I've been using AWS Bedrock, specifically the Nova model, because I want a more secure option compared to ChatGPT. Lately, I've uploaded some bank statements to have the model reference them for my business data. However, I keep running into an issue where I receive the error message, 'The generated text has been blocked by our content filters.' This is frustrating since I chose AWS Bedrock for its privacy features, and now these filters are getting in my way. I'm looking for help on a few things: any ways to remove or work around these content filters, fixes for this problem, or even alternative models that don't have such strict limitations. Just a heads up, my budget is pretty tight, so I can't afford to host a more powerful model myself.
2 Answers
Hey! Sorry to hear you’re having a tough time with the content filters. I found a couple of AWS docs that might help you out: https://go.aws/4f5B713 and https://go.aws/45l46dW. If you're still stuck, you might want to check this resource: http://go.aws/get-help.
I ran into the same issue a while back. After multiple support cases, the answer I got was to switch to another provider like Anthropic. They made it clear that the content filters with the Nova model are fixed and can’t be changed. We ended up moving to Claude because the Nova model just wouldn’t provide relevant answers for certain industry-specific topics. It might be worth considering if you can’t get around these restrictions.
Related Questions
xAI Grok Token Calculator
DeepSeek Token Calculator
Google Gemini Token Calculator
Meta LLaMA Token Calculator
OpenAI Token Calculator