Visual-language assistant with Phi3-Vision and OpenVINO#

This Jupyter notebook can be launched after a local installation only.

The Phi-3-Vision is a lightweight, state-of-the-art open multimodal model built upon datasets which include - synthetic data and filtered publicly available websites - with a focus on very high-quality, reasoning dense data both on text and vision. The model belongs to the Phi-3 model family, and the multimodal version comes with 128K context length (in tokens) it can support. The model underwent a rigorous enhancement process, incorporating both supervised fine-tuning and direct preference optimization to ensure precise instruction adherence and robust safety measures. More details about model can be found in model blog post, technical report, Phi-3-cookbook

In this tutorial we consider how to launch Phi-3-vision using OpenVINO for creation multimodal chatbot. Additionally, we optimize model to low precision using NNCF

Table of contents:

Prerequisites
Select Model
Convert and Optimize model
- Compress model weights to 4-bit
Select inference device
Run OpenVINO model
Interactive demo

Installation Instructions#

This is a self-contained example that relies solely on its own code.

We recommend running the notebook in a virtual environment. You only need a Jupyter server to start. For details, please refer to Installation Guide.

Prerequisites#

install required packages and setup helper functions.

%pip install -q "torch>=2.1" "torchvision" "transformers>=4.40" "protobuf>=3.20" "gradio>=4.26" "Pillow" "accelerate" "tqdm"  --extra-index-url https://download.pytorch.org/whl/cpu
%pip install  -q "openvino>=2024.2.0" "nncf>=2.11.0"

Note: you may need to restart the kernel to use updated packages.
Note: you may need to restart the kernel to use updated packages.

import requests
from pathlib import Path

if not Path("ov_phi3_vision_helper.py").exists():
    r = requests.get(url="https://raw.githubusercontent.com/openvinotoolkit/openvino_notebooks/latest/notebooks/phi-3-vision/ov_phi3_vision_helper.py")
    open("ov_phi3_vision_helper.py", "w").write(r.text)


if not Path("gradio_helper.py").exists():
    r = requests.get(url="https://raw.githubusercontent.com/openvinotoolkit/openvino_notebooks/latest/notebooks/phi-3-vision/gradio_helper.py")
    open("gradio_helper.py", "w").write(r.text)

if not Path("notebook_utils.py").exists():
    r = requests.get(url="https://raw.githubusercontent.com/openvinotoolkit/openvino_notebooks/latest/utils/notebook_utils.py")
    open("notebook_utils.py", "w").write(r.text)

Select Model#

The tutorial supports the following models from Phi-3 model family:

You can select one from the provided options below.

import ipywidgets as widgets

# Select model
model_ids = [
    "microsoft/Phi-3.5-vision-instruct",
    "microsoft/Phi-3-vision-128k-instruct",
]

model_dropdown = widgets.Dropdown(
    options=model_ids,
    value=model_ids[0],
    description="Model:",
    disabled=False,
)

model_dropdown

Dropdown(description='Model:', options=('microsoft/Phi-3.5-vision-instruct', 'microsoft/Phi-3-vision-128k-inst…

Convert and Optimize model#

Phi-3-vision is PyTorch model. OpenVINO supports PyTorch models via conversion to OpenVINO Intermediate Representation (IR). OpenVINO model conversion API should be used for these purposes. ov.convert_model function accepts original PyTorch model instance and example input for tracing and returns ov.Model representing this model in OpenVINO framework. Converted model can be used for saving on disk using ov.save_model function or directly loading on device using core.complie_model.

The script ov_phi3_vision_helper.py contains helper function for model conversion, please check its content if you interested in conversion details.

