-
Hello,
today i tried the `ops setup devcluster` on a Ubuntu machine which has the cockpit serivce enabled.
The setup failed because the port 9090 (wanted by nuvolaris-worker) is used by default fr…
-
### 🐛 Describe the bug
torchbench_amp_bf16_training
xpu train pyhpc_turbulent_kinetic_energy
Traceback (most recent call last):
File "/home/sdp/actions-runner/_work/torch-xpu-ops/pytorch/benc…
-
**Describe the bug:**
I was trying out the devtools connection of servo for the first time. I tried to execute a minimal snippet of JS code in the console: `window.location`. This completely crashed …
-
### 🐛 Describe the bug
- [x] `timm_efficientdet`
- [ ] `simple_gpt`
- [ ] `simple_gpt_tp_manual`
Traceback (most recent call last):
File "/home/sdp/actions-runner/_work/torch-xpu-ops/pyto…
-
### 🐛 Describe the bug
torchbench_amp_bf16_inference
- [ ] `moco`
Traceback (most recent call last):
File "/home/sdp/actions-runner/_work/torch-xpu-ops/pytorch/benchmarks/dynamo/common.py", …
-
I have a question about a use case of flex attention.
I am wondering if it is possible to write to some globally scoped tensor the way that the alibi bias example in the link above reads from a glo…
-
### 🐛 Describe the bug
torchbench_amp_bf16_training
xpu train detectron2_fcos_r_50_fpn
Traceback (most recent call last):
File "/home/sdp/actions-runner/_work/torch-xpu-ops/pytorch/benc…
-
### 🐛 Describe the bug
torchbench_amp_fp16_training
- [ ] `detectron2_fcos_r_50_fpn`
- [ ] `detectron2_maskrcnn_r_50_c4`
xpu train detectron2_maskrcnn
Traceback (most recent …
-
### 🐛 Describe the bug
torchbench_amp_bf16_training
- [ ] `pyhpc_equation_of_state`
- [ ] `pyhpc_isoneutral_mixing`
- [ ] `maml`
- [ ] `maml_omniglot`
- [ ] `cm3leon_generate`
- [ ] `hf_T5_ge…
-
### 🐛 Describe the bug
torchbench_amp_bf16_training
xpu train opacus_cifar10
Traceback (most recent call last):
File "/home/sdp/actions-runner/_work/torch-xpu-ops/pytorch/benc…