tianbuwei

执行finetune.sh时出现问题

1

There were missing keys in the checkpoint model loaded: ['base_model.model.model.embed_tokens.weight', 'base_model.model.model.layers.0.self_attn.q_proj.weight', 'base_model.model.model.layers.0.self_attn.k_proj.weight', 'base_model.model.model.layers.0.self_attn.v_proj.weight', ....] 程序在训练完成后，需要加载模型参数进行预测时会报错，如上： 我看了下您的代码，好像模型只会加载 LoRA的参数，而不加载LLaMa的参数，所以才会出现这样的问题。 请问一下，我这该怎么处理

[BUG]: save model parameters are incorrect

7

### 🐛 Describe the bug After the Llama model is trained using Lora training method, the model can be saved normally. However, Lora's model parameters were not included in the...

[BUG]: The LLama model trained in Lora mode is unable to perform normal reasoning

1

### 🐛 Describe the bug In this path 'ColossalAI/applications/Chat/examples/train_sft.sh', LLama-7B model is trained with Lora training method, but there is a problem in the reasoning process, is it because Lora...

使用双卡推理出现问题

5

前提：并没有安装apex，但是设置了config中的enable_layernorm_kernel=False 启动命令：CUDA_VISIBLE_DEVICES=0,1 torchrun --standalone --nproc_per_node=gpu scripts/inference.py configs/opensora/inference/16x512x512.py --ckpt-path /data1/tianhs/repositories/model/pretrained/hpcai-tech/Open-Sora/OpenSora-v1-HQ-16x512x512.pth --prompt-path /data1/tianhs/research/Open-Sora/assets/texts/t2v_samples.txt 报错信息： 

xinference 加载lora权重后不起作用，与原始模型效果一致

2

### System Info / 系統信息 linux ### Running Xinference with Docker? / 是否使用 Docker 运行 Xinfernece？ - [x] docker / docker - [ ] pip install / 通过 pip install...