xlang-ai/instructor-embedding issues and pull requests

#129 - update instructor for sentence transformers

Pull Request - State: open - Opened by BBC-Esq about 1 month ago

#128 - Adding support for Sentence-Transformers 3.3.1

Pull Request - State: open - Opened by NoahBPeterson about 1 month ago

#127 - Resolve compatibility issues with sentence-transformers 3.3.1

Pull Request - State: open - Opened by merthmagic about 2 months ago - 3 comments

#126 - How to combine instructor-embedding x bge-m3

Issue - State: open - Opened by geekodour about 2 months ago - 2 comments

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

Issue - State: open - Opened by devin-liu 5 months ago - 5 comments

#124 - correct quantization in readme

Issue - State: closed - Opened by BBC-Esq 6 months ago

#123 - prompt parameters cannot be used?

Issue - State: open - Opened by Huangouzm 7 months ago - 1 comment

#122 - Update _load_sbert_model Parameters and Fix Tokenize Padding

Pull Request - State: open - Opened by pascalhuerten 7 months ago

#121 - NameError: name 'disabled_tqdm' is not defined

Issue - State: open - Opened by marybloodyzz 7 months ago - 1 comment

#120 - set sentence_transformer =2.6.0

Issue - State: open - Opened by Soberaice 8 months ago - 2 comments

#119 - in kaggle

Issue - State: open - Opened by Soberaice 8 months ago - 1 comment

#118 - Redundant code & Tokenize issue

Issue - State: open - Opened by JonathanZha47 9 months ago - 3 comments

#117 - 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

Issue - State: open - Opened by EricPaul03 9 months ago

#116 - 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)

Issue - State: closed - Opened by EricPaul03 10 months ago

#115 - Update to newer sentence-transformers with offline usage and renaming of classes to match the released package

Pull Request - State: closed - Opened by racinmat 10 months ago - 3 comments

#114 - Fine-tune cross- encoder by "Instruction-Finetuned Text"

Issue - State: open - Opened by Lahaina936 10 months ago

#113 - please accept this pull request ASAP PLEASE!

Pull Request - State: closed - Opened by BBC-Esq 10 months ago - 10 comments

#112 - Updated to work with the latest version of sentence transformers

Pull Request - State: closed - Opened by SilasMarvin 11 months ago - 14 comments

#111 - Are document instructions not used for evaluation?

Issue - State: closed - Opened by orionw 11 months ago - 2 comments

#110 - Chinese support?

Issue - State: open - Opened by Yaqing2023 12 months ago - 1 comment

#109 - Improving inference time

Issue - State: open - Opened by alokpadhi about 1 year ago - 2 comments

#108 - Phrase embeddings in context

Issue - State: open - Opened by jnferfer about 1 year ago - 2 comments

#107 - Discrepancy in training data versions

Issue - State: open - Opened by vaibhavad about 1 year ago

#106 - Model not loading

Issue - State: open - Opened by laylabitar about 1 year ago - 7 comments

#105 - Pooling

Pull Request - State: closed - Opened by ashokrajab about 1 year ago

#104 - No matching distribution found for instructor

Issue - State: closed - Opened by xfzhang990 about 1 year ago - 1 comment

#103 - Are these embeddings Contextualized Embeddings ?

Issue - State: open - Opened by jbdatascience about 1 year ago - 1 comment

#102 - LICENSE is missing copyright owner name(s) and year

Issue - State: closed - Opened by MB-Finski about 1 year ago - 1 comment

#101 - KeyError `task_name`

Issue - State: open - Opened by zanussbaum about 1 year ago - 3 comments

#100 - pythons > 3.7?

Issue - State: closed - Opened by mgrosso over 1 year ago - 2 comments

#99 - Save and load dynamically quantized model

Issue - State: open - Opened by roman-dobrov over 1 year ago - 4 comments

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

Issue - State: closed - Opened by sunzhaoyang1 over 1 year ago - 3 comments

#97 - Why I cannot save model?

