Skip to content

Instantly share code, notes, and snippets.

@yzhangcs

yzhangcs/chunk_split.py

Created December 17, 2024 16:37
Show Gist options
  • Save yzhangcs/404155039bc906de30eb015ff37bebd5 to your computer and use it in GitHub Desktop.
Save yzhangcs/404155039bc906de30eb015ff37bebd5 to your computer and use it in GitHub Desktop.
Download ZIP
Raw
chunk_split.py
# -*- coding: utf-8 -*-
# Copyright (c) 2024, Songlin Yang, Yu Zhang
from typing import Optional, Tuple
import torch
import triton
import triton.language as tl
from fla.ops.common.chunk_h_split import chunk_bwd_dh, chunk_fwd_h
from fla.ops.utils import chunk_local_cumsum
from fla.utils import contiguous
@triton.autotune(
configs=[
triton.Config({'BK': BK}, num_warps=num_warps, num_stages=num_stages)
for BK in [32, 64]
for num_warps in [1, 2, 4, 8]
for num_stages in [2, 3, 4]
],
key=["BC"],
)
@triton.heuristics({'USE_OFFSETS': lambda args: args['offsets'] is not None})
@triton.jit
def chunk_gla_fwd_A_kernel_intra_sub_inter(
q,
k,
g,
A,
offsets,
indices,
scale,
T: tl.constexpr,
H: tl.constexpr,
K: tl.constexpr,
BT: tl.constexpr,
BC: tl.constexpr,
BK: tl.constexpr,
NC: tl.constexpr,
USE_OFFSETS: tl.constexpr,
HEAD_FIRST: tl.constexpr
):
i_t, i_c, i_bh = tl.program_id(0), tl.program_id(1), tl.program_id(2)
i_b, i_h = i_bh // H, i_bh % H
i_i, i_j = i_c // NC, i_c % NC
if USE_OFFSETS:
i_n, i_t = tl.load(indices + i_t * 2).to(tl.int32), tl.load(indices + i_t * 2 + 1).to(tl.int32)
bos, eos = tl.load(offsets + i_n).to(tl.int32), tl.load(offsets + i_n + 1).to(tl.int32)
T = eos - bos
else:
bos, eos = i_b * T, i_b * T + T
if i_t * BT + i_i * BC >= T:
return
if i_i <= i_j:
return
b_A = tl.zeros([BC, BC], dtype=tl.float32)
for i_k in range(tl.cdiv(K, BK)):
o_k = i_k * BK + tl.arange(0, BK)
m_k = o_k < K
if HEAD_FIRST:
p_q = tl.make_block_ptr(q + i_bh * T*K, (T, K), (K, 1), (i_t * BT + i_i * BC, i_k * BK), (BC, BK), (1, 0))
p_g = tl.make_block_ptr(g + i_bh * T*K, (T, K), (K, 1), (i_t * BT + i_i * BC, i_k * BK), (BC, BK), (1, 0))
p_k = tl.make_block_ptr(k + i_bh * T*K, (K, T), (1, K), (i_k * BK, i_t * BT + i_j * BC), (BK, BC), (0, 1))
p_gk = tl.make_block_ptr(g + i_bh * T*K, (K, T), (1, K), (i_k * BK, i_t * BT + i_j * BC), (BK, BC), (0, 1))
p_gn = tl.max_contiguous(tl.multiple_of(g + (i_bh * T + i_t * BT + i_i * BC) * K + o_k, BK), BK)
else:
p_q = tl.make_block_ptr(q + (bos*H+i_h)*K, (T, K), (H*K, 1), (i_t * BT + i_i * BC, i_k * BK), (BC, BK), (1, 0))
p_g = tl.make_block_ptr(g + (bos*H+i_h)*K, (T, K), (H*K, 1), (i_t * BT + i_i * BC, i_k * BK), (BC, BK), (1, 0))
p_k = tl.make_block_ptr(k + (bos*H+i_h)*K, (K, T), (1, H*K), (i_k * BK, i_t * BT + i_j * BC), (BK, BC), (0, 1))
p_gk = tl.make_block_ptr(g + (bos*H+i_h)*K, (K, T), (1, H*K), (i_k * BK, i_t * BT + i_j * BC), (BK, BC), (0, 1))
p_gn = tl.max_contiguous(tl.multiple_of(g + (bos + i_t * BT + i_i * BC) * H*K + i_h * K + o_k, BK), BK)
# [BK,]
b_gn = tl.load(p_gn, mask=m_k, other=0)
# [BC, BK]
b_q = tl.load(p_q, boundary_check=(0, 1))
b_g = tl.load(p_g, boundary_check=(0, 1))
b_qg = b_q * tl.exp(b_g - b_gn[None, :]) * scale
# [BK, BC]
b_k = tl.load(p_k, boundary_check=(0, 1))
b_gk = tl.load(p_gk, boundary_check=(0, 1))
b_kg = b_k * tl.exp(b_gn[:, None] - b_gk)
# [BC, BC] using tf32 to improve precision here.
b_A += tl.dot(b_qg, b_kg)
if HEAD_FIRST:
p_A = tl.make_block_ptr(A + i_bh*T*BT, (T, BT), (BT, 1), (i_t * BT + i_i * BC, i_j * BC), (BC, BC), (1, 0))
else:
p_A = tl.make_block_ptr(A + (bos*H + i_h)*BT, (T, BT), (H*BT, 1), (i_t * BT + i_i * BC, i_j * BC), (BC, BC), (1, 0))
tl.store(p_A, b_A.to(A.dtype.element_ty), boundary_check=(0, 1))
@triton.autotune(
configs=[
triton.Config({}, num_warps=1),
triton.Config({}, num_warps=2),
triton.Config({}, num_warps=4),
triton.Config({}, num_warps=8),
],
key=["BK", "BT"],
)
@triton.heuristics({'USE_OFFSETS': lambda args: args['offsets'] is not None})
@triton.jit
def chunk_gla_fwd_A_kernel_intra_sub_intra(
q,
k,
g,
A,
offsets,
indices,
scale,
T: tl.constexpr,
H: tl.constexpr,
K: tl.constexpr,
BT: tl.constexpr,
BC: tl.constexpr,
BK: tl.constexpr,
USE_OFFSETS: tl.constexpr,
HEAD_FIRST: tl.constexpr
):
i_t, i_i, i_bh = tl.program_id(0), tl.program_id(1), tl.program_id(2)
i_b, i_h = i_bh // H, i_bh % H
i_j = i_i
if USE_OFFSETS:
i_n, i_t = tl.load(indices + i_t * 2).to(tl.int32), tl.load(indices + i_t * 2 + 1).to(tl.int32)
bos, eos = tl.load(offsets + i_n).to(tl.int32), tl.load(offsets + i_n + 1).to(tl.int32)
T = eos - bos
else:
bos, eos = i_b * T, i_b * T + T
if i_t * BT + i_i * BC >= T:
return
o_i = tl.arange(0, BC)
o_k = tl.arange(0, BK)
m_k = o_k < K
m_A = (i_t * BT + i_i * BC + tl.arange(0, BC)) < T
if HEAD_FIRST:
o_A = i_bh * T*BT + (i_t * BT + i_i * BC + tl.arange(0, BC)) * BT + i_j * BC
p_q = tl.make_block_ptr(q + i_bh * T*K, (T, K), (K, 1), (i_t * BT + i_i * BC, 0), (BC, BK), (1, 0))
p_g = tl.make_block_ptr(g + i_bh * T*K, (T, K), (K, 1), (i_t * BT + i_i * BC, 0), (BC, BK), (1, 0))
p_k = tl.max_contiguous(tl.multiple_of(k + (i_bh * T + i_t * BT + i_j * BC) * K + o_k, BK), BK)
p_gk = tl.max_contiguous(tl.multiple_of(g + (i_bh * T + i_t * BT + i_j * BC) * K + o_k, BK), BK)
else:
o_A = (bos + i_t * BT + i_i * BC + tl.arange(0, BC)) * H*BT + i_h * BT + i_j * BC
p_q = tl.make_block_ptr(q + (bos * H + i_h) * K, (T, K), (H*K, 1), (i_t * BT + i_i * BC, 0), (BC, BK), (1, 0))
p_g = tl.make_block_ptr(g + (bos * H + i_h) * K, (T, K), (H*K, 1), (i_t * BT + i_i * BC, 0), (BC, BK), (1, 0))
p_k = tl.max_contiguous(tl.multiple_of(k + (bos + i_t * BT + i_j * BC) * H*K + i_h * K + o_k, BK), BK)
p_gk = tl.max_contiguous(tl.multiple_of(g + (bos + i_t * BT + i_j * BC) * H*K + i_h * K + o_k, BK), BK)
b_q = tl.load(p_q, boundary_check=(0, 1))
b_g = tl.load(p_g, boundary_check=(0, 1))
for j in range(0, min(BC, T - i_t * BT - i_i * BC)):
b_k = tl.load(p_k, mask=m_k, other=0).to(tl.float32)
b_gk = tl.load(p_gk, mask=m_k, other=0).to(tl.float32)
b_A = tl.sum(b_q * b_k[None, :] * tl.exp(b_g - b_gk[None, :]), 1)
b_A = tl.where(o_i >= j, b_A * scale, 0.)
