finalRLEnv / app.py

Commit History

  1. fix: install nvidia-cudnn-cu12 alongside CUDA torch (removes --no-deps)

    0c3df82
  2. fix: force-reinstall CUDA torch before gradio import; populate logs on page load

    82cc86d
  3. fix: eliminate API leaks in training loop, set 30k steps for A100

    a5c7dd0
  4. Fix CPU training + 1902ms step speed bottleneck

    e8360fd
  5. fix: push reward curve + logs every 10k steps, not just at training end

    a798c01
  6. perf: add encode+step benchmark to logs, validate CUDA speed

    812dbcc
  7. perf: force ST to CUDA, disable LLM tasks, reduce to 100k steps

    b71b25c
  8. fix: force-install CUDA torch in app.py before training starts

    83cf44b
  9. pre-deployment fixes for HF Spaces

    aa154f2

  1. Remove gr.Timer: caused C-level crash with no Python output⁠

    garvitsachdeva +1: Claude Sonnet 4.6
    278b691
  2. Fix logs not showing: use file-based state instead of in-memory globals⁠

    garvitsachdeva +1: Claude Sonnet 4.6
    52e0424
  3. Add auto-refresh timer and dark theme CSS⁠

    garvitsachdeva +1: Claude Sonnet 4.6
    0c2a6e7
  4. Fix theme/css: keep in Blocks(), not launch()⁠

    garvitsachdeva +1: Claude Sonnet 4.6
    c0c63ff
  5. Improve Gradio UI: fix theme, add CSS, clean up layout⁠

    garvitsachdeva +1: Claude Sonnet 4.6
    c0997b4
  6. Fix create_repo naming error with huggingface_hub 1.x⁠

    garvitsachdeva +1: Claude Sonnet 4.6
    4bad83e
  7. Add startup diagnostics

    171b6d6
  8. SpindleFlow RL — periodic push + log persistence

    02ff91f