-
Notifications
You must be signed in to change notification settings - Fork 22
Add changes used in Devoxx Demo #54
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
Changes from 1 commit
File filter
Filter by extension
Conversations
Jump to
Diff view
Diff view
There are no files selected for viewing
| Original file line number | Diff line number | Diff line change |
|---|---|---|
| @@ -0,0 +1,56 @@ | ||
| #!/bin/bash | ||
|
|
||
| models=( | ||
| "../models/beehive-llama-3.2-1b-instruct-fp16.gguf" | ||
| "../models/Phi-3-mini-4k-instruct-fp16.gguf" | ||
| "../models/Qwen3-1.7B-f16.gguf" | ||
| "../models/Mistral-7B-Instruct-v0.3.fp16.gguf" | ||
| "../models/Qwen3-8B-f16.gguf" | ||
| ) | ||
|
|
||
| for model in "${models[@]}"; do | ||
| name=$(basename "$model" .gguf) | ||
| # file size (human readable, GB/MB) | ||
| if command -v numfmt &> /dev/null; then | ||
| size=$(stat -c%s "$model" 2>/dev/null | numfmt --to=iec --suffix=B) # Linux | ||
| [ -z "$size" ] && size=$(stat -f%z "$model" 2>/dev/null | numfmt --to=iec --suffix=B) # macOS | ||
| else | ||
| size=$(stat -c%s "$model" 2>/dev/null || stat -f%z "$model" 2>/dev/null) | ||
| size="${size} bytes" | ||
| fi | ||
|
|
||
| # colors | ||
| CYAN="\033[1;36m" | ||
| YELLOW="\033[1;33m" | ||
| RESET="\033[0m" | ||
|
|
||
| width=$(tput cols) # get terminal width | ||
| line=$(printf '━%.0s' $(seq 1 $width)) | ||
|
|
||
| echo -e "\n${CYAN}${line}${RESET}" | ||
| echo -e " 🚀 Running Model: ${YELLOW}$name${RESET} (size: ${YELLOW}$size${RESET}) 🚀" | ||
| # echo -e " 🚀 Running Model: ${YELLOW}$name${RESET} 🚀" | ||
| echo -e "${CYAN}${line}${RESET} \n" | ||
|
|
||
| cmd=( | ||
| java @argfile | ||
| -cp /home/devoxx2025-demo/java-ai-demos/GPULlama3.java/target/gpu-llama3-0.2.2.jar | ||
| org.beehive.gpullama3.LlamaApp | ||
| --model "$model" | ||
| --stream true | ||
| --echo false | ||
| -p "Who are you?" | ||
| --instruct | ||
| ) | ||
|
|
||
| # Pretty print the command (one-liner) | ||
| echo -e "java @argfile -cp /home/devoxx2025-demo/java-ai-demos/GPULlama3.java/target/gpu-llama3-0.2.2.jar org.beehive.gpullama3.LlamaApp --model \"$model\" --stream true --echo false -p \"Who are you?\" --instruct \n" | ||
|
|
||
| # Execute it | ||
| "${cmd[@]}" | ||
|
|
||
| #java @argfile -cp /home/devoxx2025-demo/java-ai-demos/GPULlama3.java/target/gpu-llama3-0.2.2.jar org.beehive.gpullama3.LlamaApp --model "$model" --stream true --echo false -p "Who are you?" --instruct | ||
|
|
||
| #./llama-tornado --gpu --opencl --model "$model" --prompt "Who are you?" | ||
| done | ||
|
|
| Original file line number | Diff line number | Diff line change |
|---|---|---|
| @@ -0,0 +1,184 @@ | ||
| -server | ||
| -XX:+UnlockExperimentalVMOptions | ||
| -XX:+EnableJVMCI | ||
| -Xms20g | ||
| -Xmx20g | ||
| --enable-preview | ||
|
|
||
| # -Dgraal.Dump=*:verbose -Dgraal.PrintGraph=Network -Dgraal.PrintBackendCFG=true | ||
|
|
||
| -Djava.library.path=/home/devoxx2025-demo/java-ai-demos/TornadoVM/bin/sdk/lib | ||
| -Djdk.module.showModuleResolution=false | ||
| --module-path .:/home/devoxx2025-demo/java-ai-demos/TornadoVM/bin/sdk/share/java/tornado | ||
| -Dtornado.load.api.implementation=uk.ac.manchester.tornado.runtime.tasks.TornadoTaskGraph | ||
| -Dtornado.load.runtime.implementation=uk.ac.manchester.tornado.runtime.TornadoCoreRuntime | ||
| -Dtornado.load.tornado.implementation=uk.ac.manchester.tornado.runtime.common.Tornado | ||
| -Dtornado.load.annotation.implementation=uk.ac.manchester.tornado.annotation.ASMClassVisitor | ||
