Skip to content

Update int4pack related in torchchat gguf #1404

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.

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

Merged
merged 5 commits into from
Dec 17, 2024
Merged
Changes from all commits
Commits
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
66 changes: 41 additions & 25 deletions torchchat/utils/gguf_loader.py
Original file line number Diff line number Diff line change
Expand Up @@ -24,6 +24,8 @@
pack_scales_and_zeros,
)

from torchao.dtypes.utils import is_device


logger: logging.Logger = logging.getLogger(__name__)

Expand Down Expand Up @@ -128,6 +130,7 @@ def linear_int4(input, weight_int4pack, scales_and_zeros, out_features, groupsiz
groupsize,
scales_and_zeros,
)

new_shape = origin_input_size[:-1] + (out_features,)
c = c.reshape(new_shape)
return c
Expand Down Expand Up @@ -178,16 +181,27 @@ def __init__(
), "must specify both weights and scales_and_zeros, or neither"

if weight is None:
weight = torch.empty(
(
out_features // 8,
in_features // (inner_k_tiles * 16),
32,
inner_k_tiles // 2,
),
dtype=torch.int32,
device=device,
)
if is_device(device, "cpu"):
weight = torch.empty(
(
out_features,
in_features // 2,
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

nice

),
dtype=torch.uint8,
device=device,
)
else:
weight = torch.empty(
(
out_features // 8,
in_features // (inner_k_tiles * 16),
32,
inner_k_tiles // 2,
),
dtype=torch.int32,
device=device,
)

scales_and_zeros = torch.empty(
(in_features // groupsize, out_features, 2),
dtype=get_precision(),
Expand Down Expand Up @@ -223,12 +237,17 @@ def _prepare_weight_and_scales_and_zeros(
weight_int32, scales_and_zeros = group_quantize_tensor(
weight_bf16, n_bit=4, groupsize=groupsize
)
weight_uint8 = (weight_int32[::, ::2] << 4 | weight_int32[::, 1::2]).to(
torch.uint8
)
weight_int4pack = torch.ops.aten._convert_weight_to_int4pack(
weight_uint8, inner_k_tiles
)
if is_device(weight_int32.device.type, "cpu"):
weight_int4pack = torch.ops.aten._convert_weight_to_int4pack_for_cpu(
weight_int32, inner_k_tiles
)
else:
weight_uint8 = (weight_int32[::, ::2] << 4 | weight_int32[::, 1::2]).to(
torch.uint8
)
weight_int4pack = torch.ops.aten._convert_weight_to_int4pack(
weight_uint8, inner_k_tiles
)
return weight_int4pack, scales_and_zeros

@classmethod
Expand Down Expand Up @@ -609,17 +628,14 @@ def load_model_and_state_dict(
if load_state_dict:
q, s, z = Q4_0.unpack(t)
scales_and_zeros = pack_scales_and_zeros(s, z)
q_uint8 = (q[::, ::2] << 4 | q[::, 1::2]).to(torch.uint8)

if torch.device(device).type == "cpu":
weight_int4pack = (
torch.ops.aten._convert_weight_to_int4pack_for_cpu(
q, inner_k_tiles
)
if is_device(q.device.type, "cpu"):
weight_int4pack = torch.ops.aten._convert_weight_to_int4pack_for_cpu(
q, inner_k_tiles
)
else:
q_tmp = (q[::, ::2] << 4 | q[::, 1::2]).to(torch.uint8)
weight_int4pack = torch.ops.aten._convert_weight_to_int4pack(
q_uint8, inner_k_tiles
q_tmp, inner_k_tiles
)
state_dict[f"{fqn}.weight"] = weight_int4pack
state_dict[f"{fqn}.scales_and_zeros"] = scales_and_zeros
Expand All @@ -632,7 +648,7 @@ def load_model_and_state_dict(
in_features=in_features,
out_features=out_features,
bias=False,
device="meta",
device="cpu",
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Let's keep this as a meta device as long as we can

Copy link
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

groupsize=Q4_0.groupsize,
inner_k_tiles=inner_k_tiles,
),
Expand Down
Loading