-
I have test latest master branch for hf trainer, llama2 example run right, then I run the qwen example, it hang when save to memory.
env:
one node with 8 gpus
train scripts:
https://github.com…
-
### What happened?
A bug happened!
Few custom Smoker recipes broken after /cc reload or Server restart
### How can the bug be reproduced?
Create & Save 2 couple of custom item using same mat…
-
Hi
I have Smart Plug device which operates almost as smartplugv2_energyv3, but without light control.
so I prepared smartplugv2_energyv4 with corresponding product id
```yaml
name: Energy monito…
-
torch.cuda.OutOfMemoryError: CUDA out of memory. Tried to allocate 16.68 GiB (GPU 0; 79.35 GiB total capacity; 47.98 GiB already allocated; 13.28 GiB free; 64.89 GiB reserved in total by PyTorch) If r…
-
Hey, is it possible to send commands to a DL24? For example reset measurements, start and stop run, set backlight brightness. Would be very cool.
BR
-
background: I want to implement an op to fused type convert and matmul, base on triton matmul, the different with matmul:
1. force C's dtype is torch.float16, dot_out_dtype is tl.float16
2. accep…
-
```
[21:30:23][W][atorch_dl24:211]: CRC check failed (8F != 0D). Skipping frame
[21:30:23][D][atorch_dl24:212]: Payload: FF.55.01.03.00.07.EF.00.00.23.00.01.5A.00.00.02.BF.00.09.00.09.00.1F.00.00.26…
-
On DL24 it shows incorrect readings - voltage is 10 times smaller and current is 10 times larger, and other ones seem off on initial look.
![Screenshot 2023-01-17 at 16 06 45](https://user-images.…
-
@syssi you're everywhere mate :) Thankyou for your amazing contributions. I am now using 3 of them - your dl24 one, your JBD one and your JK-BMS one. Now I just have to work out how to combine them al…
-
Hi,
I have this product, it's an outlet with a temperature controller on it, I'm thinking to add this support to tinytuya, should I create a new category ?
```
"category": "wk",
…