Instructions to use edededdy/LittleKedi-medium-wide-sparse-instruct with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Notebooks
Google Colab
Kaggle
Local Apps Settings

How to use edededdy/LittleKedi-medium-wide-sparse-instruct with llama.cpp:

Install (macOS, Linux)

curl -LsSf https://llama.app/install.sh | sh
# Start a local OpenAI-compatible server with a web UI:
llama serve -hf edededdy/LittleKedi-medium-wide-sparse-instruct:F16
# Run inference directly in the terminal:
llama cli -hf edededdy/LittleKedi-medium-wide-sparse-instruct:F16

Install from WinGet (Windows)

winget install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama serve -hf edededdy/LittleKedi-medium-wide-sparse-instruct:F16
# Run inference directly in the terminal:
llama cli -hf edededdy/LittleKedi-medium-wide-sparse-instruct:F16

Build from source code

git clone https://github.com/ggml-org/llama.cpp.git
cd llama.cpp
cmake -B build
cmake --build build -j --target llama-app
# Start a local OpenAI-compatible server with a web UI:
./build/bin/llama serve -hf edededdy/LittleKedi-medium-wide-sparse-instruct:F16
# Run inference directly in the terminal:
./build/bin/llama cli -hf edededdy/LittleKedi-medium-wide-sparse-instruct:F16

Use Docker

docker model run hf.co/edededdy/LittleKedi-medium-wide-sparse-instruct:F16

Llama new
LM Studio
Jan

vLLM

How to use edededdy/LittleKedi-medium-wide-sparse-instruct with vLLM:

Install from pip and serve model

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

Use Docker

docker model run hf.co/edededdy/LittleKedi-medium-wide-sparse-instruct:F16

Ollama
How to use edededdy/LittleKedi-medium-wide-sparse-instruct with Ollama:
```
ollama run hf.co/edededdy/LittleKedi-medium-wide-sparse-instruct:F16
```
Unsloth Desktop
Docker Model Runner
How to use edededdy/LittleKedi-medium-wide-sparse-instruct with Docker Model Runner:
```
docker model run hf.co/edededdy/LittleKedi-medium-wide-sparse-instruct:F16
```

Lemonade

How to use edededdy/LittleKedi-medium-wide-sparse-instruct with Lemonade:

Pull the model

# Download Lemonade from https://lemonade-server.ai/
lemonade pull edededdy/LittleKedi-medium-wide-sparse-instruct:F16

Run and chat with the model

lemonade run user.LittleKedi-medium-wide-sparse-instruct-F16

List all available models

lemonade list

Atomic Chat

LittleKedi-medium-wide-sparse-instruct / config.json

edededdy

Upload folder using huggingface_hub

75cf7b9 verified 4 days ago

Raw History Blame Contribute Delete

806 Bytes

{
  "vocab_size": 16384,
  "max_seq_len": 2048,
  "dim": 512,
  "n_layers": 8,
  "n_dense_layers": 2,
  "dense_hidden_dim": 1408,
  "norm_eps": 1e-06,
  "init_std": 0.02,
  "tie_embeddings": true,
  "n_heads": 8,
  "q_lora_rank": 256,
  "kv_lora_rank": 128,
  "qk_nope_head_dim": 64,
  "qk_rope_head_dim": 32,
  "v_head_dim": 64,
  "rope_theta": 10000.0,
  "rope_original_seq_len": 0,
  "rope_factor": 1.0,
  "rope_beta_fast": 32.0,
  "rope_beta_slow": 1.0,
  "rope_mscale": 1.0,
  "n_routed_experts": 240,
  "n_shared_experts": 2,
  "n_activated_experts": 8,
  "moe_hidden_dim": 128,
  "n_expert_groups": 8,
  "n_limited_groups": 4,
  "route_scale": 1.0,
  "bias_update_speed": 0.001,
  "seq_aux_loss_alpha": 0.0001,
  "moe_backend": "capacity",
  "capacity_factor": 2.0
}

	{
	"vocab_size": 16384,
	"max_seq_len": 2048,
	"dim": 512,
	"n_layers": 8,
	"n_dense_layers": 2,
	"dense_hidden_dim": 1408,
	"norm_eps": 1e-06,
	"init_std": 0.02,
	"tie_embeddings": true,
	"n_heads": 8,
	"q_lora_rank": 256,
	"kv_lora_rank": 128,
	"qk_nope_head_dim": 64,
	"qk_rope_head_dim": 32,
	"v_head_dim": 64,
	"rope_theta": 10000.0,
	"rope_original_seq_len": 0,
	"rope_factor": 1.0,
	"rope_beta_fast": 32.0,
	"rope_beta_slow": 1.0,
	"rope_mscale": 1.0,
	"n_routed_experts": 240,
	"n_shared_experts": 2,
	"n_activated_experts": 8,
	"moe_hidden_dim": 128,
	"n_expert_groups": 8,
	"n_limited_groups": 4,
	"route_scale": 1.0,
	"bias_update_speed": 0.001,
	"seq_aux_loss_alpha": 0.0001,
	"moe_backend": "capacity",
	"capacity_factor": 2.0
	}