Update generation.ts to fix TOGETHER/LLAMACLOUD image generation

[prophet10x]

Relates to:

Risks

Background

For use of LLAMACLOUD/TOGETHER via imageModelProvider: ModelProviderName.TOGETHER the API didn't appear to be correctly implemented.

What does this PR do?

Fixes the implementation of the together api in packages/core/src/generation.ts

What kind of change is this?

Bug fixes (non-breaking change which fixes an issue)

together API call was not fetching the generated image and storing it correctly

Documentation changes needed?

none

Testing

i added " imageModelProvider: ModelProviderName.TOGETHER," to defaultcharacter.ts
added a together API key to the .env file along with openai key

set the following
DISCORD_APPLICATION_ID=
DISCORD_API_TOKEN=
IMAGE_GEN=TRUE
USE_OPENAI_EMBEDDING=TRUE
USE_OLLAMA_EMBEDDING=TRUE

Where should a reviewer start?

1. .env file keys
2. add " imageModelProvider: ModelProviderName.TOGETHER," to defaultcharacter.ts
3. set up discord bot stuff (see online instructions for this) and add to chat room
4. build & start
5. test discord agent to gen an image

Detailed testing steps

1. .env file keys
2. add " imageModelProvider: ModelProviderName.TOGETHER," to defaultcharacter.ts
3. set up discord bot stuff (see online instructions for this) and add to chat room
4. build & start
5. test discord agent to gen an image

then proceeded to go to discord and asked the agent to generate an image of a cat girl

which resulted in the following in the console log

◎ LOGS
Generating image with prompt:
<@&1303451439787544669> draw me some anime cat girls with black hair and sparkly digital eyes

ℹ INFORMATIONS
Generating image with options:
{"imageModelProvider":"black-forest-labs/FLUX.1-schnell"}

◎ LOGS
Image generation successful, number of images:
1

◎ LOGS
Processing image 1:
generated_1733131072769_0

◎ LOGS
Generated caption for image 1: