4 Pull requests merged by 2 people

• feat: add merge base model with all adapter ckpts and update README

  #363 merged Jan 17, 2025

• Update utils.py

  #361 merged Jan 16, 2025

• add sft lora script and add LORA adapter configuration support

  #362 merged Jan 16, 2025

• feat: update lora config

  #359 merged Jan 15, 2025

14 Issues closed by 3 people

• Aider benchmark, DeepSeek-6.7B-Instruct model hardly generates SEARCH/REPLACE blocks, leading to very low pass rates

  #192 closed Feb 11, 2025

• 微调（dpo）的数据格式（json）能否给一个例子

  #224 closed Feb 11, 2025

• Train a model for a new language

  #200 closed Feb 11, 2025

• 什么时候支持多模态的代码大模型

  #218 closed Feb 11, 2025

• Qwen2.5-Coder-Instruct-C model release

  #215 closed Feb 11, 2025

• torch.cuda.OutOfMemoryError: CUDA out of memory

  #228 closed Feb 11, 2025

• Tareesh

  #208 closed Feb 11, 2025

• Facing issues in finetuning JSON data using qwen 2.5 1.5b coder instruct

  #229 closed Feb 11, 2025

• 代码补全功能无法正常使用。

  #342 closed Feb 11, 2025

• A consultation about code completion format: FIM (Fill In the Middle) + repository level information

  #343 closed Feb 11, 2025

• Is the ExecRepoBench dataset file missing?

  #346 closed Feb 11, 2025

• failed to install deepspeed==0.12.6+c00388a2 while finetuning using DPO method

  #347 closed Feb 11, 2025

• 2张V100部署，报错OOM

  #375 closed Feb 7, 2025

• vllm部署qwen2.5-coder-32b-instruct-AWQ 模型长上下文下（8k)输出结果不正确，且无法正常停止

  #354 closed Jan 13, 2025

7 Issues opened by 7 people

• Reproducibility issues with ExecRepoBench

  #374 opened Feb 4, 2025

• السلام عليكم

  #373 opened Feb 2, 2025

• License missing for this repo

  #367 opened Jan 22, 2025

• Inquiry About Prompt Format for qwen-coder-base Model in Text-to-SQL Tasks

  #366 opened Jan 21, 2025

• Other language support

  #365 opened Jan 19, 2025

• LM Studio加载qwen2.5-coder-14b-instruct-q4_0.gguf模型文件运行回复乱码

  #364 opened Jan 19, 2025

• 大家预训练或精调这个模型时，一般使用什么数据集呢？能分享下直接用于sft精调的如下格式的数据集吗？感谢！What datasets are generally used when pre-training or fine-tuning this model? Could you share a dataset in the following format that can be directly used for SFT fine-tuning? Thank you!

  #358 opened Jan 13, 2025

1 Unresolved conversation

Sometimes conversations happen on old items that aren’t yet closed. Here is a list of all the Issues and Pull Requests with unresolved conversations.

• Qwen2.5-Code-Instruct Performance without DPO

  #357 commented on Feb 11, 2025 • 0 new comments