Cerebras implementation and training recipes related to multimodal LLaVA models
Cerebras
company
Verified
AI & ML interests
None defined yet.
Organization Card
About org cards
Cerebras is the inventor of the Wafer-Scale Engine – the revolutionary processor at the heart of our Cerebras CS-2 system. Our co-designed hardware/software stack is designed to train large language models upward of 1 trillion parameters using only data parallelism. This is a collection of models we trained on Cerebras CS-2 systems.
Join the Cerebras Discord to discuss our work and research!
Collections
1
models
14
cerebras/Cerebras-GPT-Intermediate
Text Generation
•
Updated
cerebras/Cerebras-LLaVA-13B
Text Generation
•
Updated
•
9
•
2
cerebras/Cerebras-ViT-L-336-patch14-llava13b-ShareGPT4V
Updated
•
10
cerebras/Cerebras-ViT-L-336-patch14-llava7b-ShareGPT4V
Updated
•
6
cerebras/Cerebras-LLaVA-7B
Text Generation
•
Updated
•
12
•
2
cerebras/btlm-3b-8k-chat
Text Generation
•
Updated
•
63
•
12
cerebras/Cerebras-GPT-13B
Text Generation
•
Updated
•
4.44k
•
636
cerebras/Cerebras-GPT-6.7B
Text Generation
•
Updated
•
3.96k
•
65
cerebras/Cerebras-GPT-111M
Text Generation
•
Updated
•
6k
•
70
cerebras/Cerebras-GPT-256M
Text Generation
•
Updated
•
4.42k
•
24