Click here for more detailed explanation of conversion steps Phi-3-vision is autoregressive transformer generative model, it means that each next model step depends from model output from previous step. The generation approach is based on the assumption that the probability distribution of a word sequence can be decomposed into the product of conditional next word distributions. In other words, model predicts the next token in the loop guided by previously generated tokens until the stop-condition will be not reached (generated sequence of maximum length or end of string token obtained). The way the next token will be selected over predicted probabilities is driven by the selected decoding methodology. You can find more information about the most popular decoding methods in this blog. The entry point for the generation process for models from the Hugging Face Transformers library is the generate method. You can find more information about its parameters and configuration in the documentation. To preserve flexibility in the selection decoding methodology, we will convert only model inference for one step.

The inference flow has difference on first step and for the next. On the first step, model accept preprocessed input instruction and image, that transformed to the unified embedding space using input_embedding and image_encoder models, after that language model, LLM-based part of model, runs on input embeddings to predict probability of next generated tokens. On the next step, language_model accepts only next token id selected based on sampling strategy and processed by input_embedding model and cached attention key and values. Since the output side is auto-regressive, an output token hidden state remains the same once computed for every further generation step. Therefore, recomputing it every time you want to generate a new token seems wasteful. With the cache, the model saves the hidden state once it has been computed. The model only computes the one for the most recently generated output token at each time step, re-using the saved ones for hidden tokens. This reduces the generation complexity from \(O(n^3)\) to \(O(n^2)\) for a transformer model. More details about how it works can be found in this article. For improving support images of various resolution, input image separated on patches and processed by image feature extractor and image projector that are part of image encoder.

To sum up above, model consists of 4 parts:

Image feature extractor and Image projector for encoding input images into embedding space.
Input Embedding for conversion input text tokens into embedding space
Language Model for generation answer based on input embeddings provided by Image Encoder and Input Embedding models.

Compress model weights to 4-bit#

For reducing memory consumption, weights compression optimization can be applied using NNCF.

Click here for more details about weight compression Weight compression aims to reduce the memory footprint of a model. It can also lead to significant performance improvement for large memory-bound models, such as Large Language Models (LLMs). LLMs and other models, which require extensive memory to store the weights during inference, can benefit from weight compression in the following ways:

enabling the inference of exceptionally large models that cannot be accommodated in the memory of the device;
improving the inference performance of the models by reducing the latency of the memory access when computing the operations with weights, for example, Linear layers.

Neural Network Compression Framework (NNCF) provides 4-bit / 8-bit mixed weight quantization as a compression method primarily designed to optimize LLMs. The main difference between weights compression and full model quantization (post-training quantization) is that activations remain floating-point in the case of weights compression which leads to a better accuracy. Weight compression for LLMs provides a solid inference performance improvement which is on par with the performance of the full model quantization. In addition, weight compression is data-free and does not require a calibration dataset, making it easy to use.

nncf.compress_weights function can be used for performing weights compression. The function accepts an OpenVINO model and other compression parameters. Compared to INT8 compression, INT4 compression improves performance even more, but introduces a minor drop in prediction quality.

More details about weights compression, can be found in OpenVINO documentation.

from ov_phi3_vision_helper import convert_phi3_model

# uncomment these lines to see model conversion code
# convert_phi3_model??

INFO:nncf:NNCF initialized successfully. Supported frameworks detected: torch, tensorflow, onnx, openvino

2024-11-05 02:18:20.797755: I tensorflow/core/util/port.cc:110] oneDNN custom operations are on. You may see slightly different numerical results due to floating-point round-off errors from different computation orders. To turn them off, set the environment variable TF_ENABLE_ONEDNN_OPTS=0.
2024-11-05 02:18:20.831817: I tensorflow/core/platform/cpu_feature_guard.cc:182] This TensorFlow binary is optimized to use available CPU instructions in performance-critical operations.
To enable the following instructions: AVX2 AVX512F AVX512_VNNI FMA, in other operations, rebuild TensorFlow with the appropriate compiler flags.
2024-11-05 02:18:21.377210: W tensorflow/compiler/tf2tensorrt/utils/py_utils.cc:38] TF-TRT Warning: Could not find TensorRT

from pathlib import Path
import nncf


model_id = model_dropdown.value
out_dir = Path("model") / Path(model_id).name / "INT4"
compression_configuration = {
    "mode": nncf.CompressWeightsMode.INT4_SYM,
    "group_size": 64,
    "ratio": 0.6,
}
convert_phi3_model(model_id, out_dir, compression_configuration)

