-
直接使用官方样例数据 https://github.com/FlagOpen/FlagEmbedding/blob/master/FlagEmbedding/llm_reranker/toy_finetune_data.jsonl
对 bge-reranker-v2-m3 进行微调
命令为
`torchrun --nproc_per_node 1 -m FlagEmbedding.l…
-
While looking at the reason why one of the pipelines was stuck in the `LABEL_INFERENCE`, I found that we appear to re-load the full trip model before each trip inference.
This is:
- unncessary, si…
-
### 🐛 Describe the bug
Internal xref: https://www.internalfb.com/intern/anp/view/?id=4913315
```
from typing import List, Tuple
import torch
from torchrec.sparse.jagged_tensor import KeyedJag…
-
(pytorch) root@node03:/mnt/pengfeng/perfxcloud/enpoints/solar-10.7b-instruct# CUDA_VISIBLE_DEVICES=0,1 bentoml serve . --port 2999
2024-04-28T04:44:35+0000 [WARNING] [cli] Converting SOLAR-10.7B-Inst…
-
I was able to download the 7B weights on Mac OS Monterey. I get the following errors when I try to call the example from the README in my Terminal: ```torchrun --nproc_per_node 1 example.py --ckpt_dir…
-
## Your Environment
* Operating System and version: Ubuntu 14.04
* Compiler: GCC v4.8.4
* PCL Version: 1.8.1
* C++ Version: C++11
## Context
I'm currently trying to find a point cl…
-
### Buggy output
```
WARNING:torch.distributed.run:
*****************************************
Setting OMP_NUM_THREADS environment variable for each process to be 1 in default, to avoid your system…
-
That doesn't allow arbitrary unpickling and thus arbitrary code execution. Maybe an option for torch.load?
Yes, one should not load/run code from unknown locations, but sometimes intermediate cont…
-
### 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 …
-
### 一、BackGround 📚
飞桨新IR（PIR）功能建设已经基本完成，当前CI流水线上静态图依然是以飞桨旧IR运行，我们想将CI默认运行的IR切换至PIR，从而能顺利支持未来飞桨基于PIR下的代码提交与验证。但是当前依然有很多单测在PIR模式下的运行会存在问题，修复这些问题成为实现默认切换PIR的必要条件。
### 二、Task Introduction📚
对于存在问题的单测，我们已统一…