tl.store(A + o_A + j, b_A, mask=m_A)
p_k += K if HEAD_FIRST else H*K
p_gk += K if HEAD_FIRST else H*K
@triton.autotune(
configs=[
triton.Config({}, num_warps=1),
triton.Config({}, num_warps=2),
triton.Config({}, num_warps=4),
triton.Config({}, num_warps=8),
],
key=["BC", "BK"],
)
@triton.heuristics({'USE_OFFSETS': lambda args: args['offsets'] is not None})
@triton.jit
def chunk_gla_fwd_A_kernel_intra_sub_intra_split(
q,
k,
g,
A,
offsets,
indices,
scale,
B: tl.constexpr,
T: tl.constexpr,
H: tl.constexpr,
K: tl.constexpr,
BT: tl.constexpr,
BC: tl.constexpr,
BK: tl.constexpr,
NC: tl.constexpr,
USE_OFFSETS: tl.constexpr,
HEAD_FIRST: tl.constexpr
):
i_k, i_tc, i_bh = tl.program_id(0), tl.program_id(1), tl.program_id(2)
i_b, i_h = i_bh // H, i_bh % H
i_t, i_i = i_tc // NC, i_tc % NC
i_j = i_i
if USE_OFFSETS:
i_n, i_t = tl.load(indices + i_t * 2).to(tl.int32), tl.load(indices + i_t * 2 + 1).to(tl.int32)
bos, eos = tl.load(offsets + i_n).to(tl.int32), tl.load(offsets + i_n + 1).to(tl.int32)
all = T
T = eos - bos
else:
bos, eos = i_b * T, i_b * T + T
all = B * T
if i_t * BT + i_i * BC >= T:
return
o_i = tl.arange(0, BC)
o_k = i_k * BK + tl.arange(0, BK)
m_k = o_k < K
m_A = (i_t * BT + i_i * BC + tl.arange(0, BC)) < T
if HEAD_FIRST:
o_A = (i_k * B*H + i_bh) * T * BC + (i_t * BT + i_i * BC + tl.arange(0, BC)) * BC
p_q = tl.make_block_ptr(q + i_bh * T*K, (T, K), (K, 1), (i_t * BT + i_i * BC, i_k * BK), (BC, BK), (1, 0))
p_g = tl.make_block_ptr(g + i_bh * T*K, (T, K), (K, 1), (i_t * BT + i_i * BC, i_k * BK), (BC, BK), (1, 0))
p_k = tl.max_contiguous(tl.multiple_of(k + (i_bh * T + i_t * BT + i_j * BC) * K + o_k, BK), BK)
p_gk = tl.max_contiguous(tl.multiple_of(g + (i_bh * T + i_t * BT + i_j * BC) * K + o_k, BK), BK)
else:
o_A = (i_k * all + bos + i_t * BT + i_i * BC + tl.arange(0, BC)) * H*BC + i_h * BC
p_q = tl.make_block_ptr(q + (bos * H + i_h) * K, (T, K), (H*K, 1), (i_t * BT + i_i * BC, i_k * BK), (BC, BK), (1, 0))
p_g = tl.make_block_ptr(g + (bos * H + i_h) * K, (T, K), (H*K, 1), (i_t * BT + i_i * BC, i_k * BK), (BC, BK), (1, 0))
p_k = tl.max_contiguous(tl.multiple_of(k + (bos + i_t * BT + i_j * BC) * H*K + i_h * K + o_k, BK), BK)
p_gk = tl.max_contiguous(tl.multiple_of(g + (bos + i_t * BT + i_j * BC) * H*K + i_h * K + o_k, BK), BK)
b_q = tl.load(p_q, boundary_check=(0, 1))
b_g = tl.load(p_g, boundary_check=(0, 1))
for j in range(0, min(BC, T - i_t * BT - i_i * BC)):
b_A = tl.zeros([BC], dtype=tl.float32)
b_k = tl.load(p_k, mask=m_k, other=0).to(tl.float32)
b_gk = tl.load(p_gk, mask=m_k, other=0).to(tl.float32)
b_A += tl.sum(b_q * b_k[None, :] * tl.exp(b_g - b_gk[None, :]), 1)
b_A = tl.where(o_i >= j, b_A * scale, 0.)
tl.store(A + o_A + j, b_A, mask=m_A)
p_k += K if HEAD_FIRST else H*K
p_gk += K if HEAD_FIRST else H*K
@triton.autotune(
configs=[
triton.Config({}, num_warps=1),
triton.Config({}, num_warps=2),
triton.Config({}, num_warps=4),
triton.Config({}, num_warps=8),
],
key=["BC"],
)
@triton.heuristics({'USE_OFFSETS': lambda args: args['offsets'] is not None})
@triton.jit
def chunk_gla_fwd_A_kernel_intra_sub_intra_merge(
A,
A2,
offsets,
indices,
B: tl.constexpr,
T: tl.constexpr,
H: tl.constexpr,
BT: tl.constexpr,
BC: tl.constexpr,
NK: tl.constexpr,
USE_OFFSETS: tl.constexpr,
HEAD_FIRST: tl.constexpr
):
i_t, i_c, i_bh = tl.program_id(0), tl.program_id(1), tl.program_id(2)
i_b, i_h = i_bh // H, i_bh % H
if USE_OFFSETS:
i_n, i_t = tl.load(indices + i_t * 2).to(tl.int32), tl.load(indices + i_t * 2 + 1).to(tl.int32)
bos, eos = tl.load(offsets + i_n).to(tl.int32), tl.load(offsets + i_n + 1).to(tl.int32)
all = T
T = eos - bos
else:
bos, eos = i_b * T, i_b * T + T
all = B * T
if i_t * BT + i_c * BC >= T:
return
b_A = tl.zeros([BC, BC], dtype=tl.float32)
for i_k in range(0, NK):
if HEAD_FIRST:
p_A = tl.make_block_ptr(A + (i_k*B*H+i_bh)*T*BC, (T, BC), (BC, 1), (i_t*BT + i_c*BC, 0), (BC, BC), (1, 0))
else:
p_A = tl.make_block_ptr(A + (i_k*all+bos)*H*BC+i_h*BC, (T, BC), (H*BC, 1), (i_t*BT + i_c*BC, 0), (BC, BC), (1, 0))
b_A += tl.load(p_A, boundary_check=(0, 1))
if HEAD_FIRST:
p_A2 = tl.make_block_ptr(A2 + i_bh*T*BT, (T, BT), (BT, 1), (i_t * BT + i_c * BC, i_c * BC), (BC, BC), (1, 0))
else:
p_A2 = tl.make_block_ptr(A2 + (bos*H+i_h)*BT, (T, BT), (H*BT, 1), (i_t * BT + i_c * BC, i_c * BC), (BC, BC), (1, 0))
tl.store(p_A2, b_A.to(A2.dtype.element_ty), boundary_check=(0, 1))
@triton.autotune(
configs=[
triton.Config({'BK': BK, 'BV': BV}, num_warps=num_warps)
for BK in [32, 64]
for BV in [64, 128]
for num_warps in [2, 4, 8]
],
key=["BT"],
)
@triton.heuristics({'USE_OFFSETS': lambda args: args['offsets'] is not None})
@triton.jit
def chunk_gla_fwd_kernel_o(
q,
k,
v,
g,
h,
o,
A,
offsets,
split_indices,
scale,
T: tl.constexpr,
S: tl.constexpr,
H: tl.constexpr,
K: tl.constexpr,
V: tl.constexpr,
BT: tl.constexpr,
BK: tl.constexpr,
BV: tl.constexpr,
USE_OFFSETS: tl.constexpr,
HEAD_FIRST: tl.constexpr
):
# handle one split at a time
# i_h: head index
# i_n: sequence index
# i_s: local split index inside a sequence
i_v, i_i, i_sh = tl.program_id(0), tl.program_id(1), tl.program_id(2)
i_ss, i_h = i_sh // H, i_sh % H
if USE_OFFSETS:
i_n, i_s = tl.load(split_indices + i_ss * 2).to(tl.int32), tl.load(split_indices + i_ss * 2 + 1).to(tl.int32)
bos, eos = tl.load(offsets + i_n).to(tl.int32), tl.load(offsets + i_n + 1).to(tl.int32)
T = eos - bos
NS = tl.cdiv(T, S)
else:
NS = tl.cdiv(T, S)
i_n, i_s = i_ss // NS, i_ss % NS
bos, eos = i_n * T, i_n * T + T
i_nh = i_n * H + i_h
i_i = tl.cdiv(i_s * S, BT) + i_i
if i_i >= tl.cdiv(T, BT):
return
b_o = tl.zeros([BT, BV], dtype=tl.float32)
for i_k in range(tl.cdiv(K, BK)):
o_k = i_k * BK + tl.arange(0, BK)
m_k = o_k < K
if HEAD_FIRST:
p_q = tl.make_block_ptr(q + i_nh * T*K, (T, K), (K, 1), (i_i * BT, i_k * BK), (BT, BK), (1, 0))
p_g = tl.make_block_ptr(g + i_nh * T*K, (T, K), (K, 1), (i_i * BT, i_k * BK), (BT, BK), (1, 0))
else:
p_q = tl.make_block_ptr(q + (bos * H + i_h) * K, (T, K), (H*K, 1), (i_i * BT, i_k * BK), (BT, BK), (1, 0))
p_g = tl.make_block_ptr(g + (bos * H + i_h) * K, (T, K), (H*K, 1), (i_i * BT, i_k * BK), (BT, BK), (1, 0))
# [BT, BK]
b_q = tl.load(p_q, boundary_check=(0, 1))
b_q = (b_q * scale).to(b_q.dtype)
# [BT, BK]
b_g = tl.load(p_g, boundary_check=(0, 1))
b_qg = (b_q * tl.exp(b_g)).to(b_q.dtype)
