Laurent Mazare fb1c2ac535 Add flash-attn support. (#912) ...

* Add flash-attn support.

* Add the use-flash-attn flag.

* Re-enable flash-attn.

2023-09-20 14:07:55 +01:00

..

segment_anything

Return the low res mask in the wasm segment-anything module. (#798)

2023-09-10 13:03:02 +01:00

stable_diffusion

Line-up the wuerstchen model with the python implementation. (#901)

2023-09-19 21:59:44 +01:00

whisper

Use softmax-last-dim in whisper. (#810)

2023-09-11 11:05:05 +01:00

wuerstchen

Add flash-attn support. (#912)

2023-09-20 14:07:55 +01:00

bert.rs

Move some models to candle-transformers so that it's easier to re-use. (#794)

2023-09-10 09:40:27 +01:00

bigcode.rs

Move some models to candle-transformers so that it's easier to re-use. (#794)

2023-09-10 09:40:27 +01:00

dinov2.rs

Move more models to candle-transformers (#796)

2023-09-10 10:20:18 +01:00

efficientnet.rs

Move more models to candle-transformers (#796)

2023-09-10 10:20:18 +01:00

falcon.rs

Move some models to candle-transformers so that it's easier to re-use. (#794)

2023-09-10 09:40:27 +01:00

llama.rs

Move some models to candle-transformers so that it's easier to re-use. (#794)

2023-09-10 09:40:27 +01:00

mod.rs

Start adding the Wuerstchen diffusion pipeline (#843)

2023-09-14 10:56:07 +01:00

quantized_llama.rs

Use softmax-last-dim in the quantized example. (#848)

2023-09-14 17:29:24 +01:00

t5.rs

Flan T5: Read lm_head when word embeddings are not tied (#903)

2023-09-19 22:36:47 +01:00