Instructions to use HuggingFaceBio/Carbon-500M with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use HuggingFaceBio/Carbon-500M with Transformers:

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

pipe = pipeline("text-generation", model="HuggingFaceBio/Carbon-500M")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("HuggingFaceBio/Carbon-500M")
model = AutoModelForCausalLM.from_pretrained("HuggingFaceBio/Carbon-500M")
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 HuggingFaceBio/Carbon-500M with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "HuggingFaceBio/Carbon-500M"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "HuggingFaceBio/Carbon-500M",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/HuggingFaceBio/Carbon-500M

SGLang

How to use HuggingFaceBio/Carbon-500M 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 "HuggingFaceBio/Carbon-500M" \
    --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": "HuggingFaceBio/Carbon-500M",
		"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 "HuggingFaceBio/Carbon-500M" \
        --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": "HuggingFaceBio/Carbon-500M",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use HuggingFaceBio/Carbon-500M with Docker Model Runner:
```
docker model run hf.co/HuggingFaceBio/Carbon-500M
```

Carbon-500M

Commit History

Update README.md

1223795
verified

lewtun HF Staff commited on about 17 hours ago

Update README.md

021dd2b
verified

lewtun HF Staff commited on about 17 hours ago

Update README.md

57127eb
verified

lewtun HF Staff commited on about 17 hours ago

fix: correct bos/eos/pad token IDs in generation_config (eos=pad=151643, bos=null)

4d2ebe0
verified

kashif HF Staff commited on about 17 hours ago

Update README.md

dff0b7f
verified

lewtun HF Staff commited on about 17 hours ago

Upload figures/carbon-500m-banner.png

6643225
verified

lewtun HF Staff commited on about 17 hours ago

Update README.md

f2c5de1
verified

loubnabnl HF Staff commited on about 18 hours ago

docs: fix dtype, remove trust_remote_code for model, clean up internal comments

c933421
verified

kashif HF Staff commited on 5 days ago

revert: restore LlamaForCausalLM in config.json

9d803d5
verified

kashif HF Staff commited on 5 days ago

revert: remove auto_map, restore LlamaForCausalLM (no trust_remote_code needed)

75822fe
verified

kashif HF Staff commited on 5 days ago

tokenizer: fix decode() to handle torch tensor input via .tolist()

6411d45
verified

kashif HF Staff commited on 6 days ago

modeling_carbon: replace _sample() override with stable LogitsProcessor API

6300b57
verified

kashif HF Staff commited on 6 days ago

Update tokenizer.py

e1cc2f4
verified

GenerTeam commited on 6 days ago

Upload modeling_carbon.py

6aad06c
verified

GenerTeam commited on 6 days ago

Update config.json

1f08d98
verified

GenerTeam commited on 6 days ago

model card: fix typos, remove TODOs, 48->49 kbp, add skip_special_tokens=True

ddc127f
verified

kashif HF Staff commited on 7 days ago

tokenizer: fix auto_dna_tags None -> False

098b919
verified

kashif HF Staff commited on 7 days ago

tokenizer: add auto_dna_tags to dna_config.json

8da7a58
verified

kashif HF Staff commited on 7 days ago

tokenizer: fix EOS append bug, decode skip_special_tokens=True, add auto_dna_tags

a5f56cd
verified

kashif HF Staff commited on 7 days ago

Update model card namespace references (#1)

a955d9a

lewtun HF Staff commited on 7 days ago

Update README.md

805e7ba
verified

loubnabnl HF Staff commited on 7 days ago

Promote hybrid step-286000 to main (300B CE + 300B FNS, total 600B tokens)

401b752
verified

loubnabnl HF Staff commited on 7 days ago

Update README.md

c3a251e
verified

loubnabnl HF Staff commited on 8 days ago

initial commit

fbabbaf
verified

loubnabnl HF Staff commited on 8 days ago

Commit History

Update README.md 1223795 verified

Update README.md 021dd2b verified

Update README.md 57127eb verified

fix: correct bos/eos/pad token IDs in generation_config (eos=pad=151643, bos=null) 4d2ebe0 verified

Update README.md dff0b7f verified

Upload figures/carbon-500m-banner.png 6643225 verified

Update README.md f2c5de1 verified

docs: fix dtype, remove trust_remote_code for model, clean up internal comments c933421 verified

revert: restore LlamaForCausalLM in config.json 9d803d5 verified

revert: remove auto_map, restore LlamaForCausalLM (no trust_remote_code needed) 75822fe verified

tokenizer: fix decode() to handle torch tensor input via .tolist() 6411d45 verified

modeling_carbon: replace _sample() override with stable LogitsProcessor API 6300b57 verified

Update tokenizer.py e1cc2f4 verified

Upload modeling_carbon.py 6aad06c verified

Update config.json 1f08d98 verified

model card: fix typos, remove TODOs, 48->49 kbp, add skip_special_tokens=True ddc127f verified

tokenizer: fix auto_dna_tags None -> False 098b919 verified

tokenizer: add auto_dna_tags to dna_config.json 8da7a58 verified

tokenizer: fix EOS append bug, decode skip_special_tokens=True, add auto_dna_tags a5f56cd verified

Update model card namespace references (#1) a955d9a

Update README.md 805e7ba verified

Promote hybrid step-286000 to main (300B CE + 300B FNS, total 600B tokens) 401b752 verified

Update README.md c3a251e verified

initial commit fbabbaf verified

Update README.md

1223795
verified

Update README.md

021dd2b
verified

Update README.md

57127eb
verified

fix: correct bos/eos/pad token IDs in generation_config (eos=pad=151643, bos=null)

4d2ebe0
verified

Update README.md

dff0b7f
verified

Upload figures/carbon-500m-banner.png

6643225
verified

Update README.md

f2c5de1
verified

docs: fix dtype, remove trust_remote_code for model, clean up internal comments

c933421
verified

revert: restore LlamaForCausalLM in config.json

9d803d5
verified

revert: remove auto_map, restore LlamaForCausalLM (no trust_remote_code needed)

75822fe
verified

tokenizer: fix decode() to handle torch tensor input via .tolist()

6411d45
verified

modeling_carbon: replace _sample() override with stable LogitsProcessor API

6300b57
verified

Update tokenizer.py

e1cc2f4
verified

Upload modeling_carbon.py

6aad06c
verified

Update config.json

1f08d98
verified

model card: fix typos, remove TODOs, 48->49 kbp, add skip_special_tokens=True

ddc127f
verified

tokenizer: fix auto_dna_tags None -> False

098b919
verified

tokenizer: add auto_dna_tags to dna_config.json

8da7a58
verified

tokenizer: fix EOS append bug, decode skip_special_tokens=True, add auto_dna_tags

a5f56cd
verified

Update model card namespace references (#1)

a955d9a

Update README.md

805e7ba
verified

Promote hybrid step-286000 to main (300B CE + 300B FNS, total 600B tokens)

401b752
verified

Update README.md

c3a251e
verified

initial commit

fbabbaf
verified