⌛ Phi-3.5-vision-instruct conversion started. Be patient, it may takes some time.
⌛ Load Original model

Loading checkpoint shards:   0%|          | 0/2 [00:00<?, ?it/s]

/opt/home/k8sworker/ci-ai/cibuilds/jobs/ov-notebook/jobs/OVNotebookOps/builds/810/archive/.workspace/scm/ov-notebook/.venv/lib/python3.8/site-packages/transformers/models/auto/image_processing_auto.py:520: FutureWarning: The image_processor_class argument is deprecated and will be removed in v4.42. Please use slow_image_processor_class, or fast_image_processor_class instead
  warnings.warn(

✅ Original model successfully loaded
⌛ Convert Input embedding model
WARNING:tensorflow:Please fix your imports. Module tensorflow.python.training.tracking.base has been moved to tensorflow.python.trackable.base. The old module will be deleted in version 2.11.

[ WARNING ]  Please fix your imports. Module %s has been moved to %s. The old module will be deleted in version %s.

WARNING:nncf:NNCF provides best results with torch==2.4.*, while current torch version is 2.2.2+cpu. If you encounter issues, consider switching to torch==2.4.*

/opt/home/k8sworker/ci-ai/cibuilds/jobs/ov-notebook/jobs/OVNotebookOps/builds/810/archive/.workspace/scm/ov-notebook/.venv/lib/python3.8/site-packages/transformers/modeling_utils.py:5006: FutureWarning: _is_quantized_training_enabled is going to be deprecated in transformers 4.39.0. Please use model.hf_quantizer.is_trainable instead
  warnings.warn(
loss_type=None was set in the config but it is unrecognised.Using the default loss: ForCausalLMLoss.

✅ Input embedding model successfully converted
⌛ Convert Image embedding model

/opt/home/k8sworker/ci-ai/cibuilds/jobs/ov-notebook/jobs/OVNotebookOps/builds/810/archive/.workspace/scm/ov-notebook/.venv/lib/python3.8/site-packages/transformers/models/clip/modeling_clip.py:243: TracerWarning: Converting a tensor to a Python boolean might cause the trace to be incorrect. We can't record the data flow of Python values, so this value will be treated as a constant in the future. This means that the trace might not generalize to other inputs!
  if not interpolate_pos_encoding and (height != self.image_size or width != self.image_size):

✅ Image embedding model successfully converted
⌛ Convert Image projection model

You are not running the flash-attention implementation, expect numerical differences.

