-
主题 berry 启用方式：启动的时候添加环境变量 THEME=berry 即可，未来会允许直接在后台改
感谢主题作者：https://github.com/MartialBE
欢迎在此提交相关问题。
---
2024年4月5日：berry 主题目前相比默认主题缺失的功能：
1. 令牌维度模型限制；
2. 令牌维度 ip 限制；
3. 渠道批量添加 key；
4. …
-
Bloom models have a more permissive license than llama models and are also multilingual in nature. While there is a project [based on llama.cpp](https://github.com/NouamaneTazi/bloomz.cpp) which can p…
-
-
I tried to use the load_8bit argument to try and train large models but it seems that the Trainer is not recognizing the PEFT adapters and is giving the following error:
`ValueError: You cannot perfo…
-
老师您好！关于您llm-action/llm-train/peft/clm/peft_prompt_tuning_clm.ipynb中 in4第一行提到的
tokenizer = AutoTokenizer.from_pretrained(model_name_or_path)
这部分代码运行一直报错Exception: expected value at line 1 column 1，一直…
-
I got this error when finetuning "EleutherAI/gpt-j-6B" using load_in_8bit and LoRA on 8×2080ti:
`RuntimeError: mat1 and mat2 shapes cannot be multiplied`
I'm using data parallelism and not using mod…
-
Can the project be run on a computer without a graphics card?
-
### System Info
peft: 0.4.0
### Who can help?
@pacman100 @younesbelkada @sayakpaul
### Information
- [X] The official example scripts
- [ ] My own modified scripts
### Tasks
- [X] An officially…
-
Solution should provide a script for prompt llama 7b with a given template prompt. The script should take in CLI arguments and be transferable to other models like Mistral and bloomz. See [[2307.16833…
owos updated
10 months ago
-
Hello all, and thank you for your great work!
Earlier this week, we announced [X-LoRA](https://arxiv.org/abs/2402.07148), a flexible MoE approach for LoRA adapters. We implement deep layer- and tok…