This is an automated email from the ASF dual-hosted git repository. github-merge-queue[bot] pushed a commit to branch gh-readonly-queue/main/pr-5628-ae3128d999c48d0f014be2eea2eb4fc0f4ee9e66 in repository https://gitbox.apache.org/repos/asf/texera.git
commit 1abb3002647da47446790eb87cab2a2023ff092a Author: Matthew B. <[email protected]> AuthorDate: Thu Jun 11 11:33:28 2026 -0700 docs: backport content updates from incubator-texera-site (#5628) ### What changes were proposed in this PR? - Backport doc content that was edited directly on the website (`apache/incubator-texera-site`, `content/docs/latest/`) back into this repo's `docs/` source, so `docs/` is the source of truth for the docs sync. - `getting-started/installing-using-docker.md`: use the released download link instead of the RC5 dev-dist URL, add "Enable the Texera Agent" and "Add more LLMs or providers" sections, and fix the stop command to `docker compose --profile examples stop`. - `tutorials/migrate-jupyter-notebook.md`: correct the `hub.texera.io/dashboard/...` URLs. - Each file keeps its existing local front matter; website-only fields (`aliases`) and the `_index.md` `/docs/latest/` routing differences are intentionally not pulled, since they are site-specific. ### Any related issues, documentation, or discussions? Closes: #5627 ### How was this PR tested? - Docs-only change, no code paths affected, so no automated tests apply. - To verify content: open `docs/getting-started/installing-using-docker.md` and confirm the download link points to `downloads.apache.org/incubator/texera/1.1.0-incubating/...`, the "Enable the Texera Agent" and "Add more LLMs or providers" sections are present, and the stop command reads `docker compose --profile examples stop`. - To verify content: open `docs/tutorials/migrate-jupyter-notebook.md` and confirm the example links use the `hub.texera.io/dashboard/...` paths. - To confirm parity: diff `docs/` against the website's `content/docs/latest/` bodies and confirm the only remaining differences are website-only front matter and `_index.md` routing. ### Was this PR authored or co-authored using generative AI tooling? Generated-by: Claude Opus 4.8 --- docs/getting-started/installing-using-docker.md | 75 +++++++++++++++++++++---- docs/tutorials/migrate-jupyter-notebook.md | 16 +----- 2 files changed, 66 insertions(+), 25 deletions(-) diff --git a/docs/getting-started/installing-using-docker.md b/docs/getting-started/installing-using-docker.md index b269e98068..179ec87916 100644 --- a/docs/getting-started/installing-using-docker.md +++ b/docs/getting-started/installing-using-docker.md @@ -50,10 +50,10 @@ If either command produces output, that port is occupied by another process. You --- +## Download Texera -## Download the docker compose tarball from the release +Download the [docker compose tarball](https://downloads.apache.org/incubator/texera/1.1.0-incubating/apache-texera-1.1.0-incubating-docker-compose.tar.gz) and extract it. -Download by clicking [here](https://dist.apache.org/repos/dist/dev/incubator/texera/1.1.0-incubating-RC5/apache-texera-1.1.0-incubating-docker-compose.tar.gz) and extract it. ## Launch Texera @@ -62,19 +62,13 @@ Enter the extracted directory and run the following command to start Texera: docker compose --profile examples up ``` -This command will start docker containers that host the Texera services, and pre-create two example workflows and datasets. +This command will start docker containers that host the Texera services, and pre-create two example workflows and datasets. If you don't want to have these examples pre-created, run the following command instead: ```bash docker compose up ``` -To enable the AI copilot panel, also pass your LLM provider key inline. For example, with Anthropic: -```bash -export ANTHROPIC_API_KEY=<your-api-key> -docker compose --profile examples up -``` - > If you see the error message like `unable to get image 'nginx:alpine': > Cannot connect to the Docker daemon at > unix:///Users/kunwoopark/.docker/run/docker.sock. Is the docker daemon > running?