Instructions to use Banaxi-Tech/BananaMind-v1 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- llama-cpp-python
How to use Banaxi-Tech/BananaMind-v1 with llama-cpp-python:
# !pip install llama-cpp-python from llama_cpp import Llama llm = Llama.from_pretrained( repo_id="Banaxi-Tech/BananaMind-v1", filename="BananaMind-V1.gguf", )
output = llm( "Once upon a time,", max_tokens=512, echo=True ) print(output)
- Notebooks
- Google Colab
- Kaggle
- Local Apps
- llama.cpp
How to use Banaxi-Tech/BananaMind-v1 with llama.cpp:
Install from brew
brew install llama.cpp # Start a local OpenAI-compatible server with a web UI: llama-server -hf Banaxi-Tech/BananaMind-v1 # Run inference directly in the terminal: llama-cli -hf Banaxi-Tech/BananaMind-v1
Install from WinGet (Windows)
winget install llama.cpp # Start a local OpenAI-compatible server with a web UI: llama-server -hf Banaxi-Tech/BananaMind-v1 # Run inference directly in the terminal: llama-cli -hf Banaxi-Tech/BananaMind-v1
Use pre-built binary
# Download pre-built binary from: # https://github.com/ggerganov/llama.cpp/releases # Start a local OpenAI-compatible server with a web UI: ./llama-server -hf Banaxi-Tech/BananaMind-v1 # Run inference directly in the terminal: ./llama-cli -hf Banaxi-Tech/BananaMind-v1
Build from source code
git clone https://github.com/ggerganov/llama.cpp.git cd llama.cpp cmake -B build cmake --build build -j --target llama-server llama-cli # Start a local OpenAI-compatible server with a web UI: ./build/bin/llama-server -hf Banaxi-Tech/BananaMind-v1 # Run inference directly in the terminal: ./build/bin/llama-cli -hf Banaxi-Tech/BananaMind-v1
Use Docker
docker model run hf.co/Banaxi-Tech/BananaMind-v1
- LM Studio
- Jan
- Ollama
How to use Banaxi-Tech/BananaMind-v1 with Ollama:
ollama run hf.co/Banaxi-Tech/BananaMind-v1
- Unsloth Studio new
How to use Banaxi-Tech/BananaMind-v1 with Unsloth Studio:
Install Unsloth Studio (macOS, Linux, WSL)
curl -fsSL https://unsloth.ai/install.sh | sh # Run unsloth studio unsloth studio -H 0.0.0.0 -p 8888 # Then open http://localhost:8888 in your browser # Search for Banaxi-Tech/BananaMind-v1 to start chatting
Install Unsloth Studio (Windows)
irm https://unsloth.ai/install.ps1 | iex # Run unsloth studio unsloth studio -H 0.0.0.0 -p 8888 # Then open http://localhost:8888 in your browser # Search for Banaxi-Tech/BananaMind-v1 to start chatting
Using HuggingFace Spaces for Unsloth
# No setup required # Open https://huggingface.co/spaces/unsloth/studio in your browser # Search for Banaxi-Tech/BananaMind-v1 to start chatting
- Docker Model Runner
How to use Banaxi-Tech/BananaMind-v1 with Docker Model Runner:
docker model run hf.co/Banaxi-Tech/BananaMind-v1
- Lemonade
How to use Banaxi-Tech/BananaMind-v1 with Lemonade:
Pull the model
# Download Lemonade from https://lemonade-server.ai/ lemonade pull Banaxi-Tech/BananaMind-v1
Run and chat with the model
lemonade run user.BananaMind-v1-{{QUANT_TAG}}List all available models
lemonade list
π BananaMind V1
β οΈ Honest disclaimer: This model is not very good. It's a small experimental model trained as a personal project. Manage your expectations accordingly.
BananaMind V1 is a 125M parameter causal language model fine-tuned on conversational data. It is uploaded in GGUF format for use with llama.cpp and compatible runtimes.
Model Details
| Property | Value |
|---|---|
| Model Name | BananaMind V1 |
| Parameters | 125M |
| Context Length | 512 tokens |
| Format | GGUF |
| License | Apache 2.0 |
| Language | English |
Training Data
| Dataset | License |
|---|---|
| OpenAssistant/oasst1 | Apache 2.0 |
Intended Use
- Experimenting with small language models
- Local inference on CPU via llama.cpp
- Educational / research purposes
- Tinkering and testing
Not intended for: Production use, safety-critical applications, or anything requiring reliable or accurate outputs.
How to Use
With llama.cpp
./llama-cli -m BananaMind-V1.gguf \
-p "User: Hello! How are you?\nAssistant:" \
-n 128 \
--ctx-size 512
With Python (via llama-cpp-python)
from llama_cpp import Llama
llm = Llama(
model_path="BananaMind-V1.gguf",
n_ctx=512,
)
output = llm(
"User: What is the capital of France?\nAssistant:",
max_tokens=128,
stop=["User:"],
)
print(output["choices"][0]["text"])
With Ollama
ollama run banaxitech/bananamind-v1
``
---
## Limitations & Known Issues
- **Small capacity:** 125M parameters is quite limited β don't expect GPT-4 quality.
- **Short context:** 512 token context window means it forgets things quickly.
- **Hallucinations:** Will confidently make things up.
- **Inconsistency:** Outputs can vary wildly between runs.
- **Generally not very good** β as stated upfront. Consider this a starting point, not a destination.
---
## Evaluation
No formal benchmarks were run. Informal testing suggests the model can hold basic conversations but struggles with anything requiring reasoning, factual recall, or multi-turn coherence.
---
## License
This model is released under the **Apache 2.0** license, consistent with its training data. See [LICENSE](https://www.apache.org/licenses/LICENSE-2.0) for details.
---
## Citation
If you somehow use this model in research (why?), you can cite it informally as:
BananaMind V1 (2024). A small 125M parameter GGUF language model. Personal project.
---
*Made with π and questionable judgment.*
- Downloads last month
- 10
We're not able to determine the quantization variants.