matdmiller commited on
Commit
d98b634
β€’
1 Parent(s): 636fd12

update readme with instructions

Browse files
Files changed (2) hide show
  1. README.md +44 -2
  2. app.ipynb +1 -1
README.md CHANGED
@@ -1,5 +1,5 @@
1
  ---
2
- title: Tts Openai
3
  emoji: πŸ”₯
4
  colorFrom: green
5
  colorTo: red
@@ -13,5 +13,47 @@ hf_oauth: true
13
  # optional, default duration is 8 hours/480 minutes. Max duration is 30 days/43200 minutes.
14
  hf_oauth_expiration_minutes: 480
15
  ---
 
 
16
 
17
- Check out the configuration reference at https://huggingface.co/docs/hub/spaces-config-reference
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
  ---
2
+ title: TTS
3
  emoji: πŸ”₯
4
  colorFrom: green
5
  colorTo: red
 
13
  # optional, default duration is 8 hours/480 minutes. Max duration is 30 days/43200 minutes.
14
  hf_oauth_expiration_minutes: 480
15
  ---
16
+ # Getting Started
17
+ > This spaces allows you to create speech audio from text. It currently works with OpenAI and Cartesia AI.
18
 
19
+ ## Running Locally
20
+ ```bash
21
+ git clone https://huggingface.co/spaces/matdmiller/tts-openai
22
+ cd tts-openai
23
+ pip install -r requirements.txt
24
+ ```
25
+ Next create a file `tts_openai_secrets.py` to hold the secrets or set them up as environment variables yourself.
26
+ ```python
27
+ import os
28
+ os.environ['OPENAI_API_KEY'] = 'sk-xxx'
29
+ os.environ["CARTESIA_API_KEY"]= 'xxx'
30
+ os.environ["ALLOWED_OAUTH_PROFILE_USERNAMES"]= '<your huggingface username>' #comma delimited
31
+ os.environ["REQUIRE_AUTH"]= 'false'
32
+ ```
33
+ This file is included in the `.gitignore` so it will not be included when you publish to your repo. This is very important because the πŸ€— Spaces repos are public by default and you don't want to leak your API keys.
34
+
35
+ Finally to run locally simply run:
36
+ ```bash
37
+ python app.py
38
+ ```
39
+
40
+ ## Cloning to run in your own πŸ€— Space.
41
+
42
+ Click the `...` in the top right hand corder of my Space and click `Duplicate Space`.
43
+
44
+ ![A photo showing how to duplicate a Hugging Face Space.](./hf_duplicate_space.png "Duplicate Space")
45
+
46
+ This will bring up a form that you need to fill out with the appropriate keys and settings.
47
+
48
+ ![A photo showing the form for duplicating a Hugging Face Space.](./hf_duplicate_space_form.png "Duplicate Space Form")
49
+
50
+ - Owner: Your HF Username
51
+ - Visibility: Public (Or Private if you know what you're doing). By default this app is secured with πŸ€— OAuth and only allows usernames you specify in the `ALLOWED_OAUTH_PROFILE_USERNAMES` property.
52
+ - Space Hardware: Free CPU Basic is plenty of horsepower for this app becuase all of the TTS work is happening behind the API.
53
+ - `OPENAI_API_KEY`: Input your own OpenAI API key which you can get by signing up for an account on [https://platform.openai.com](https://platform.openai.com). This is completely separate from ChatGPT.
54
+ - `CARTESIA_API_KEY`: Input your own Cartesia AI API key which you can obtain from [https://play.cartesia.ai/subscription](https://play.cartesia.ai/subscription)
55
+ - `ALLOWED_OAUTH_PROFILE_USERNAMES`: Add a comma separated list of Hugging Face usernames that are allowed to use the space. Only usernames listed here will be allowed generate speech audio, though anyone can view the site. Ex: `santa1225,clem2024`. If you have a single username, don't add a comma.
56
+
57
+ Click `Duplicate Space`. Your new space should be created within a couple of minutes.
58
+
59
+ Additional HF Spaced Config Info: https://huggingface.co/docs/hub/spaces-config-reference
app.ipynb CHANGED
@@ -979,7 +979,7 @@
979
  },
980
  {
981
  "cell_type": "code",
982
- "execution_count": 34,
983
  "id": "0420310d-930b-4904-8bd4-3458ad8bdbd3",
984
  "metadata": {},
985
  "outputs": [],
 
979
  },
980
  {
981
  "cell_type": "code",
982
+ "execution_count": 35,
983
  "id": "0420310d-930b-4904-8bd4-3458ad8bdbd3",
984
  "metadata": {},
985
  "outputs": [],