Skip to content

[BugFix][Quant] Guard scale tails and FP32 absmax #15

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
JayceSu98 wants to merge 1 commit into
base: main
Choose a base branch
from
Open

[BugFix][Quant] Guard scale tails and FP32 absmax #15

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
9 changes: 7 additions & 2 deletions tile_kernels/quant/cast_back_kernel.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -60,11 +60,16 @@ def cast_back_kernel(
out_fragment = T.alloc_fragment((TILE_M, TILE_K), out_dtype)

T.copy(x[pid_token * TILE_M, pid_hidden * TILE_K], x_shared, disable_tma=True)
num_sf_token_blocks = T.ceildiv(num_tokens, num_per_tokens)
num_sf_channel_blocks = T.ceildiv(hidden, num_per_channels)
for i, j in T.Parallel(T.ceildiv(TILE_M, num_per_tokens), T.ceildiv(TILE_K, num_per_channels)):
token_index = pid_token * TILE_M // num_per_tokens + i
channel_index = pid_hidden * TILE_K // num_per_channels + j
sf = load_sf(x_sf, token_index, channel_index, in_config)
sf_shared[i, j] = transform_sf(sf, in_config)
if token_index < num_sf_token_blocks and channel_index < num_sf_channel_blocks:
sf = load_sf(x_sf, token_index, channel_index, in_config)
sf_shared[i, j] = transform_sf(sf, in_config)
else:
sf_shared[i, j] = 0

for i, j in T.Parallel(TILE_M, TILE_K):
out_fragment[i, j] = x_shared[i, j] * sf_shared[i // num_per_tokens, j // num_per_channels]
Expand Down
18 changes: 12 additions & 6 deletions tile_kernels/quant/per_block_cast_lossless_kernel.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -88,10 +88,13 @@ def per_block_cast_lossless_kernel(

# Load scaling factor of x to fragment
T.fill(x_sf_fragment, 0)
num_in_sf_blocks_m = T.ceildiv(num_tokens, in_config.sf_block[0])
num_in_sf_blocks_k = T.ceildiv(hidden, in_config.sf_block[1])
for i, j in T.Parallel(num_in_sf_per_block_m, num_in_sf_per_block_k):
m_idx = pid_token * block_m // in_config.sf_block[0] + i
k_idx = pid_hidden * block_k // in_config.sf_block[1] + j
x_sf_fragment[i, j] = load_sf(x_sf, m_idx, k_idx, in_config)
if m_idx < num_in_sf_blocks_m and k_idx < num_in_sf_blocks_k:
x_sf_fragment[i, j] = load_sf(x_sf, m_idx, k_idx, in_config)

# Alloc fragments
x_sf_uint32_fragment = T.alloc_fragment((num_in_sf_per_block_m, num_in_sf_per_block_k), T.uint32)
Expand Down Expand Up @@ -137,14 +140,17 @@ def per_block_cast_lossless_kernel(
x_out_fragment[i, j] = T.cast(T.float32(x_in_shared[i, j]) * sf, out_config.dtype)

# Store scaling factor back to global memory
num_out_sf_blocks_m = T.ceildiv(num_tokens, out_config.sf_block[0])
num_out_sf_blocks_k = T.ceildiv(hidden, out_config.sf_block[1])
for i, j in T.Parallel(num_out_sf_per_block_m, num_out_sf_per_block_k):
sf_m_idx = pid_token * num_out_sf_per_block_m + i
sf_k_idx = pid_hidden * num_out_sf_per_block_k + j
if out_config.use_packed_ue8m0:
sf = T.uint8(out_sf_uint32_fragment[i, j])
else:
sf = transform_sf_to_fp32(out_sf_uint32_fragment[i, j])
store_sf(out_sf, sf, sf_m_idx, sf_k_idx, out_config)
if sf_m_idx < num_out_sf_blocks_m and sf_k_idx < num_out_sf_blocks_k:
if out_config.use_packed_ue8m0:
sf = T.uint8(out_sf_uint32_fragment[i, j])
else:
sf = transform_sf_to_fp32(out_sf_uint32_fragment[i, j])
store_sf(out_sf, sf, sf_m_idx, sf_k_idx, out_config)

