feat: implement Unsloth fine-tuning backend phases 2-7

- Add Python unsloth backend with QLoRA/LoRA/full fine-tuning support
- Implement Go service layer for job management
- Add HTTP API endpoints for fine-tuning jobs
- Integrate GGUF export functionality
- Update CI/build configuration

Signed-off-by: localai-bot <localai-bot@users.noreply.github.com>

Author: localai-bot

.github/workflows/backend.yml

@@ -79,6 +79,19 @@ jobs:
             dockerfile: "./backend/Dockerfile.python"
             context: "./"
             ubuntu-version: '2404'
+          - build-type: ''
+            cuda-major-version: ""
+            cuda-minor-version: ""
+            platforms: 'linux/amd64'
+            tag-latest: 'auto'
+            tag-suffix: '-cpu-unsloth'
+            runs-on: 'ubuntu-latest'
+            base-image: "ubuntu:24.04"
+            skip-drivers: 'true'
+            backend: "unsloth"
+            dockerfile: "./backend/Dockerfile.python"
+            context: "./"
+            ubuntu-version: '2404'
           - build-type: ''
             cuda-major-version: ""
             cuda-minor-version: ""
@@ -444,6 +457,19 @@ jobs:
             dockerfile: "./backend/Dockerfile.python"
             context: "./"
             ubuntu-version: '2404'
+          - build-type: 'cublas'
+            cuda-major-version: "12"
+            cuda-minor-version: "8"
+            platforms: 'linux/amd64'
+            tag-latest: 'auto'
+            tag-suffix: '-gpu-nvidia-cuda-12-unsloth'
+            runs-on: 'ubuntu-latest'
+            base-image: "ubuntu:24.04"
+            skip-drivers: 'false'
+            backend: "unsloth"
+            dockerfile: "./backend/Dockerfile.python"
+            context: "./"
+            ubuntu-version: '2404'
           - build-type: 'cublas'
             cuda-major-version: "12"
             cuda-minor-version: "8"
@@ -965,6 +991,19 @@ jobs:
             dockerfile: "./backend/Dockerfile.python"
             context: "./"
             ubuntu-version: '2404'
+          - build-type: 'cublas'
+            cuda-major-version: "13"
+            cuda-minor-version: "0"
+            platforms: 'linux/amd64'
+            tag-latest: 'auto'
+            tag-suffix: '-gpu-nvidia-cuda-13-unsloth'
+            runs-on: 'ubuntu-latest'
+            base-image: "ubuntu:24.04"
+            skip-drivers: 'false'
+            backend: "unsloth"
+            dockerfile: "./backend/Dockerfile.python"
+            context: "./"
+            ubuntu-version: '2404'
           - build-type: 'cublas'
             cuda-major-version: "13"
             cuda-minor-version: "0"

Makefile

@@ -1,5 +1,5 @@
 # Disable parallel execution for backend builds
-.NOTPARALLEL: backends/diffusers backends/llama-cpp backends/outetts backends/piper backends/stablediffusion-ggml backends/whisper backends/faster-whisper backends/silero-vad backends/local-store backends/huggingface backends/rfdetr backends/kitten-tts backends/kokoro backends/chatterbox backends/llama-cpp-darwin backends/neutts build-darwin-python-backend build-darwin-go-backend backends/mlx backends/diffuser-darwin backends/mlx-vlm backends/mlx-audio backends/mlx-distributed backends/stablediffusion-ggml-darwin backends/vllm backends/vllm-omni backends/moonshine backends/pocket-tts backends/qwen-tts backends/faster-qwen3-tts backends/qwen-asr backends/nemo backends/voxcpm backends/whisperx backends/ace-step backends/acestep-cpp backends/fish-speech backends/voxtral backends/opus
+.NOTPARALLEL: backends/diffusers backends/llama-cpp backends/outetts backends/piper backends/stablediffusion-ggml backends/whisper backends/faster-whisper backends/silero-vad backends/local-store backends/huggingface backends/rfdetr backends/kitten-tts backends/kokoro backends/chatterbox backends/llama-cpp-darwin backends/neutts build-darwin-python-backend build-darwin-go-backend backends/mlx backends/diffuser-darwin backends/mlx-vlm backends/mlx-audio backends/mlx-distributed backends/stablediffusion-ggml-darwin backends/vllm backends/vllm-omni backends/moonshine backends/pocket-tts backends/qwen-tts backends/faster-qwen3-tts backends/qwen-asr backends/nemo backends/voxcpm backends/whisperx backends/ace-step backends/acestep-cpp backends/fish-speech backends/voxtral backends/opus backends/unsloth
 
