The prebuilt vllm 0.14.1+cpu wheel from GitHub releases is compiled with
SIMD instructions (AVX-512 VNNI/BF16 or AMX-BF16) that not every CPU
supports. GitHub Actions ubuntu-latest runners SIGILL when vllm spawns
the model_executor.models.registry subprocess for introspection, so
LoadModel never reaches the actual inference path.
- install.sh: when FROM_SOURCE=true on a CPU build, temporarily hide
requirements-cpu-after.txt so installRequirements installs the base
deps + torch CPU without pulling the prebuilt wheel, then clone vllm
and compile it with VLLM_TARGET_DEVICE=cpu. The resulting binaries
target the host's actual CPU.
- backend/Dockerfile.python: accept a FROM_SOURCE build-arg and expose
it as an ENV so install.sh sees it during `make`.
- Makefile docker-build-backend: forward FROM_SOURCE as --build-arg
when set, so backends that need source builds can opt in.
- Makefile test-extra-backend-vllm: call docker-build-vllm via a
recursive $(MAKE) invocation so FROM_SOURCE flows through.
- .github/workflows/test-extra.yml: set FROM_SOURCE=true on the
tests-vllm-grpc job. Slower but reliable — the prebuilt wheel only
works on hosts that share the build-time SIMD baseline.
Answers 'did you test locally?': yes, end-to-end on my local machine
with the prebuilt wheel (CPU supports AVX-512 VNNI). The CI runner CPU
gap was not covered locally — this commit plugs that gap.
Extends tests/e2e-backends to cover backends that:
- Resolve HuggingFace model ids natively (vllm, vllm-omni) instead of
loading a local file: BACKEND_TEST_MODEL_NAME is passed verbatim as
ModelOptions.Model with no download/ModelFile.
- Parse tool calls into ChatDelta.tool_calls: new "tools" capability
sends a Predict with a get_weather function definition and asserts
the Reply contains a matching ToolCallDelta. Uses UseTokenizerTemplate
with OpenAI-style Messages so the backend can wire tools into the
model's chat template.
- Need backend-specific Options[]: BACKEND_TEST_OPTIONS lets a test set
e.g. "tool_parser:hermes,reasoning_parser:qwen3" at LoadModel time.
Adds make target test-extra-backend-vllm that:
- docker-build-vllm
- loads Qwen/Qwen2.5-0.5B-Instruct
- runs health,load,predict,stream,tools with tool_parser:hermes
Drops backend/python/vllm/test_{cpu_inference,tool_calls}.py — those
standalone scripts were scaffolding used while bringing up the Python
backend; the e2e-backends harness now covers the same ground uniformly
alongside llama-cpp and ik-llama-cpp.
Concrete vLLM tool parsers override the abstract base's __init__ and
drop the tools kwarg (e.g. Hermes2ProToolParser only takes tokenizer).
Instantiating with tools= raised TypeError which was silently caught,
leaving chat_deltas.tool_calls empty.
Retry the constructor without the tools kwarg on TypeError — tools
aren't required by these parsers since extract_tool_calls finds tool
syntax in the raw model output directly.
Validated with Qwen/Qwen2.5-0.5B-Instruct + hermes parser on CPU:
the backend correctly returns ToolCallDelta{name='get_weather',
arguments='{"location": "Paris, France"}'} in ChatDelta.
test_tool_calls.py is a standalone smoke test that spawns the gRPC
backend, sends a chat completion with tools, and asserts the response
contains a structured tool call.
Validated end-to-end on CPU with Qwen2.5-0.5B-Instruct (LoadModel, Predict,
TokenizeString, Free all working).
- requirements-cpu-after.txt: pin vllm to 0.14.1+cpu (pre-built wheel from
GitHub releases) for x86_64 and aarch64. vllm 0.14.1 is the newest CPU
wheel whose torch dependency resolves against published PyTorch builds
(torch==2.9.1+cpu). Later vllm CPU wheels currently require
torch==2.10.0+cpu which is only available on the PyTorch test channel
with incompatible torchvision.
- requirements-cpu.txt: bump torch to 2.9.1+cpu, add torchvision/torchaudio
so uv resolves them consistently from the PyTorch CPU index.
