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.

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

example: add example of Phi 4 using torch.compile backend #2906

Open
wants to merge 2 commits into
base: main
Choose a base branch
from
Open
Show file tree
Hide file tree
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
1 change: 1 addition & 0 deletions examples/dynamo/README.rst
Original file line number Diff line number Diff line change
Expand Up @@ -18,6 +18,7 @@ Model Zoo
* :ref:`torch_compile_transformer`: Compiling a Transformer model using ``torch.compile``
* :ref:`torch_compile_stable_diffusion`: Compiling a Stable Diffusion model using ``torch.compile``
* :ref:`_torch_compile_gpt2`: Compiling a GPT2 model using ``torch.compile``
* :ref:`_torch_compile_phi4`: Compiling a Phi4 model from Hugging Face using ``torch.compile``
* :ref:`_torch_export_gpt2`: Compiling a GPT2 model using AOT workflow (`ir=dynamo`)
* :ref:`_torch_export_llama2`: Compiling a Llama2 model using AOT workflow (`ir=dynamo`)
* :ref:`_torch_export_sam2`: Compiling SAM2 model using AOT workflow (`ir=dynamo`)
Expand Down
65 changes: 65 additions & 0 deletions examples/dynamo/torch_compile_phi4.py
Original file line number Diff line number Diff line change
@@ -0,0 +1,65 @@
"""
.. _torch_compile_phi4:

Compiling Phi 4 model from Hugging Face using the Torch-TensorRT `torch.compile` Backend
======================================================

This script is intended as a sample of the Torch-TensorRT workflow with `torch.compile` on a Phi 4 model from Hugging Face.
"""

# %%
# Imports and Model Definition
# ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^

import requests
import torch
import torch_tensorrt
from PIL import Image
from transformers import AutoModelForCausalLM, AutoProcessor

# %%
# Load the pre-trained model weights from Hugging Face
# ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^

model_id = "microsoft/Phi-4-multimodal-instruct"
processor = AutoProcessor.from_pretrained(model_id, trust_remote_code=True)
model = AutoModelForCausalLM.from_pretrained(
model_id, trust_remote_code=True, torch_dtype="auto"
).cuda()

# %%
# Compile the model with torch.compile, using Torch-TensorRT backend
# ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^

model = torch.compile(model, backend="tensorrt")

# %%
# Write prompt and load image
# ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^

user_prompt = "<|user|>\n"
assistant_prompt = "<|assistant|>\n"
prompt_suffix = "<|end|>\n"

# single-image prompt
prompt = f"{user_prompt}<|image_1|>\nWhat is shown in this image?{prompt_suffix}{assistant_prompt}"
url = "https://www.ilankelman.org/stopsigns/australia.jpg"
print(f">>> Prompt\n{prompt}")

image = Image.open(requests.get(url, stream=True).raw)
inputs = processor(prompt, image, return_tensors="pt").to("cuda:0")

# %%
# Inference
# ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^

generate_ids = model.generate(
**inputs,
max_new_tokens=1000,
eos_token_id=processor.tokenizer.eos_token_id,
)
generate_ids = generate_ids[:, inputs["input_ids"].shape[1] :]
response = processor.batch_decode(
generate_ids, skip_special_tokens=True, clean_up_tokenization_spaces=False
)[0]
print(f">>> Response\n{response}")
Loading