Multimodal RAG in Prompt Engineering / GenAI - Model Pipeline Trace

Model Pipeline - Multimodal RAG

Multimodal RAG combines text and images to answer questions by retrieving relevant information and generating answers using both types of data.

Data Flow - 5 Stages

1Input Data

1000 samples with text and images→Collect paired text and image data for questions and documents→1000 samples with text and image data

Question: 'What is shown in this picture?' + Image of a cat

↓

2Preprocessing

1000 samples with text and images→Clean text, resize images, and normalize both→1000 samples with cleaned text and processed images

Text: 'What is shown?' -> 'what is shown'; Image resized to 224x224 pixels

↓

3Feature Extraction

1000 samples with cleaned text and processed images→Convert text to embeddings and images to feature vectors→1000 samples with text embeddings (768 dims) and image embeddings (512 dims)

Text embedding vector: [0.12, -0.05, ...]; Image embedding vector: [0.34, 0.78, ...]

↓

4Retrieval

1000 samples with text and image embeddings→Retrieve top 5 relevant documents using combined embeddings→1000 samples with 5 retrieved documents each

Retrieved docs: ['Doc1 text', 'Doc2 text', ...]

↓

5Fusion and Generation

1000 samples with retrieved documents and embeddings→Fuse multimodal info and generate answer using a language model→1000 samples with generated text answers

Answer: 'The image shows a cat sitting on a sofa.'

Training Trace - Epoch by Epoch

Epoch 1: ************ (1.2)
Epoch 2: *********    (0.9)
Epoch 3: *******      (0.7)
Epoch 4: *****        (0.55)
Epoch 5: ****         (0.45)

Epoch	Loss ↓	Accuracy ↑	Observation
1	1.2	0.45	Model starts learning, loss high, accuracy low
2	0.9	0.60	Loss decreases, accuracy improves
3	0.7	0.72	Model learns better multimodal relations
4	0.55	0.80	Loss continues to drop, accuracy rises
5	0.45	0.85	Good convergence, model ready for predictions

Prediction Trace - 5 Layers

Layer 1: Input

Layer 2: Preprocessing

Layer 3: Feature Extraction

Layer 4: Retrieval

Layer 5: Fusion and Generation

Model Quiz - 3 Questions

What happens to the data shape after feature extraction?

A Data shape increases to include raw pixels

B Text and image converted to embeddings with fixed dimensions

C Text is removed and only images remain

D Data shape stays the same as input

Key Insight

Multimodal RAG effectively combines text and image data by converting both into embeddings, retrieving relevant documents, and generating accurate answers. The training shows steady improvement, highlighting the model's ability to learn from combined data types.

Practice

1. What is the main purpose of Multimodal RAG in AI systems?

easy

A. To generate images from text descriptions without retrieval

B. To translate languages using only text data

C. To combine text and images for better information retrieval and generation

D. To classify images into categories without text input

Solution

1. Step 1: Understand the components of Multimodal RAG

   Multimodal RAG uses both text and image data to improve retrieval and generation tasks.

2. Step 2: Identify the main goal

   The goal is to combine these data types to find and generate better answers than using text or images alone.

3. Final Answer:

   To combine text and images for better information retrieval and generation -> Option C

4. Quick Check:

   Multimodal RAG = combine text + images [OK]

Hint: Remember: Multimodal means multiple data types combined [OK]

Common Mistakes:

• Thinking it only works with text
• Confusing it with image-only models
• Assuming it only generates images

2. Which of the following is the correct component setup for a Multimodal RAG system?

easy

A. Single encoder for both text and images, no retriever

B. Separate encoders for text and images, plus a retriever and a generator

C. Only a text encoder and a generator, no image processing

D. Only an image encoder and a retriever, no text input

Solution

1. Step 1: Recall the architecture of Multimodal RAG

   It uses separate encoders for text and images to handle each data type properly.

2. Step 2: Understand the role of retriever and generator

   The retriever finds relevant data, and the generator creates the final output combining both modalities.

3. Final Answer:

   Separate encoders for text and images, plus a retriever and a generator -> Option B

4. Quick Check:

   Separate encoders + retriever + generator = B [OK]

Hint: Look for separate encoders and both retriever and generator [OK]

Common Mistakes:

• Assuming one encoder handles both text and images
• Ignoring the retriever component
• Thinking image processing is optional

3. Given the following pseudocode for a Multimodal RAG retrieval step, what will be the output type?

text_embedding = text_encoder(text_input)
image_embedding = image_encoder(image_input)
combined_embedding = concatenate(text_embedding, image_embedding)
retrieved_docs = retriever.retrieve(combined_embedding)
print(type(retrieved_docs))

medium

A. <class 'int'>

B. <class 'dict'>

C. <class 'str'>

D. <class 'list'>

Solution

1. Step 1: Understand the retriever output

   The retriever typically returns a list of documents or data items relevant to the query embedding.

2. Step 2: Identify the output type printed

   Since retrieved_docs holds multiple documents, its type is a list.

3. Final Answer:

   <class 'list'> -> Option D

4. Quick Check:

   Retriever output = list of documents [OK]

Hint: Retriever returns a list of relevant documents [OK]

Common Mistakes:

• Assuming output is a string or dictionary
• Confusing embedding types with retrieval output
• Expecting a single document instead of a list

4. You have this code snippet for a Multimodal RAG generator:

def generate_answer(text, image):
    text_emb = text_encoder(text)
    image_emb = image_encoder(image)
    combined = text_emb + image_emb
    docs = retriever.retrieve(combined)
    answer = generator.generate(docs)
    return answer

What is the main error in this code?

medium

A. Using '+' to combine embeddings instead of concatenation

B. Missing image encoder call

C. Retriever should not be called before generator

D. Generator cannot take documents as input

Solution

1. Step 1: Check how embeddings are combined

   Embeddings from different modalities should be concatenated, not added, to preserve information.

2. Step 2: Understand the impact of using '+' operator

   Adding embeddings sums values element-wise, which can lose modality-specific features.

3. Final Answer:

   Using '+' to combine embeddings instead of concatenation -> Option A

4. Quick Check:

   Combine embeddings = concatenate, not add [OK]

Hint: Use concatenate, not plus, to combine embeddings [OK]

Common Mistakes:

• Thinking '+' merges embeddings correctly
• Ignoring the need for separate encoders
• Assuming retriever or generator order is wrong

5. You want to improve a Multimodal RAG system that sometimes misses relevant images when answering questions. Which approach is best to fix this?

hard

A. Train the image encoder with more diverse image-text pairs to improve embedding quality

B. Remove the retriever and rely only on the generator

C. Use only text data and ignore images to simplify the model

D. Replace the text encoder with a simpler model to speed up processing

Solution

1. Step 1: Identify the cause of missing relevant images

   Low-quality image embeddings can cause the retriever to miss relevant images.

2. Step 2: Choose the best fix

   Training the image encoder with more diverse data improves embedding quality and retrieval accuracy.

3. Final Answer:

   Train the image encoder with more diverse image-text pairs to improve embedding quality -> Option A

4. Quick Check:

   Better image encoder training = better retrieval [OK]

Hint: Improve encoder training with diverse data for better retrieval [OK]

Common Mistakes:

• Removing retriever loses retrieval benefits
• Ignoring images reduces multimodal power
• Simplifying text encoder won't fix image retrieval