Ecosyste.ms: Issues

An open API service for providing issue and pull request metadata for open source projects.

GitHub / mymusise/ChatGLM-Tuning issues and pull requests

#85 - 'Parameter' object has no attribute 'named_children'

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

#84 - 怎么使用多卡进行训练？

Issue - State: open - Opened by 1a2cjitenfei over 1 year ago - 22 comments

#83 - support tensorboard

Pull Request - State: closed - Opened by xxw1995 over 1 year ago - 1 comment

#82 - support tensorboard for train

Pull Request - State: closed - Opened by xxw1995 over 1 year ago

#81 - rm CastOutputToFloat when finetuning

Pull Request - State: open - Opened by maybeluo over 1 year ago - 1 comment

#80 - 请问怎样在finetune好的模型上进行量化呀

Issue - State: open - Opened by aiaiyueq11 over 1 year ago

#79 - finetune数据样本不需要增加[MASK]标记吗？

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

#78 - 有没有在代码上解除chatglm限制的办法

Issue - State: open - Opened by YIZXIY over 1 year ago - 8 comments

#77 - 怎么使用训练好的lora做int8推理？

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

#76 - peft最高直到0.2.0版本，请问0.3.0.dev0版本怎么安装啊？

Issue - State: open - Opened by rayvzn119 over 1 year ago - 4 comments

#75 - fp参数和load_in_8bit都设置False后出现以下报错，V100环境

Issue - State: open - Opened by qingzhan0511 over 1 year ago - 9 comments

#74 - 做了中文Belle的实验

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

#73 - Fix eos_token_id error

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

#72 - 微调训练到16%，OOM

Issue - State: closed - Opened by yxk9810 over 1 year ago - 3 comments

#71 - 微调需要最少得资源是多少啊?

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

#70 - 试了下中文instrcut数据微调，效果好像还行

Issue - State: open - Opened by 27182812 over 1 year ago - 14 comments

#69 - Fix save_model

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

#68 - infer finetune完推理输出为空

Issue - State: open - Opened by nuass over 1 year ago - 9 comments

#67 - Fix eos_token_id error

Pull Request - State: closed - Opened by ypwhs over 1 year ago - 3 comments

#66 - 新版的finetune报错：KeyError: 'seq_len'

Issue - State: closed - Opened by baibaiw5 over 1 year ago - 3 comments

#65 - code about RewardMode and PPO

Issue - State: open - Opened by zhangsanfeng86 over 1 year ago - 1 comment

#64 - 想问下这个在中文小数据集上finetune后效果怎么样？

Issue - State: closed - Opened by rayvzn119 over 1 year ago - 11 comments

#63 - 关于训练数据

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

#62 - 训练完以后不会说中文了

Issue - State: closed - Opened by szpnygo over 1 year ago - 14 comments

#61 - 如何指定GPU进行训练？

Issue - State: open - Opened by dsh54054 over 1 year ago - 2 comments

#60 - label attention_mask构造问题

Issue - State: closed - Opened by JaheimLee over 1 year ago - 15 comments

#58 - 基于ChatGLM-6B-INT4的模型进行fineturning应该怎么做？

Issue - State: open - Opened by genius0182 over 1 year ago - 1 comment

#57 - 按步骤运行报错，RuntimeError: expected scalar type Half but found Float

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

#55 - 代码基本已经验证成功了，但会输出无关问题

Issue - State: open - Opened by archwolf118 over 1 year ago - 8 comments

#54 - hotfix max_seq_length && add skip_overlength

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

#53 - RuntimeError: GET was unable to find an engine to execute this computation

Issue - State: open - Opened by zhou888888 over 1 year ago - 5 comments

#52 - 24G显存oom了

Issue - State: open - Opened by hjyMM2018 over 1 year ago - 5 comments

#51 - Merge LoRa with original weights

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

#50 - infer和cover_alpaca2jsonl的数据格式不一致

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

#49 - key error 'seq_len'

