Age | Commit message (Expand) | Author |
2023-06-05 | ggml : fix internal overflow in ggml_time_us on Windows (#1702) | grahameth |
2023-06-05 | ggml : add SOTA 2,3,4,5,6 bit k-quantizations (#1684) | Kawrakow |
2023-06-04 | llama : Metal inference (#1642) | Georgi Gerganov |
2023-06-04 | OpenCL: Fix duplication of layers in VRAM and RAM, add GPU mul kernel (#1653) | 0cc4m |
2023-05-29 | ggml : sync cgraph import / export API | Georgi Gerganov |
2023-05-29 | ggml : fix bug in ggml_alibi | Georgi Gerganov |
2023-05-27 | ggml : add support for the RISCV architecture (#1616) | apcameron |
2023-05-27 | ggml : add ggml_tensor_overhead() | Georgi Gerganov |
2023-05-27 | ggml : sync ggml core (minor additions, e.g. ggml_get_tensor_by_name()) | Georgi Gerganov |
2023-05-23 | OpenCL Token Generation Acceleration (#1459) | 0cc4m |
2023-05-21 | ggml : output 3d sizes in ggml_graph_dump_dot() | Georgi Gerganov |
2023-05-20 | ggml : update WASM SIMD | Georgi Gerganov |
2023-05-20 | ggml : add ggml_clamp() (#1539) | Georgi Gerganov |
2023-05-20 | cuda : loading models directly into VRAM, norm calculation on GPU, broadcasti... | Johannes Gäßler |
2023-05-20 | llama : fix name shadowing and C4146 (#1526) | Maxime |
2023-05-20 | ggml : fix scalar implementation of Q4_1 dot | Georgi Gerganov |
2023-05-19 | ggml : use F16 instead of F32 in Q4_0, Q4_1, Q8_0 (#1508) | Georgi Gerganov |
2023-05-16 | ~7% faster Q5_1 AVX2 code (#1477) | Ilya Kurdyukov |
2023-05-14 | ggml : alternative fix for race condition bug in non-inplace ggml_compute_for... | xaedes |
2023-05-14 | ggml : various fixes (#1450) | Georgi Gerganov |
2023-05-14 | ggml : add AVX support based on AVX2 code (#1430) | katsu560 |
2023-05-13 | ggml : multi-thread mul and diag_mask ops (#1428) | Georgi Gerganov |
2023-05-13 | ggml : GPU-accelerated token generation (#1412) | Johannes Gäßler |
2023-05-13 | ggml : implement backward pass for llama + small training-llama-from-scratch ... | xaedes |
2023-05-13 | ggml : sync alibi fix from ggml repo | Georgi Gerganov |
2023-05-13 | Adding SSE instructions to ggml_vec_dot_q4_0_q8_0 (#1413) | 3ooabkhxtn |
2023-05-12 | ggml : remove bit shuffling (#1405) | Georgi Gerganov |
2023-05-09 | use pause asm insn in busyloop to run the CPU (13600K) 10 °C cooler (#1314) | Sami Farin |
2023-05-06 | ggml : Allow usage of CLBlast alongside Accelerate.framework (#1336) | swittk |
2023-05-04 | ggml : change immintrin.h to intrin.h for compatibility (#1307) | Ron Jailall |
2023-05-03 | ggml : vectorize Q8_0 quantization | Georgi Gerganov |
2023-05-02 | ggml : fix 32-bit ARM | Georgi Gerganov |
2023-05-02 | ggml : fix ppc64le build error and make cmake detect Power processors (#1284) | Marvin Gießing |
2023-05-02 | ggml: add names to tensors (#1268) | slaren |
2023-05-01 | cuBLAS: refactor and optimize f16 mat mul performance (#1259) | slaren |
2023-05-01 | ggml : fix ggml_used_mem() (#1264) | Kerfuffle |
2023-04-30 | ggml : fix UB (int << 31) | Georgi Gerganov |
2023-04-30 | ggml : add Q5 WASM SIMD + GGML_FTYPE | Georgi Gerganov |
2023-04-30 | ggml : fix labels for GGML_OP_ALIBI | Georgi Gerganov |
2023-04-29 | ggml : fix 32-bit ARM NEON | Georgi Gerganov |
2023-04-29 | ggml : use vzip instead of vuzp for consistency | Georgi Gerganov |
2023-04-29 | ggml : fix visibility and unused warnings | Georgi Gerganov |
2023-04-29 | ggml : fix #if for f32_f32 mul_mat (CLBlast) (#1229) | Georgi Gerganov |
2023-04-29 | ggml : adjust mul_mat_f16 work memory (#1226) | Georgi Gerganov |
2023-04-29 | cuBLAS: use host pinned memory and dequantize while copying (#1207) | slaren |
2023-04-29 | cuBLAS: non-contiguous tensor support (#1215) | Henri Vasserman |
2023-04-28 | Remove Q4_3 which is no better than Q5 (#1218) | Stephan Walter |
2023-04-28 | ggml : sync ggml (ggml_alibi) | Georgi Gerganov |
2023-04-28 | ggml : add helper debug printf in soft_max | Georgi Gerganov |
2023-04-28 | ggml : add CLBlast support (#1164) | 0cc4m |