Instructions to use AMKCode/gemma-2-2b-it-q4f32_1-MLC with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use AMKCode/gemma-2-2b-it-q4f32_1-MLC with Transformers:

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

pipe = pipeline("text-generation", model="AMKCode/gemma-2-2b-it-q4f32_1-MLC")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# pip install -U transformers accelerate
# Load model directly
from transformers import AutoModel
model = AutoModel.from_pretrained("AMKCode/gemma-2-2b-it-q4f32_1-MLC", device_map="auto")

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use AMKCode/gemma-2-2b-it-q4f32_1-MLC with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "AMKCode/gemma-2-2b-it-q4f32_1-MLC"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "AMKCode/gemma-2-2b-it-q4f32_1-MLC",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/AMKCode/gemma-2-2b-it-q4f32_1-MLC

SGLang

How to use AMKCode/gemma-2-2b-it-q4f32_1-MLC 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 "AMKCode/gemma-2-2b-it-q4f32_1-MLC" \
    --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": "AMKCode/gemma-2-2b-it-q4f32_1-MLC",
		"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 "AMKCode/gemma-2-2b-it-q4f32_1-MLC" \
        --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": "AMKCode/gemma-2-2b-it-q4f32_1-MLC",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use AMKCode/gemma-2-2b-it-q4f32_1-MLC with Docker Model Runner:
```
docker model run hf.co/AMKCode/gemma-2-2b-it-q4f32_1-MLC
```

gemma-2-2b-it-q4f32_1-MLC / mlc-chat-config.json

AMKCode

Add files using upload-large-folder tool

beec542 verified about 2 years ago

Raw History Blame

2.09 kB

	{
	"version": "0.1.0",
	"model_type": "gemma2",
	"quantization": "q4f32_1",
	"model_config": {
	"hidden_size": 2304,
	"intermediate_size": 9216,
	"attention_bias": false,
	"num_attention_heads": 8,
	"num_key_value_heads": 4,
	"head_dim": 256,
	"num_hidden_layers": 26,
	"rms_norm_eps": 1e-06,
	"vocab_size": 256000,
	"hidden_activation": "gelu_pytorch_tanh",
	"position_embedding_base": 10000.0,
	"context_window_size": 4096,
	"prefill_chunk_size": 4096,
	"tensor_parallel_shards": 1,
	"max_batch_size": 128,
	"attn_logit_softcapping": 50.0,
	"final_logit_softcapping": 30.0,
	"query_pre_attn_scalar": 256,
	"sliding_window": 4096
	},
	"vocab_size": 256000,
	"context_window_size": 4096,
	"sliding_window_size": -1,
	"prefill_chunk_size": 4096,
	"attention_sink_size": -1,
	"tensor_parallel_shards": 1,
	"pipeline_parallel_stages": 1,
	"temperature": 1.0,
	"presence_penalty": 0.0,
	"frequency_penalty": 0.0,
	"repetition_penalty": 1.0,
	"top_p": 1.0,
	"tokenizer_files": [
	"tokenizer.model",
	"tokenizer.json",
	"tokenizer_config.json"
	],
	"tokenizer_info": {
	"token_postproc_method": "byte_fallback",
	"prepend_space_in_encode": false,
	"strip_space_in_decode": false
	},
	"conv_template": {
	"name": "gemma_instruction",
	"system_template": "{system_message}",
	"system_message": "",
	"system_prefix_token_ids": [
	2
	],
	"add_role_after_system_message": true,
	"roles": {
	"user": "<start_of_turn>user",
	"assistant": "<start_of_turn>model"
	},
	"role_templates": {
	"user": "{user_message}",
	"assistant": "{assistant_message}",
	"tool": "{tool_message}"
	},
	"messages": [],
	"seps": [
	"<end_of_turn>\n"
	],
	"role_content_sep": "\n",
	"role_empty_sep": "\n",
	"stop_str": [
	"<end_of_turn>"
	],
	"stop_token_ids": [
	1,
	107
	],
	"function_string": "",
	"use_function_calling": false
	},
	"pad_token_id": 0,
	"bos_token_id": 2,
	"eos_token_id": [
	1,
	107
	]
	}