Instructions to use yusufcelebi/qwen3-4B-dynamic-layer-selected-step180-adapter with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use yusufcelebi/qwen3-4B-dynamic-layer-selected-step180-adapter with PEFT:

from peft import PeftModel
from transformers import AutoModelForCausalLM

base_model = AutoModelForCausalLM.from_pretrained("unsloth/Qwen3-4B-Base")
model = PeftModel.from_pretrained(base_model, "yusufcelebi/qwen3-4B-dynamic-layer-selected-step180-adapter")

Transformers

How to use yusufcelebi/qwen3-4B-dynamic-layer-selected-step180-adapter with Transformers:

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

pipe = pipeline("text-generation", model="yusufcelebi/qwen3-4B-dynamic-layer-selected-step180-adapter")

# Load model directly
from transformers import AutoModel
model = AutoModel.from_pretrained("yusufcelebi/qwen3-4B-dynamic-layer-selected-step180-adapter", dtype="auto")

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use yusufcelebi/qwen3-4B-dynamic-layer-selected-step180-adapter with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "yusufcelebi/qwen3-4B-dynamic-layer-selected-step180-adapter"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "yusufcelebi/qwen3-4B-dynamic-layer-selected-step180-adapter",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker

docker model run hf.co/yusufcelebi/qwen3-4B-dynamic-layer-selected-step180-adapter

SGLang

How to use yusufcelebi/qwen3-4B-dynamic-layer-selected-step180-adapter 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 "yusufcelebi/qwen3-4B-dynamic-layer-selected-step180-adapter" \
    --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": "yusufcelebi/qwen3-4B-dynamic-layer-selected-step180-adapter",
		"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 "yusufcelebi/qwen3-4B-dynamic-layer-selected-step180-adapter" \
        --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": "yusufcelebi/qwen3-4B-dynamic-layer-selected-step180-adapter",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Unsloth Studio

How to use yusufcelebi/qwen3-4B-dynamic-layer-selected-step180-adapter with Unsloth Studio:

Install Unsloth Studio (macOS, Linux, WSL)

curl -fsSL https://unsloth.ai/install.sh | sh
# Run unsloth studio
unsloth studio -H 0.0.0.0 -p 8888
# Then open http://localhost:8888 in your browser
# Search for yusufcelebi/qwen3-4B-dynamic-layer-selected-step180-adapter to start chatting

Install Unsloth Studio (Windows)

irm https://unsloth.ai/install.ps1 | iex
# Run unsloth studio
unsloth studio -H 0.0.0.0 -p 8888
# Then open http://localhost:8888 in your browser
# Search for yusufcelebi/qwen3-4B-dynamic-layer-selected-step180-adapter to start chatting

Using HuggingFace Spaces for Unsloth

# No setup required
# Open https://huggingface.co/spaces/unsloth/studio in your browser
# Search for yusufcelebi/qwen3-4B-dynamic-layer-selected-step180-adapter to start chatting

Load model with FastModel

pip install unsloth
from unsloth import FastModel
model, tokenizer = FastModel.from_pretrained(
    model_name="yusufcelebi/qwen3-4B-dynamic-layer-selected-step180-adapter",
    max_seq_length=2048,
)

Docker Model Runner
How to use yusufcelebi/qwen3-4B-dynamic-layer-selected-step180-adapter with Docker Model Runner:
```
docker model run hf.co/yusufcelebi/qwen3-4B-dynamic-layer-selected-step180-adapter
```

