Instructions to use ByteDance-Seed/Cola2 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use ByteDance-Seed/Cola2 with Transformers:

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

pipe = pipeline("text-generation", model="ByteDance-Seed/Cola2", trust_remote_code=True)
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# pip install -U transformers accelerate
# Load model directly
from transformers import AutoModel
model = AutoModel.from_pretrained("ByteDance-Seed/Cola2", trust_remote_code=True, device_map="auto")

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use ByteDance-Seed/Cola2 with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "ByteDance-Seed/Cola2"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "ByteDance-Seed/Cola2",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/ByteDance-Seed/Cola2

SGLang

How to use ByteDance-Seed/Cola2 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 "ByteDance-Seed/Cola2" \
    --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": "ByteDance-Seed/Cola2",
		"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 "ByteDance-Seed/Cola2" \
        --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": "ByteDance-Seed/Cola2",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use ByteDance-Seed/Cola2 with Docker Model Runner:
```
docker model run hf.co/ByteDance-Seed/Cola2
```

Cola2

File size: 2,163 Bytes

a310630

{
  "architectures": [
    "Cola2ForConditionalGeneration"
  ],
  "attention_dropout": 0.0,
  "auto_map": {
    "AutoConfig": "configuration_cola2.Cola2Config",
    "AutoModel": "modeling_cola2.Cola2ForConditionalGeneration",
    "AutoModelForCausalLM": "modeling_cola2.Cola2ForConditionalGeneration"
  },
  "bos_token_id": 151643,
  "boundary_score_alpha": 10.0,
  "boundary_score_threshold": 0.5,
  "cola2_format_version": 1,
  "condition_num_attention_heads": 20,
  "condition_rope_theta": 500000.0,
  "decoder_condition_num_hidden_layers": 2,
  "decoder_head_dim": 128,
  "decoder_hidden_size": 2560,
  "decoder_intermediate_size": 9728,
  "decoder_lm_head_bias": true,
  "decoder_num_attention_heads": 32,
  "decoder_num_hidden_layers": 36,
  "decoder_num_key_value_heads": 8,
  "decoder_rope_theta": 1000000.0,
  "diffusion_time": 1000.0,
  "dit_block_size": 8,
  "dit_head_dim": 128,
  "dit_hidden_size": 2560,
  "dit_intermediate_size": 10240,
  "dit_layer_norm_eps": 1e-05,
  "dit_num_attention_heads": 20,
  "dit_num_hidden_layers": 34,
  "dit_rope_dim": 96,
  "dit_rope_theta": 10000.0,
  "duration_hidden_size": 1280,
  "encoder_embedding_size": 2560,
  "encoder_hidden_size": 1536,
  "encoder_intermediate_size": 6144,
  "encoder_num_attention_heads": 12,
  "encoder_num_hidden_layers": 0,
  "encoder_rope_theta": 500000.0,
  "eos_token_id": 151643,
  "hidden_act": "silu",
  "hidden_dropout": 0.0,
  "im_end_token_id": 151645,
  "im_start_token_id": 151644,
  "initializer_range": 0.02,
  "latent_block_causal": false,
  "latent_block_size": 8,
  "latent_decoder_num_hidden_layers": 0,
  "latent_dim": 64,
  "latent_encoder_num_hidden_layers": 2,
  "layer_norm_eps": 1e-06,
  "local_position_dim": 128,
  "mask_token_id": 151936,
  "max_position_embeddings": 32768,
  "max_span_length": 8,
  "min_span_length": 1,
  "model_type": "cola2",
  "pad_token_id": 151936,
  "pool_num_hidden_layers": 2,
  "preset": "cola2",
  "rope_theta": 1000000.0,
  "scaling_factor": 1.0,
  "shifting_factor": 0.0,
  "target_compression_rate": 0.5,
  "tie_word_embeddings": false,
  "torch_dtype": "float32",
  "use_variational_encoder": true,
  "vocab_size": 151936
}