p_h = tl.make_block_ptr(h + i_sh * K*V, (K, V), (V, 1), (i_k * BK, i_v * BV), (BK, BV), (1, 0))
# [BK, BV]
b_h = tl.load(p_h, boundary_check=(0, 1)).to(tl.float32)
for i_j in range(tl.cdiv(i_s * S, BT), i_i):
if HEAD_FIRST:
p_k = tl.make_block_ptr(k + i_nh * T*K, (K, T), (1, K), (i_k * BK, i_j * BT), (BK, BT), (0, 1))
p_v = tl.make_block_ptr(v + i_nh * T*V, (T, V), (V, 1), (i_j * BT, i_v * BV), (BT, BV), (1, 0))
p_gk = tl.make_block_ptr(g + i_nh * T*K, (K, T), (1, K), (i_k * BK, i_j * BT), (BK, BT), (0, 1))
p_gn = g + i_nh * T*K + (min(i_j * BT + BT, T) - 1) * K + o_k
else:
p_k = tl.make_block_ptr(k + (bos * H + i_h)*K, (K, T), (1, H*K), (i_k * BK, i_j * BT), (BK, BT), (0, 1))
p_v = tl.make_block_ptr(v + (bos * H + i_h)*V, (T, V), (H*V, 1), (i_j * BT, i_v * BV), (BT, BV), (1, 0))
p_gk = tl.make_block_ptr(g + (bos * H + i_h)*K, (K, T), (1, H*K), (i_k * BK, i_j * BT), (BK, BT), (0, 1))
p_gn = g + (bos + min(i_j * BT + BT, T) - 1) * H*K + i_h * K + o_k
p_gn = tl.max_contiguous(tl.multiple_of(p_gn, BK), BK)
# [BK,]
b_gn = tl.load(p_gn, mask=m_k, other=0)
# [BK, BT]
b_k = tl.load(p_k, boundary_check=(0, 1))
b_gk = tl.load(p_gk, boundary_check=(0, 1))
b_kg = (b_k * tl.exp(b_gn[:, None] - b_gk)).to(b_k.dtype)
# [BT, BV]
b_v = tl.load(p_v, boundary_check=(0, 1))
# [BK, BV]
b_h = b_h * tl.exp(b_gn[:, None]) + tl.dot(b_kg, b_v)
# [BT, BV]
if i_k >= 0:
b_o += tl.dot(b_qg, b_h.to(b_qg.dtype))
if HEAD_FIRST:
p_A = tl.make_block_ptr(A + i_nh * T*BT, (T, BT), (BT, 1), (i_i * BT, 0), (BT, BT), (1, 0))
p_v = tl.make_block_ptr(v + i_nh * T*V, (T, V), (V, 1), (i_i * BT, i_v * BV), (BT, BV), (1, 0))
p_o = tl.make_block_ptr(o + i_nh * T*V, (T, V), (V, 1), (i_i * BT, i_v * BV), (BT, BV), (1, 0))
else:
p_A = tl.make_block_ptr(A + (bos * H + i_h) * BT, (T, BT), (H*BT, 1), (i_i * BT, 0), (BT, BT), (1, 0))
p_v = tl.make_block_ptr(v + (bos * H + i_h) * V, (T, V), (H*V, 1), (i_i * BT, i_v * BV), (BT, BV), (1, 0))
p_o = tl.make_block_ptr(o + (bos * H + i_h) * V, (T, V), (H*V, 1), (i_i * BT, i_v * BV), (BT, BV), (1, 0))
# [BT, BT]
b_A = tl.load(p_A, boundary_check=(0, 1))
b_A = tl.where(tl.arange(0, BT)[:, None] >= tl.arange(0, BT)[None, :], b_A, 0.)
# [BT, BV]
b_v = tl.load(p_v, boundary_check=(0, 1))
b_o += tl.dot(b_A.to(b_v.dtype), b_v, allow_tf32=False)
tl.store(p_o, b_o.to(p_o.dtype.element_ty), boundary_check=(0, 1))
@triton.autotune(
configs=[
triton.Config({}, num_warps=1),
triton.Config({}, num_warps=2),
triton.Config({}, num_warps=4),
triton.Config({}, num_warps=8),
],
key=["BK", "NC", "BT"],
)
@triton.heuristics({'USE_OFFSETS': lambda args: args['offsets'] is not None})
@triton.jit
def chunk_gla_bwd_kernel_intra(
q,
k,
g,
dA,
dq,
dk,
offsets,
indices,
T: tl.constexpr,
H: tl.constexpr,
K: tl.constexpr,
BT: tl.constexpr,
BC: tl.constexpr,
BK: tl.constexpr,
NC: tl.constexpr,
USE_OFFSETS: tl.constexpr,
HEAD_FIRST: tl.constexpr
):
i_k, i_c, i_bh = tl.program_id(0), tl.program_id(1), tl.program_id(2)
i_b, i_h = i_bh // H, i_bh % H
i_t, i_i = i_c // NC, i_c % NC
if USE_OFFSETS:
i_n, i_t = tl.load(indices + i_t * 2).to(tl.int32), tl.load(indices + i_t * 2 + 1).to(tl.int32)
bos, eos = tl.load(offsets + i_n).to(tl.int32), tl.load(offsets + i_n + 1).to(tl.int32)
else:
bos, eos = i_b * T, i_b * T + T
T = eos - bos
if i_t * BT + i_i * BC >= T:
return
o_k = i_k * BK + tl.arange(0, BK)
m_k = o_k < K
if HEAD_FIRST:
p_g = tl.make_block_ptr(g + i_bh * T*K, (T, K), (K, 1), (i_t * BT + i_i * BC, i_k * BK), (BC, BK), (1, 0))
else:
p_g = tl.make_block_ptr(g + (bos*H + i_h) * K, (T, K), (H*K, 1), (i_t * BT + i_i * BC, i_k * BK), (BC, BK), (1, 0))
# [BC, BK]
b_g = tl.load(p_g, boundary_check=(0, 1))
b_dq = tl.zeros([BC, BK], dtype=tl.float32)
if i_i > 0:
if HEAD_FIRST:
p_gn = tl.max_contiguous(tl.multiple_of(g + (i_bh * T + i_t * BT + i_i * BC) * K + o_k, BK), BK)
else:
p_gn = tl.max_contiguous(tl.multiple_of(g + (bos + i_t * BT + i_i * BC) * H*K + i_h*K + o_k, BK), BK)
# [BK,]
b_gn = tl.load(p_gn, mask=m_k, other=0)
for i_j in range(0, i_i):
if HEAD_FIRST:
p_k = tl.make_block_ptr(k + i_bh * T*K, (T, K), (K, 1), (i_t * BT + i_j * BC, i_k * BK), (BC, BK), (1, 0))
p_gk = tl.make_block_ptr(g + i_bh * T*K, (T, K), (K, 1), (i_t * BT + i_j * BC, i_k * BK), (BC, BK), (1, 0))
p_dA = tl.make_block_ptr(dA + i_bh * T*BT, (T, BT), (BT, 1), (i_t * BT + i_i * BC, i_j * BC), (BC, BC), (1, 0))
else:
p_k = tl.make_block_ptr(k+(bos*H+i_h)*K, (T, K), (H*K, 1), (i_t*BT+i_j*BC, i_k * BK), (BC, BK), (1, 0))
p_gk = tl.make_block_ptr(g+(bos*H+i_h)*K, (T, K), (H*K, 1), (i_t*BT+i_j*BC, i_k * BK), (BC, BK), (1, 0))
p_dA = tl.make_block_ptr(dA+(bos*H+i_h)*BT, (T, BT), (H*BT, 1), (i_t*BT+i_i*BC, i_j * BC), (BC, BC), (1, 0))
# [BC, BK]
b_k = tl.load(p_k, boundary_check=(0, 1))
b_gk = tl.load(p_gk, boundary_check=(0, 1))
b_kg = (b_k * tl.exp(b_gn[None, :] - b_gk))
# [BC, BC]
b_dA = tl.load(p_dA, boundary_check=(0, 1))
# [BC, BK]
b_dq += tl.dot(b_dA, b_kg)
b_dq *= tl.exp(b_g - b_gn[None, :])
o_i = tl.arange(0, BC)
m_dA = (i_t * BT + i_i * BC + tl.arange(0, BC)) < T
if HEAD_FIRST:
o_dA = i_bh * T*BT + (i_t * BT + i_i * BC + tl.arange(0, BC)) * BT + i_i * BC
p_kj = tl.max_contiguous(tl.multiple_of(k + (i_bh * T + i_t * BT + i_i * BC) * K + o_k, BK), BK)
p_gkj = tl.max_contiguous(tl.multiple_of(g + (i_bh * T + i_t * BT + i_i * BC) * K + o_k, BK), BK)
p_dq = tl.make_block_ptr(dq + i_bh * T*K, (T, K), (K, 1), (i_t * BT + i_i * BC, i_k * BK), (BC, BK), (1, 0))
else:
o_dA = bos*H*BT + (i_t * BT + i_i * BC + tl.arange(0, BC)) * H*BT + i_h * BT + i_i * BC
p_kj = tl.max_contiguous(tl.multiple_of(k + (bos + i_t * BT + i_i * BC) * H*K + i_h * K + o_k, BK), BK)
p_gkj = tl.max_contiguous(tl.multiple_of(g + (bos + i_t * BT + i_i * BC) * H*K + i_h * K + o_k, BK), BK)
p_dq = tl.make_block_ptr(dq + (bos*H + i_h) * K, (T, K), (H*K, 1), (i_t * BT + i_i * BC, i_k * BK), (BC, BK), (1, 0))
for j in range(0, min(BC, T - i_t * BT - i_i * BC)):
# [BC,]
b_dA = tl.load(dA + o_dA + j, mask=m_dA, other=0)
# [BK,]
b_kj = tl.load(p_kj, mask=m_k, other=0).to(tl.float32)
b_gkj = tl.load(p_gkj, mask=m_k, other=0).to(tl.float32)
# [BC, BK]
m_i = o_i[:, None] >= j
# [BC, BK]
# (SY 09/17) important to not use bf16 here to have a good precision.
b_dq += tl.where(m_i, b_dA[:, None] * b_kj[None, :] * tl.exp(b_g - b_gkj[None, :]), 0.)
