root
/
tabby
mirror of https://gitee.com/zhang_1334717033/tabby.git
1
0
Fork 0
Code Issues Packages Projects Releases Wiki Activity

docs: update documentation to prepare for 0.2 release (#502) 

* docs: fix installation emoji

* docs: set StarCoder-1B to be default model for docker install

* docs: add `--chat-model` in model directory
r0.2
Meng Zhang 2023-10-04 04:11:07 +08:00 committed by GitHub
parent 0e5128e8fb
commit 6ad99cb23d
No known key found for this signature in database
GPG Key ID: 4AEE18F83AFDEB23
4 changed files with 21 additions and 37 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

2
website/docs/getting-started.md
View File

@ -13,7 +13,7 @@ Tabby is an open-source, self-hosted AI coding assistant. With Tabby, every team
| Section | Goal | | 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 | | [💻 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 | | [🧑‍🔬 Models Directory](./models) | A curated list of models that we recommend using with Tabby |
| [🏷️ API References](/api) | API Documentation | | [🏷️ API References](/api) | API Documentation |

4
website/docs/installation/docker-compose.mdx
View File

@ -18,7 +18,7 @@ services:
tabby: tabby:
restart: always restart: always
image: tabbyml/tabby image: tabbyml/tabby
command: serve --model TabbyML/SantaCoder-1B command: serve --model TabbyML/StarCoder-1B
volumes: volumes:
- "$HOME/.tabby:/data" - "$HOME/.tabby:/data"
ports: ports:
@ -34,7 +34,7 @@ services:
tabby: tabby:
restart: always restart: always
image: tabbyml/tabby image: tabbyml/tabby
command: serve --model TabbyML/SantaCoder-1B --device cuda command: serve --model TabbyML/StarCoder-1B --device cuda
volumes: volumes:
- "$HOME/.tabby:/data" - "$HOME/.tabby:/data"
ports: ports:

4
website/docs/installation/docker.mdx
View File

@ -15,7 +15,7 @@ import TabItem from '@theme/TabItem';
```bash title="run.sh" ```bash title="run.sh"
docker run -it \ docker run -it \
-p 8080:8080 -v $HOME/.tabby:/data \ -p 8080:8080 -v $HOME/.tabby:/data \
tabbyml/tabby serve --model TabbyML/SantaCoder-1B tabbyml/tabby serve --model TabbyML/StarCoder-1B
``` ```
</TabItem> </TabItem>
@ -25,7 +25,7 @@ import TabItem from '@theme/TabItem';
docker run -it \ docker run -it \
--gpus all -p 8080:8080 -v $HOME/.tabby:/data \ --gpus all -p 8080:8080 -v $HOME/.tabby:/data \
tabbyml/tabby \ tabbyml/tabby \
serve --model TabbyML/SantaCoder-1B --device cuda serve --model TabbyML/StarCoder-1B --device cuda
``` ```
</TabItem> </TabItem>

50
website/docs/models/index.md
View File

@ -4,43 +4,27 @@ sidebar_position: 4
# 🧑‍🔬 Models Directory # 🧑‍🔬 Models Directory
## Completion models (For `--model`)
We recommend using We recommend using
* **small models (less than 400M)** for **CPU devices**. * **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 **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**. * For **7B to 13B models**, we recommend using **NVIDIA V100, A100, 30 Series, or 40 Series GPUs**.
| 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/) | ❌ | ❌ |
## Chat models (For `--chat-model`)
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.
| Model ID | License | | Model ID | License |
| ---------------------------------------------------------------------- | ------------------------------------------------------------------------------------------- | | ------------------------------------------------------------------------- | :---------------------------------------------------------------------------------: |
| [TabbyML/CodeLlama-13B](https://huggingface.co/TabbyML/CodeLlama-13B) | [Llama2](https://github.com/facebookresearch/llama/blob/main/LICENSE) | | [TabbyML/WizardCoder-15B](https://huggingface.co/TabbyML/WizardCoder-15B) | [OpenRAIL-M](https://huggingface.co/spaces/bigcode/bigcode-model-license-agreement) |
| [TabbyML/CodeLlama-7B](https://huggingface.co/TabbyML/CodeLlama-7B) | [Llama2](https://github.com/facebookresearch/llama/blob/main/LICENSE) | | [TabbyML/Mistral-7B](https://huggingface.co/TabbyML/Mistral-7B) | [Apache 2.0](https://opensource.org/licenses/Apache-2.0) |
| [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-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/) |
### CodeLlama-7B / CodeLlama-13B <span title="Apple GPU Support"></span>
Code Llama is a collection of pretrained and fine-tuned generative text models. Theses model is designed for general code synthesis and understanding.
### StarCoder-1B / StarCoder-3B / StarCoder-7B <span title="Apple GPU Support"></span>
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 <span title="Apple GPU Support"></span>
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.