Ecosyste.ms: Issues
An open API service for providing issue and pull request metadata for open source projects.
GitHub / FlagAI-Open/FlagAI issues and pull requests
#570 - [Question]: Why not call optimizer.zero_grad() in train_step_pytorch and train_step_pytorchDDP function?
Issue -
State: open - Opened by THU-Kingmin 5 months ago
Labels: question
#569 - [Question]:ALtCLIp : local variable 'model_id' referenced before assignment
Issue -
State: open - Opened by molu-ggg 6 months ago
Labels: question
#568 - [Question]: TypeError: __init__() got multiple values for argument 'text_config'
Issue -
State: open - Opened by molu-ggg 6 months ago
Labels: question
#567 - Aquila2推理优化:请问Aquila2-34b在量化后还想使用单机多卡和多机多卡并行推理以提高效率,该如何实现呢?
Issue -
State: open - Opened by monkeyZhy 6 months ago
Labels: question
#566 - [Question]: 请问目前支持Langchain框架吗
Issue -
State: open - Opened by PlatinumDiors 6 months ago
Labels: question
#565 - Windows bug ModuleNotFoundError: No module named 'torch._six'
Issue -
State: open - Opened by MGzhou 8 months ago
Labels: bug
#564 - Update python-app.yml
Pull Request -
State: closed - Opened by zacliu2023 8 months ago
#563 - Update python-app.yml
Pull Request -
State: closed - Opened by zacliu2023 8 months ago
#562 - [Question]: alm的阿拉伯语有没有文本分类的例子?
Issue -
State: open - Opened by handywc 8 months ago
Labels: question
#561 - AltCLIP has no effect on CIFAR10 after finetune
Issue -
State: open - Opened by youyuge34 8 months ago
- 1 comment
Labels: question
#560 - Do you have this dataset?[Question]:
Issue -
State: closed - Opened by kopyl 9 months ago
Labels: question
#559 - [Question]: 请问AltCLIP有没有提供使用Flickr30k微调及评估的脚本
Issue -
State: open - Opened by qyliuAI 9 months ago
Labels: question
#558 - Update README.md
Pull Request -
State: open - Opened by eltociear 11 months ago
#557 - How does one obtain training data suitable for fine-tuning Aquila models?
Issue -
State: closed - Opened by Micla-SHL 11 months ago
- 2 comments
#556 - TypeError: AquilaPreTrainedModel._set_gradient_checkpointing() got an unexpected keyword argument 'enable'
Issue -
State: closed - Opened by Chenjingliang1 11 months ago
- 2 comments
Labels: bug
#555 - Can't run examples/vit_cifar100/train_deepspeed.py
Issue -
State: open - Opened by twmht 11 months ago
- 1 comment
Labels: bug
#554 - [Question]: 想问下AquilaDecoderLayer里的传参为什么是float的?这样是用的float类型的attention
Issue -
State: open - Opened by Chenjingliang1 11 months ago
- 2 comments
Labels: question
#553 - torchrun command not found
Issue -
State: closed - Opened by Micla-SHL 11 months ago
- 5 comments
#552 - [Question]: No module named 'packaging' This question is after install NVIDIA's apex
Issue -
State: closed - Opened by Micla-SHL 11 months ago
- 1 comment
Labels: question
#551 - fixed yaml setup issue
Pull Request -
State: closed - Opened by BAAI-OpenPlatform 12 months ago
#550 - Collecting PyYAML==5.4.1 (from flagai)报错
Issue -
State: closed - Opened by Darrenzeng 12 months ago
- 4 comments
#549 - 运行对话脚本aquilachat2-34b-16k 报错误
Issue -
State: closed - Opened by gaord 12 months ago
- 1 comment
Labels: bug
#548 - [Question]: 使用flagai AutoLoader加载AquilaChat2-7B-16K,输入英文 输出乱码
Issue -
State: closed - Opened by muzian666 12 months ago
- 2 comments
Labels: question
#547 - AquilaChat2 34B推理问题
Issue -
State: closed - Opened by ningpengtao-coder 12 months ago
- 1 comment
Labels: bug
#546 - Update conversation.py
Pull Request -
State: closed - Opened by Anhforth 12 months ago
#545 - updated version
Pull Request -
State: closed - Opened by Anhforth 12 months ago
#544 - changed aquila prediction
Pull Request -
State: closed - Opened by Anhforth 12 months ago
#543 - added quantize
Pull Request -
State: closed - Opened by Anhforth 12 months ago
#542 - add aquila sql
Pull Request -
State: closed - Opened by xiaofengShi 12 months ago
#541 - fixed bug
Pull Request -
State: closed - Opened by Anhforth 12 months ago
#540 - Added aquila2 finetuning
Pull Request -
State: closed - Opened by Anhforth 12 months ago
#539 - add flagai fit for aquila2
Pull Request -
State: closed - Opened by Anhforth about 1 year ago
#538 - [Question]: How to train the altclip model
Issue -
State: open - Opened by gybwyy-art about 1 year ago
Labels: question
#537 - add aquila2 modeling file
Pull Request -
State: closed - Opened by shunxing1234 about 1 year ago
#536 - [Question]: altclip直接运行example的inference.py报错KeyError: 'text_config'
Issue -
State: closed - Opened by susht3 about 1 year ago
- 2 comments
Labels: question
#535 - fix Update grads.py Torch does not support torch._six anymore
Pull Request -
State: open - Opened by fenglui about 1 year ago
#534 - [Question]: Model hub is not reachable!
