File editing problems

File editing problems

Sometimes the LLM will reply with some code changes that don't get applied to your local files. In these cases, aider might say something like "Failed to apply edit to filename" or other error messages.

This usually happens because the LLM is disobeying the system prompts and trying to make edits in a format that aider doesn't expect. Aider makes every effort to get the LLM to conform, and works hard to deal with LLM edits that are "almost" correctly formatted.

But sometimes the LLM just won't cooperate. In these cases, here are some things you might try.

Don't add too many files

Many LLMs now have very large context windows, but filling them with irrelevant code or conversation can confuse the model. Above about 25k tokens of context, most models start to become distracted and become less likely to conform to their system prompt.

• Don't add too many files to the chat, just add the files you think need to be edited. Aider also sends the LLM a map of your entire git repo, so other relevant code will be included automatically.
• Use /drop to remove files from the chat session which aren't needed for the task at hand. This will reduce distractions and may help the LLM produce properly formatted edits.
• Use /clear to remove the conversation history, again to help the LLM focus.
• Use /tokens to see how many tokens you are using for each message.

Use a more capable model

If possible try using GPT-4o, o3-mini, Claude 3.7 Sonnet, DeepSeek V3 or DeepSeek R1. They are the strong and capable models.

Weaker models are more prone to disobeying the system prompt instructions. Most local models are just barely capable of working with aider, so editing errors are probably unavoidable.

Local models: context window and quantization

Be especially careful about the Ollama context window when working with local models. It defaults to be very small and silently discards data if you exceed it.

Local models which have been quantized are more likely to have editing problems because they are not capable enough to follow aider's system prompts.

Try the whole edit format

Run aider with --edit-format whole if were using a different edit format. You can see which edit format it is using in the announce lines:

Aider v0.50.2-dev
Models: claude-3-5-sonnet-20240620 with ♾️ diff edit format

Try architect mode

Run aider with --architect or /chat-mode architect to enable architect mode. This mode first proposes changes, then uses a separate model to handle the file edits. This two-step process often produces more reliable edits, especially with models that have trouble following edit format instructions.

More help

{% include help.md %}