Update raycast-ollama extension

- Merge pull request raycast#9 from MassimilianoPasquini97/dev
- Updated README.md and CHANGELOG.md files.
- Load PDF and Text based files on prompt for query chains.
- Code splitted
- Merge pull request raycast#6 from MassimilianoPasquini97/OllamaResponseApiFix
- Updated changelog.md
- TextField apper if ModelsOnRegistry is undefined
- Fixed ModelsOnRegistry.lengh undefined
- Deleted \'error\' message on event emitter
- Deleted field no longher used by Ollama Generate Response API
- CHANGELOG.md update
- Merge pull request raycast#5 from MassimilianoPasquini97/clipboard_fallback
- New Preference \'Enable Input Source Fallback\'.
- New Preference \'Input Source\'
- Removed unused embedding from codebase
- Merge branch \'contributions/merge-1697691587951609000\'
- Pull contributions
- Updated CHANGELOG.md
- Updated README.md with new model name.
- Metrics metadata now available on Chat Command.
- Ollama Host is now configurable throw Preferences.
- New Action.Open for quickly go to \'Manage Models\'.
- Moved Model preferences to LocalStorage.
- Updated Models Library link on README.md
- Reduced re-rendering on models downloading.
- Deleted navigationTitle from Form.
- Last fixes before publish.
- Improvement on \'Chat With Ollama\' ActionPanel
- New command \'Manage Models\'
- OllamaApiTags() function returns Promise<OllamaApiTagsResponse>
- Error Handling for \'ollama-custom-create\'
- Implemented new command \'Create Custom Command\'
- Multiple chat saving feature
- Convertation is now saved only when inference is done.
- Chat is now saved on LocalStorage
- First implementation of a chat command.
- [Improvement and BugFix] - 2023-08-12
- Merge pull request raycast#2 from suhaildawood/main
- Changed CHANGELOG.md file.
- feat: support for llama2:70b
- Import optimized images
- Pull contributions
- CHANGELOG Update
- Updated README and minor fix
- Fixed CHANGELOG and README
- [Improvement] - 2023-07-31
- Improvement] - 2023-07-30 v2
- [Improvement] - 2023-07-30
- [Code Improvement and BugFix] - 2023-07-29
- ray lint --fix runned
- Added git repository
- Initial commit
- Initial commit

extensions/raycast-ollama/CHANGELOG.md

@@ -1,5 +1,10 @@
 # raycast-ollama Changelog
 
+## [Improvement] - 2023-11-30
+
+- [Improvement] Query you pdf or text based file with Ollama. More information on how to use is on README.md.
+- [Improvement] On Command 'Manage Models' is now possible to view all Modelfile parameters. If a specific parameter isn't set on Modelfile it display the default value.
+
 ## [BugFix] - 2023-11-5
 
 - [BugFix] Fixed error `ModelsOnRegistry.lengh undefined`.

extensions/raycast-ollama/README.md

@@ -1,21 +1,25 @@
-# Raycast Ollama
+<p align="center">
+    <img src="assets/icon.png" height="128">
+    <h1 align="center">Raycast Ollama</h1>
+</p>
 
 Use [Ollama](https://ollama.ai) for local llama inference on Raycast.
 
 ## Requirements
 
-1. Ollama installed and running.
-2. At least one model installed. Use 'Manage Models' commands for pulling images or ollama cli.
+[Ollama](https://ollama.ai) installed and running on your mac. At least one model need to be installed throw Ollama cli tools or with 'Manage Models' Command. You can find all available model [here](https://ollama.ai/library).
 
-```bash
-ollama pull orca-mini
-ollama pull llama2
-```
+## How to Use
 
-## Use a different model
+### Command: Chat With Ollama
 
-This plugin allows you to select a different model for each command. Keep in mind that you need to have the corresponding model installed on your machine. You can find all available model [here](https://ollama.ai/library).
+Chat with your preferred model from Raycast, with the following features:
 
-## Create your own custom commands
+- Save conversation with `CMD+S` keyboard shortcut. You can access your saved conversation with `CMD+P` keyboard shortcut.
+- Change model with `CMD+M` keyboard shortcuts. For embedding is recommended to use a lower parameters model for better performance.
+- Copy your Question, Answer or even the entire Chat to the clipboard.
+- Is now possible to ask the model about one or more files. Select the files using `CMD+F`, at this stage only text based files and PDF are supported. And finally use tag `/file` for query about selected files. By default it use 'Stuff' Chain, you can change Chain type from 'Document Loader' submenu. This feature is currently experimental.
 
-With '***Create Custom Command***' you can create your own custom command or chatbot using whatever model you want.
+### Command: Create Custom Commands
+
+With '***Create Custom Command***' you can create your own custom Command using whatever model you want.