qwen3-4B-dynamic-layer-selected-step180-adapter / adapter_config.json

yusufcelebi

Upload folder using huggingface_hub

45d0e49 verified 5 months ago

Raw

History Blame Contribute Delete

12.3 kB

	{
	"alora_invocation_tokens": null,
	"alpha_pattern": {
	"model.layers.15.mlp.down_proj": 128,
	"model.layers.15.mlp.gate_proj": 128,
	"model.layers.15.mlp.up_proj": 128,
	"model.layers.15.self_attn.k_proj": 128,
	"model.layers.15.self_attn.o_proj": 128,
	"model.layers.15.self_attn.q_proj": 128,
	"model.layers.15.self_attn.v_proj": 128,
	"model.layers.16.mlp.down_proj": 32,
	"model.layers.16.mlp.gate_proj": 32,
	"model.layers.16.mlp.up_proj": 32,
	"model.layers.16.self_attn.k_proj": 32,
	"model.layers.16.self_attn.o_proj": 32,
	"model.layers.16.self_attn.q_proj": 32,
	"model.layers.16.self_attn.v_proj": 32,
	"model.layers.18.mlp.down_proj": 32,
	"model.layers.18.mlp.gate_proj": 32,
	"model.layers.18.mlp.up_proj": 32,
	"model.layers.18.self_attn.k_proj": 32,
	"model.layers.18.self_attn.o_proj": 32,
	"model.layers.18.self_attn.q_proj": 32,
	"model.layers.18.self_attn.v_proj": 32,
	"model.layers.20.mlp.down_proj": 64,
	"model.layers.20.mlp.gate_proj": 64,
	"model.layers.20.mlp.up_proj": 64,
	"model.layers.20.self_attn.k_proj": 64,
	"model.layers.20.self_attn.o_proj": 64,
	"model.layers.20.self_attn.q_proj": 64,
	"model.layers.20.self_attn.v_proj": 64,
	"model.layers.23.mlp.down_proj": 32,
	"model.layers.23.mlp.gate_proj": 32,
	"model.layers.23.mlp.up_proj": 32,
	"model.layers.23.self_attn.k_proj": 32,
	"model.layers.23.self_attn.o_proj": 32,
	"model.layers.23.self_attn.q_proj": 32,
	"model.layers.23.self_attn.v_proj": 32,
	"model.layers.24.mlp.down_proj": 64,
	"model.layers.24.mlp.gate_proj": 64,
	"model.layers.24.mlp.up_proj": 64,
	"model.layers.24.self_attn.k_proj": 64,
	"model.layers.24.self_attn.o_proj": 64,
	"model.layers.24.self_attn.q_proj": 64,
	"model.layers.24.self_attn.v_proj": 64,
	"model.layers.25.mlp.down_proj": 32,
	"model.layers.25.mlp.gate_proj": 32,
	"model.layers.25.mlp.up_proj": 32,
	"model.layers.25.self_attn.k_proj": 32,
	"model.layers.25.self_attn.o_proj": 32,
	"model.layers.25.self_attn.q_proj": 32,
	"model.layers.25.self_attn.v_proj": 32,
	"model.layers.26.mlp.down_proj": 128,
	"model.layers.26.mlp.gate_proj": 128,
	"model.layers.26.mlp.up_proj": 128,
	"model.layers.26.self_attn.k_proj": 128,
	"model.layers.26.self_attn.o_proj": 128,
	"model.layers.26.self_attn.q_proj": 128,
	"model.layers.26.self_attn.v_proj": 128,
	"model.layers.29.mlp.down_proj": 64,
	"model.layers.29.mlp.gate_proj": 64,
	"model.layers.29.mlp.up_proj": 64,
	"model.layers.29.self_attn.k_proj": 64,
	"model.layers.29.self_attn.o_proj": 64,
	"model.layers.29.self_attn.q_proj": 64,
	"model.layers.29.self_attn.v_proj": 64,
	"model.layers.32.mlp.down_proj": 64,
	"model.layers.32.mlp.gate_proj": 64,
	"model.layers.32.mlp.up_proj": 64,
	"model.layers.32.self_attn.k_proj": 64,
	"model.layers.32.self_attn.o_proj": 64,
	"model.layers.32.self_attn.q_proj": 64,
	"model.layers.32.self_attn.v_proj": 64,
	"model.layers.33.mlp.down_proj": 128,
	"model.layers.33.mlp.gate_proj": 128,
	"model.layers.33.mlp.up_proj": 128,
	"model.layers.33.self_attn.k_proj": 128,
	"model.layers.33.self_attn.o_proj": 128,
	"model.layers.33.self_attn.q_proj": 128,
	"model.layers.33.self_attn.v_proj": 128,
	"model.layers.7.mlp.down_proj": 128,
	"model.layers.7.mlp.gate_proj": 128,
	"model.layers.7.mlp.up_proj": 128,
	"model.layers.7.self_attn.k_proj": 128,
	"model.layers.7.self_attn.o_proj": 128,
	"model.layers.7.self_attn.q_proj": 128,
	"model.layers.7.self_attn.v_proj": 128,
	"model.layers.8.mlp.down_proj": 32,
	"model.layers.8.mlp.gate_proj": 32,
	"model.layers.8.mlp.up_proj": 32,
	"model.layers.8.self_attn.k_proj": 32,
	"model.layers.8.self_attn.o_proj": 32,
	"model.layers.8.self_attn.q_proj": 32,
	"model.layers.8.self_attn.v_proj": 32
	},
	"arrow_config": null,
	"auto_mapping": {
	"base_model_class": "Qwen3ForCausalLM",
	"parent_library": "transformers.models.qwen3.modeling_qwen3",
	"unsloth_fixed": true
	},
	"base_model_name_or_path": "unsloth/Qwen3-4B-Base",
	"bias": "none",
	"corda_config": null,
	"ensure_weight_tying": false,
	"eva_config": null,
	"exclude_modules": null,
	"fan_in_fan_out": false,
	"inference_mode": true,
	"init_lora_weights": true,
	"layer_replication": null,
	"layers_pattern": null,
	"layers_to_transform": null,
	"loftq_config": {},
	"lora_alpha": 64,
	"lora_bias": false,
	"lora_dropout": 0.0,
	"megatron_config": null,
	"megatron_core": "megatron.core",
	"modules_to_save": null,
	"peft_type": "LORA",
	"peft_version": "0.18.1",
	"qalora_group_size": 16,
	"r": 32,
	"rank_pattern": {
	"model.layers.15.mlp.down_proj": 64,
	"model.layers.15.mlp.gate_proj": 64,
	"model.layers.15.mlp.up_proj": 64,
	"model.layers.15.self_attn.k_proj": 64,
	"model.layers.15.self_attn.o_proj": 64,
	"model.layers.15.self_attn.q_proj": 64,
	"model.layers.15.self_attn.v_proj": 64,
	"model.layers.16.mlp.down_proj": 16,
	"model.layers.16.mlp.gate_proj": 16,
	"model.layers.16.mlp.up_proj": 16,
	"model.layers.16.self_attn.k_proj": 16,
	"model.layers.16.self_attn.o_proj": 16,
	"model.layers.16.self_attn.q_proj": 16,
	"model.layers.16.self_attn.v_proj": 16,
	"model.layers.18.mlp.down_proj": 16,
	"model.layers.18.mlp.gate_proj": 16,
	"model.layers.18.mlp.up_proj": 16,
	"model.layers.18.self_attn.k_proj": 16,
	"model.layers.18.self_attn.o_proj": 16,
	"model.layers.18.self_attn.q_proj": 16,
	"model.layers.18.self_attn.v_proj": 16,
	"model.layers.20.mlp.down_proj": 32,
	"model.layers.20.mlp.gate_proj": 32,
	"model.layers.20.mlp.up_proj": 32,
	"model.layers.20.self_attn.k_proj": 32,
	"model.layers.20.self_attn.o_proj": 32,
	"model.layers.20.self_attn.q_proj": 32,
	"model.layers.20.self_attn.v_proj": 32,
	"model.layers.23.mlp.down_proj": 16,
	"model.layers.23.mlp.gate_proj": 16,
