From 342271841573cb4d554f8542c3088f439640a620 Mon Sep 17 00:00:00 2001 From: Paul Gauthier Date: Tue, 28 Jan 2025 11:22:09 -0800 Subject: [PATCH] copy --- aider/website/_posts/2025-01-28-deepseek-down.md | 9 +++++---- aider/website/docs/llms/ollama.md | 2 ++ 2 files changed, 7 insertions(+), 4 deletions(-) diff --git a/aider/website/_posts/2025-01-28-deepseek-down.md b/aider/website/_posts/2025-01-28-deepseek-down.md index 20340595f..d4694ffe0 100644 --- a/aider/website/_posts/2025-01-28-deepseek-down.md +++ b/aider/website/_posts/2025-01-28-deepseek-down.md @@ -119,8 +119,11 @@ setx OLLAMA_API_BASE http://127.0.0.1:11434 # Windows, restart shell after set aider --model ollama/deepseek-v3 ``` -It's important to provide model settings, especially the `num_ctx` parameter. +It's important to provide model settings, especially the `num_ctx` parameter to +set the context window. Ollama uses a 2k context window by default, which is very small for working with aider. +Larger context windows will allow you to work with larger amounts of code, +but will use memory and increase latency. Unlike most other LLM servers, Ollama does not throw an error if you submit a request that exceeds the context window. Instead, it just silently truncates the request by discarding the “oldest” messages in the chat to make it fit within the context window. @@ -137,15 +140,13 @@ Create a `.aider.model.settings.yml` file in your home directory or git project lazy: false reminder: sys examples_as_sys_msg: true - extra_params: - max_tokens: 8192 cache_control: false caches_by_default: true use_system_prompt: true use_temperature: true streaming: true extra_params: - num_ctx: 8192 + num_ctx: 8192 # How large a context window? ``` ## Other providers diff --git a/aider/website/docs/llms/ollama.md b/aider/website/docs/llms/ollama.md index ba034a6cf..5207656f5 100644 --- a/aider/website/docs/llms/ollama.md +++ b/aider/website/docs/llms/ollama.md @@ -56,6 +56,8 @@ you added to the chat. That's because ollama is silently discarding them because they exceed the context window. Aider sets Ollama's context window to 8k by default. +Larger context windows will allow you to work with larger amounts of code, +but will use memory and increase latency. If you would like a larger context window you can use a