v.1.5.3

[ggerganov]

Overview

Minor maintenance release:

• Fix CUDA issues where the transcription produces garbage
• FIX quantized models to work with CUDA backend
• Allow to use whisper.cpp and llama.cpp together in SwiftUI projects

What's Changed

• Update bench.py by @ForkedInTime in Update bench.py #1655
• cmake : Resolve quantized model issue when CUBLAS enabled by @bobqianic in cmake : Resolve quantized model issue when CUBLAS enabled #1667
• examples : Revert CMakeLists.txt for talk-llama by @bobqianic in examples : Revert CMakeLists.txt for talk-llama #1669
• CI : Add coverage for talk-llama when WHISPER_CUBLAS=1 by @bobqianic in CI : Add coverage for talk-llama when WHISPER_CUBLAS=1 #1672
• ci: build and push docker image by @OpenWaygate in ci: build and push docker image #1674
• sync : ggml (ggml_scale, ggml_row_size, etc.) by @ggerganov in sync : ggml (ggml_scale, ggml_row_size, etc.) #1677
• Replace WHISPER_PRINT_DEBUG with WHISPER_LOG_DEBUG by @bobqianic in Replace WHISPER_PRINT_DEBUG with WHISPER_LOG_DEBUG #1681
• download: Fix large q5 model name by @dimopep in download: Fix large q5 model name #1695
• sync : ggml (VMM, sync-ggml-am.sh, dotprod ARM fixes) by @ggerganov in sync : ggml (VMM, sync-ggml-am.sh, dotprod ARM fixes) #1691
• whisper : replace tensor->n_dims with ggml_n_dims(tensor) by @bobqianic in whisper : replace tensor->n_dims with ggml_n_dims(tensor) #1694
• Build with CLBlast by @tamo in Build with CLBlast #1576
• docker : Fix the Publishing of the CUDA Docker Image by @bobqianic in docker : Fix the Publishing of the CUDA Docker Image #1704
• emscripten: fix "Stack Overflow!" by @Huguet57 in emscripten: fix "Stack Overflow!" #1713
• sync : ggml by @ggerganov in sync : ggml #1717
• Add error handling to graph_compute by @finnvoor in Add error handling to graph_compute #1714
• Updates Package.swift to use ggml as package dependency by @1-ashraful-islam in Updates Package.swift to use ggml as package dependency #1701

New Contributors

• @ForkedInTime made their first contribution in Update bench.py #1655
• @OpenWaygate made their first contribution in ci: build and push docker image #1674
• @dimopep made their first contribution in download: Fix large q5 model name #1695
• @Huguet57 made their first contribution in emscripten: fix "Stack Overflow!" #1713
• @1-ashraful-islam made their first contribution in Updates Package.swift to use ggml as package dependency #1701

Full Changelog: v1.5.2...v1.5.3

---

This discussion was created from the release v.1.5.3.