Instructions to use Brucamian/70-qwen70-24-R with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use Brucamian/70-qwen70-24-R with Transformers:

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

pipe = pipeline("text-generation", model="Brucamian/70-qwen70-24-R")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoModelForCausalLM
model = AutoModelForCausalLM.from_pretrained("Brucamian/70-qwen70-24-R", dtype="auto")

Notebooks
Google Colab
Kaggle
Local Apps

vLLM

How to use Brucamian/70-qwen70-24-R with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "Brucamian/70-qwen70-24-R"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "Brucamian/70-qwen70-24-R",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/Brucamian/70-qwen70-24-R

SGLang

How to use Brucamian/70-qwen70-24-R 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 "Brucamian/70-qwen70-24-R" \
    --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": "Brucamian/70-qwen70-24-R",
		"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 "Brucamian/70-qwen70-24-R" \
        --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": "Brucamian/70-qwen70-24-R",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use Brucamian/70-qwen70-24-R with Docker Model Runner:
```
docker model run hf.co/Brucamian/70-qwen70-24-R
```

70-qwen70-24-R

File size: 2,341 Bytes

159aca2

{
  "_name_or_path": "/data/cambrian-u/ckpt/20251012_mix70960k_SPMD_7b_finetune_diffloss_token256_dit2_4_cfg0_1_context512_ep1_v6-rescue",
  "architectures": [
    "CambrianQwenForCausalLM"
  ],
  "attention_dropout": 0.0,
  "aux_regression": false,
  "aux_regression_coef": 1.0,
  "bos_token_id": 151643,
  "connector_only": true,
  "ddt_encoder_depth": 2,
  "diff_head_lr": 0.000565,
  "diffusion_base_dim": null,
  "diffusion_class_dropout_prob": 0.1,
  "diffusion_model_channels": 1152,
  "diffusion_model_depth": 32,
  "diffusion_model_heads": 32,
  "diffusion_model_hidden_size": 2048,
  "diffusion_model_z_channels": 2048,
  "diffusion_norm_stats_path": null,
  "diffusion_split_per_token": 256,
  "diffusion_timesteps_per_sample": 1,
  "eos_token_id": 151645,
  "freeze_mm_mlp_adapter": false,
  "hidden_act": "silu",
  "hidden_size": 3584,
  "image_aspect_ratio": "square",
  "image_position": 35,
  "initializer_range": 0.02,
  "intermediate_size": 18944,
  "max_position_embeddings": 32768,
  "max_window_layers": 28,
  "miv_token_len": 0,
  "mm_hidden_size": 1152,
  "mm_projector_lr": null,
  "mm_projector_type": "mlp2x_gelu",
  "mm_use_im_patch_token": false,
  "mm_use_im_start_end": true,
  "mm_vision_sampler_lr": null,
  "mm_vision_select_feature": "patch",
  "mm_vision_select_layer": -1,
  "mm_vision_tower_aux_list": [
    "google/siglip2-so400m-patch14-224"
  ],
  "mm_vision_tower_aux_token_len_list": [
    256
  ],
  "mm_vision_tower_lr": 2e-06,
  "model_type": "cambrian_qwen",
  "num_attention_heads": 28,
  "num_hidden_layers": 28,
  "num_key_value_heads": 4,
  "pretrain_adapter_and_vision_head": null,
  "rms_norm_eps": 1e-06,
  "rope_scaling": null,
  "rope_theta": 1000000.0,
  "si_token_len": 729,
  "sliding_window": 131072,
  "tie_word_embeddings": false,
  "tokenizer_model_max_length": 512,
  "tokenizer_padding_side": "right",
  "torch_dtype": "bfloat16",
  "transformers_version": "4.37.0",
  "tune_adapter_and_vision_head": false,
  "tune_mm_mlp_adapter": false,
  "tune_vision_head": false,
  "unfreeze_mm_vision_tower": false,
  "use_cache": false,
  "use_mm_proj": true,
  "use_sliding_window": false,
  "vision_coef": 2.0,
  "vision_hidden_size": 1152,
  "vision_loss": "diffusion-loss",
  "vision_loss_mode": "query",
  "vision_tower_aux_token_len_list": [
    256
  ],
  "vocab_size": 151667
}