---
license: apache-2.0
base_model: InternScience/Agents-A1-4B
library_name: mlx
tags:
- mlx
- vision
- agentic
pipeline_tag: image-text-to-text
---

# ToPo-ToPo/Agents-A1-4B-mlx-4bit

MLX **4bit** conversion of [`InternScience/Agents-A1-4B`](https://huggingface.co/InternScience/Agents-A1-4B) for Apple Silicon (mlx-vlm).

`Agents-A1-4B` is a `Qwen3_5ForConditionalGeneration` (Qwen3.5-VL family) vision-language
model. The text tower uses a hybrid of linear attention and full attention
(`full_attention_interval: 4`) with interleaved MRoPE.

## Provenance (self-converted from official weights)
- Source: [`InternScience/Agents-A1-4B`](https://huggingface.co/InternScience/Agents-A1-4B) (license: apache-2.0)
- Tool: `mlx-vlm 0.6.6` — `mlx_vlm.convert --hf-path InternScience/Agents-A1-4B --mlx-path . -q --q-bits 4 --q-group-size 64`
- Effective: **5.347 bits/weight**
- Size on disk: 2.8 GB — peak memory 4.2 GB, ~165 tok/s generation (M-series, short prompt)

Effective bit width is higher than nominal because embeddings and parts of the
vision tower are kept at higher precision (mlx default behaviour).

## Validation
Checked after conversion: correctly described a synthetic test image (red square /
blue circle, both shape and colour), and answered a text-only arithmetic prompt
correctly. This is a smoke test, not a benchmark — no eval suite was run.

## Usage
```python
from mlx_vlm import load, generate
model, processor = load("ToPo-ToPo/Agents-A1-4B-mlx-4bit")
```

CLI:
```bash
mlx_vlm.generate --model ToPo-ToPo/Agents-A1-4B-mlx-4bit \
  --image path/to/image.png --prompt "Describe this image." --max-tokens 200
```

## Requirements
`mlx-vlm >= 0.6.6`. Earlier versions do not implement the `qwen3_5`
architecture and cannot load or convert this model.

## License
Apache-2.0, inherited from the base model. Weights were converted/quantized to
MLX format; no other modification to the weights.

## Note on config files
`tokenizer_config.json` and `processor_config.json` were re-serialized by
`transformers 5.14.1` during conversion, so they differ textually from upstream:
the chat template now lives in `chat_template.jinja` (byte-identical to upstream),
special tokens moved into `model_specific_special_tokens`, and the image processor
uses `min_pixels`/`max_pixels` instead of `size`/`resample`. The tokenizer was
verified functionally identical to upstream (identical vocab, and identical
encode/decode on a spot-check set). Users on `transformers 4.x` should prefer the
upstream config files.