HKUDS/GraphGPT issues and pull requests

#64 - Issue in train_graph.py fail to load the model

Issue - State: closed - Opened by GoldyMoon 10 months ago - 2 comments

#63 - AttributeError: 'GraphLlamaConfig' object has no attribute 'pretrain_graph_model_path'

Issue - State: closed - Opened by AstroCIEL 10 months ago - 2 comments

#62 - graphgpt/train文件夹下不同的train_[*].py文件的作用？

Issue - State: open - Opened by tigerbunny2023 10 months ago - 1 comment

#61 - AttributeError: 'str' object has no attribute 'requires_grad_'和config.json missing报错

Issue - State: open - Opened by Melo-1017 10 months ago - 1 comment

#60 - 关于graph matching

Issue - State: open - Opened by zhuiyue233 10 months ago

#59 - ERROR: Could not build wheels for mpi4py, Pillow, which is required to install pyproject.toml-based projects

Issue - State: open - Opened by HelloWorld4747 10 months ago - 1 comment

#58 - out of memory

Issue - State: open - Opened by xxrrnn 10 months ago - 1 comment

#57 - AssertionError: config.json missing

Issue - State: closed - Opened by xxrrnn 10 months ago - 1 comment

#56 - No module named 'graphgpt'

Issue - State: closed - Opened by xxrrnn 10 months ago - 21 comments

#55 - CUDA error: invalid device ordinal when running graphgpt_stage1.sh

Issue - State: closed - Opened by MaoPopovich 10 months ago - 1 comment

#54 - CUDA error: invalid device ordinal

Issue - State: closed - Opened by zhuiyue233 10 months ago - 4 comments

#53 - Some questions about input embedding of graph tokens

Issue - State: closed - Opened by brothermaster 10 months ago - 3 comments

#52 - The results of baselines are not reasonable and not correct (Clarified by author)

Issue - State: closed - Opened by AGTSAAA 11 months ago - 13 comments

#51 - I can not find this file: ./arxiv_ti_ab.json

Issue - State: closed - Opened by MissyLee2018 11 months ago - 1 comment

#50 - A problem occurred when I tried to run stage1.sh

Issue - State: closed - Opened by Andarkness1 11 months ago - 4 comments

#49 - CUDA out of memory.

Issue - State: closed - Opened by daixixiwang 12 months ago

#48 - 这个预训练模型可以用来做图结构的下游预测任务么？有相关示例么？

Issue - State: closed - Opened by dlmty 12 months ago - 1 comment

#47 - ModuleNotFoundError: No module named 'graphgpt'

Issue - State: closed - Opened by lhkhiem28 12 months ago - 2 comments

#46 - 关于图编码器的问题

Issue - State: closed - Opened by serendipity800 about 1 year ago - 1 comment

#45 - The error occurred while installing the packages listed in requirements.txt.

Issue - State: closed - Opened by TrendBoyedc about 1 year ago - 2 comments

#44 - `graphgpt_stage1_lightning` 在第一个epoch训练结束后发生nccl超时错误。

Issue - State: closed - Opened by smurf-1119 about 1 year ago - 1 comment

#43 - bug in docs

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

#42 - Ask for the construction of datasets

Issue - State: closed - Opened by linwhitehat about 1 year ago - 4 comments

#41 - > It is a very nice work and inspires me a lot. How do you evaluate the predictions generated from LLM? The paper claims that evaluating with Acc and F1. However it can be hard to evaluate the text sometimes.

Issue - State: closed - Opened by linwhitehat about 1 year ago - 6 comments

#40 - About baseline codes

Issue - State: closed - Opened by W-rudder about 1 year ago - 2 comments

#39 - hugging face访问不了，能不能加一个备选的访问链接

Issue - State: closed - Opened by daixixiwang about 1 year ago - 2 comments

#38 - AttributeError: 'str' object has no attribute 'requires_grad_'

Issue - State: closed - Opened by zhihui-shao about 1 year ago - 7 comments

#37 - Extract the Trained Projector

Issue - State: closed - Opened by bamboo-boy about 1 year ago - 6 comments

#36 - How to calculate the Accuracy and Macro-f1 from the evaluation output?

