HappyZ

happyz synced commits to refs/pull/7642/merge at happyz/llama.cpp from mirror 2024-06-03 06:20:23 -07:00

628ec291a2 Merge 4d78cff3ed into 3d7ebf6312

3d7ebf6312 Vulkan Mixture of Experts (MoE) support (#7628)

a10cda58d3 cmake : add pkg-config spec file for llama.cpp (#7702)

6f28a333c1 llama : MiniCPM support tied embeddings (#7664)

549279d804 llama : avoid double token-to-piece cache (#7654)

Compare 6 commits »

happyz synced commits to refs/pull/7634/merge at happyz/llama.cpp from mirror 2024-06-03 06:20:23 -07:00

11962e1da5 Merge ddac1ef681 into 3d7ebf6312

3d7ebf6312 Vulkan Mixture of Experts (MoE) support (#7628)

a10cda58d3 cmake : add pkg-config spec file for llama.cpp (#7702)

6f28a333c1 llama : MiniCPM support tied embeddings (#7664)

549279d804 llama : avoid double token-to-piece cache (#7654)

Compare 6 commits »

happyz synced commits to refs/pull/7640/head at happyz/llama.cpp from mirror 2024-06-03 06:20:23 -07:00

243a3e4bb2 Update llama.cpp

464c75c00e add rpc-server to Makefile

805cd78c5a - fix copy_tensor being called on the src buffer instead of the dst buffer

369213eda9 llama : offload to RPC in addition to other backends

549279d804 llama : avoid double token-to-piece cache (#7654)

Compare 28 commits »

happyz synced commits to refs/pull/7640/merge at happyz/llama.cpp from mirror 2024-06-03 06:20:23 -07:00

5d102ace52 Merge 243a3e4bb2 into 3d7ebf6312

243a3e4bb2 Update llama.cpp

3d7ebf6312 Vulkan Mixture of Experts (MoE) support (#7628)

a10cda58d3 cmake : add pkg-config spec file for llama.cpp (#7702)

6f28a333c1 llama : MiniCPM support tied embeddings (#7664)

Compare 10 commits »

happyz synced commits to refs/pull/7606/head at happyz/llama.cpp from mirror 2024-06-03 06:20:22 -07:00

e0b077d4da Update ggml.c

happyz synced commits to refs/pull/7606/merge at happyz/llama.cpp from mirror 2024-06-03 06:20:22 -07:00

2394e671f4 Merge e0b077d4da into 3d7ebf6312

e0b077d4da Update ggml.c

3d7ebf6312 Vulkan Mixture of Experts (MoE) support (#7628)

a10cda58d3 cmake : add pkg-config spec file for llama.cpp (#7702)

6f28a333c1 llama : MiniCPM support tied embeddings (#7664)

Compare 7 commits »

happyz synced commits to refs/pull/7553/merge at happyz/llama.cpp from mirror 2024-06-03 06:20:22 -07:00

d6febdcffa Merge 5833323754 into 3d7ebf6312

3d7ebf6312 Vulkan Mixture of Experts (MoE) support (#7628)

a10cda58d3 cmake : add pkg-config spec file for llama.cpp (#7702)

6f28a333c1 llama : MiniCPM support tied embeddings (#7664)

549279d804 llama : avoid double token-to-piece cache (#7654)

Compare 6 commits »

happyz synced commits to refs/pull/7582/merge at happyz/llama.cpp from mirror 2024-06-03 06:20:22 -07:00

08adeb1b90 Merge 243b5efe0586c1e6fff749fbb52981e01d557bc7 into 6f28a333c1

6f28a333c1 llama : MiniCPM support tied embeddings (#7664)

549279d804 llama : avoid double token-to-piece cache (#7654)

9e405b6e2e kompute : implement op_getrows_f32 (#6403)

Compare 4 commits »

happyz synced commits to refs/pull/7596/merge at happyz/llama.cpp from mirror 2024-06-03 06:20:22 -07:00

a3d02baf3b Merge 96a6f55222 into 3d7ebf6312

3d7ebf6312 Vulkan Mixture of Experts (MoE) support (#7628)

a10cda58d3 cmake : add pkg-config spec file for llama.cpp (#7702)

6f28a333c1 llama : MiniCPM support tied embeddings (#7664)

549279d804 llama : avoid double token-to-piece cache (#7654)

Compare 8 commits »

happyz synced commits to refs/pull/7599/merge at happyz/llama.cpp from mirror 2024-06-03 06:20:22 -07:00

aec6c51662 Merge a913ca4cb9 into 3d7ebf6312

3d7ebf6312 Vulkan Mixture of Experts (MoE) support (#7628)

a10cda58d3 cmake : add pkg-config spec file for llama.cpp (#7702)

6f28a333c1 llama : MiniCPM support tied embeddings (#7664)

549279d804 llama : avoid double token-to-piece cache (#7654)

Compare 6 commits »

happyz synced commits to refs/pull/7514/merge at happyz/llama.cpp from mirror 2024-06-03 06:20:21 -07:00

