Instructions to use midnigter/CryptoGemma-4B-v1 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use midnigter/CryptoGemma-4B-v1 with Transformers:

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

pipe = pipeline("text-generation", model="midnigter/CryptoGemma-4B-v1")
messages = [
    {
        "role": "user",
        "content": [
            {"type": "image", "url": "https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/p-blog/candy.JPG"},
            {"type": "text", "text": "What animal is on the candy?"}
        ]
    },
]
pipe(text=messages)

# Load model directly
from transformers import AutoProcessor, AutoModelForImageTextToText

processor = AutoProcessor.from_pretrained("midnigter/CryptoGemma-4B-v1")
model = AutoModelForImageTextToText.from_pretrained("midnigter/CryptoGemma-4B-v1")
messages = [
    {
        "role": "user",
        "content": [
            {"type": "image", "url": "https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/p-blog/candy.JPG"},
            {"type": "text", "text": "What animal is on the candy?"}
        ]
    },
]
inputs = processor.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(processor.decode(outputs[0][inputs["input_ids"].shape[-1]:]))

MLX

How to use midnigter/CryptoGemma-4B-v1 with MLX:

# Make sure mlx-lm is installed
# pip install --upgrade mlx-lm

# Generate text with mlx-lm
from mlx_lm import load, generate

model, tokenizer = load("midnigter/CryptoGemma-4B-v1")

prompt = "Write a story about Einstein"
messages = [{"role": "user", "content": prompt}]
prompt = tokenizer.apply_chat_template(
    messages, add_generation_prompt=True
)

text = generate(model, tokenizer, prompt=prompt, verbose=True)

Notebooks
Google Colab
Kaggle
Local Apps
LM Studio

vLLM

How to use midnigter/CryptoGemma-4B-v1 with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "midnigter/CryptoGemma-4B-v1"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "midnigter/CryptoGemma-4B-v1",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/midnigter/CryptoGemma-4B-v1

SGLang

How to use midnigter/CryptoGemma-4B-v1 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 "midnigter/CryptoGemma-4B-v1" \
    --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": "midnigter/CryptoGemma-4B-v1",
		"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 "midnigter/CryptoGemma-4B-v1" \
        --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": "midnigter/CryptoGemma-4B-v1",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Pi new

How to use midnigter/CryptoGemma-4B-v1 with Pi:

Start the MLX server

# Install MLX LM:
uv tool install mlx-lm
# Start a local OpenAI-compatible server:
mlx_lm.server --model "midnigter/CryptoGemma-4B-v1"

Configure the model in Pi

# Install Pi:
npm install -g @mariozechner/pi-coding-agent
# Add to ~/.pi/agent/models.json:
{
  "providers": {
    "mlx-lm": {
      "baseUrl": "http://localhost:8080/v1",
      "api": "openai-completions",
      "apiKey": "none",
      "models": [
        {
          "id": "midnigter/CryptoGemma-4B-v1"
        }
      ]
    }
  }
}

Run Pi

# Start Pi in your project directory:
pi

Hermes Agent new

How to use midnigter/CryptoGemma-4B-v1 with Hermes Agent:

Start the MLX server

# Install MLX LM:
uv tool install mlx-lm
# Start a local OpenAI-compatible server:
mlx_lm.server --model "midnigter/CryptoGemma-4B-v1"

Configure Hermes

# Install Hermes:
curl -fsSL https://hermes-agent.nousresearch.com/install.sh | bash
hermes setup
# Point Hermes at the local server:
hermes config set model.provider custom
hermes config set model.base_url http://127.0.0.1:8080/v1
hermes config set model.default midnigter/CryptoGemma-4B-v1

Run Hermes

hermes

MLX LM

How to use midnigter/CryptoGemma-4B-v1 with MLX LM:

