-
Thanks for your impressive work. There is still one thing that I can't figure out. According to the setting of continual learning, when learning on the current task, the data from previous tasks is un…
-
export HF_HOME="/data/kmei1/huggingface/"
export DISK_DIR="/data/kmei1/huggingface/cache"
export MODEL_DIR="stabilityai/stable-diffusion-2-1"
export OUTPUT_DIR="canny_model"
export DATASET_NAME="j…
-
### Willingness to contribute
Yes. I would be willing to contribute this feature with guidance from the MLflow community.
### Proposal Summary
Adding a new feature to MLflow for enhanced prompt man…
-
未做数据增强,划分为7:3 7份为训练集,3份用作evaluate/test set, evaluate时关闭train(其实就是测试状态了)
-
跑了qwen2 72B的PPO出现的OOM异常
环境:4机32卡,80G显存,2T内存。理论上应该不会oom才对。
运行代码:
```
ray job submit --address="http://127.0.0.1:8265" \
--runtime-env-json='{"working_dir": "mycode/OpenRLHF-new","excludes"…
-
Les options
- avec machine learning ( créant u prompt engineering ?) mais opeanAI va couer trop cher ( faire autrement)
- avec la liste de toutes les villes liés au cameroun et en les répérant
- f…
-
- https://arxiv.org/abs/2109.01134
- 2021
視覚言語の事前学習は、近年、表現学習の有望な選択肢として浮上している。
これは、固定された重みのセットを学習するために画像と個別のラベルを使用するという伝統的な方法から、2つの別々のエンコーダのために画像と生のテキストを揃える方法へと移行します。
このようなパラダイムは、より広範な監視源の恩恵を受け、視…
e4exp updated
2 years ago
-
Getting this problem on an i9-9900k, RTX 2080 + RTX 3060 Ti system. Any suggestions to fix?
```A module that was compiled using NumPy 1.x cannot be run in
NumPy 2.0.1 as it may crash. To support b…
-
Hi, I've been playing with moonshot and I'm trying to figure out how to implement my own custom attack module using the provided sample template `sample_attack_module.py`.
In my case, I'm trying to…
-
### Is there an existing issue for this?
- [X] I have searched the existing issues
### Feature Description
Kaggle dataset : https://www.kaggle.com/datasets/ishigamisenku10/hiv-prediction
(The fo…