763abdbd77 Merge 15d5c257a0 into 3d7ebf6312

3d7ebf6312 Vulkan Mixture of Experts (MoE) support (#7628)

a10cda58d3 cmake : add pkg-config spec file for llama.cpp (#7702)

6f28a333c1 llama : MiniCPM support tied embeddings (#7664)

549279d804 llama : avoid double token-to-piece cache (#7654)

Compare 6 commits »

happyz synced commits to refs/pull/7531/head at happyz/llama.cpp from mirror 2024-06-03 06:20:21 -07:00

8fb57ac0fb llama : use im2col and mul_mat to perform convolution for Mamba

happyz synced commits to refs/pull/7531/merge at happyz/llama.cpp from mirror 2024-06-03 06:20:21 -07:00

5c346b2674 Merge 8fb57ac0fb into 3d7ebf6312

3d7ebf6312 Vulkan Mixture of Experts (MoE) support (#7628)

a10cda58d3 cmake : add pkg-config spec file for llama.cpp (#7702)

6f28a333c1 llama : MiniCPM support tied embeddings (#7664)

549279d804 llama : avoid double token-to-piece cache (#7654)

Compare 7 commits »

happyz synced commits to refs/pull/7547/merge at happyz/llama.cpp from mirror 2024-06-03 06:20:21 -07:00

a3d881c8a9 Merge 150111f419 into a10cda58d3

a10cda58d3 cmake : add pkg-config spec file for llama.cpp (#7702)

6f28a333c1 llama : MiniCPM support tied embeddings (#7664)

549279d804 llama : avoid double token-to-piece cache (#7654)

9e405b6e2e kompute : implement op_getrows_f32 (#6403)

Compare 8 commits »

happyz synced commits to refs/pull/7526/merge at happyz/llama.cpp from mirror 2024-06-03 06:20:21 -07:00

c6d35a2f97 Merge 6a8aa22aaa into 3d7ebf6312

3d7ebf6312 Vulkan Mixture of Experts (MoE) support (#7628)

a10cda58d3 cmake : add pkg-config spec file for llama.cpp (#7702)

6f28a333c1 llama : MiniCPM support tied embeddings (#7664)

549279d804 llama : avoid double token-to-piece cache (#7654)

Compare 6 commits »

happyz synced commits to refs/pull/7487/merge at happyz/llama.cpp from mirror 2024-06-03 06:20:20 -07:00

8b1a904586 Merge 0adedd712e into a10cda58d3

a10cda58d3 cmake : add pkg-config spec file for llama.cpp (#7702)

6f28a333c1 llama : MiniCPM support tied embeddings (#7664)

549279d804 llama : avoid double token-to-piece cache (#7654)

9e405b6e2e kompute : implement op_getrows_f32 (#6403)

Compare 13 commits »

happyz synced commits to refs/pull/7499/merge at happyz/llama.cpp from mirror 2024-06-03 06:20:20 -07:00

e3b4ae95aa Merge 197eab2dd5e95387269c4a69ff9731917bc055bc into 3d7ebf6312

3d7ebf6312 Vulkan Mixture of Experts (MoE) support (#7628)

a10cda58d3 cmake : add pkg-config spec file for llama.cpp (#7702)

6f28a333c1 llama : MiniCPM support tied embeddings (#7664)

549279d804 llama : avoid double token-to-piece cache (#7654)

Compare 6 commits »

happyz synced commits to refs/pull/7477/merge at happyz/llama.cpp from mirror 2024-06-03 06:20:20 -07:00

c03ee0a6cd Merge 74716945ff66f2a8b437bac69367aced99099e06 into 3d7ebf6312

3d7ebf6312 Vulkan Mixture of Experts (MoE) support (#7628)

a10cda58d3 cmake : add pkg-config spec file for llama.cpp (#7702)

6f28a333c1 llama : MiniCPM support tied embeddings (#7664)

549279d804 llama : avoid double token-to-piece cache (#7654)

Compare 8 commits »

happyz synced commits to refs/pull/7458/merge at happyz/llama.cpp from mirror 2024-06-03 06:20:20 -07:00

eb3595aa3c Merge 590720fa38 into 3413ae2193

3413ae2193 fix bug introduced in using calloc (#7701)

1669810d7c flake.lock: Update (#7686)

7c4e5b7eae chore : add ignore rule for generated server themes (#7689)

9422c5e34b [SYCL] Update rpc-server.cpp to include SYCL backend (#7682)

Compare 10 commits »

happyz synced commits to refs/pull/7392/merge at happyz/llama.cpp from mirror 2024-06-03 06:20:20 -07:00

1177c3821b Merge 8404d201afd9fb0fe7dbc2acda6a2d4fd6309b4c into a10cda58d3

a10cda58d3 cmake : add pkg-config spec file for llama.cpp (#7702)

6f28a333c1 llama : MiniCPM support tied embeddings (#7664)

549279d804 llama : avoid double token-to-piece cache (#7654)

9e405b6e2e kompute : implement op_getrows_f32 (#6403)

Compare 8 commits »