SQLCoder Models by Defog.ai
About
SQLCoder is a family of large language models developed by Defog.ai, tailored for transforming natural language queries into SQL commands. These models stand out for their capability to enable non-technical users to easily interact with databases by converting human-readable questions into accurate SQL queries. Fine-tuned on vast datasets of human-curated questions and SQL queries, SQLCoder achieves state-of-the-art performance across various benchmarks. The family includes various versions such as the 7B parameter model, optimized for different hardware from high-end GPUs to consumer-grade devices. Designed specifically for read-only database interactions, these models are not suitable for tasks involving write access. Defog.ai is committed to continuously enhancing SQLCoder, refining its features, and addressing limitations based on user feedback 10 5 13.
Current Variants
Use-when guidance is derived from seed capabilities, context, release, and replacement fields.
Use when the workload needs 8k context and 8B parameters.
Use when the workload needs 32k context and 7B parameters.
Use when the workload needs 16k context and 70B parameters.
Use when the workload needs 100k context and 34B parameters.
Use when the workload needs 100k context and 7B parameters.
Use when the workload needs 100k context and 15B parameters.
Use when the workload needs 8k context and 15B parameters.
| Model | Use when | Released | Signals | Status |
|---|---|---|---|---|
| Llama 3 SQLCoder 8B | Use when the workload needs 8k context and 8B parameters. | 2024-05 | 8k context8B parameters | Current |
| SQLCoder 7B 2 | Use when the workload needs 32k context and 7B parameters. | 2024-02 | 32k context7B parameters | Current |
| SQLCoder 70B Alpha | Use when the workload needs 16k context and 70B parameters. | 2024-01 | 16k context70B parameters | Current |
| SQLCoder 34B Alpha | Use when the workload needs 100k context and 34B parameters. | 2023-11 | 100k context34B parameters | Current |
| SQLCoder 7B | Use when the workload needs 100k context and 7B parameters. | 2023-10 | 100k context7B parameters | Current |
| SQLCoder 2 | Use when the workload needs 100k context and 15B parameters. | 2023-10 | 100k context15B parameters | Current |
| SQLCoder | Use when the workload needs 8k context and 15B parameters. | 2023-08 | 8k context15B parameters | Current |
Release Timeline
6 release groupsSpecifications(7 models)
| Model | Released | Context | Parameters |
|---|---|---|---|
| Llama 3 SQLCoder 8B | 2024-05 | 8k | 8B |
| SQLCoder 7B 2 | 2024-02 | 32k | 7B |
| SQLCoder 70B Alpha | 2024-01 | 16k | 70B |
| SQLCoder 34B Alpha | 2023-11 | 100k | 34B |
| SQLCoder 7B | 2023-10 | 100k | 7B |
| SQLCoder 2 | 2023-10 | 100k | 15B |
| SQLCoder | 2023-08 | 8k | 15B |
Available From(1 provider)
Frequently Asked Questions
- What is SQLCoder used for?
- SQLCoder is used for coding. The family description and listed model capabilities point to those workloads as the best fit.
- How does SQLCoder compare to StarCoder?
- SQLCoder by Defog.ai is strongest where you need coding, while StarCoder by ServiceNow Research is the closest related family to check for coding. SQLCoder has 7 listed variants and reaches up to 100k context, while StarCoder reaches up to 8k context, so compare the specs and pricing tables before choosing a production model.
- Which SQLCoder model should I use?
- If price is the main constraint, use the pricing table first because SQLCoder does not have complete provider pricing in the local data. For the most capable/latest local choice, evaluate SQLCoder 34B Alpha with 100k context.
