Text Generation
Transformers
Safetensors
gemma4
image-text-to-text
emotional-support
companion
on-device
conversational
emotion
doi:10.5281/zenodo.23179311
Eval Results
Instructions to use Lythri/Lythri-Wink-2B with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use Lythri/Lythri-Wink-2B with Transformers:
# Use a pipeline as a high-level helper from transformers import pipeline pipe = pipeline("text-generation", model="Lythri/Lythri-Wink-2B") messages = [ { "role": "user", "content": [ {"type": "image", "url": "https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/p-blog/candy.JPG"}, {"type": "text", "text": "What animal is on the candy?"} ] }, ] pipe(text=messages)# pip install -U transformers accelerate # Load model directly from transformers import AutoProcessor, AutoModelForMultimodalLM processor = AutoProcessor.from_pretrained("Lythri/Lythri-Wink-2B") model = AutoModelForMultimodalLM.from_pretrained("Lythri/Lythri-Wink-2B", device_map="auto") messages = [ { "role": "user", "content": [ {"type": "image", "url": "https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/p-blog/candy.JPG"}, {"type": "text", "text": "What animal is on the candy?"} ] }, ] inputs = processor.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(processor.decode(outputs[0][inputs["input_ids"].shape[-1]:])) - Notebooks
- Google Colab
- Kaggle
- Local Apps Settings
- vLLM
How to use Lythri/Lythri-Wink-2B with vLLM:
Install from pip and serve model
# Install vLLM from pip: pip install vllm # Start the vLLM server: vllm serve "Lythri/Lythri-Wink-2B" # Call the server using curl (OpenAI-compatible API): curl -X POST "http://localhost:8000/v1/chat/completions" \ -H "Content-Type: application/json" \ --data '{ "model": "Lythri/Lythri-Wink-2B", "messages": [ { "role": "user", "content": "What is the capital of France?" } ] }'Use Docker
docker model run hf.co/Lythri/Lythri-Wink-2B
- SGLang
How to use Lythri/Lythri-Wink-2B 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 "Lythri/Lythri-Wink-2B" \ --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": "Lythri/Lythri-Wink-2B", "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 "Lythri/Lythri-Wink-2B" \ --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": "Lythri/Lythri-Wink-2B", "messages": [ { "role": "user", "content": "What is the capital of France?" } ] }' - Docker Model Runner
How to use Lythri/Lythri-Wink-2B with Docker Model Runner:
docker model run hf.co/Lythri/Lythri-Wink-2B
|
Download README.md from Lythri/Lythri-Wink-2B: direct link, hf CLI and curl.
- Browser
- Download file 7.72 kB
-
https://huggingface.co/Lythri/Lythri-Wink-2B/resolve/main/README.md
- Command line
-
hf download hf://Lythri/Lythri-Wink-2B/README.md
-
curl -L -o README.md https://huggingface.co/Lythri/Lythri-Wink-2B/resolve/main/README.md
7.72 kB
| license: apache-2.0 | |
| base_model: google/gemma-4-E2B | |
| tags: | |
| - emotional-support | |
| - companion | |
| - on-device | |
| - conversational | |
| - emotion | |
| - doi:10.5281/zenodo.23179311 | |
| library_name: transformers | |
| pipeline_tag: text-generation | |
| <p align="center"> | |
| <img src="https://huggingface.co/spaces/Lythri/README/resolve/main/Banner_flat.png" width="600" alt="Lythri"/> | |
| </p> | |
| <p align="center"> | |
| <a href="https://huggingface.co/Lythri">Hugging Face</a> | | |
| <a href="https://www.modelscope.cn/profile/Spike8086">ModelScope</a> | | |
| <a href="https://zenodo.org/records/23179312">Technical Report</a> | |
| </p> | |
| ## Introduction | |
| **Lythri-Wink-2B** is part of the **Lythri 2.0** family, with **2.52B total parameters** and **1.48B activated parameters**. It is a compact, direct-response model optimized for on-device conversation and emotional support. | |