| -Dtornado.load.annotation.parallel=uk.ac.manchester.tornado.api.annotations.Parallel | ||
| -Dtornado.tvm.maxbytecodesize=65536 | ||
| -Duse.tornadovm=true | ||
| -Dtornado.threadInfo=false | ||
| -Dtornado.debug=false | ||
| -Dtornado.fullDebug=false | ||
| -Dtornado.printKernel=false | ||
| -Dtornado.print.bytecodes=false | ||
| -Dtornado.device.memory=20GB | ||
| -Dtornado.profiler=false | ||
| -Dtornado.log.profiler=false | ||
| -Dtornado.profiler.dump.dir=/home/mikepapadim/repos/gpu-llama3.java/prof.json | ||
| -Dtornado.enable.fastMathOptimizations=true | ||
| -Dtornado.enable.mathOptimizations=false | ||
| -Dtornado.enable.nativeFunctions=true | ||
| "-Dtornado.loop.interchange=true -Dtornado.dump.bytecodes.dir=/home/devoxx2025-demo/java-ai-demos/GPULlama3.java" | ||
| -Dtornado.eventpool.maxwaitevents=32000 | ||
| "-Dtornado.opencl.compiler.flags=-cl-denorms-are-zero -cl-no-signed-zeros -cl-finite-math-only" | ||
| --upgrade-module-path /home/devoxx2025-demo/java-ai-demos/TornadoVM/bin/sdk/share/java/graalJars | ||
| --add-exports jdk.internal.vm.compiler/org.graalvm.compiler.nodes.cfg=tornado.runtime | ||
| --add-exports jdk.internal.vm.ci/jdk.vm.ci.common=jdk.internal.vm.compiler | ||
| --add-exports jdk.internal.vm.compiler/org.graalvm.compiler.hotspot.meta=tornado.runtime | ||
| --add-exports jdk.internal.vm.compiler/org.graalvm.compiler.core.common.util=tornado.runtime | ||
| --add-exports jdk.internal.vm.compiler/org.graalvm.compiler.lir=tornado.runtime,tornado.drivers.common | ||
| --add-exports jdk.internal.vm.ci/jdk.vm.ci.meta=tornado.runtime,tornado.annotation,tornado.drivers.common,jdk.internal.vm.compiler | ||
| --add-exports jdk.internal.vm.ci/jdk.vm.ci.code=tornado.runtime,tornado.drivers.common,jdk.internal.vm.compiler | ||
| --add-exports jdk.internal.vm.compiler/org.graalvm.compiler.graph=tornado.runtime | ||
| --add-exports jdk.internal.vm.compiler/org.graalvm.compiler.graph.spi=tornado.runtime | ||
| --add-exports jdk.internal.vm.compiler/org.graalvm.compiler.lir.gen=tornado.runtime | ||
| --add-exports jdk.internal.vm.compiler/org.graalvm.compiler.nodeinfo=tornado.runtime | ||
| --add-exports jdk.internal.vm.compiler/org.graalvm.compiler.nodes=tornado.runtime,tornado.drivers.common | ||
| --add-exports jdk.internal.vm.compiler/org.graalvm.compiler.nodes.calc=tornado.runtime | ||
| --add-exports jdk.internal.vm.compiler/org.graalvm.compiler.nodes.spi=tornado.runtime,tornado.drivers.common | ||
| --add-exports jdk.internal.vm.compiler/org.graalvm.compiler.api.runtime=tornado.runtime | ||
| --add-exports jdk.internal.vm.compiler/org.graalvm.compiler.code=tornado.runtime | ||
| --add-exports jdk.internal.vm.compiler/org.graalvm.compiler.core=tornado.runtime | ||
| --add-exports jdk.internal.vm.compiler/org.graalvm.compiler.core.common=tornado.runtime,tornado.drivers.common | ||
| --add-exports jdk.internal.vm.compiler/org.graalvm.compiler.core.target=tornado.runtime | ||
| --add-exports jdk.internal.vm.compiler/org.graalvm.compiler.debug=tornado.runtime,tornado.drivers.common | ||
| --add-exports jdk.internal.vm.compiler/org.graalvm.compiler.hotspot=tornado.runtime | ||
| --add-exports jdk.internal.vm.compiler/org.graalvm.compiler.java=tornado.runtime | ||
| --add-exports jdk.internal.vm.compiler/org.graalvm.compiler.lir.asm=tornado.runtime | ||
| --add-exports jdk.internal.vm.compiler/org.graalvm.compiler.lir.phases=tornado.runtime | ||
