github/langchain
1
0
Fork 0
mirror of https://github.com/hwchase17/langchain.git synced 2025-09-17 23:41:46 +00:00
Code Issues Packages Projects Releases Wiki Activity

community[patch]: Added support for Ollama's num_predict option in ChatOllama (#16633)

Browse Source 
Just a simple default addition to the options payload for a ollama
generate call to support a max_new_tokens parameter.

Should fix issue: https://github.com/langchain-ai/langchain/issues/14715
This commit is contained in:
Micah Parker
2024-01-26 17:00:19 -06:00
committed by GitHub
parent 6a75ef74ca
commit 6543e585a5
2 changed files with 7 additions and 0 deletions
Download Patch File Download Diff File Expand all files Collapse all files

2
libs/community/tests/unit_tests/llms/test_ollama.py
View File

@@ -88,6 +88,7 @@ def test_handle_kwargs_top_level_parameters(monkeypatch: MonkeyPatch) -> None:
"num_ctx": None,
"num_gpu": None,
"num_thread": None,
"num_predict": None,
"repeat_last_n": None,
"repeat_penalty": None,
"stop": [],
@@ -133,6 +134,7 @@ def test_handle_kwargs_with_unknown_param(monkeypatch: MonkeyPatch) -> None:
"num_ctx": None,
"num_gpu": None,
"num_thread": None,
"num_predict": None,
"repeat_last_n": None,
"repeat_penalty": None,
"stop": [],