- install.sh: add --index-strategy=unsafe-best-match for CPU builds so uv
can mix the PyTorch index and PyPI for transitive deps (matches the
existing intel profile behaviour).
- backend.py LoadModel: vllm >= 0.14 removed AsyncLLMEngine.get_model_config
so the old code path errored out with AttributeError on model load.
Switch to the new get_tokenizer()/tokenizer accessor with a fallback
to building the tokenizer directly from request.Model.
- Split vllm install per acceleration: move generic `vllm` out of
requirements-after.txt into per-profile after files (cublas12, hipblas,
intel) and add CPU wheel URL for cpu-after.txt
- requirements-cpu.txt now pulls torch==2.7.0+cpu from PyTorch CPU index
- backend/index.yaml: register cpu-vllm / cpu-vllm-development variants
- New backend/python/common/vllm_utils.py: shared parse_options,
messages_to_dicts, setup_parsers helpers (used by both vllm backends)
- vllm-omni: replace hardcoded chat template with tokenizer.apply_chat_template,
wire native parsers via shared utils, emit ChatDelta with token counts,
add TokenizeString and Free RPCs, detect CPU and set VLLM_TARGET_DEVICE
- Add test_cpu_inference.py: standalone script to validate CPU build with
a small model (Qwen2.5-0.5B-Instruct)
- Use vLLM's ToolParserManager/ReasoningParserManager to extract structured
output (tool calls, reasoning content) instead of reimplementing parsing
- Convert proto Messages to dicts and pass tools to apply_chat_template
- Emit ChatDelta with content/reasoning_content/tool_calls in Reply
- Extract prompt_tokens, completion_tokens, and logprobs from output
- Replace boolean GuidedDecoding with proper GuidedDecodingParams from Grammar
- Add TokenizeString and Free RPC methods
- Fix missing `time` import used by load_video()
* feat: add distributed mode (experimental)
Signed-off-by: Ettore Di Giacinto <mudler@localai.io>
* fix data races, mutexes, transactions
Signed-off-by: Ettore Di Giacinto <mudler@localai.io>
* refactorings
Signed-off-by: Ettore Di Giacinto <mudler@localai.io>
* fixups
Signed-off-by: Ettore Di Giacinto <mudler@localai.io>
* fix events and tool stream in agent chat
Signed-off-by: Ettore Di Giacinto <mudler@localai.io>
* use ginkgo
Signed-off-by: Ettore Di Giacinto <mudler@localai.io>
* refactoring and consolidation
Signed-off-by: Ettore Di Giacinto <mudler@localai.io>
* refactoring and consolidation
Signed-off-by: Ettore Di Giacinto <mudler@localai.io>
* refactoring and consolidation
Signed-off-by: Ettore Di Giacinto <mudler@localai.io>
* refactoring and consolidation
Signed-off-by: Ettore Di Giacinto <mudler@localai.io>
* refactoring and consolidation
Signed-off-by: Ettore Di Giacinto <mudler@localai.io>
* refactoring and consolidation
Signed-off-by: Ettore Di Giacinto <mudler@localai.io>
* refactoring and consolidation
Signed-off-by: Ettore Di Giacinto <mudler@localai.io>
* refactoring and consolidation
Signed-off-by: Ettore Di Giacinto <mudler@localai.io>
* fix(cron): compute correctly time boundaries avoiding re-triggering
Signed-off-by: Ettore Di Giacinto <mudler@localai.io>
* enhancements, refactorings
Signed-off-by: Ettore Di Giacinto <mudler@localai.io>
* do not flood of healthy checks
Signed-off-by: Ettore Di Giacinto <mudler@localai.io>
* do not list obvious backends as text backends
Signed-off-by: Ettore Di Giacinto <mudler@localai.io>
* tests fixups
Signed-off-by: Ettore Di Giacinto <mudler@localai.io>
* refactoring and consolidation
Signed-off-by: Ettore Di Giacinto <mudler@localai.io>
* Drop redundant healthcheck
Signed-off-by: Ettore Di Giacinto <mudler@localai.io>
* enhancements, refactorings
Signed-off-by: Ettore Di Giacinto <mudler@localai.io>
---------
Signed-off-by: Ettore Di Giacinto <mudler@localai.io>
* chore(transformers): bump to >5.0
Signed-off-by: Ettore Di Giacinto <mudler@localai.io>
* chore: refactor to use generic model loading
Signed-off-by: Ettore Di Giacinto <mudler@localai.io>
---------
Signed-off-by: Ettore Di Giacinto <mudler@localai.io>
* feat: add fine-tuning endpoint
Signed-off-by: Ettore Di Giacinto <mudler@localai.io>
* feat(experimental): add fine-tuning endpoint and TRL support
This changeset defines new GRPC signatues for Fine tuning backends, and
add TRL backend as initial fine-tuning engine. This implementation also
supports exporting to GGUF and automatically importing it to LocalAI
after fine-tuning.
