Instructions to use AhiskaAI/AhiskaAI-308m-Base-v0.2 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use AhiskaAI/AhiskaAI-308m-Base-v0.2 with Transformers:

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

pipe = pipeline("text-generation", model="AhiskaAI/AhiskaAI-308m-Base-v0.2")

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("AhiskaAI/AhiskaAI-308m-Base-v0.2")
model = AutoModelForCausalLM.from_pretrained("AhiskaAI/AhiskaAI-308m-Base-v0.2", device_map="auto")

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use AhiskaAI/AhiskaAI-308m-Base-v0.2 with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "AhiskaAI/AhiskaAI-308m-Base-v0.2"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "AhiskaAI/AhiskaAI-308m-Base-v0.2",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker

docker model run hf.co/AhiskaAI/AhiskaAI-308m-Base-v0.2

SGLang

How to use AhiskaAI/AhiskaAI-308m-Base-v0.2 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 "AhiskaAI/AhiskaAI-308m-Base-v0.2" \
    --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": "AhiskaAI/AhiskaAI-308m-Base-v0.2",
		"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 "AhiskaAI/AhiskaAI-308m-Base-v0.2" \
        --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": "AhiskaAI/AhiskaAI-308m-Base-v0.2",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Docker Model Runner
How to use AhiskaAI/AhiskaAI-308m-Base-v0.2 with Docker Model Runner:
```
docker model run hf.co/AhiskaAI/AhiskaAI-308m-Base-v0.2
```

AhiskaAI-308m-Base-v0.2

AhiskaAI-308m-Base-v0.2 is a 308 million parameter Small Language Model (SLM) built entirely from scratch. As the largest model in the AhiskaAI v0.2 family, it is designed to deliver stronger Turkish language understanding while maintaining efficient deployment on consumer hardware.

Model Details

Architecture: Llama-based architecture.
Parameters: 308M.
Context Window: 1024 tokens.
Tokenizer: Custom BPE Tokenizer (Vocabulary Size: 32,000).
Training Framework: PyTorch & Transformers.

Data Curation (The "Quality over Quantity" Approach)

The v0.2 release adopts a data-centric training strategy, prioritizing corpus quality over dataset size.

Raw Data (v0.1): 5GB of raw Turkish corpus.
Curated Data (v0.2): 1.2GB of carefully filtered, high-quality Turkish text.
Process: Approximately 75% of noisy, duplicated, and low-quality samples were removed to improve linguistic quality and training efficiency.

Key Improvements from v0.1

Architecture Shift: Migrated from GPT-2 to a modern Llama-based architecture.
Normalization: RMSNorm.
Positional Encoding: RoPE (Rotary Positional Embeddings).
Activation: SiLU.
Precision: Trained using bfloat16 for efficient consumer GPU training.

Design Goal

The 308M model serves as the flagship base model of the AhiskaAI v0.2 family.

Its primary objectives are:

Strong Turkish language modeling.
Improved semantic understanding.
Better contextual consistency.
A research foundation for future instruction tuning and alignment.

Training Logs

The graph above demonstrates the training convergence of AhiskaAI-308m-Base-v0.2. The stable decline in loss indicates effective optimization throughout pretraining.

Usage

from transformers import AutoModelForCausalLM, AutoTokenizer

model = AutoModelForCausalLM.from_pretrained("AhiskaAI/AhiskaAI-308m-Base-v0.2")
tokenizer = AutoTokenizer.from_pretrained("AhiskaAI/AhiskaAI-308m-Base-v0.2")

text = "Türkiye Cumhuriyeti"

inputs = tokenizer(text, return_tensors="pt")

outputs = model.generate(**inputs, max_new_tokens=50)

print(tokenizer.decode(outputs[0], skip_special_tokens=True))

Training Hardware

Trained on NVIDIA RTX 4050 6GB Laptop GPU.

Future Plans

Instruction-tuned (IT) version.
Preference alignment with DPO.
Larger and more diverse Turkish datasets.
Future AhiskaAI v0.3 model family.

About AhiskaAI

AhiskaAI is an independent open-source initiative dedicated to developing efficient Turkish Small Language Models trained completely from scratch.

Downloads last month: 1

Safetensors

Model size

0.3B params

Tensor type

BF16

Collection including AhiskaAI/AhiskaAI-308m-Base-v0.2

AhıskaAI-v0.2-Collection

Collection

The v0.2 release of AhiskaAI models, trained from scratch on curated Turkish datasets. This series features a Llama-based architecture, optimized data • 6 items • Updated 3 days ago • 1