| --add-exports jdk.internal.vm.compiler/org.graalvm.compiler.nodes.graphbuilderconf=tornado.runtime | ||
| --add-exports jdk.internal.vm.compiler/org.graalvm.compiler.options=tornado.runtime | ||
| --add-exports jdk.internal.vm.compiler/org.graalvm.compiler.phases=tornado.runtime,tornado.drivers.common | ||
| --add-exports jdk.internal.vm.compiler/org.graalvm.compiler.phases.tiers=tornado.runtime | ||
| --add-exports jdk.internal.vm.compiler/org.graalvm.compiler.phases.util=tornado.runtime | ||
| --add-exports jdk.internal.vm.compiler/org.graalvm.compiler.printer=tornado.runtime | ||
| --add-exports jdk.internal.vm.compiler/org.graalvm.compiler.runtime=tornado.runtime | ||
| --add-exports jdk.internal.vm.ci/jdk.vm.ci.runtime=tornado.runtime | ||
| --add-exports jdk.internal.vm.compiler/org.graalvm.compiler.graph.iterators=tornado.runtime | ||
| --add-exports jdk.internal.vm.compiler/org.graalvm.compiler.nodes.java=tornado.runtime | ||
| --add-exports jdk.internal.vm.compiler/org.graalvm.compiler.bytecode=tornado.runtime | ||
| --add-exports jdk.internal.vm.compiler/org.graalvm.compiler.phases.common=tornado.runtime | ||
| --add-exports jdk.internal.vm.compiler/org.graalvm.compiler.core.common.spi=tornado.runtime,tornado.drivers.common | ||
| --add-exports jdk.internal.vm.compiler/org.graalvm.compiler.api.replacements=tornado.runtime | ||
| --add-exports jdk.internal.vm.compiler/org.graalvm.compiler.replacements=tornado.runtime | ||
| --add-exports jdk.internal.vm.compiler/org.graalvm.compiler.phases.common.inlining=tornado.runtime | ||
| --add-exports jdk.internal.vm.compiler/org.graalvm.compiler.core.phases=tornado.runtime | ||
| --add-exports jdk.internal.vm.compiler/org.graalvm.compiler.core.common.type=tornado.runtime | ||
| --add-exports jdk.internal.vm.compiler/org.graalvm.compiler.nodes.extended=tornado.runtime | ||
| --add-exports jdk.internal.vm.compiler/org.graalvm.compiler.nodes.loop=tornado.runtime | ||
| --add-exports jdk.internal.vm.compiler/org.graalvm.compiler.phases.common.inlining.info=tornado.runtime | ||
| --add-exports jdk.internal.vm.compiler/org.graalvm.compiler.phases.common.inlining.policy=tornado.runtime | ||
| --add-exports jdk.internal.vm.compiler/org.graalvm.compiler.phases.common.inlining.walker=tornado.runtime | ||
| --add-exports jdk.internal.vm.compiler/org.graalvm.compiler.loop.phases=tornado.runtime | ||
| --add-exports jdk.internal.vm.compiler/org.graalvm.compiler.nodes.debug=tornado.runtime | ||
| --add-exports jdk.internal.vm.compiler/org.graalvm.compiler.nodes.memory=tornado.runtime,tornado.drivers.common | ||
| --add-exports jdk.internal.vm.compiler/org.graalvm.compiler.nodes.util=tornado.runtime | ||
| --add-exports jdk.internal.vm.compiler/org.graalvm.compiler.nodes.virtual=tornado.runtime | ||
| --add-exports jdk.internal.vm.compiler/org.graalvm.compiler.lir.constopt=tornado.runtime | ||
| --add-opens jdk.internal.vm.ci/jdk.vm.ci.hotspot=tornado.runtime | ||
| --add-exports jdk.internal.vm.ci/jdk.vm.ci.hotspot=tornado.runtime | ||
| --add-exports jdk.internal.vm.compiler/org.graalvm.compiler.nodes.gc=tornado.runtime,tornado.drivers.common | ||
| --add-exports jdk.internal.vm.compiler/org.graalvm.compiler.nodes.memory.address=tornado.runtime,tornado.drivers.common | ||
| --add-exports jdk.internal.vm.compiler/org.graalvm.compiler.replacements.nodes=tornado.runtime | ||
| --add-exports jdk.internal.vm.compiler/org.graalvm.compiler.word=tornado.drivers.common | ||
| --add-exports jdk.internal.vm.compiler/org.graalvm.compiler.phases.util=tornado.drivers.common | ||
| --add-exports jdk.internal.vm.compiler/org.graalvm.compiler.lir.framemap=tornado.runtime | ||
| --add-exports jdk.internal.vm.compiler/org.graalvm.compiler.core.common.alloc=tornado.runtime | ||
| --add-exports jdk.internal.vm.compiler/org.graalvm.compiler.core.common.memory=tornado.drivers.common | ||
| --add-exports jdk.internal.vm.compiler/org.graalvm.compiler.graph=tornado.runtime,tornado.drivers.common | ||
| --add-exports jdk.internal.vm.compiler/org.graalvm.compiler.graph.iterators=tornado.drivers.common | ||
| --add-exports jdk.internal.vm.compiler/org.graalvm.compiler.nodes.java=tornado.drivers.common | ||
| --add-exports jdk.internal.vm.compiler/org.graalvm.compiler.nodes.extended=tornado.drivers.common | ||
| --add-exports jdk.internal.vm.compiler/org.graalvm.compiler.nodes.loop=tornado.drivers.common | ||
| --add-exports jdk.internal.vm.compiler/org.graalvm.compiler.nodes.calc=tornado.drivers.common | ||
| --add-exports jdk.internal.vm.compiler/org.graalvm.compiler.options=tornado.drivers.common | ||
| --add-exports jdk.internal.vm.compiler/org.graalvm.compiler.nodes.debug=tornado.drivers.common | ||
| --add-exports jdk.internal.vm.compiler/org.graalvm.compiler.nodes.util=tornado.drivers.common | ||
| --add-exports jdk.internal.vm.compiler/org.graalvm.compiler.nodes.virtual=tornado.drivers.common | ||
| --add-exports jdk.internal.vm.compiler/org.graalvm.compiler.loop.phases=tornado.drivers.common | ||
| --add-exports jdk.internal.vm.compiler/org.graalvm.compiler.core.common.util=tornado.drivers.common | ||
| --add-exports jdk.internal.vm.compiler/org.graalvm.compiler.phases.tiers=tornado.drivers.common | ||
| --add-exports jdk.internal.vm.compiler/org.graalvm.compiler.phases.common=tornado.drivers.common | ||
|
|
||
| --add-opens java.base/java.lang=tornado.drivers.opencl | ||
| --add-exports jdk.internal.vm.ci/jdk.vm.ci.common=tornado.drivers.opencl | ||
| --add-exports jdk.internal.vm.ci/jdk.vm.ci.amd64=tornado.drivers.opencl | ||
| --add-exports jdk.internal.vm.compiler/org.graalvm.compiler.hotspot.meta=tornado.drivers.opencl | ||
| --add-exports jdk.internal.vm.compiler/org.graalvm.compiler.replacements.classfile=tornado.drivers.opencl | ||
| --add-exports jdk.internal.vm.compiler/org.graalvm.compiler.core.common.alloc=tornado.drivers.opencl | ||
| --add-exports jdk.internal.vm.compiler/org.graalvm.compiler.core.common.util=tornado.drivers.opencl | ||
| --add-exports jdk.internal.vm.compiler/org.graalvm.compiler.core.common.cfg=tornado.drivers.opencl | ||
| --add-exports jdk.internal.vm.compiler/org.graalvm.compiler.lir=tornado.drivers.opencl | ||
| --add-exports jdk.internal.vm.compiler/org.graalvm.compiler.lir.framemap=tornado.drivers.opencl | ||
| --add-exports jdk.internal.vm.ci/jdk.vm.ci.meta=tornado.drivers.opencl | ||
| --add-exports jdk.internal.vm.ci/jdk.vm.ci.code=tornado.drivers.opencl | ||
| --add-exports jdk.internal.vm.compiler/org.graalvm.compiler.graph=tornado.drivers.opencl | ||
| --add-exports jdk.internal.vm.compiler/org.graalvm.compiler.graph.spi=tornado.drivers.opencl | ||
| --add-exports jdk.internal.vm.compiler/org.graalvm.compiler.lir.gen=tornado.drivers.opencl | ||
| --add-exports jdk.internal.vm.compiler/org.graalvm.compiler.nodeinfo=tornado.drivers.opencl | ||
| --add-exports jdk.internal.vm.compiler/org.graalvm.compiler.nodes=tornado.drivers.opencl | ||
