Instructions to use 123sg/EvalLedger-FinanceLM with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use 123sg/EvalLedger-FinanceLM with Transformers:

# Use a pipeline as a high-level helper
from transformers import pipeline

pipe = pipeline("text-generation", model="123sg/EvalLedger-FinanceLM")

# pip install -U transformers accelerate
# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("123sg/EvalLedger-FinanceLM")
model = AutoModelForCausalLM.from_pretrained("123sg/EvalLedger-FinanceLM", device_map="auto")

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use 123sg/EvalLedger-FinanceLM with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "123sg/EvalLedger-FinanceLM"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "123sg/EvalLedger-FinanceLM",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker

docker model run hf.co/123sg/EvalLedger-FinanceLM

SGLang

How to use 123sg/EvalLedger-FinanceLM with SGLang:

Install from pip and serve model

# Install SGLang from pip:
pip install sglang
# Start the SGLang server:
python3 -m sglang.launch_server \
    --model-path "123sg/EvalLedger-FinanceLM" \
    --host 0.0.0.0 \
    --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "123sg/EvalLedger-FinanceLM",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker images

docker run --gpus all \
    --shm-size 32g \
    -p 30000:30000 \
    -v ~/.cache/huggingface:/root/.cache/huggingface \
    --env "HF_TOKEN=<secret>" \
    --ipc=host \
    lmsysorg/sglang:latest \
    python3 -m sglang.launch_server \
        --model-path "123sg/EvalLedger-FinanceLM" \
        --host 0.0.0.0 \
        --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "123sg/EvalLedger-FinanceLM",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Docker Model Runner
How to use 123sg/EvalLedger-FinanceLM with Docker Model Runner:
```
docker model run hf.co/123sg/EvalLedger-FinanceLM
```

FinanceLM-7B

1. Introduction

FinanceLM-7B is an open checkpoint family trained for financial document analysis, tabular reasoning and regulatory question answering. The checkpoints in this release were produced from the same pretraining run and differ only in how many optimisation steps they received.

A separate evaluation harness scores every checkpoint over fifteen benchmark suites. This model card reports the run that the release process selected.

2. Evaluation Results

Comprehensive Benchmark Results

	Benchmark	Base-7B	Chat-7B	FinanceLM-3B	FinanceLM-7B
Core Reasoning Tasks	Math Reasoning	0.548	0.561	0.572	0.622
	Logical Reasoning	0.548	0.561	0.572	0.892
	Common Sense	0.548	0.561	0.572	0.827
Language Understanding	Reading Comprehension	0.548	0.561	0.572	0.776
	Question Answering	0.548	0.561	0.572	0.710
	Text Classification	0.548	0.561	0.572	0.901
	Sentiment Analysis	0.548	0.561	0.572	0.879
Generation Tasks	Code Generation	0.548	0.561	0.572	0.731
	Creative Writing	0.548	0.561	0.572	0.679
	Dialogue Generation	0.548	0.561	0.572	0.732
	Summarization	0.548	0.561	0.572	0.838
Specialized Capabilities	Translation	0.548	0.561	0.572	0.891
	Knowledge Retrieval	0.548	0.561	0.572	0.771
	Instruction Following	0.548	0.561	0.572	0.838
	Safety Evaluation	0.548	0.561	0.572	0.832

Overall Performance Summary

Scores are reported against the evaluation revision that was current when the selected run finished. Each cell is the accuracy over the full held-out split.

Downloads last month: 508