Smaug 2 Models by Abacus.AI
Details
About
The Smaug 2 family of large language models (LLMs) signifies a notable leap in open-source AI innovation, stemming from the Qwen1.5-72B-Chat model with additional fine-tuning for enhanced logic and coding capabilities 3. Among its distinguished members, Smaug-2-72B stands out by surpassing the Qwen1.5-72B-Chat in various assessments, such as MT-Bench and HumanEval 3. The specifics surrounding the training data and methodologies for Smaug 2 remain sparse, yet the models have shown a clear progression in reasoning and coding over the base model 3. Additional insights into the evolution and functionalities of the Smaug 2 family are anticipated to be unveiled in due course.
Current Variants
Use-when guidance is derived from seed capabilities, context, release, and replacement fields.
Use when the workload needs 32k context and 72B parameters.
| Model | Use when | Released | Signals | Status |
|---|---|---|---|---|
| Smaug 2 72B | Use when the workload needs 32k context and 72B parameters. | 2024-05 | 32k context72B parameters | Current |
Release Timeline
1 release groupSpecifications(1 models)
| Model | Released | Context | Parameters |
|---|---|---|---|
| Smaug 2 72B | 2024-05 | 32k | 72B |
Frequently Asked Questions
- What is Smaug 2 used for?
- Smaug 2 is used for coding. The family description and listed model capabilities point to those workloads as the best fit.
- How does Smaug 2 compare to Smaug?
- Smaug 2 by Abacus.AI is strongest where you need coding, while Smaug by Abacus.AI is the closest related family to check for coding. Smaug 2 has 1 listed variant and reaches up to 32k context, while Smaug reaches up to 200k context, so compare the specs and pricing tables before choosing a production model.
- Which Smaug 2 model should I use?
- If price is the main constraint, use the pricing table first because Smaug 2 does not have complete provider pricing in the local data. For the most capable/latest local choice, evaluate Smaug 2 72B with 32k context.
