Ecosyste.ms: Issues

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

GitHub / magpie-align/magpie issues and pull requests

#38 - Great job! Where is the default config to generate CoT data?

Issue - State: closed - Opened by eyuansu62 about 1 month ago - 3 comments

#37 - configs in the paper

Issue - State: open - Opened by KimJaehee0725 about 1 month ago - 4 comments

#36 - How to select DPO subset?

Issue - State: open - Opened by qychen2001 about 2 months ago - 11 comments

#35 - 是否支持商用api

Issue - State: closed - Opened by xxlxms 3 months ago - 1 comment

#34 - qwen2.5-math系列模型生成问题

Issue - State: closed - Opened by Moyhub 3 months ago - 2 comments

#33 - Script for Generating Magpie DPO Data

Issue - State: closed - Opened by slark-prime 3 months ago - 1 comment

#32 - Wrap gen_ins.py under if __name__ == "__main__":

Pull Request - State: closed - Opened by ikergarcia1996 4 months ago - 1 comment

#31 - function call data

Issue - State: open - Opened by Watebear 4 months ago - 2 comments

#30 - fix links for Qwen2.5-72B

Pull Request - State: closed - Opened by huliang2016 4 months ago - 1 comment

#27 - Can this approach generate data for multi-turn conversations?

Issue - State: closed - Opened by Tendo33 5 months ago - 3 comments

#26 - Improve distance threshold check readability

Pull Request - State: closed - Opened by mandlinsarah 5 months ago

#25 - 单条训练数据超过了token的限制

Issue - State: closed - Opened by rogerslh 6 months ago - 3 comments

#24 - KeyError: 'uuid' during filtering

Issue - State: closed - Opened by rickyang1114 6 months ago - 1 comment

#23 - Are there any duplicates in these datasets?

Issue - State: closed - Opened by Tendo33 6 months ago - 4 comments

#22 - qwen2-math 有计划支持吗？

Issue - State: closed - Opened by yucheng-zeng 6 months ago - 3 comments

#21 - Can I extend 3rd, 4th turns using pre-generated magpie-mt dataset?

Issue - State: closed - Opened by hyunwoongko 6 months ago - 1 comment

#20 - 生成dis以后执行过滤，data_filter.ipynb中的uuid字段存在bug

Issue - State: closed - Opened by FlyCarrot 6 months ago - 1 comment

#18 - Update model_configs.json

Pull Request - State: closed - Opened by andresuribe87 6 months ago - 1 comment

#17 - 如何使用多卡生成数据？

Issue - State: closed - Opened by blackblue9 6 months ago - 8 comments

#16 - gen_res.py代码问题

Issue - State: closed - Opened by blackblue9 6 months ago

#15 - Question About Using API version

Issue - State: closed - Opened by DylanDDeng 7 months ago - 3 comments

#14 - 关于magpie微调的一个疑惑

Issue - State: closed - Opened by rangehow 7 months ago - 10 comments

#13 - Have you thought about generating the SFT data for Deepseek-Coder-V2?

Issue - State: open - Opened by tangzhy 7 months ago - 2 comments

#11 - The evaluation of input_quality

Issue - State: closed - Opened by wwwadx 7 months ago - 1 comment

#10 - 为何可以产生prompts

Issue - State: open - Opened by lyravv 7 months ago - 6 comments

#9 - Gemma OOM

Issue - State: closed - Opened by choco9966 7 months ago - 7 comments

#8 - Use online LLM inference

Issue - State: closed - Opened by mkaratayev 7 months ago - 1 comment

#7 - language support

Issue - State: open - Opened by sazzadhrz 8 months ago - 5 comments

#6 - I use qwen2-7b to generate instructions and responses, but error. thanks!

Issue - State: closed - Opened by 20184490 8 months ago - 3 comments

#5 - How to guide the LLM to generate a specific instruction-response pair?

Issue - State: closed - Opened by AliButtar 8 months ago - 1 comment

#4 - multi-turn support

Issue - State: closed - Opened by ZhaoKangkang0572 8 months ago - 1 comment

#3 - how to exstract qwen72B chinese instruction?

Issue - State: closed - Opened by moseshu 8 months ago - 2 comments

#2 - More details about controling LLMs to generate instructions

Issue - State: closed - Opened by DeyangKong 8 months ago - 1 comment

#1 - Question about Step1 and Step2 being separated steps.

Issue - State: closed - Opened by xinyazha92 8 months ago - 2 comments