-
首先，非常牛逼的项目，1024 x 1204 出来的图挺细腻的。点赞！👍👍👍
代码质量也非常高，注释非常到位，说明大佬是非常高标准要求的牛人。
测试了一下，下面是几点问题和建议：
1. 显存使用能否优化？像 SDXL、Kolors、PixArt，原始模型能跑 2048 x 2048，24GB显卡不会爆显存，大佬的一跑就爆了，不知有没有优化的空间？要不如果想跑 4096 x 4096 的话，需…
-
### 🐛 Describe the bug
the latest verison pytorch[2023-08-01] has support aten::scatter_reduce operator, when export the model to ONNX. but occurs error:
raise errors.OnnxExporterError(
"ONNX d…
-
我对qwen1half-moe-2.7B-chat使用常规lora微调后尝试了gptq的4bit量化，但在重新推理时出现了：
[rank0]: File "/root/miniconda3/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1709, in __getattr__
[rank0]: rais…
-
### 🐛 Describe the bug
While batching the BBBP dataset, there is one graph that is not associated with any node. This causes a discrepancy in the number of graph labels in the batch and output shape …
-
### Your current environment
Libc version: glibc-2.35
Python version: 3.10.12 (main, Jun 11 2023, 05:26:28) [GCC 11.4.0] (64-bit runtime)
Python platform: Linux-5.4.250-2-velinux1u1-amd64-x86_64-…
-
### 🐛 Describe the bug
[inductor_single_run.sh](https://github.com/chuanqi129/inductor-tools/blob/main/scripts/modelbench/inductor_single_run.sh)
static shape:
```
/workspace/pytorch# bash inducto…
-
### 🐛 Describe the bug
I ran into this issue when applying torch.compile transformer blocks in torchtitain. When running llama_70b w/ local batch_size=4, both eager mode and torch.compile works. Howe…
-
### 🐛 Describe the bug
Compare with the 2023-03-29, there is a performance regression on huggingface model**BartForCausalLM & MBartForCausalLM & OPTForCausalLM & PLBartForCausalLM** on [TorchInductor…
-
#pip list
Package Version
-------------------------- --------------
absl-py 2.1.0
accelerate 0.22.0
aniso8601 9.0.1
annotat…
-
### Your current environment
```text
Collecting environment information...
PyTorch version: 2.3.1+cu121
Is debug build: False
CUDA used to build PyTorch: 12.1
ROCM used to build PyTorch: N/A
…