Releases: mattvr/ShellGPT
Releases · mattvr/ShellGPT
0.4.1
0.4.0
Installation
Install this version by running the following command:
deno run -rA https://deno.land/x/shellgpt/install.ts
Changelog
- Adds support for generating embeddings, use
gpt --embed "The quick brown fox jumped over the lazy dog."
- Supports
--model
and --dims` parameters.
- Supports
- [2nd attempt] Always throw errors if returned #23
Full Changelog: 0.3.9...0.4.0
0.3.9
0.3.8
0.3.6
Installation
Install this version by running the following command:
deno run -rA https://deno.land/x/shellgpt/install.ts
Changelog
- Potential fix for windows install issue
- Support sending requests to custom API URLs (which support the OpenAI schema) using
OPENAI_CHAT_URL
andOPENAI_IMG_URL
env vars.
Full Changelog: 0.3.5...0.3.6
0.3.5
Installation
Install this version by running the following command:
deno run -rA https://deno.land/x/shellgpt/install.ts
Changelog
- New: Adds
--img
flag which generates an image using OpenAI DALL-E 2 endpoint.- e.g.
gpt --img "a corgi wearing cashmere pants hacking into myspace.com"
- e.g.
- New: Adds
--lang
flag, which allows customizing the language output by--code
or other requests. - Improved: Default to GPT-4-turbo (or GPT-3.5-turbo in fast mode).
- Fix: Resolves an issue with early cutoff, which has started occurring more often recently.
- Fix: Tweaks prompts for better compliance using latest GPT models.
Full Changelog: 0.3.3...0.3.5
0.3.3
0.3.2
0.3.1
Installation
Install this version by running the following command:
deno run -rA https://deno.land/x/shellgpt/install.ts
Changelog
- Allows bypassing
wpm
buffer, for instant output.- Use
--wpm=-1
(or any number <= 0).
- Use
- Adds support for
wpm
key in config.- Manually edit
~/.gpt/config.json
and add a line like"wpm": 1234
.
- Manually edit
- Adds
--as-sys
to send a message to GPT assystem
rather thanuser
. - Adds
--bump-sys
to bump the most recent system message to the most recent position.- This generally improves adherence to the system prompt.