Skip to content

armgit / Soup

Public Imported from GitHub
0

Fine-tune LLMs from one YAML. Layer streaming trains an 8B model on a 4 GB laptop GPU.

197 B
Raw Blame
{
  "identical_responses": true,
  "rows": 24,
  "wall_seconds": {
    "a1": 282.2454737,
    "b1": 101.5466282,
    "b2": 96.2936871,
    "a2": 131.3516435
  },
  "d_pass": true,
  "pass": true
}