Financial NLP
FinGPT Sentiment
Instruction-tuning financial news and social-media sentiment. Used to ask whether a small fine-tune matches a frontier model on tone.
As of 13 August 2026. Conditional pack. FinGPT sentiment instruction set. LoRA papers claim small models can match GPT-4 on this distribution. Those claims stay in their papers.
FinGPT Sentiment is the instruction-tuning set behind FinGPT, covering financial news and social media. Accuracy and macro F1 stay separate. A LoRA versus GPT-4 claim is a paper result, not a 2026 Fable 5 cell.
Recheck terms before commercial use. No official Fable 5 or GPT-5.6 table is cited.
- Scale
- FinGPT sentiment instruction set. Pin the hosted revision at acquisition.
- Provenance
- FinGPT sentiment instruction set. Recheck terms before commercial use.
- Contamination
- High. Public instruction set behind FinGPT.
- Cannot establish
- GPT-4 or Fable 5 superiority without a frozen replay. Not filing QA and not alpha.
Metrics, kept separate
- Classification accuracy
- Macro F1
- Per-class F1
No cited public figure is installed for this pack. Method, scale, and gates still apply. A missing chart is not a pass rate.
What the figures bound
The pack is listed because buyers still ask whether sentiment is worth frontier prices. That question needs a frozen replay, not a catalog headline.
No chart is installed without a cited exact split.
What this page does not conclude
- GPT-4, FinGPT, Fable 5, or GPT-5.6 superiority on this set
- Trading-signal quality
- A commercial run before terms are approved
- Filing-QA performance
Cited as calibration on these memos — not a score of those products.
Not attached to a software, advisory, or automation memo yet.