Instructions to use InvestmentResearchAI/LLM-ADE_tiny-v0.001 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use InvestmentResearchAI/LLM-ADE_tiny-v0.001 with Transformers:

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

pipe = pipeline("text-generation", model="InvestmentResearchAI/LLM-ADE_tiny-v0.001")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("InvestmentResearchAI/LLM-ADE_tiny-v0.001")
model = AutoModelForCausalLM.from_pretrained("InvestmentResearchAI/LLM-ADE_tiny-v0.001", device_map="auto")
messages = [
    {"role": "user", "content": "Who are you?"},
]
inputs = tokenizer.apply_chat_template(
	messages,
	add_generation_prompt=True,
	tokenize=True,
	return_dict=True,
	return_tensors="pt",
).to(model.device)

outputs = model.generate(**inputs, max_new_tokens=40)
print(tokenizer.decode(outputs[0][inputs["input_ids"].shape[-1]:]))

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use InvestmentResearchAI/LLM-ADE_tiny-v0.001 with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "InvestmentResearchAI/LLM-ADE_tiny-v0.001"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "InvestmentResearchAI/LLM-ADE_tiny-v0.001",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/InvestmentResearchAI/LLM-ADE_tiny-v0.001

SGLang

How to use InvestmentResearchAI/LLM-ADE_tiny-v0.001 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 "InvestmentResearchAI/LLM-ADE_tiny-v0.001" \
    --host 0.0.0.0 \
    --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "InvestmentResearchAI/LLM-ADE_tiny-v0.001",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

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 "InvestmentResearchAI/LLM-ADE_tiny-v0.001" \
        --host 0.0.0.0 \
        --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "InvestmentResearchAI/LLM-ADE_tiny-v0.001",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use InvestmentResearchAI/LLM-ADE_tiny-v0.001 with Docker Model Runner:
```
docker model run hf.co/InvestmentResearchAI/LLM-ADE_tiny-v0.001
```

LLM-ADE_tiny-v0.001 / README.md

WilliamGazeley

Update prompt template in examples

93e33c7 verified about 2 years ago

preview code

Raw

History Blame Contribute Delete

1.17 kB

metadata

language:
  - en
license: mit
tags:
  - finance
pipeline_tag: text-generation
widget:
  - example_title: Easy
    text: |
      <|im_start|>user
      How do call options benefit the buyer?<|im_end|>
      <|im_start|>assistant
  - example_title: Medium
    text: >
      <|im_start|>user

      Why might a trader choose to quickly exit a losing position, even if they
      still believe in the original trade idea?<|im_end|>

      <|im_start|>assistant
  - example_title: Hard
    text: >
      <|im_start|>user

      In the context of Harry Markowitz's Portfolio Selection theory, what does
      an 'efficient' portfolio refer to?<|im_end|>

      <|im_start|>assistant
inference:
  parameters:
    temperature: 0.2
    min_new_tokens: 20
    max_new_tokens: 250

AlphaBlind Tiny v0.001

Our Proof-of-Concept (POC) for the LLM-ADE framework (https://arxiv.org/abs/2404.13028). A very early, initial version of TinyLlama processing and ingesting llm-ade-fin_data-subset-earnings-10k and other financial data with the LLM-ADE framework.

Note: This model has not been thoroughly tested, and is very small - it can run on a Macbook Pro. Please do not use this version of the model as is.