Generate or start a chat session

# Install MLX LM
uv tool install mlx-lm
# Interactive chat REPL
mlx_lm.chat --model "midnigter/CryptoGemma-4B-v1"

Run an OpenAI-compatible server

# Install MLX LM
uv tool install mlx-lm
# Start the server
mlx_lm.server --model "midnigter/CryptoGemma-4B-v1"
# Calling the OpenAI-compatible server with curl
curl -X POST "http://localhost:8000/v1/chat/completions" \
   -H "Content-Type: application/json" \
   --data '{
     "model": "midnigter/CryptoGemma-4B-v1",
     "messages": [
       {"role": "user", "content": "Hello"}
     ]
   }'

Docker Model Runner
How to use midnigter/CryptoGemma-4B-v1 with Docker Model Runner:
```
docker model run hf.co/midnigter/CryptoGemma-4B-v1
```

CryptoGemma-4B-v1

The first open-source crypto-specialized LLM. Fine-tuned from Google's Gemma 4 E4B on 56,000+ crypto-specific examples.

What is CryptoGemma?

CryptoGemma is a 4B parameter language model specialized for cryptocurrency and blockchain analysis. Built by SkillForge, it understands:

Token analysis — fundamental and on-chain metrics
Scam detection — smart contract risk assessment
Portfolio advice — allocation and rebalancing strategies
Technical analysis — chart patterns, indicators, market structure
DeFi protocols — liquidity, yield farming, lending
Blockchain security — audit patterns, vulnerability detection
Market sentiment — social signals, fear & greed analysis

Training Details

Parameter	Value
Base model	google/gemma-4-E4B-it (4B params)
Dataset size	56,612 examples
Method	LoRA (8 layers)
Iterations	500
Learning rate	1e-5
Train loss	0.120
Validation loss	0.153
Training time	~1.5 hours on Mac Studio M2 Ultra
Framework	mlx-lm

Dataset Composition

Source	Examples
Cogneo Crypto Sentiment	43,027
Forta Malicious Contracts	12,000
HuggingFace open datasets (Bitcoin, smart contracts, blockchain)	3,267
ChainGPT-generated Q&A (16 categories)	~300
Total (after dedup)	56,612

Usage

With Ollama

# Download and run
ollama run cryptogemma

# Example
>>> Analyze Bitcoin's current market structure and key support levels

With Transformers

from transformers import AutoTokenizer, AutoModelForCausalLM

model = AutoModelForCausalLM.from_pretrained("SkillForge/CryptoGemma-4B-v1")
tokenizer = AutoTokenizer.from_pretrained("SkillForge/CryptoGemma-4B-v1")

prompt = "Analyze the risk factors of this DeFi protocol..."
inputs = tokenizer(prompt, return_tensors="pt")
outputs = model.generate(**inputs, max_new_tokens=512)
print(tokenizer.decode(outputs[0], skip_special_tokens=True))

With MLX (Apple Silicon)

from mlx_lm import load, generate

model, tokenizer = load("SkillForge/CryptoGemma-4B-v1")
response = generate(model, tokenizer, prompt="What are the key metrics for evaluating a DeFi token?", max_tokens=512)

Skills

CryptoGemma powers 4 ready-to-use skills:

Token Analyzer — ticker to full breakdown
Scam Detector — contract address to risk score
Portfolio Advisor — positions to recommendations
TA Generator — trading pair to technical analysis

License

Apache 2.0 (same as base Gemma model)

Citation

@misc{cryptogemma2026,
  title={CryptoGemma-4B-v1: Open-Source Crypto-Specialized LLM},
  author={SkillForge Team},
  year={2026},
  url={https://huggingface.co/SkillForge/CryptoGemma-4B-v1}
}

Downloads last month: 226

Safetensors

Model size

8B params

Tensor type

BF16

MLX

Hardware compatibility

Quantized

Model tree for midnigter/CryptoGemma-4B-v1

Base model

google/gemma-4-E4B

Finetuned

google/gemma-4-E4B-it