Llama-cpp-python

- feat: Update llama.cpp to ggerganov/llama.cpp794fe23f29fb40104975c91fe19f23798f7c726e
- fix(ci): Fix the CUDA workflow by oobabooga in 1894
- fix: error showing time spent in llama perf context print, adds `no_perf` flag to `Llama` class by shakalaca in 1898

- feat: Update llama.cpp to ggerganov/llama.cppf7cd13301c2a88f97073fd119072b4cc92c08df1
- fix(server): streaming resource lock by gjpower in 1879

- feat: Update llama.cpp to ggerganov/llama.cpp26a8406ba9198eb6fdd8329fa717555b4f77f05f
- fix(ci): Fix release by updating macos runner image to non-deprecated version by abetlen in afedfc888462f9a6e809dc9455eb3b663764cc3f
- fix(server): add missing await statements for async exit_stack handling by gjpower in 1858

- fix(ci): Build wheels for macos 13-15, cuda 12.1-12.4 by abetlen in ca808028bd16b8327bd84128d48015a4b1304690

- feat: Update llama.cpp to ggerganov/llama.cppce8784bdb153ff7794dde5a50b0ebfa51baa6171
- fix: chat API logprobs format by domdomegg in 1788
- feat: Add support for CUDA 12.6, fix CUDA 12.5 by Smartappli in 1775
- fix: Make content not required in ChatCompletionRequestAssistantMessage by feloy in 1807
- fix: Fix pickling of Llama class by setting seed from _seed member by abetlen in 2523472c3eccb9ab9277117cc4ff705212b6888a
- fix: Fix logit-bias type hint by ddh0 in 1802
- fix(server): Avoid thread starvation on many concurrent requests by making use of asyncio to lock llama_proxy context by gjpower in 1798
- fix(server): Added missing exit_stack.close() to /v1/chat/completions by Ian321 in 1796
- fix(examples): Refactor Batching notebook to use new sampler chain API by lukestanley in 1793
- fix(docs): Update development instructions by Florents-Tselai in 1833
- fix(docs): Remove ref to llama_eval in llama_cpp.py docs by richdougherty in 1819

- feat: Update llama.cpp to ggerganov/llama.cpp74d73dc85cc2057446bf63cc37ff649ae7cebd80