FinanceGPT Developer Platform
Adapters, PEFT & fine-tuning recipes
Prepare pinned FinanceGPT transformer models for LoRA and PEFT workflows using reproducible local recipes. HF9 does not execute training jobs or publish adapters automatically.
Base models
1
Transformer models in the public catalog
Adapter-capable
1
Pinned repository + revision available
Published adapters
0
HF9 does not invent or auto-publish adapters
Preferred output
Safetensors
Adapter and merged weight files
Governed adaptation policy
Recipes are developer guidance, not training authority.
Pinned base revisions
Required
Dataset review
Required before training
Evaluation
Required before publication
Publication
Explicit action only
Adapter-ready base models
LoRA is the preferred initial method; prefix and prompt tuning remain documented compatibility methods.
| Model | Task | Revision | Methods | Status | Action |
|---|---|---|---|---|---|
all-MiniLM-L6-v2 sentence-transformers/all-MiniLM-L6-v2 |
feature-extraction | ea78891063587eb050ed... |
lora, prefix-tuning, prompt-tuning | Ready | Open recipe |