Text Generation
MLX
Safetensors
English
small-language-model
from-scratch
conversational
apple-silicon
research
Instructions to use OpenSML/Solstice-1 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- MLX
How to use OpenSML/Solstice-1 with MLX:
# Make sure mlx-lm is installed # pip install --upgrade mlx-lm # Generate text with mlx-lm from mlx_lm import load, generate model, tokenizer = load("OpenSML/Solstice-1") prompt = "Write a story about Einstein" messages = [{"role": "user", "content": prompt}] prompt = tokenizer.apply_chat_template( messages, add_generation_prompt=True ) text = generate(model, tokenizer, prompt=prompt, verbose=True) - Notebooks
- Google Colab
- Kaggle
- Local Apps Settings
- LM Studio
- MLX LM
How to use OpenSML/Solstice-1 with MLX LM:
Generate or start a chat session
# Install MLX LM uv tool install mlx-lm # Interactive chat REPL mlx_lm.chat --model "OpenSML/Solstice-1"
Run an OpenAI-compatible server
# Install MLX LM uv tool install mlx-lm # Start the server mlx_lm.server --model "OpenSML/Solstice-1" # Calling the OpenAI-compatible server with curl curl -X POST "http://localhost:8000/v1/chat/completions" \ -H "Content-Type: application/json" \ --data '{ "model": "OpenSML/Solstice-1", "messages": [ {"role": "user", "content": "Hello"} ] }' - Atomic Chat
Commit History
4 days ago
-
Add sourced SmolLM2 scores and highlight benchmark leaders
Verified signatureUse the independently published SmolLM2 result set consistently, bold the highest reported metric values, and remove redundant evaluation graphics. Model weights and inference runtime are unchanged.632e406 -
Expand model card with sourced comparisons and training figures
Verified signatureCompare published OpenSML-150M and SmolLM2 results, include IFEval, simplify metric tables, and add training details, recorded examples and reproducible figures. Model weights and inference runtime are unchanged.04436f5 -
Prepare Solstice-1 175M private release package
Verified signatureMerged FP32 checkpoint, frozen tokenizer, native MLX runner, Apache-2.0 license, technical documentation, training provenance and benchmark evidence.7bcd331