:source: fortios_llm_profile.py
:orphan:
.. fortios_llm_profile:
fortios_llm_profile -- Configure LLM Proxy profiles in Fortinet's FortiOS and FortiGate.
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
.. versionadded:: 2.0.0
.. contents::
:local:
:depth: 1
Synopsis
--------
- This module is able to configure a FortiGate or FortiOS (FOS) device by allowing the user to set and modify llm feature and profile category. Examples include all parameters and values need to be adjusted to datasources before usage. Tested with FOS v6.0.0
Requirements
------------
The below requirements are needed on the host that executes this module.
- ansible>=2.16
Tips
----
Using member operation to add an element to an existing object.
FortiOS Version Compatibility
-----------------------------
Supported Version Ranges: v8.0.0
Parameters
----------
.. raw:: html
- access_token - Token-based authentication. Generated from GUI of Fortigate. type: str required: false
- enable_log - Enable/Disable logging for task. type: bool required: false default: False
- vdom - Virtual domain, among those defined previously. A vdom is a virtual instance of the FortiGate that can be configured and used as a different unit. type: str default: root
- member_path - Member attribute path to operate on. type: str
- member_state - Add or delete a member under specified attribute path. type: str choices: present, absent
- state - Indicates whether to create or remove the object. type: str required: true choices: present, absent
- llm_profile - Configure LLM Proxy profiles. type: dict
more...
| Supported Version Ranges |
| llm_profile |
v8.0.0 -> 8.0.0 |
- chat - LLM Proxy chat completions API (/v1/chat/completions). type: dict
more...
| Supported Version Ranges |
| chat |
v8.0.0 -> 8.0.0 |
- max_completion_tokens - Maximum number of completion tokens (0 - 10000000). type: int
more...
|
Supported Version Ranges |
| max_completion_tokens |
v8.0.0 -> 8.0.0 |
- max_req_len - Max size of chat completions request body, (0 - 65535 KiB). type: int
more...
|
Supported Version Ranges |
| max_req_len |
v8.0.0 -> 8.0.0 |
- status - Support chat completions API. type: str choices: enable, disable
more...
|
Supported Version Ranges |
| status |
v8.0.0 -> 8.0.0 |
| [enable] |
v8.0.0 -> 8.0.0 |
| [disable] |
v8.0.0 -> 8.0.0 |
- stream - support chat completions stream mode. type: str choices: bypass, block
more...
|
Supported Version Ranges |
| stream |
v8.0.0 -> 8.0.0 |
| [bypass] |
v8.0.0 -> 8.0.0 |
| [block] |
v8.0.0 -> 8.0.0 |
- system_prompt - Replace/Append chat completions system prompt. type: str
more...
|
Supported Version Ranges |
| system_prompt |
v8.0.0 -> 8.0.0 |
- system_prompt_mode - System prompt processing mode. type: str choices: bypass, replace, prepend, append
more...
|
Supported Version Ranges |
| system_prompt_mode |
v8.0.0 -> 8.0.0 |
| [bypass] |
v8.0.0 -> 8.0.0 |
| [replace] |
v8.0.0 -> 8.0.0 |
| [prepend] |
v8.0.0 -> 8.0.0 |
| [append] |
v8.0.0 -> 8.0.0 |
- image_gen - LLM Proxy image generation API (/v1/images/generations). type: dict
more...
| Supported Version Ranges |
| image_gen |
v8.0.0 -> 8.0.0 |
- status - Support image generation API. type: str choices: enable, disable
more...
|
Supported Version Ranges |
| status |
v8.0.0 -> 8.0.0 |
| [enable] |
v8.0.0 -> 8.0.0 |
| [disable] |
v8.0.0 -> 8.0.0 |
- list_models - LLM Proxy list models API (/v1/models). type: dict
more...
| Supported Version Ranges |
| list_models |
v8.0.0 -> 8.0.0 |
- status - Support list models API. type: str choices: enable, disable
more...
|
Supported Version Ranges |
| status |
v8.0.0 -> 8.0.0 |
| [enable] |
v8.0.0 -> 8.0.0 |
| [disable] |
v8.0.0 -> 8.0.0 |
- log - Log option. type: str choices: none, blocked, all
more...
|
Supported Version Ranges |
| log |
v8.0.0 -> 8.0.0 |
| [none] |
v8.0.0 -> 8.0.0 |
| [blocked] |
v8.0.0 -> 8.0.0 |
| [all] |
v8.0.0 -> 8.0.0 |
- name - LLM Proxy profile name. type: str required: true
more...
|
Supported Version Ranges |
| name |
v8.0.0 -> 8.0.0 |
- unknown_api - Support unknown API. type: str choices: enable, disable
more...
|
Supported Version Ranges |
| unknown_api |
v8.0.0 -> 8.0.0 |
| [enable] |
v8.0.0 -> 8.0.0 |
| [disable] |
v8.0.0 -> 8.0.0 |
Notes
-----
.. note::
- Legacy fortiosapi has been deprecated, httpapi is the preferred way to run playbooks
Examples
--------
.. code-block:: yaml+jinja
- name: Configure LLM Proxy profiles.
fortinet.fortios.fortios_llm_profile:
vdom: "{{ vdom }}"
state: "present"
access_token: ""
llm_profile:
chat:
max_completion_tokens: "0"
max_req_len: "1024"
status: "enable"
stream: "bypass"
system_prompt: ""
system_prompt_mode: "bypass"
image_gen:
status: "enable"
list_models:
status: "enable"
log: "none"
name: "default_name_15"
unknown_api: "enable"
Return Values
-------------
Common return values are documented: https://docs.ansible.com/ansible/latest/reference_appendices/common_return_values.html#common-return-values, the following are the fields unique to this module:
.. raw:: html
- build - Build number of the fortigate image returned: always type: str sample: 1547
- http_method - Last method used to provision the content into FortiGate returned: always type: str sample: PUT
- http_status - Last result given by FortiGate on last operation applied returned: always type: str sample: 200
- mkey - Master key (id) used in the last call to FortiGate returned: success type: str sample: id
- name - Name of the table used to fulfill the request returned: always type: str sample: urlfilter
- path - Path of the table used to fulfill the request returned: always type: str sample: webfilter
- revision - Internal revision number returned: always type: str sample: 17.0.2.10658
- serial - Serial number of the unit returned: always type: str sample: FGVMEVYYQT3AB5352
- status - Indication of the operation's result returned: always type: str sample: success
- vdom - Virtual domain used returned: always type: str sample: root
- version - Version of the FortiGate returned: always type: str sample: v5.6.3
Status
------
- This module is not guaranteed to have a backwards compatible interface.
Authors
-------
- Link Zheng (@chillancezen)
- Jie Xue (@JieX19)
- Hongbin Lu (@fgtdev-hblu)
- Frank Shen (@frankshen01)
- Miguel Angel Munoz (@mamunozgonzalez)
- Nicolas Thomas (@thomnico)
.. hint::
If you notice any issues in this documentation, you can create a pull request to improve it.