Mortdecai

Seth/Mortdecai

Fork 0

Commit Graph

Author	SHA1	Message	Date
Mortdecai	da8f557219	GPU scheduler, 14-tool architecture, plugin deployment, event dispatcher GPU Scheduler (gpu.sethpc.xyz): - Live dashboard with 4 GPUs, training monitor, loss sparklines - Preset-based job scheduler with 3 triggers (time, finish_training, cost) - Model selection per GPU, pipeline configuration - Tool self-play and training pipeline types - Behind Google OAuth, live-refresh without page reload Tool Architecture (14 tools): - 3 new tools: world.nearby_entities, memory.read, memory.write - 7 script.* tools: write, validate, execute, read, list, delete, schedule - ScriptManager: full mcfunction datapack CRUD with RCON validation - Training data: 1,430 tool examples (up from 1,159) Plugin Deployment (paper-ai-25567): - WorldGuard 7.0.12, CoreProtect CE 23.1, EssentialsX 2.21.2, Vault 1.7.3 - Fresh greenfield world reset - 104 RCON-validated plugin training examples Event Dispatcher: - Watches server log for deaths, joins, advancements, PvP kills - Configurable trigger probability and cooldowns per event type - Deployed to dev server, fires god_system prompts on events - 21 event-response training examples Training Infrastructure: - train_lora.py: --save-steps 50, --resume from checkpoint - run_training.sh: stops Ollama, activates conda, restarts after - Passwordless sudo for ollama services on steel141 - Dev server added to MCSManager with autoStart Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>	2026-03-21 03:14:45 -04:00
Seth	ee764cd22a	Tool-calling training: 1,159 multi-turn examples with error correction Tool schemas (agent/tools/tool_schemas.py): - rcon.execute: execute commands, get success/error results - minecraft.wiki_lookup: look up syntax and item info - world.player_info: player health, position, inventory - world.server_state: time, weather, online players - 10 RCON error patterns with corrections - 12 common error scenarios for training Training data generator (training/scripts/generate_tool_training.py): - Converts seed dataset to multi-turn tool conversations - Error correction: model tries wrong command → gets error → self-corrects - Wiki/player/server lookups for uncertainty scenarios - Qwen3 native tool-calling format with <tool_call> tags 1,159 examples: 1043 success, 79 error correction, 24 error scenarios, 13 tool lookups. Ready for v4 training. Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>	2026-03-19 18:49:08 -04:00

Author

SHA1

Message

Date

Mortdecai

da8f557219

GPU scheduler, 14-tool architecture, plugin deployment, event dispatcher

GPU Scheduler (gpu.sethpc.xyz):
- Live dashboard with 4 GPUs, training monitor, loss sparklines
- Preset-based job scheduler with 3 triggers (time, finish_training, cost)
- Model selection per GPU, pipeline configuration
- Tool self-play and training pipeline types
- Behind Google OAuth, live-refresh without page reload

Tool Architecture (14 tools):
- 3 new tools: world.nearby_entities, memory.read, memory.write
- 7 script.* tools: write, validate, execute, read, list, delete, schedule
- ScriptManager: full mcfunction datapack CRUD with RCON validation
- Training data: 1,430 tool examples (up from 1,159)

Plugin Deployment (paper-ai-25567):
- WorldGuard 7.0.12, CoreProtect CE 23.1, EssentialsX 2.21.2, Vault 1.7.3
- Fresh greenfield world reset
- 104 RCON-validated plugin training examples

Event Dispatcher:
- Watches server log for deaths, joins, advancements, PvP kills
- Configurable trigger probability and cooldowns per event type
- Deployed to dev server, fires god_system prompts on events
- 21 event-response training examples

Training Infrastructure:
- train_lora.py: --save-steps 50, --resume from checkpoint
- run_training.sh: stops Ollama, activates conda, restarts after
- Passwordless sudo for ollama services on steel141
- Dev server added to MCSManager with autoStart

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>

2026-03-21 03:14:45 -04:00

Seth

ee764cd22a

Tool-calling training: 1,159 multi-turn examples with error correction

Tool schemas (agent/tools/tool_schemas.py):
- rcon.execute: execute commands, get success/error results
- minecraft.wiki_lookup: look up syntax and item info
- world.player_info: player health, position, inventory
- world.server_state: time, weather, online players
- 10 RCON error patterns with corrections
- 12 common error scenarios for training

Training data generator (training/scripts/generate_tool_training.py):
- Converts seed dataset to multi-turn tool conversations
- Error correction: model tries wrong command → gets error → self-corrects
- Wiki/player/server lookups for uncertainty scenarios
- Qwen3 native tool-calling format with <tool_call> tags

1,159 examples: 1043 success, 79 error correction, 24 error scenarios,
13 tool lookups. Ready for v4 training.

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>

2026-03-19 18:49:08 -04:00

2 Commits