#38 · Primary category： Foundation Models
ChatGLM2-6B
ChatGLM2-6B: An Open Bilingual Chat LLM
Project last updated：06/27/24
GitHub Stars
15.5K
Forks
1.8K
Contributors
12
License
Other
Why we included this project
ChatGLM2-6B is a solid choice for teams that want a bilingual Chinese-English chat model that runs on hardware they already own. At 6B parameters it keeps the low deployment bar of the original ChatGLM-6B while delivering much stronger reasoning and math scores, which is what made the second generation worth the upgrade. FlashAttention pushes the context window to 32K, so long documents stay usable, and Multi-Query Attention cuts both latency and memory, letting INT4 quantization hold long conversations on a 6GB GPU. The repo bundles the weights, tokenizer, and inference code, plus P-Tuning v2 and full fine-tuning scripts, so you can go from download to a customized assistant without stitching together several projects. For Chinese-language work it is a practical base to build on, and the surrounding ecosystem of acceleration and UI projects makes production integration cheaper.
Articles for this project
No articles for this project yet.
To suggest a topic or contribute an article, contact us.
Related projects in this category
transformers
🤗 Transformers: the model-definition framework for state-of-the-art machine learning models in text, vision, audio, and multimodal models, for both inference and training.
CLIP
CLIP (Contrastive Language-Image Pretraining), Predict the most relevant text snippet given an image
MiniCPM-V
A Pocket-Sized MLLM for Ultra-Efficient Image and Video Understanding on Your Phone
generative-models
Generative Models by Stability AI
unilm
Large-scale Self-supervised Pre-training Across Tasks, Languages, and Modalities