diff --git a/website/docs/getting-started.md b/website/docs/getting-started.md index 0362832..267fe8e 100644 --- a/website/docs/getting-started.md +++ b/website/docs/getting-started.md @@ -13,7 +13,7 @@ Tabby is an open-source, self-hosted AI coding assistant. With Tabby, every team | Section | Goal | | ------------------------------------------- | --------------------------------------------------------------------------- | -| [🔧 Installation](./installation) | Everything deployment: Docker, Homebrew, Hugging Face Space and many others | +| [📚 Installation](./installation) | Everything deployment: Docker, Homebrew, Hugging Face Space and many others | | [💻 IDE / Editor Extensions](./extensions) | IDE/Editor extensions that can be seamlessly integrated with Tabby | | [🧑‍🔬 Models Directory](./models) | A curated list of models that we recommend using with Tabby | | [🏷️ API References](/api) | API Documentation | diff --git a/website/docs/installation/docker-compose.mdx b/website/docs/installation/docker-compose.mdx index 23ef7ef..9f79d67 100644 --- a/website/docs/installation/docker-compose.mdx +++ b/website/docs/installation/docker-compose.mdx @@ -18,7 +18,7 @@ services: tabby: restart: always image: tabbyml/tabby - command: serve --model TabbyML/SantaCoder-1B + command: serve --model TabbyML/StarCoder-1B volumes: - "$HOME/.tabby:/data" ports: @@ -34,7 +34,7 @@ services: tabby: restart: always image: tabbyml/tabby - command: serve --model TabbyML/SantaCoder-1B --device cuda + command: serve --model TabbyML/StarCoder-1B --device cuda volumes: - "$HOME/.tabby:/data" ports: diff --git a/website/docs/installation/docker.mdx b/website/docs/installation/docker.mdx index ab8cfb5..b073726 100644 --- a/website/docs/installation/docker.mdx +++ b/website/docs/installation/docker.mdx @@ -15,7 +15,7 @@ import TabItem from '@theme/TabItem'; ```bash title="run.sh" docker run -it \ -p 8080:8080 -v $HOME/.tabby:/data \ - tabbyml/tabby serve --model TabbyML/SantaCoder-1B + tabbyml/tabby serve --model TabbyML/StarCoder-1B ``` @@ -25,7 +25,7 @@ import TabItem from '@theme/TabItem'; docker run -it \ --gpus all -p 8080:8080 -v $HOME/.tabby:/data \ tabbyml/tabby \ - serve --model TabbyML/SantaCoder-1B --device cuda + serve --model TabbyML/StarCoder-1B --device cuda ``` diff --git a/website/docs/models/index.md b/website/docs/models/index.md index 1d73800..de2d34d 100644 --- a/website/docs/models/index.md +++ b/website/docs/models/index.md @@ -4,43 +4,27 @@ sidebar_position: 4 # 🧑‍🔬 Models Directory +## Completion models (For `--model`) We recommend using * **small models (less than 400M)** for **CPU devices**. * For **1B to 7B models**, it's advisable to have at least **NVIDIA T4, 10 Series, or 20 Series GPUs**. * For **7B to 13B models**, we recommend using **NVIDIA V100, A100, 30 Series, or 40 Series GPUs**. -| Model ID | License | -| ---------------------------------------------------------------------- | ------------------------------------------------------------------------------------------- | -| [TabbyML/CodeLlama-13B](https://huggingface.co/TabbyML/CodeLlama-13B) | [Llama2](https://github.com/facebookresearch/llama/blob/main/LICENSE) | -| [TabbyML/CodeLlama-7B](https://huggingface.co/TabbyML/CodeLlama-7B) | [Llama2](https://github.com/facebookresearch/llama/blob/main/LICENSE) | -| [TabbyML/StarCoder-7B](https://huggingface.co/TabbyML/StarCoder-7B) | [BigCode-OpenRAIL-M](https://huggingface.co/spaces/bigcode/bigcode-model-license-agreement) | -| [TabbyML/StarCoder-3B](https://huggingface.co/TabbyML/StarCoder-3B) | [BigCode-OpenRAIL-M](https://huggingface.co/spaces/bigcode/bigcode-model-license-agreement) | -| [TabbyML/StarCoder-1B](https://huggingface.co/TabbyML/StarCoder-1B) | [BigCode-OpenRAIL-M](https://huggingface.co/spaces/bigcode/bigcode-model-license-agreement) | -| [TabbyML/SantaCoder-1B](https://huggingface.co/TabbyML/SantaCoder-1B) | [BigCode-OpenRAIL-M](https://huggingface.co/spaces/bigcode/bigcode-model-license-agreement) | -| [TabbyML/WizardCoder-3B](https://huggingface.co/TabbyML/WizardCoder-3B)| [OpenRAIL-M](https://huggingface.co/spaces/bigcode/bigcode-model-license-agreement) | -| [TabbyML/WizardCoder-1B](https://huggingface.co/TabbyML/WizardCoder-1B)| [OpenRAIL-M](https://huggingface.co/spaces/bigcode/bigcode-model-license-agreement) | -| [TabbyML/J-350M](https://huggingface.co/TabbyML/J-350M) | [BSD-3](https://opensource.org/license/bsd-3-clause/) | -| [TabbyML/T5P-220M](https://huggingface.co/TabbyML/T5P-220M) | [BSD-3](https://opensource.org/license/bsd-3-clause/) | +| Model