Issue - State: closed - Opened by superhg over 1 year ago - 9 comments

#47 - Update tokenize_dataset_rows.py

Pull Request - State: closed - Opened by kinghuin over 1 year ago - 2 comments

#45 - 是否考虑多卡并行训练？

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

#44 - 无法正确生成 eos token

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

#43 - 多轮对话如何处理input和output数据呢？

Issue - State: open - Opened by js0316 over 1 year ago - 5 comments

#40 - 环境配置问题

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

#39 - 由于代码更新后不涉及此问题，所以关闭

Issue - State: closed - Opened by archwolf118 over 1 year ago

#38 - 3090训练时报显存溢出

Issue - State: closed - Opened by wxh0 over 1 year ago - 2 comments

#37 - Sync official code

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

#36 - 加载模型该用哪个pt文件

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

#35 - bug: _IncompatibleKeys(missing_keys:.......................

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

#34 - 在colab上报错ModuleNotFoundError: No module named 'modeling_chatglm'

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

#33 - 问一下大家，设备和环境问题，no kernel image is available

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

#32 - NameError: name 'init_empty_weights' is not defined

Issue - State: open - Opened by xiamaozi11 over 1 year ago - 2 comments

#31 - output中存下的checkpoint怎么用

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

#30 - update mask_position & position_ids

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

#29 - jsonl文件如何处理获得

Issue - State: open - Opened by Mr-lonely0 over 1 year ago - 1 comment

#28 - infer的时候使用int8报错

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

#27 - 请问这个项目是Lora，还是原始结构的finetune?

Issue - State: open - Opened by lulia0228 over 1 year ago - 3 comments

#26 - 必须finetune后infer？

Issue - State: open - Opened by June-Li over 1 year ago - 2 comments

#25 - ValueError: Please specify `target_modules` in `peft_config`

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

#24 - bs设置为1和3的时候，loss都是0， 是哪里有问题吗？

Issue - State: open - Opened by phantommlin over 1 year ago - 14 comments

#22 - 多卡执行finetune报错

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

#21 - infer报错

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

#20 - infer 时报错

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

#19 - 大概5小时可以训练完，但是loss一直是0，是正常的吗

Issue - State: closed - Opened by TccccD over 1 year ago - 14 comments

#18 - 感谢所有贡献者，希望共享微调模型

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

#17 - Where is define "VOCAB_FILES_NAMES" in tokenization_chatglm.py

Issue - State: closed - Opened by zhangsanfeng86 over 1 year ago - 2 comments

#16 - finetune完之后又进行效果评估吗？

Issue - State: open - Opened by zhhongzhi over 1 year ago - 6 comments

#15 - 在推理时的temperature参数似乎无效

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

#14 - 如何在上一次的结果上继续训练

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

#13 - 关于微调代码的疑问

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

#12 - 有没有可用于指令微调的中文数据集？

Issue - State: open - Opened by carbonz0 over 1 year ago - 5 comments

#11 - finetune指定--per_device_train_batch_size 大于1时报错

Issue - State: closed - Opened by cokuehuang over 1 year ago - 15 comments

#10 - 运行finetune.py时报错

Issue - State: open - Opened by zxcvbn114514 over 1 year ago - 2 comments

#9 - Fix inhomogeneous array

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

#8 - 数据预处理报错

Issue - State: closed - Opened by MEIZU16 over 1 year ago - 3 comments

#7 - requirements问题

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

#6 - 预处理数据集时报错

Issue - State: open - Opened by zxcvbn114514 over 1 year ago - 6 comments

#5 - infer报错

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

#4 - 问答数据集

Issue - State: closed - Opened by huangtao36 over 1 year ago - 3 comments

#3 - 加载预训练报错

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

#2 - perf安装

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

#1 - no attribute 'enable_input_require_grads'

Issue - State: closed - Opened by huangtao36 over 1 year ago - 10 comments