Instructions to use LiquidAI/LFM3-1.7B with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use LiquidAI/LFM3-1.7B with Transformers:

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

pipe = pipeline("text-generation", model="LiquidAI/LFM3-1.7B", trust_remote_code=True)
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# pip install -U transformers accelerate
# Load model directly
from transformers import AutoModelForCausalLM
model = AutoModelForCausalLM.from_pretrained("LiquidAI/LFM3-1.7B", trust_remote_code=True, device_map="auto")

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use LiquidAI/LFM3-1.7B with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "LiquidAI/LFM3-1.7B"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "LiquidAI/LFM3-1.7B",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/LiquidAI/LFM3-1.7B

SGLang

How to use LiquidAI/LFM3-1.7B 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 "LiquidAI/LFM3-1.7B" \
    --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": "LiquidAI/LFM3-1.7B",
		"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 "LiquidAI/LFM3-1.7B" \
        --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": "LiquidAI/LFM3-1.7B",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use LiquidAI/LFM3-1.7B with Docker Model Runner:
```
docker model run hf.co/LiquidAI/LFM3-1.7B
```

Got error when running and finetuning

by mahdisml - opened about 18 hours ago

Discussion

mahdisml

about 18 hours ago

•

edited about 18 hours ago

hi , i got this error when finetuning with unsloth :
Unsloth: Failed to load model. Both AutoConfig and PeftConfig loading failed. AutoConfig error: Class 'Lfm3Config' must be a dataclass before applying @strict . PeftConfig error: Can't find 'adapter_config.json' at 'D:\aimodels\models\LiquidAILFM3-1.7B'

i got this error when i want to run :
Failed to load model: Unsloth: Failed to load model. Both AutoConfig and PeftConfig loading failed.
AutoConfig error: Class 'Lfm3Config' must be a dataclass before applying @strict .

PeftConfig error: Can't find 'adapter_config.json' at 'D:/aimodels/models/LiquidAILFM3-1.7B'

mahdisml changed discussion title from Got error when finetuning to Got error when running and finetuning about 18 hours ago

Tdamre

about 16 hours ago

did you ask sonnet 5.5?

alphabell-research

about 4 hours ago

•

edited about 4 hours ago

@mahdisml The adapter_config.json message is PEFT's fallback after AutoConfig failed; this base checkpoint is not supposed to contain an adapter config. Unsloth's LFM3 support issue is still open. First try AutoConfig.from_pretrained('LiquidAI/LFM3-1.7B', trust_remote_code=True) with plain Transformers, then share exact Unsloth and Transformers versions and the first full exception. Liquid's model card points to TRL; its linked TRL examples still show LFM2.5, so I would not copy them verbatim for LFM3.

Update from a config-only check: in clean Python 3.12 environments, plain Transformers 5.19.0 and 5.21.0 both loaded Lfm3Config with 24 layers from official revision 40ebe0aa6662c9d40389c2f5d565b6a2c6aebacd using trust_remote_code=True. This did not load weights or test training. Comparing the Hub model ID and your local directory with the same Transformers version would help isolate the failing path.

Upload images, audio, and videos by dragging in the text input, pasting, or clicking here.

Tap or paste here to upload images

· Sign up or log in to comment