psp_dada

第二版电子书-第二版序-勘误中 [issue列表](https://github.com/flutterchina/flutter-in-action-2nd/issues) 的链接错了

将 clash-verge 的链接换成当前仍在维护的仓库

[Bug]: NCCL get stuck when instantiating the LLM class.

5

### Your current environment The output of `python collect_env.py` ```text Collecting environment information... PyTorch version: 2.4.0+cu121 Is debug build: False CUDA used to build PyTorch: 12.1 ROCM used to build...

[Bug]: PyNcclCommunicator error when inferencing

3

### Your current environment The output of `python collect_env.py` ```text Collecting environment information... PyTorch version: 2.4.0 Is debug build: False CUDA used to build PyTorch: 12.1 ROCM used to build...

Use MiniGPT4 for batch inference without using Chat model

1

How to use MiniGPT4 for batch inference without using Chat model? I can't find a method to do so. This is necessary for us to test on new datasets.

The inference on the AMBER dataset is very slow.

6

The AMBER dataset is a yes-or-no dataset, but the questions in this dataset do not explicitly state: "Please answer yes or no." As a result, during evaluation, the model generates...

By comparing web pages [a](https://composingprograms.netlify.app/3/2) and [b](https://www.composingprograms.com/pages/32-functional-programming.html), it can be seen that Section 3.2 does not have the runtime results. It is hoped that they can be added.  

'TinyLlavaForConditionalGeneration' object has no attribute 'peft_config'

2

I followed the instructions in the `CUSTOM_FINETUNE.md` file to run the Pokémon dataset for LoRA fine-tuning and encountered an issue when using `bash scripts/train/custom_finetune.sh`. The error message is as follows:...

Support DPO training?

3

Does this repo support dpo training?

flash_attn_2 undefined symbol

2

I follow the instruction to set up env: ``` bash git clone https://github.com/TinyLLaVA/TinyLLaVA_Factory.git cd TinyLLaVA_Factory conda create -n tinyllava_factory python=3.10 -y conda activate tinyllava_factory pip install --upgrade pip # enable...