DeepSeek R1 Local Image Upload - I Finally fixed it, Upload images locally!
Вставка
- Опубліковано 9 лют 2025
- #deepseek #ai #openai
Check out my Udemy course to create more advanced AI apps: www.udemy.com/...
Image upload has been down for almost 4 days and they never fixed it, so I created a local model that allows you to upload your images to DeepSeek R1
Code: github.com/mur...
#deepseek #ai #deepseekr1 #openai #chatgpt #artificialintelligence
Sick !!!
Thanks man
Thanks for sharing Murat, please correct me if I'm mistaken;
OCR can only get you the text so it wont work if your file is an artwork therefore you are using a paid-service via API key (I recommend excluding it from your code btw via using enviroment variables because you show them in the video even if its not complete on this one) my question is wouldn't it better feed the image directly to locally running Image-Text-to-Text model like LLava? (Başarılar)
When running DeepSeek locally with powershell & Ollama can you upload images from Ollama CLI (command line interface). Without using Python?
unfortunately no, the only way is, using OCR libraries to make the image compatible and ready for the model to read them. Deepseek itself is also using the same method on its web interface
@@murataavcu Do you have step-by-step instructions on what needs to be downloaded to replicate what is done in this video
@@karivirtanen3255 Yes you can check the installation process for the local deepseek model in my previous video. after that point on you can easily copy what's happening in this video of local image upload. Local deepseek installation: ua-cam.com/video/FVA-Rvet6B4/v-deo.html , You can find the code in the video here: github.com/muratali016/DeepSeek-R1-Codes , if you have more questions don't hesitate to ask, please
@@murataavcu Is this example totally local (no-internet)? For example does using ocr and openai require a connection afte a key has been gotten?