Skip to content

Experimental Ollama support with new tutorial and backend integration #118

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Draft
barvhaim wants to merge 2 commits into
base: main
Choose a base branch
from
Draft

Experimental Ollama support with new tutorial and backend integration #118

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
6a6b29e
Add Ollama support with new tutorial and backend integration
barvhaim Aug 16, 2026
1f7178b
Add experimental RAG flow tutorial for Ollama integration and update …
barvhaim Aug 16, 2026
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
10 changes: 10 additions & 0 deletions .gitignore
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -4,6 +4,9 @@
# Claude Code
.claude/

# IDEs
.idea/

# Python cache files
*.pyc
__pycache__/
Expand Down Expand Up @@ -35,6 +38,13 @@ tmp/
# Model/adapter download cache (regression tests)
.cache/

# Ollama tutorial: cached RAG adapter io.yaml configs (fetched at runtime)
.rag_io_cache/

# RAG tutorials: govt corpus + ChromaDB index built at runtime
govt.jsonl
govt_chroma/

# Build outputs (multi-GB model checkpoints)
modular-granite/
my-modular-model/
Expand Down
16 changes: 16 additions & 0 deletions pyproject.toml
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -36,6 +36,22 @@ tutorials = [
"sentence-transformers>=3.0.0",
"datasets>=2.0.0",
]
# Ollama-backed tutorials (macOS / no-CUDA). Deliberately excludes vLLM so the
# environment installs on Apple Silicon: adapters are driven through a local
# `ollama serve` via mellea instead of a vLLM server. Sync with:
# uv sync --extra ollama --no-default-groups
# (--no-default-groups skips the default vllm19 group, which pulls CUDA.)
ollama = [
"granite-switch[compose]",
"requests>=2.31.0",
"rich>=13.0.0",
"mellea==0.6.0",
"ipython>=8.10.0",
"python-dotenv>=1.0.0",
"chromadb>=0.4.0",
"httpx>=0.24.0",
"sentence-transformers>=3.0.0",
]

[project.entry-points."vllm.general_plugins"]
register_granite_switch = "granite_switch.vllm:register"
Expand Down
Loading
Loading