-
### Issue summary
I construct a new classification forward instance in main thread like this:
```
// construct a new forward class
Caffe::set_mode(Caffe::GPU);
net_.reset(new Net(model_file, TE…
-
### 🐛 Describe the bug
Working with LLMs, I got a strangely large CUDA OOM error. I was using torch.svd_lowrank, which again calls on torch._lowrank.get_approximate_basis. Below I paste the minimal…
-
#### Description
An API can only be given entire GPU units. Add support for fractional values for the GPU resource. Here's an example:
```yaml
# cortex.yaml
- name: # API name (required)
# ...
…
-
### Your current environment
```text
Collecting environment information...
PyTorch version: N/A
Is debug build: N/A
CUDA used to build PyTorch: N/A
ROCM used to build PyTorch: N/A
OS: Ubuntu …
-
Hello,
Thanks for the great package!
I'd like to do multi-GPU parallel sweeps. I have 4 GPUs and I'd like to do a sweep on, say 16 configs. I have this code:
```python
wandb.require("core"…
-
### Suggestion Description
My laptop is an HP Omen. It has a gfx90c (not supported) and a gfx1032 (supported). The gfx90c cannot be disabled, even in the BIOS. I cannot run rocprof because it detects…
-
Epoch GPU_mem box_loss cls_loss dfl_loss Instances Size
6/99 5.66G 7.085 13.34 5.611 11 640: 100%|██████████| 30/30 00:12
…
-
Hi,
I faced the issue when I tried to run **6.1 normal inference** and **6.2 inference with mixed precision** as your indications. But something was wrong:
**For 6.1 normal inference:**
(viditq…
-
Look into:
* Access to fine-tuning: API access for closed-source models
* Code availability for OS models
* Rate limits
* API costs for close-source models
* Cloud compute costs for OS …
-
**Please provide an in-depth description of the question you have**:
I was experimenting with Nvidia g4dn.xlarge instance type with HAMi device plugin Daemonset with these options set
The sched…