ID | License | Infilling Support | Apple M1/M2 Supports | +| --------------------------------------------------------------------- | :-----------------------------------------------------------------------------------------: | :---------------: | :------------: | +| [TabbyML/CodeLlama-13B](https://huggingface.co/TabbyML/CodeLlama-13B) | [Llama2](https://github.com/facebookresearch/llama/blob/main/LICENSE) | ✅ | ✅ | +| [TabbyML/CodeLlama-7B](https://huggingface.co/TabbyML/CodeLlama-7B) | [Llama2](https://github.com/facebookresearch/llama/blob/main/LICENSE) | ✅ | ✅ | +| [TabbyML/StarCoder-7B](https://huggingface.co/TabbyML/StarCoder-7B) | [BigCode-OpenRAIL-M](https://huggingface.co/spaces/bigcode/bigcode-model-license-agreement) | ✅ | ✅ | +| [TabbyML/StarCoder-3B](https://huggingface.co/TabbyML/StarCoder-3B) | [BigCode-OpenRAIL-M](https://huggingface.co/spaces/bigcode/bigcode-model-license-agreement) | ✅ | ✅ | +| [TabbyML/StarCoder-1B](https://huggingface.co/TabbyML/StarCoder-1B) | [BigCode-OpenRAIL-M](https://huggingface.co/spaces/bigcode/bigcode-model-license-agreement) | ✅ | ✅ | +| [TabbyML/J-350M](https://huggingface.co/TabbyML/J-350M) | [BSD-3](https://opensource.org/license/bsd-3-clause/) | ❌ | ❌ | -### CodeLlama-7B / CodeLlama-13B +## Chat models (For `--chat-model`) -Code Llama is a collection of pretrained and fine-tuned generative text models. Theses model is designed for general code synthesis and understanding. +To ensure optimal response quality, and given that latency requirements are not stringent in this scenario, we recommend using a model with at least 3B parameters. -### StarCoder-1B / StarCoder-3B / StarCoder-7B - -StarCoder series model are trained on 80+ programming languages from The Stack (v1.2), with opt-out requests excluded. The model uses Multi Query Attention, a context window of 8192 tokens, and was trained using the Fill-in-the-Middle objective on 1 trillion tokens. - -### WizardCoder-1B / WizardCoder-3B - -WizardCoder [(arXiv)](https://arxiv.org/abs/2306.08568) series model are finetuned on StarCoder models with the Evol-Instruct method to adapt to coding tasks. Note that WizardCoder models have used GPT-4 generated data for finetuning, and thus adhere to [OpenAI's limitations](https://openai.com/policies/terms-of-use) for model usage. - -### SantaCoder-1B - -SantaCoder is the smallest member of the BigCode family of models, boasting just 1.1 billion parameters. This model is specifically trained with a fill-in-the-middle objective, enabling it to efficiently auto-complete function parameters. It offers support for three programming languages: Python, Java, and JavaScript. - -### J-350M - -Derived from [Salesforce/codegen-350M-multi](https://huggingface.co/Salesforce/codegen-350M-multi), a model of CodeGen family. - -### T5P-220M -Derived from [Salesforce/codet5p-220m](https://huggingface.co/Salesforce/codet5p-220m), a model of CodeT5+ family. +| Model ID | License | +| ------------------------------------------------------------------------- | :---------------------------------------------------------------------------------: | +| [TabbyML/WizardCoder-15B](https://huggingface.co/TabbyML/WizardCoder-15B) | [OpenRAIL-M](https://huggingface.co/spaces/bigcode/bigcode-model-license-agreement) | +| [TabbyML/Mistral-7B](https://huggingface.co/TabbyML/Mistral-7B) | [Apache 2.0](https://opensource.org/licenses/Apache-2.0) | +| [TabbyML/WizardCoder-3B](https://huggingface.co/TabbyML/WizardCoder-3B) | [OpenRAIL-M](https://huggingface.co/spaces/bigcode/bigcode-model-license-agreement) | \ No newline at end of file