From 9719f513c0174ed3afade9facdca0b63cde9f6ef Mon Sep 17 00:00:00 2001 From: Neo Zhang <14088817+arthw@users.noreply.github.com> Date: Sat, 15 Jun 2024 10:23:07 +0800 Subject: [PATCH 1/4] Update README-sycl.md --- README-sycl.md | 16 ++++++++++++++++ 1 file changed, 16 insertions(+) diff --git a/README-sycl.md b/README-sycl.md index 93b623daf6a1a..cf978a618bf5a 100644 --- a/README-sycl.md +++ b/README-sycl.md @@ -1,6 +1,7 @@ # llama.cpp for SYCL - [Background](#background) +- [Recommended Release](#recommended-release) - [News](#news) - [OS](#os) - [Hardware](#hardware) @@ -31,8 +32,23 @@ When targeting **Intel CPU**, it is recommended to use llama.cpp for [Intel oneM It has the similar design of other llama.cpp BLAS-based paths such as *OpenBLAS, cuBLAS, etc..*. In beginning work, the oneAPI's [SYCLomatic](https://github.com/oneapi-src/SYCLomatic) open-source migration tool (Commercial release [Intel® DPC++ Compatibility Tool](https://www.intel.com/content/www/us/en/developer/tools/oneapi/dpc-compatibility-tool.html)) was used for this purpose. +## Recommended Release + +The SYCL backend would be broken by some PRs due to no online CI. + +Here is the releases with good quality: + +|Commit ID|Tag|Release| +|-|-|-| +|fb76ec31a9914b7761c1727303ab30380fd4f05c|b3038 |[llama-b3038-bin-win-sycl-x64.zip](https://github.com/ggerganov/llama.cpp/releases/download/b3038/llama-b3038-bin-win-sycl-x64.zip) | + + ## News +- 2024.5 + - Performance is increased: 34 -> 37 tokens/s of llama-2-7b.Q4_0 on Arc770. + - Arch Linux is verified successfully. + - 2024.4 - Support data types: GGML_TYPE_IQ4_NL, GGML_TYPE_IQ4_XS, GGML_TYPE_IQ3_XXS, GGML_TYPE_IQ3_S, GGML_TYPE_IQ2_XXS, GGML_TYPE_IQ2_XS, GGML_TYPE_IQ2_S, GGML_TYPE_IQ1_S, GGML_TYPE_IQ1_M. From c8cfc963f09d6fca1567e2faa08d545798469cb2 Mon Sep 17 00:00:00 2001 From: Neo Zhang Date: Sat, 15 Jun 2024 13:34:29 +0800 Subject: [PATCH 2/4] Update README-sycl.md --- README-sycl.md | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/README-sycl.md b/README-sycl.md index cf978a618bf5a..58da1228b19ff 100644 --- a/README-sycl.md +++ b/README-sycl.md @@ -48,7 +48,7 @@ Here is the releases with good quality: - 2024.5 - Performance is increased: 34 -> 37 tokens/s of llama-2-7b.Q4_0 on Arc770. - Arch Linux is verified successfully. - + - 2024.4 - Support data types: GGML_TYPE_IQ4_NL, GGML_TYPE_IQ4_XS, GGML_TYPE_IQ3_XXS, GGML_TYPE_IQ3_S, GGML_TYPE_IQ2_XXS, GGML_TYPE_IQ2_XS, GGML_TYPE_IQ2_S, GGML_TYPE_IQ1_S, GGML_TYPE_IQ1_M. From 59dc0acbacb315d5eaa68ab82998c300cda70b29 Mon Sep 17 00:00:00 2001 From: Neo Zhang Date: Mon, 17 Jun 2024 09:44:18 +0800 Subject: [PATCH 3/4] Update README-sycl.md --- README-sycl.md | 8 ++++---- 1 file changed, 4 insertions(+), 4 deletions(-) diff --git a/README-sycl.md b/README-sycl.md index 58da1228b19ff..e08d48a4eaed5 100644 --- a/README-sycl.md +++ b/README-sycl.md @@ -36,11 +36,11 @@ It has the similar design of other llama.cpp BLAS-based paths such as *OpenBLAS, The SYCL backend would be broken by some PRs due to no online CI. -Here is the releases with good quality: +The following release is verified with good quality: -|Commit ID|Tag|Release| -|-|-|-| -|fb76ec31a9914b7761c1727303ab30380fd4f05c|b3038 |[llama-b3038-bin-win-sycl-x64.zip](https://github.com/ggerganov/llama.cpp/releases/download/b3038/llama-b3038-bin-win-sycl-x64.zip) | +|Commit ID|Tag|Release|Verify Platform| +|-|-|-|-| +|fb76ec31a9914b7761c1727303ab30380fd4f05c|b3038 |[llama-b3038-bin-win-sycl-x64.zip](https://github.com/ggerganov/llama.cpp/releases/download/b3038/llama-b3038-bin-win-sycl-x64.zip) |Arc770/Linux/oneAPI 2024.1
MTL Arc GPU/Windows 11/oneAPI 2024.1| ## News From 5b7b7e0894fb8696ba81050fe04d1ec3a615ebbd Mon Sep 17 00:00:00 2001 From: Neo Zhang Date: Mon, 17 Jun 2024 09:54:48 +0800 Subject: [PATCH 4/4] Update README-sycl.md --- README-sycl.md | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/README-sycl.md b/README-sycl.md index e08d48a4eaed5..bd1984706225f 100644 --- a/README-sycl.md +++ b/README-sycl.md @@ -38,7 +38,7 @@ The SYCL backend would be broken by some PRs due to no online CI. The following release is verified with good quality: -|Commit ID|Tag|Release|Verify Platform| +|Commit ID|Tag|Release|Verified Platform| |-|-|-|-| |fb76ec31a9914b7761c1727303ab30380fd4f05c|b3038 |[llama-b3038-bin-win-sycl-x64.zip](https://github.com/ggerganov/llama.cpp/releases/download/b3038/llama-b3038-bin-win-sycl-x64.zip) |Arc770/Linux/oneAPI 2024.1
MTL Arc GPU/Windows 11/oneAPI 2024.1|