niki/jrnl
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

update working solution for llamacpp for non thinking models

Browse Source
This commit is contained in:
Niki Vihtola
2026-03-13 15:22:37 +02:00
parent 3a13161ec6
commit 0ee34eb56d
15 changed files with 736 additions and 277 deletions
Download Patch File Download Diff File Expand all files Collapse all files

24
README.md
View File

@@ -37,6 +37,26 @@ jrnl config set anthropic api_key YOUR_API_KEY
jrnl config set-provider ollama
```
Configuration file: `~/.jrnl/config.json`
After installation, configure your LLM provider:
```bash
# For Anthropic Claude
jrnl config set anthropic api_key YOUR_API_KEY
# For Ollama (local)
jrnl config set-provider ollama
# For llama.cpp (local)
jrnl config set-provider llamacpp
# Set llama.cpp binary path (if not default)
jrnl config set llamacpp llamacpp_bin /path/to/llamacpp/bin
# Set model directory (optional)
jrnl config set llamacpp model_dir ~/.llamacpp/models
# Set model name (optional)
jrnl config set llamacpp model_name Qwen/Qwen3-8B-GGUF
```
Configuration file: `~/.jrnl/config.json`
## Usage
@@ -83,6 +103,10 @@ jrnl config
# Switch LLM provider
jrnl config set-provider ollama
# Switch LLM provider
jrnl config set-provider ollama
jrnl config set-provider llamacpp
# Set provider-specific settings
jrnl config set anthropic model claude-3-5-sonnet-20241022