[WIP] InferenceClient.post is deprecated, but Sentence Ranking tasks are not implemented

[Copilot]

Thanks for assigning this issue to me. I'm starting to work on it and will keep this PR's description up to date as I form a plan and make progress.

Original issue description:

Describe the bug

The InferenceClient API still does not support many of the tasks that can be hosted at inference endpoints, but gives a deprecation warning when using .post to get around this.

Reproduction

from huggingface_hub import InferenceClient, get_inference_endpoint
import json

# Get endpoint and create client
MODEL_NAME = "YOUR_MODEL_NAME_OR_ENDPOINT_NAME"
NAMESPACE = "YOUR_NAMESPACE"
endpoint = get_inference_endpoint(MODEL_NAME, namespace=NAMESPACE)
client = InferenceClient(endpoint.url, timeout=10)

# Test data
query = "What is the capital of France?"
document = "Paris is the capital of France."
sentence_ranking_style_inputs = [[query, document]]
text_classification_style_inputs = [{"text": query, "text_pair": document}]

# 1. Using post method
response_bytes = client.post(json={"inputs": sentence_ranking_style_inputs})
print(json.loads(response_bytes))
# Problem: post method has deprecation warning

# 2. Using text_classification task
try:
    result = client.text_classification(text_classification_style_inputs) # There's no way to inject the inputs format that would have worked on this task for reranking
    print(result)
except Exception as e:
    print(f"text_classification error: {e}")
# Problem: text_classification doesn't properly support text pairs format needed for reranking/cross-encoding

# 3. Using sentence_similarity task
try:
    result = client.sentence_similarity(sentence_ranking_style_inputs) # There's no direct way to inject the inputs format that would have worked on this task for reranking
    print(result)
except Exception as e:
    print(f"sentence_similarity error: {e}")
# Problem: No direct support for sentence ranking despite endpoint supporting this task

Ideally, the sentence_ranking task is supported.

Logs

System info

- huggingface_hub version: 0.30.2
- Platform: macOS-15.4-arm64-arm-64bit
- Python version: 3.12.10
- Running in iPython ?: No
- Running in notebook ?: No
- Running in Google Colab ?: No
- Running in Google Colab Enterprise ?: No
- Token path ?: redacted
- Has saved token ?: False
- Configured git credential helpers: redacted
- FastAI: N/A
- Tensorflow: N/A
- Torch: 2.7.0
- Jinja2: 3.1.6
- Graphviz: N/A
- keras: N/A
- Pydot: N/A
- Pillow: 11.2.1
- hf_transfer: N/A
- gradio: N/A
- tensorboard: N/A
- numpy: 1.26.4
- pydantic: 2.11.4
- aiohttp: 3.11.18
- hf_xet: N/A
- ENDPOINT: https://huggingface.co
- HF_HUB_CACHE: redacted
- HF_ASSETS_CACHE: redacted
- HF_TOKEN_PATH: redacted
- HF_STORED_TOKENS_PATH: redacted
- HF_HUB_OFFLINE: False
- HF_HUB_DISABLE_TELEMETRY: False
- HF_HUB_DISABLE_PROGRESS_BARS: None
- HF_HUB_DISABLE_SYMLINKS_WARNING: False
- HF_HUB_DISABLE_EXPERIMENTAL_WARNING: False
- HF_HUB_DISABLE_IMPLICIT_TOKEN: False
- HF_HUB_ENABLE_HF_TRANSFER: False
- HF_HUB_ETAG_TIMEOUT: 10
- HF_HUB_DOWNLOAD_TIMEOUT: 10

Fixes #3055.

---

💡 You can make Copilot smarter by setting up custom instructions, customizing its development environment and configuring Model Context Protocol (MCP) servers. Learn more Copilot coding agent tips in the docs.