| --add-exports jdk.internal.vm.compiler/org.graalvm.compiler.nodes.calc=tornado.drivers.opencl | ||
| --add-exports jdk.internal.vm.compiler/org.graalvm.compiler.nodes.spi=tornado.drivers.opencl | ||
| --add-exports jdk.internal.vm.compiler/org.graalvm.compiler.code=tornado.drivers.opencl | ||
| --add-exports jdk.internal.vm.compiler/org.graalvm.compiler.core=tornado.drivers.opencl | ||
| --add-exports jdk.internal.vm.compiler/org.graalvm.compiler.core.common=tornado.drivers.opencl | ||
| --add-exports jdk.internal.vm.compiler/org.graalvm.compiler.debug=tornado.drivers.opencl | ||
| --add-exports jdk.internal.vm.compiler/org.graalvm.compiler.hotspot=tornado.drivers.opencl | ||
| --add-exports jdk.internal.vm.compiler/org.graalvm.compiler.java=tornado.drivers.opencl | ||
| --add-exports jdk.internal.vm.compiler/org.graalvm.compiler.lir.asm=tornado.drivers.opencl | ||
| --add-exports jdk.internal.vm.compiler/org.graalvm.compiler.lir.phases=tornado.drivers.opencl | ||
| --add-exports jdk.internal.vm.compiler/org.graalvm.compiler.nodes.graphbuilderconf=tornado.drivers.opencl | ||
| --add-exports jdk.internal.vm.compiler/org.graalvm.compiler.options=tornado.drivers.opencl | ||
| --add-exports jdk.internal.vm.compiler/org.graalvm.compiler.phases=tornado.drivers.opencl | ||
| --add-exports jdk.internal.vm.compiler/org.graalvm.compiler.phases.tiers=tornado.drivers.opencl | ||
| --add-exports jdk.internal.vm.compiler/org.graalvm.compiler.phases.util=tornado.drivers.opencl | ||
| --add-exports jdk.internal.vm.compiler/org.graalvm.compiler.printer=tornado.drivers.opencl | ||
| --add-exports jdk.internal.vm.ci/jdk.vm.ci.runtime=tornado.drivers.opencl | ||
| --add-exports jdk.internal.vm.compiler/org.graalvm.compiler.graph.iterators=tornado.drivers.opencl | ||
| --add-exports jdk.internal.vm.compiler/org.graalvm.compiler.nodes.java=tornado.drivers.opencl | ||
| --add-exports jdk.internal.vm.compiler/org.graalvm.compiler.bytecode=tornado.drivers.opencl | ||
| --add-exports jdk.internal.vm.compiler/org.graalvm.compiler.phases.common=tornado.drivers.opencl | ||
| --add-exports jdk.internal.vm.compiler/org.graalvm.compiler.core.common.spi=tornado.drivers.opencl | ||
| --add-exports jdk.internal.vm.compiler/org.graalvm.compiler.api.replacements=tornado.drivers.opencl | ||
| --add-exports jdk.internal.vm.compiler/org.graalvm.compiler.replacements=tornado.drivers.opencl | ||
| --add-exports jdk.internal.vm.compiler/org.graalvm.compiler.phases.common.inlining=tornado.drivers.opencl | ||
| --add-exports jdk.internal.vm.compiler/org.graalvm.compiler.core.phases=tornado.drivers.opencl | ||
| --add-exports jdk.internal.vm.compiler/org.graalvm.compiler.core.common.type=tornado.drivers.opencl | ||
| --add-exports jdk.internal.vm.compiler/org.graalvm.compiler.nodes.extended=tornado.drivers.opencl | ||
| --add-exports jdk.internal.vm.compiler/org.graalvm.compiler.nodes.loop=tornado.drivers.opencl | ||
| --add-exports jdk.internal.vm.compiler/org.graalvm.compiler.loop.phases=tornado.drivers.opencl | ||
| --add-exports jdk.internal.vm.compiler/org.graalvm.compiler.nodes.debug=tornado.drivers.opencl | ||
| --add-exports jdk.internal.vm.compiler/org.graalvm.compiler.nodes.memory=tornado.drivers.opencl | ||
| --add-exports jdk.internal.vm.compiler/org.graalvm.compiler.nodes.util=tornado.drivers.opencl | ||
| --add-opens jdk.internal.vm.ci/jdk.vm.ci.hotspot=tornado.drivers.opencl | ||
| --add-exports jdk.internal.vm.ci/jdk.vm.ci.hotspot=tornado.drivers.opencl | ||