`, please make sure Docker Desktop is installed and running > When you start Texera for the first time, it will take around 5 minutes to > download needed images. @@ -103,7 +97,7 @@ Press `Ctrl+C` in the terminal to stop Texera. If you already closed the terminal, you can go to the installation folder and run: ```bash -docker compose stop +docker compose --profile examples stop ``` to stop Texera. @@ -118,6 +112,28 @@ docker compose --profile examples down -v > ⚠️ Warning: This will permanently delete all the data used by Texera. +## Enable the Texera Agent + +The Texera agent is powered by a large language model (LLM). By default, Texera uses [Claude Haiku 4.5](https://www.anthropic.com/claude/haiku) as the LLM and queries it through [LiteLLM](https://docs.litellm.ai/). Without an API key, the Texera agent panel still appears but model calls will fail with a provider auth error. + +To enable it: + +1. [Stop Texera](#stop) if it is already running. +2. Get an API key for the LLM. Since Claude Haiku 4.5 is enabled by default, you need an [Anthropic API key](https://console.anthropic.com/settings/keys). +3. Export the key and restart Texera: + ```bash + export ANTHROPIC_API_KEY=sk-ant-... + docker compose --profile examples up + ``` + +Once Texera is up, create a new workflow and open the Texera agent panel at the bottom right. Type a task like: + +> For /texera/popular-movies-of-imdb/v1/TMDb_updated.csv, visualize the top 10 most-voted movies. + +To switch providers or add more LLMs, see [Add more LLMs or providers](#add-more-llms-or-providers). + + + ## Advanced Settings Before making any of the changes below, please [stop Texera](#stop) first. Once you finish the changes, [restart Texera](#restart) to apply them. @@ -161,6 +177,45 @@ $ docker compose up y // answer y to this prompt ``` +### Add more LLMs or providers +Only Claude Haiku 4.5 is enabled by default. To add more LLMs, open `litellm-config.yaml` in the installation folder and append entries under `model_list`. Each entry follows this shape: +```diff + model_list: + ... ++ - model_name: <name shown in Texera> ++ litellm_params: ++ model: <provider model id> ++ api_key: "os.environ/<API_KEY_ENV_VAR>" +``` +For example, to add OpenAI's GPT-5.2 and Google's Gemini 2.5 Pro: +```diff + model_list: + ... ++ - model_name: gpt-5.2 ++ litellm_params: ++ model: gpt-5.2 ++ api_key: "os.environ/OPENAI_API_KEY" ++ ++ - model_name: gemini-2.5-pro ++ litellm_params: ++ model: gemini/gemini-2.5-pro ++ api_key: "os.environ/GEMINI_API_KEY" +``` +Make sure to set the corresponding API key environment variable when you launch Texera (see [Enable the Texera Agent](#enable-the-texera-agent)). Get keys from each provider's console — for example, [OpenAI](https://platform.openai.com/api-keys) or [Google](https://aistudio.google.com/apikey). + +If your provider is not Anthropic, OpenAI, or Google, also pass its key into the LiteLLM container by editing `docker-compose.yml`: +```diff + litellm: + ... + environment: + ANTHROPIC_API_KEY: ${ANTHROPIC_API_KEY:-} + OPENAI_API_KEY: ${OPENAI_API_KEY:-} + GEMINI_API_KEY: ${GEMINI_API_KEY:-} ++ <NEW_API_KEY>: ${<NEW_API_KEY>:-} +``` + +For the full list of supported providers and model IDs, see the [LiteLLM proxy config docs](https://docs.litellm.ai/docs/providers). + ## Troubleshooting ### Port conflicts diff --git a/docs/tutorials/migrate-jupyter-notebook.md b/docs/tutorials/migrate-jupyter-notebook.md index eb1b3cea74..5635270414 100644 --- a/docs/tutorials/migrate-jupyter-notebook.md +++ b/docs/tutorials/migrate-jupyter-notebook.md @@ -16,7 +16,7 @@ Migrating notebook code into Texera operators, then wiring those operators with ## 2. Example: convert a "tweet analysis" notebook into a workflow -> The [notebook](https://hub.texera.io/user/dataset/124), [dataset](https://hub.texera.io/user/dataset/124) and [workflow](https://hub.texera.io/user/workflow/1162) in this example are available on [TexeraHub](https://hub.texera.io/about). +> The [notebook](https://hub.texera.io/dashboard/user/dataset/124), [dataset](https://hub.texera.io/dashboard/user/dataset/124) and [workflow](https://hub.texera.io/dashboard/user/workspace/1162) in this example are available on [TexeraHub](https://hub.texera.io/dashboard/about). ### Notebook Overview We will use a Tweet-Analysis notebook to demonstrate the migration process. The notebook has three cells: @@ -174,17 +174,3 @@ Texera contains more than 110 built-in operators that cover data loading, cleani - **Identify the data dependencies in the Python code in order to connect operators** In Texera, data flows along links. Before wiring operators, review the notebook to understand which variables feed which; then reproduce those dependencies via links so the executions matches the original notebook. - - - - - - - - - - - - - -
