Dynamo 8B outperforms Mistral 7B, Llama2 13B, Bloom 7B, and PolyLM 13B on most multilingual benchmarks we tested (i.e. PAWS and XCOPA).
Dynamo 8B has not been instruction fine-tuned and has not undergone alignment using techniques like reinforcement learning from human feedback. The intention behind crafting this model is to provide the research community with a model to explore vital multilingual capabilities that enable widespread use of LLMs globally.
1 comment
[ 0.23 ms ] story [ 16.0 ms ] threadDynamo 8B has not been instruction fine-tuned and has not undergone alignment using techniques like reinforcement learning from human feedback. The intention behind crafting this model is to provide the research community with a model to explore vital multilingual capabilities that enable widespread use of LLMs globally.