| --add-exports jdk.internal.vm.compiler/org.graalvm.compiler.asm=tornado.drivers.opencl | ||
| --add-exports jdk.internal.vm.compiler/org.graalvm.compiler.nodes.cfg=tornado.drivers.opencl | ||
| --add-exports jdk.internal.vm.compiler/org.graalvm.compiler.phases.schedule=tornado.drivers.opencl | ||
| --add-exports jdk.internal.vm.compiler/org.graalvm.compiler.virtual.phases.ea=tornado.drivers.opencl | ||
| --add-exports jdk.internal.vm.compiler/org.graalvm.compiler.lir.ssa=tornado.drivers.opencl | ||
| --add-exports jdk.internal.vm.compiler/org.graalvm.compiler.core.common.calc=tornado.drivers.opencl | ||
| --add-exports jdk.internal.vm.compiler/org.graalvm.compiler.core.gen=tornado.drivers.opencl | ||
| --add-exports jdk.internal.vm.compiler/org.graalvm.compiler.core.match=tornado.drivers.opencl | ||
| --add-exports jdk.internal.vm.compiler/org.graalvm.compiler.nodes.memory.address=tornado.drivers.opencl | ||
| --add-exports jdk.internal.vm.compiler/org.graalvm.compiler.nodes.type=tornado.drivers.opencl | ||
| --add-exports jdk.internal.vm.compiler/org.graalvm.compiler.phases.graph=tornado.drivers.opencl | ||
| --add-exports jdk.internal.vm.compiler/org.graalvm.compiler.phases.common.util=tornado.drivers.opencl | ||
| --add-exports jdk.internal.vm.compiler/org.graalvm.compiler.phases.common.util=tornado.drivers.opencl | ||
| --add-exports jdk.internal.vm.compiler/org.graalvm.compiler.phases.graph=tornado.drivers.opencl | ||
| --add-exports jdk.internal.vm.compiler/org.graalvm.compiler.word=tornado.drivers.opencl | ||
| --add-exports jdk.internal.vm.compiler/org.graalvm.compiler.core.common.memory=tornado.drivers.opencl | ||
|
|
||
| --add-modules | ||
| ALL-SYSTEM,jdk.incubator.vector,tornado.runtime,tornado.annotation,tornado.drivers.common,tornado.drivers.opencl |
| Original file line number | Diff line number | Diff line change |
|---|---|---|
|
|
@@ -24,7 +24,7 @@ public MistralModelLoader(FileChannel fileChannel, GGUF gguf, int contextLength, | |
| // @formatter:off | ||
| @Override | ||
| public Mistral loadModel() { | ||
| try (var ignored = Timer.log("Load Mistral model")) { | ||
| try { | ||
|
||
| Map<String, Object> metadata = gguf.getMetadata(); | ||
|
|
||
| Vocabulary vocabulary = Vocabulary.loadMistralVocabulary(metadata); | ||
|
|
||
| Original file line number | Diff line number | Diff line change |
|---|---|---|
|
|
@@ -34,14 +34,16 @@ public Phi3ModelLoader(FileChannel fileChannel, GGUF gguf, int contextLength, bo | |
| // @formatter:off | ||
| @Override | ||
| public Phi3 loadModel() { | ||
| try (var ignored = Timer.log("Load Phi3 model")) { | ||
| try { | ||
|
||
| Map<String, Object> metadata = gguf.getMetadata(); | ||
| final String modelPrefix = "phi3."; | ||
|
|
||
| Vocabulary vocabulary = Vocabulary.loadPhi3Vocabulary(metadata); | ||
| Tokenizer tokenizer = new Phi3Tokenizer(metadata, vocabulary); | ||
|
|
||
| System.out.println("Tokenizer: " + tokenizer.getClass().getSimpleName()); | ||
| if (TornadoVMMasterPlan.ENABLE_TORNADOVM_INIT_TIME) { | ||
| System.out.println("Tokenizer: " + tokenizer.getClass().getSimpleName()); | ||
| } | ||
|
|
||
| int modelContextLength = (int) metadata.get(modelPrefix + "context_length"); | ||
| if (contextLength < 0 || modelContextLength < contextLength) { | ||
|
|
||
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
This line contains two JVM arguments concatenated in a single string. They should be separated into individual array elements for proper argument parsing.