forked from Mirror/ollama4j
Refactor tool handling and add model unload support
Replaces the 'clientHandlesTools' flag with 'useTools' for tool execution control, defaulting to true. Adds support for model unloading via keep_alive parameter. Updates chat request and builder to use the new flag. Improves integration tests and documentation to reflect these changes. Fixes constructor order in OllamaGenerateStreamObserver and adds ignoreUnknown to ModelsProcessResponse.
This commit is contained in:
File diff suppressed because it is too large
Load Diff
@@ -1,4 +1,5 @@
|
||||
USE_EXTERNAL_OLLAMA_HOST=true
|
||||
OLLAMA_HOST=http://192.168.29.229:11434/
|
||||
REQUEST_TIMEOUT_SECONDS=120
|
||||
NUMBER_RETRIES_FOR_MODEL_PULL=3
|
||||
NUMBER_RETRIES_FOR_MODEL_PULL=3
|
||||
MODEL_KEEP_ALIVE_TIME=0
|
||||
Reference in New Issue
Block a user