Ru-adapted phi-2 (change the tokenizer and the embeddings layer like here: https://russiansuperglue.com/login/submit_info/2899). Unfortunately, they left no code behind so I can only hope my code's correct.

Model hasn't been trained yet. It will undergo full parameter second-stage pre-training.

Downloads last month: 6

Safetensors

Model size

2.78B params

Tensor type

F32

Inference Providers NEW

Text Generation

This model is not currently available via any of the supported Inference Providers.