ROCm/FastFlowLM v0.1.3

🚀 FastFlowLM v0.1.3 — Metadata, File Prompts, and Code Readiness

on GitHub

🚀 FastFlowLM v0.1.3 — Metadata, File Prompts, and Code Readiness

FastFlowLM v0.1.3 brings important improvements for usability, WebUI integration, and future extensibility.

✨ What's New

1. REST API Now Includes Metadata

   • Added generation metadata (e.g., generation speed, token count) in the API response.
   • This enables proper integration with WebUI clients that rely on these fields.

2. Prompt File Input for flm run

   • You can now pass a prompt text file directly:

     flm run llama3.2 prompt.txt

3. Code Refactoring for Future Models

   • Internal code structure has been reorganized to support upcoming models more cleanly and modularly.

4. Improved Model Download (Pull) Interface

   • The model pulling mechanism is more robust and user-friendly for downloading models on demand.

✅ Usage Recap

Interactive CLI:

flm run llama3.2
flm run llama3.2 prompt.txt

Server Mode:

flm serve

Includes /v1/api/completions with metadata support.