NumPy

Visualize any NumPy array with instant dimension awareness. Support for high-dimensional, large arrays.

import svetoviz_webgpu as sv

data = np.random.rand(4, 4, 32, 32).astype(np.float32)

sv.array(data, gap_size=10)

Learn More

torch.Tensor

Load any torch.Tensor

import svetoviz_webgpu as sv

embedding_tensor = model.transformer.wte.weight

sv.tensor(embedding_tensor)

Learn More

PyTorch Debugger

Directly load your PyTorch module, interact, and debug activations in real-time.

import svetoviz_webgpu as sv

img = Image.open("sample_image.jpg").convert("RGB")
img_np = np.array(img).astype(np.float32) / 255.0

# Rearrange from (H, W, C) to (C, H, W) and add batch dim
input_tensor = torch.from_numpy(img_np).permute(2, 0, 1).unsqueeze(0)

# 2. Define the 2D Convolutional module
conv2d = nn.Conv2d(in_channels=3, out_channels=16, kernel_size=3, stride=1, padding=1)

def terminal_callback(buffer, message, images, files):
    # 3. Process the image through the module
    output = conv2d(input_tensor)

    # Log the feature map shape to the Svetoviz terminal
    buffer.send_system_message(f"Processed image. Shape: {list(output.shape)}")

# 4. Start the interactive session
sv.pytorch(module=conv2d, terminal_callback=terminal_callback)

Learn More

Model Analysis

Deep-dive into model architecture. Explore layers down to a single parameter.

import svetoviz_webgpu as sv

image_processor = DetrImageProcessor.from_pretrained(
    "facebook/detr-resnet-50",
    revision="no_timm",
    device_map="cpu")
model = DetrForObjectDetection.from_pretrained(
    "facebook/detr-resnet-50",
    revision="no_timm",
    device_map="cpu")

sv.model(
    id="detr-resnet-50",
    model=model,
    image_processor=image_processor
)

Learn More

Model Debugger

Interact with your model and investigate activations in real-time.

import svetoviz_webgpu as sv

tokenizer = GPT2Tokenizer.from_pretrained('gpt2', device_map="cpu")
model = GPT2LMHeadModel.from_pretrained('gpt2', device_map="cpu")

def terminal_callback(buffer, message, images, files):
    text = message  # "Replace me by any text you'd like."
    buffer.send_user_message(text)
    inputs = tokenizer(text, return_tensors="pt")

    # Generate continuation
    outputs = model.generate(**inputs, max_new_tokens=50)

    # Decode to string
    decoded = tokenizer.decode(outputs[0], skip_special_tokens=True)
    buffer.send_system_message(decoded)

sv.model(
    id="gpt2",
    model=model,
    tokenizer=tokenizer,
    terminal_callback=terminal_callback
)

Learn More

Dataset Explorer

Efficiently stream and visualize large-scale HuggingFace datasets from local storage or remote server.

import svetoviz_webgpu as sv

hf_name = "eltorio/ROCOv2-radiology"
dataset = load_dataset(hf_name, split="train")
view = DatasetView(
    id=hf_name,
    dataset=dataset,
    image_columns=["image"],
    cell_width=400,
    cell_height=400
)
sv.save_to_disc(view=view,
                compression="jpeg_0",
                directory=f"/Volumes/Untitled/{hf_name}")

sv.load_from_disc(directory=f"/Volumes/Untitled/{hf_name}")

Learn More

Image Browser

Browse through massive collections of images regardless of resolution.

import svetoviz_webgpu as sv

view = DirectorySpiralView(
    id="my_images"
    directory="/Desktop/nasaimages"
)

sv.save_to_disc(view=view,
                compression="jpeg_0",
                directory="/Volumes/Untitled/universe")

sv.load_from_disc(directory="/Volumes/Untitled/universe")

Learn More

Guide

Explore installation guides and user manuals for Svetoviz

Next