Instructions to use CultriX/MergeStage1v3 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use CultriX/MergeStage1v3 with Transformers:

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

pipe = pipeline("text-generation", model="CultriX/MergeStage1v3")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("CultriX/MergeStage1v3")
model = AutoModelForCausalLM.from_pretrained("CultriX/MergeStage1v3")
messages = [
    {"role": "user", "content": "Who are you?"},
]
inputs = tokenizer.apply_chat_template(
	messages,
	add_generation_prompt=True,
	tokenize=True,
	return_dict=True,
	return_tensors="pt",
).to(model.device)

outputs = model.generate(**inputs, max_new_tokens=40)
print(tokenizer.decode(outputs[0][inputs["input_ids"].shape[-1]:]))

Notebooks
Google Colab
Kaggle
Local Apps

vLLM

How to use CultriX/MergeStage1v3 with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "CultriX/MergeStage1v3"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "CultriX/MergeStage1v3",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/CultriX/MergeStage1v3

SGLang

How to use CultriX/MergeStage1v3 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 "CultriX/MergeStage1v3" \
    --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": "CultriX/MergeStage1v3",
		"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 "CultriX/MergeStage1v3" \
        --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": "CultriX/MergeStage1v3",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use CultriX/MergeStage1v3 with Docker Model Runner:
```
docker model run hf.co/CultriX/MergeStage1v3
```

Improve language tag

by lbourdois - opened Apr 28, 2025

base: refs/heads/main

←

from: refs/pr/2

Discussion Files changed

+61

-48

Files changed (1) hide show

README.md +61 -48

README.md CHANGED Viewed

@@ -1,48 +1,61 @@
----
-base_model:
-- Qwen/Qwen2.5-14B-Instruct
-- Qwen/Qwen2.5-14B
-- CultriX/Qwen2.5-14B_Virtuoso-small-v2-LoRA_r128
-library_name: transformers
-tags:
-- mergekit
-- merge
----
-# MergeStage3v1
-This is a merge of pre-trained language models created using [mergekit](https://github.com/cg123/mergekit).
-## Merge Details
-### Merge Method
-This model was merged using the [DARE TIES](https://arxiv.org/abs/2311.03099) merge method using [Qwen/Qwen2.5-14B](https://huggingface.co/Qwen/Qwen2.5-14B) as a base.
-### Models Merged
-The following models were included in the merge:
-* [Qwen/Qwen2.5-14B-Instruct](https://huggingface.co/Qwen/Qwen2.5-14B-Instruct)
-* [CultriX/Qwen2.5-14B_Virtuoso-small-v2-LoRA_r128](https://huggingface.co/CultriX/Qwen2.5-14B_Virtuoso-small-v2-LoRA_r128)
-### Configuration
-The following YAML configuration was used to produce this model:
-```yaml
-name: Virtuoso-Infused-Qwen
-merge_method: dare_ties
-base_model: Qwen/Qwen2.5-14B # New Base Model
-tokenizer_source: base
-dtype: float32
-out_dtype: bfloat16
-models:
-  - model: Qwen/Qwen2.5-14B-Instruct
-    parameters:
-      weight: 1.0
-      density: 1.0
-  - model: CultriX/Qwen2.5-14B_Virtuoso-small-v2-LoRA_r128
-    parameters:
-      weight: 1.0
-      density: 1.0
-```

+---
+base_model:
+- Qwen/Qwen2.5-14B-Instruct
+- Qwen/Qwen2.5-14B
+- CultriX/Qwen2.5-14B_Virtuoso-small-v2-LoRA_r128
+library_name: transformers
+tags:
+- mergekit
+- merge
+language:
+- zho
+- eng
+- fra
+- spa
+- por
+- deu
+- ita
+- rus
+- jpn
+- kor
+- vie
+- tha
+- ara
+---
+# MergeStage3v1
+This is a merge of pre-trained language models created using [mergekit](https://github.com/cg123/mergekit).
+## Merge Details
+### Merge Method
+This model was merged using the [DARE TIES](https://arxiv.org/abs/2311.03099) merge method using [Qwen/Qwen2.5-14B](https://huggingface.co/Qwen/Qwen2.5-14B) as a base.
+### Models Merged
+The following models were included in the merge:
+* [Qwen/Qwen2.5-14B-Instruct](https://huggingface.co/Qwen/Qwen2.5-14B-Instruct)
+* [CultriX/Qwen2.5-14B_Virtuoso-small-v2-LoRA_r128](https://huggingface.co/CultriX/Qwen2.5-14B_Virtuoso-small-v2-LoRA_r128)
+### Configuration
+The following YAML configuration was used to produce this model:
+```yaml
+name: Virtuoso-Infused-Qwen
+merge_method: dare_ties
+base_model: Qwen/Qwen2.5-14B # New Base Model
+tokenizer_source: base
+dtype: float32
+out_dtype: bfloat16
+models:
+  - model: Qwen/Qwen2.5-14B-Instruct
+    parameters:
+      weight: 1.0
+      density: 1.0
+  - model: CultriX/Qwen2.5-14B_Virtuoso-small-v2-LoRA_r128
+    parameters:
+      weight: 1.0
+      density: 1.0
+```