| Compared with the larger Lythri-Tada-4B, Wink focuses on three practical features: | |
| - **Smaller deployment footprint.** Wink substantially reduces both total and activated parameters. | |
| - **Multi-turn conversations.** Wink is trained on multi-turn companion conversations and earlier-turn context. | |
| - **System prompt support.** Like Lythri-Tada-4B, Wink follows system prompts, so you can set a persona, tone or rules for your app directly in the system message. | |
| Wink trades knowledge, reasoning, math and coding performance for its smaller deployment footprint; it is not intended to match the larger Lythri-Tada-4B on general benchmarks. | |
| ## Writing Naturalness | |
| <p align="center"> | |
| <img src="https://huggingface.co/spaces/Lythri/README/resolve/main/Tada_Slop.png" width="800" alt="EQ-Bench Slop Score"/> | |
| </p> | |
| Within the same family, **Lythri-Tada-4B scores 39.0 on the [EQ-Bench Slop Score](https://eqbench.com/slop-score.html)** and writes more naturally than GPT-4o (47.7) and DeepSeek-R1 (67.9), far ahead of Qwen3-4B (62.2) and Gemma-3-4B (85.2). | |
| ## General Benchmarks | |
| <table width="100%" style="display: table; width: 100%; border-collapse: collapse; text-align: center;"> | |
| <thead> | |
| <tr style="border-bottom: 2px solid #338f65;"> | |
| <th width="40%" style="text-align: left; padding: 10px 16px;">Benchmark</th> | |
| <th width="30%" style="color: #338f65; white-space: nowrap; padding: 10px 16px;">Lythri-Tada-4B</th> | |
| <th width="30%" style="background: rgba(51,143,101,0.12); color: #338f65; white-space: nowrap; padding: 10px 16px;">Lythri-Wink-2B</th> | |
| </tr> | |
| </thead> | |
| <tbody> | |
| <tr style="background: rgba(51,143,101,0.06);"> | |
| <td colspan="3" style="text-align: left; color: #338f65; font-weight: bold;">Knowledge</td> | |
| </tr> | |
| <tr><td style="text-align: left;">MMLU</td><td>55.41</td><td style="background: rgba(51,143,101,0.12); color: #338f65; font-weight: bold;">39.32</td></tr> | |
| <tr><td style="text-align: left;">ARC-C</td><td>56.57</td><td style="background: rgba(51,143,101,0.12); color: #338f65; font-weight: bold;">41.13</td></tr> | |
| <tr style="background: rgba(51,143,101,0.06);"> | |
| <td colspan="3" style="text-align: left; color: #338f65; font-weight: bold;">Reasoning</td> | |
| </tr> | |
| <tr><td style="text-align: left;">HellaSwag</td><td>74.23</td><td style="background: rgba(51,143,101,0.12); color: #338f65; font-weight: bold;">58.65</td></tr> | |
| <tr style="background: rgba(51,143,101,0.06);"> | |
| <td colspan="3" style="text-align: left; color: #338f65; font-weight: bold;">General</td> | |
| </tr> | |
| <tr><td style="text-align: left;">SocialIQA</td><td>51.43</td><td style="background: rgba(51,143,101,0.12); color: #338f65; font-weight: bold;">45.24</td></tr> | |
| <tr style="background: rgba(51,143,101,0.06);"> | |
| <td colspan="3" style="text-align: left; color: #338f65; font-weight: bold;">Science</td> | |
| </tr> | |
| <tr><td style="text-align: left;">GPQA Diamond</td><td>21.21</td><td style="background: rgba(51,143,101,0.12); color: #338f65; font-weight: bold;">23.23</td></tr> | |
| <tr style="background: rgba(51,143,101,0.06);"> | |
| <td colspan="3" style="text-align: left; color: #338f65; font-weight: bold;">Math</td> | |
| </tr> | |
| <tr><td style="text-align: left;">GSM8K</td><td>27.45</td><td style="background: rgba(51,143,101,0.12); color: #338f65; font-weight: bold;">13.04</td></tr> | |
| <tr style="background: rgba(51,143,101,0.06);"> | |