p_kj += K if HEAD_FIRST else H*K
p_gkj += K if HEAD_FIRST else H*K
tl.store(p_dq, b_dq.to(p_dq.dtype.element_ty), boundary_check=(0, 1))
tl.debug_barrier()
if HEAD_FIRST:
p_k = tl.make_block_ptr(k + i_bh * T*K, (T, K), (K, 1), (i_t * BT + i_i * BC, i_k * BK), (BC, BK), (1, 0))
p_gk = tl.make_block_ptr(g + i_bh * T*K, (T, K), (K, 1), (i_t * BT + i_i * BC, i_k * BK), (BC, BK), (1, 0))
else:
p_k = tl.make_block_ptr(k + (bos*H + i_h) * K, (T, K), (H*K, 1), (i_t * BT + i_i * BC, i_k * BK), (BC, BK), (1, 0))
p_gk = tl.make_block_ptr(g + (bos*H + i_h) * K, (T, K), (H*K, 1), (i_t * BT + i_i * BC, i_k * BK), (BC, BK), (1, 0))
# [BC, BK]
b_k = tl.load(p_k, boundary_check=(0, 1))
b_gk = tl.load(p_gk, boundary_check=(0, 1))
b_dk = tl.zeros([BC, BK], dtype=tl.float32)
NC = min(NC, tl.cdiv(T - i_t * BT, BC))
if i_i < NC - 1:
if HEAD_FIRST:
p_gn = tl.max_contiguous(tl.multiple_of(g + i_bh*T*K + (i_t * BT + i_i * BC + BC - 1)*K + o_k, BK), BK)
else:
p_gn = tl.max_contiguous(tl.multiple_of(g + bos*H*K + (i_t * BT + i_i * BC + BC - 1)*H*K + i_h*K + o_k, BK), BK)
# [BK,]
b_gn = tl.load(p_gn, mask=m_k, other=0)
for i_j in range(i_i + 1, NC):
if HEAD_FIRST:
p_q = tl.make_block_ptr(q + i_bh * T*K, (T, K), (K, 1), (i_t * BT + i_j * BC, i_k * BK), (BC, BK), (1, 0))
p_g = tl.make_block_ptr(g + i_bh * T*K, (T, K), (K, 1), (i_t * BT + i_j * BC, i_k * BK), (BC, BK), (1, 0))
p_dA = tl.make_block_ptr(dA + i_bh * T * BT, (BT, T), (1, BT), (i_i*BC, i_t*BT + i_j*BC), (BC, BC), (0, 1))
else:
p_q = tl.make_block_ptr(q + (bos*H+i_h)*K, (T, K), (H*K, 1), (i_t * BT + i_j * BC, i_k * BK), (BC, BK), (1, 0))
p_g = tl.make_block_ptr(g + (bos*H+i_h)*K, (T, K), (H*K, 1), (i_t * BT + i_j * BC, i_k * BK), (BC, BK), (1, 0))
p_dA = tl.make_block_ptr(dA + (bos*H+i_h)*BT, (BT, T), (1, H*BT), (i_i*BC, i_t*BT + i_j*BC), (BC, BC), (0, 1))
# [BC, BK]
b_q = tl.load(p_q, boundary_check=(0, 1))
b_g = tl.load(p_g, boundary_check=(0, 1))
b_qg = (b_q * tl.exp(b_g - b_gn[None, :]))
# [BC, BC]
b_dA = tl.load(p_dA, boundary_check=(0, 1))
# [BC, BK]
# (SY 09/17) important to not use bf16 here to have a good precision.
b_dk += tl.dot(b_dA, b_qg)
b_dk *= tl.exp(b_gn[None, :] - b_gk)
if HEAD_FIRST:
o_dA = i_bh * T * BT + (i_t * BT + i_i * BC) * BT + i_i * BC + tl.arange(0, BC)
p_qj = tl.max_contiguous(tl.multiple_of(q + (i_bh * T + i_t * BT + i_i * BC) * K + o_k, BK), BK)
p_gqj = tl.max_contiguous(tl.multiple_of(g + (i_bh * T + i_t * BT + i_i * BC) * K + o_k, BK), BK)
p_dk = tl.make_block_ptr(dk + i_bh*T*K, (T, K), (K, 1), (i_t * BT + i_i * BC, i_k * BK), (BC, BK), (1, 0))
else:
o_dA = bos*H*BT + (i_t * BT + i_i * BC) * H*BT + i_h * BT + i_i * BC + tl.arange(0, BC)
p_qj = tl.max_contiguous(tl.multiple_of(q + (bos + i_t * BT + i_i * BC) * H*K + i_h * K + o_k, BK), BK)
p_gqj = tl.max_contiguous(tl.multiple_of(g + (bos + i_t * BT + i_i * BC) * H*K + i_h * K + o_k, BK), BK)
p_dk = tl.make_block_ptr(dk + (bos*H+i_h)*K, (T, K), (H*K, 1), (i_t * BT + i_i * BC, i_k * BK), (BC, BK), (1, 0))
for j in range(0, min(BC, T - i_t * BT - i_i * BC)):
# [BC,]
b_dA = tl.load(dA + o_dA + j * (1 if HEAD_FIRST else H) * BT)
# [BK,]
b_qj = tl.load(p_qj, mask=m_k, other=0).to(tl.float32)
b_gqj = tl.load(p_gqj, mask=m_k, other=0).to(tl.float32)
# [BC, BK]
m_i = o_i[:, None] <= j
b_dk += tl.where(m_i, b_dA[:, None] * b_qj[None, :] * tl.exp(b_gqj[None, :] - b_gk), 0.)
p_qj += K if HEAD_FIRST else H*K
p_gqj += K if HEAD_FIRST else H*K
tl.store(p_dk, b_dk.to(p_dk.dtype.element_ty), boundary_check=(0, 1))
@triton.autotune(
configs=[
triton.Config({}, num_warps=1),
triton.Config({}, num_warps=2),
triton.Config({}, num_warps=4),
triton.Config({}, num_warps=8),
],
key=["BV", "BT"],
)
@triton.heuristics({'USE_OFFSETS': lambda args: args['offsets'] is not None})
@triton.jit
def chunk_gla_bwd_kernel_dA(
v,
do,
dA,
offsets,
indices,
scale,
T: tl.constexpr,
H: tl.constexpr,
V: tl.constexpr,
BT: tl.constexpr,
BV: tl.constexpr,
USE_OFFSETS: tl.constexpr,
HEAD_FIRST: tl.constexpr
):
i_t, i_bh = tl.program_id(0), tl.program_id(1)
i_b, i_h = i_bh // H, i_bh % H
if USE_OFFSETS:
i_n, i_t = tl.load(indices + i_t * 2).to(tl.int32), tl.load(indices + i_t * 2 + 1).to(tl.int32)
bos, eos = tl.load(offsets + i_n).to(tl.int32), tl.load(offsets + i_n + 1).to(tl.int32)
else:
bos, eos = i_b * T, i_b * T + T
T = eos - bos
b_dA = tl.zeros([BT, BT], dtype=tl.float32)
for i_v in range(tl.cdiv(V, BV)):
if HEAD_FIRST:
p_do = tl.make_block_ptr(do + i_bh * T*V, (T, V), (V, 1), (i_t * BT, i_v * BV), (BT, BV), (1, 0))
p_v = tl.make_block_ptr(v + i_bh * T*V, (V, T), (1, V), (i_v * BV, i_t * BT), (BV, BT), (0, 1))
else:
p_do = tl.make_block_ptr(do + (bos*H + i_h) * V, (T, V), (H*V, 1), (i_t * BT, i_v * BV), (BT, BV), (1, 0))
p_v = tl.make_block_ptr(v + (bos*H + i_h) * V, (V, T), (1, H*V), (i_v * BV, i_t * BT), (BV, BT), (0, 1))
b_v = tl.load(p_v, boundary_check=(0, 1))
b_do = tl.load(p_do, boundary_check=(0, 1))
b_dA += tl.dot(b_do, b_v)
if HEAD_FIRST:
p_dA = tl.make_block_ptr(dA + i_bh * T*BT, (T, BT), (BT, 1), (i_t * BT, 0), (BT, BT), (1, 0))
else:
p_dA = tl.make_block_ptr(dA + (bos * H + i_h) * BT, (T, BT), (H*BT, 1), (i_t * BT, 0), (BT, BT), (1, 0))
m_s = tl.arange(0, BT)[:, None] >= tl.arange(0, BT)[None, :]
b_dA = tl.where(m_s, b_dA * scale, 0.)
