-
### Describe the bug
When clicking buttons in the gradio web interface, there's a noticeable delay before the button press is actually received by the python server. I am not sure whether this dela…
-
问题：通过webui.py运行，推理模式选择预训练音色，点击生成音频报错，服务端显示：RuntimeError: "addmm_impl_cpu_" not implemented for 'Half'，具体报错信息如下：
2024-09-27 16:32:01,942 INFO get sft inference request
tn 我是通义实验室语音团队全新推出的生成式语音大模型，提…
-
**Describe the bug**
4xA100 gpu fine-tuning llama-3.1-8b-instruct (also tried llama2-13b-ms, same error), cli
```
CUDA_VISIBLE_DEVICES=0,1,2,3 \
NPROC_PER_NODE=4 \
LOCAL_WORLD_SIZE=4 \
swift…
-
Hello,
I installed privateGPT using:
```
salloc --job-name "InteractiveJob" --cpus-per-task 4 --mem-per-cpu 50gb --time 01:00:10 -p batch_gpu --gres=gpu:1
module load Mamba git CUDA && conda…
-
-
代码是直接从项目那边拷贝下来的：
```python
from qwen_agent.agents.doc_qa import ParallelDocQA
from qwen_agen…
wuybo updated
4 months ago
-
PS C:\ai\TensorLLM> pip install tensorrt_llm --extra-index-url https://pypi.nvidia.com --extra-index-url https://download.pytorch.org/whl/cu121
pip : DEPRECATION: Loading egg at c:\users\real\appdata…
-
### What I've done:
1. conda create -n gptq python=3.9 -y
2. conda activate gptq
3. conda install pytorch torchvision torchaudio pytorch-cuda=11.7 -c pytorch -c nvidia
4. git clone https://githu…
-
Starting from: https://github.com/h2oai/h2ogpt/pull/111
- [ ] Add template user can choose/upload to constrain output
- [ ] Avoid odd chops of chunks, see starts off with `ure\n\nquestion\n\nplaye…
-
I ran glm4 on MTL iGPU , and it reported this error:
oneAPI: l_BaseKit_p_2024.0.1.46_offline.sh
My env …