| <td colspan="3" style="text-align: left; color: #338f65; font-weight: bold;">Code / Instruction</td> | |
| </tr> | |
| <tr><td style="text-align: left;">HumanEval</td><td>24.39</td><td style="background: rgba(51,143,101,0.12); color: #338f65; font-weight: bold;">7.93</td></tr> | |
| <tr><td style="text-align: left;">IFEval</td><td>39.19</td><td style="background: rgba(51,143,101,0.12); color: #338f65; font-weight: bold;">23.48</td></tr> | |
| </tbody> | |
| </table> | |
| The listed benchmarks use their standard task settings and few-shot counts with the chat template applied. Lythri-Wink-2B is evaluated as a Q8_0 GGUF. | |
| ## Quickstart | |
| ```python | |
| from transformers import AutoModelForCausalLM, AutoTokenizer | |
| import torch | |
| model_path = "Lythri/Lythri-Wink-2B" | |
| tok = AutoTokenizer.from_pretrained(model_path) | |
| model = AutoModelForCausalLM.from_pretrained( | |
| model_path, dtype=torch.bfloat16, device_map="auto" | |
| ) | |
| messages = [{"role": "user", "content": "My friend just lost their job and seems really down. What should I say to them?"}] | |
| chat = tok.apply_chat_template(messages, tokenize=False, add_generation_prompt=True) | |
| inputs = tok(chat, return_tensors="pt", add_special_tokens=False).to(model.device) | |
| with torch.inference_mode(): | |
| out = model.generate( | |
| **inputs, | |
| max_new_tokens=1024, | |
| do_sample=True, | |
| temperature=0.8, | |
| top_p=0.95, | |
| eos_token_id=[1, 106], | |
| ) | |
| print(tok.decode(out[0][inputs.input_ids.shape[1]:], skip_special_tokens=True)) | |
| ``` | |
| ### Recommended Generation Config | |
| ```python | |
| generation_config = { | |
| "temperature": 0.8, | |
| "top_p": 0.95, | |
| "top_k": 64, | |
| "max_new_tokens": 1024, | |
| "repetition_penalty": 1.05, | |
| "do_sample": True, | |
| "eos_token_id": [1, 106], | |
| } | |
| out = model.generate(**inputs, **generation_config) | |
| ``` | |
| ### System Prompt (Optional) | |
| Lythri-Wink-2B supports system prompts. It works well without one; to set a persona or rules for your app, add a system message: | |
| ```python | |
| messages = [ | |
| {"role": "system", "content": "You are Lythri, a warm and caring companion. Reply in the same language as the user."}, | |
| {"role": "user", "content": "I had a really rough day at work."}, | |
| ] | |
| ``` | |
| ## Compute | |
| Lythri-Wink-2B is structurally sheared from the Gemma4-E2B PT model and recovered through teacher distillation. It is then trained with CoT SFT, followed by instant-reply SFT with multi-turn companion conversations and a small writing-only online preference-optimization stage mainly judged by Gemma2-27B,and Gemma4-E4B as a language judge model for language consistency. | |
| ## Limitations | |
| - Lythri is optimized for conversation and emotional understanding, not for math, coding or complex reasoning. | |
| - Lythri is not a substitute for professional mental health support. If you or someone you know is in crisis, please contact local emergency services or a crisis helpline. | |
| - Like all language models, it can produce inaccurate or inappropriate content. | |
| ## License | |
| Lythri is built on Gemma 4 and is released under the Apache License 2.0. | |
| ## Citation | |
| ```bibtex | |
| @techreport{li2026lythri, | |
| title = {Lythri Technical Report}, | |
| author = {Li, Jiawen}, | |
| year = {2026}, | |
| institution = {Zenodo}, | |
| doi = {10.5281/zenodo.23179311}, | |
| url = {https://doi.org/10.5281/zenodo.23179311} | |
| } | |
| ``` | |
| ## Support | |
| **The whole training process is self-funded. If you like our model, please click a free like — that means a lot to me as a student!** | |