tl.store(p_dA, b_dA.to(p_dA.dtype.element_ty), boundary_check=(0, 1))
@triton.autotune(
configs=[
triton.Config({'BK': BK, 'BV': BV}, num_warps=num_warps)
for BK in [32, 64]
for BV in [64, 128]
for num_warps in [2, 4, 8]
],
key=["BT"],
)
@triton.heuristics({'USE_OFFSETS': lambda args: args['offsets'] is not None})
@triton.jit
def chunk_gla_bwd_kernel_dv(
q,
k,
g,
A,
do,
dh,
dv,
offsets,
split_indices,
scale,
T: tl.constexpr,
S: tl.constexpr,
H: tl.constexpr,
K: tl.constexpr,
V: tl.constexpr,
BT: tl.constexpr,
BK: tl.constexpr,
BV: tl.constexpr,
USE_OFFSETS: tl.constexpr,
HEAD_FIRST: tl.constexpr
):
i_v, i_i, i_sh = tl.program_id(0), tl.program_id(1), tl.program_id(2)
i_ss, i_h = i_sh // H, i_sh % H
if USE_OFFSETS:
i_n, i_s = tl.load(split_indices + i_ss * 2).to(tl.int32), tl.load(split_indices + i_ss * 2 + 1).to(tl.int32)
bos, eos = tl.load(offsets + i_n).to(tl.int32), tl.load(offsets + i_n + 1).to(tl.int32)
T = eos - bos
NS = tl.cdiv(T, S)
else:
NS = tl.cdiv(T, S)
i_n, i_s = i_ss // NS, i_ss % NS
bos, eos = i_n * T, i_n * T + T
i_nh = i_n * H + i_h
i_i = tl.cdiv(i_s * S, BT) + i_i
if i_i >= tl.cdiv(T, BT):
return
b_dv = tl.zeros([BT, BV], dtype=tl.float32)
for i_k in range(tl.cdiv(K, BK)):
o_k = i_k * BK + tl.arange(0, BK)
m_k = o_k < K
if HEAD_FIRST:
p_k = tl.make_block_ptr(k + i_nh * T*K, (T, K), (K, 1), (i_i * BT, i_k * BK), (BT, BK), (1, 0))
p_gk = tl.make_block_ptr(g + i_nh * T*K, (T, K), (K, 1), (i_i * BT, i_k * BK), (BT, BK), (1, 0))
p_gn = g + i_nh * T*K + (min(i_i * BT + BT, T) - 1) * K + o_k
else:
p_k = tl.make_block_ptr(k + (bos * H + i_h) * K, (T, K), (H*K, 1), (i_i * BT, i_k * BK), (BT, BK), (1, 0))
p_gk = tl.make_block_ptr(g + (bos * H + i_h) * K, (T, K), (H*K, 1), (i_i * BT, i_k * BK), (BT, BK), (1, 0))
p_gn = g + (bos + min(i_i * BT + BT, T) - 1)*H*K + i_h * K + o_k
p_gn = tl.max_contiguous(tl.multiple_of(p_gn, BK), BK)
b_gn = tl.load(p_gn, mask=m_k, other=0)
# [BT, BK]
b_k = tl.load(p_k, boundary_check=(0, 1))
b_gk = tl.load(p_gk, boundary_check=(0, 1))
b_kg = (b_k * tl.exp(b_gn[None, :] - b_gk)).to(b_k.dtype)
p_dh = tl.make_block_ptr(dh + i_sh * K*V, (K, V), (V, 1), (i_k * BK, i_v * BV), (BK, BV), (1, 0))
# [BK, BV]
b_dh = tl.load(p_dh, boundary_check=(0, 1)).to(tl.float32)
for i_j in range(tl.cdiv(min(i_s * S + S, T), BT) - 1, i_i, -1):
if HEAD_FIRST:
p_q = tl.make_block_ptr(q + i_nh * T*K, (K, T), (1, K), (i_k * BK, i_j * BT), (BK, BT), (0, 1))
p_g = tl.make_block_ptr(g + i_nh * T*K, (K, T), (1, K), (i_k * BK, i_j * BT), (BK, BT), (0, 1))
p_do = tl.make_block_ptr(do + i_nh * T*V, (T, V), (V, 1), (i_j * BT, i_v * BV), (BT, BV), (1, 0))
p_gn = g + (i_n * T + min(i_j * BT + BT, T) - 1) * K + o_k
else:
p_q = tl.make_block_ptr(q + (bos*H + i_h) * K, (K, T), (1, H*K), (i_k * BK, i_j * BT), (BK, BT), (0, 1))
p_g = tl.make_block_ptr(g + (bos*H + i_h) * K, (K, T), (1, H*K), (i_k * BK, i_j * BT), (BK, BT), (0, 1))
p_do = tl.make_block_ptr(do + (bos*H + i_h) * V, (T, V), (H*V, 1), (i_j * BT, i_v * BV), (BT, BV), (1, 0))
p_gn = g + (bos + min(i_j * BT + BT, T) - 1) * H*K + i_h * K + o_k
p_gn = tl.max_contiguous(tl.multiple_of(p_gn, BK), BK)
b_gn = tl.load(p_gn, mask=m_k, other=0.)
# [BK, BT]
b_q = tl.load(p_q, boundary_check=(0, 1))
b_g = tl.load(p_g, boundary_check=(0, 1))
b_q = (b_q * scale * tl.exp(b_g)).to(b_q.dtype)
# [BT, BV]
b_do = tl.load(p_do, boundary_check=(0, 1))
# [BK, BV]
b_dh = b_dh * tl.exp(b_gn)[:, None] + tl.dot(b_q, b_do)
# [BT, BV]
# (SY 09/17) it is ok to have bf16 interchunk gradient contribution here
b_dv += tl.dot(b_kg, b_dh.to(b_k.dtype))
if HEAD_FIRST:
p_A = tl.make_block_ptr(A + i_nh * T * BT, (BT, T), (1, BT), (0, i_i * BT), (BT, BT), (0, 1))
p_do = tl.make_block_ptr(do + i_nh * T*V, (T, V), (V, 1), (i_i * BT, i_v * BV), (BT, BV), (1, 0))
p_dv = tl.make_block_ptr(dv + i_nh * T*V, (T, V), (V, 1), (i_i * BT, i_v * BV), (BT, BV), (1, 0))
else:
p_A = tl.make_block_ptr(A + (bos * H + i_h) * BT, (BT, T), (1, H*BT), (0, i_i * BT), (BT, BT), (0, 1))
p_do = tl.make_block_ptr(do + (bos * H + i_h) * V, (T, V), (H*V, 1), (i_i * BT, i_v * BV), (BT, BV), (1, 0))
p_dv = tl.make_block_ptr(dv + (bos * H + i_h) * V, (T, V), (H*V, 1), (i_i * BT, i_v * BV), (BT, BV), (1, 0))
b_A = tl.load(p_A, boundary_check=(0, 1))
b_A = tl.where(tl.arange(0, BT)[:, None] <= tl.arange(0, BT)[None, :], b_A, 0.)
b_do = tl.load(p_do, boundary_check=(0, 1))
# (SY 09/17) important to disallow tf32 here to maintain a good precision.
b_dv += tl.dot(b_A, b_do.to(b_A.dtype), allow_tf32=False)
tl.store(p_dv, b_dv.to(p_dv.dtype.element_ty), boundary_check=(0, 1))
@triton.heuristics({
'USE_OFFSETS': lambda args: args['offsets'] is not None
})
@triton.autotune(
configs=[
triton.Config({'BK': BK, 'BV': BV}, num_warps=num_warps)
for BK in [32, 64]
for BV in [64, 128]
for num_warps in [2, 4, 8]
],
key=["BT"],
)
@triton.jit
def chunk_gla_bwd_kernel_inter(
q,
k,
v,
h,
g,
do,
dh,
dq,
dk,
dq2,
dk2,
dg,
offsets,
split_indices,
scale,
T: tl.constexpr,
S: tl.constexpr,
H: tl.constexpr,
K: tl.constexpr,
V: tl.constexpr,
BT: tl.constexpr,
BK: tl.constexpr,
BV: tl.constexpr,
USE_OFFSETS: tl.constexpr,
HEAD_FIRST: tl.constexpr
):
i_k, i_i, i_sh = tl.program_id(0), tl.program_id(1), tl.program_id(2)
i_ss, i_h = i_sh // H, i_sh % H
if USE_OFFSETS:
i_n, i_s = tl.load(split_indices + i_ss * 2).to(tl.int32), tl.load(split_indices + i_ss * 2 + 1).to(tl.int32)
bos, eos = tl.load(offsets + i_n).to(tl.int32), tl.load(offsets + i_n + 1).to(tl.int32)
T = eos - bos
NS = tl.cdiv(T, S)
else:
NS = tl.cdiv(T, S)
i_n, i_s = i_ss // NS, i_ss % NS
bos, eos = i_n * T, i_n * T + T
i_nh = i_n * H + i_h
i_i = tl.cdiv(i_s * S, BT) + i_i
o_k = i_k * BK + tl.arange(0, BK)
m_k = o_k < K
b_dq = tl.zeros([BT, BK], dtype=tl.float32)
b_dk = tl.zeros([BT, BK], dtype=tl.float32)
b_dgk = tl.zeros([BK,], dtype=tl.float32)
for i_v in range(tl.cdiv(V, BV)):
p_h = tl.make_block_ptr(h + i_sh * K*V, (K, V), (V, 1), (i_k * BK, i_v * BV), (BK, BV), (1, 0))
# [BK, BV]
b_h = tl.load(p_h, boundary_check=(0, 1)).to(tl.float32)
for i_j in range(tl.cdiv(i_s * S, BT), i_i):
if HEAD_FIRST:
p_k = tl.make_block_ptr(k + i_nh * T*K, (K, T), (1, K), (i_k * BK, i_j * BT), (BK, BT), (0, 1))
p_v = tl.make_block_ptr(v + i_nh * T*V, (T, V), (V, 1), (i_j * BT, i_v * BV), (BT, BV), (1, 0))
p_gk = tl.make_block_ptr(g + i_nh * T*K, (K, T), (1, K), (i_k * BK, i_j * BT), (BK, BT), (0, 1))
p_gn = g + i_nh * T*K + (min(i_j * BT + BT, T) - 1) * K + o_k
else:
p_k = tl.make_block_ptr(k + (bos * H + i_h)*K, (K, T), (1, H*K), (i_k * BK, i_j * BT), (BK, BT), (0, 1))
p_v = tl.make_block_ptr(v + (bos * H + i_h)*V, (T, V), (H*V, 1), (i_j * BT, i_v * BV), (BT, BV), (1, 0))
p_gk = tl.make_block_ptr(g + (bos * H + i_h)*K, (K, T), (1, H*K), (i_k * BK, i_j * BT), (BK, BT), (0, 1))
p_gn = g + (bos + min(i_j * BT + BT, T) - 1) * H*K + i_h * K + o_k
p_gn = tl.max_contiguous(tl.multiple_of(p_gn, BK), BK)
# [BK,]
b_gn = tl.load(p_gn, mask=m_k, other=0)
# [BK, BT]
b_k = tl.load(p_k, boundary_check=(0, 1))
b_gk = tl.load(p_gk, boundary_check=(0, 1))
b_kg = (b_k * tl.exp(b_gn[:, None] - b_gk)).to(b_k.dtype)
# [BT, BV]
b_v = tl.load(p_v, boundary_check=(0, 1))
# [BK, BV]
b_h = b_h * tl.exp(b_gn[:, None]) + tl.dot(b_kg, b_v)
p_dh = tl.make_block_ptr(dh + i_sh * K*V, (K, V), (V, 1), (i_k * BK, i_v * BV), (BK, BV), (1, 0))
# [BK, BV]
b_dh = tl.load(p_dh, boundary_check=(0, 1)).to(tl.float32)
for i_j in range(tl.cdiv(min(i_s * S + S, T), BT) - 1, i_i, -1):
if HEAD_FIRST:
p_q = tl.make_block_ptr(q + i_nh * T*K, (K, T), (1, K), (i_k * BK, i_j * BT), (BK, BT), (0, 1))
p_g = tl.make_block_ptr(g + i_nh * T*K, (K, T), (1, K), (i_k * BK, i_j * BT), (BK, BT), (0, 1))
p_do = tl.make_block_ptr(do + i_nh * T*V, (T, V), (V, 1), (i_j * BT, i_v * BV), (BT, BV), (1, 0))
p_gn = g + (i_n * T + min(i_j * BT + BT, T) - 1) * K + o_k
else:
p_q = tl.make_block_ptr(q + (bos*H + i_h) * K, (K, T), (1, H*K), (i_k * BK, i_j * BT), (BK, BT), (0, 1))
p_g = tl.make_block_ptr(g + (bos*H + i_h) * K, (K, T), (1, H*K), (i_k * BK, i_j * BT), (BK, BT), (0, 1))
p_do = tl.make_block_ptr(do + (bos*H + i_h) * V, (T, V), (H*V, 1), (i_j * BT, i_v * BV), (BT, BV), (1, 0))
p_gn = g + (bos + min(i_j * BT + BT, T) - 1) * H*K + i_h * K + o_k
p_gn = tl.max_contiguous(tl.multiple_of(p_gn, BK), BK)
b_gn = tl.load(p_gn, mask=m_k, other=0.)
