Instructions to use canbingol/gemma3_1B_base-tr-cpt-1epoch_stage3 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use canbingol/gemma3_1B_base-tr-cpt-1epoch_stage3 with Transformers:

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

pipe = pipeline("text-generation", model="canbingol/gemma3_1B_base-tr-cpt-1epoch_stage3")

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("canbingol/gemma3_1B_base-tr-cpt-1epoch_stage3")
model = AutoModelForCausalLM.from_pretrained("canbingol/gemma3_1B_base-tr-cpt-1epoch_stage3", device_map="auto")

Inference
Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use canbingol/gemma3_1B_base-tr-cpt-1epoch_stage3 with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "canbingol/gemma3_1B_base-tr-cpt-1epoch_stage3"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "canbingol/gemma3_1B_base-tr-cpt-1epoch_stage3",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker

docker model run hf.co/canbingol/gemma3_1B_base-tr-cpt-1epoch_stage3

SGLang

How to use canbingol/gemma3_1B_base-tr-cpt-1epoch_stage3 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 "canbingol/gemma3_1B_base-tr-cpt-1epoch_stage3" \
    --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": "canbingol/gemma3_1B_base-tr-cpt-1epoch_stage3",
		"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 "canbingol/gemma3_1B_base-tr-cpt-1epoch_stage3" \
        --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": "canbingol/gemma3_1B_base-tr-cpt-1epoch_stage3",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Docker Model Runner
How to use canbingol/gemma3_1B_base-tr-cpt-1epoch_stage3 with Docker Model Runner:
```
docker model run hf.co/canbingol/gemma3_1B_base-tr-cpt-1epoch_stage3
```

A newer version of this model is available: canbingol/gemma3_1B_base-tr-cpt-2nd_epoch_stage1

Model Card: Gemma3-1B Turkish CPT (100K–150K Subset, 1 Epoch – Stage 3)

Overview

This model is the Stage 3 Turkish Continued Pretraining (CPT) variant of Gemma-3-1B.

Unlike Stage 1, which was initialized from google/gemma-3-1b-pt,
this model was initialized from:

canbingol/gemma3_1B_base-tr-cpt-1epoch_stage2

Stage 3 continues domain adaptation by exposing the model to new data rather than repeating the same subset.

The model was trained for 1 epoch on samples 100,000 to 150,000 of the Turkish web corpus.

Importantly, this model is a direct continuation of Stage 2.
Therefore, cumulatively it has been trained on samples 0–150,000 of the corpus (Stage 1: 0–50K, Stage 2: 50K–100K, Stage 3: 100K–150K).

Training Lineage

Stage 0: google/gemma-3-1b-pt
Stage 1: Samples 0–50,000 (1 epoch)
Stage 2: Samples 50,000–100,000 (1 epoch)
Stage 3 (this release): Samples 100,000–150,000 (1 epoch)

Cumulative data exposure: 0–150,000 samples

This represents sequential CPT across disjoint data shards.

Training Setup

Dataset: canbingol/vngrs-web-corpus-200k
Subset Used: Samples 100,000–150,000
Initialization: Stage 2 checkpoint
Training Objective: Continued Pretraining
Epochs: 1
Data Regime: Plain text
Token Count: ~21.5M tokens
Cumulative Token Exposure (Stage 1 + Stage 2 + Stage 3): ~64.5M tokens (approximate)

Usage Example

import torch
from transformers import AutoModelForCausalLM, AutoTokenizer

model_id = "canbingol/gemma3_1B_base-tr-cpt-1epoch_stage3"

device = "cuda" if torch.cuda.is_available() else "cpu"

model = AutoModelForCausalLM.from_pretrained(model_id)
tokenizer = AutoTokenizer.from_pretrained(model_id)

model = model.to(device)

prompt = "bundan böyle"
inputs = tokenizer(prompt, return_tensors="pt").to(device)

outputs = model.generate(
    **inputs,
    max_new_tokens=50,
    do_sample=True,
    temperature=0.8,
    top_p=0.9
)

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