Ecosyste.ms: Issues
An open API service for providing issue and pull request metadata for open source projects.
GitHub / deepseek-ai/deepseek-math issues and pull requests
#31 - Official fine-tuning code
Issue -
State: open - Opened by beichenzbc 27 days ago
#30 - minif2f-Isabella acc
Issue -
State: open - Opened by wangzhihao-coder 4 months ago
- 1 comment
#29 - Any Plan to release the code of GRPO?
Issue -
State: open - Opened by Viper403 4 months ago
- 1 comment
#28 - My environment is something wrong with flash-atten, can I drop it when finetune DeepSeek-Math?
Issue -
State: open - Opened by AceCHQ 4 months ago
#27 - Should we need to add "You are an AI assistant, developed by DeepSeek Company...." when further finetune MATH-7B-instruct?
Issue -
State: open - Opened by AceCHQ 4 months ago
#26 - GRPO as part of HF TRL?
Issue -
State: open - Opened by idobenshaul10 5 months ago
#25 - Why adding "hey\n" before model output staring with "```python"?
Issue -
State: open - Opened by tongyx361 5 months ago
#24 - Paper 第二节预训练 2.2 节:为什么对不同 size 的数据集都要训练至高达 150B tokens?
Issue -
State: open - Opened by yucc-leon 6 months ago
#23 - 关于sft阶段中数据拼接的问题
Issue -
State: open - Opened by SymbolZH 7 months ago
- 1 comment
#22 - Access to data set?
Issue -
State: open - Opened by brando90 7 months ago
- 2 comments
#21 - Unable to get evaluation results
Issue -
State: open - Opened by ViperVille007 7 months ago
- 1 comment
#20 - Are you planning to release the training dataset?
Issue -
State: open - Opened by Stefano-retinize 7 months ago
- 1 comment
#19 - RuntimeError: cutlassF: no kernel found to launch!
Issue -
State: open - Opened by BlackTea-c 7 months ago
#19 - RuntimeError: cutlassF: no kernel found to launch!
Issue -
State: open - Opened by BlackTea-c 7 months ago
#18 - Question about the way to extract text from CC HTML
Issue -
State: open - Opened by voladorlu 7 months ago
#17 - [fixed] the merging output is incorrect, when parallel_num=1
Pull Request -
State: open - Opened by Dylancer1998 8 months ago
#16 - apply_chat_template()报错,请问如何修改代码
Issue -
State: open - Opened by FreeYiran 8 months ago
- 1 comment
#15 - 数学中英语料占比
Issue -
State: open - Opened by youweihao-tal 8 months ago
#14 - how to sample 64 output from old policy model?
Issue -
State: open - Opened by mohhao 8 months ago
- 2 comments
#13 - Ask about the evaluation of deepseek-math-rl
Issue -
State: closed - Opened by ChengpengLi1003 9 months ago
- 2 comments
#12 - About raw common crawl data
Issue -
State: open - Opened by jordane95 9 months ago
#11 - SFT的数据分布
Issue -
State: open - Opened by cyzhh 9 months ago
- 1 comment
#10 - [Question] SFT Data Curation
Issue -
State: closed - Opened by choco9966 9 months ago
- 1 comment
#9 - 代码数据应该怎么用呢
Issue -
State: open - Opened by songge25 9 months ago
#8 - What is your chat template for huggingface chat ui?
Issue -
State: open - Opened by houghtonweihu 10 months ago
- 1 comment
#7 - Any plan to provide local Web UI like this: https://github.com/imoneoi/openchat?
Issue -
State: open - Opened by houghtonweihu 10 months ago
#6 - Add Replicate demo and API
Pull Request -
State: closed - Opened by chenxwh 10 months ago
#5 - Path Issue when running evals
Issue -
State: closed - Opened by yapdianang 10 months ago
- 2 comments
#4 - Publish on Ollama
Issue -
State: open - Opened by ThatOneCalculator 10 months ago
- 1 comment
#3 - MATH Test Score reproduce acc=43.6
Issue -
State: closed - Opened by GanjinZero 10 months ago
- 5 comments
#2 - 建议检查数据
Issue -
State: closed - Opened by hzwer 10 months ago
- 15 comments
#1 - Request to add SeaLLM-7B-v2 in your paper tables.
Issue -
State: closed - Opened by nxphi47 10 months ago
- 1 comment