# [BK, BT]
b_q = tl.load(p_q, boundary_check=(0, 1))
b_g = tl.load(p_g, boundary_check=(0, 1))
b_q = (b_q * scale * tl.exp(b_g)).to(b_q.dtype)
# [BT, BV]
b_do = tl.load(p_do, boundary_check=(0, 1))
# [BK, BV]
b_dh = b_dh * tl.exp(b_gn)[:, None] + tl.dot(b_q, b_do)
if HEAD_FIRST:
p_v = tl.make_block_ptr(v + i_nh * T*V, (T, V), (V, 1), (i_i * BT, i_v * BV), (BT, BV), (1, 0))
p_do = tl.make_block_ptr(do + i_nh * T*V, (T, V), (V, 1), (i_i * BT, i_v * BV), (BT, BV), (1, 0))
else:
p_v = tl.make_block_ptr(v + (bos*H + i_h) * V, (T, V), (H*V, 1), (i_i * BT, i_v * BV), (BT, BV), (1, 0))
p_do = tl.make_block_ptr(do + (bos*H + i_h) * V, (T, V), (H*V, 1), (i_i * BT, i_v * BV), (BT, BV), (1, 0))
# [BT, BV]
b_v = tl.load(p_v, boundary_check=(0, 1))
b_do = tl.load(p_do, boundary_check=(0, 1))
# [BK]
b_dgk += tl.sum(b_h * b_dh, axis=1)
# [BT, BK]
b_dq += tl.dot(b_do, tl.trans(b_h).to(b_do.dtype))
b_dk += tl.dot(b_v, tl.trans(b_dh).to(b_v.dtype))
if HEAD_FIRST:
p_gk = tl.make_block_ptr(g + i_nh * T*K, (T, K), (K, 1), (i_i * BT, i_k * BK), (BT, BK), (1, 0))
p_gn = g + i_nh * T*K + (min(T, i_i * BT + BT)-1) * K + o_k
else:
p_gk = tl.make_block_ptr(g + (bos*H+i_h)*K, (T, K), (H*K, 1), (i_i * BT, i_k * BK), (BT, BK), (1, 0))
p_gn = g + (bos + min(T, i_i * BT + BT)-1) * H*K + i_h * K + o_k
p_gn = tl.max_contiguous(tl.multiple_of(p_gn, BK), BK)
b_gn = tl.load(p_gn, mask=m_k, other=0)
# [BT, BK]
b_gk = tl.load(p_gk, boundary_check=(0, 1))
b_dq = b_dq * tl.exp(b_gk) * scale
b_dk = b_dk * tl.exp(b_gn[None, :] - b_gk)
b_dgk *= tl.exp(b_gn)
if HEAD_FIRST:
p_q = tl.make_block_ptr(q + i_nh * T*K, (T, K), (K, 1), (i_i * BT, i_k * BK), (BT, BK), (1, 0))
p_k = tl.make_block_ptr(k + i_nh * T*K, (T, K), (K, 1), (i_i * BT, i_k * BK), (BT, BK), (1, 0))
p_dq = tl.make_block_ptr(dq + i_nh * T*K, (T, K), (K, 1), (i_i * BT, i_k * BK), (BT, BK), (1, 0))
p_dk = tl.make_block_ptr(dk + i_nh * T*K, (T, K), (K, 1), (i_i * BT, i_k * BK), (BT, BK), (1, 0))
else:
p_q = tl.make_block_ptr(q + (bos*H+i_h)*K, (T, K), (H*K, 1), (i_i * BT, i_k * BK), (BT, BK), (1, 0))
p_k = tl.make_block_ptr(k + (bos*H+i_h)*K, (T, K), (H*K, 1), (i_i * BT, i_k * BK), (BT, BK), (1, 0))
p_dq = tl.make_block_ptr(dq + (bos*H+i_h)*K, (T, K), (H*K, 1), (i_i * BT, i_k * BK), (BT, BK), (1, 0))
p_dk = tl.make_block_ptr(dk + (bos*H+i_h)*K, (T, K), (H*K, 1), (i_i * BT, i_k * BK), (BT, BK), (1, 0))
b_q = tl.load(p_q, boundary_check=(0, 1))
b_k = tl.load(p_k, boundary_check=(0, 1))
b_dgk += tl.sum(b_dk * b_k, axis=0)
b_dq += tl.load(p_dq, boundary_check=(0, 1))
b_dk += tl.load(p_dk, boundary_check=(0, 1))
b_dg = b_q * b_dq - b_k * b_dk
# tl.debug_barrier()
b_dg = b_dg - tl.cumsum(b_dg, axis=0) + tl.sum(b_dg, axis=0)[None, :] + b_dgk[None, :]
