{"id":25121347,"url":"https://github.com/angellozan/local_deep_seek","last_synced_at":"2025-04-02T15:22:55.619Z","repository":{"id":276361973,"uuid":"929040961","full_name":"AngelLozan/Local_Deep_Seek","owner":"AngelLozan","description":"How to run DeepSeek on your local Mac","archived":false,"fork":false,"pushed_at":"2025-02-08T00:54:54.000Z","size":283,"stargazers_count":0,"open_issues_count":0,"forks_count":0,"subscribers_count":1,"default_branch":"main","last_synced_at":"2025-02-08T01:28:38.953Z","etag":null,"topics":["ai","deepseek","deepseek-ai","deepseek-chat","deepseek-coder","deepseek-llm","deepseek-r1","open-web","vscode","vscode-extension"],"latest_commit_sha":null,"homepage":"","language":null,"has_issues":true,"has_wiki":null,"has_pages":null,"mirror_url":null,"source_name":null,"license":null,"status":null,"scm":"git","pull_requests_enabled":true,"icon_url":"https://github.com/AngelLozan.png","metadata":{"files":{"readme":"README.md","changelog":null,"contributing":null,"funding":null,"license":null,"code_of_conduct":null,"threat_model":null,"audit":null,"citation":null,"codeowners":null,"security":null,"support":null,"governance":null,"roadmap":null,"authors":null,"dei":null,"publiccode":null,"codemeta":null}},"created_at":"2025-02-07T17:42:43.000Z","updated_at":"2025-02-08T00:54:57.000Z","dependencies_parsed_at":"2025-02-07T19:39:24.982Z","dependency_job_id":null,"html_url":"https://github.com/AngelLozan/Local_Deep_Seek","commit_stats":null,"previous_names":["angellozan/deepseeklocal","angellozan/local_deep_seek"],"tags_count":0,"template":false,"template_full_name":null,"repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/AngelLozan%2FLocal_Deep_Seek","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/AngelLozan%2FLocal_Deep_Seek/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/AngelLozan%2FLocal_Deep_Seek/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/AngelLozan%2FLocal_Deep_Seek/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/AngelLozan","download_url":"https://codeload.github.com/AngelLozan/Local_Deep_Seek/tar.gz/refs/heads/main","host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":246838308,"owners_count":20842049,"icon_url":"https://github.com/github.png","version":null,"created_at":"2022-05-30T11:31:42.601Z","updated_at":"2022-07-04T15:15:14.044Z","host_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub","repositories_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories","repository_names_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repository_names","owners_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners"}},"keywords":["ai","deepseek","deepseek-ai","deepseek-chat","deepseek-coder","deepseek-llm","deepseek-r1","open-web","vscode","vscode-extension"],"created_at":"2025-02-08T06:17:55.578Z","updated_at":"2025-04-02T15:22:55.593Z","avatar_url":"https://github.com/AngelLozan.png","language":null,"funding_links":[],"categories":[],"sub_categories":[],"readme":"# Run DeepSeek Locally with Docker-Compose\n\u003cp align=\"center\"\u003e\n    \u003cimg src=\"image-13.png\" alt=\"alt text\"\u003e\n\u003c/p\u003e\n\nRunning DeepSeek locally with Docker-Compose is possible with a Mac, though a lighter-weight implementation of the model is recommended.\n\nThis will take you through how to run DeepSeek on localhost with a web-ui interface. \n\n\u003cp align=\"center\"\u003e\n    \u003cimg src=\"image.png\" alt=\"alt text\"\u003e\n\u003c/p\u003e\n\n\u003cp align=\"center\"\u003e\n    \u003cimg src=\"image-1.png\" alt=\"alt text\"\u003e\n\u003c/p\u003e\n\n\u003chr/\u003e\n\n## Run with web interface\n\n**These steps require internet connection**\n\n1. Install [Ollama](https://martech.org/how-to-run-deepseek-locally-on-your-computer/)\n\n2.  