Instructions to use HuggingFaceBio/Carbon-3B-GGUF with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use HuggingFaceBio/Carbon-3B-GGUF with llama-cpp-python:

# !pip install llama-cpp-python

from llama_cpp import Llama

llm = Llama.from_pretrained(
	repo_id="HuggingFaceBio/Carbon-3B-GGUF",
	filename="carbon-3b-bf16.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 HuggingFaceBio/Carbon-3B-GGUF with llama.cpp:

Install from brew

brew install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama-server -hf HuggingFaceBio/Carbon-3B-GGUF:BF16
# Run inference directly in the terminal:
llama-cli -hf HuggingFaceBio/Carbon-3B-GGUF:BF16

Install from WinGet (Windows)

winget install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama-server -hf HuggingFaceBio/Carbon-3B-GGUF:BF16
# Run inference directly in the terminal:
llama-cli -hf HuggingFaceBio/Carbon-3B-GGUF:BF16

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 HuggingFaceBio/Carbon-3B-GGUF:BF16
# Run inference directly in the terminal:
./llama-cli -hf HuggingFaceBio/Carbon-3B-GGUF:BF16

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 HuggingFaceBio/Carbon-3B-GGUF:BF16
# Run inference directly in the terminal:
./build/bin/llama-cli -hf HuggingFaceBio/Carbon-3B-GGUF:BF16

Use Docker

docker model run hf.co/HuggingFaceBio/Carbon-3B-GGUF:BF16

LM Studio
Jan
Ollama
How to use HuggingFaceBio/Carbon-3B-GGUF with Ollama:
```
ollama run hf.co/HuggingFaceBio/Carbon-3B-GGUF:BF16
```

Unsloth Studio new

How to use HuggingFaceBio/Carbon-3B-GGUF 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 HuggingFaceBio/Carbon-3B-GGUF 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 HuggingFaceBio/Carbon-3B-GGUF to start chatting

Using HuggingFace Spaces for Unsloth

# No setup required
# Open https://huggingface.co/spaces/unsloth/studio in your browser
# Search for HuggingFaceBio/Carbon-3B-GGUF to start chatting

Docker Model Runner
How to use HuggingFaceBio/Carbon-3B-GGUF with Docker Model Runner:
```
docker model run hf.co/HuggingFaceBio/Carbon-3B-GGUF:BF16
```

Lemonade

How to use HuggingFaceBio/Carbon-3B-GGUF with Lemonade:

Pull the model

# Download Lemonade from https://lemonade-server.ai/
lemonade pull HuggingFaceBio/Carbon-3B-GGUF:BF16

Run and chat with the model

lemonade run user.Carbon-3B-GGUF-BF16

List all available models

lemonade list

kashif HF Staff commited on 1 day ago

Commit

12c3c9f

verified ·

1 Parent(s): 1428e66

card: add download step, point parity to ggml-org/vocabs

Browse files

Files changed (1) hide show

README.md +7 -1

README.md CHANGED Viewed

@@ -35,6 +35,12 @@ cd llama.cpp && cmake -B build && cmake --build build -j
 ## Usage
 ### Basic DNA completion
 ```bash
@@ -105,7 +111,7 @@ The `revision="fns"` example from the source card needs custom modeling code (fa
 ## Tokenization parity
-For every prompt in the [test fixture](https://github.com/kashif/llama.cpp/blob/carbon-3b-tokenizer/models/ggml-vocab-hybriddna.gguf.inp), llama.cpp produces byte-for-byte identical token IDs to the Python `HybridDNATokenizer` (loaded with `trust_remote_code=True`).
 ## See also

 ## Usage
+### Download
+```bash
+hf download HuggingFaceBio/Carbon-3B-GGUF carbon-3b-bf16.gguf --local-dir .
+```
 ### Basic DNA completion
 ```bash
 ## Tokenization parity
+llama.cpp produces byte-for-byte identical token IDs to the Python `HybridDNATokenizer` (loaded with `trust_remote_code=True`) on the standard `<dna>`/metadata/edge-case fixtures shipped in [`ggml-org/vocabs`](https://huggingface.co/ggml-org/vocabs).
 ## See also