diff --git a/website/_posts/2024-05-13-models-over-time.md b/website/_posts/2024-05-13-models-over-time.md index a30a538c2..3a6b5fa53 100644 --- a/website/_posts/2024-05-13-models-over-time.md +++ b/website/_posts/2024-05-13-models-over-time.md @@ -15,7 +15,7 @@ nav_exclude: true I recently wanted to draw a graph showing how LLM code editing skill has been changing over time as new models have been released by OpenAI, Anthropic and others. I have all the -[data in a yaml file](https://github.com/paul-gauthier/aider/blob/main/_data/edit_leaderboard.yml) that is used to render +[data in a yaml file](https://github.com/paul-gauthier/aider/blob/main/website/_data/edit_leaderboard.yml) that is used to render [aider's LLM leaderboards](https://aider.chat/docs/leaderboards/). Below is the aider chat transcript, which shows: diff --git a/website/docs/benchmarks-0125.md b/website/docs/benchmarks-0125.md index 0de49c947..94a42e64f 100644 --- a/website/docs/benchmarks-0125.md +++ b/website/docs/benchmarks-0125.md @@ -28,7 +28,7 @@ the new `gpt-4-0125-preview` model seems lazier than the November `gpt-4-1106-preview` model: - It gets worse benchmark scores when using the [unified diffs](https://aider.chat/docs/unified-diffs.html) code editing format. -- Using aider's older [SEARCH/REPLACE block](https://github.com/paul-gauthier/aider/blob/9033be74bf74ae70459013e54b2ae6a97c47c2e6/aider/coders/editblock_prompts.py#L75-L80) editing format, the new January model outperforms the older November model. But it still performs worse than both models using unified diffs. +- Using aider's older SEARCH/REPLACE block editing format, the new January model outperforms the older November model. But it still performs worse than both models using unified diffs. ## Related reports diff --git a/website/docs/ctags.md b/website/docs/ctags.md index b53c921cf..b57f8de33 100644 --- a/website/docs/ctags.md +++ b/website/docs/ctags.md @@ -61,7 +61,7 @@ To code with GPT-4 using the techniques discussed here: - Install [aider](https://aider.chat/docs/install.html). - - Install [universal ctags](https://aider.chat/docs/install.html#install-universal-ctags-optional). + - Install universal ctags. - Run `aider` inside your repo, and it should say "Repo-map: universal-ctags using 1024 tokens". ## The problem: code context @@ -246,5 +246,5 @@ specific language(s) of interest. To use this experimental repo map feature: - Install [aider](https://aider.chat/docs/install.html). - - Install [universal ctags](https://aider.chat/docs/install.html#install-universal-ctags-optional). + - Install ctags. - Run `aider` inside your repo, and it should say "Repo-map: universal-ctags using 1024 tokens". diff --git a/website/docs/faq.md b/website/docs/faq.md index c7c89543c..1744e969d 100644 --- a/website/docs/faq.md +++ b/website/docs/faq.md @@ -128,4 +128,4 @@ When experimenting with coder backends, it helps to run aider with `--verbose -- all the raw information being sent to/from the LLM in the conversation. You can also refer to the -[instructions for installing a development version of aider](https://aider.chat/docs/install.html#install-development-versions-of-aider-optional). +[instructions for installing a development version of aider](https://aider.chat/docs/install/optional.html#install-the-development-version-of-aider). diff --git a/website/docs/leaderboards/index.md b/website/docs/leaderboards/index.md index 28892dabb..097c542f1 100644 --- a/website/docs/leaderboards/index.md +++ b/website/docs/leaderboards/index.md @@ -221,4 +221,4 @@ See the [benchmark README](https://github.com/paul-gauthier/aider/blob/main/benchmark/README.md) for information on running aider's code editing benchmarks. Submit results by opening a PR with edits to the -[benchmark results data files](https://github.com/paul-gauthier/aider/blob/main/_data/). +[benchmark results data files](https://github.com/paul-gauthier/aider/blob/main/website/_data/). diff --git a/website/docs/unified-diffs.md b/website/docs/unified-diffs.md index 5e0ecc3a3..9f745969d 100644 --- a/website/docs/unified-diffs.md +++ b/website/docs/unified-diffs.md @@ -307,7 +307,7 @@ radically increases the number of hunks which fail to apply. ## Refactoring benchmark Aider has long used a -[benchmark suite based on 133 Exercism python exercises](). +[benchmark suite based on 133 Exercism python exercises](https://aider.chat/2023/07/02/benchmarks.html). But these are mostly small coding problems, usually requiring only a few dozen lines of code. GPT-4 Turbo is typically only lazy on 2-3 of these exercises: diff --git a/website/docs/voice.md b/website/docs/voice.md index 64e85fca3..ee63dfb3c 100644 --- a/website/docs/voice.md +++ b/website/docs/voice.md @@ -19,7 +19,7 @@ Your voice coding instructions will be transcribed and sent to GPT, as if you had typed them into the aider chat session. -See the [installation instructions](https://aider.chat/docs/install.html#install-portaudio-optional) for +See the [installation instructions](https://aider.chat/docs/install/optional.html#enable-voice-coding) for information on how to enable the `/voice` command.