✅ Image projection model successfully converted
⌛ Convert Language model

/opt/home/k8sworker/ci-ai/cibuilds/jobs/ov-notebook/jobs/OVNotebookOps/builds/810/archive/.workspace/scm/ov-notebook/.venv/lib/python3.8/site-packages/transformers/cache_utils.py:458: TracerWarning: Using len to get tensor shape might cause the trace to be incorrect. Recommended usage would be tensor.shape[0]. Passing a tensor of different shape might lead to errors or silently give incorrect results.
  or len(self.key_cache[layer_idx]) == 0  # the layer has no cache
/opt/home/k8sworker/ci-ai/cibuilds/jobs/ov-notebook/jobs/OVNotebookOps/builds/810/archive/.workspace/scm/ov-notebook/.venv/lib/python3.8/site-packages/transformers/modeling_attn_mask_utils.py:116: TracerWarning: Converting a tensor to a Python boolean might cause the trace to be incorrect. We can't record the data flow of Python values, so this value will be treated as a constant in the future. This means that the trace might not generalize to other inputs!
  if (input_shape[-1] > 1 or self.sliding_window is not None) and self.is_causal:
/opt/home/k8sworker/ci-ai/cibuilds/jobs/ov-notebook/jobs/OVNotebookOps/builds/810/archive/.workspace/scm/ov-notebook/.venv/lib/python3.8/site-packages/transformers/modeling_attn_mask_utils.py:164: TracerWarning: Converting a tensor to a Python boolean might cause the trace to be incorrect. We can't record the data flow of Python values, so this value will be treated as a constant in the future. This means that the trace might not generalize to other inputs!
  if past_key_values_length > 0:
/opt/home/k8sworker/.cache/huggingface/modules/transformers_modules/microsoft/Phi-3.5-vision-instruct/4a0d683eba9f1d0cbfb6151705d1ee73c25a80ca/modeling_phi3_v.py:444: TracerWarning: Converting a tensor to a Python boolean might cause the trace to be incorrect. We can't record the data flow of Python values, so this value will be treated as a constant in the future. This means that the trace might not generalize to other inputs!
  seq_len = seq_len or torch.max(position_ids) + 1
/opt/home/k8sworker/.cache/huggingface/modules/transformers_modules/microsoft/Phi-3.5-vision-instruct/4a0d683eba9f1d0cbfb6151705d1ee73c25a80ca/modeling_phi3_v.py:445: TracerWarning: Converting a tensor to a Python boolean might cause the trace to be incorrect. We can't record the data flow of Python values, so this value will be treated as a constant in the future. This means that the trace might not generalize to other inputs!
  if seq_len > self.original_max_position_embeddings:
/opt/home/k8sworker/ci-ai/cibuilds/jobs/ov-notebook/jobs/OVNotebookOps/builds/810/archive/.workspace/scm/ov-notebook/.venv/lib/python3.8/site-packages/nncf/torch/dynamic_graph/wrappers.py:86: TracerWarning: torch.tensor results are registered as constants in the trace. You can safely ignore this warning if you use this function to create tensors out of constant variables that would be the same every time you call this function. In any other case, this might cause the trace to be incorrect.
  op1 = operator(*args, **kwargs)
/opt/home/k8sworker/ci-ai/cibuilds/jobs/ov-notebook/jobs/OVNotebookOps/builds/810/archive/.workspace/scm/ov-notebook/.venv/lib/python3.8/site-packages/transformers/cache_utils.py:443: TracerWarning: Using len to get tensor shape might cause the trace to be incorrect. Recommended usage would be tensor.shape[0]. Passing a tensor of different shape might lead to errors or silently give incorrect results.
  elif len(self.key_cache[layer_idx]) == 0:  # fills previously skipped layers; checking for tensor causes errors
/opt/home/k8sworker/.cache/huggingface/modules/transformers_modules/microsoft/Phi-3.5-vision-instruct/4a0d683eba9f1d0cbfb6151705d1ee73c25a80ca/modeling_phi3_v.py:683: TracerWarning: Converting a tensor to a Python boolean might cause the trace to be incorrect. We can't record the data flow of Python values, so this value will be treated as a constant in the future. This means that the trace might not generalize to other inputs!
  if attn_weights.size() != (bsz, self.num_heads, q_len, kv_seq_len):
/opt/home/k8sworker/.cache/huggingface/modules/transformers_modules/microsoft/Phi-3.5-vision-instruct/4a0d683eba9f1d0cbfb6151705d1ee73c25a80ca/modeling_phi3_v.py:690: TracerWarning: Converting a tensor to a Python boolean might cause the trace to be incorrect. We can't record the data flow of Python values, so this value will be treated as a constant in the future. This means that the trace might not generalize to other inputs!
  if attention_mask.size() != (bsz, 1, q_len, kv_seq_len):
/opt/home/k8sworker/.cache/huggingface/modules/transformers_modules/microsoft/Phi-3.5-vision-instruct/4a0d683eba9f1d0cbfb6151705d1ee73c25a80ca/modeling_phi3_v.py:702: TracerWarning: Converting a tensor to a Python boolean might cause the trace to be incorrect. We can't record the data flow of Python values, so this value will be treated as a constant in the future. This means that the trace might not generalize to other inputs!
  if attn_output.size() != (bsz, self.num_heads, q_len, self.head_dim):
/opt/home/k8sworker/ci-ai/cibuilds/jobs/ov-notebook/jobs/OVNotebookOps/builds/810/archive/.workspace/scm/ov-notebook/.venv/lib/python3.8/site-packages/torch/jit/_trace.py:165: UserWarning: The .grad attribute of a Tensor that is not a leaf Tensor is being accessed. Its .grad attribute won't be populated during autograd.backward(). If you indeed want the .grad field to be populated for a non-leaf Tensor, use .retain_grad() on the non-leaf Tensor. If you access the non-leaf Tensor by mistake, make sure you access the leaf Tensor instead. See github.com/pytorch/pytorch/pull/30531 for more informations. (Triggered internally at aten/src/ATen/core/TensorBody.h:489.)
  if a.grad is not None:

