Ecosyste.ms: Issues
An open API service for providing issue and pull request metadata for open source projects.
GitHub / FlagOpen/FlagEmbedding issues and pull requests
#1375 - bge-m3,vllm和FlagEmbedding,同一query向量不一样
Issue -
State: open - Opened by hynnn 7 days ago
#1375 - bge-m3,vllm和FlagEmbedding,同一query向量不一样
Issue -
State: open - Opened by hynnn 7 days ago
#1374 - fix "\\n" issue: replace "\\n" with "\n"
Pull Request -
State: closed - Opened by hanhainebula 8 days ago
#1374 - fix "\\n" issue: replace "\\n" with "\n"
Pull Request -
State: closed - Opened by hanhainebula 8 days ago
#1374 - fix "\\n" issue: replace "\\n" with "\n"
Pull Request -
State: closed - Opened by hanhainebula 8 days ago
#1373 - bge-reranker-v2.5-gemma2-lightweight输出的分数如何归一化呢
Issue -
State: open - Opened by gaopeiyang 8 days ago
#1373 - bge-reranker-v2.5-gemma2-lightweight输出的分数如何归一化呢
Issue -
State: open - Opened by gaopeiyang 8 days ago
#1373 - bge-reranker-v2.5-gemma2-lightweight输出的分数如何归一化呢
Issue -
State: open - Opened by gaopeiyang 8 days ago
#1372 - 微调bge-reranker-v2-minicpm-layerwise,80G的A100卡,batchsize最大只能设置16吗
Issue -
State: closed - Opened by zhangying950 11 days ago
#1371 - 微调bge-reranker-v2-minicpm-layerwise,添加教师分数脚本用的compute score是不是应该跟推理原生模型时一样?
Issue -
State: open - Opened by zhangying950 11 days ago
#1370 - 微调阶段,fp16训练loss值为0,fp32训练loss正常
Issue -
State: open - Opened by zhangxiangchn 11 days ago
- 1 comment
#1369 - finetune bge-m3 效果差
Issue -
State: closed - Opened by hynnn 11 days ago
#1368 - How to evaluate bge-en-icl?
Issue -
State: open - Opened by puritysarah 11 days ago
#1367 - BGE-reranker用于图文检索重排
Issue -
State: open - Opened by liumingzhu6060 11 days ago
#1366 - optimize evaluation process in evaluator.py
Pull Request -
State: closed - Opened by hanhainebula 11 days ago
#1365 - AttributeError: module 'pyarrow.lib' has no attribute 'ListViewType'
Issue -
State: open - Opened by LinXin04 11 days ago
#1364 - Using scatter_reduce instead of scatter and max
Issue -
State: open - Opened by lsrock1 12 days ago
#1363 - Fix typos in LLARA
Pull Request -
State: open - Opened by whybe-choi 12 days ago
#1362 - 构造微调时使用的正负样本的疑问
Issue -
State: open - Opened by LawsonAbs 13 days ago
- 1 comment
#1361 - BGE-M3 Inference 多卡报错
Issue -
State: open - Opened by xcvil 14 days ago
- 1 comment
#1360 - del dp code
Pull Request -
State: closed - Opened by hanhainebula 14 days ago
#1359 - Fix a bug in beir evaluation and release v1.3.4
Pull Request -
State: closed - Opened by hanhainebula 14 days ago
#1358 - 进程退出 cleanup 时抛出异常
Issue -
State: open - Opened by patricksuo 14 days ago
- 4 comments
#1357 - Where is the code for 'A Multi-Task Embedder For Retrieval Augmented LLMs'?
Issue -
State: closed - Opened by Ask-sola 17 days ago
#1357 - Where is the code for 'A Multi-Task Embedder For Retrieval Augmented LLMs'?
