Implement chat with image for ai_automators

Thanks @fishfree for reporting this, I didn't even realize it was missing. It is added now and can be tested in dev release.

In the bottom of the advanced settings you will have a field for the image that you can choose and also if you choose an image field, you can choose an image style to apply to lower costs (recommended).

Please test if you have time and set to Fixed if it works.

@marcus Thank you! I can select image field. But it request failed. The AI log said:

Prompt
user
Please extract text from this image:

Do not include any explanations, only provide a RFC8259 compliant JSON response following this format without deviation.
[{"value": "requested value"}]

Operation Type
chat
Provider
ollama
Model
llava:latest
Tags
chat
Response
{"id":"chatcmpl-291","object":"chat.completion","created":1727049200,"model":"llava:latest","systemFingerprint":"fp_ollama","choices":[{"index":0,"message":{"role":"assistant","content":" ```json\n}[{\"value\": \"No request was made.\"}]\n``` ","toolCalls":[],"functionCall":null},"finishReason":"stop"}],"usage":{"promptTokens":57,"completionTokens":19,"totalTokens":76}}
Configuration
max_tokens:
1024

temperature:
1

top_p:
1

Extra Data
[]
Authored on
Mon, 09/23/2024 - 07:53

If you succeeded, pls share me a screenshot of configuration.

@fishfree - this is another issue, I created a ticket for it here: https://www.drupal.org/project/ai/issues/3476097 ✨ Add image Chat input possibility with Ollama Active

We simply didn't add the image support in Ollama yet.

@fishfree - the images in Ollama is testable here if you have time: https://www.drupal.org/project/ai/issues/3476097# →

Automatically closed - issue fixed for 2 weeks with no activity.