This model was released on 2025-06-18 and added to Hugging Face Transformers on 2025-06-24.

</div>

Arcee

Arcee is a decoder-only transformer model based on the Llama architecture with a key modification: it uses ReLU² (ReLU-squared) activation in the MLP blocks instead of SiLU, following recent research showing improved training efficiency with squared activations. This architecture is designed for efficient training and inference while maintaining the proven stability of the Llama design.

The Arcee model is architecturally similar to Llama but uses x * relu(x) in MLP layers for improved gradient flow and is optimized for efficiency in both training and inference scenarios.

[!TIP] The Arcee model supports extended context with RoPE scaling and all standard transformers features including Flash Attention 2, SDPA, gradient checkpointing, and quantization support.

The example below demonstrates how to generate text with Arcee using [Pipeline] or the [AutoModel].

from transformers import pipeline


pipeline = pipeline(
    task="text-generation",
    model="arcee-ai/AFM-4.5B",
    device=0
)

output = pipeline("The key innovation in Arcee is")
print(output[0]["generated_text"])

import torch

from transformers import ArceeForCausalLM, AutoTokenizer


tokenizer = AutoTokenizer.from_pretrained("arcee-ai/AFM-4.5B")
model = ArceeForCausalLM.from_pretrained(
    "arcee-ai/AFM-4.5B",
    device_map="auto"
)

inputs = tokenizer("The key innovation in Arcee is", return_tensors="pt").to(model.device)
with torch.no_grad():
    outputs = model.generate(**inputs, max_new_tokens=50)
print(tokenizer.decode(outputs[0], skip_special_tokens=True))

ArceeConfig

[[autodoc]] ArceeConfig

ArceeModel

[[autodoc]] ArceeModel - forward

ArceeForCausalLM

[[autodoc]] ArceeForCausalLM - forward

ArceeForSequenceClassification

[[autodoc]] ArceeForSequenceClassification - forward

ArceeForQuestionAnswering

[[autodoc]] ArceeForQuestionAnswering - forward

ArceeForTokenClassification

[[autodoc]] ArceeForTokenClassification - forward