Supported Models in Refact
Cloud Version of Refact
Completion models
- Refact/1.6B
Chat models
- GPT 3.5
- GPT 4 (Pro plan)
Self-Hosted Version of Refact
In Refact self-hosted you can select between the following models:
Completion models
| Model Name | Fine-tuning support |
|---|---|
| Refact/1.6B | ✓ |
| Refact/1.6B/vllm | ✓ |
| starcoder/1b/base | ✓ |
| starcoder/1b/vllm | |
| starcoder/3b/base | ✓ |
| starcoder/3b/vllm | |
| starcoder/7b/base | ✓ |
| starcoder/7b/vllm | |
| starcoder/15b/base | |
| starcoder/15b/plus | |
| starcoder2/3b/base | ✓ |
| starcoder2/3b/vllm | ✓ |
| starcoder2/7b/base | ✓ |
| starcoder2/7b/vllm | ✓ |
| starcoder2/15b/base | ✓ |
| deepseek-coder/1.3b/base | ✓ |
| deepseek-coder/1.3b/vllm | ✓ |
| deepseek-coder/5.7b/mqa-base | ✓ |
| deepseek-coder/5.7b/vllm | ✓ |
| codellama/7b | ✓ |
| stable/3b/code | |
| wizardcoder/15b |
Chat models
| Model Name |
|---|
| starchat/15b/beta |
| deepseek-coder/33b/instruct |
| deepseek-coder/6.7b/instruct |
| deepseek-coder/6.7b/instruct-finetune |
| deepseek-coder/6.7b/instruct-finetune/vllm |
| wizardlm/7b |
| wizardlm/13b |
| wizardlm/30b |
| llama2/7b |
| llama2/13b |
| magicoder/6.7b |
| mistral/7b/instruct-v0.1 |
| mixtral/8x7b/instruct-v0.1 |
| llama3/8b/instruct |
| llama3/8b/instruct/vllm |
For an up-to-date list of models, see the Known Models file on GitHub.
Integrations
Refact.ai offers OpenAI and Anthropic API integrations.
To enable these integrations, navigate to the Model Hosting page activate the OpenAI and/or Anthropic integrations by pressing the switch button in the 3rd Party APIs section.
Press API Keys tab link, you will be redirected to the integrations page. Alternatively, you can access the integrations page by clicking on the Settings dropdown menu in the header and selecting Credentials.
In the Credentials page, you can specify your OpenAI and/or Anthropic API keys.