🛫 Runway

Model deployment pipeline · GGUF library · LoRA fine-tunes · Fleet distribution

6
Deployed Models
36.7 GB
Total VRAM
1,884
Requests/hr
77 models
GGUF Library
14
LoRA Adapters

Deployed Models

ModelNodeStatusVRAM UsageRequests/hrAvg Latency
axe-phantom-8bJL1 (Nova)running6.2 GB142320ms
axe-blade-4bJL3 (Vigil)running3.1 GB89180ms
axe-oracle-27bJL1 (Nova)running18.4 GB341.2s
nomic-embed-textJL3 (Vigil)running0.8 GB1,24012ms
axe-nano-3bJL2 (Forge)running2.4 GB6795ms
partner:latestJL2 (Forge)running5.8 GB312240ms

📦 GGUF Library

77 GGUF models · 14 LoRA adapters

axe-phantom-8b
4.7 GBQ4_K_M
Downloaded
axe-oracle-27b
15.2 GBQ4_K_M
Downloaded
axe-blade-4b
2.8 GBQ5_K_M
Downloaded
axe-sage-14b
8.1 GBQ4_K_M
Downloaded
axe-sentinel-8b
5.9 GBQ5_K_M
Downloaded
axe-cipher-4b
4.2 GBQ8_0
Downloaded
axe-titan-70b
38.5 GBQ4_K_M
Available
axe-colossus-405b
220.0 GBF16
Available

🔥 Fine-Tune Pipeline

📊
Data Collection
active

Conversation pairs from team chat and daemon interactions stored in training_pairs table

4,821
training pairs
Last run
2h ago
🔥
Training
idle

LoRA fine-tuning on collected data with automatic hyperparameter selection

14
LoRA adapters
Last run
3d ago
🚀
Deployment
idle

Push trained models to fleet nodes with automatic VRAM allocation and health checks

6
deployed models
Last run
5d ago