Skip to content

docs: update README for 26B MoE support and 40 tools - #6

Open
ToniBirat7 wants to merge 1 commit into
masterfrom
docs/readme-model-and-tool-count
Open

ToniBirat7 wants to merge 1 commit into
masterfrom
docs/readme-model-and-tool-count

Conversation

@ToniBirat7

Copy link
Copy Markdown
Owner

The README was behind the code in two places.

  • 26B MoE: the README called the Gemma 4 26B A4B MoE a "next milestone". install.sh already offers it as option 2, and scripts/local_up.py and scripts/detect_hardware.py offload its expert layers to system RAM with --n-cpu-moe when the model does not fit in VRAM. The intro, the callout, the model table and the requirements line now say so. E2B stays documented as the default.
  • Tool count: services/mcp/main.py registers 40 @mcp.tool functions; the README said 32 in four places.

Not changed:

  • The .wiki/ pages still say 32 tools. .wiki/mcp-tools.md lists tools individually, so it needs the eight newer tools written up, not only a new number.
  • No throughput figure is added for the 26B model, because the repo does not record one for a 6 GB GPU.
  • The Next.js web UI lines stay: apps/frontend is still in the tree.

The README still called the 26B A4B MoE a next milestone and counted 32
MCP tools. The installer has offered the 26B model as option 2 since the
launcher gained expert offload (--n-cpu-moe), and services/mcp/main.py
registers 40 tools.
Copilot AI balanced review requested due to automatic review settings October 1, 2026 13:02

Copilot AI left a comment

Copy link
Copy Markdown

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Copilot was unable to review this pull request because the user who requested the review has reached their quota limit.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants