Ecosyste.ms: Issues
An open API service for providing issue and pull request metadata for open source projects.
GitHub / facebookresearch/DPR issues and pull requests
#251 - nanoDPR: simple replication of DPR
Issue -
State: open - Opened by Hannibal046 over 1 year ago
#250 - mismatch between encoded results and wiki passages
Issue -
State: open - Opened by Hannibal046 over 1 year ago
#249 - 'dpr_all_documents' is not defined
Issue -
State: open - Opened by golubovic over 1 year ago
- 2 comments
#248 - A possible bug: Wrong hard_neg_ctx_indices
Issue -
State: closed - Opened by zhiyuanpeng over 1 year ago
- 1 comment
#247 - Fix setup.py for resolving error
Pull Request -
State: open - Opened by naem1023 almost 2 years ago
Labels: CLA Signed
#246 - setup.py error during installation
Pull Request -
State: closed - Opened by naem1023 almost 2 years ago
- 1 comment
#245 - error about python dense_retriever.py
Issue -
State: open - Opened by CodingPeasantzgl almost 2 years ago
#244 - question: reader loss
Issue -
State: open - Opened by PaulLerner almost 2 years ago
#243 - Retriever does not train properly on a custom dataset with empty negative_ctxs and hard_negative_ctxs
Issue -
State: open - Opened by yu-rovikov about 2 years ago
- 2 comments
#242 - Having error while running train_extractive_reader.py
Issue -
State: open - Opened by tkabir1 about 2 years ago
- 1 comment
#241 - How to enter model_file in dense_retriever.py?
Issue -
State: open - Opened by tkabir1 about 2 years ago
- 1 comment
#240 - Training DPR with different language
Issue -
State: open - Opened by shahad2099 about 2 years ago
- 1 comment
#239 - Can we get wiki page url by passage id?
Issue -
State: open - Opened by heya5 about 2 years ago
#238 - About dataset generation
Issue -
State: open - Opened by Heisenberg-Yin about 2 years ago
#237 - Missing required positional arguments
Issue -
State: open - Opened by grossmanm over 2 years ago
- 6 comments
#236 - Unable to reproduce table 3 (Gold-127N) from scratch
Issue -
State: closed - Opened by xhluca over 2 years ago
- 1 comment
#235 - Make module installable using pip, even from GH
Pull Request -
State: open - Opened by syats over 2 years ago
Labels: CLA Signed
#234 - Explicitly added py_modules in setup.py
Pull Request -
State: closed - Opened by syats over 2 years ago
- 3 comments
Labels: CLA Signed
#233 - Passages with multiple answer spans
Issue -
State: open - Opened by TZeng20 over 2 years ago
#232 - `cosine_scores` defined in biencoder.py does not work
Issue -
State: open - Opened by xhluca over 2 years ago
#231 - Question in training reader model.
Issue -
State: open - Opened by gaishun over 2 years ago
- 1 comment
#230 - Doubt regarding all_gather_list in case of DDP
Issue -
State: open - Opened by bhattg over 2 years ago
#229 - Data from retriever is not working with reader
Issue -
State: open - Opened by UmerTariq1 over 2 years ago
- 1 comment
#228 - RuntimeError: Missing key(s) in state_dict: "encoder.embeddings.position_ids"
Issue -
State: open - Opened by NehaNishikant over 2 years ago
#227 - [Feature Request] Allowing unix-style pattern matching in `download_data.py`
Issue -
State: open - Opened by xhluca over 2 years ago
#226 - Fix `pip install .` by adding empty py_modules
Pull Request -
State: open - Opened by xhluca over 2 years ago
- 1 comment
Labels: CLA Signed
#225 - IndexError: Dimension out of range (expected to be in range of [-1, 0], but got 1)
Issue -
State: open - Opened by ZiluLii over 2 years ago
- 3 comments
#224 - How do the passage embeddings use the 'title' of the passage
Issue -
State: open - Opened by jigsaw2212 over 2 years ago
- 1 comment
#223 - Unable to install dependencies
Issue -
State: closed - Opened by jigsaw2212 over 2 years ago
- 2 comments
#222 - Question about Wikipedia Corpus Preprocessing
Issue -
State: open - Opened by manveertamber over 2 years ago
#221 - NQ Dataset
Issue -
State: open - Opened by varshakishore over 2 years ago
- 1 comment
#220 - Retriever Training Error with train_dense_encoder.py
Issue -
State: open - Opened by AIML-YOGI over 2 years ago
- 1 comment
#219 - Seek Wikipedia corresponding passage ID or dump
Issue -
State: closed - Opened by jzhoubu over 2 years ago
- 1 comment
#218 - Loss across the nodes....[multi-gpu training]
Issue -
State: closed - Opened by gabinguo over 2 years ago
- 2 comments
#217 - how to test the prediction?
