Ecosyste.ms: Issues

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

GitHub / ymcui/Chinese-LLaMA-Alpaca issues and pull requests

#900 - 数据预处理合并文本到指定文本长度的问题

Issue - State: closed - Opened by Arcmoon-Hu 4 months ago - 2 comments
Labels: stale

#899 - 请问在哪下载您33B模型的全部权重，并且有无quick start？

Issue - State: closed - Opened by 141forever 4 months ago - 7 comments
Labels: stale

#422 - 回答的问题一直重复

Issue - State: closed - Opened by nicole828 over 1 year ago - 10 comments
Labels: stale

#390 - 输出内容后面会一直重复

Issue - State: closed - Opened by happy-zhangbo over 1 year ago - 10 comments
Labels: stale

#161 - Vocabulary expansion pre-training code for Chinese language

Issue - State: closed - Opened by abdoelsayed2016 over 1 year ago - 22 comments
Labels: stale

#102 - 怎么暴露成api啊 通过界面请求

Issue - State: closed - Opened by controZheng over 1 year ago - 5 comments
Labels: stale

#101 - 答案和输出示例的答案差距有点大

Issue - State: closed - Opened by hcl242 over 1 year ago - 9 comments

#100 - can not do batch generate atfer loading lora weights

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

#99 - CPU有点慢，有什么办法能在GPU上运行吗?

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

#98 - how to start with adapter_model.bin in Chinese-Alpaca-13B

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

#96 - llama-7b和Lora权重合并之后，在自己的数据上继续fine-tuning

Issue - State: closed - Opened by Tungsong over 1 year ago - 18 comments
Labels: stale

#95 - Add script to merge lora directly to hf format and works for latest peft

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

#94 - 本地部署可以做那种类似GPT4all的release吗？

Issue - State: closed - Opened by Gong1suwu over 1 year ago - 3 comments

#93 - 可以给一个不做量化直接推理的代码吗？

Issue - State: closed - Opened by puppet101 over 1 year ago - 1 comment

#92 - 重复“下面是一道例题”

Issue - State: closed - Opened by lingfengchencn over 1 year ago - 5 comments
Labels: stale

#91 - Size([49952, 5120]) from checkpoint V.S. Size([49953, 5120]). from model

Issue - State: closed - Opened by cArlIcon over 1 year ago - 5 comments
Labels: stale

#90 - 找不到params.json 文件，模型文件里面也没有这个文件呀

Issue - State: closed - Opened by Song367 over 1 year ago - 4 comments

#89 - 请教全量模型merged model不生成任何回复是怎么会事呢？

Issue - State: closed - Opened by mikeda100 over 1 year ago - 13 comments
Labels: stale

#86 - llama.cpp编译失败

Issue - State: closed - Opened by controZheng over 1 year ago - 9 comments
Labels: stale

#84 - X86 Linux服务器 运行非常缓慢

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

#83 - 合并lora_13b时出错

Issue - State: closed - Opened by zlht812 over 1 year ago - 9 comments

#82 - python 3.11 已经可以运行 llama.cpp

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

#80 - 指令精调

Issue - State: closed - Opened by tkone2018 over 1 year ago

#79 - 分享百度网盘下载7b/13b-hf的链接，有需要的请自取

Issue - State: closed - Opened by shmily91 over 1 year ago - 1 comment

#78 - windows4位量化

Issue - State: closed - Opened by ghost over 1 year ago - 5 comments
Labels: stale

#77 - Add steps to convert consolidated to Huggingface format

Pull Request - State: closed - Opened by wb14123 over 1 year ago

#76 - 教程有误

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

#75 - Fix text links.

