-
-
### 🐛 Describe the bug
This example started to fail yesterday with the nightly build.
```python
import torch
class UpdateModel(torch.nn.Module):
def __init__(self):
super()._…
-
-
Hi, I'm running the [prune_ldm.py](https://github.com/VainF/Diff-Pruning/blob/a3a2a2646cb89f6a977f92059ded122bda620b23/ldm_exp/prune_ldm.py), and I got the following error in Line pruner.step(). Coul…
-
Consider the following model and export where we use the same tensor as the sample input
```python
import torch
from typing import List
class ListAdd(torch.nn.Module):
def __init__(self…
-
### 🚀 The feature, motivation and pitch
I use torch 2.3.1 on MacOS 15.1 beta
I profiling my model and observed that aten::bernoulli is taking 34% of the runtime on cpu while dropout time is marg…
-
### 🐛 Describe the bug
I am trying to build pytroch from source with support for rocm.
System is ubuntu 24
Cpu is ryzen 5 2600
Gpu is rx 580
I am following instructions from there: https://github…
-
### Comment:
A conda-forge environment with nothing but pytorch for cuda in it currently ways in at 7.2GB, which is perceived to be rather on the heavy side of things.
Looking at potential for slimm…
-
**Describe the bug**
https://github.com/NVIDIA/Megatron-LM/blob/01ca03f11e89f4f85682dcac647c2b913b25fcee/examples/run_simple_mcore_train_loop.py#L118
When I moditied `tensor_model_parallel_size` in `r…
-
im currently training a model with vctk from scratch on colab but everytime it loads a checkpoint, the training just restarts from 0 anyways
```
INFO:vctk_base:{'train': {'log_interval': 250, 'ev…