Issue -
State: closed - Opened by Ask-sola 17 days ago
#1356 - cannot import name 'Gemma2FlashAttention2' from 'transformers.models.gemma2.modeling_gemma2'
Issue -
State: open - Opened by mxchinegod 20 days ago
- 2 comments
#1355 - docs: fix link to https://bge-model.com/ within NEWS section
Pull Request -
State: closed - Opened by bufferoverflow 25 days ago
- 1 comment
#1354 - pass kwargs into AutoModel.from_pretrained
Pull Request -
State: open - Opened by ayang 28 days ago
- 1 comment
#1353 - jina-embeddings-v2-base-zh 评测指标很低?显著低于排行榜数值
Issue -
State: closed - Opened by zhangxiangchn 29 days ago
#1352 - no_in_batch_neg_flag设置问题
Issue -
State: open - Opened by meu98 29 days ago
- 1 comment
#1351 - 更新transformers版本?
Issue -
State: open - Opened by Davidyz 29 days ago
- 1 comment
#1350 - support musa backend in FlagEmbedding
Pull Request -
State: closed - Opened by qiyulei-mt 30 days ago
- 4 comments
#1349 - Fix Bug: OOM
Pull Request -
State: closed - Opened by 545999961 about 1 month ago
#1348 - Problem with fine-tuning bge-reranker-v2-gemma
Issue -
State: open - Opened by Ask-sola about 1 month ago
- 1 comment
#1347 - 单卡训练
Issue -
State: open - Opened by JiaXinLI98 about 1 month ago
- 1 comment
#1346 - Finetune BGE-M3
Issue -
State: open - Opened by tenafrangelos about 1 month ago
- 2 comments
#1345 - Wrong similarity score for identical embeddings
Issue -
State: open - Opened by SaidKhudoyan about 1 month ago
- 1 comment
#1344 - how to get embedding with finetuned model[encoder-only]
Issue -
State: open - Opened by liulizuel about 1 month ago
- 1 comment
#1343 - fix transformers 4.48.0
Pull Request -
State: closed - Opened by Hypothesis-Z about 1 month ago
- 1 comment
#1342 - update embedder finetune code
Pull Request -
State: closed - Opened by hanhainebula about 1 month ago
#1341 - fix beir data_loader.py: dev -> validation
Pull Request -
State: closed - Opened by hanhainebula about 1 month ago
#1340 - Fix bugs
Pull Request -
State: closed - Opened by hanhainebula about 1 month ago
#1339 - Error occurred in hard negative examples mining
Issue -
State: open - Opened by LLLLucensus about 1 month ago
- 1 comment
#1338 - bge-multilingual-gemma2微调OOM
Issue -
State: open - Opened by LLLLucensus about 1 month ago
- 3 comments
#1337 - 想用一个模型,同时对相似查询和文档这两类数据进行排序,会不会出现给的相似问的分数比文档的分数高的情况啊?这种问题要怎么解决呢?
Issue -
State: open - Opened by ZHAOFEGNSHUN about 1 month ago
- 1 comment
#1336 - 使用bge-rerank-large对两个段落重排两次,两次重排前会对所有段落进行strip(),结果大不相同,为什么会有这么大的差异?
Issue -
State: open - Opened by yumingmin88 about 1 month ago
- 1 comment
#1335 - 用rerank来提供服务, cpu的占用率过高,开了线程和多work,占用率能到400%。如何在computer的计算得分的时候,完全使用gpu 或者降低cpu。谢谢
Issue -
State: open - Opened by hn529945497 about 1 month ago
#1334 - 想请问一下微调bge-reranker-v2-minicpm-layerwise模型过程中,所用的prompt是什么?在哪里看呀~
Issue -
State: open - Opened by ZHAOFEGNSHUN about 1 month ago
- 2 comments
#1333 - Is there a smaller-sized reranker model? Is it based on the secondary training of xlm-roberta trained by the Facebook team, or is it initialized from 0?
Issue -
State: open - Opened by oslijunw about 1 month ago
#1332 - FlagEmbedding/evaluation/air_bench/arguments.py 导入错误from air_benchmark import EvalArgs as AIRBenchEvalArgs
Issue -
State: open - Opened by Marverlises about 1 month ago
- 3 comments
#1331 - 微调mincpm-layerwise报错torch.distributed.elastic.multiprocessing.errors.ChildFailedError
Issue -
State: closed - Opened by ZHAOFEGNSHUN about 1 month ago
#1330 - fix a bug in icl/dataset.py
Pull Request -
State: closed - Opened by hanhainebula about 1 month ago
#1329 - Reranker微调好后,做推理时加载原始minicpm_layerwise报错,是否可以调用本地已下好的模型?
