update `bos_token_id` and `eos_token_id`

Update LICENSE

Update README.md

Update README.md

Update config.json

Update modeling_phi.py

Update README.md

Update modeling_phi.py

Update modeling_phi.py

Update modeling_phi.py

Update config.json

Upload modeling_phi.py

Delete Research License.docx

Upload 5 files

Update README.md (#69)

Update config.json

chore(root): Updates files to internal transformers implementation.

Update README.md

Upload 4 files

Update README.md

Update README.md

Disables inference API to prevent mismatch with HF implementation.

fix(modeling_phi): Fixes initial generation with length larger than context length.

fix(modeling_phi): Fixes cached generation when above maximum context length.

Fixes exceeding maximum sequence length when using generate().

Delete modeling_mixformer_sequential.py

Delete configuration_mixformer_sequential.py

Update to new model interface.

Improves type hinting on configuration arguments.

Enables to toggle fused_dense, flash_rotary and attn_pdrop in the configuration.

Fixes flash-attn import with a try/except statement

Adds support for flash-attn rotary embedding and fused dense layers.

Adds support for MQA/GQA and attention mask during training.

Update modeling_mixformer_sequential.py

Adding _set_gradient_checkpointing for compatibility (#22)

Upload modeling_mixformer_sequential.py

Add more precise license metadata (UI will be cleaner!) (#35)

Upload README.md

fix(phi-1_5): Checks length of `attention_mask`if it is passed as direct tensor.

Support for `attention_mask` in forward pass.

add _no_split_modules property (#17)

Update README.md

Update README.md

Link paper to arXiv (#5)

Update README.md

Update README.md

Update README.md

Upload tokenizer

Upload MixFormerSequentialForCausalLM

Update README.md