T.copy(x_out_fragment, out[pid_token * block_m: (pid_token + 1) * block_m, pid_hidden * block_k: (pid_hidden + 1) * block_k])

Expand Down
8 changes: 5 additions & 3 deletions tile_kernels/quant/per_token_cast_kernel.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -115,7 +115,8 @@ def per_token_cast_kernel(
# Store SF
m_idx = pid_token * block_m + i
k_idx = pid_hidden * num_groups + j
store_sf(out_sf, sf, m_idx, k_idx, out_config)
if m_idx < num_tokens:
store_sf(out_sf, sf, m_idx, k_idx, out_config)
sf_inv_fragment[i, j] = sf_inv

# Store casted values
Expand All @@ -131,7 +132,7 @@ def per_token_cast_kernel(
sf = load_sf(out_sf, pid_token * block_m + i, pid_hidden * num_groups + j, out_config)
sf_inv_fragment[i, j] = 1 / sf
else:
amax_fragment = T.alloc_fragment((block_m, num_groups), in_config.dtype)
amax_fragment = T.alloc_fragment((block_m, num_groups), T.float32)
x_fragment_reshaped = T.reshape(x_fragment, [block_m, num_groups, num_per_channels])
# Reduce SF
T.reduce_absmax(x_fragment_reshaped, amax_fragment, dim=2)
Expand All @@ -142,7 +143,8 @@ def per_token_cast_kernel(
# Store SF
m_idx = pid_token * block_m + i
k_idx = pid_hidden * num_groups + j
store_sf(out_sf, sf, m_idx, k_idx, out_config)
if m_idx < num_tokens:
store_sf(out_sf, sf, m_idx, k_idx, out_config)
sf_inv_fragment[i, j] = sf_inv

# Store casted values
Expand Down
12 changes: 8 additions & 4 deletions tile_kernels/quant/per_token_cast_to_e5m6_kernel.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -119,7 +119,7 @@ def per_token_cast_to_e5m6_kernel(
# Copy input into registers
T.copy(x[pid_token * block_m, pid_hidden * block_k], x_fragment)

amax_fragment = T.alloc_fragment((block_m, num_groups), in_config.dtype)
amax_fragment = T.alloc_fragment((block_m, num_groups), T.float32)
x_fragment_reshaped = T.reshape(x_fragment, [block_m, num_groups, num_per_channels])
# Reduce SF
T.reduce_absmax(x_fragment_reshaped, amax_fragment, dim=2)
Expand All @@ -130,7 +130,8 @@ def per_token_cast_to_e5m6_kernel(
# Store SF
m_idx = pid_token * block_m + i
k_idx = pid_hidden * num_groups + j
store_sf(out_sf, sf, m_idx, k_idx, out_config)
if m_idx < num_tokens:
store_sf(out_sf, sf, m_idx, k_idx, out_config)
sf_inv_fragment[i, j] = sf_inv

T.annotate_layout({
Expand All @@ -150,8 +151,11 @@ def per_token_cast_to_e5m6_kernel(
for j in T.serial(8):
in_local[j] = out_fragment[x, y * 8 + j]
float_to_e5m6(in_local, out_local)
for j in T.serial(3):
out[pid_token * block_m + x, pid_hidden * (block_k // 8 * 3) + y * 3 + j] = out_local[j]
m_idx = pid_token * block_m + x
k_idx = pid_hidden * block_k + y * 8
if m_idx < num_tokens and k_idx < hidden:
for j in T.serial(3):
out[m_idx, (k_idx // 8) * 3 + j] = out_local[j]

return per_token_cast_to_e5m6_kernel

Expand Down