Instructions to use GrizleeBer/gs-test-2 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use GrizleeBer/gs-test-2 with Transformers:

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

pipe = pipeline("text-generation", model="GrizleeBer/gs-test-2")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoModelForCausalLM
model = AutoModelForCausalLM.from_pretrained("GrizleeBer/gs-test-2", dtype="auto")

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use GrizleeBer/gs-test-2 with vLLM:

Install from pip and serve model

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

Use Docker

docker model run hf.co/GrizleeBer/gs-test-2

SGLang

How to use GrizleeBer/gs-test-2 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 "GrizleeBer/gs-test-2" \
    --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": "GrizleeBer/gs-test-2",
		"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 "GrizleeBer/gs-test-2" \
        --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": "GrizleeBer/gs-test-2",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use GrizleeBer/gs-test-2 with Docker Model Runner:
```
docker model run hf.co/GrizleeBer/gs-test-2
```

gs-test-2 / config.json

GrizleeBer

Upload folder using huggingface_hub

211db0a verified about 1 month ago

raw

history blame contribute delete

4.54 kB

	{
	"adapter_names": [
	"citations",
	"query_rewrite",
	"query_clarification",
	"context_relevance",
	"hallucination_detection",
	"answerability",
	"uncertainty",
	"requirement-check",
	"context-attribution",
	"factuality-detection",
	"policy-guardrails",
	"factuality-correction",
	"guardian-core"
	],
	"adapter_ranks": [
	16,
	32,
	32,
	16,
	16,
	16,
	32,
	64,
	16,
	32,
	16,
	32,
	32
	],
	"adapter_third_party": [
	"citations",
	"query_rewrite",
	"query_clarification",
	"context_relevance",
	"hallucination_detection",
	"answerability",
	"uncertainty",
	"requirement-check",
	"context-attribution",
	"factuality-detection",
	"policy-guardrails",
	"factuality-correction",
	"guardian-core"
	],
	"adapter_token_ids": [
	100352,
	100353,
	100354,
	100355,
	100356,
	100357,
	100358,
	100359,
	100360,
	100361,
	100362,
	100363,
	100364
	],
	"architectures": [
	"GraniteSwitchForCausalLM"
	],
	"attention_bias": false,
	"attention_dropout": 0.0,
	"attention_multiplier": 0.015625,
	"bos_token_id": 100257,
	"control_dims": 32,
	"control_token_gain": 15.0,
	"dtype": "bfloat16",
	"embedding_multiplier": 12.0,
	"eos_token_id": 100257,
	"fused_add_norm": false,
	"hidden_act": "silu",
	"hidden_size": 2560,
	"hiding_groups": {
	"all_controls": [
	"citations",
	"query_rewrite",
	"query_clarification",
	"context_relevance",
	"hallucination_detection",
	"answerability",
	"uncertainty",
	"requirement-check",
	"context-attribution",
	"factuality-detection",
	"policy-guardrails",
	"factuality-correction",
	"guardian-core"
	]
	},
	"hiding_policy": {
	"answerability": [
	"all_controls"
	],
	"base": [
	"all_controls"
	],
	"citations": [
	"all_controls"
	],
	"context-attribution": [
	"all_controls"
	],
	"context_relevance": [
	"all_controls"
	],
	"factuality-correction": [
	"all_controls"
	],
	"factuality-detection": [
	"all_controls"
	],
	"guardian-core": [
	"all_controls"
	],
	"hallucination_detection": [
	"all_controls"
	],
	"policy-guardrails": [
	"all_controls"
	],
	"query_clarification": [
	"all_controls"
	],
	"query_rewrite": [
	"all_controls"
	],
	"requirement-check": [
	"all_controls"
	],
	"uncertainty": [
	"all_controls"
	]
	},
	"initializer_range": 0.1,
	"intermediate_size": 8192,
	"layer_types": [
	"attention",
	"attention",
	"attention",
	"attention",
	"attention",
	"attention",
	"attention",
	"attention",
	"attention",
	"attention",
	"attention",
	"attention",
	"attention",
	"attention",
	"attention",
	"attention",
	"attention",
	"attention",
	"attention",
	"attention",
	"attention",
	"attention",
	"attention",
	"attention",
	"attention",
	"attention",
	"attention",
	"attention",
	"attention",
	"attention",
	"attention",
	"attention",
	"attention",
	"attention",
	"attention",
	"attention",
	"attention",
	"attention",
	"attention",
	"attention",
	"attention"
	],
	"logits_scaling": 10.0,
	"lora_target_modules": [
	"o_proj",
	"qkv_proj",
	"shared_input_linear",
	"shared_output_linear"
	],
	"mamba_chunk_size": 256,
	"mamba_conv_bias": true,
	"mamba_d_conv": 4,
	"mamba_d_head": 40,
	"mamba_d_state": 256,
	"mamba_expand": 2,
	"mamba_n_groups": 1,
	"mamba_n_heads": 128,
	"mamba_proj_bias": false,
	"max_lora_rank": 64,
	"max_position_embeddings": 131072,
	"model_type": "granite_switch",
	"num_adapters": 13,
	"num_attention_heads": 40,
	"num_experts_per_tok": 2,
	"num_hidden_layers": 41,
	"num_key_value_heads": 8,
	"num_local_experts": 0,
	"output_router_logits": false,
	"pad_token_id": 100256,
	"position_embedding_type": "rope",
	"projection_head_dim": 64,
	"residual_multiplier": 0.22,
	"rms_norm_eps": 1e-05,
	"rope_parameters": {
	"rope_theta": 10000000,
	"rope_type": "default"
	},
	"router_aux_loss_coef": 0.001,
	"shared_intermediate_size": 8192,
	"switch_head_dim": 32,
	"tie_word_embeddings": true,
	"time_step_limit": [
	0.0,
	{
	"__float__": "Infinity"
	}
	],
	"time_step_max": 0.1,
	"time_step_min": 0.001,
	"transformers_version": "5.6.2",
	"use_cache": true,
	"vocab_size": 100365
	}