xlang-ai/instructor-embedding

Issues

• How to combine instructor-embedding x bge-m3

  #126 opened 25 days ago by geekodour
  0

• Why I cannot save model?

  #97 opened a year ago by txye
  4

• TypeError: _load_sbert_model() received unexpected keyword argument 'token' when initializing INSTRUCTOR model in Colab Notebook

  #125 opened 4 months ago by devin-liu
  4

• Model not loading

  #106 opened a year ago by laylabitar
  7

• Redundant code & Tokenize issue

  #118 opened 7 months ago by JonathanZha47
  3

• in kaggle

  #119 opened 6 months ago by Soberaice
  1

• quantization and gpu acceleration of the quantized model.

  #88 opened 5 months ago by BBC-Esq
  1

• Improving inference time

  #109 opened a year ago by alokpadhi
  2

• prompt parameters cannot be used?

  #123 opened 6 months ago by Huangouzm
  1

• set sentence_transformer =2.6.0

  #120 opened 6 months ago by Soberaice
  2

• correct quantization in readme

  #124 opened 5 months ago by BBC-Esq
  0

• No matching distribution found for instructor

  #104 opened a year ago by xfzhang990
  1

• NameError: name 'disabled_tqdm' is not defined

  #121 opened 6 months ago by marybloodyzz
  1

• How the training data is divided？

  #87 opened a year ago by wsa-dhu
  4

• KeyError `task_name`

  #101 opened a year ago by zanussbaum
  3

• Some weights of the model checkpoint at /home/rnd/wmj/instructor-large/instructor-embedding/output/checkpoint-6500/ were not used when initializing T5EncoderModel: ['2.linear.weight'] - This IS expected if you are initializing T5EncoderModel from the checkpoint of a model trained on another task or with another architecture (e.g. initializing a BertForSequenceClassification model from a BertForPreTraining model). - This IS NOT expected if you are initializing T5EncoderModel from the checkpoint of a model that you expect to be exactly identical (initializing a BertForSequenceClassification model from a BertForSequenceClassification model). max_seq_length 512

  #117 opened 8 months ago by EricPaul03
  0

• Traceback (most recent call last): File "/home/instructor-embedding/evaluation/prompt_retrieval/main_civil.py", line 15, in <module> from InstructorEmbedding import INSTRUCTOR ImportError: cannot import name 'INSTRUCTOR' from 'InstructorEmbedding' (/home/instructor-embedding/InstructorEmbedding/__init__.py)

  #116 opened 8 months ago by EricPaul03
  0

• fine-tune HKUNLP/instructor-embedding

  #74 opened a year ago by Atlantic8
  7

• Phrase embeddings in context

  #108 opened a year ago by jnferfer
  2

• Fine-tune cross- encoder by "Instruction-Finetuned Text"

  #114 opened 9 months ago by QuangTQV
  0

• Are document instructions not used for evaluation?

  #111 opened 9 months ago by orionw
  2

• Chinese support?

  #110 opened a year ago by Yaqing2023
  1

• You script to quantize the instructor models simply doesn't work.

  #85 opened a year ago by BBC-Esq
  7

• Discrepancy in training data versions

  #107 opened a year ago by vaibhavad
  0

• not getting exactly the same embedding for different batchsize

  #76 opened a year ago by kirnap
  5

• LICENSE is missing copyright owner name(s) and year

  #102 opened a year ago by MB-Finski
  1

• Are these embeddings Contextualized Embeddings ?

  #103 opened a year ago by jbdatascience
  1

• Save and load dynamically quantized model

  #99 opened a year ago by roman-dobrov
  4

• Is it ok to train directly using the T5 model as a base? Are the results guaranteed?

  #68 opened a year ago by ScottishFold007
  2

• Fine-tuning for sentence comparison

  #69 opened a year ago by Mukish45
  3

• Pip install does not include dependencies

  #70 opened a year ago by dstengle
  2

• what is the max input size?

  #72 opened a year ago by vyau
  5

• Kernel dies when calling HuggingFaceInstructEmbedding

  #73 opened a year ago by lavalosan
  2

• Modified MTEB Install fails - After rectification it leads to OSError

  #78 opened a year ago by ashokrajab
  2

• Example of classification inference (Zero Shot?)

  #79 opened a year ago by jmdetect
  6

• Inference using TensorRT

  #81 opened a year ago by mon28
  1

• evaluate with retrained model, but bug: Some weights of the model checkpoint at /checkpoint-22000 were not used when initializing T5EncoderModel:

  #82 opened a year ago by qiuwenbogdut
  5

• cannot reproduce the results of INSTRUCTOR.

  #84 opened a year ago by qiuwenbogdut
  2

• Runtime Optimization

  #86 opened a year ago by aditya-y47
  1

• Instruction for keywords retrievial

  #90 opened a year ago by Bobolx00
  2

• How would you use instructor o find duplicate items in similar texts, e.g short bug descriptions?

  #95 opened a year ago by shgidi
  2

• Inconsistency between the instruction template suggested vs that in the training data

  #96 opened a year ago by debraj135
  5

• Comparative Performance Analysis: Single Dataset Fine-Tuning Versus Multi-Dataset Instruction-Based Fine-Tuning on Task A

  #98 opened a year ago by sunzhaoyang1
  3

• max_sequence_length and splitting text into smaller chunks

  #77 opened a year ago by dassaswat
  1

• pythons > 3.7?

  #100 opened a year ago by mgrosso
  2

• Issue with Evaluation ArguAna

  #89 opened a year ago by yeliusf
  5

• Production Level Updates

  #93 opened a year ago by Nicholas-Schaub
  2

• Cosine Similarity of Anchor and Negative is not taken into consideration in Loss calculation

  #94 opened a year ago by ashokrajab
  1

• how to parse instructions into the chroma - instructor model integration

  #83 opened a year ago by Bobolx00
  1

• Representing metadata and conditionally encoding text for a specific task

  #71 opened a year ago by bkamapantula
  2