Skip to content
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

Update llama-3.2-webgpu demo (GQA support) #11

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

Update llama-3.2-webgpu demo (GQA support) #11

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
16 changes: 8 additions & 8 deletions llama-3.2-webgpu/package-lock.json
View file
Open in desktop

Some generated files are not rendered by default. Learn more about how customized files appear on GitHub.

2 changes: 1 addition & 1 deletion llama-3.2-webgpu/package.json
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -10,7 +10,7 @@
"preview": "vite preview"
},
"dependencies": {
"@huggingface/transformers": "3.0.0",
"@huggingface/transformers": "3.0.1",
"dompurify": "^3.1.2",
"marked": "^12.0.2",
"react": "^18.3.1",
Expand Down
4 changes: 2 additions & 2 deletions llama-3.2-webgpu/src/App.jsx
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -10,7 +10,7 @@ const STICKY_SCROLL_THRESHOLD = 120;
const EXAMPLES = [
"Give me some tips to improve my time management skills.",
"What is the difference between AI and ML?",
"Write python code to compute the nth fibonacci number.",
"Write Python code to perform merge sort.",
];

function App() {
Expand Down Expand Up @@ -209,7 +209,7 @@ function App() {
<br />
You are about to load{" "}
<a
href="https://huggingface.co/onnx-community/Llama-3.2-1B-Instruct-q4f16"
href="https://huggingface.co/onnx-community/Llama-3.2-1B-Instruct-onnx-web-gqa"
target="_blank"
rel="noreferrer"
className="font-medium underline"
Expand Down
12 changes: 5 additions & 7 deletions llama-3.2-webgpu/src/worker.js
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -9,15 +9,14 @@ import {
* This class uses the Singleton pattern to enable lazy-loading of the pipeline
*/
class TextGenerationPipeline {
static model_id = "onnx-community/Llama-3.2-1B-Instruct-q4f16";
static model_id = "onnx-community/Llama-3.2-1B-Instruct-onnx-web-gqa";

static async getInstance(progress_callback = null) {
this.tokenizer ??= AutoTokenizer.from_pretrained(this.model_id, {
progress_callback,
});

this.model ??= AutoModelForCausalLM.from_pretrained(this.model_id, {
dtype: "q4f16",
device: "webgpu",
progress_callback,
});
Expand Down Expand Up @@ -69,18 +68,17 @@ async function generate(messages) {

const { past_key_values, sequences } = await model.generate({
...inputs,
// TODO: Add when model is fixed
// past_key_values: past_key_values_cache,
past_key_values: past_key_values_cache,

// Sampling
do_sample: false,
// do_sample: true,

max_new_tokens: 1024,
streamer,
stopping_criteria,
return_dict_in_generate: true,
});
// past_key_values_cache = past_key_values;
past_key_values_cache = past_key_values;

const decoded = tokenizer.batch_decode(sequences, {
skip_special_tokens: true,
Expand Down Expand Up @@ -153,7 +151,7 @@ self.addEventListener("message", async (e) => {
break;

case "reset":
// past_key_values_cache = null;
past_key_values_cache = null;
stopping_criteria.reset();
break;
}
Expand Down