Issue -
State: open - Opened by liuchiente over 2 years ago
#216 - How to get ctx_encoder from the checkpoint?
Issue -
State: closed - Opened by ali-abz almost 3 years ago
- 2 comments
#215 - Typo fix (Hugh -> high)
Pull Request -
State: open - Opened by LouisCastricato almost 3 years ago
- 4 comments
Labels: CLA Signed
#214 - error when python dense_retriever.py
Issue -
State: open - Opened by ztx313 almost 3 years ago
- 1 comment
#213 - IndexError: Dimension out of range
Issue -
State: open - Opened by mary-octavia almost 3 years ago
- 12 comments
#212 - Hard Negative Indices Issue
Issue -
State: open - Opened by khuangaf almost 3 years ago
- 3 comments
#211 - an error occurred in Retriever inference
Issue -
State: closed - Opened by juyoung228 almost 3 years ago
- 2 comments
#210 - Why does the tokenization part need to set the last token position to sep_token_id?
Issue -
State: open - Opened by kwang2049 almost 3 years ago
- 1 comment
#209 - About 'speech_data.py' file in dpr/data
Issue -
State: closed - Opened by xyltt almost 3 years ago
- 1 comment
#208 - Cannot fully reproduce the embeddings with the nq-single checkpoint
Issue -
State: open - Opened by kwang2049 almost 3 years ago
- 1 comment
#207 - Converting checkpoints to HuggingFace format
Issue -
State: closed - Opened by tanay2001 about 3 years ago
- 1 comment
#206 - Importerror
Issue -
State: closed - Opened by Facico about 3 years ago
- 2 comments
Labels: bug
#205 - Weird discontinuities in training
Issue -
State: closed - Opened by emilysilcock about 3 years ago
- 2 comments
#204 - Question about reader training
Issue -
State: closed - Opened by TZeng20 about 3 years ago
- 1 comment
#203 - Fix minor typo in download_data.py script
Pull Request -
State: closed - Opened by juyongjiang about 3 years ago
- 2 comments
Labels: CLA Signed
#202 - Question: Is there a way to reduce GPU's ram consumption?
Issue -
State: closed - Opened by ali-abz about 3 years ago
- 2 comments
#201 - Question on 'gradient_accumulation_steps'
Issue -
State: closed - Opened by jzhoubu over 3 years ago
- 2 comments
#200 - Question about answer quality quantification
Issue -
State: closed - Opened by churinga over 3 years ago
- 4 comments
#199 - Questions about the path of the dev files
Issue -
State: closed - Opened by unw9527 over 3 years ago
- 2 comments
#198 - TypeError: string indices must be integers
Issue -
State: closed - Opened by Suki-m730026038 over 3 years ago
- 1 comment
#197 - Fix: incorrect reference for gradient_acc_steps.
Pull Request -
State: closed - Opened by zake7749 over 3 years ago
- 1 comment
Labels: CLA Signed
#196 - about fix_ctx_encoder
Issue -
State: closed - Opened by wuyaoxuehun over 3 years ago
- 1 comment
#195 - Refactor into retriever
Pull Request -
State: closed - Opened by yyiilluu over 3 years ago
- 2 comments
#194 - jsonlines is missing as declared dependency in setup.py
Issue -
State: closed - Opened by jobergum over 3 years ago
- 2 comments
#193 - About the NQ model trained with mined hard negatives.
Issue -
State: closed - Opened by shunyuzh over 3 years ago
- 1 comment
#192 - Problem with multi-gpu training
Issue -
State: closed - Opened by jigsaw2212 over 3 years ago
- 4 comments
#191 - docs: fix typo in readme
Pull Request -
State: closed - Opened by l-yohai over 3 years ago
- 1 comment
Labels: CLA Signed
#190 - docs: fix typo in README
Pull Request -
State: closed - Opened by l-yohai over 3 years ago
- 1 comment
#189 - Retrieval time changes (dramatically) over trials
Issue -
State: closed - Opened by solomon-han over 3 years ago
- 1 comment
#188 - Siamese vs BiEncoder
Issue -
State: closed - Opened by zhiqihuang over 3 years ago
- 1 comment
#187 - fine-tune with hard negative ctxs
Issue -
State: closed - Opened by AndrewHYC over 3 years ago
- 2 comments
#186 - Seeking KILT meta-data for DPR
Issue -
State: closed - Opened by jzhoubu over 3 years ago
- 11 comments
#185 - Reproduce Table 3 in paper
Issue -
State: closed - Opened by yxliu-ntu over 3 years ago
- 9 comments
#184 - Question about position_embeddings
Issue -
State: closed - Opened by mug2mag over 3 years ago
- 1 comment
#183 - Support/Explanations for the usage of the file tables.py
Issue -
State: closed - Opened by estella98 over 3 years ago
- 1 comment
#182 - How to replicate the wiki preprocessing
Issue -
State: closed - Opened by kevinxyc1 over 3 years ago
- 1 comment
#181 - Possibly a bug on hard_neg_ctx_indices
Issue -
State: closed - Opened by thongnt99 over 3 years ago
- 1 comment
#180 - Generating Question's Dense Embeddings
Issue -
State: closed - Opened by hot-cheeto over 3 years ago
- 2 comments
#179 - problem with retriever training on xlm-roberta
Issue -
State: open - Opened by ffaisal93 over 3 years ago
- 3 comments
#178 - RuntimeError: Expected object of device type cuda but got device type cpu for argument #3 'index' in call to _th_index_select
Issue -
State: closed - Opened by XY2323819551 over 3 years ago
- 3 comments
#177 - Trivia Checkpoint Requirement
Issue -
State: closed - Opened by chengzhipanpan over 3 years ago
- 1 comment
#176 - Ambiguity regards hard_neg_ctx_indices in dpr/models/biencoder.py
Issue -
State: closed - Opened by ali-abz over 3 years ago
- 3 comments
#175 - Did you use regex during to calculate top-k retrieval ?