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

#74 - fix typo

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

#73 - Update README.md

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

#72 - Self-instruction数据

Issue - State: closed - Opened by Lisennlp over 1 year ago - 2 comments
Labels: stale

#71 - 关于第一阶段预训练

Issue - State: closed - Opened by xxr11 over 1 year ago - 4 comments

#70 - 在合并模型的时候出现 Killed 输出

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

#68 - Add Chinese LLaMA/Alpaca 13B version

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

#67 - 合并模型的时候报错

Issue - State: closed - Opened by Log4J-Ops over 1 year ago - 7 comments

#65 - DeepSpeed和Lora的结合

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

#64 - 中文LLaMA是通过lora来训练的？

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

#63 - 中文二次预训练

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

#62 - 是否可以使输出 格式化？

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

#60 - 执行Q4量化时出现数组下标越界的情况

Issue - State: closed - Opened by GeorgeDong32 over 1 year ago - 1 comment

#59 - 请问int4量化后的模型可以进一步finetune么

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

#57 - 多轮对话训练的语料的问题

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

#56 - 最后一步执行报错是什么原因？

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

#55 - 请问运行Q4量化模型和ggml-FP16模型在效果上有何区别吗？

Issue - State: closed - Opened by 12lxr over 1 year ago - 9 comments

#54 - tokenizer的special token和原版的区别

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

#53 - 合并模型的时候报错

Issue - State: closed - Opened by Log4J-Ops over 1 year ago - 7 comments
Labels: stale

#52 - 请问如何实现chat的形式

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

#51 - win10的cmd输入不了中文吗？输入英文回答中文

Issue - State: closed - Opened by wscsjnhboy over 1 year ago - 6 comments

#48 - 同样的问题，为什么有时候是 英文回答，有时候是 中文

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

#47 - 关于load_in_8bit

Issue - State: closed - Opened by weilx2267 over 1 year ago - 1 comment

#46 - 请问该模型推理时使用的是CPU还是GPU？

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

#45 - 关于alpaca和lora

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

#44 - 模型的理解中文文档的能力

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

#43 - 如何继续fine-tuning

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

#41 - Prompt 最长可以有多少？

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

#40 - 模型合并过程中，step2没有生成 .pth文件

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

#38 - 精调耗时

Issue - State: closed - Opened by wl1320310292 over 1 year ago - 3 comments

#37 - 关于模型中lora结构

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

#36 - About the model output.

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

#35 - 预训练数据量

Issue - State: closed - Opened by lw3259111 over 1 year ago - 10 comments

#34 - Correct Chinese corpus error

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

#32 - 如何合并safetenors模型？

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

#31 - 中英互译效果怎么样呢？

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

#29 - 按照你这样处理以后, 能绕开license, 模型能用于商业化?

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

#28 - 尝试了用belle数据集在此版基础上进行微调

Issue - State: closed - Opened by 27182812 over 1 year ago

#27 - llama_model_load: tensor 'tok_embeddings.weight' has wrong size in model file

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

#26 - 建议

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

#24 - 复读机和胡言乱语情况

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

#22 - 本地部署运行起来之后中英文掺杂着输出

Issue - State: closed - Opened by gaoerrong over 1 year ago - 12 comments

#21 - 关于合并后的完整权重模型能否进行finetune

Issue - State: closed - Opened by fanxiangshangfen over 1 year ago - 6 comments

#20 - 无法复现

Issue - State: closed - Opened by howard0su over 1 year ago - 1 comment

#19 - 新版本中quantize.py被取消，更改了量化方式

Pull Request - State: closed - Opened by fireindark707 over 1 year ago - 1 comment
Labels: stale

#18 - DefaultCPUAllocator: can't allocate memory

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

#17 - 有和chatglm-6b做过对比吗

Issue - State: closed - Opened by zhongtao93 over 1 year ago - 1 comment

#16 - 关于词表扩充以及垂直应用微调的问题

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

#15 - Easy merge

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

#13 - Add backslash for python cmd

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

#12 - 全量模型(.pth)转huggingface参数模型遇到的问题和解决方法

Issue - State: closed - Opened by LeonG7 over 1 year ago - 1 comment

#11 - Windows本地部署经过量化的模型貌似无法理解中文Prompt?

Issue - State: closed - Opened by LainNya over 1 year ago - 37 comments

#10 - 再请教一下您对预训练阶段所起作用的见解

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

#9 - 请教，预训练阶段的具体细节

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

#8 - 报错：无法加载"LlamaForCausalLM"

Issue - State: closed - Opened by IloveTTLHz2023 over 1 year ago - 2 comments

#7 - Finetune时添加的prompts还是原版的英文的吗？

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