Add support for HF tag mappings, starting with Flux LoRA #9
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
zeke
changed the title
Member
Author
|
I ran this locally and it looks like it worked! ⚡ |
Wauplin
approved these changes
May 1, 2025
There was a problem hiding this comment.
Love this tool!
(+ seems to work according to https://huggingface.co/api/partners/replicate/models and https://huggingface.co/api/models/fofr/flux-80s-cyberpunk?expand[]=inferenceProviderMapping)
src/hf.ts
Outdated
| @@ -1,8 +1,11 @@ | |||
| interface MappingItem { | |||
There was a problem hiding this comment.
nit-picking here but an interface like that would better describe the 2 only combinations that should be supported
interface ModelMappingItem {
task: string;
hfModel: string;
providerModel: string;
status?: 'live' | 'staging';
}
interface TagFilterMappingItem {
task: string;
providerModel: string;
status?: 'live' | 'staging';
type: 'tag-filter';
tags: string[];
adapterType: 'lora';
}
type MappingItem = ModelMappingItem | TagFilterMappingItem;
zeke
force-pushed
the
add-support-tag-mappings,-starting-with-flux-lora
branch
from
e73a3cd to
54d23c0
Compare
hanouticelina
added a commit
to huggingface/huggingface.js
that referenced
this pull request
May 5, 2025
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 - 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" ``` --------- Co-authored-by: Simon Brandeis <[email protected]> Co-authored-by: célina <[email protected]> Co-authored-by: Lucain <[email protected]>
Deep-unlearning
pushed a commit
to huggingface/huggingface.js
that referenced
this pull request
May 13, 2025
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 - 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" ``` --------- Co-authored-by: Simon Brandeis <[email protected]> Co-authored-by: célina <[email protected]> Co-authored-by: Lucain <[email protected]>
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
This PR adds support the new HuggingFace tag mappings feature.
Tag mappings allow you to map multiple Hugging Face models to a single Replicate model based on their tags. This is useful when you have a single Replicate model that can handle multiple variations of a model (like different LoRA adapters).
Companion PR in the inference client: