Instructions to use Alsamir/Abjad_0.2 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use Alsamir/Abjad_0.2 with MLX:

# Make sure mlx-vlm is installed
# pip install --upgrade mlx-vlm

from mlx_vlm import load, generate
from mlx_vlm.prompt_utils import apply_chat_template
from mlx_vlm.utils import load_config

# Load the model
model, processor = load("Alsamir/Abjad_0.2")
config = load_config("Alsamir/Abjad_0.2")

# Prepare input
image = ["http://images.cocodataset.org/val2017/000000039769.jpg"]
prompt = "Describe this image."

# Apply chat template
formatted_prompt = apply_chat_template(
    processor, config, prompt, num_images=1
)

# Generate output
output = generate(model, processor, formatted_prompt, image)
print(output)

Notebooks
Google Colab
Kaggle
Local Apps Settings
LM Studio

How to use Alsamir/Abjad_0.2 with Pi:

Start the MLX server

# Install MLX LM:
uv tool install mlx-lm
# Start a local OpenAI-compatible server:
mlx_lm.server --model "Alsamir/Abjad_0.2"

Configure the model in Pi

# Install Pi:
npm install -g @earendil-works/pi-coding-agent
# Add to ~/.pi/agent/models.json:
{
  "providers": {
    "mlx-lm": {
      "baseUrl": "http://localhost:8080/v1",
      "api": "openai-completions",
      "apiKey": "none",
      "models": [
        {
          "id": "Alsamir/Abjad_0.2"
        }
      ]
    }
  }
}

Run Pi

# Start Pi in your project directory:
pi

Hermes Agent

How to use Alsamir/Abjad_0.2 with Hermes Agent:

Start the MLX server

# Install MLX LM:
uv tool install mlx-lm
# Start a local OpenAI-compatible server:
mlx_lm.server --model "Alsamir/Abjad_0.2"

Configure Hermes

# Install Hermes:
curl -fsSL https://hermes-agent.nousresearch.com/install.sh | bash
hermes setup
# Point Hermes at the local server:
hermes config set model.provider custom
hermes config set model.base_url http://127.0.0.1:8080/v1
hermes config set model.default Alsamir/Abjad_0.2

Run Hermes

hermes

Atomic Chat

OpenClaw

How to use Alsamir/Abjad_0.2 with OpenClaw:

Start the MLX server

# Install MLX LM:
uv tool install mlx-lm
# Start a local OpenAI-compatible server:
mlx_lm.server --model "Alsamir/Abjad_0.2"

Configure OpenClaw

# Install OpenClaw:
npm install -g openclaw@latest
# Register the local server and set it as the default model:
openclaw onboard --non-interactive --mode local \
  --auth-choice custom-api-key \
  --custom-base-url http://127.0.0.1:8080/v1 \
  --custom-model-id "Alsamir/Abjad_0.2" \
  --custom-provider-id mlx-lm \
  --custom-compatibility openai \
  --custom-text-input \
  --accept-risk \
  --skip-health

Run OpenClaw

openclaw agent --local --agent main --message "Hello from Hugging Face"

Abjad_0.2 / docs /evaluation.md

Alsamir

Add Hunyuan-style repository documentation and tooling

92134b1 verified about 1 month ago

preview code

Raw History Blame Contribute Delete

1.35 kB

# Evaluation guide

Compare models on the same rendered pages, prompt, resolution, and output budget. A single successful page is not
enough to rank OCR systems.

## Recommended protocol

1. Keep a held-out set of Arabic scans with verified text.
2. Render each PDF page once and use the same image for every model.
3. Use deterministic decoding and the same instruction: exact visible transcription, preserve RTL order and digit
   shapes, ignore watermarks and signatures, and never repeat text.
4. Compare normalized character error rate (CER), word error rate (WER), digit accuracy, and table-cell accuracy.
5. Review layout separately from text accuracy. A model can have good text but poor reading order.

## Reproducible Abjad command

```bash
.venv/bin/python mlx_ocr_pipeline.py page.pdf \
  -o outputs/eval_abjad.txt \
  --model /path/to/Abjad_0.2 \
  --no-adapter \
  --regions \
  --preserve-digit-shapes \
  --max-tokens 4096
```

Keep the raw output. Post-processing that collapses a runaway repetition is a safety filter, not evidence that the
model recognized the original page correctly.

## Reporting

Report the page count, source type, whether labels were human verified, prompt, resolution, model revision, and
the exact metrics. If no ground truth exists, describe the result qualitatively and do not assign a numeric score.

	# Evaluation guide

	Compare models on the same rendered pages, prompt, resolution, and output budget. A single successful page is not
	enough to rank OCR systems.

	## Recommended protocol

	1. Keep a held-out set of Arabic scans with verified text.
	2. Render each PDF page once and use the same image for every model.
	3. Use deterministic decoding and the same instruction: exact visible transcription, preserve RTL order and digit
	shapes, ignore watermarks and signatures, and never repeat text.
	4. Compare normalized character error rate (CER), word error rate (WER), digit accuracy, and table-cell accuracy.
	5. Review layout separately from text accuracy. A model can have good text but poor reading order.

	## Reproducible Abjad command

	```bash
	.venv/bin/python mlx_ocr_pipeline.py page.pdf \
	-o outputs/eval_abjad.txt \
	--model /path/to/Abjad_0.2 \
	--no-adapter \
	--regions \
	--preserve-digit-shapes \
	--max-tokens 4096
	```

	Keep the raw output. Post-processing that collapses a runaway repetition is a safety filter, not evidence that the
	model recognized the original page correctly.

	## Reporting

	Report the page count, source type, whether labels were human verified, prompt, resolution, model revision, and
	the exact metrics. If no ground truth exists, describe the result qualitatively and do not assign a numeric score.