Signed-off-by: Ettore Di Giacinto <mudler@localai.io>
* commit TRL backend, stop by killing process
Signed-off-by: Ettore Di Giacinto <mudler@localai.io>
* move fine-tune to generic features
Signed-off-by: Ettore Di Giacinto <mudler@localai.io>
* add evals, reorder menu
Signed-off-by: Ettore Di Giacinto <mudler@localai.io>
* Fix tests
Signed-off-by: Ettore Di Giacinto <mudler@localai.io>
---------
Signed-off-by: Ettore Di Giacinto <mudler@localai.io>
* fix: add missing bufio.Flush in processImageFile
The processImageFile function writes decoded image data (from base64
or URL download) through a bufio.NewWriter but never calls Flush()
before closing the underlying file. Since bufio's default buffer is
4096 bytes, small images produce 0-byte files and large images are
truncated — causing PIL to fail with "cannot identify image file".
This breaks all image input paths: file, files, and ref_images
parameters in /v1/images/generations, making img2img, inpainting,
and reference image features non-functional.
Signed-off-by: Attila Györffy <attila+git@attilagyorffy.com>
* fix: merge options into kwargs in diffusers GenerateImage
The GenerateImage method builds a local `options` dict containing the
source image (PIL), negative_prompt, and num_inference_steps, but
never merges it into `kwargs` before calling self.pipe(**kwargs).
This causes img2img to fail with "Input is in incorrect format"
because the pipeline never receives the image parameter.
Signed-off-by: Attila Györffy <attila+git@attilagyorffy.com>
* test: add unit test for processImageFile base64 decoding
Verifies that a base64-encoded PNG survives the write path
(encode → decode → bufio.Write → Flush → file on disk) with
byte-for-byte fidelity. The test image is small enough to fit
entirely in bufio's 4096-byte buffer, which is the exact scenario
where the missing Flush() produced a 0-byte file.
Also tests that invalid base64 input is handled gracefully.
Signed-off-by: Attila Györffy <attila+git@attilagyorffy.com>
* test: verify GenerateImage merges options into pipeline kwargs
Mocks the diffusers pipeline and calls GenerateImage with a source
image and negative prompt. Asserts that the pipeline receives the
image, negative_prompt, and num_inference_steps via kwargs — the
exact parameters that were silently dropped before the fix.
Signed-off-by: Attila Györffy <attila+git@attilagyorffy.com>
* fix: move kwargs.update(options) earlier in GenerateImage
Move the options merge right after self.options merge (L742) so that
image, negative_prompt, and num_inference_steps are available to all
downstream code paths including img2vid and txt2vid.
Signed-off-by: Attila Györffy <attila+git@attilagyorffy.com>
* test: convert processImageFile tests to ginkgo
Replace standard testing with ginkgo/gomega to be consistent with
the rest of the test suites in the project.
Signed-off-by: Attila Györffy <attila+git@attilagyorffy.com>
---------
Signed-off-by: Attila Györffy <attila+git@attilagyorffy.com>
Co-authored-by: Ettore Di Giacinto <mudler@users.noreply.github.com>
* feat(mlx-distributed): add new MLX-distributed backend
Add new MLX distributed backend with support for both TCP and RDMA for
model sharding.
This implementation ties in the discovery implementation already in
place, and re-uses the same P2P mechanism for the TCP MLX-distributed
inferencing.
The Auto-parallel implementation is inspired by Exo's
ones (who have been added to acknowledgement for the great work!)
