-
### 🐛 Describe the bug
Float32 dynamic shape CPP wrapper
suite
name
thread
batch_size_new
speed_up_new
inductor_new
eager_new
compila…
-
### Your current environment
```
# python collect_env.py
WARNING 11-08 16:53:16 cuda.py:23] You are using a deprecated `pynvml` package. Please install `nvidia-ml-py` instead, and make sure to un…
-
I am running onediff/benchmarks/image_to_video.py
### Your current environment information
Collecting environment information...
PyTorch version: 2.1.0a0+29c30b1
Is debug build: False
CUDA …
-
### 🐛 Describe the bug
After using torch.compile, I noticed a change in my output:
```
AssertionError: Tensor-likes are not close!
Mismatched elements: 8309 / 967680 (0.9%)
Greatest absolute di…
-
Hi,
As much of a technical guy I am I still cannot get past this error.
Can somebody tell me what I need to install manually and with what command ?|
Checking installs and venv + autodebug …
-
### 🐛 Describe the bug
Test Case
```
python -u -m pytest -s -v test/inductor/test_cpu_select_algorithm.py -k test_linear_amx_batch_size_1024_in_features_1024_out_features_1025_bias_True_cpu_float16…
-
### Bug description
When restarting from an existing checkpoint with a trainer that has a `max_steps` value set, the trainer does a single validation step before actually restarting the training epoc…
-
### 🐛 Describe the bug
I’m trying to use TorchDispatchMode to hook grad_fn for every tensor created within a dispatch mode context. I can successfully access grad_fn when running on a CPU device. How…
-
### 🐛 Describe the bug
torch.fake_quantize_per_tensor_affine quantizes inf, -inf and nan to quant_min (after dequantize).
This is expected for -inf, but not expected behavior for inf.
```
import…
-
### 🐛 Describe the bug
Hello,
I'm not sure whether it is intended, but autocast seems not working on embedding module.
below is the link of a colab notebook that reproduce the issue
https://…