	"model.layers.23.mlp.up_proj": 16,
	"model.layers.23.self_attn.k_proj": 16,
	"model.layers.23.self_attn.o_proj": 16,
	"model.layers.23.self_attn.q_proj": 16,
	"model.layers.23.self_attn.v_proj": 16,
	"model.layers.24.mlp.down_proj": 32,
	"model.layers.24.mlp.gate_proj": 32,
	"model.layers.24.mlp.up_proj": 32,
	"model.layers.24.self_attn.k_proj": 32,
	"model.layers.24.self_attn.o_proj": 32,
	"model.layers.24.self_attn.q_proj": 32,
	"model.layers.24.self_attn.v_proj": 32,
	"model.layers.25.mlp.down_proj": 16,
	"model.layers.25.mlp.gate_proj": 16,
	"model.layers.25.mlp.up_proj": 16,
	"model.layers.25.self_attn.k_proj": 16,
	"model.layers.25.self_attn.o_proj": 16,
	"model.layers.25.self_attn.q_proj": 16,
	"model.layers.25.self_attn.v_proj": 16,
	"model.layers.26.mlp.down_proj": 64,
	"model.layers.26.mlp.gate_proj": 64,
	"model.layers.26.mlp.up_proj": 64,
	"model.layers.26.self_attn.k_proj": 64,
	"model.layers.26.self_attn.o_proj": 64,
	"model.layers.26.self_attn.q_proj": 64,
	"model.layers.26.self_attn.v_proj": 64,
	"model.layers.29.mlp.down_proj": 32,
	"model.layers.29.mlp.gate_proj": 32,
	"model.layers.29.mlp.up_proj": 32,
	"model.layers.29.self_attn.k_proj": 32,
	"model.layers.29.self_attn.o_proj": 32,
	"model.layers.29.self_attn.q_proj": 32,
	"model.layers.29.self_attn.v_proj": 32,
	"model.layers.32.mlp.down_proj": 32,
	"model.layers.32.mlp.gate_proj": 32,
	"model.layers.32.mlp.up_proj": 32,
	"model.layers.32.self_attn.k_proj": 32,
	"model.layers.32.self_attn.o_proj": 32,
	"model.layers.32.self_attn.q_proj": 32,
	"model.layers.32.self_attn.v_proj": 32,
	"model.layers.33.mlp.down_proj": 64,
	"model.layers.33.mlp.gate_proj": 64,
	"model.layers.33.mlp.up_proj": 64,
	"model.layers.33.self_attn.k_proj": 64,
	"model.layers.33.self_attn.o_proj": 64,
	"model.layers.33.self_attn.q_proj": 64,
	"model.layers.33.self_attn.v_proj": 64,
	"model.layers.7.mlp.down_proj": 64,
	"model.layers.7.mlp.gate_proj": 64,
	"model.layers.7.mlp.up_proj": 64,
	"model.layers.7.self_attn.k_proj": 64,
	"model.layers.7.self_attn.o_proj": 64,
	"model.layers.7.self_attn.q_proj": 64,
	"model.layers.7.self_attn.v_proj": 64,
	"model.layers.8.mlp.down_proj": 16,
	"model.layers.8.mlp.gate_proj": 16,
	"model.layers.8.mlp.up_proj": 16,
	"model.layers.8.self_attn.k_proj": 16,
	"model.layers.8.self_attn.o_proj": 16,
	"model.layers.8.self_attn.q_proj": 16,
	"model.layers.8.self_attn.v_proj": 16
	},
	"revision": null,
	"target_modules": [
	"model.layers.33.self_attn.v_proj",
	"model.layers.20.mlp.gate_proj",
	"model.layers.29.mlp.down_proj",
	"model.layers.16.mlp.gate_proj",
	"model.layers.8.self_attn.q_proj",
	"model.layers.24.self_attn.q_proj",
	"model.layers.15.self_attn.k_proj",
	"model.layers.26.mlp.gate_proj",
	"model.layers.7.mlp.up_proj",
	"model.layers.18.self_attn.o_proj",
	"model.layers.24.self_attn.k_proj",
	"model.layers.16.self_attn.v_proj",
	"model.layers.20.mlp.down_proj",
