---
language: afr
license: apache-2.0
tags:
  - trimmed
  - qwen3.5
base_model: Qwen/Qwen3.5-0.8B
base_model_relation: quantized
datasets:
  - Lumberjackk/fineweb-2-trimming
---

# Qwen3.5-0.8B-afr-32768

This model is a **25.88% smaller** version of [Qwen/Qwen3.5-0.8B](https://huggingface.co/Qwen/Qwen3.5-0.8B) optimized for Afrikaans language via vocabulary size reduction using the [trimming](https://huggingface.co/blog/introduction-to-trimming) method.  

This trimmed model should perform similarly to the original model with only **32,768 tokens** and a much smaller memory footprint. However, it may not perform well for other languages as tokens not commonly used in Afrikaans were removed from the vocabulary.

Note: Qwen3.5 is a multimodal (vision-language) model. This trimmed version retains the vision encoder but reduces only the text vocabulary.

## Model Statistics

| Metric | Original | Trimmed | Reduction |
|--------|----------|---------|-----------|
| **Vocabulary size** | 248,044 tokens | 32,768 tokens | **86.79%** |
| **Model size** | 852,985,920 params | 632,260,672 params | **25.88%** |


## Mining Dataset Statistics

- **Number of texts used for mining**: 200,000 texts  
- **Dataset**: [Lumberjackk/fineweb-2-trimming](https://huggingface.co/datasets/Lumberjackk/fineweb-2-trimming)

## Usage

```python
from transformers import AutoModelForCausalLM, AutoTokenizer

model_name = "AlphaEdge-AI/Qwen3.5-0.8B-afr-32768"

# load the tokenizer and the model
tokenizer = AutoTokenizer.from_pretrained(model_name, trust_remote_code=True)
model = AutoModelForCausalLM.from_pretrained(
    model_name,
    torch_dtype="auto",
    device_map="auto",
    trust_remote_code=True
)

# prepare the model input
prompt = "Your prompt in Afrikaans."
messages = [
    {"role": "user", "content": prompt}
]
text = tokenizer.apply_chat_template(
    messages,
    tokenize=False,
    add_generation_prompt=True
)
model_inputs = tokenizer([text], return_tensors="pt").to(model.device)

# conduct text completion
generated_ids = model.generate(
    **model_inputs,
    max_new_tokens=32768
)
output_ids = generated_ids[0][len(model_inputs.input_ids[0]):]
content = tokenizer.decode(output_ids, skip_special_tokens=True)

print("content:", content)

```

## Citation

#### Qwen3.5

```bibtex
@misc{qwen3.5,
    title  = {Qwen3.5: Towards Native Multimodal Agents},
    author = {Qwen Team},
    month  = {February},
    year   = {2026},
    url    = {https://qwen.ai/blog?id=qwen3.5}
}
```