Issue -
State: closed - Opened by susht3 about 1 year ago
- 4 comments
Labels: question
#533 - [Question]: UnboundLocalError: local variable 'model_id' referenced before assignment
Issue -
State: closed - Opened by susht3 about 1 year ago
Labels: question
#532 - [Question]: modehub中的altclip与huggingface上的altclip参数不一致
Issue -
State: closed - Opened by Jake-wei about 1 year ago
- 4 comments
Labels: question
#531 - [Question]: Lora微调
Issue -
State: closed - Opened by Bit-sjw about 1 year ago
- 2 comments
Labels: question
#530 - add aquila flagai2hf weight script
Pull Request -
State: closed - Opened by 920232796 about 1 year ago
#529 - Update docs
Pull Request -
State: closed - Opened by Anhforth about 1 year ago
#528 - Update docs
Pull Request -
State: closed - Opened by Anhforth about 1 year ago
#527 - 怎么加载altclip模型?
Issue -
State: open - Opened by susht3 about 1 year ago
- 6 comments
Labels: question
#526 - [Question]: 如何进行模型并行训练
Issue -
State: closed - Opened by kunden0612 about 1 year ago
- 2 comments
Labels: question
#525 - [Question]: 如何实现多语言tokenizer?
Issue -
State: closed - Opened by BaiqingL about 1 year ago
- 1 comment
Labels: question
#524 - updated docs
Pull Request -
State: closed - Opened by Anhforth about 1 year ago
#523 - Fit torch
Pull Request -
State: closed - Opened by Anhforth about 1 year ago
#522 - [Question]: A little question about the lm loss computation
Issue -
State: closed - Opened by nb003pl about 1 year ago
- 1 comment
Labels: question
#521 - fix messy code
Pull Request -
State: closed - Opened by 920232796 about 1 year ago
#520 - [Question]: 听说AquliaCode要在Aqulia语言模型基础上重新增量预训练,如何保证不会降低原语言模型的能力
Issue -
State: closed - Opened by kingcca about 1 year ago
- 2 comments
Labels: question
#519 - [Question]: Altclip inference, Why use inner product instead of cosine similarity?
Issue -
State: closed - Opened by MobiusDai about 1 year ago
- 4 comments
Labels: question
#518 - [Question]:预训练过程中,可以处理文本长度是多少
Issue -
State: closed - Opened by whl120901318 about 1 year ago
- 1 comment
Labels: question
#517 - Update docs
Pull Request -
State: closed - Opened by Anhforth about 1 year ago
#516 - fix messy code bug
Pull Request -
State: closed - Opened by 920232796 about 1 year ago
#515 - [Question]: 集成到vscode的具体指导文件请问在哪里
Issue -
State: closed - Opened by YIXIANREN about 1 year ago
- 2 comments
Labels: question
#514 - [Question]: 流式生成出现乱码
Issue -
State: closed - Opened by yinzhang809 about 1 year ago
- 3 comments
Labels: question
#513 - add aquila server huggingface
Pull Request -
State: closed - Opened by 920232796 about 1 year ago
#512 - [Question]: 天数GPU版本怎么没有了?模型训练后,推理或微调还区分GPU厂商吗?
Issue -
State: closed - Opened by kingcca about 1 year ago
- 3 comments
Labels: question
#511 - [Question]: aquila_pretrain.py: error: unrecognized arguments: --local-rank=1
Issue -
State: closed - Opened by caiweige about 1 year ago
- 3 comments
Labels: question
#510 - docs updated
Pull Request -
State: closed - Opened by Anhforth about 1 year ago
#509 - updated docs
Pull Request -
State: closed - Opened by Anhforth about 1 year ago
#508 - [Question]: 如何通过deepspeed+huggingface transformers的方式进行微调训练?
Issue -
State: closed - Opened by xgySTATISICT about 1 year ago
- 2 comments
Labels: question
#507 - [Question]: 断点续训出现错误
Issue -
State: closed - Opened by whl120901318 about 1 year ago
- 2 comments
Labels: question
#506 - [Question]: FlagAI\examples\AltCLIP>python altclip_inference.py 报错
Issue -
State: closed - Opened by panghongwei17 about 1 year ago
- 2 comments
Labels: question
#505 - fix readme
Pull Request -
State: closed - Opened by shunxing1234 about 1 year ago
#504 - fix readme
Pull Request -
State: closed - Opened by shunxing1234 about 1 year ago
#503 - [Question]:预训练的模型输出结果乱码
Issue -
State: closed - Opened by whl120901318 about 1 year ago
- 4 comments
Labels: question
#502 - 用自己的数据完成pretrain后,怎样运行pretrain 的model
Issue -
State: closed - Opened by whl120901318 about 1 year ago
- 3 comments
#501 - [Question]:I have finish pretrain,how run the myself pretrained model ?
