wuyang/agent
2
2
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

migrate knowledge atlas to llama.cpp

Browse Source
This commit is contained in:
wuyang
2026-07-28 17:16:34 +08:00
parent 88fb212985
commit efbd82532f
9 changed files with 152 additions and 109 deletions
Download Patch File Download Diff File Expand all files Collapse all files
AGENTS.md
+8 -7
View File
@@ -78,23 +78,24 @@ The existing evaluation contract must be revised against that research before ru
- Keep generated cache and PID/log files under `web/cache/`; they are intentionally ignored.
- Never commit credentials. Git authentication is configured outside the repository.
## Ollama Contract
## Model API Contract
Default endpoint:
Default internal llama.cpp endpoints:
```text
http://192.168.1.10:11434
Luna: http://192.168.1.10:11435/v1
Terra/Sol: http://192.168.1.10:11436/v1
```
Current model tiers are configurable with:
```text
OLLAMA_MODEL_LIGHT=ChatGPT-5.6:Luna
OLLAMA_MODEL_FAST=ChatGPT-5.6:Terra
OLLAMA_MODEL_LARGE=ChatGPT-5.6:Sol
MODEL_LIGHT=ChatGPT-5.6:Luna
MODEL_FAST=ChatGPT-5.6:Terra
MODEL_LARGE=ChatGPT-5.6:Sol
```
The application serializes Ollama calls because all models share one V100 32G GPU. Application requests must not send `num_ctx` or `keep_alive`; the Ollama server owns context size and residency. Expensive `Sol` actions should remain explicit, not automatic.
Luna runs on the NAS AMD iGPU. Terra and Sol are aliases of one resident llama.cpp runner on the V100 32G. The application serializes model calls and must not override server context sizing. Expensive `Sol` actions should remain explicit, not automatic.
## Common Commands