	"model.layers.24.mlp.up_proj",
	"model.layers.15.mlp.gate_proj",
	"model.layers.16.mlp.down_proj",
	"model.layers.32.mlp.up_proj",
	"model.layers.20.self_attn.o_proj",
	"model.layers.8.self_attn.k_proj",
	"model.layers.18.self_attn.q_proj",
	"model.layers.15.self_attn.o_proj",
	"model.layers.24.self_attn.v_proj",
	"model.layers.29.mlp.gate_proj",
	"model.layers.26.self_attn.o_proj",
	"model.layers.8.mlp.up_proj",
	"model.layers.20.self_attn.k_proj",
	"model.layers.33.self_attn.k_proj",
	"model.layers.18.mlp.gate_proj",
	"model.layers.7.self_attn.k_proj",
	"model.layers.26.self_attn.k_proj",
	"model.layers.33.mlp.down_proj",
	"model.layers.29.self_attn.k_proj",
	"model.layers.16.self_attn.k_proj",
	"model.layers.23.self_attn.k_proj",
	"model.layers.7.mlp.down_proj",
	"model.layers.16.self_attn.q_proj",
	"model.layers.25.self_attn.k_proj",
	"model.layers.29.mlp.up_proj",
	"model.layers.18.self_attn.k_proj",
	"model.layers.29.self_attn.v_proj",
	"model.layers.7.mlp.gate_proj",
	"model.layers.23.mlp.down_proj",
	"model.layers.24.self_attn.o_proj",
	"model.layers.26.mlp.down_proj",
	"model.layers.25.self_attn.q_proj",
	"model.layers.16.self_attn.o_proj",
	"model.layers.24.mlp.down_proj",
	"model.layers.15.mlp.down_proj",
	"model.layers.7.self_attn.q_proj",
	"model.layers.32.self_attn.q_proj",
	"model.layers.23.self_attn.v_proj",
	"model.layers.26.mlp.up_proj",
	"model.layers.8.mlp.gate_proj",
	"model.layers.15.self_attn.q_proj",
	"model.layers.20.self_attn.v_proj",
	"model.layers.33.self_attn.o_proj",
	"model.layers.32.self_attn.v_proj",
	"model.layers.33.mlp.up_proj",
	"model.layers.32.mlp.down_proj",
	"model.layers.26.self_attn.v_proj",
	"model.layers.23.self_attn.q_proj",
	"model.layers.25.self_attn.v_proj",
	"model.layers.32.self_attn.k_proj",
	"model.layers.16.mlp.up_proj",
	"model.layers.33.self_attn.q_proj",
	"model.layers.8.self_attn.v_proj",
	"model.layers.23.mlp.gate_proj",
	"model.layers.8.mlp.down_proj",
	"model.layers.8.self_attn.o_proj",
	"model.layers.25.self_attn.o_proj",
	"model.layers.32.mlp.gate_proj",
	"model.layers.20.self_attn.q_proj",
	"model.layers.26.self_attn.q_proj",
	"model.layers.18.self_attn.v_proj",
	"model.layers.7.self_attn.v_proj",
	"model.layers.23.self_attn.o_proj",
	"model.layers.25.mlp.down_proj",
	"model.layers.25.mlp.up_proj",
	"model.layers.24.mlp.gate_proj",
	"model.layers.33.mlp.gate_proj",
	"model.layers.32.self_attn.o_proj",
	"model.layers.7.self_attn.o_proj",
	"model.layers.20.mlp.up_proj",
	"model.layers.29.self_attn.q_proj",
	"model.layers.15.mlp.up_proj",
	"model.layers.23.mlp.up_proj",
	"model.layers.15.self_attn.v_proj",
	"model.layers.25.mlp.gate_proj",
	"model.layers.29.self_attn.o_proj",
	"model.layers.18.mlp.down_proj",
	"model.layers.18.mlp.up_proj"
	],
	"target_parameters": null,
	"task_type": "CAUSAL_LM",
	"trainable_token_indices": null,
	"use_dora": false,
	"use_qalora": false,
	"use_rslora": false
	}