Fill Mask

Our fill mask endpoint predicts masked tokens in text sequences, similar to BERT-style masked language modeling. Here is an example of fill mask: Parameters:

model: The model ID of the model you want to use
inputs: The text sequence(s) with masked tokens (typically using <mask> tokens)
targets: Specific target tokens to consider (optional)
top_k: The number of top predictions to return (default: 5)

from synapsai import SynapsAI

model = SynapsAI("your-api-key")

response = model.fill_mask.create(
    model="model-id",
    inputs="The quick brown <mask> jumps over the lazy dog.",
    targets=None,
    top_k=5
)

for result in response.data:
    print("Masks:", result.masks)
    print("Score:", result.score)

curl -X POST https://api.synapsai.cloud/v1/fill-mask \
  -H "Authorization: Bearer $SYNAPSAI_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
        "model": "model-id",
        "inputs": "The quick brown <mask> jumps over the lazy dog.",
        "targets": null,
        "top_k": 5
    }'

Here is an example with multiple inputs and specific targets:

from synapsai import SynapsAI

model = SynapsAI("your-api-key")

response = model.fill_mask.create(
    model="model-id",
    inputs=[
        "The capital of France is <mask>.",
        "Machine learning is a subset of <mask>."
    ],
    targets=["Paris", "artificial intelligence", "AI", "computer science"],
    top_k=3
)

for result in response.data:
    print("Masks:", result.masks)
    print("Score:", result.score)
    print("---")

curl -X POST https://api.synapsai.cloud/v1/fill-mask \
  -H "Authorization: Bearer $SYNAPSAI_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
        "model": "model-id",
        "inputs": [
          "The capital of France is <mask>.",
          "Machine learning is a subset of <mask>."
        ],
        "targets": ["Paris", "artificial intelligence", "AI", "computer science"],
        "top_k": 3
    }'

For deeper insights into the fill mask endpoint, see our API reference.

Getting started

Guides

Examples

Manage

Platform

Resources

Support

Legal