Ecosyste.ms: Issues

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

GitHub / stanleylsx/llms_tool issues and pull requests

#76 - sft train报错

Issue - State: open - Opened by langgege-cqu 8 months ago

#74 - SFT数据集字段

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

#73 - QLora似乎不能和deepspeed zero3一起使用？

Issue - State: closed - Opened by shaomai00 10 months ago - 1 comment

#72 - deepspeed报错

Issue - State: open - Opened by wq343580510 10 months ago

#71 - 设置QLoRA的target_modules为全部线性层

Pull Request - State: closed - Opened by Mxoder 11 months ago

#69 - 大佬，能尝试做训练用的WebUI吗?

Issue - State: closed - Opened by win10ogod 11 months ago - 1 comment

#68 - 指定扩充词表中加载模型的精度

Pull Request - State: closed - Opened by Mxoder 11 months ago

#67 - 关于模型预训练

Issue - State: closed - Opened by clclclaiggg 11 months ago - 2 comments

#66 - 解决中文加入不生效的问题

Pull Request - State: closed - Opened by tiandiweizun 11 months ago

#65 - 扩展词表代码需优化

Issue - State: closed - Opened by tiandiweizun 11 months ago

#64 - 支持词表扩充功能

Pull Request - State: closed - Opened by Mxoder 11 months ago - 2 comments

#63 - 笔误 + 删除重复实现

Pull Request - State: closed - Opened by tiandiweizun 11 months ago - 1 comment

#61 - 大神能建个微信群或者留个联系方式吗？

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

#60 - sft时不输出eval loss

Issue - State: closed - Opened by shaomai00 12 months ago - 3 comments

#59 - 期待預訓練代碼

Issue - State: closed - Opened by indiejoseph 12 months ago - 1 comment
Labels: enhancement

