Bumps llama-cpp-python from 0.3.14 to 0.3.23.

Sourced from llama-cpp-python's changelog.

[0.3.23]

• feat: Update llama.cpp to ggml-org/llama.cpp@7d442abf
• fix: Correct batched embedding outputs for multi-sequence embed() calls by @​Anai-Guo in #2205
• fix: Configure embedding contexts with enough sequence slots for batched embed() calls
• fix: Mark all embedding input tokens as outputs to avoid llama.cpp override warnings by @​Anai-Guo in #2212

[0.3.22]

• feat: Update llama.cpp to ggml-org/llama.cpp@63d93d173
• feat(ci): Re-enable Windows CUDA wheels and add CUDA 12.5.1 wheel builds

[0.3.21]

• feat: Update llama.cpp to ggml-org/llama.cpp@f53577432 and sync Python bindings
• fix(ci): Build one arm64 release wheel for py3-none wheel publishing

[0.3.20]

• refactor: Replace deprecated llama.cpp references in library, docs, and examples by @​abetlen in #2170
• feat: Update llama.cpp to ggml-org/llama.cpp@f49e917 by @​abetlen in #2169
• feat(server): Add model-load chat_template_kwargs support and document the CLI/config usage by @​abetlen in #2168
• ci: Publish release wheels as py3-none by @​Bing-su in #2166
• fix(ci): Publish distinct manylinux and musllinux CPU wheels by @​abetlen in #2165

[0.3.19]

• feat: Update llama.cpp to ggml-org/llama.cpp@c0159f9 by @​abetlen in #2161
• fix: Handle embedding models without KV memory and test embeddings with a real GGUF embedding model by @​abetlen in #2160
• fix(ci): Shrink CUDA wheel fatbins so CUDA releases stay under GitHub's asset size limit by @​abetlen in #2158

[0.3.18]

• feat: Expose attention_type in Llama.__init__ for non-causal embedding models by @​jamesbiederbeck in #2143
• fix(ci): Build Docker images from the checked-out source and sanitize branch tags by @​abetlen in #2156
• fix(ci): Fix the CUDA wheel workflow and keep release tags aligned with the built toolkit by @​abetlen in #2155
• fix(ci): Speed up release wheel builds by moving arm64 off QEMU and parallelizing riscv64 by @​abetlen in #2154

[0.3.17]

• feat: Update llama.cpp to ggml-org/llama.cpp@49bfdde and sync Python bindings by @​abetlen in #2151
• fix: Handle Qwen 3.5 hybrid prefix reuse by @​codavidgarcia and @​r-dh in #2152
• chore(dev): Add Ruff-based formatting and a safe lint baseline, and run it in CI for pull requests and pushes to main
• fix(ci): Run macOS CI on supported Apple Silicon and Intel runners by @​abetlen in #2150
• fix(ci): Use the hf CLI instead of the deprecated huggingface-cli name in GitHub Actions and docs by @​abetlen in #2149
• ci: add riscv64 wheel builds to release workflow by @​gounthar in #2139

[0.3.16]

• feat: Update llama.cpp to ggml-org/llama.cpp@4227c9b

... (truncated)

• 4a1a8ec chore: bump version to 0.3.23 (#2215)
• 5684112 feat: update llama.cpp to 7d442abf (#2214)
• f8c1f36 fix(embed): mark all tokens as output to suppress llama.cpp 'overriding' INFO...
• f774690 feat: update llama.cpp to 5d6f18a63 (#2207)
• 128c331 fix: configure n_seq_max for batched embeddings (#2206)
• 90e8df9 fix(internals): use n_tokens0 offset when enabling last-token logits in add...
• 14d7846 fix(ci): skip unsupported Windows CUDA versions (#2204)
• bc6ff9f fix(ci): install CUDA CCCL headers for wheel builds (#2203)
• 04a3638 fix(ci): pass CUDA compiler arg for Windows detection (#2202)
• 2bfd80c fix(ci): pass CUDA unsupported compiler flag during detection (#2201)
• Additional commits viewable in compare view

Dependabot will resolve any conflicts with this PR as long as you don't alter it yourself. You can also trigger a rebase manually by commenting @dependabot rebase.

You can trigger Dependabot actions by commenting on this PR:

• @dependabot rebase will rebase this PR
• @dependabot recreate will recreate this PR, overwriting any edits that have been made to it
• @dependabot show <dependency name> ignore conditions will show all of the ignore conditions of the specified dependency
• @dependabot ignore this major version will close this PR and stop Dependabot creating any more for this major version (unless you reopen the PR or upgrade to it yourself)
• @dependabot ignore this minor version will close this PR and stop Dependabot creating any more for this minor version (unless you reopen the PR or upgrade to it yourself)
• @dependabot ignore this dependency will close this PR and stop Dependabot creating any more for this dependency (unless you reopen the PR or upgrade to it yourself)