illuin-tech/colpali

Issues

• Why not remove describe text embedding in image_embeddings output

  #97 opened 2 months ago by kiendn1
  2

• How to run eval on a finetuned model?

  #100 opened 2 months ago by VoVAllen
  2

• [question] usage in Rust/ONNX?

  #103 opened 2 months ago by louis030195
  2

• source pdf files for benchmark

  #90 opened 2 months ago by roipony
  1

• quantized version of the model fo CPU inference

  #78 opened 2 months ago by akshayballal95
  2

• Align internal methods with other vision models, e.g. Clip. Improvements to the AutoProcessor

  #92 opened 2 months ago by michaelfeil
  5

• Offline model path produces different results when compared to standard name path

  #95 opened 2 months ago by venki-lfc
  3

• Fine tuning the model with hard negatives: compute_harnegs.py fails

  #80 opened 2 months ago by CommissarSilver
  5

• Add `AutoModel` and `AutoProcessor` to libary

  #91 opened 2 months ago by michaelfeil
  3

• Removal of interpretability

  #81 opened 3 months ago by tilo-schroeder
  3

• Where to store the embeddings for production

  #85 opened 2 months ago by iamkhalidbashir
  5

• Error: Processor should be provided for vision collator

  #89 opened 2 months ago by vicaasas
  2

• Faster unit tests via smaller dummy models

  #93 opened 2 months ago by michaelfeil
  2

• Replicating ColQwen2 results

  #84 opened 2 months ago by oxpie
  5

• Integrate late chunking for (potentially) missing context problem.

  #82 opened 3 months ago by muazhari
  1

• Custom text prompt

  #77 opened 2 months ago by nuschandra
  2

• High memory usage

  #87 opened 2 months ago by iamkhalidbashir
  4

• CustomCollator Problem

  #88 opened 2 months ago by vicaasas
  1

• Doubt regarding Gemma 2b used in the model

  #86 opened 2 months ago by Papz2k
  1

• ColQwen2 generates zero vectors for doc embeddings

  #79 opened 3 months ago by jbellis
  4

• Figure 1

  #74 opened 3 months ago by warm345
  3

• current code for `clopali1.2` is not runnable/trainable

  #66 opened 3 months ago by CommissarSilver
  2

• Training details `vidore/colpali-v1.2`

  #65 opened 3 months ago by d-rau
  1

• What is granularity unit of input chunks?

  #42 opened 3 months ago by muazhari
  1

• OCR with colpali

  #68 opened 3 months ago by lukiod
  1

• How the chunk comparison works?

  #41 opened 3 months ago by muazhari
  1

• A question related to the dataset refactor

  #70 opened 3 months ago by XMHZZ2018
  1

• Possible to normalize colpali_engine.models.ColPaliProcessor.score_multi_vector() to value between 0-1?

  #64 opened 3 months ago by plamb-viso
  5

• dose it can using the other LLM ? like Qwen2-vl or MiniCPMV, Llava, any VLM

  #35 opened 4 months ago by Lbaiall
  3

• Re-indexing issue?

  #53 opened 3 months ago by skyxiaobai
  2

• Fix Up Inference Code with colpali_engine==0.3.0

  #57 opened 3 months ago by DrChrisLevy
  1

• Code for "similarity of the image patches w.r.t. the underlined token"

  #58 opened 3 months ago by roipony
  1

• Hugging Face Model Card Inference Code not working after recent refactor

  #55 opened 3 months ago by DrChrisLevy
  3

• #Restructuring cause code breaks

  #43 opened 3 months ago by Jaykumaran
  2

• Facing error with importing utility function in Colab after recent code restructuring.

  #44 opened 3 months ago by Jaykumaran
  1

• Multilingual Colpali Adapter Training

  #19 opened 4 months ago by contrebande-labs
  11

• Dataset Release

  #34 opened 4 months ago by AdiscoUSR
  5

• Training for any language

  #13 opened 5 months ago by GriffithV
  5

• Missing training dataset

  #16 opened 5 months ago by VoVAllen
  5

• Generalize the training configuration

  #32 opened 4 months ago by efenocchi
  3

• Base Model Linear Projection Layer Weights

  #31 opened 4 months ago by contrebande-labs
  5

• Inconsistent results - hosted demo vs. running locally

  #17 opened 4 months ago by RahulNET
  44

• Using different VLM

  #26 opened 4 months ago by sky-2002
  1

• Different embedding for same image when model reloads

  #30 opened 4 months ago by paniabhisek
  1

• Fine tuning code

  #29 opened 4 months ago by nico2rdj
  0

• Model quantization

  #21 opened 4 months ago by sky-2002
  1

• it would be better if it can involved the type class "text"，such like the text embeding

  #20 opened 4 months ago by Lbaiall
  1

• More detailed steps to install this application or Dockerfile

  #18 opened 4 months ago by dev3py
  1

• Query preprocessing

  #12 opened 5 months ago by hannah348
  4

• Need a python file

  #14 opened 5 months ago by UCASlemontea
  2