Instructions to use rafat234/compassai-geo-nlp with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use rafat234/compassai-geo-nlp with Transformers:

# Use a pipeline as a high-level helper
from transformers import pipeline

pipe = pipeline("text-generation", model="rafat234/compassai-geo-nlp")

# Load model directly
from transformers import AutoTokenizer, AutoModelForSeq2SeqLM

tokenizer = AutoTokenizer.from_pretrained("rafat234/compassai-geo-nlp")
model = AutoModelForSeq2SeqLM.from_pretrained("rafat234/compassai-geo-nlp")

PEFT
How to use rafat234/compassai-geo-nlp with PEFT:
```
Task type is invalid.
```
Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use rafat234/compassai-geo-nlp with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "rafat234/compassai-geo-nlp"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "rafat234/compassai-geo-nlp",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker

docker model run hf.co/rafat234/compassai-geo-nlp

SGLang

How to use rafat234/compassai-geo-nlp with SGLang:

Install from pip and serve model

# Install SGLang from pip:
pip install sglang
# Start the SGLang server:
python3 -m sglang.launch_server \
    --model-path "rafat234/compassai-geo-nlp" \
    --host 0.0.0.0 \
    --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "rafat234/compassai-geo-nlp",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker images

docker run --gpus all \
    --shm-size 32g \
    -p 30000:30000 \
    -v ~/.cache/huggingface:/root/.cache/huggingface \
    --env "HF_TOKEN=<secret>" \
    --ipc=host \
    lmsysorg/sglang:latest \
    python3 -m sglang.launch_server \
        --model-path "rafat234/compassai-geo-nlp" \
        --host 0.0.0.0 \
        --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "rafat234/compassai-geo-nlp",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Docker Model Runner
How to use rafat234/compassai-geo-nlp with Docker Model Runner:
```
docker model run hf.co/rafat234/compassai-geo-nlp
```

YAML Metadata Warning:The pipeline tag "text2text-generation" is not in the official list: text-classification, token-classification, table-question-answering, question-answering, zero-shot-classification, translation, summarization, feature-extraction, text-generation, fill-mask, sentence-similarity, text-to-speech, text-to-audio, automatic-speech-recognition, audio-to-audio, audio-classification, audio-text-to-text, voice-activity-detection, depth-estimation, image-classification, object-detection, image-segmentation, text-to-image, image-to-text, image-to-image, image-to-video, unconditional-image-generation, video-classification, reinforcement-learning, robotics, tabular-classification, tabular-regression, tabular-to-text, table-to-text, multiple-choice, text-ranking, text-retrieval, time-series-forecasting, text-to-video, image-text-to-text, image-text-to-image, image-text-to-video, visual-question-answering, document-question-answering, zero-shot-image-classification, graph-ml, mask-generation, zero-shot-object-detection, text-to-3d, image-to-3d, image-feature-extraction, video-text-to-text, keypoint-detection, visual-document-retrieval, any-to-any, video-to-video, other

Model Card: CompassAI Geo-NLP (v2)

An English and Arabic bilingual Geo-NLP Semantic Parser fine-tuned from google/flan-t5-base using PEFT (LoRA).

This model functions as a Text-to-GIS API Agent designed specifically to translate natural language user queries (in both English and Arabic) into structured JSON payloads. These payloads map directly to operations, spatial/attribute filters, and parameters within the ArcGIS Maps SDK for JavaScript.

Model Details

Model Description

Developed by: Rafat Kamel
Model type: Encoder-Decoder (Text2Text Generation)
Language(s) (NLP): English (en), Arabic (ar)
License: Apache 2.0
Finetuned from model: google/flan-t5-base
Training Method: PEFT - Parameter-Efficient Fine-Tuning via LoRA (Low-Rank Adaptation)

Model Sources

Repository: rafat234/compassai-geo-nlp

Uses

Direct Use

The model is built to serve as the core natural language interface for Web GIS applications. It accurately parses raw textual instructions into semantic components including:

Intent Identification (select_by_attribute, select_by_location, symbology, zoom_navigate, layer_toggle, etc.)
ArcGIS API Mapping (FeatureLayer.queryFeatures, geometryEngine.buffer, ClassBreaksRenderer, etc.)
SQL Where Clauses and structured parameter extraction for spatial analysis.

Intent & Schema Preview

The model maps user commands into a robust JSON schema:

{
  "input": "User natural language query",
  "intent": "GIS operational intent",
  "where_clause": "SQL filter expression or null",
  "primary_api": "ArcGIS JS API class/method",
  "api_call": "JavaScript executable code snippet",
  "api_inputs": { "extracted_parameters": "values" },
  "layer": "target_layer_name",
  "language": "en/ar"
}

Downloads last month: 34

Safetensors

Model size

0.2B params

Tensor type

F32

Model tree for rafat234/compassai-geo-nlp

Base model

google/flan-t5-base

Finetuned

(913)

this model