| | | |||||
| * | late: add packed maxsim scoring | Onur Gündüz | 2026-06-14 | 1 | -4/+8 |
| | | |||||
| * | kernels: vectorize bf16 widening on arm | Onur Gündüz | 2026-06-14 | 1 | -0/+14 |
| | | |||||
| * | kernels: use sgemv for single-row f32 linear | Onur Gündüz | 2026-06-14 | 1 | -4/+10 |
| | | |||||
| * | kernels: add neon rope apply path | Onur Gündüz | 2026-06-14 | 1 | -0/+21 |
| | | |||||
| * | cpu: reuse tiled attention scratch buffers | Onur Gündüz | 2026-06-14 | 1 | -12/+46 |
| | | |||||
| * | cpu: tile packed attention through blas | Onur Gündüz | 2026-06-14 | 1 | -54/+110 |
| | | |||||
| * | kernels,cpu: fuse tiny BF16 gate up projections | Onur Gündüz | 2026-06-14 | 1 | -0/+93 |
| | | |||||
| * | kernels,cpu: fuse small BF16 QKV projections | Onur Gündüz | 2026-06-14 | 1 | -15/+40 |
| | | |||||
| * | kernels: validate real BF16 artifacts | Onur Gündüz | 2026-06-14 | 1 | -0/+52 |
| | | |||||
| * | cpu: fuse CPU embedding hot paths | Onur Gündüz | 2026-06-14 | 1 | -0/+26 |
| | | |||||
| * | kernels: avoid tiny packed attention dispatch | Onur Gündüz | 2026-06-14 | 1 | -1/+7 |
| | | |||||
| * | all: support batched embedding attention | Onur Gündüz | 2026-06-14 | 1 | -0/+95 |
| | | |||||
| * | Refactor and enhance pplx_embed and pplx_embed_mlx for improved model support | Onur Gündüz | 2026-03-28 | 1 | -5/+0 |
| | | | | | | | | | | | | | - Updated `pplx_embed.h` to support new embedding models: `pplx-embed-v1-4b` and `pplx-embed-context-v1-4b`. - Simplified model configuration structure and constants for better maintainability. - Enhanced comments for clarity on model architecture and configuration. - Modified `pplx_embed_mlx.c` to streamline loading of model weights and configuration from `config.json`. - Improved memory management and error handling in the MLX backend. - Updated embedding computation to support dynamic hidden sizes based on model configuration. - Added `pplx_mlx_config` function to retrieve model configuration details. - Updated README.md - Added .gitignore | ||||
| * | Init | Onur Gündüz | 2026-03-28 | 1 | -0/+1247 |
