diff --git a/omp-docs/agentic-web-research/index.md b/omp-docs/agentic-web-research/index.md index 4cdee48..7120308 100644 --- a/omp-docs/agentic-web-research/index.md +++ b/omp-docs/agentic-web-research/index.md @@ -2,7 +2,12 @@ Agentic web research means that an AI drives the search loop. The agent formulates queries, reads results, decides what to read next, and writes a synthesis. It works best when search, extraction, and reasoning each run in the right place. -> **Your setup:** Pi Web Access calls SearXNG at `192.168.0.143:8081` (dankclaw) and serves models through llama.cpp at `192.168.0.235:7770`, `192.168.0.143:7770`, and `localhost:8080`. The pages below assume that topology. +> **Fleet topology (2026-10):** SearXNG lives at `192.168.0.143:8081` (UI) / +> `:8082` (JSON) on dankclaw; model endpoints are `frank-archon` +> (dankserver :7770, radiance), `frank-dragoon` (dankclaw :7770, +> halogen-flash-server), `frank-zealot` (dankclaw :7771, llama.cpp PrismML). +> The pi-era `localhost:8080` llama.cpp endpoint is retired. See +> [System hosts](/system/index.md). This section connects four decisions: diff --git a/omp-docs/agentic-web-research/open-models.md b/omp-docs/agentic-web-research/open-models.md index 13b81ce..e3bffd0 100644 --- a/omp-docs/agentic-web-research/open-models.md +++ b/omp-docs/agentic-web-research/open-models.md @@ -32,9 +32,15 @@ Qwen3 is the current state of the art for open function calling and tool use. It The 8B and 14B models run on one consumer GPU. The 30B-A3B MoE runs on a single A100-class card or equivalent. -## Best Vision Model: Qwen2.5-VL-7B +## Best Vision Model: Qwen3-VL (Apache 2.0) -Many web tasks need a model that can read a page image, not just text. Qwen2.5-VL-7B is the best open vision model for this. It reads text, charts, icons, and layout in screenshots. It supports computer-use and phone-use agents. Apache 2.0 license. The 7B model fits on a single GPU with 8 GB or more. +Many web tasks need a model that can read a page image, not just text. +**Qwen3-VL** supersedes Qwen2.5-VL as the strongest open-weights vision family +(4B/8B dense through 30B-A3B/235B-A22B MoE): text, chart, icon, and layout +reading in screenshots, plus computer-use agent support. Small dense sizes +ship Apache 2.0; a quantized 4B/8B runs on a single consumer GPU. On this +fleet, vision duty is carried by the mmproj vision towers on the endpoints +(dragoon, zealot) — see [System hosts](/system/index.md). Llama 4 Scout is the larger multimodal option. Verify the claims before you trust them: @@ -47,7 +53,7 @@ Llama 4 Scout is the larger multimodal option. Verify the claims before you trus ```mermaid graph TD A[Consumer GPU 16-24GB] --> B[Qwen3-8B / 14B] - A --> C[Qwen2.5-VL-7B] + A --> C[Qwen3-VL-4B / 8B] D[Single server GPU 80GB] --> E[Qwen3-30B-A3B] D --> F[Llama 4 Scout INT4] G[Multi-GPU cluster] --> H[Qwen3-235B-A22B] @@ -71,7 +77,7 @@ llama-server -m qwen3-8b-q4_k_m.gguf -c 131072 --port 7770 | Need | Pick | |------|------| | Best tool calling, one consumer GPU | Qwen3-8B or 14B | -| Visually read a web page | Qwen2.5-VL-7B | +| Visually read a web page | Qwen3-VL-4B / 8B | | Server-scale reliable agent | Qwen3-235B-A22B | | Cheapest acceptable agent | Qwen3-8B on a mid GPU | @@ -79,7 +85,8 @@ llama-server -m qwen3-8b-q4_k_m.gguf -c 131072 --port 7770 - Qwen3 technical report: - Qwen3 repository: -- Qwen2.5-VL-7B model card: +- Qwen3-VL repository: (tech report: ) +- Qwen2.5-VL-7B model card (predecessor): - Meta AI Llama 4 announcement: - Ollama library: - Browser Use supported models: diff --git a/omp-docs/agentic-web-research/self-hosted-web-agents.md b/omp-docs/agentic-web-research/self-hosted-web-agents.md index 8803c88..7bcb949 100644 --- a/omp-docs/agentic-web-research/self-hosted-web-agents.md +++ b/omp-docs/agentic-web-research/self-hosted-web-agents.md @@ -40,7 +40,7 @@ agent = Agent( task="Search for the latest AI news", llm=ChatOpenAI( base_url="http://192.168.0.235:7770/v1", - model="frank-default-small", + model="frank-archon", api_key="EMPTY", ), ) diff --git a/omp-docs/research/chezmoi-evaluation.md b/omp-docs/research/chezmoi-evaluation.md index d3aacab..d9233df 100644 --- a/omp-docs/research/chezmoi-evaluation.md +++ b/omp-docs/research/chezmoi-evaluation.md @@ -170,7 +170,7 @@ This allows plugging in any custom secrets backend. — [Keychain and Windows Credentials Manager][keyring] -**This is relevant for the workstation** — CachyOS uses GNOME Keyring (KDE uses kwallet, but the system has GNOME components). On the Strix Halo (Arch Linux) and RTX 3090 (FreeToken) boxes, GNOME Keyring or kwallet would be available depending on desktop environment. +**This is relevant for the workstation** — CachyOS uses GNOME Keyring (KDE uses kwallet, but the system has GNOME components). On the Ubuntu 26.04 server boxes (dankclaw, dankserver), GNOME Keyring or kwallet would be available depending on desktop environment. --- @@ -336,9 +336,9 @@ Use symlinks for configs that external programs modify and must write back to (V ### Setup -- **Machine fleet:** the workstation (Intel Ultra 7, CachyOS, niri), dankclaw (Strix Halo, llama.cpp frank-default-large on port 7770), dankserver (RTX 3090, FreeToken, frank-default-small) +- **Machine fleet:** the dev workstation (AMD Ryzen 5 PRO 5650U laptop, CachyOS, niri) plus client boxes; dankclaw (Strix Halo + RTX 3090, Ubuntu 26.04, fleet services); dankserver (2x Radeon AI PRO R9700, Ubuntu 26.04) — fleet detail: [/system/index.md](/system/index.md) - **Central config repo:** `~/.omp` — already a git repo with agent config, model manifests, and the KB -- **OS mix:** All Linux, but different distros (CachyOS, Arch, custom FreeToken) +- **OS mix:** All Linux, but different distros (CachyOS on clients, Ubuntu 26.04 LTS on both servers) ### Verdict: **Recommended with moderate reservations**