Skip to content

Improve timer loging for init #1

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

Merged
mikepapadim merged 3 commits into from
May 23, 2025
Merged

Improve timer loging for init #1

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
3 commits
Select commit Hold shift + click to select a range
a8c2c50
Enable TornadoVM timing and optimize task graph handling.
mikepapadim May 23, 2025
ce4d7bb
Refine logging messages for TornadoVM processes.
mikepapadim May 23, 2025
98dda0d
Disable TornadoVM init timing by default, add verbose flag
mikepapadim May 23, 2025
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
6 changes: 6 additions & 0 deletions llama-tornado
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -73,6 +73,9 @@ class LlamaRunner:
if args.use_gpu:
cmd.append("-Duse.tornadovm=true")

if args.verbose_init:
cmd.append("-Dllama.EnableTimingForTornadoVMInit=true")

# Debug options
debug_config = []

Expand Down Expand Up @@ -289,6 +292,9 @@ def create_parser() -> argparse.ArgumentParser:
help="Print kernel information (tornado.printKernel=true)")
verbose_group.add_argument("--full-dump", dest="full_dump", action="store_true",
help="Enable full debug dump (tornado.fullDebug=true)")
verbose_group.add_argument("--verbose-init", dest="verbose_init", action="store_true",
help="Enable timers for TornadoVM initialization (llama.EnableTimingForTornadoVMInit=true)")


# Command display options
command_group = parser.add_argument_group("Command Display Options")
Expand Down
2 changes: 1 addition & 1 deletion src/main/java/com/example/loader/weights/ModelLoader.java
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -89,7 +89,7 @@ public static Weights loadWeights(Map<String, GGMLTensorEntry> tensorEntries, Co
GGMLTensorEntry outputWeight = tensorEntries.getOrDefault("output.weight", tokenEmbeddings);

if (LlamaApp.USE_TORNADOVM) {
System.out.println("Loading weights in TornadoVM format");
System.out.println("Loading model weights in TornadoVM format (converting " + outputWeight.ggmlType() + " -> " + GGMLType.F16 + ")");
return createTornadoVMWeights(tensorEntries, config, ropeFreqs, tokenEmbeddings, outputWeight);
} else {
return createStandardWeights(tensorEntries, config, ropeFreqs, tokenEmbeddings, outputWeight);
Expand Down
Loading