llama.cpp/CHANGELOG.md

Changelog

All notable changes to this project will be documented in this file.

The format is based on Keep a Changelog, and this project adheres to Semantic Versioning.

[Unreleased]

[0.2.2]

• Fix bug in pip install of v0.2.1 due to scikit-build-core removing all .metal files in the source distribution (see #701)

[0.2.1]

• Fix bug in pip install of v0.2.0 due to .git folder being included in the source distribution (see #701)

[0.2.0]

• Migrated to scikit-build-core build system by @abetlen in #499
• Use numpy views for LogitsProcessor and StoppingCriteria instead of python lists by @abetlen in #499
• Drop support for end-of-life Python3.7 by @abetlen in #499
• Convert low level llama.cpp constants to use basic python types instead of ctypes types by @abetlen in #499

[0.1.85]

• Add llama_cpp.__version__ attribute by @janvdp in #684
• Fix low level api examples by @jbochi in #680

[0.1.84]

• Update llama.cpp

[0.1.83]

• Update llama.cpp

[0.1.82]

• Update llama.cpp

[0.1.81]

• Update llama.cpp

[0.1.80]

• Update llama.cpp

[0.1.79]

• GGUF Support (breaking change requiring new model format)

[0.1.78]

• Grammar based sampling via LlamaGrammar which can be passed to completions
• Make n_gpu_layers == -1 offload all layers

[0.1.77]

• (llama.cpp) Update llama.cpp add support for LLaMa 2 70B
• (server) Add temporary n_gqa and rms_norm_eps parameters required for LLaMa 2 70B

[0.1.76]

• (llama.cpp) Update llama.cpp add support for LLaMa 2 70B

[0.1.75]

• Update llama.cpp

[0.1.74]

• (server) OpenAI style error responses

[0.1.73]

• (server) Add rope parameters to server settings

[0.1.72]

• (llama.cpp) Update llama.cpp added custom_rope for extended context lengths

[0.1.71]

• (llama.cpp) Update llama.cpp

• (server) Fix several pydantic v2 migration bugs

[0.1.70]

• (Llama.create_completion) Revert change so that max_tokens is not truncated to context_size in create_completion
• (server) Fixed changed settings field names from pydantic v2 migration

[0.1.69]

• (server) Streaming requests can are now interrupted pre-maturely when a concurrent request is made. Can be controlled with the interrupt_requests setting.
• (server) Moved to fastapi v0.100.0 and pydantic v2
• (docker) Added a new "simple" image that builds llama.cpp from source when started.
• (server) performance improvements by avoiding unnecessary memory allocations during sampling

[0.1.68]

• (llama.cpp) Update llama.cpp

[0.1.67]

• Fix performance bug in Llama model by pre-allocating memory tokens and logits.
• Fix bug in Llama model where the model was not free'd after use.

[0.1.66]

• (llama.cpp) New model API

• Performance issue during eval caused by looped np.concatenate call

• State pickling issue when saving cache to disk

[0.1.65]

• (llama.cpp) Fix struct misalignment bug

[0.1.64]

• (llama.cpp) Update llama.cpp
• Fix docs for seed. Set -1 for random.

[0.1.63]

• (llama.cpp) Add full gpu utilisation in CUDA
• (llama.cpp) Add get_vocab
• (llama.cpp) Add low_vram parameter
• (server) Add logit_bias parameter

[0.1.62]

• Metal support working
• Cache re-enabled

[0.1.61]

• Fix broken pip installation

[0.1.60]

NOTE: This release was deleted due to a bug with the packaging system that caused pip installations to fail.

• Truncate max_tokens in create_completion so requested tokens doesn't exceed context size.
• Temporarily disable cache for completion requests

[v0.1.59]

• (llama.cpp) k-quants support
• (server) mirostat sampling parameters to server
• Support both .so and .dylib for libllama on MacOS

[v0.1.58]

• (llama.cpp) Metal Silicon support

[v0.1.57]

• (llama.cpp) OpenLlama 3B support

[v0.1.56]

• (misc) Added first version of the changelog
• (server) Use async routes
• (python-api) Use numpy for internal buffers to reduce memory usage and improve performance.
• (python-api) Performance bug in stop sequence check slowing down streaming.