Issue -
State: closed - Opened by whl120901318 about 1 year ago
- 1 comment
Labels: question
#500 - chage aquila readme & add aquila hf model
Pull Request -
State: closed - Opened by shunxing1234 about 1 year ago
#499 - updated docs for v0.8 of checkpoints
Pull Request -
State: closed - Opened by Anhforth about 1 year ago
#498 - [Question]: lora微调后出现输出重复和乱码问题
Issue -
State: closed - Opened by Mandy0016 about 1 year ago
- 4 comments
Labels: question
#497 - [Question]: 悟道·天鹰的最大上下文长度是多少?
Issue -
State: closed - Opened by ifsheldon about 1 year ago
- 2 comments
Labels: question
#496 - Update aquila_pretrain.py
Pull Request -
State: closed - Opened by Anhforth about 1 year ago
#495 - [Question]: 是否支持断点续训?
Issue -
State: closed - Opened by Mandy0016 about 1 year ago
- 3 comments
Labels: question
#494 - add aquilachat model for multigpu inference
Pull Request -
State: closed - Opened by 920232796 about 1 year ago
#493 - AquilaChat HF 版 tokenizer 没有带上 BOS, EOS
Issue -
State: closed - Opened by huntzhan about 1 year ago
- 4 comments
Labels: bug
#492 - [Question]:怎么使用deepspeed方式运行Aquila-pretrain,官方可以提供相关运行脚本吗
Issue -
State: closed - Opened by zt1556329495 about 1 year ago
- 2 comments
Labels: question
#491 - Fixed an accuracy calculation error in test_eval_gsm.py
Pull Request -
State: closed - Opened by NoahRe1 about 1 year ago
#490 - docs update for new model version
Pull Request -
State: closed - Opened by Anhforth about 1 year ago
#488 - [Question]: 跑AltDiffusion-m18模型推理时报错
Issue -
State: closed - Opened by smallhackeryifa about 1 year ago
- 1 comment
Labels: question
#486 - [Question]: ModuleNotFoundError: No module named 'flagai.model.tools.peft'
Issue -
State: closed - Opened by linsida1 about 1 year ago
- 3 comments
Labels: question
#485 - AttributeError: 'AutoLoader' object has no attribute 'model'
Issue -
State: closed - Opened by tdf1995 about 1 year ago
- 2 comments
Labels: bug
#484 - fixed deepspeed bugs
Pull Request -
State: closed - Opened by Anhforth about 1 year ago
#483 - changed requirements and fixed an issue in docs
Pull Request -
State: closed - Opened by Anhforth about 1 year ago
#482 - changed to cuda loading for Aquila base model
Pull Request -
State: closed - Opened by Anhforth about 1 year ago
#481 - name fix
Pull Request -
State: closed - Opened by Anhforth about 1 year ago
#480 - Add lora input
Pull Request -
State: closed - Opened by Anhforth about 1 year ago
#479 - [Question]: No module named 'megatron.data'
Issue -
State: closed - Opened by smallhackeryifa about 1 year ago
- 4 comments
Labels: question
#478 - [Question]: 模型加载速度慢
Issue -
State: closed - Opened by ningpengtao-coder about 1 year ago
- 3 comments
Labels: question
#477 - [Question]: A100 执行 python generate.py 特别慢,模型用的aquila-7b
Issue -
State: closed - Opened by smallhackeryifa about 1 year ago
Labels: question
#476 - [Question]: 什么时候开放33b的模型了
Issue -
State: closed - Opened by litetoooooom about 1 year ago
- 5 comments
Labels: question
#475 - [Question]: 训练LORA后模型输出错误
Issue -
State: closed - Opened by NoahRe1 about 1 year ago
- 2 comments
Labels: question
#474 - [Question]: 天鹰,是否支持windows单卡和centos系统?如何先下载模型到本地,加载本地模型?
Issue -
State: closed - Opened by chengzhen123 about 1 year ago
- 2 comments
Labels: question
#473 - 为什么按照官方提供方法做aquilachat-7b lora微调,training loss不收敛
Issue -
State: closed - Opened by ylhou about 1 year ago
- 2 comments
Labels: question
#472 - 提供的dockerfile无法打镜像
Issue -
State: closed - Opened by ToviHe about 1 year ago
- 1 comment
Labels: bug
#471 - Update env_args.py
Pull Request -
State: closed - Opened by Anhforth about 1 year ago
#470 - modified an argument type
Pull Request -
State: closed - Opened by Anhforth about 1 year ago
#469 - add filter for model output
Pull Request -
State: closed - Opened by 920232796 about 1 year ago