sleepy 3799240d74 fix: head node objectively judges all responses using quality metrics ...

- Removed biased self-reported confidence voting
- Head node now collects ALL responses and scores them objectively
- Uses quality scoring (length, structure, completeness) to compare
- Shows quality scores for all nodes so user can see comparison
- Prevents overconfident small models from beating better large models
- 3B models will only win if they actually produce better quality output
- All 41 tests passing

2026-02-25 22:24:00 +01:00

..

api

fix: use actual consensus confidence for peers instead of hardcoded 0.8

2026-02-25 22:21:19 +01:00

backends

feat: CUDA/Android support and federation metrics (#7)

2026-02-25 00:53:07 +01:00

cli

fix: OpenAI API compatibility for hollama and other clients

2026-02-25 19:39:05 +01:00

hardware

feat: CUDA/Android support and federation metrics (#7)

2026-02-25 00:53:07 +01:00

interactive

refactor(interactive): create modular structure for interactive module

2026-02-25 13:13:21 +01:00

models

fix: OpenAI API compatibility for hollama and other clients

2026-02-25 19:39:05 +01:00

network

fix: head node objectively judges all responses using quality metrics

2026-02-25 22:24:00 +01:00

swarm

refactor(swarm): add orchestrator module for swarm generation

2026-02-25 13:14:47 +01:00

tools

fix: allow absolute and ~ paths to access files outside working directory

2026-02-25 21:13:02 +01:00

utils

refactor(cli): break down main.py into modular CLI components

2026-02-25 12:57:28 +01:00

__init__.py

Initial commit: Local Swarm project structure and documentation

2026-02-23 16:46:31 +01:00

chatlog.py

feat: add chatlogger for tool execution debugging - logs to chatlog.md when LOCAL_SWARM_CHATLOG=1

2026-02-25 19:52:52 +01:00

interactive.py

feat: add command line example to startup summary

2026-02-24 13:02:18 +01:00

mcp_server.py

feat: Apple Silicon MLX support, sequential workers, live status display, worker names

2026-02-23 22:57:38 +01:00