fix(modeling_phi): Fixes cached generation when above maximum context length. 37527ba gugarosa commited on Dec 5, 2023
Fixes exceeding maximum sequence length when using generate(). 5fd430c gugarosa commited on Nov 20, 2023
Enables to toggle fused_dense, flash_rotary and attn_pdrop in the configuration. 45f4b21 gugarosa commited on Nov 1, 2023
Adds support for flash-attn rotary embedding and fused dense layers. 0bbd68a gugarosa commited on Nov 1, 2023
Adds support for MQA/GQA and attention mask during training. de35f90 gugarosa commited on Oct 30, 2023
Adding _set_gradient_checkpointing for compatibility (#22) 8091327 gugarosa vriveras commited on Oct 17, 2023
Add more precise license metadata (UI will be cleaner!) (#35) 8ab0f29 gugarosa julien-c HF staff commited on Sep 27, 2023
fix(phi-1_5): Checks length of `attention_mask`if it is passed as direct tensor. f9f2ac7 gugarosa commited on Sep 26, 2023
