Skip to content

[Tracing] Decouple vision tower from first layer #1710

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

[Tracing] Decouple vision tower from first layer #1710

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
4 changes: 1 addition & 3 deletions examples/quantization_w8a8_fp8/fp8_block_example.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -15,9 +15,7 @@
# In this case, we:
# * quantize the weights to fp8 with per channel via ptq
# * quantize the activations to fp8 with dynamic per token
recipe = QuantizationModifier(
targets="Linear", scheme="FP8_BLOCK", ignore=["lm_head"]
)
recipe = QuantizationModifier(targets="Linear", scheme="FP8_BLOCK", ignore=["lm_head"])

# Apply quantization.
oneshot(model=model, recipe=recipe)
Expand Down
8 changes: 4 additions & 4 deletions src/llmcompressor/pipelines/sequential/helpers.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -277,14 +277,14 @@ def topological_partition(graph: GraphModule, targets: Set[Module]) -> List[List
while len(queue) > 0:
node = queue.popleft()

# assign to partition
partitions[partition_index].append(node)

# guarantee targets are assigned to disjoint partitions
if node in target_nodes:
if node in target_nodes and len(partitions[partition_index]) > 0:
partition_index += 1
partitions.append([])

# assign to partition
partitions[partition_index].append(node)

# recurse on last indegree only in order to guarantee that
# the node is assigned to maximal partition
for user in node.users:
Expand Down
Loading