[Feature]: Integrate DeepSeek-V3 as an AI provider

[dhuettner]

Description

I propose integrating DeepSeek-V3 as a new AI provider in OpenCommit. DeepSeek-V3 is a powerful AI model that offers high performance and speed, making it a great alternative to existing providers like OpenAI, Anthropic, or Ollama.

This integration would allow users to use DeepSeek-V3 for generating commit messages, providing more options and flexibility in choosing AI models.

Suggested Solution

1. Add DeepSeek-V3 as a supported AI provider in the OpenCommit configuration.
2. Allow users to set DeepSeek-V3 as the default provider via the OCO_AI_PROVIDER config.
3. Support DeepSeek-V3's API endpoint and model options in the configuration.

Example configuration:

oco config set OCO_AI_PROVIDER=deepseek-v3
oco config set OCO_API_KEY=<DeepSeek_API_Key>
oco config set OCO_API_URL=<DeepSeek_API_URL>
oco config set OCO_MODEL=deepseek-v3

### Alternatives

_No response_

### Additional Context

_No response_

[di-sukharev]

do you want to make a PR? this is quite a straightforward feature to add if you look into code

[satoru2727]

DeepSeek has implemented a Context Caching feature (see: https://api-docs.deepseek.com/news/news0802). This should help reduce API usage costs. Therefore, I support adding this feature.

[di-sukharev]

@satoru2727 nice, do you want to make a PR then?