llama-3.1-8b-financial-reasoner-v1
A fine-tune of meta-llama/Llama-3.1-8B-Instruct for two-stage financial analysis of a stock ticker:
- Task A – classify how the market reacted to a news item given the price move (
good,bad,neutral,overreaction_down,overreaction_up). - Task B – given that reaction, a valuation gap and fundamentals, write the reasoning to find a recommendation (BUY/SELL/HOLD) and answer as JSON.
Format
Full merged checkpoint in bfloat16 (~16 GB, 4 safetee-quantized — load it in 4-bit at runtime if you need tofit a small GPU:
from unsloth import FastLanguageModel
model, tokenizer = FastLanguageModel.from_pretrained(
"gaparecido/llama-3.1-8b-financial-reasoner-v1",
max_seq_length=2048,
load_in_4bit=True,
)
FastLanguageModel.for_inference(model)
Serving notes
- Runs fine in fp16 on a T4 (16 GB) when loaded in 4-bit — no bf16 GPU required.
- use_cache is already true in the config; no pin needed
- Uses the standard Llama 3.1 chat template shipped in tokenizer_config.json.
Training
Trained with Unsloth (https://github.com/unslothai/unsloth) (QLoRA, 4-bit) and Hugging Face TRL, loaded from Unsloth's 4-bit mirror of the
base model.