 GOCMD=go
 GOTEST=$(GOCMD) test
@@ -421,6 +421,7 @@ prepare-test-extra: protogen-python
 	$(MAKE) -C backend/python/voxcpm
 	$(MAKE) -C backend/python/whisperx
 	$(MAKE) -C backend/python/ace-step
+	$(MAKE) -C backend/python/unsloth
 
 test-extra: prepare-test-extra
 	$(MAKE) -C backend/python/transformers test
@@ -440,6 +441,7 @@ test-extra: prepare-test-extra
 	$(MAKE) -C backend/python/voxcpm test
 	$(MAKE) -C backend/python/whisperx test
 	$(MAKE) -C backend/python/ace-step test
+	$(MAKE) -C backend/python/unsloth test
 
 DOCKER_IMAGE?=local-ai
 IMAGE_TYPE?=core
@@ -571,6 +573,7 @@ BACKEND_NEMO = nemo|python|.|false|true
 BACKEND_VOXCPM = voxcpm|python|.|false|true
 BACKEND_WHISPERX = whisperx|python|.|false|true
 BACKEND_ACE_STEP = ace-step|python|.|false|true
+BACKEND_UNSLOTH = unsloth|python|.|false|true
 BACKEND_MLX_DISTRIBUTED = mlx-distributed|python|./|false|true
 
 # Helper function to build docker image for a backend
@@ -627,14 +630,15 @@ $(eval $(call generate-docker-build-target,$(BACKEND_NEMO)))
 $(eval $(call generate-docker-build-target,$(BACKEND_VOXCPM)))
 $(eval $(call generate-docker-build-target,$(BACKEND_WHISPERX)))
 $(eval $(call generate-docker-build-target,$(BACKEND_ACE_STEP)))
+$(eval $(call generate-docker-build-target,$(BACKEND_UNSLOTH)))
 $(eval $(call generate-docker-build-target,$(BACKEND_ACESTEP_CPP)))
 $(eval $(call generate-docker-build-target,$(BACKEND_MLX_DISTRIBUTED)))
 
 # Pattern rule for docker-save targets
 docker-save-%: backend-images
 	docker save local-ai-backend:$* -o backend-images/$*.tar
 
-docker-build-backends: docker-build-llama-cpp docker-build-rerankers docker-build-vllm docker-build-vllm-omni docker-build-transformers docker-build-outetts docker-build-diffusers docker-build-kokoro docker-build-faster-whisper docker-build-coqui docker-build-chatterbox docker-build-vibevoice docker-build-moonshine docker-build-pocket-tts docker-build-qwen-tts docker-build-fish-speech docker-build-faster-qwen3-tts docker-build-qwen-asr docker-build-nemo docker-build-voxcpm docker-build-whisperx docker-build-ace-step docker-build-acestep-cpp docker-build-voxtral docker-build-mlx-distributed
+docker-build-backends: docker-build-llama-cpp docker-build-rerankers docker-build-vllm docker-build-vllm-omni docker-build-transformers docker-build-outetts docker-build-diffusers docker-build-kokoro docker-build-faster-whisper docker-build-coqui docker-build-chatterbox docker-build-vibevoice docker-build-moonshine docker-build-pocket-tts docker-build-qwen-tts docker-build-fish-speech docker-build-faster-qwen3-tts docker-build-qwen-asr docker-build-nemo docker-build-voxcpm docker-build-whisperx docker-build-ace-step docker-build-acestep-cpp docker-build-voxtral docker-build-mlx-distributed docker-build-unsloth
 
