Shirdel-Finance-E4B : GGUF

Shirdel-Finance-E4B is a finance-specialized language model built on Google Gemma-4 E4B and optimized for local inference, financial reasoning, cryptocurrency discussions, market analysis, and structured financial conversations.

The model is designed to provide strong performance across finance-related tasks while maintaining efficient execution on consumer hardware and mobile devices.

This release is distributed in GGUF format for compatibility with local inference ecosystems including llama.cpp, Ollama, LM Studio, Jan, and Open WebUI.

Model Details

  • Base model: Google Gemma-4 E4B

  • Fine-tuned model: Shirdel-Finance-E4B

  • Architecture: Gemma-4

  • Format: GGUF

  • Context window: 131072 tokens

  • Domain: Finance / Cryptocurrency / Trading

  • Intended use:

    • Financial question answering
    • Cryptocurrency analysis
    • Trading education
    • Market discussions
    • Fintech assistance
    • Structured financial reasoning

Supported Platforms

  • llama.cpp
  • Ollama
  • LM Studio
  • Jan
  • Open WebUI
  • KoboldCpp
  • llama-cpp-python

Available Model Files

  • Shirdel-Finance-E4B.Q2_K.gguf
  • Shirdel-Finance-E4B.Q3_K_M.gguf
  • Shirdel-Finance-E4B.Q4_K_M.gguf
  • Shirdel-Finance-E4B.Q5_K_M.gguf
  • Shirdel-Finance-E4B.Q6_K.gguf
  • Shirdel-Finance-E4B.Q8_0.gguf

Quantization recommendations:

Quant Recommended usage
Q2_K Very low memory devices
Q3_K_M Mobile devices
Q4_K_M Balanced quality and speed
Q5_K_M Higher quality
Q6_K Strong quality-performance balance
Q8_0 Maximum quality

How to use with llama-cpp-python

from llama_cpp import Llama

llm = Llama.from_pretrained(
    repo_id="Tesleum/Shirdel-Finance-E4B",
    filename="Shirdel-Finance-E4B.Q4_K_M.gguf"
)

response = llm.create_chat_completion(
    messages=[
        {
            "role":"user",
            "content":"Explain Bitcoin halving."
        }
    ]
)

print(response)

How to use with llama.cpp

Install

curl -LsSf https://llama.cpp/install.sh | sh

Run directly

llama-cli -hf Tesleum/Shirdel-Finance-E4B:Q4_K_M

Run server mode

llama-server -hf Tesleum/Shirdel-Finance-E4B:Q4_K_M

How to use with Ollama

Run directly:

ollama run hf.co/Tesleum/Shirdel-Finance-E4B:Q4_K_M

Or create manually:

FROM ./Shirdel-Finance-E4B.Q4_K_M.gguf

TEMPLATE """{{ .Prompt }}"""

PARAMETER temperature 0.7
PARAMETER top_p 0.95
PARAMETER num_ctx 131072

Then:

ollama create shirdel-finance -f Modelfile
ollama run shirdel-finance

Recommended Tasks

  • Cryptocurrency explanations
  • Trading concepts
  • Financial education
  • Stablecoin discussions
  • Tokenomics analysis
  • Market sentiment discussions
  • Economic reasoning
  • Fintech assistance

Limitations

  • Model outputs may contain inaccuracies.
  • Financial outputs should not be considered investment advice.
  • Market predictions may be incorrect.
  • Always independently verify important information.

Training Information

Fine-tuned using finance-oriented instruction datasets and domain-specific conversational examples designed to improve:

  • Financial terminology understanding
  • Cryptocurrency knowledge
  • Structured reasoning
  • Instruction following
  • Context retention

License

Please follow the original Google Gemma license and applicable downstream usage restrictions.

Author

Tesleum

Downloads last month
65
GGUF
Model size
8B params
Architecture
gemma4
Hardware compatibility
Log In to add your hardware

6-bit

Video Preview
loading

Model tree for Tesleum/Shirdel-Finance-E4B

Quantized
(264)
this model

Datasets used to train Tesleum/Shirdel-Finance-E4B