Issue - State: closed - Opened by tyfeld about 1 year ago - 2 comments

#35 - 关于对比学习中的一些疑问

Issue - State: closed - Opened by Hongqiong12 about 1 year ago - 8 comments

#34 - A question about training weights of embedding

Issue - State: closed - Opened by AriKing11 about 1 year ago - 1 comment

#33 - Could this model used for other Graph Custom data Finetuning

Issue - State: closed - Opened by WEIYanbin1999 about 1 year ago - 9 comments

#32 - 关于OOM错误

Issue - State: closed - Opened by bamboo-boy about 1 year ago - 3 comments

#31 - 对比学习的一个细节问题

Issue - State: closed - Opened by serendipity800 about 1 year ago - 2 comments

#30 - About text feature

Issue - State: closed - Opened by W-rudder about 1 year ago - 2 comments

#29 - CUDA out of memory

Issue - State: closed - Opened by bamboo-boy about 1 year ago - 3 comments

#28 - 关于文本特征

Issue - State: closed - Opened by serendipity800 about 1 year ago - 1 comment

#27 - Performance of LLM baselines too low

Issue - State: closed - Opened by W-rudder about 1 year ago - 3 comments

#26 - 模型节点识别和关系推理能力

Issue - State: closed - Opened by Brook-L about 1 year ago - 2 comments

#25 - Can the GraphGPT be used for LLM inference on other domain graph without retrainng?

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

#24 - Would you share your code for generating prompting files?

Issue - State: closed - Opened by QiaoYRan about 1 year ago - 4 comments

#23 - Would you provide the modified config.json of 'vicuna-7b-v1.5'?

Issue - State: closed - Opened by xin-ran-w about 1 year ago - 1 comment

#22 - Why do you align GNN with a transformer training from scratch instead of LLaMA?

Issue - State: closed - Opened by yzhHoward about 1 year ago - 2 comments

#21 - model.get_graph_tower().to(dtype=torch.float16, device=training_args.device) AttributeError: 'str' object has no attribute 'to'

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

#20 - Missing the code for structure-text grounding on many graphs

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

#19 - Requesting for the code for evaluation

Issue - State: closed - Opened by yzhHoward about 1 year ago - 2 comments

#18 - where is ./arxiv_ti_ab.json

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

#17 - Error in Self-Supervised Instruction Tuning

Issue - State: closed - Opened by aiwen7 about 1 year ago - 3 comments

#16 - 关于运行graphgpt_eval.sh找不到模块的问题

Issue - State: closed - Opened by nulinuli about 1 year ago - 6 comments

#15 - 已解决

Issue - State: closed - Opened by nulinuli about 1 year ago

#14 - 关于找不到config.json文件

Issue - State: closed - Opened by nulinuli about 1 year ago

#13 - ImportError: libc10_cuda.so: cannot open shared object file: No such file or directory ImportError: ImportErrorlibc10_cuda.so: cannot open shared object file: No such file or directory: libc10_cuda.so: cannot open shared object file: No such file or directory

Issue - State: closed - Opened by kunupup about 1 year ago - 4 comments

#12 - solved now

Issue - State: closed - Opened by kunupup about 1 year ago - 3 comments

#11 - No module named 'fastchat'

Issue - State: closed - Opened by wangjade54241 about 1 year ago - 2 comments

#10 - docs: available path tune script misspell comments README

Pull Request - State: closed - Opened by guspan-tanadi about 1 year ago - 1 comment

#9 - The user requested protobuf==3.20.0 icetk 0.0.7 depends on protobuf<3.19

Issue - State: closed - Opened by kunupup about 1 year ago - 9 comments

#8 - Requesting for the code for structure-text grounding

Issue - State: closed - Opened by SpaceLearner about 1 year ago - 2 comments

#7 - pretrain_graph_model_path is not defined

Issue - State: closed - Opened by AriKing11 over 1 year ago - 8 comments

#6 - Why does the paper say that only the parameters of the alignment projector are optimized, but in fact the parameters tuned also include the input embedding of the language model?

GitHub / HKUDS/GraphGPT issues and pull requests