 ########################################################
 ### Mock Backend for E2E Tests

backend/index.yaml

@@ -530,6 +530,25 @@
     nvidia-cuda-12: "cuda12-chatterbox"
     nvidia-l4t-cuda-12: "nvidia-l4t-arm64-chatterbox"
     nvidia-l4t-cuda-13: "cuda13-nvidia-l4t-arm64-chatterbox"
+- &unsloth
+  urls:
+    - https://github.com/unslothai/unsloth
+  description: |
+    Fine-tuning backend powered by Unsloth. Supports QLoRA, LoRA, and full fine-tuning
+    of LLMs with 2x faster training and 70% less memory. Includes GGUF export support.
+  tags:
+    - fine-tuning
+    - training
+    - lora
+    - qlora
+  license: apache-2.0
+  name: "unsloth"
+  alias: "unsloth"
+  capabilities:
+    nvidia: "cuda12-unsloth"
+    default: "cpu-unsloth"
+    nvidia-cuda-13: "cuda13-unsloth"
+    nvidia-cuda-12: "cuda12-unsloth"
 - &vibevoice
   urls:
     - https://github.com/microsoft/VibeVoice
@@ -3029,3 +3048,41 @@
   uri: "quay.io/go-skynet/local-ai-backends:master-metal-darwin-arm64-voxtral"
   mirrors:
     - localai/localai-backends:master-metal-darwin-arm64-voxtral
+## unsloth
+- !!merge <<: *unsloth
+  name: "unsloth-development"
+  capabilities:
+    nvidia: "cuda12-unsloth-development"
+    default: "cpu-unsloth-development"
+    nvidia-cuda-13: "cuda13-unsloth-development"
+    nvidia-cuda-12: "cuda12-unsloth-development"
+- !!merge <<: *unsloth
+  name: "cpu-unsloth"
+  uri: "quay.io/go-skynet/local-ai-backends:latest-cpu-unsloth"
+  mirrors:
+    - localai/localai-backends:latest-cpu-unsloth
+- !!merge <<: *unsloth
+  name: "cpu-unsloth-development"
+  uri: "quay.io/go-skynet/local-ai-backends:master-cpu-unsloth"
+  mirrors:
+    - localai/localai-backends:master-cpu-unsloth
+- !!merge <<: *unsloth
+  name: "cuda12-unsloth"
+  uri: "quay.io/go-skynet/local-ai-backends:latest-gpu-nvidia-cuda-12-unsloth"
+  mirrors:
+    - localai/localai-backends:latest-gpu-nvidia-cuda-12-unsloth
+- !!merge <<: *unsloth
+  name: "cuda12-unsloth-development"
+  uri: "quay.io/go-skynet/local-ai-backends:master-gpu-nvidia-cuda-12-unsloth"
+  mirrors:
+    - localai/localai-backends:master-gpu-nvidia-cuda-12-unsloth
+- !!merge <<: *unsloth
+  name: "cuda13-unsloth"
+  uri: "quay.io/go-skynet/local-ai-backends:latest-gpu-nvidia-cuda-13-unsloth"
+  mirrors:
+    - localai/localai-backends:latest-gpu-nvidia-cuda-13-unsloth
+- !!merge <<: *unsloth
+  name: "cuda13-unsloth-development"
+  uri: "quay.io/go-skynet/local-ai-backends:master-gpu-nvidia-cuda-13-unsloth"
+  mirrors:
+    - localai/localai-backends:master-gpu-nvidia-cuda-13-unsloth

backend/python/unsloth/Makefile

@@ -0,0 +1,23 @@
+.PHONY: unsloth
+unsloth:
+	bash install.sh
+
+.PHONY: run
+run: unsloth
+	@echo "Running unsloth..."
+	bash run.sh
+	@echo "unsloth run."
+
+.PHONY: test
+test: unsloth
+	@echo "Testing unsloth..."
+	bash test.sh
+	@echo "unsloth tested."
+
+.PHONY: protogen-clean
+protogen-clean:
+	$(RM) backend_pb2_grpc.py backend_pb2.py
+
+.PHONY: clean
+clean: protogen-clean
+	rm -rf venv __pycache__