✅ Language model successfully converted
⌛ Weights compression with int4_sym mode started

Output()

INFO:nncf:Statistics of the bitwidth distribution:
┍━━━━━━━━━━━━━━━━┯━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┯━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┑
│   Num bits (N) │ % all parameters (layers)   │ % ratio-defining parameters (layers)   │
┝━━━━━━━━━━━━━━━━┿━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┿━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┥
│              8 │ 42% (54 / 129)              │ 40% (53 / 128)                         │
├────────────────┼─────────────────────────────┼────────────────────────────────────────┤
│              4 │ 58% (75 / 129)              │ 60% (75 / 128)                         │
┕━━━━━━━━━━━━━━━━┷━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┷━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┙

Output()

✅ Weights compression finished
✅ Phi-3.5-vision-instruct model conversion finished. You can find results in model/Phi-3.5-vision-instruct/INT4

Select inference device#

from notebook_utils import device_widget

device = device_widget(default="AUTO", exclude=["NPU"])

device

Dropdown(description='Device:', index=1, options=('CPU', 'AUTO'), value='AUTO')

Run OpenVINO model#

OvPhi3vison class provides convenient way for running model. It accepts directory with converted model and inference device as arguments. For running model we will use generate method.

from ov_phi3_vision_helper import OvPhi3Vision

# Uncomment below lines to see the model inference class code

# OvPhi3Vision??

model = OvPhi3Vision(out_dir, device.value)

import requests
from PIL import Image

url = "https://github.com/openvinotoolkit/openvino_notebooks/assets/29454499/d5fbbd1a-d484-415c-88cb-9986625b7b11"
image = Image.open(requests.get(url, stream=True).raw)

print("Question:\n What is unusual on this picture?")
image

Question:
 What is unusual on this picture?

../_images/phi-3-vision-with-output_14_1.png

from transformers import AutoProcessor, TextStreamer

messages = [
    {"role": "user", "content": "<|image_1|>\nWhat is unusual on this picture?"},
]

processor = AutoProcessor.from_pretrained(out_dir, trust_remote_code=True)

prompt = processor.tokenizer.apply_chat_template(messages, tokenize=False, add_generation_prompt=True)

inputs = processor(prompt, [image], return_tensors="pt")

generation_args = {"max_new_tokens": 50, "do_sample": False, "streamer": TextStreamer(processor.tokenizer, skip_prompt=True, skip_special_tokens=True)}

print("Answer:")
generate_ids = model.generate(**inputs, eos_token_id=processor.tokenizer.eos_token_id, **generation_args)

Answer:
Nothing unusual, it's a cat lying in a box.

Interactive demo#

from gradio_helper import make_demo

demo = make_demo(model, processor)

try:
    demo.launch(debug=False, height=600)
except Exception:
    demo.launch(debug=False, share=True, height=600)
# if you are launching remotely, specify server_name and server_port
# demo.launch(server_name='your server name', server_port='server port in int')
# Read more in the docs: https://gradio.app/docs/

Running on local URL:  http://127.0.0.1:7860

To create a public link, set share=True in launch().