Instructions to use anthonym21/pinescript-v5-instructions-merged with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use anthonym21/pinescript-v5-instructions-merged with Transformers:

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

pipe = pipeline("text-generation", model="anthonym21/pinescript-v5-instructions-merged")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("anthonym21/pinescript-v5-instructions-merged")
model = AutoModelForCausalLM.from_pretrained("anthonym21/pinescript-v5-instructions-merged")
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

vLLM

How to use anthonym21/pinescript-v5-instructions-merged with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "anthonym21/pinescript-v5-instructions-merged"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "anthonym21/pinescript-v5-instructions-merged",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/anthonym21/pinescript-v5-instructions-merged

SGLang

How to use anthonym21/pinescript-v5-instructions-merged 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 "anthonym21/pinescript-v5-instructions-merged" \
    --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": "anthonym21/pinescript-v5-instructions-merged",
		"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 "anthonym21/pinescript-v5-instructions-merged" \
        --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": "anthonym21/pinescript-v5-instructions-merged",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use anthonym21/pinescript-v5-instructions-merged with Docker Model Runner:
```
docker model run hf.co/anthonym21/pinescript-v5-instructions-merged
```

PineScript v5 Code Generator

A fine-tuned CodeGemma 7B model specialized in generating TradingView PineScript v5 code for trading indicators, strategies, and libraries.

Model Description

This model was fine-tuned on 5,000+ PineScript v5 examples from the PineScripts-Permissive dataset, which contains high-quality, permissively-licensed scripts from TradingView.

Base Model: google/codegemma-7b-it
Fine-tuning Method: QLoRA (4-bit quantization + LoRA adapters)
Training Data: 4,774 instruction/response pairs
Context Length: 4096 tokens

Intended Use

Generate PineScript v5 code for:

Technical indicators (RSI, MACD, Bollinger Bands, custom indicators)
Trading strategies with backtesting
Reusable libraries
Alert conditions
Custom visualizations

Usage

from transformers import AutoModelForCausalLM, AutoTokenizer
import torch

model_id = "anthonym21/pinescript-v5-instructions-merged"

tokenizer = AutoTokenizer.from_pretrained(model_id)
model = AutoModelForCausalLM.from_pretrained(
    model_id,
    torch_dtype=torch.bfloat16,
    device_map="auto",
)

def generate_pinescript(prompt, max_tokens=1024):
    formatted = f"### Human: {prompt}\n### Assistant:"
    inputs = tokenizer(formatted, return_tensors="pt").to(model.device)

    outputs = model.generate(
        **inputs,
        max_new_tokens=max_tokens,
        temperature=0.7,
        top_p=0.9,
        do_sample=True,
        pad_token_id=tokenizer.eos_token_id,
    )

    response = tokenizer.decode(outputs[0], skip_special_tokens=True)
    return response.split("### Assistant:")[-1].strip()

# Example
code = generate_pinescript("Write a PineScript v5 RSI indicator with overbought/oversold zones")
print(code)

Example Prompts

Prompt	Description
"Write a PineScript v5 indicator that shows RSI with dynamic overbought/oversold levels"	RSI with adaptive levels
"Create a MACD crossover strategy with stop loss and take profit"	Complete trading strategy
"Write a Bollinger Bands indicator with squeeze detection"	Volatility indicator
"Create a multi-timeframe moving average indicator"	MTF analysis tool

Training Details

Parameter	Value
Epochs	3
Batch Size	2
Gradient Accumulation	8
Learning Rate	2e-4
LoRA r	64
LoRA alpha	128
Max Seq Length	4096
Quantization	4-bit (nf4)

Limitations

Generates PineScript v5 syntax; may not be compatible with older versions
Code should be reviewed and tested before live trading
Complex multi-indicator strategies may require refinement
Does not provide financial advice

Dataset

Training data sourced from mrmegatelo/PineScripts-Permissive:

5,848 PineScript v5 scripts
Filtered to 4,774 high-quality examples
Includes indicators, strategies, and libraries
All scripts under permissive licenses (MPL-2.0, Apache-2.0)

Citation

@misc{pinescript-v5-instructions-merged,
  author = {Anthony Maio},
  title = {PineScript v5 Code Generator},
  year = {2025},
  publisher = {HuggingFace},
  url = {https://huggingface.co/anthonym21/pinescript-v5-instructions-merged}
}

License

Apache 2.0 (same as base model)

Downloads last month: 18

Safetensors

Model size

9B params

Tensor type

BF16

Model tree for anthonym21/pinescript-v5-instructions-merged

Base model

google/codegemma-7b-it

Finetuned

(8)

this model

anthonym21
/

pinescript-v5-instructions-merged