Instructions to use hf-internal-testing/tiny-DeepseekV4ForCausalLM-for-CI with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use hf-internal-testing/tiny-DeepseekV4ForCausalLM-for-CI with Transformers:

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

pipe = pipeline("text-generation", model="hf-internal-testing/tiny-DeepseekV4ForCausalLM-for-CI")

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("hf-internal-testing/tiny-DeepseekV4ForCausalLM-for-CI")
model = AutoModelForCausalLM.from_pretrained("hf-internal-testing/tiny-DeepseekV4ForCausalLM-for-CI")

Notebooks
Google Colab
Kaggle
Local Apps

vLLM

How to use hf-internal-testing/tiny-DeepseekV4ForCausalLM-for-CI with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "hf-internal-testing/tiny-DeepseekV4ForCausalLM-for-CI"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "hf-internal-testing/tiny-DeepseekV4ForCausalLM-for-CI",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker

docker model run hf.co/hf-internal-testing/tiny-DeepseekV4ForCausalLM-for-CI

SGLang

How to use hf-internal-testing/tiny-DeepseekV4ForCausalLM-for-CI 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 "hf-internal-testing/tiny-DeepseekV4ForCausalLM-for-CI" \
    --host 0.0.0.0 \
    --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "hf-internal-testing/tiny-DeepseekV4ForCausalLM-for-CI",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

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 "hf-internal-testing/tiny-DeepseekV4ForCausalLM-for-CI" \
        --host 0.0.0.0 \
        --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "hf-internal-testing/tiny-DeepseekV4ForCausalLM-for-CI",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Docker Model Runner
How to use hf-internal-testing/tiny-DeepseekV4ForCausalLM-for-CI with Docker Model Runner:
```
docker model run hf.co/hf-internal-testing/tiny-DeepseekV4ForCausalLM-for-CI
```

tiny-DeepseekV4ForCausalLM-for-CI

File size: 2,499 Bytes

5ff624d

{
  "architectures": [
    "DeepseekV4ForCausalLM"
  ],
  "attention_bias": false,
  "attention_dropout": 0.0,
  "bos_token_id": 0,
  "compress_rates": {
    "compressed_sparse_attention": 4,
    "heavily_compressed_attention": 128
  },
  "compress_rope_theta": 160000,
  "dtype": "float32",
  "eos_token_id": 1,
  "expert_dtype": "fp4",
  "hc_eps": 1e-06,
  "hc_mult": 4,
  "hc_sinkhorn_iters": 20,
  "head_dim": 512,
  "hidden_act": "silu",
  "hidden_size": 4096,
  "index_head_dim": 128,
  "index_n_heads": 2,
  "index_topk": 512,
  "initializer_range": 0.02,
  "layer_types": [
    "sliding_attention",
    "sliding_attention",
    "compressed_sparse_attention",
    "heavily_compressed_attention"
  ],
  "max_position_embeddings": 1048576,
  "mlp_bias": false,
  "mlp_layer_types": [
    "hash_moe",
    "hash_moe",
    "hash_moe",
    "moe"
  ],
  "model_type": "deepseek_v4",
  "moe_intermediate_size": 2048,
  "n_routed_experts": 4,
  "n_shared_experts": 1,
  "norm_topk_prob": true,
  "num_attention_heads": 4,
  "num_experts_per_tok": 2,
  "num_hidden_layers": 4,
  "num_key_value_heads": 1,
  "num_nextn_predict_layers": 1,
  "o_groups": 8,
  "o_lora_rank": 1024,
  "output_router_logits": false,
  "pad_token_id": null,
  "partial_rotary_factor": 0.125,
  "q_lora_rank": 1024,
  "qk_rope_head_dim": 64,
  "quantization_config": {
    "activation_scheme": "dynamic",
    "fmt": "e4m3",
    "quant_method": "fp8",
    "scale_fmt": "ue8m0",
    "weight_block_size": [
      128,
      128
    ]
  },
  "rms_norm_eps": 1e-06,
  "rope_parameters": {
    "compress": {
      "beta_fast": 32,
      "beta_slow": 1,
      "factor": 16,
      "original_max_position_embeddings": 65536,
      "partial_rotary_factor": 0.125,
      "rope_theta": 160000,
      "rope_type": "yarn",
      "type": "yarn"
    },
    "main": {
      "beta_fast": 32,
      "beta_slow": 1,
      "factor": 16,
      "original_max_position_embeddings": 65536,
      "partial_rotary_factor": 0.125,
      "rope_theta": 10000,
      "rope_type": "yarn",
      "type": "yarn"
    },
    "partial_rotary_factor": 0.125,
    "rope_theta": 10000,
    "rope_type": "default"
  },
  "rope_theta": 10000,
  "routed_scaling_factor": 1.5,
  "router_aux_loss_coef": 0.001,
  "router_jitter_noise": 0.0,
  "scoring_func": "sqrtsoftplus",
  "sliding_window": 128,
  "swiglu_limit": 10.0,
  "tie_word_embeddings": false,
  "topk_method": "noaux_tc",
  "transformers_version": "5.8.0.dev0",
  "use_cache": true,
  "vocab_size": 129280
}