Instructions to use ceyda/Qwen3-1.7B-Base-trim-koen-32768 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use ceyda/Qwen3-1.7B-Base-trim-koen-32768 with Transformers:

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

pipe = pipeline("text-generation", model="ceyda/Qwen3-1.7B-Base-trim-koen-32768")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("ceyda/Qwen3-1.7B-Base-trim-koen-32768")
model = AutoModelForCausalLM.from_pretrained("ceyda/Qwen3-1.7B-Base-trim-koen-32768", device_map="auto")
messages = [
    {"role": "user", "content": "Who are you?"},
]
inputs = tokenizer.apply_chat_template(
	messages,
	add_generation_prompt=True,
	tokenize=True,
	return_dict=True,
	return_tensors="pt",
).to(model.device)

outputs = model.generate(**inputs, max_new_tokens=40)
print(tokenizer.decode(outputs[0][inputs["input_ids"].shape[-1]:]))

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use ceyda/Qwen3-1.7B-Base-trim-koen-32768 with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "ceyda/Qwen3-1.7B-Base-trim-koen-32768"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "ceyda/Qwen3-1.7B-Base-trim-koen-32768",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/ceyda/Qwen3-1.7B-Base-trim-koen-32768

SGLang

How to use ceyda/Qwen3-1.7B-Base-trim-koen-32768 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 "ceyda/Qwen3-1.7B-Base-trim-koen-32768" \
    --host 0.0.0.0 \
    --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "ceyda/Qwen3-1.7B-Base-trim-koen-32768",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

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 "ceyda/Qwen3-1.7B-Base-trim-koen-32768" \
        --host 0.0.0.0 \
        --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "ceyda/Qwen3-1.7B-Base-trim-koen-32768",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use ceyda/Qwen3-1.7B-Base-trim-koen-32768 with Docker Model Runner:
```
docker model run hf.co/ceyda/Qwen3-1.7B-Base-trim-koen-32768
```

Qwen3-1.7B-Base — vocabulary-trimmed (Korean + English, 32,768)

A vocabulary-trimmed version of Qwen/Qwen3-1.7B-Base: the byte-level BPE vocabulary is reduced from ~151.7k to 32768 tokens covering Korean + English, and the (tied) embedding matrix is sliced to match. No retraining — weights are copied verbatim for kept tokens, so on kept tokens the model is numerically identical to the original.

Method: Introduction to Trimming.

What changed

	Original	Trimmed
Parameters	1,720,574,976	1,476,518,912 (-14.2%)
Vocab size	151,669	32,768
Merges	151,387	32,486
Embedding	tied	tied (sliced)

All 256 byte-level tokens + all 26 special tokens are kept, plus the most frequent Korean/English tokens and their full BPE merge-derivation closure (so multi-byte Korean stays reachable).
Verified: lossless round-trip on KO/EN/code; teacher-forced logit equivalence max|Δ| = 0.0; greedy generation on natural Korean/English is token-for-token identical to the base model.

Usage

from transformers import AutoModelForCausalLM, AutoTokenizer

repo = "ceyda/Qwen3-1.7B-Base-trim-koen-32768"
tok = AutoTokenizer.from_pretrained(repo)
model = AutoModelForCausalLM.from_pretrained(repo)

ids = tok("대한민국의 수도는", return_tensors="pt")
print(tok.decode(model.generate(**ids, max_new_tokens=20)[0], skip_special_tokens=True))

Limitations

Trimmed for Korean + English prose. Code and rare jargon still encode losslessly but may split into a few extra tokens. Text in other languages falls back to byte tokens (longer sequences).
This is a base (non-instruction-tuned) model.

Attribution

Derived from Qwen/Qwen3-1.7B-Base (Apache-2.0). Trimming method by Loïck Bourdois.

Downloads last month: 325

Safetensors

Model size

1B params

Tensor type

BF16

Model tree for ceyda/Qwen3-1.7B-Base-trim-koen-32768

Base model

Qwen/Qwen3-1.7B-Base

Finetuned

(452)

this model