Signed-off-by: Ettore Di Giacinto <mudler@localai.io>
* expose a CLI to facilitate backend starting
Signed-off-by: Ettore Di Giacinto <mudler@localai.io>
* feat: make manual rank0 configurable via model configs
Signed-off-by: Ettore Di Giacinto <mudler@localai.io>
* Add missing features from mlx backend
Signed-off-by: Ettore Di Giacinto <mudler@localai.io>
* Apply suggestion from @mudler
Signed-off-by: Ettore Di Giacinto <mudler@users.noreply.github.com>
---------
Signed-off-by: Ettore Di Giacinto <mudler@localai.io>
Signed-off-by: Ettore Di Giacinto <mudler@users.noreply.github.com>
Don't pass instruct because it is added to kwargs
Fixes the error `qwen_tts.inference.qwen3_tts_model.Qwen3TTSModel.generate_voice_design() got multiple values for keyword argument 'instruct'`
Signed-off-by: Weathercold <weathercold.scr@proton.me>
* Add support for multiple voice clones in Qwen TTS
Signed-off-by: Andres Smith <andressmithdev@pm.me>
* Add voice prompt caching and generation logs to see generation time
---------
Signed-off-by: Andres Smith <andressmithdev@pm.me>
Co-authored-by: Ettore Di Giacinto <mudler@users.noreply.github.com>
- Add trailing newline to all requirements*.txt files in qwen-tts backend
- This ensures proper file formatting and prevents potential issues with
package installation tools that expect newline-terminated files
Adding debug logging to help investigate the pocket-tts custom voice
finding issue (Issue #8244). This is a first step to understand how
voices are being loaded and where the failure occurs.
Signed-off-by: localai-bot <localai-bot@users.noreply.github.com>
Co-authored-by: localai-bot <localai-bot@users.noreply.github.com>
* feat(backends): add faster-qwen3-tts
Signed-off-by: Ettore Di Giacinto <mudler@localai.io>
* fix: this backend is CUDA only
Signed-off-by: Ettore Di Giacinto <mudler@localai.io>
* fix: add requirements-install.txt with setuptools for build isolation
The faster-qwen3-tts backend requires setuptools to build packages
like sox that have setuptools as a build dependency. This ensures
the build completes successfully in CI.
Signed-off-by: LocalAI Bot <localai-bot@users.noreply.github.com>
---------
Signed-off-by: Ettore Di Giacinto <mudler@localai.io>
Signed-off-by: LocalAI Bot <localai-bot@users.noreply.github.com>
Co-authored-by: Ettore Di Giacinto <mudler@localai.io>
* add experimental support for sd_embed-style prompt embedding
Signed-off-by: Austen Dicken <cvpcsm@gmail.com>
* add doc equivalent to compel
Signed-off-by: Austen Dicken <cvpcsm@gmail.com>
* need to use flux1 embedding function for flux model
Signed-off-by: Austen Dicken <cvpcsm@gmail.com>
---------
Signed-off-by: Austen Dicken <cvpcsm@gmail.com>
* feat(musicgen): add ace-step and UI interface
Signed-off-by: Ettore Di Giacinto <mudler@localai.io>
* Correctly handle model dir
Signed-off-by: Ettore Di Giacinto <mudler@localai.io>
* Drop auto-download
Signed-off-by: Ettore Di Giacinto <mudler@localai.io>
* Fixups
Signed-off-by: Ettore Di Giacinto <mudler@localai.io>
* Add to models, fixup UIs icons
Signed-off-by: Ettore Di Giacinto <mudler@localai.io>
* fixups
Signed-off-by: Ettore Di Giacinto <mudler@localai.io>
* Update docs
Signed-off-by: Ettore Di Giacinto <mudler@localai.io>
* l4t13 is incompatbile
Signed-off-by: Ettore Di Giacinto <mudler@localai.io>
* avoid pinning version for cuda12
Signed-off-by: Ettore Di Giacinto <mudler@localai.io>
* Drop l4t12
Signed-off-by: Ettore Di Giacinto <mudler@localai.io>
---------
Signed-off-by: Ettore Di Giacinto <mudler@localai.io>
