Using OLMo 7B Twin-2T on Together AI
Implementation guide · OLMo · Allen Institute for Artificial Intelligence (AI2)
Quick Start
- 1
- 2Use the Together AI SDK or REST API to call
olmo-7b-twin-2t— see the documentation for request format. - 3
Code Examples
pip install togetherTOGETHER_API_KEYolmo-7b-twin-2tTogether uses "organization/model-name" format, e.g. "meta-llama/Llama-4-Scout-17B-16E-Instruct" or "Qwen/QwQ-32B". See the Together model catalog for the exact ID.
from together import Together
client = Together() # reads TOGETHER_API_KEY from env
response = client.chat.completions.create(
model="olmo-7b-twin-2t",
messages=[{"role": "user", "content": "Hello"}]
)
print(response.choices[0].message.content)About Together AI
Platform for running open-source and proprietary LLMs
Together AI is a platform for running open-source and proprietary LLMs with fast serverless and dedicated endpoints at competitive inference pricing.
Pricing on Together AI
| Type | Price (per 1M) |
|---|---|
| Input tokens | $0.20 |
| Output tokens | $0.20 |
Capabilities
About OLMo 7B Twin-2T
The OLMo 7B Twin-2T is a robust open-source large language model that implements a decoder-only transformer architecture with enhancements for greater stability and performance. It features non-parametric layer normalization and SwiGLU activation functions, along with Rotary positional embeddings for better sequence handling. The model, comprising 32 layers and 32 attention heads, was trained on approximately 2 trillion tokens and supports a context length of 2048. It is notable for its transparency in AI research, as all training data, code, and evaluations are publicly accessible, promoting collaborative advancements. The model excels in various NLP tasks and has options for fine-tuning, while its developers advocate for responsible AI usage to mitigate risks of bias and inaccuracies.