-
```
(Unidepth) hygx@DESKTOP-47Q8A9V:~/code/UniDepth$ python ./scripts/demo.py
Triton is not available, some optimizations will not be enabled.
This is just a warning: triton is not available
Torch…
-
### System Info
### what i want
So I want a solution that can quickly generate AI output by efficiently using precompute kv caches of text and images of all previous prompts!
### by using…
-
- https://arxiv.org/abs/2110.07811
- 2021
自然言語によるセマンティックコード検索の目的は、自然言語のクエリを用いて、固定された候補のセットから意味的に関連するコードスニペットを取り出すことです。
既存のアプローチは、実用的なセマンティックコード検索システムに向けて十分な効果も効率もありません。
本論文では、高速モデルと低速モデルをカスケード接続…
e4exp updated
2 years ago
-
- [link to OpenReview](https://openreview.net/forum?id=U1edbV4kNu_)
- 2021 ICLR 2022 Conference Blind Submission
概要:
多くの深層学習アプリケーションでは、何十億ものパラメータを持つ大規模なモデルを使用することで利益を得ています。
これらのモデルは、低レイテンシーかつ高…
e4exp updated
2 years ago
-
TensorRT-LLM has great potential for allowing people to run larger models efficiently with limited hardware resources. Unfortunately, the current quantization workflow requires significant computation…
-
-
# Papers
- Sapiens: Foundation for Human Vision Models
- 메타에서 나온 Human foundation model ㄷㄷㄷ
- 2D pose estimation, body-part segmentation, depth prediction and normal prediction이 하나의 모델에서 …
-
- Deepeval: https://github.com/confident-ai/deepeval
- RAGAS: https://docs.ragas.io/en/stable/
- Figure out the best way to run this: a) in front-end using Pyodide or b) run it as back-end Flask API…
-
imported this into ggmax, its a construction frame built to construct Titanic in 1910s/1911s. gave it a good clean over in blender, exported and imported into ggmax, the model kills my frame rate in …
-
**Introduction or background of this discussion**:
OSPP project: "Real-Time IoT Perception Systems Based on Edge-Cloud Collaboration with Large Foundation Models"
**Contents of this discussion…