Hugging Face's logo Hugging Face
  • Models
  • Datasets
  • Spaces
  • Buckets new
  • Docs
  • Enterprise
  • Pricing

  • Log In
  • Sign Up

Spaces:
remcohendriks
/
metrollm
Running on Zero

App Files Files Community
Fetching metadata from the HF Docker repository...
metrollm
1.5 MB
Ctrl+K
Ctrl+K
  • 3 contributors
History: 33 commits
Remco Hendriks
Claude Opus 4.7 (1M context)
roll back torch 2.6 + fla path (TTFT + duration cap crashes)
dc22384 5 days ago
  • dashboard
    init: MetroLLM-Bench kiosk demo (2B + LoRA, free CPU) 9 days ago
  • data
    flash-attn for ZeroGPU + regional scenario flavor + subtitle 6 days ago
  • harness
    fix: normalize disruption segment to list at /set_disruptions 8 days ago
  • .gitattributes
    1.52 kB
    initial commit 9 days ago
  • .gitignore
    58 Bytes
    init: MetroLLM-Bench kiosk demo (2B + LoRA, free CPU) 9 days ago
  • README.md
    833 Bytes
    deploy: 9B + ZeroGPU (low_cpu_mem_usage for module-load) 9 days ago
  • app.py
    113 kB
    revert regex throttle (TTFT regressed to ~20s) 5 days ago
  • model.py
    14.1 kB
    probe runtime attn_implementation post-PEFT merge 6 days ago
  • prompts.py
    4.36 kB
    init: MetroLLM-Bench kiosk demo (2B + LoRA, free CPU) 9 days ago
  • requirements.txt
    1.35 kB
    roll back torch 2.6 + fla path (TTFT + duration cap crashes) 5 days ago
  • tools.py
    4.05 kB
    init: MetroLLM-Bench kiosk demo (2B + LoRA, free CPU) 9 days ago