-
https://github.com/InternLM/lmdeploy/issues/1991#issue-2402071158 问题跟这里类似,问一些比较短的问题能正常输出,但是问一些比较长的问题(超过1万字,不超过session-len里面设置的长度)结果返回空。
模型:qwen1.5-7b-chat
启动脚本:lmdeploy serve api_server qwen1half-7b…
-
### Your current environment
Collecting environment information...
PyTorch version: 2.1.2+cu121
Is debug build: False
CUDA used to build PyTorch: 12.1
ROCM used to build PyTorch: N/A
OS: Ubunt…
-
## Expected Behavior
When importing a list of devices from the provided CSV template file, the portal should check the type of device authentication before creating it, and only set values fo…
-
Greeting,
I am encountering difficulties with GPS synchronization in the Lora gateway. My hardware setup is: Raspberry Pi 3B+ and SX1302 LoRaWAN Gateway HAT by waveshare. I've successfully connected…
-
I run this:
```python -m slora.server.api_server --max_total_token_num 64 --model meta-llama/Llama-2-13b-hf --tokenizer_mode auto --lora /playground/slora-playground/adapters-1/ --lora /playground/s…
-
### Is there an existing issue for this?
- [X] I have searched the existing issues and checked the recent builds/commits
### What happened?
I started 3 WebUI servers on one Nvidia GPU L40, the gen…
-
### Is this a bug or a feature request?
bug
### What did you expect?
build lora-app-server on Raspberry pi 3 , running alpine arm64
### What happened?
```
Generating combined Swagger JSON
[…
-
这是我输入的提示词:
我想让你扮演Java开发工程师的面试官,我将作为候选人回答问题。请按照以下要求进行:
1.你负责提问Java开发工程师职位的面试问题。
2.每次只问一个问题。
3.只提问,不要解释你提的问题,不要回答你提的问题。
请问第一个问题
这是大模型给出的回答:
请简述一下Java中垃圾回收机制(Garbage Collection)的工作原理和你在项目中如何优化内存…
xqxls updated
2 weeks ago
-
**Is your feature request related to a problem? Please describe.**
This is a feature related to how to deploy a model with LoRA supported.
**Describe the solution you'd like**
I have a UNet model…
-
I encountered an issue with an older RAK v1.5 where the concentrator reset pin was 17 instead of 25, which is the default specified in the variant file for COMP-RAKHM. This was causing an endless loop…