# Buggy due to strange triton compiler issue.
# m_s = tl.where(tl.arange(0, BT)[:, None] <= tl.arange(0, BT)[None, :], 1., 0.)
# b_dg = tl.dot(m_s, b_dg, allow_tf32=False) + b_dgk[None, :]
if HEAD_FIRST:
p_dq = tl.make_block_ptr(dq2 + i_nh * T*K, (T, K), (K, 1), (i_i * BT, i_k * BK), (BT, BK), (1, 0))
p_dk = tl.make_block_ptr(dk2 + i_nh * T*K, (T, K), (K, 1), (i_i * BT, i_k * BK), (BT, BK), (1, 0))
p_dg = tl.make_block_ptr(dg + i_nh * T*K, (T, K), (K, 1), (i_i * BT, i_k * BK), (BT, BK), (1, 0))
else:
p_dq = tl.make_block_ptr(dq2 + (bos * H + i_h) * K, (T, K), (H*K, 1), (i_i * BT, i_k * BK), (BT, BK), (1, 0))
p_dk = tl.make_block_ptr(dk2 + (bos * H + i_h) * K, (T, K), (H*K, 1), (i_i * BT, i_k * BK), (BT, BK), (1, 0))
p_dg = tl.make_block_ptr(dg + (bos * H + i_h) * K, (T, K), (H*K, 1), (i_i * BT, i_k * BK), (BT, BK), (1, 0))
tl.store(p_dq, b_dq.to(p_dq.dtype.element_ty), boundary_check=(0, 1))
tl.store(p_dk, b_dk.to(p_dk.dtype.element_ty), boundary_check=(0, 1))
tl.store(p_dg, b_dg.to(p_dg.dtype.element_ty), boundary_check=(0, 1))
def chunk_gla_fwd_intra_gk(
q: torch.Tensor,
k: torch.Tensor,
g: torch.Tensor,
scale: float,
offsets: Optional[torch.LongTensor] = None,
indices: Optional[torch.LongTensor] = None,
head_first: bool = True,
chunk_size: int = 64
):
if head_first:
B, H, T, K = k.shape
else:
B, T, H, K = k.shape
BT = min(chunk_size, max(16, triton.next_power_of_2(T)))
NT = triton.cdiv(T, BT) if offsets is None else len(indices)
BC = min(16, BT)
NC = triton.cdiv(BT, BC)
A = q.new_empty(B, *((H, T) if head_first else (T, H)), BT, dtype=torch.float)
grid = (NT, NC * NC, B * H)
chunk_gla_fwd_A_kernel_intra_sub_inter[grid](
q,
k,
g,
A,
offsets,
indices,
scale,
T=T,
H=H,
K=K,
BT=BT,
BC=BC,
NC=NC,
HEAD_FIRST=head_first
)
grid = (NT, NC, B * H)
# load the entire [BC, K] blocks into SRAM at once
if K <= 256:
BK = triton.next_power_of_2(K)
chunk_gla_fwd_A_kernel_intra_sub_intra[grid](
q,
k,
g,
A,
offsets,
indices,
scale,
T=T,
H=H,
K=K,
BT=BT,
BC=BC,
BK=BK,
HEAD_FIRST=head_first
)
# split then merge
else:
BK = min(128, triton.next_power_of_2(K))
NK = triton.cdiv(K, BK)
A_intra = q.new_empty(NK, B, *((H, T) if head_first else (T, H)), BC, dtype=torch.float)
grid = (NK, NT * NC, B * H)
chunk_gla_fwd_A_kernel_intra_sub_intra_split[grid](
q,
k,
g,
A_intra,
offsets,
indices,
scale,
B=B,
T=T,
H=H,
K=K,
BT=BT,
BC=BC,
BK=BK,
NC=NC,
HEAD_FIRST=head_first
)
grid = (NT, NC, B * H)
chunk_gla_fwd_A_kernel_intra_sub_intra_merge[grid](
A_intra,
A,
offsets,
indices,
B=B,
T=T,
H=H,
BT=BT,
BC=BC,
NK=NK,
HEAD_FIRST=head_first
)
return A
def chunk_gla_fwd_o_gk(
q: torch.Tensor,
k: torch.Tensor,
v: torch.Tensor,
g: torch.Tensor,
A: torch.Tensor,
h: torch.Tensor,
scale: float,
offsets: Optional[torch.LongTensor] = None,
split_offsets: Optional[torch.LongTensor] = None,
split_indices: Optional[torch.LongTensor] = None,
head_first: bool = True,
chunk_size: int = 64,
split_size: int = 256
):
if head_first:
B, H, T, K, V = *q.shape, v.shape[-1]
else:
B, T, H, K, V = *q.shape, v.shape[-1]
S = split_size
BT = min(chunk_size, max(16, triton.next_power_of_2(T)))
NS = (B * triton.cdiv(T, S)) if offsets is None else split_offsets[-1]
NC = triton.cdiv(S, BT)
o = torch.empty_like(v)
def grid(meta): return (triton.cdiv(V, meta['BV']), NC, NS * H)
chunk_gla_fwd_kernel_o[grid](
q,
k,
v,
g,
h,
o,
A,
offsets,
split_indices,
scale,
T=T,
S=S,
H=H,
K=K,
V=V,
BT=BT,
HEAD_FIRST=head_first
)
return o
def chunk_gla_bwd_dA(
v: torch.Tensor,
do: torch.Tensor,
scale: float,
offsets: Optional[torch.LongTensor] = None,
indices: Optional[torch.LongTensor] = None,
head_first: bool = True,
chunk_size: int = 64
):
if head_first:
B, H, T, V = v.shape
else:
B, T, H, V = v.shape
BT = min(chunk_size, max(16, triton.next_power_of_2(T)))
NT = triton.cdiv(T, BT) if offsets is None else len(indices)
BV = min(64, triton.next_power_of_2(V))
dA = v.new_empty(B, *((H, T) if head_first else (T, H)), BT, dtype=torch.float)
grid = (NT, B * H)
chunk_gla_bwd_kernel_dA[grid](
v,
do,
dA,
offsets,
indices,
scale,
T=T,
H=H,
V=V,
BT=BT,
BV=BV,
HEAD_FIRST=head_first
)
return dA
def chunk_gla_bwd_dv(
q: torch.Tensor,
k: torch.Tensor,
g: torch.Tensor,
A: torch.Tensor,
do: torch.Tensor,
dh: torch.Tensor,
scale: float,
offsets: Optional[torch.LongTensor] = None,
split_offsets: Optional[torch.LongTensor] = None,
split_indices: Optional[torch.LongTensor] = None,
head_first: bool = True,
chunk_size: int = 64,
split_size: int = 256
):
if head_first:
B, H, T, K, V = *k.shape, do.shape[-1]
else:
B, T, H, K, V = *k.shape, do.shape[-1]
S = split_size
BT = min(chunk_size, max(16, triton.next_power_of_2(T)))
NS = (B * triton.cdiv(T, S)) if offsets is None else split_offsets[-1]
NC = triton.cdiv(S, BT)
dv = torch.empty_like(do)
def grid(meta): return (triton.cdiv(V, meta['BV']), NC, NS * H)
chunk_gla_bwd_kernel_dv[grid](
q,
k,
g,
A,
do,
dh,
dv,
offsets,
split_indices,
scale,
T=T,
S=S,
H=H,
K=K,
V=V,
BT=BT,
HEAD_FIRST=head_first
)
return dv
def chunk_gla_bwd_dqk_intra(
q: torch.Tensor,
k: torch.Tensor,
g: torch.Tensor,
dA: torch.Tensor,
offsets: Optional[torch.LongTensor] = None,
indices: Optional[torch.LongTensor] = None,
head_first: bool = True,
chunk_size: int = 64
):
if head_first:
B, H, T, K = q.shape
else:
B, T, H, K = q.shape
BT = min(chunk_size, max(16, triton.next_power_of_2(T)))
BC = min(16, BT)
BK = min(64, triton.next_power_of_2(K))
NT = triton.cdiv(T, BT) if offsets is None else len(indices)
NC = triton.cdiv(BT, BC)
NK = triton.cdiv(K, BK)
dq = torch.empty_like(q, dtype=torch.float)
dk = torch.empty_like(k, dtype=torch.float)
grid = (NK, NT * NC, B * H)
chunk_gla_bwd_kernel_intra[grid](
q,
k,
g,
dA,
dq,
dk,
offsets,
indices,
T=T,
H=H,
K=K,
BT=BT,
BC=BC,
BK=BK,
NC=NC,
HEAD_FIRST=head_first
)
return dq, dk
def chunk_gla_bwd_dqkg(
q: torch.Tensor,
k: torch.Tensor,
v: torch.Tensor,
h: torch.Tensor,
g: torch.Tensor,
do: torch.Tensor,
dh: torch.Tensor,
dq: torch.Tensor,
dk: torch.Tensor,
scale: float,
offsets: Optional[torch.LongTensor] = None,
split_offsets: Optional[torch.LongTensor] = None,
split_indices: Optional[torch.LongTensor] = None,
head_first: bool = True,
chunk_size: int = 64,
split_size: int = 256
):
if head_first:
B, H, T, K, V = *k.shape, v.shape[-1]
else:
B, T, H, K, V = *k.shape, v.shape[-1]
S = split_size
BT = min(chunk_size, max(16, triton.next_power_of_2(T)))
NS = (B * triton.cdiv(T, S)) if offsets is None else split_offsets[-1]
NC = triton.cdiv(S, BT)
dg = torch.empty_like(g)
