Ecosyste.ms: Issues

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

GitHub / CVI-SZU/Linly issues and pull requests

#100 - 7b模型性能和billa对比

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

#99 - readme上的加群二维码过期了

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

#98 - 如何cite？

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

#97 - 请问大佬65B的模型何时能够放出

Issue - State: open - Opened by Expert68 over 1 year ago

#96 - 关于33B模型预训练语料长度

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

#95 - 请问openllama 13b怎么转成HF格式

Issue - State: open - Opened by lin1490188 over 1 year ago

#93 - 关于平行语料的预处理

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

#92 - openllama性能评估

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

#90 - Chinese-LLaMA-33B (hf格式)的模型如何部署，进行推理？

Issue - State: open - Opened by xfg0913 over 1 year ago - 1 comment

#89 - 33b Huggingface 格式怎么转成TencentPretrain 格式

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

#88 - Is it possible to support OPT models

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

#87 - 博主群二维码过期了,可以更新一个新的二维码吗

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

#85 - 我在运行时，发生了一个报错。已经换了好几个cuda版本了

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

#84 - 如何在A100 40G上多卡部署 Chinese-LLaMA-33B？

Issue - State: open - Opened by xfg0913 over 1 year ago - 2 comments

#83 - CPU out of memory during initialization

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

#82 - 既然用了LLAMA, 多轮对话为什么不用 Alpaca ?

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

#79 - lora训练得到的lora权重如何融合原模型

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

#77 - 关于扩充词表再增量预训练的疑问

Issue - State: open - Opened by qiancheng99 over 1 year ago

#76 - 运行llama-7b微服务时，报错

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

#75 - 多轮对话回答结果的一些疑问

Issue - State: open - Opened by hepj987 over 1 year ago - 5 comments

#74 - 请问Linly-OpenLLaMA什么时候发布？

Issue - State: open - Opened by chk4991 over 1 year ago

#72 - 能不能测试一下这几个问题

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

#70 - Linly-ChatFlow-7B可以基于peft实现LoRA微调吗？

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

#69 - AttributeError: 'Namespace' object has no attribute 'emb_size'

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

#67 - 13B的模型下载为什么没有config文件和tokenizer.model文件？

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

#66 - huggingface online demo

Pull Request - State: closed - Opened by fengyh3 over 1 year ago

#64 - 在hugging face上下载的 33B 模型，怎么在llama.cpp上面运行？

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

#61 - zero3预训练相关问题

Issue - State: closed - Opened by wang99711123 over 1 year ago - 3 comments

#60 - 预训练输入文本处理注意事项

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

#59 - llama 编译错误

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

#58 - 二维码失效了，请问可以更新下不

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

#57 - 训练参数和时长

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

#56 - llama_7b_config.json在哪

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

#55 - 基础模型使用

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

#54 - which is faster ? int4 or int8?

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

#53 - fix typo in generate_chatllama.py

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

#51 - pretrain.py failed when deal with vocabulary special tokens

Issue - State: open - Opened by better629 over 1 year ago - 2 comments

#50 - chain of thought data link is wrong?

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

#49 - 能不能提供一份量化模型int4的方法或者脚本？

Issue - State: open - Opened by doctor1984 over 1 year ago - 5 comments

#48 - 33b模型权重少了五份

Issue - State: open - Opened by genggui001 over 1 year ago - 1 comment

#45 - bitsandbytes报错

Issue - State: open - Opened by yuanzhiyong1999 over 1 year ago - 1 comment

#44 - 腾讯格式的权重转换成HF格式的转换脚本在哪里？

Issue - State: open - Opened by riverzhou over 1 year ago - 12 comments

#43 - 'Parameter' object has no attribute 'CB' 使用int8 启动时报错

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

#42 - 13B大概需要多少显存？

Issue - State: closed - Opened by fireice009 over 1 year ago - 1 comment

#40 - 如何进行多轮问答的测试呢？有固定的格式要求么

Issue - State: open - Opened by heyboy905 over 1 year ago - 2 comments

#39 - 基础模型LLaMA-zh-7B 和 ChatLLaMA-zh-7B的区别是什么？

Issue - State: open - Opened by cangyi071 over 1 year ago - 1 comment

#38 - 指令微调readme的数据下载repo里没有数据，后期会上传吗？

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

#37 - chatllama_zh_13b.bin如何转成chatllama-ggml-q4_0.bin格式？谢谢

Issue - State: open - Opened by doctor1984 over 1 year ago - 1 comment

#35 - 这个预训练最低需要多大的显存哦，什么配置可以跑？

Issue - State: closed - Opened by ze00ro over 1 year ago - 4 comments

#34 - 请问ChatLLaMa需要GCC哪个版本？

Issue - State: open - Opened by Yesgo1220 over 1 year ago

#33 - 关于instances_buffer_size参数的问题

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

#32 - Is there a Baidu Cloud download link for the model?

Issue - State: closed - Opened by zay95 over 1 year ago

#31 - 指令微调中preprocess.py文件的位置

Issue - State: closed - Opened by yingweima2022 over 1 year ago

#30 - 使用zero3以后的模型加载问题

Issue - State: open - Opened by hepj987 over 1 year ago - 2 comments

#29 - 如何搭配 text-generation-webui 运行？

Issue - State: open - Opened by zms351 over 1 year ago - 3 comments

#28 - 增量预训练超参数设置

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

#27 - 基础模型 LLaMA_zh下载页面404

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

#26 - DeepSpeed ZeRO-3预训练

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

#25 - 'NoneType' object cannot be interpreted as an integer

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

#24 - generate_chatllama.py 这找不到

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

#23 - trainer里是否需要开启fp16才能达到混合精度目的？

Issue - State: closed - Opened by 15810856129 over 1 year ago - 1 comment

#22 - 运行7B模型32G的T4不够嘛，每次都被kill了

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

#20 - 词表扩充

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

#19 - llama 自己数据增量预训练数据格式

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

#18 - 中文预训练模型有开源的计划吗

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

#17 - 能不能提供 llama.cpp 的 FP16 格式的权重文件下载。

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

#16 - 中文增量预训练程序长时间没有输出

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

#14 - 怎么转换成hf的格式，我想转换成GPTQ的量化格式

Issue - State: open - Opened by edsky over 1 year ago - 1 comment
Labels: enhancement

#13 - 中文预训练

Issue - State: open - Opened by lw3259111 over 1 year ago

#12 - ChatLLaMA-zh-7B-int4模型加载报错问题

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

#10 - 领域语料增量预训练怎么做？

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

#9 - 推理速度比原版的llama要慢一些

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

#8 - 运行脚本generate_chatllama.py后，tokenizer报错

Issue - State: open - Opened by tianmala over 1 year ago - 10 comments

#7 - 训练时 gpu的耗时能透露一下吗

Issue - State: open - Opened by zhangxueren9 over 1 year ago - 8 comments

#6 - 生成结果与readme中不符（试过快十次）

Issue - State: open - Opened by pan-2001 over 1 year ago - 4 comments

#5 - 解码失败'BertTokenizer' object has no attribute 'decode'

Issue - State: closed - Opened by pan-2001 over 1 year ago - 1 comment

#4 - generate_chatllama.py 86行应该去掉True

Issue - State: closed - Opened by pan-2001 over 1 year ago - 1 comment

#3 - 按照 README.md 来根本跑不出预期的效果

Issue - State: closed - Opened by ToSev7en over 1 year ago - 11 comments

#1 - 直接加载ChatLLaMA-zh-7B模型失败

Issue - State: open - Opened by 15810856129 over 1 year ago - 12 comments