Create README.md
Browse files
README.md
ADDED
@@ -0,0 +1,41 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
---
|
2 |
+
inference: false
|
3 |
+
license: llama2
|
4 |
+
language:
|
5 |
+
- en
|
6 |
+
tags:
|
7 |
+
- information retrieval
|
8 |
+
- reranker
|
9 |
+
---
|
10 |
+
|
11 |
+
# RankVicuna (FP16) Model Card
|
12 |
+
|
13 |
+
## Model Details
|
14 |
+
|
15 |
+
RankVicuna is a chat assistant trained by fine-tuning Llama 2 on user-shared conversations collected from ShareGPT.
|
16 |
+
|
17 |
+
- **Developed by:** [Castorini](https://github.com/castorini)
|
18 |
+
- **Model type:** An auto-regressive language model based on the transformer architecture
|
19 |
+
- **License:** Llama 2 Community License Agreement
|
20 |
+
- **Finetuned from base model:** [Llama 2](https://arxiv.org/abs/2307.09288)
|
21 |
+
|
22 |
+
This specific model is a 7B variant and is trained with data augmentation.
|
23 |
+
It is also worth noting that it is converted to FP16.
|
24 |
+
|
25 |
+
### Model Sources
|
26 |
+
|
27 |
+
- **Repository:** https://github.com/castorini/rank_llm
|
28 |
+
- **Paper:** https://arxiv.org/abs/2309.15088
|
29 |
+
|
30 |
+
## Uses
|
31 |
+
|
32 |
+
The primary use of RankVicuna is research at the intersection of large language models and retrieval.
|
33 |
+
The primary intended users of the model are researchers and hobbyists in natural language processing and information retrieval.
|
34 |
+
|
35 |
+
## Training Details
|
36 |
+
|
37 |
+
RankVicuna is finetuned from `lmsys/vicuna-7b-v1.5` with supervised instruction fine-tuning.
|
38 |
+
|
39 |
+
## Evaluation
|
40 |
+
|
41 |
+
RankVicuna is currently evaluated on DL19/DL20. See more details in our [paper](https://arxiv.org/pdf/2309.15088.pdf).
|