V4.8.20
FastGPT V4.8.20 release notes covering container image updates, upgrade steps, and feature changes.
Upgrade Guide
1. Back up your database
2. Update environment variables
If you are an early adopter who configured ONEAPI_URL, you need to change it to OPENAI_BASE_URL.
3. Update images:
- Update the fastgpt image tag to v4.8.20-fix2
- Update the fastgpt-pro commercial edition image tag to v4.8.20-fix2
- Sandbox image update is not required
4. Run migration script
From any terminal, send an HTTP request. Replace {{rootkey}} with the rootkey from your environment variables, and {{host}} with your FastGPT domain:
curl --location --request POST 'https://{{host}}/api/admin/initv4820' \
--header 'rootkey: {{rootkey}}' \
--header 'Content-Type: application/json'The script automatically loads models from the old configuration file into the new model configuration.
Full Release Notes
- New - Visual model parameter configuration, replacing the old config file approach. Over 100 model presets included, with one-click testing for all model types. (Full in-page channel configuration is planned for the next version.) View model configuration guide
- New - DeepSeek Reasoner model supports outputting the thinking process.
- New - Usage record export and dashboard.
- New - Markdown syntax extension supporting audio and video (via
audioandvideocode blocks). - New - Adjusted
max_tokenscalculation logic. FastGPT reserves the configuredmax_tokensvalue for the response and trims conversation history when the request would exceed the context window. For example, settingmax_tokensto 8000 leaves 8000 fewer tokens available for input context. - Improved - Query rewriting now includes context filtering to prevent exceeding context limits.
- Improved - Page component extraction to reduce page component routing.
- Improved - Full-text search is now case-insensitive.
- Improved - QA generation and enhanced indexing switched to streaming output to prevent timeouts with some models.
- Improved - Automatically adds
nullto empty assistantcontent, and merges consecutive text assistant messages to prevent errors from some models. - Improved - Image URLs are completed with
FE_DOMAINimmediately before a chat request is sent. This keeps previously uploaded images usable after a domain change. - Fixed - Member list not triggering bottom-loading in some scenarios.
- Fixed - Workflow recursive execution failing under certain conditions.
Edit on GitHub
Last updated