Pick a model based on your hardware:\n```\nollama pull deepseek-r1:8b  # Fast, lightweight  \n\nollama pull deepseek-r1:14b # Balanced performance  \n\nollama pull deepseek-r1:32b # Heavy processing  \n\nollama pull deepseek-r1:70b # Max reasoning, slowest\n\nollama pull deepseek-coder:1.3b # Code completion assist\n```\n3. Test the model locally via the terminal \n```\nollama run deepseek-r1:8b\n```\n\n4. Install [Docker](https://www.docker.com/get-started)\n\n5. Install [Docker-Compose](https://formulae.brew.sh/formula/docker-compose)\n\n6. Create Docker-Compose file as seen in this repo. If you wish to use an internet connect, you can simply uncomment the image for the open-webui service and remove the build. \n\n7. Open the docker app and run `docker-compose up --build`\n\n8. Visit `http://localhost:3000` to see your chat. \n\n\u003chr/\u003e\n\n## Run with VScode (offline):\n\n1. Follow steps 1-2 in the [Steps to run with a web interface](#run-with-web-interface), then you can also install the CodeGPT for VScode extension. \n![alt text](image-2.png)\n![alt text](image-3.png)\n\n2. Navigate to the Local LLMs section. This is likely accessed from the initial model selection drop down (pictured with claude selected).\n![alt text](image-9.png)\n![alt text](image-10.png)\n\n3. From the available options, select 'Ollama' as the local LLM provider.\n![alt text](image-11.png)\n\n4. Select your DeepSeek Model and you're done.\n![alt text](image-12.png)\n\n5. You can now turn off internet and using Local LLMs, continue to chat/analyze code. \n\n\u003chr/\u003e\n\n## Running Open-webui locally without internet\n\n1. Follow steps 1-2 in the [Steps to run with a web interface](#run-with-web-interface)\n\n2. Install `uv` `curl -LsSf https://astral.sh/uv/install.sh | sh`\n\n3. Create uv env:  `mkdir ~/\u003c project root \u003e/\u003c your directory name\u003e \u0026\u0026 uv venv --python 3.11`\n\n4. Install open-webui: `cd ~/\u003c project root \u003e/\u003c your directory name \u003e \u0026\u0026 uv pip install open-webui`\n\n5. Start open-webui: `DATA_DIR=~/.open-webui uv run open-webui serve`\n\n6. Visit localhost and start chatting!\n\n\u003chr/\u003e\n\n## Running locally via docker without internet with Open-webui\n\n1. Follow steps 1-6 in the [steps to run with a web interface](#run-with-web-interface)\n\n2. Next, follow steps 1-4 in the [steps to running open-webui locally without internet](#running-open-webui-locally-without-internet)\n\n3. Once this is done, create a `Dockerfile` in your chosen directory where the open-webui deps live like that seen in this project, to mimic the setup and install in the docker container of all dependencies for open-webui.\n\n4. Next, start the app: `docker-compose up --build`. If you do not wish to see logs: `docker-compose up --build -d`\n\n5. Visit localhost and start chatting!\n\n6. If models are not available to select, turn on your internet temporarily, go back to the terminal and run `docker exec -it ollama bash`\n\n7. Download the model to your service using the `ollama pull` commands seen earlier in step 1. \n\n8. Verify models are installed with `ollama list` while still in the cli. If so, you can turn off internet again and exit the cli with `ctrl + d` or `exit`\n\n9. Restart your `open-webui` container with `docker-compose restart open-webui`\n\n\u003chr/\u003e\n\n## Troubleshooting\n\n1. Inspect the network: `docker network ls` then `docker network inspect \u003c network \u003e`\n\n2. Inspect Ollama and models: \n\n`curl http://localhost:11434/api/tags`\n\n- or -\n\n`docker exec -it ollama ollama list`\n\n3. Restart open-webui container: `docker-compose restart open-webui`\n\n4. Depending on your hardware, running `docker-compose down` then `docker-compose up -d` to restart built containers can take a moment. Check progress with `docker logs \u003c service name \u003e`","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fangellozan%2Flocal_deep_seek","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Fangellozan%2Flocal_deep_seek","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fangellozan%2Flocal_deep_seek/lists"}