#58 - falcon的pad_token_id=eos_token_id

Pull Request - State: closed - Opened by stanleylsx about 1 year ago

#57 - 更新说明

Pull Request - State: closed - Opened by stanleylsx about 1 year ago

#56 - 支持使用Rope编码方式的Falcon使用ntk方式

Pull Request - State: closed - Opened by stanleylsx about 1 year ago

#55 - baichuan2的prompt_template配置fix

Pull Request - State: closed - Opened by stanleylsx about 1 year ago

#54 - 更新readme

Pull Request - State: closed - Opened by stanleylsx about 1 year ago

#53 - 支持Baichuan2

Pull Request - State: closed - Opened by stanleylsx about 1 year ago

#52 - fix

Pull Request - State: closed - Opened by stanleylsx about 1 year ago

#51 - 修改Readme

Pull Request - State: closed - Opened by stanleylsx about 1 year ago

#50 - 支持部分模型在推断的时候使用ntk方法

Pull Request - State: closed - Opened by stanleylsx about 1 year ago

#49 - 取消rwkv支持

Pull Request - State: closed - Opened by stanleylsx about 1 year ago

#48 - Develop

Pull Request - State: closed - Opened by stanleylsx about 1 year ago

#47 - 使用prefix-tuning微调Qwen模型时报错

Issue - State: closed - Opened by FelixZhang7 about 1 year ago - 1 comment

#46 - dpo训练异常

Issue - State: closed - Opened by tuzeao about 1 year ago - 16 comments

#45 - 只支持lora和adalora训练后的权重合并

Pull Request - State: closed - Opened by stanleylsx about 1 year ago

#44 - 只支持lora和adalora训练后的权重合并

Pull Request - State: closed - Opened by stanleylsx about 1 year ago

#43 - Develop

Pull Request - State: closed - Opened by stanleylsx about 1 year ago

#42 - 关于权重合并

Issue - State: closed - Opened by FelixZhang7 about 1 year ago - 1 comment

#41 - 关于权重合并

Issue - State: closed - Opened by FelixZhang7 about 1 year ago

#40 - 构造数据的说明

Pull Request - State: closed - Opened by stanleylsx about 1 year ago

#38 - fix

Pull Request - State: closed - Opened by stanleylsx about 1 year ago

#37 - 支持deepspeed训练

Pull Request - State: closed - Opened by stanleylsx about 1 year ago

#36 - 添加中文的配置说明

Pull Request - State: closed - Opened by stanleylsx about 1 year ago

#35 - 修改注释

Pull Request - State: closed - Opened by stanleylsx about 1 year ago

#34 - Develop

Pull Request - State: closed - Opened by stanleylsx about 1 year ago

#33 - 支持DPO的训练

Pull Request - State: closed - Opened by stanleylsx about 1 year ago

#32 - dpo强化学习

Pull Request - State: closed - Opened by stanleylsx about 1 year ago

#31 - Develop

Pull Request - State: closed - Opened by stanleylsx about 1 year ago

#30 - Develop

Pull Request - State: closed - Opened by stanleylsx about 1 year ago

#29 - Develop

Pull Request - State: closed - Opened by stanleylsx about 1 year ago

#28 - readme

Pull Request - State: closed - Opened by stanleylsx about 1 year ago

#27 - Develop

Pull Request - State: closed - Opened by stanleylsx about 1 year ago

#26 - 支持对各个模型的RLHF训练(ppo方法)

Pull Request - State: closed - Opened by stanleylsx about 1 year ago

#25 - 修改Readme

Pull Request - State: closed - Opened by stanleylsx about 1 year ago

#24 - Develop

Pull Request - State: closed - Opened by stanleylsx about 1 year ago

#23 - reward model训练报错

Pull Request - State: closed - Opened by stanleylsx about 1 year ago

#22 - demo config

Pull Request - State: closed - Opened by stanleylsx about 1 year ago

#21 - Develop

Pull Request - State: closed - Opened by stanleylsx about 1 year ago

#20 - 增加奖励模型的测试集例子

Pull Request - State: closed - Opened by stanleylsx about 1 year ago

#19 - 加载奖励模型进行批量训练

Pull Request - State: closed - Opened by stanleylsx about 1 year ago

#18 - 奖励模型批量测试

Pull Request - State: closed - Opened by stanleylsx about 1 year ago

#17 - unwrap_model

Pull Request - State: closed - Opened by stanleylsx about 1 year ago

#16 - bug-fix

Pull Request - State: closed - Opened by stanleylsx about 1 year ago

#15 - modify code

Pull Request - State: closed - Opened by stanleylsx about 1 year ago

#14 - Develop

Pull Request - State: closed - Opened by stanleylsx about 1 year ago

#13 - 解决qwen不停输出的问题

Pull Request - State: closed - Opened by stanleylsx about 1 year ago

#12 - bug-fix

Pull Request - State: closed - Opened by stanleylsx about 1 year ago

#11 - Develop

Pull Request - State: closed - Opened by stanleylsx about 1 year ago

#10 - Develop

Pull Request - State: closed - Opened by stanleylsx about 1 year ago

#9 - Develop

Pull Request - State: closed - Opened by stanleylsx about 1 year ago

#8 - 奖励模型训练完成

Pull Request - State: closed - Opened by stanleylsx about 1 year ago

#7 - 千问模型输出batch first

Pull Request - State: closed - Opened by stanleylsx about 1 year ago

#6 - aquila预测不中断问题

Pull Request - State: closed - Opened by stanleylsx about 1 year ago

#5 - 奖励模型训练70%

Pull Request - State: closed - Opened by stanleylsx about 1 year ago

#4 - Pls support RWKV world model

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

#3 - qwen tokenizer

Pull Request - State: closed - Opened by stanleylsx about 1 year ago

#2 - bug-fix

Pull Request - State: closed - Opened by stanleylsx about 1 year ago

#1 - 支持通义千问7B的web_demo及SFT微调。

Pull Request - State: closed - Opened by stanleylsx about 1 year ago