Comments (2)
您好,确实目前avoid injection attacks主要体现在我们对chatml格式中特殊分隔token(im_start、im_end)的处理机制上,这部分逻辑是预训练涉及不到的,应该也不会对目前的预训练生成造成干扰。对于您提出的具体问题,我们已经去除相关描述哈。
from qwen.
嗯,预训练生成没影响。可能有一点点🤏不方便的地方在于,如果想使用 chatml 格式对 Qwen/Qwen-7B
进行 SFT 训练,在有 avoid injection attacks 的情况下,为了让 <im_start>
是一个 token,训练代码中就要加上就要手动拼 token_ids 的代码,而不能直接把拼接好的完整 prompt 送给 tokenizer。不过问题不大,了解了机制训练代码对应改一下就好了
from qwen.
Related Issues (20)
- [BUG] <title> wrong system prompt check? HOT 2
- [BUG] <title>batch_infer报错:'tuple' object has no attribute 'dtype' HOT 2
- 如何添加`LogitsProcessor`控制结果输出?
- [BUG] <title>lora微调loss异常? HOT 5
- tokenizer.decoder 抛出'utf-8' codec can't decode bytes in position 1-2: unexpected end of data异常 HOT 2
- [BUG] lora微调后,合并成一个模型。这种方式如何加载且推理 HOT 4
- [BUG] Qwen/Qwen-72B-Chat-Int8,不能多GPU并行计算 HOT 1
- Qwen/eval中的评测CEval和CMMLU,开大推理的batchsize评测指标会显著降低 HOT 1
- 请问基于qwen-72b-chat,基于怎样的配置可以在一台4090上训练起来? HOT 4
- 💡 [REQUEST] - <title> 关于lora 模型合并的几个问题 HOT 3
- [BUG] <关于model.generate时发现的源码错误> HOT 2
- [BUG] <Qwen-14B-Chat 输入长文本时无输出结果> HOT 5
- [BUG] Function Calling 示例有错误,最新的 openai sdk 运行时提示 api 已经废弃 HOT 1
- 请问哪里可以找到qwen用于vllm的jinja template? HOT 1
- [BUG] <title>执行eval中的eval_plugin进行评测 有一个agent从huggingface_hub拉包错误 HOT 1
- 请问可以使用高通的npu进行部署和推理吗? HOT 1
- 微调完成后使用llama_factory的vllm和qwen官方的vllm部署方式启动返回的不一样 HOT 1
- 💡 [REQUEST] - <使用ollama来调用qwen:14B时,怎么设置输出文本长度呢> HOT 1
- [BUG] <title>fastchat + vLLM +OpenAI API 调用qwen模型,数据不需要预先处理吗 HOT 1
- 本地部署后,运行很慢啊 HOT 3
Recommend Projects
-
React
A declarative, efficient, and flexible JavaScript library for building user interfaces.
-
Vue.js
🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
-
Typescript
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
-
TensorFlow
An Open Source Machine Learning Framework for Everyone
-
Django
The Web framework for perfectionists with deadlines.
-
Laravel
A PHP framework for web artisans
-
D3
Bring data to life with SVG, Canvas and HTML. 📊📈🎉
-
Recommend Topics
-
javascript
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
-
web
Some thing interesting about web. New door for the world.
-
server
A server is a program made to process requests and deliver data to clients.
-
Machine learning
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
-
Visualization
Some thing interesting about visualization, use data art
-
Game
Some thing interesting about game, make everyone happy.
Recommend Org
-
Facebook
We are working to build community through open source technology. NB: members must have two-factor auth.
-
Microsoft
Open source projects and samples from Microsoft.
-
Google
Google ❤️ Open Source for everyone.
-
Alibaba
Alibaba Open Source for everyone
-
D3
Data-Driven Documents codes.
-
Tencent
China tencent open source team.
from qwen.