Implementation of Flash Attention in Jax. It will likely not be as performant as with the official CUDA version, given lack of ability for fine memory management. But just for educational purposes as well as to see how clever XLA compiler is (or is not).
$ pip install flash-attention-jaxfrom jax import random
from flash_attention_jax import flash_attention
rng_key = random.PRNGKey(42)
q = random.normal(rng_key, (1, 2, 131072, 512)) # (batch, heads, seq, dim)
k = random.normal(rng_key, (1, 2, 131072, 512))
v = random.normal(rng_key, (1, 2, 131072, 512))
mask = random.randint(rng_key, (1, 131072,), 0, 2) # (batch, seq)
out, _ = flash_attention(q, k, v, mask)
out.shape # (1, 2, 131072, 512) - (batch, heads, seq, dim)Quick sanity check
from flash_attention_jax import plain_attention, flash_attention, value_and_grad_difference
diff, (dq_diff, dk_diff, dv_diff) = value_and_grad_difference(
plain_attention,
flash_attention,
seed = 42
)
print('shows differences between normal and flash attention for output, dq, dk, dv')
print(f'o: {diff}') # < 1e-4
print(f'dq: {dq_diff}') # < 1e-6
print(f'dk: {dk_diff}') # < 1e-6
print(f'dv: {dv_diff}') # < 1e-6Autoregressive Flash Attention - GPT-like decoder attention
from jax import random
from flash_attention_jax import causal_flash_attention
rng_key = random.PRNGKey(42)
q = random.normal(rng_key, (131072, 512))
k = random.normal(rng_key, (131072, 512))
v = random.normal(rng_key, (131072, 512))
out, _ = causal_flash_attention(q, k, v)
out.shape # (131072, 512)-
leading dimensions for causal flash attention variant
-
figure out issue with jit and static argnums
-
comment with references to paper algorithms and explanations
-
make sure it can work one-headed key / values, as in PaLM
@article{Dao2022FlashAttentionFA,
title = {FlashAttention: Fast and Memory-Efficient Exact Attention with IO-Awareness},
author = {Tri Dao and Daniel Y. Fu and Stefano Ermon and Atri Rudra and Christopher R'e},
journal = {ArXiv},
year = {2022},
volume = {abs/2205.14135}
}@article{Rabe2021SelfattentionDN,
title = {Self-attention Does Not Need O(n2) Memory},
author = {Markus N. Rabe and Charles Staats},
journal = {ArXiv},
year = {2021},
volume = {abs/2112.05682}
}
React
Typescript
Django
Laravel
Facebook
Microsoft
Google
Alibaba
D3
Tencent