# work around triton compiler bugs.
dq2 = torch.empty_like(dq)
dk2 = torch.empty_like(dk)
def grid(meta): return (triton.cdiv(K, meta['BK']), NC, NS * H)
chunk_gla_bwd_kernel_inter[grid](
q,
k,
v,
h,
g,
do,
dh,
dq,
dk,
dq2,
dk2,
dg,
offsets,
split_indices,
scale,
T=T,
S=S,
H=H,
K=K,
V=V,
BT=BT,
HEAD_FIRST=head_first
)
return dq2, dk2, dg
def chunk_gla_fwd(
q: torch.Tensor,
k: torch.Tensor,
v: torch.Tensor,
g: torch.Tensor,
g_cumsum: Optional[torch.Tensor],
scale: float,
initial_state: torch.Tensor,
output_final_state: bool,
offsets: Optional[torch.LongTensor] = None,
indices: Optional[torch.LongTensor] = None,
split_offsets: Optional[torch.LongTensor] = None,
split_indices: Optional[torch.LongTensor] = None,
head_first: bool = True,
chunk_size: int = 64,
split_size: int = 256
) -> Tuple[torch.Tensor, torch.Tensor, torch.Tensor]:
T = q.shape[2] if head_first else q.shape[1]
S = split_size
BT = min(chunk_size, max(16, triton.next_power_of_2(T)))
if g_cumsum is None:
g_cumsum = chunk_local_cumsum(g, BT, offsets=offsets, head_first=head_first)
h, ht = chunk_fwd_h(
k=k,
v=v,
g=None,
gk=g_cumsum,
gv=None,
h0=initial_state,
output_final_state=output_final_state,
offsets=offsets,
split_offsets=split_offsets,
split_indices=split_indices,
head_first=head_first,
chunk_size=BT,
split_size=S,
states_in_fp32=True
)
# the intra A is kept in fp32
# the computation has very marginal effect on the entire throughput
A = chunk_gla_fwd_intra_gk(
q=q,
k=k,
g=g_cumsum,
scale=scale,
offsets=offsets,
indices=indices,
head_first=head_first,
chunk_size=BT
)
o = chunk_gla_fwd_o_gk(
q=q,
k=k,
v=v,
g=g_cumsum,
A=A,
h=h,
scale=scale,
offsets=offsets,
split_offsets=split_offsets,
split_indices=split_indices,
head_first=head_first,
chunk_size=BT,
split_size=S
)
return g_cumsum, A, h, ht, o
def chunk_gla_bwd(
q: torch.Tensor,
k: torch.Tensor,
v: torch.Tensor,
g: torch.Tensor,
g_cumsum: Optional[torch.Tensor],
scale: float,
initial_state: torch.Tensor,
h: torch.Tensor,
A: torch.Tensor,
do: torch.Tensor,
dht: torch.Tensor,
offsets: Optional[torch.LongTensor] = None,
indices: Optional[torch.LongTensor] = None,
split_offsets: Optional[torch.LongTensor] = None,
split_indices: Optional[torch.LongTensor] = None,
head_first: bool = True,
chunk_size: int = 64,
split_size: int = 256
):
T = q.shape[2] if head_first else q.shape[1]
S = split_size
BT = min(chunk_size, max(16, triton.next_power_of_2(T)))
if g_cumsum is None:
g_cumsum = chunk_local_cumsum(g, BT, offsets=offsets, head_first=head_first)
if h is None:
h, ht = chunk_fwd_h(
k=k,
v=v,
g=None,
gk=g_cumsum,
gv=None,
h0=initial_state,
output_final_state=False,
offsets=offsets,
split_offsets=split_offsets,
split_indices=split_indices,
head_first=head_first,
chunk_size=BT,
split_size=S,
states_in_fp32=True
)
dh, dh0 = chunk_bwd_dh(
q=q,
k=k,
v=v,
g=None,
gk=g_cumsum,
gv=None,
do=do,
h0=initial_state,
dht=dht,
scale=scale,
offsets=offsets,
split_offsets=split_offsets,
split_indices=split_indices,
head_first=head_first,
chunk_size=BT,
split_size=S,
states_in_fp32=True
)
dv = chunk_gla_bwd_dv(
q=q,
k=k,
g=g_cumsum,
A=A,
do=do,
dh=dh,
scale=scale,
offsets=offsets,
split_offsets=split_offsets,
split_indices=split_indices,
head_first=head_first,
chunk_size=BT,
split_size=S
)
# dq dk in fp32
dA = chunk_gla_bwd_dA(
v=v,
do=do,
scale=scale,
offsets=offsets,
indices=indices,
head_first=head_first,
chunk_size=BT
)
dq, dk = chunk_gla_bwd_dqk_intra(
q=q,
k=k,
g=g_cumsum,
dA=dA,
offsets=offsets,
indices=indices,
head_first=head_first,
chunk_size=BT
)
dq, dk, dg = chunk_gla_bwd_dqkg(
q=q,
k=k,
v=v,
h=h,
g=g_cumsum,
do=do,
dh=dh,
dq=dq,
dk=dk,
scale=scale,
offsets=offsets,
split_offsets=split_offsets,
split_indices=split_indices,
head_first=head_first,
chunk_size=BT,
split_size=S
)
return dq, dk, dv, dg, dh0
class ChunkGLAFunction(torch.autograd.Function):
@staticmethod
@contiguous
def forward(
ctx,
q,
k,
v,
g,
scale,
initial_state,
output_final_state,
offsets,
head_first
):
T = q.shape[2] if head_first else q.shape[1]
chunk_size = min(64, max(16, triton.next_power_of_2(T)))
split_size = max(chunk_size, min(256, triton.next_power_of_2(T)))
# 2-d indices denoting the offsets of chunks in each sequence
# for example, if the passed `offsets` is [0, 100, 356] and `chunk_size` is 64,
# then there are 2 and 4 chunks in the 1st and 2nd sequences respectively, and `indices` will be
# [[0, 0], [0, 1], [1, 0], [1, 1], [1, 2], [1, 3]]
indices, split_offsets, split_indices = None, None, None
if offsets is not None:
indices = torch.cat([torch.arange(n) for n in triton.cdiv(offsets[1:] - offsets[:-1], chunk_size).tolist()])
indices = torch.stack([indices.eq(0).cumsum(0) - 1, indices], 1).to(offsets)
split_offsets = offsets.new_tensor([0] + [triton.cdiv(T, split_size) for _ in range(offsets[-1])]).cumsum(0)
split_indices = torch.cat([torch.arange(n) for n in (split_offsets[1:] - split_offsets[:-1]).tolist()])
split_indices = torch.stack([split_indices.eq(0).cumsum(0) - 1, split_indices], 1).to(split_offsets)
g_cumsum, A, h, ht, o = chunk_gla_fwd(
q=q,
k=k,
v=v,
g=g,
g_cumsum=None,
scale=scale,
initial_state=initial_state,
output_final_state=output_final_state,
offsets=offsets,
indices=indices,
split_offsets=split_offsets,
split_indices=split_indices,
head_first=head_first,
chunk_size=chunk_size,
split_size=split_size
)
# recompute g_cumsum in bwd pass
if g.dtype != torch.float:
g_cumsum = None
else:
g = None
ctx.save_for_backward(q, k, v, g, g_cumsum, initial_state, A)
ctx.chunk_size = chunk_size
ctx.scale = scale
ctx.offsets = offsets
ctx.indices = indices
ctx.head_first = head_first
return o, ht
@staticmethod
@contiguous
def backward(ctx, do, dht):
q, k, v, g, g_cumsum, initial_state, A = ctx.saved_tensors
chunk_size, scale, offsets, indices, head_first = ctx.chunk_size, ctx.scale, ctx.offsets, ctx.indices, ctx.head_first
dq, dk, dv, dg, dh0 = chunk_gla_bwd(
q=q,
k=k,
v=v,
g=g,
g_cumsum=g_cumsum,
scale=scale,
h=None,
A=A,
initial_state=initial_state,
do=do,
dht=dht,
offsets=offsets,
indices=indices,
head_first=head_first,
chunk_size=chunk_size
)
return dq.to(q), dk.to(k), dv.to(v), dg, None, dh0, None, None, None
def chunk_gla(
q: torch.Tensor,
k: torch.Tensor,
v: torch.Tensor,
g: torch.Tensor,
scale: Optional[int] = None,
initial_state: torch.Tensor = None,
output_final_state: bool = False,
offsets: Optional[torch.LongTensor] = None,
head_first: bool = True
) -> Tuple[torch.Tensor, torch.Tensor]:
r"""
Args:
q (torch.Tensor):
queries of shape `[B, H, T, K]` if `head_first=True` else `[B, T, H, K]`.
k (torch.Tensor):
keys of shape `[B, H, T, K]` if `head_first=True` else `[B, T, H, K]`.
v (torch.Tensor):
values of shape `[B, H, T, V]` if `head_first=True` else `[B, T, H, V]`.
g (torch.Tensor):
Forget gates of shape `[B, H, T, K]` if `head_first=True` else `[B, T, H, K]` applied to keys.
scale (Optional[int]):
Scale factor for the attention scores.
If not provided, it will default to `1 / sqrt(K)`. Default: `None`.
initial_state (Optional[torch.Tensor]):
Initial state of shape `[N, H, K, V]` for `N` input sequences.
For equal-length input sequences, `N` equals the batch size `B`.
Default: `None`.
output_final_state (Optional[bool]):
Whether to output the final state of shape `[N, H, K, V]`. Default: `False`.
offsets (Optional[torch.LongTensor]):
Offsets of shape `[N+1]` defining the bos/eos positions of `N` variable-length sequences in the batch.
For example,
if `offsets` is `[0, 1, 3, 6, 10, 15]`, there are `N=5` sequences with lengths 1, 2, 3, 4 and 5 respectively.
If provided, the inputs are concatenated and the batch size `B` is expected to be 1.
Default: `None`.
head_first (Optional[bool]):
Whether the inputs are in the head-first format, which is not supported for variable-length inputs.
Default: `True`.
Returns:
o (torch.Tensor):
Outputs of shape `[B, H, T, V]` if `head_first=True` else `[B, T, H, V]`.
final_state (torch.Tensor):
Final state of shape `[N, H, K, V]` if `output_final_state=True` else `None`.
Examples::
>>> import torch
>>> import torch.nn.functional as F
>>> from einops import rearrange
>>> from fla.ops.gla import chunk_gla
# inputs with equal lengths
>>> B, T, H, K, V = 4, 2048, 4, 512, 512
>>> q = torch.randn(B, T, H, K, device='cuda')
>>> k = torch.randn(B, T, H, K, device='cuda')
>>> v = torch.randn(B, T, H, V, device='cuda')
>>> g = F.logsigmoid(torch.randn(B, T, H, K, device='cuda'))
>>> h0 = torch.randn(B, H, K, V, device='cuda')
>>> o, ht = chunk_gla(q, k, v, g,
initial_state=h0,
output_final_state=True,
head_first=False)
# for variable-length inputs, the batch size `B` is expected to be 1 and `offsets` is required
>>> q, k, v, g = map(lambda x: rearrange(x, 'b t h d -> 1 (b t) h d'), (q, k, v, g))
# for a batch with 4 sequences, offsets with 5 start/end positions are expected
>>> offsets = q.new_tensor([0, 2048, 4096, 6144, 8192], dtype=torch.long)
>>> o_var, ht_var = chunk_gla(q, k, v, g,
initial_state=h0,
output_final_state=True,
offsets=offsets,
head_first=False)
>>> assert o.allclose(o_var.view(o.shape))
>>> assert ht.allclose(ht_var)
"""
if offsets is not None:
if q.shape[0] != 1:
raise ValueError(f"The batch size is expected to be 1 rather than {q.shape[0]} when using `offsets`."
f"Please flatten variable-length inputs before processing.")
if head_first:
raise RuntimeError("Sequences with variable lengths are not supported for head-first mode")
if initial_state is not None and initial_state.shape[0] != len(offsets) - 1:
raise ValueError(f"The number of initial states is expected to be equal to the number of input sequences, "
f"i.e., {len(offsets) - 1} rather than {initial_state.shape[0]}.")
if scale is None:
scale = q.shape[-1] ** -0.5
o, final_state = ChunkGLAFunction.apply(q, k, v, g, scale, initial_state, output_final_state, offsets, head_first)
return o, final_state
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment