Instructions to use dhanesh-hf/Jarvis-Titan-M4-Activated with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use dhanesh-hf/Jarvis-Titan-M4-Activated with Transformers:

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

pipe = pipeline("text-generation", model="dhanesh-hf/Jarvis-Titan-M4-Activated", trust_remote_code=True)
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoModelForCausalLM
model = AutoModelForCausalLM.from_pretrained("dhanesh-hf/Jarvis-Titan-M4-Activated", trust_remote_code=True, device_map="auto")

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use dhanesh-hf/Jarvis-Titan-M4-Activated with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "dhanesh-hf/Jarvis-Titan-M4-Activated"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "dhanesh-hf/Jarvis-Titan-M4-Activated",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/dhanesh-hf/Jarvis-Titan-M4-Activated

SGLang

How to use dhanesh-hf/Jarvis-Titan-M4-Activated 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 "dhanesh-hf/Jarvis-Titan-M4-Activated" \
    --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": "dhanesh-hf/Jarvis-Titan-M4-Activated",
		"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 "dhanesh-hf/Jarvis-Titan-M4-Activated" \
        --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": "dhanesh-hf/Jarvis-Titan-M4-Activated",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use dhanesh-hf/Jarvis-Titan-M4-Activated with Docker Model Runner:
```
docker model run hf.co/dhanesh-hf/Jarvis-Titan-M4-Activated
```

Jarvis-Titan-M4-Activated / config.json

dhanesh-hf

Sync config.json from Phase 0 base

2ac4b1f verified 13 days ago

Raw History Blame

2.91 kB

	{
	"transformers_version": "5.12.1",
	"architectures": [
	"JarvisTitanMoEForCausalLM"
	],
	"output_hidden_states": false,
	"return_dict": true,
	"dtype": "bfloat16",
	"chunk_size_feed_forward": 0,
	"is_encoder_decoder": false,
	"id2label": {
	"0": "LABEL_0",
	"1": "LABEL_1"
	},
	"label2id": {
	"LABEL_0": 0,
	"LABEL_1": 1
	},
	"problem_type": null,
	"vocab_size": 152064,
	"hidden_size": 3584,
	"intermediate_size": 18944,
	"num_hidden_layers": 28,
	"num_attention_heads": 28,
	"num_key_value_heads": 4,
	"hidden_act": "silu",
	"max_position_embeddings": 32768,
	"initializer_range": 0.02,
	"rms_norm_eps": 1e-06,
	"use_cache": true,
	"tie_word_embeddings": false,
	"rope_parameters": {
	"rope_theta": 1000000.0,
	"rope_type": "default"
	},
	"use_sliding_window": false,
	"sliding_window": 2048,
	"max_window_layers": 28,
	"layer_types": [
	"full_attention",
	"full_attention",
	"full_attention",
	"full_attention",
	"full_attention",
	"full_attention",
	"full_attention",
	"full_attention",
	"full_attention",
	"full_attention",
	"full_attention",
	"full_attention",
	"full_attention",
	"full_attention",
	"full_attention",
	"full_attention",
	"full_attention",
	"full_attention",
	"full_attention",
	"full_attention",
	"full_attention",
	"full_attention",
	"full_attention",
	"full_attention",
	"full_attention",
	"full_attention",
	"full_attention",
	"full_attention"
	],
	"attention_dropout": 0.0,
	"pad_token_id": null,
	"bos_token_id": 151643,
	"eos_token_id": 151645,
	"_name_or_path": "dhanesh-hf/Jarvis-Titan-V12-SFT-Merged",
	"model_type": "jarvis_titan_moe",
	"output_attentions": false,
	"num_routed_experts": 8,
	"num_shared_experts": 1,
	"num_experts_per_tok": 2,
	"expert_intermediate_size": 4736,
	"routed_scaling_factor": 1.5,
	"auto_map": {
	"AutoConfig": "modeling_jarvis_titan_moe.JarvisTitanMoEConfig",
	"AutoModelForCausalLM": "modeling_jarvis_titan_moe.JarvisTitanMoEForCausalLM"
	},
	"m4_calibrated": true,
	"architecture_milestone": "M4",
	"injection_layers": [
	3,
	7,
	11,
	15,
	19,
	23,
	27
	],
	"memory_dim": 512,
	"d_reservoir": 512,
	"sliding_window_size": 2048,
	"tribrid_target_distribution": [
	0.55,
	0.25,
	0.2
	],
	"upstream_m4_adapter": "dhanesh-hf/Jarvis-Titan-M4-Calibrated-Adapter",
	"base_backbone": "dhanesh-hf/Jarvis-Titan-V14-MoE-Merged",
	"m4_architecture": "tribrid_dha3_mtp",
	"dha3_holographic_decoupling": true,
	"dha3_anchor_interval": 4,
	"csa3_delta_caching": true,
	"csa3_anchor_interval": 4,
	"m4_bridge_layers": [
	3,
	7,
	11,
	15,
	19,
	23,
	27
	],
	"dynamic_mag_gating": true,
	"dynamic_mag_soft_floor": 0.05,
	"fim_regularization_lambda": 0.005,
	"mtp_num_targets": 1,
	"license": "other",
	"license_name": "jtrl-v1.0"
	}