Issue -
State: open - Opened by zjx623 about 1 month ago
- 6 comments
#1328 - fix bugs for embedder finetune
Pull Request -
State: closed - Opened by hanhainebula about 1 month ago
#1327 - What data was used to train bge-en-v1.5 models?
Issue -
State: open - Opened by abhinand5 about 1 month ago
#1326 - Some configuration files are missing after finetuning bge-m3 model
Issue -
State: open - Opened by chansonzhang about 1 month ago
- 2 comments
#1326 - Some configuration files are missing after finetuning bge-m3 model
Issue -
State: open - Opened by chansonzhang about 1 month ago
- 1 comment
#1325 - GPU利用率不稳定问题
Issue -
State: open - Opened by xiayouran about 1 month ago
- 1 comment
#1325 - GPU利用率不稳定问题
Issue -
State: open - Opened by xiayouran about 1 month ago
#1324 - os.path.basename will get empty model_name if path is given end with "/"
Issue -
State: open - Opened by Taogreen about 1 month ago
#1323 - What's the difference between encode_corpus and encode?
Issue -
State: open - Opened by chansonzhang about 1 month ago
- 1 comment
#1323 - What's the difference between encode_corpus and encode?
Issue -
State: open - Opened by chansonzhang about 1 month ago
- 1 comment
#1322 - encoder类型的模型用的是cls进行相似度分的输出 decoder类型的模型也输出相似度分吗?是怎么输出的?
Issue -
State: open - Opened by ZHAOFEGNSHUN about 1 month ago
- 1 comment
#1322 - encoder类型的模型用的是cls进行相似度分的输出 decoder类型的模型也输出相似度分吗?是怎么输出的?
Issue -
State: open - Opened by ZHAOFEGNSHUN about 1 month ago
- 1 comment
#1321 - RuntimeError: Cannot re-initialize CUDA in forked subprocess. To use CUDA with multiprocessing, you must use the 'spawn' start method
Issue -
State: open - Opened by andrewzheng1618 about 1 month ago
- 1 comment
#1321 - RuntimeError: Cannot re-initialize CUDA in forked subprocess. To use CUDA with multiprocessing, you must use the 'spawn' start method
Issue -
State: open - Opened by andrewzheng1618 about 1 month ago
- 1 comment
#1320 - 用自己的评估集评估各类reranker模型的效果,发现recall指标完全相同,是评估集太简单了吗?
Issue -
State: open - Opened by ZHAOFEGNSHUN about 1 month ago
- 1 comment
#1319 - 关于bge-multilingual-gemma2的lora微调的显存问题
Issue -
State: open - Opened by sunyifeng6 about 1 month ago
- 3 comments
#1319 - 关于bge-multilingual-gemma2的lora微调的显存问题
Issue -
State: open - Opened by sunyifeng6 about 1 month ago
- 2 comments
#1318 - bge-reranker-v2.5-gemma2-lightweight计算分数不存在参数compress_layers和compress_ratio
Issue -
State: closed - Opened by ZHAOFEGNSHUN about 2 months ago
- 2 comments
#1317 - 使用Flag Embedding对gte-base-en-v1,5和 e5-mistral-7b-instruct评估时出现的问题
Issue -
State: closed - Opened by yzx220 about 2 months ago
- 2 comments
#1316 - ModenBERT Support
Issue -
State: open - Opened by prince14322 about 2 months ago
#1315 - BGE-M3无监督训练阶段所使用的时间以及训练了多少个epoch
Issue -
State: open - Opened by WangCC-77 about 2 months ago
#1314 - Fix torch.OutofMemoryError --> torch.OutOfMemoryError
Pull Request -
State: closed - Opened by mrdmnd about 2 months ago
- 2 comments
#1313 - fix/reranking tutorial typos
Pull Request -
State: closed - Opened by rendyfebry about 2 months ago
- 1 comment
#1312 - AutoModel produces different results with SentenceTransformer and FlagEmbedding
Issue -
State: closed - Opened by yuzhichang about 2 months ago
- 2 comments
#1311 - Update model_mapping.py
Pull Request -
State: closed - Opened by pengjunfeng11 about 2 months ago
- 1 comment
#1310 - FlagEmbedding因transformers==4.44.2不兼容,没法很好在同一个env下并用其他模型
Issue -
State: open - Opened by firezym about 2 months ago
- 2 comments
#1309 - 如何提升passage_max_len到8192来Finetune BGE-M3?