Issue - State: open - Opened by txye over 1 year ago - 4 comments

#96 - Inconsistency between the instruction template suggested vs that in the training data

Issue - State: open - Opened by debraj135 over 1 year ago - 5 comments

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

Issue - State: closed - Opened by shgidi over 1 year ago - 2 comments

#94 - Cosine Similarity of Anchor and Negative is not taken into consideration in Loss calculation

Issue - State: closed - Opened by ashokrajab over 1 year ago - 1 comment

#93 - Production Level Updates

Issue - State: open - Opened by Nicholas-Schaub over 1 year ago - 2 comments

#92 - Modifed masking before pooling - Fixes issue in ONNX conversion

Pull Request - State: closed - Opened by ashokrajab over 1 year ago - 4 comments

#91 - Update README.md

Pull Request - State: closed - Opened by TuanaCelik over 1 year ago

#90 - Instruction for keywords retrievial

Issue - State: closed - Opened by Bobolx00 over 1 year ago - 2 comments

#89 - Issue with Evaluation ArguAna

Issue - State: open - Opened by yeliusf over 1 year ago - 5 comments

#88 - quantization and gpu acceleration of the quantized model.

Issue - State: closed - Opened by BBC-Esq over 1 year ago - 1 comment

#87 - How the training data is divided？

Issue - State: open - Opened by wsa-dhu over 1 year ago - 4 comments

#86 - Runtime Optimization

Issue - State: open - Opened by aditya-y47 over 1 year ago - 1 comment

#85 - You script to quantize the instructor models simply doesn't work.

Issue - State: closed - Opened by BBC-Esq over 1 year ago - 7 comments

#84 - cannot reproduce the results of INSTRUCTOR.

Issue - State: closed - Opened by qiuwenbogdut over 1 year ago - 2 comments

#83 - how to parse instructions into the chroma - instructor model integration

Issue - State: closed - Opened by Bobolx00 over 1 year ago - 1 comment

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

Issue - State: closed - Opened by qiuwenbogdut over 1 year ago - 5 comments

#81 - Inference using TensorRT

Issue - State: open - Opened by mon28 over 1 year ago - 1 comment

#80 - Bugfix AbsTaskRetrieval for DREPS evaluation

Pull Request - State: closed - Opened by ashokrajab over 1 year ago

#79 - Example of classification inference (Zero Shot?)

Issue - State: closed - Opened by jmdetect over 1 year ago - 6 comments

#78 - Modified MTEB Install fails - After rectification it leads to OSError

Issue - State: closed - Opened by ashokrajab over 1 year ago - 2 comments

#77 - max_sequence_length and splitting text into smaller chunks

Issue - State: closed - Opened by dassaswat over 1 year ago - 1 comment

#76 - not getting exactly the same embedding for different batchsize

Issue - State: open - Opened by kirnap over 1 year ago - 5 comments

#75 - updated the colon missing in readme

Pull Request - State: closed - Opened by venkatesh-pro over 1 year ago

#74 - fine-tune HKUNLP/instructor-embedding

Issue - State: open - Opened by Atlantic8 over 1 year ago - 7 comments

#73 - Kernel dies when calling HuggingFaceInstructEmbedding

Issue - State: closed - Opened by lavalosan over 1 year ago - 2 comments

#72 - what is the max input size?

Issue - State: closed - Opened by vyau over 1 year ago - 5 comments

#71 - Representing metadata and conditionally encoding text for a specific task

Issue - State: closed - Opened by bkamapantula over 1 year ago - 2 comments

#70 - Pip install does not include dependencies

Issue - State: closed - Opened by dstengle over 1 year ago - 2 comments

#69 - Fine-tuning for sentence comparison

Issue - State: open - Opened by Mukish45 over 1 year ago - 3 comments

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

Issue - State: closed - Opened by ScottishFold007 over 1 year ago - 2 comments

#67 - Evaluation settings of INSTRUCTOR

Issue - State: open - Opened by EliverQ over 1 year ago - 11 comments

#66 - Reproduction of training INSTRUCTOR

Issue - State: open - Opened by EliverQ over 1 year ago - 4 comments

#65 - gettin oom on train

Issue - State: closed - Opened by erlakshmi123 over 1 year ago - 1 comment

#64 - Different types of embedding used internally inside the tool?

Issue - State: closed - Opened by pradeepdev-1995 over 1 year ago - 2 comments

#63 - What if instruction+text merged?

