Instructions to use valleprata-experience/valleprata-one with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use valleprata-experience/valleprata-one with Transformers:

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

pipe = pipeline("text-generation", model="valleprata-experience/valleprata-one")

# pip install -U transformers accelerate
# Load model directly
from transformers import AutoModel
model = AutoModel.from_pretrained("valleprata-experience/valleprata-one", device_map="auto")

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use valleprata-experience/valleprata-one with vLLM:

Install from pip and serve model

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

Use Docker

docker model run hf.co/valleprata-experience/valleprata-one

SGLang

How to use valleprata-experience/valleprata-one 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 "valleprata-experience/valleprata-one" \
    --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": "valleprata-experience/valleprata-one",
		"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 "valleprata-experience/valleprata-one" \
        --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": "valleprata-experience/valleprata-one",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Docker Model Runner
How to use valleprata-experience/valleprata-one with Docker Model Runner:
```
docker model run hf.co/valleprata-experience/valleprata-one
```

Qwen3.5-4B-RAG-Specialist

Qwen3.5-4B-RAG-Specialist is a post-trained version of Qwen/Qwen3.5-4B, fine-tuned specifically for Retrieval-Augmented Generation (RAG) pipelines.

The model is optimized for:

Precise answer extraction strictly grounded in the provided context passages.
Hallucination mitigation (explicit refusal to answer when the required information is absent from the context).
Accurate source attribution/citation and resilience against noisy or irrelevant distractor documents.

Model Details

Developed by: [Your Name / Organization]
Model Type: Causal Language Model
Base Model: Qwen/Qwen3.5-4B
Parameters: ~4 Billion
Context Length: Supports up to 32k/128k tokens (inherited from the Qwen base architecture)
Language(s): English, Italian (multilingual capabilities inherited from base)
License: Apache-2.0

Intended Use & Best Practices

Intended Use

Retrieval-Augmented Generation (RAG) in enterprise or local deployments.
Question Answering over internal documentation, manuals, and reports.
Synthesis of heterogeneous document contexts containing distractor/irrelevant passages.

Recommended Prompt Template

The model performs best when prompts adhere to the standard Qwen ChatML format:

<|im_start|>system
You are an assistant specialized in RAG. Answer the user's question based EXCLUSIVELY on the passages provided in [Context]. If the information is missing or insufficient, state clearly that you cannot answer.<|im_end|>
<|im_start|>user
[Context]
Document [1]: Project Chronos was initiated in May 2024 to modernize the internal storage network.
Document [2]: Peripheral server maintenance takes place on the first Tuesday of every month.

[Question]
When did Project Chronos start?
[Answer]<|im_end|>
<|im_start|>assistant

Downloads last month: -; Downloads are not tracked for this model. How to track

Model tree for valleprata-experience/valleprata-one

Base model

Qwen/Qwen3.5-4B-Base

Finetuned

Qwen/Qwen3.5-4B

Finetuned

(1007)

this model

Collection including valleprata-experience/valleprata-one

valleprata-one-1.0

Collection

1 item • Updated 5 days ago