From 3f489c64d9e55d452c255f8e2c10b0d754883dbb Mon Sep 17 00:00:00 2001 From: bobzel Date: Mon, 21 Jul 2025 13:21:25 -0400 Subject: more server side cleanup --- src/server/chunker/requirements.txt | 37 +++++++++++++++++++++++++++++-------- 1 file changed, 29 insertions(+), 8 deletions(-) (limited to 'src/server/chunker') diff --git a/src/server/chunker/requirements.txt b/src/server/chunker/requirements.txt index 20bd486e5..586bbe505 100644 --- a/src/server/chunker/requirements.txt +++ b/src/server/chunker/requirements.txt @@ -1,15 +1,36 @@ +# Prefer official CPU wheels from the PyTorch index +--extra-index-url https://download.pytorch.org/whl/cpu + +############################################################################### +# Stable env for pdf_chunker.py # +############################################################################### + +# ─── LLM clients ───────────────────────────────────────────────────────────── +openai==1.40.6 +httpx==0.27.2 # <0.28 → avoids "proxies=" crash anthropic==0.34.0 cohere==5.8.0 -python-dotenv==1.0.1 + +# ─── Torch stack (CPU) ─────────────────────────────────────────────────────── +torch<=2.7.1 +torchvision<=0.22.1 # matches torch 2.5.x +torchaudio<=2.7.1 + +# ─── Vision / OCR / PDF processing ─────────────────────────────────────────── +ultralyticsplus==0.0.28 +easyocr==1.7.0 pymupdf==1.22.2 -lxml==5.3.0 +PyPDF2==3.0.1 +pytesseract==0.3.10 +Pillow==10.4.0 layoutparser==0.3.4 +lxml==5.3.0 + +# ─── ML / maths ────────────────────────────────────────────────────────────── numpy==1.26.4 -openai==1.40.6 -Pillow==10.4.0 -pytesseract==0.3.10 -PyPDF2==3.0.1 scikit-learn==1.5.1 + +# ─── Utilities ────────────────────────────────────────────────────────────── tqdm==4.66.5 -ultralyticsplus==0.0.28 -easyocr==1.7.0 \ No newline at end of file +python-dotenv==1.0.1 +packaging==24.0 \ No newline at end of file -- cgit v1.2.3-70-g09d2