llm, stop purging partial pulls.
This commit is contained in:
@@ -194,6 +194,8 @@
|
|||||||
export OLLAMA_NUM_PARALLEL={{ llm_parallel }} # default is 1 and applies per-model (unconfirmed)
|
export OLLAMA_NUM_PARALLEL={{ llm_parallel }} # default is 1 and applies per-model (unconfirmed)
|
||||||
export OLLAMA_CONTEXT_LENGTH={{ llm_context }} # default is 4096, https://docs.ollama.com/faq#how-can-i-specify-the-context-window-size
|
export OLLAMA_CONTEXT_LENGTH={{ llm_context }} # default is 4096, https://docs.ollama.com/faq#how-can-i-specify-the-context-window-size
|
||||||
export OLLAMA_MAX_QUEUE=999 # default is 512 (unconfirmed)
|
export OLLAMA_MAX_QUEUE=999 # default is 512 (unconfirmed)
|
||||||
|
# Prevent Data Loss #
|
||||||
|
export OLLAMA_NOPRUNE=1 # Default is 0, which causes unfinished pulls to be deleeted after 1 hour. (unconfirmed)
|
||||||
exec chpst -u {{ user }} ollama serve
|
exec chpst -u {{ user }} ollama serve
|
||||||
mode: "0755"
|
mode: "0755"
|
||||||
|
|
||||||
|
|||||||
Reference in New Issue
Block a user