Issue -
State: open - Opened by Rilzob about 2 months ago
- 1 comment
#1308 - Performance for BGE-M3 inference dropped between 1.2.x and 1.3.x
Issue -
State: open - Opened by ivlcic about 2 months ago
#1307 - 想请问一下有单独评测reranker模型的代码吗?我看现在的评估方法是两阶段--先embedding模型检索之后再reranker模型排序,需要两个模型同时跑
Issue -
State: open - Opened by ZHAOFEGNSHUN about 2 months ago
- 2 comments
#1306 - encoder only 的 reranker 在微调以后如何添加 prompt 进行推理 ?
Issue -
State: open - Opened by wittenberg11 about 2 months ago
- 1 comment
#1305 - 想请问一下 这个评估代码是在哪里呀?还有就是reranker可以单独评测吗?如果可以单独评测,reranker评测数据集格式应该是什么样子的,需要包涵分数嘛?
Issue -
State: open - Opened by ZHAOFEGNSHUN about 2 months ago
- 1 comment
#1304 - FlagEmbedding/inference/reranker/decoder_only/models/gemma_model.py 存在版本问题
Issue -
State: open - Opened by moseshu about 2 months ago
- 3 comments
#1303 - 有没有一些api可以获取flagEmbedding支持的所有模型的列表?
Issue -
State: closed - Opened by pengjunfeng11 about 2 months ago
- 4 comments
#1302 - 微调bgem3 卡住不动了
Issue -
State: open - Opened by yichuxue about 2 months ago
#1301 - How to disable the progress bar
Issue -
State: open - Opened by LinXin04 2 months ago
- 2 comments
#1300 - [Bugfix] Typehint error on py38
Pull Request -
State: closed - Opened by DrDavidS 2 months ago
- 1 comment
#1299 - Python38下微调bge-reranker-v2-m3报错
Issue -
State: closed - Opened by DrDavidS 2 months ago
- 1 comment
#1298 - 模型转换
Issue -
State: closed - Opened by bjyfz 2 months ago
- 1 comment
#1297 - Cohere的rerank3.5很强啊,国内有没有可以对抗的开源模型
Issue -
State: open - Opened by datalee 2 months ago
- 2 comments
#1296 - update `C_MTEB` reference
Pull Request -
State: closed - Opened by emmanuel-ferdman 2 months ago
#1295 - 使用BGE-M3模型来对MLDR的英文数据集进行embedding很慢
Issue -
State: open - Opened by JackTan25 2 months ago
- 1 comment
#1294 - reranker加入in-batch?
Issue -
State: open - Opened by CarllllWang 2 months ago
- 1 comment
#1293 - Visualized_BGE推理时,显存一直增长不释放,最终OOM
Issue -
State: open - Opened by FoolishMao 2 months ago
#1292 - bge-multilingual-gemma2在多卡机器上部署报错
Issue -
State: open - Opened by Yao990x16 2 months ago
- 3 comments
#1291 - 这里的代码是不是有问题?deocder和cls模式都去的是序号0的embedding。
Issue -
State: open - Opened by iamreallyi9 2 months ago
#1290 - BGE-M3复现评估结果疑问
Issue -
State: open - Opened by WangCC-77 2 months ago
- 4 comments
#1289 - [Chore] Add example for ensembling multiple logits during inference time
Pull Request -
State: open - Opened by sigridjineth 2 months ago
#1288 - 预训练代码
Issue -
State: open - Opened by ymaoj 2 months ago
- 1 comment