Run Flux LoRAs on Replicate

[zeke]

This PR updates the ReplicateTextToImageTask method to support running arbitrary LoRAs on Replicate.

How it works:

1. Detect LoRAs by their tags before inference.
2. Use https://replicate.com/black-forest-labs/flux-dev-lora for inference
3. Use the requested LoRA as the weights_url input to the flux-dev-lora model

Companion PR

• Add support for HF tag mappings, starting with Flux LoRA replicate/huggingface-model-mappings#9

Testing

Remove the skip from describe.skip("InferenceClient" ... in InferenceClient.spec.ts

Then:

$ pnpx vitest run --config vitest.config.mts -t "Replicate"

[SBrandeis]

Nice and simple!
For information, mapping lora models requires to create a different kind of "inference provider mapping" (tag-based).
See the (wip) documentation PR here for more info: huggingface/hub-docs#1717

[hanouticelina]

Nice!

[zeke]

Thanks for all the helpful reviews! Made another pass of updates. I think this is ready to go. 🚀

[SBrandeis]

LGTM!
I will let @hanouticelina do a final pass on this then let's merge!

[hanouticelina]

thank you @zeke for the PR! let's first update the mapping as mentioned in this internal slack message before merging.

[zeke]

update the mapping

☝🏼 Done! replicate/huggingface-model-mappings#12