Cygnis v0.2 has been rigorously tested against internal benchmarks to ensure production-grade stability. It shows significant improvements in Instruction Following for complex, multi-step tasks compared to the v0.1 release.
1import transformers
2import torch
3
4model_id = "cygnisai/Cygnis-Alpha-2-8B-v0.2"
5
6pipeline = transformers.pipeline(
7 "text-generation",
8 model=model_id,
9 model_kwargs={"torch_dtype": torch.bfloat16},
10 device_map="auto"
11)
12
13messages = [
14 {"role": "system", "content": "You are Cygnis, the official enterprise AI assistant."},
15 {"role": "user", "content": "Analyze the scalability requirements for our LLM infrastructure."}
16]
17
18print(pipeline(messages, max_new_tokens=256))
1<|begin_of_text|><|start_header_id|>system<|end_header_id|>
2
3{system_prompt}<|eot_id|><|start_header_id|>user<|end_header_id|>
4
5{user_prompt}<|eot_id|><|start_header_id|>assistant<|end_header_id|>
1@misc{cygnis_alpha_2_v0.1,
2 author = {Simonc-44},
3 title = {The Carbon Footprint of Machine Learning Training Will Plateau, Then Shrink},
4 year = {2026},
5 publisher = {Hugging Face},
6 url = {https://huggingface.co/Simonc-44/Cygnis-Alpha-2-8B-v0.1}
7 eprint={2204.05149},
8 archivePrefix={arXiv},
9}
