Text Generation
Transformers
Safetensors
English
qwen3
muon
gqa
rope
swiglu
chatml
conversational
text-generation-inference
Instructions to use 8BitStudio/Sekai-1 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use 8BitStudio/Sekai-1 with Transformers:
# Use a pipeline as a high-level helper from transformers import pipeline pipe = pipeline("text-generation", model="8BitStudio/Sekai-1") messages = [ {"role": "user", "content": "Who are you?"}, ] pipe(messages)# pip install -U transformers accelerate # Load model directly from transformers import AutoTokenizer, AutoModelForCausalLM tokenizer = AutoTokenizer.from_pretrained("8BitStudio/Sekai-1") model = AutoModelForCausalLM.from_pretrained("8BitStudio/Sekai-1", 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=256) print(tokenizer.decode(outputs[0][inputs["input_ids"].shape[-1]:])) - Notebooks
- Google Colab
- Kaggle
- Local Apps Settings
- vLLM
How to use 8BitStudio/Sekai-1 with vLLM:
Install from pip and serve model
# Install vLLM from pip: pip install vllm # Start the vLLM server: vllm serve "8BitStudio/Sekai-1" # Call the server using curl (OpenAI-compatible API): curl -X POST "http://localhost:8000/v1/chat/completions" \ -H "Content-Type: application/json" \ --data '{ "model": "8BitStudio/Sekai-1", "messages": [ { "role": "user", "content": "What is the capital of France?" } ] }'Use Docker
docker model run hf.co/8BitStudio/Sekai-1
- SGLang
How to use 8BitStudio/Sekai-1 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 "8BitStudio/Sekai-1" \ --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": "8BitStudio/Sekai-1", "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 "8BitStudio/Sekai-1" \ --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": "8BitStudio/Sekai-1", "messages": [ { "role": "user", "content": "What is the capital of France?" } ] }' - Docker Model Runner
How to use 8BitStudio/Sekai-1 with Docker Model Runner:
docker model run hf.co/8BitStudio/Sekai-1
File size: 3,751 Bytes
0667735 fa03aac 9ecefd4 2518111 0667735 fa03aac 871f0c8 fa03aac 2518111 | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 | ---
license: apache-2.0
language:
- en
pipeline_tag: text-generation
tags:
- muon
- gqa
- rope
- swiglu
- chatml
library_name: transformers
datasets:
- karpathy/climbmix-400b-shuffle
- HuggingFaceTB/smollm-corpus
- HuggingFaceTB/finemath
- HuggingFaceTB/cosmopedia
- openbmb/UltraX-Preview
- allenai/dolma3_dolmino_mix-100B-1025
- allenai/dolma3.5_pool
- allenai/dolmino-mix-1124
- IFM/TxT360-v2
- nvidia/OpenMathInstruct-2
- nvidia/Nemotron-MIND
- openbmb/UltraData-Math
- microsoft/orca-math-word-problems-200k
- meta-math/MetaMathQA
- OpenCoder-LLM/opc-annealing-corpus
- HuggingFaceTB/smoltalk
- HuggingFaceTB/smoltalk2
- HuggingFaceH4/ultrachat_200k
- allenai/tulu-3-sft-mixture
- bigcode/commitpackft
---

Sekai-1 is the first model in the Sekai series: a small language model trained from scratch on a single RTX 5060 Ti.
The model has 95.6M parameters, with 23 layers, a hidden size of 576 and a 2,048-token context window. It uses a Qwen3-style decoder-only architecture (grouped-query attention with 9 query heads and 3 key-value heads, a SwiGLU MLP, RMSNorm, RoPE, tied embeddings) and its own 24,576-token BPE tokenizer, trained on Sekai’s data mix, that splits numbers into single digits.
Sekai-1 is pretrained on 58.5B tokens of web, educational, math, code and conversation data, using the Muon and AdamW optimizers with a warmup-stable-decay learning-rate schedule.
## Evaluation

The model scores number two in the under 100M parameters, just behind [KLM-1.5](https://huggingface.co/MinimaLabs/KLM-1.5) on the [SLM leaderboard](https://huggingface.co/spaces/AxiomicLabs/Open_SLM_Leaderboard) as seen in the image above.
All scores are zero-shot. Sekai-1's scores were measured locally on the same tasks the leaderboard uses; the other models' scores are taken from the leaderboard (November 7, 2026).
## Training
**Pretraining:** 58.5B tokens of web, educational, math, code and conversation data, with a warmup-stable-decay schedule. The final ~12B tokens were a decay phase on a higher-quality mix.
**Chat fine-tuning (SFT):** about 400M tokens of instruction following, everyday chat, short and long answers, summarizing and rewriting, code writing and code fixing, plus a small set of identity examples, so the model knows it is Sekai-1. It does not use chain-of-thought (a model this size doesn't benefit from it), and refusals were filtered out of the data.
## How to use
```bash
pip install transformers torch
```
```python
from transformers import AutoModelForCausalLM, AutoTokenizer
model_id = "8BitStudio/Sekai-1"
tokenizer = AutoTokenizer.from_pretrained(model_id)
model = AutoModelForCausalLM.from_pretrained(model_id)
messages = [{"role": "user", "content": "Who are you?"}]
inputs = tokenizer.apply_chat_template(messages, add_generation_prompt=True, return_dict=True, return_tensors="pt")
outputs = model.generate(**inputs)
print(tokenizer.decode(outputs[0][inputs["input_ids"].shape[-1]:], skip_special_tokens=True))
```
The model is small enough to run on a CPU. The default sampling settings (temperature 0.5, top-p 0.9, top-k 40, repetition penalty 1.1) are stored in `generation_config.json`; pass `do_sample=False` to `generate()` for deterministic answers.
Sekai-1 uses the ChatML format:
```
<|im_start|>user
What is the capital of France?<|im_end|>
<|im_start|>assistant
Paris<|im_end|>
```
## Limitations
Sekai-1 is a 95.6M-parameter model. It makes factual mistakes, loses track in long conversations, and is weak at multi-step reasoning and math beyond simple arithmetic.
It has no safety tuning or refusal behavior, so don't use it anywhere that needs either. It understands English only. Check responses for the things that matter. |