PaperProf / core

Commit History

  1. fix: pre-translate non-English chunks to English before LLM inference⁠

    Mehdi
    67da08d
  2. fix: enable sampling — temperature=0.8 for MCQ, 0.4 for evaluator⁠

    Mehdi
    8f2e039
  3. fix: stronger MCQ/evaluator prompts — no duplicates, no self-reference, English-only, section 2 always present⁠

    Mehdi
    c8e8b73

  1. fix: make the llama.cpp CPU runtime actually usable on the Space⁠

    Mehdi
    b5f4a07
  2. fix: truncate runaway question generation to the first question⁠

    Mehdi
    81e2542
  3. feat: amphitheater hero with mouse-tracking robot professor, fixed particles

    57ebbd6
  4. fix: force English output in all LLM prompts regardless of PDF language⁠

    Mehdi
    490c5f1

  1. fix: server error

    5ff15a8
  2. fix: increase GPU duration to 120s, prefetch FLUX weights on startup

    2bc80d0
  3. fix: increase GPU duration to 120s, prefetch FLUX weights on startup

    5158491
  4. fix: FLUX.2-klein reajust

    6f1b463
  5. fix: FLUX.2-klein prompt keyword arg, image for all verdicts

    413ab6b
  6. feat: switch to FLUX.2-klein-4B for image generation

    3e0f4f1
  7. feat: FLUX.1-schnell image generation after correct answers

    53f316c
  8. feat: add MCQ mode with 4 options, client-side evaluation and per-choice explanations

    Mehdi
    978d90d

  1. feat: Off-Brand badge - custom FastAPI+HTML frontend, zero Gradio chrome

    e984a0a

  1. fix: prompts évaluateur séparés EN/FR pour forcer la langue du feedback

    Mehdi
    5e6ff7d
  2. fix: texte question/feedback invisible + prompt langue renforcé

    Mehdi
    7c23d32
  3. fix: tighten question prompt to avoid compound multi-part questions

    Mehdi
    3e7d6d6
  4. feat: switch to MiniCPM4-8B, add robustness, score tracker, language selector⁠

    Mehdi
    83ec3f5

  1. feat: initial project structure with all core modules

    e1c0b77