Instructions to use jpacifico/Chocolatine-2-4B-Instruct-DPO-v2.1 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use jpacifico/Chocolatine-2-4B-Instruct-DPO-v2.1 with Transformers:

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

pipe = pipeline("text-generation", model="jpacifico/Chocolatine-2-4B-Instruct-DPO-v2.1")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForMultimodalLM

tokenizer = AutoTokenizer.from_pretrained("jpacifico/Chocolatine-2-4B-Instruct-DPO-v2.1")
model = AutoModelForMultimodalLM.from_pretrained("jpacifico/Chocolatine-2-4B-Instruct-DPO-v2.1")
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]:]))

Inference
Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use jpacifico/Chocolatine-2-4B-Instruct-DPO-v2.1 with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "jpacifico/Chocolatine-2-4B-Instruct-DPO-v2.1"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "jpacifico/Chocolatine-2-4B-Instruct-DPO-v2.1",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/jpacifico/Chocolatine-2-4B-Instruct-DPO-v2.1

SGLang

How to use jpacifico/Chocolatine-2-4B-Instruct-DPO-v2.1 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 "jpacifico/Chocolatine-2-4B-Instruct-DPO-v2.1" \
    --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": "jpacifico/Chocolatine-2-4B-Instruct-DPO-v2.1",
		"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 "jpacifico/Chocolatine-2-4B-Instruct-DPO-v2.1" \
        --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": "jpacifico/Chocolatine-2-4B-Instruct-DPO-v2.1",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use jpacifico/Chocolatine-2-4B-Instruct-DPO-v2.1 with Docker Model Runner:
```
docker model run hf.co/jpacifico/Chocolatine-2-4B-Instruct-DPO-v2.1
```

jpacifico commited on Mar 26

Commit

f192fd7

verified ·

1 Parent(s): 01cf917

Update README.md

Browse files

Files changed (1) hide show

README.md +10 -7

README.md CHANGED Viewed

@@ -29,11 +29,14 @@ Optimized variants (MLX, GGUF) are also available, making the model particularly
 ## Model Overview
 - **Base model:** Qwen/Qwen3-4B-Instruct-2507 (non-thinking)
-- **Architecture:** Decoder-only transformer
 - **Parameters:** 4.0B
-- **Window Context :** 256K long-context understanding
 - **Post training methods:** DPO + Model Merging
-- **Objective:** Improve French instruction-following and reasoning through post-training, while maintaining strong multilingual performance and keeping the same model size.
 **Model Variants**
@@ -45,7 +48,7 @@ Optimized variants (MLX, GGUF) are also available, making the model particularly
 Performance improves consistently across all tested FR benchmarks :
-| Benchmark | Qwen3-4B-Thinking-2507 | Chocolatine-2-4B-Instruct-DPO-v2.1 |
 |---|---:|---:|
 | french_bench_arc_challenge | 47.13 | **49.79** |
 | french_bench_grammar | 70.59 | **72.27** |
@@ -63,9 +66,9 @@ Chocolatine-2-4B-Instruct-DPO-v2.1 is derived from [Qwen/Qwen3-4B-Instruct-2507]
 **Stage 1 – DPO (Compar:IA adaptation)**
-Direct Preference Optimization (DPO) on a DPO-adapted version of **[Compar:IA](https://comparia.beta.gouv.fr/datasets)** data, derived from the preference dataset [comparia-votes](https://huggingface.co/datasets/ministere-culture/comparia-votes) of the French government initiative.
-The dataset was restructured into preference pairs and curated to build an original dataset specifically designed for DPO fine-tuning.
-Two dataset variants were constructed ([6k](https://huggingface.co/datasets/jpacifico/comparia-dpo-pairs-bt-6k) and [13k](jpacifico/comparia-dpo-pairs-bt-13k) preference pairs).
 The **6k variant** was used for the DPO training reported in this release.
 **Stage 2 – DPO (French-ORCA pairs)**

 ## Model Overview
 - **Base model:** Qwen/Qwen3-4B-Instruct-2507 (non-thinking)
 - **Parameters:** 4.0B
+- **Context Length:** 262,144 natively
 - **Post training methods:** DPO + Model Merging
+Note: This model supports only non-thinking mode and does not generate `<think></think>` blocks in its outputs.
+The choice of Qwen3 over Qwen3.5 is intentional: Qwen3 provides a more direct text-oriented generation style, better aligned with the objectives of this post-training. Empirically, my post-training on Qwen3 yielded *more consistent improvements across evaluated benchmarks* compared to Qwen3.5.
+For use cases requiring explicit reasoning traces or structured thinking outputs, Qwen3.5 (thinking mode) is recommended.
+Chocolatine-2-4B-Instruct-DPO-v2.1 is therefore optimized for direct instruction-following and practical deployment, rather than explicit CoT generation.
 **Model Variants**
 Performance improves consistently across all tested FR benchmarks :
+| Benchmark | Qwen3-4B-Thinking-2507 (base) | Chocolatine-2-4B-Instruct-DPO-v2.1 |
 |---|---:|---:|
 | french_bench_arc_challenge | 47.13 | **49.79** |
 | french_bench_grammar | 70.59 | **72.27** |
 **Stage 1 – DPO (Compar:IA adaptation)**
+Direct Preference Optimization (DPO) on a DPO-adapted version of **[Compar:IA](https://comparia.beta.gouv.fr/datasets)** data, derived from the preference dataset [comparia-votes](https://huggingface.co/datasets/ministere-culture/comparia-votes), part of a public initiative led by the Ministry of Culture (French gov). Previous iterations of the Chocolatine model series also were selected as part of this initiative.
+I constructed an original DPO dataset from these votes by transforming them into preference pairs (chosen / rejected), with additional filtering and formatting steps to make them suitable for DPO fine-tuning.
+Two dataset variants were created ([6k](https://huggingface.co/datasets/jpacifico/comparia-dpo-pairs-bt-6k) and [13k](https://huggingface.co/datasets/jpacifico/comparia-dpo-pairs-bt-13k) preference pairs).
 The **6k variant** was used for the DPO training reported in this release.
 **Stage 2 – DPO (French-ORCA pairs)**