Issue - State: closed - Opened by takiholadi over 1 year ago - 2 comments

#62 - domain adaptation method using GPL on instructor

Issue - State: closed - Opened by iavinas over 1 year ago - 3 comments

#61 - training with deepspeed got oom

Issue - State: closed - Opened by Costwen over 1 year ago

#60 - Quantization doesn't work on ARM devices

Issue - State: closed - Opened by samarthm over 1 year ago - 2 comments

#59 - Instruction guidance?

Issue - State: closed - Opened by oaustegard over 1 year ago - 4 comments

#58 - What kind of instruction would work well for prompt retrieval?

Issue - State: closed - Opened by timothylimyl over 1 year ago - 2 comments

#57 - Other model support.

Issue - State: closed - Opened by staticpunch over 1 year ago - 2 comments

#56 - special tokens in tokenizer

Issue - State: closed - Opened by erlakshmi123 over 1 year ago - 2 comments

#55 - Added quantization code

Pull Request - State: closed - Opened by RAravindDS over 1 year ago

#54 - Remove code that overrides max_seq_length

Pull Request - State: closed - Opened by michael-quinlan over 1 year ago - 1 comment

#53 - updated `README` to describe the `encode` function, and provided one example of using FAISS

Pull Request - State: closed - Opened by yuxiliu1995 over 1 year ago

#52 - docs: using the model with sagemaker

Issue - State: closed - Opened by greenpau over 1 year ago - 5 comments

#51 - Peculiar Cosine Similarity Values

Issue - State: closed - Opened by afoland over 1 year ago - 3 comments

#50 - show_progress_bar = True does not work properly in notebook

Issue - State: closed - Opened by rmovva over 1 year ago - 2 comments

#49 - file downloads on package import

Issue - State: closed - Opened by greenpau over 1 year ago - 8 comments

#48 - docs: instructions syntax

Issue - State: closed - Opened by greenpau over 1 year ago - 3 comments

#47 - Why use DataCollatorForSeq2Seq to collect data?

Issue - State: closed - Opened by tsotfsk over 1 year ago - 2 comments

#46 - Data & training details

Issue - State: closed - Opened by jordane95 over 1 year ago - 2 comments

#45 - Fine tuning: Some weights of the model checkpoint were not used when initializing T5EncoderModel

Issue - State: closed - Opened by cayden-uw over 1 year ago - 6 comments

#44 - Model Fine-tune for Classification Mission?

Issue - State: closed - Opened by MrRabbit12 over 1 year ago - 3 comments

#43 - How would you use instructor with DNA sequences?

Issue - State: closed - Opened by lborcard over 1 year ago - 8 comments

#42 - Replication of Instructor

Issue - State: open - Opened by aamir-s18 over 1 year ago - 15 comments

#41 - When does the training stop?

Issue - State: closed - Opened by qmin2 over 1 year ago - 2 comments

#40 - multilingual model

Issue - State: closed - Opened by carlesoctav over 1 year ago - 2 comments

#39 - Generating batch embeddings using hugging face datasets crashes my EC2 instance on XL model.

Issue - State: closed - Opened by bogedy over 1 year ago - 4 comments

#38 - Can i train it for legal domain based retrival task?

Issue - State: closed - Opened by addieindustry almost 2 years ago - 2 comments

#37 - GPU memory leak.

Issue - State: closed - Opened by nivibilla almost 2 years ago - 5 comments

#36 - Do other languages also work?

Issue - State: closed - Opened by FBosler almost 2 years ago - 4 comments

#35 - Is it possible to use instructor without instructions?

Issue - State: closed - Opened by jlia0 almost 2 years ago - 9 comments

#34 - What is the correct path for the cached model of the Instructor?

Issue - State: closed - Opened by thelongestusernameofall almost 2 years ago - 2 comments

#33 - Is it possible to use the Instructor model directly via sentence-transformers?

Issue - State: closed - Opened by doraui almost 2 years ago - 2 comments

#32 - Details about the Training Configurations

Issue - State: closed - Opened by Lancelot39 almost 2 years ago - 2 comments

#31 - Example for classification task

Issue - State: closed - Opened by jackiexue1993 almost 2 years ago - 2 comments

#30 - Can SetFit be incorporated InstructorEmbedding?

Issue - State: open - Opened by jackiexue1993 almost 2 years ago - 2 comments

GitHub / xlang-ai/instructor-embedding issues and pull requests