SeaLLMs
/

SeaLLM-13B-Chat

Model card Files Files and versions Community

nxphi47 commited on Feb 2

Commit

13c2080

•

1 Parent(s): 6147c97

Update README.md

Files changed (1) hide show

README.md +4 -3

README.md CHANGED Viewed

@@ -23,6 +23,7 @@ tags:
 # SeaLLMs - Large Language Models for Southeast Asia
 <p align="center">
 <a href="https://huggingface.co/SeaLLMs/SeaLLM-Chat-13b" target="_blank" rel="noopener"> 🤗 Tech Memo</a>
@@ -38,11 +39,11 @@ We introduce SeaLLMs - a family of language models optimized for Southeast Asian
 SeaLLM-13b models exhibit superior performance across a wide spectrum of linguistic tasks and assistant-style instruction-following capabilities relative to comparable open-source models. Moreover, they outperform **ChatGPT-3.5** in non-Latin languages such as Thai, Khmer, Lao, and Burmese.
-- DEMO: [SeaLLMs/SeaLLM-Chat-13b](https://huggingface.co/spaces/SeaLLMs/SeaLLM-Chat-13b) DEMO allows **batch-inference** for evaluation purposes.
 - Technical report: [Arxiv: SeaLLMs - Large Language Models for Southeast Asia](https://arxiv.org/pdf/2312.00738.pdf).
 - Model weights:
-  - [SeaLLM-7B-Hybrid](https://huggingface.co/SeaLLMs/SeaLLM-7B-Hybrid): A "base" model trained with mix of unlabeled raw texts and English supervised data in a hybrid setup, used for few-shot prompting or fine-tuning. (see [technical report](https://arxiv.org/pdf/2312.00738.pdf) for details).
-  - [SeaLLM-7B-Chat](https://huggingface.co/SeaLLMs/SeaLLM-7B-chat): Lower capability than [SeaLLM-13B-Chat](https://huggingface.co/spaces/SeaLLMs/SeaLLM-Chat-13b) but much faster and memory-efficient.
 <blockquote style="color:red">

 # SeaLLMs - Large Language Models for Southeast Asia
+# UPDATE: [SeaLLM-7B-v2](https://huggingface.co/SeaLLMs/SeaLLM-7B-v2) is released with state-of-the-art performance.
 <p align="center">
 <a href="https://huggingface.co/SeaLLMs/SeaLLM-Chat-13b" target="_blank" rel="noopener"> 🤗 Tech Memo</a>
 SeaLLM-13b models exhibit superior performance across a wide spectrum of linguistic tasks and assistant-style instruction-following capabilities relative to comparable open-source models. Moreover, they outperform **ChatGPT-3.5** in non-Latin languages such as Thai, Khmer, Lao, and Burmese.
+- DEMO: [SeaLLMs/SeaLLM-7B](https://huggingface.co/spaces/SeaLLMs/SeaLLM-7B).
 - Technical report: [Arxiv: SeaLLMs - Large Language Models for Southeast Asia](https://arxiv.org/pdf/2312.00738.pdf).
 - Model weights:
+  - (New) [SeaLLM-7B-v2](https://huggingface.co/SeaLLMs/SeaLLM-7B-v2): A new model with state-of-the-art performance in reasoning capabilities across all tasks and languages, while half the size of SeaLLM-13B
+  - [SeaLLM-7B-v1](https://huggingface.co/SeaLLMs/SeaLLM-7B-v1): Lower capability than [SeaLLM-13B-Chat](https://huggingface.co/spaces/SeaLLMs/SeaLLM-Chat-13b) but much faster and memory-efficient.
 <blockquote style="color:red">