microsoft
/

phi-1_5

Text Generation Transformers PyTorch English mixformer-sequential custom_code

Model card Files Files and versions Community

Resources

View closed (13)

Inference speed is much longer than reported

#25 opened about 3 hours ago by

Use DirectML for microsoft/phi-1_5

#24 opened 1 day ago by

raise error when `use_cache = True`

#23 opened 1 day ago by

Adding _set_gradient_checkpointing for compatibility

#22 opened 1 day ago by

Any plan to release phi-1.5 web mentioned in your paper?

#21 opened 1 day ago by

Adding `safetensors` variant of this model

#18 opened 5 days ago by

Creating a RetNet (Retentive Network) version is planned?

#16 opened 6 days ago by

Adding tf lite variant

#13 opened 7 days ago by

Could one potentially train a mini-model based on this concept on synthetic structural data?

#11 opened 7 days ago by

tokenizer.model file

#10 opened 8 days ago by

Attention mask for generation function in the future?

#7 opened 8 days ago by

Unofficial dataset

#2 opened 8 days ago by