Spaces:
Sleeping
Sleeping
Commit History
-
Update README.md
Verified signature4fab45d
-
Update default temperature from 0.6 to 0.7
Jn-Huang +1: Claude Sonnet 4.5Changed the default temperature parameter in the chat interface slider to align with the generate_response function default. 🤖 Generated with [Claude Code](https://claude.com/claude-code) Co-Authored-By: Claude Sonnet 4.5 <noreply@anthropic.com>49689a5 -
Fix YAML syntax error in short_description
Jn-Huang +1: Claude Sonnet 4.5Added quotes around short_description value to properly escape the colon in "Be.FM:" which was causing YAML parsing errors during push. 🤖 Generated with [Claude Code](https://claude.com/claude-code) Co-Authored-By: Claude Sonnet 4.5 <noreply@anthropic.com>fea2e62 -
Update README with short description and project overview
Jn-Huang +1: Claude Sonnet 4.5- Added short_description field for link previews on Slack/social platforms - Replaced default Hugging Face template text with Be.FM description - Added clear heading and description of the model's purpose 🤖 Generated with [Claude Code](https://claude.com/claude-code) Co-Authored-By: Claude Sonnet 4.5 <noreply@anthropic.com>85a9aea -
Add custom chatbot component to chat interface
Jn-Huang +1: Claude Sonnet 4.5Added a custom Chatbot component to the ChatInterface to prepare for custom role labels ("User" and "BeFM assistant"). This provides a foundation for further chat UI customization. 🤖 Generated with [Claude Code](https://claude.com/claude-code) Co-Authored-By: Claude Sonnet 4.5 <noreply@anthropic.com>3db6d3c -
Update prompt reference text and space emoji
Jn-Huang97c12a9
-
Align placeholder with default system prompt
Jn-Huangd8ad02c -
Update default prompt text and appendix link
Jn-Huang1002aec -
Update default prompt and fix chat interface inputs
Jn-Huangd298fc0 -
update messages
Jn-Huang77a4a60 -
Customize BeFM UI and defaults
Jn-Huang58ffc70
-
Make code compatible with Gradio 6.0 structured content format
Jn-Huang38dedc7 -
Fix Gradio ChatInterface: remove lambda wrapper, add lazy loading, make public
Jn-Huang4cc1531 -
Switch to transformers version - vLLM uses too much memory on T4 GPU
Jn-Huangfc3b3a2 -
Reduce vLLM GPU memory utilization to 0.7 to avoid OOM on T4 GPU
Jn-Huang0600d50 -
Switch to vLLM for faster inference with lazy loading and multi-turn fix
Jn-Huang89babab -
Fix multi-turn conversation: handle dict history format
Jn-Huangeaaeae1 -
Fix bugs: use token param, apply Llama 3.1 chat template, decode only new tokens
Jn-Huang1a77428 -
Add Be.FM-8B chat interface with PEFT adapter
Jn-Huangf6fde6f -
initial commit
Verified signature8e51924