Issue -
State: closed - Opened by shamanez over 3 years ago
- 2 comments
#174 - When will the latest checkpoint be available on Huggingface?
Issue -
State: closed - Opened by tanmaylaud over 3 years ago
- 5 comments
#173 - A problem about the device of "biencoder_input" in "validate_average_rank()"
Issue -
State: open - Opened by yxliu-ntu over 3 years ago
- 2 comments
Labels: bug
#172 - questions about in-batch negatives in multi-GPU setting
Issue -
State: closed - Opened by menghuanlater over 3 years ago
- 2 comments
#171 - Generating dense embeddings in distributed mode
Issue -
State: closed - Opened by Yu-Shi over 3 years ago
- 3 comments
#170 - Can we fine tune on unsupervised data, i.e bunch of PDF's
Issue -
State: closed - Opened by SAIVENKATARAJU over 3 years ago
- 3 comments
#169 - BM25 retrieval result
Issue -
State: closed - Opened by jzhoubu over 3 years ago
- 1 comment
#168 - Steps for reproduce processed data for training
Issue -
State: closed - Opened by huvunvidia over 3 years ago
- 3 comments
#167 - omegaconf.errors.ConfigKeyError: Key 'p' is not in struct
Issue -
State: closed - Opened by yolsever over 3 years ago
- 4 comments
#166 - Generating representation vectors for the static documents dataset
Issue -
State: closed - Opened by cn-boop over 3 years ago
- 2 comments
#165 - Unexpected key(s) in state_dict: "embeddings.position_ids".
Issue -
State: closed - Opened by yolsever over 3 years ago
- 3 comments
#160 - IndexError: list index out of range
Issue -
State: closed - Opened by XY2323819551 over 3 years ago
- 3 comments
#155 - Can prebatch negatives be utilized to train dpr, and how?
Issue -
State: closed - Opened by wuyaoxuehun over 3 years ago
- 1 comment
#150 - Question about "Selecting a truncation strategy"
Issue -
State: closed - Opened by mug2mag over 3 years ago
- 1 comment
#147 - Can't find model 'en_core_web_sm'
Issue -
State: closed - Opened by mug2mag almost 4 years ago
- 3 comments
#132 - frustrated trying to test the new modification on colab
Issue -
State: closed - Opened by sadakmed almost 4 years ago
- 7 comments
Labels: enhancement
#127 - Minor suggestion on the trivia-train.json
Issue -
State: closed - Opened by alexlimh almost 4 years ago
- 13 comments
Labels: enhancement
#124 - Retriever Inference Seems to be Hanging on retriever.generate_question_vector
Issue -
State: closed - Opened by mosicr almost 4 years ago
- 9 comments
#123 - Cannot Reproduce SQuAD Retrieval Result
Issue -
State: closed - Opened by robinsongh381 almost 4 years ago
- 16 comments
#110 - How to use in-batch negative and gold when training?
Issue -
State: closed - Opened by hongyuntw almost 4 years ago
- 11 comments
#106 - Question about transformers version
Issue -
State: closed - Opened by avikpal00 almost 4 years ago
- 11 comments
Labels: enhancement
#99 - zero learning rate when resume training
Issue -
State: closed - Opened by jzhoubu about 4 years ago
- 6 comments
#91 - how to prepare customized training data for retriever
Issue -
State: closed - Opened by luomancs about 4 years ago
- 4 comments
#52 - dpr_all_documents is not defined
Issue -
State: closed - Opened by ramybaly over 4 years ago
- 7 comments