Instructions to use vivekchakraverty/qwen3-1.7b-brandforge with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use vivekchakraverty/qwen3-1.7b-brandforge with Transformers:

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

pipe = pipeline("text-generation", model="vivekchakraverty/qwen3-1.7b-brandforge")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("vivekchakraverty/qwen3-1.7b-brandforge")
model = AutoModelForCausalLM.from_pretrained("vivekchakraverty/qwen3-1.7b-brandforge", device_map="auto")
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 Settings

vLLM

How to use vivekchakraverty/qwen3-1.7b-brandforge with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "vivekchakraverty/qwen3-1.7b-brandforge"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "vivekchakraverty/qwen3-1.7b-brandforge",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/vivekchakraverty/qwen3-1.7b-brandforge

SGLang

How to use vivekchakraverty/qwen3-1.7b-brandforge 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 "vivekchakraverty/qwen3-1.7b-brandforge" \
    --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": "vivekchakraverty/qwen3-1.7b-brandforge",
		"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 "vivekchakraverty/qwen3-1.7b-brandforge" \
        --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": "vivekchakraverty/qwen3-1.7b-brandforge",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use vivekchakraverty/qwen3-1.7b-brandforge with Docker Model Runner:
```
docker model run hf.co/vivekchakraverty/qwen3-1.7b-brandforge
```

Access BrandForge Qwen3-1.7B

These are the weights behind Brand Studio in Mr. AI Marketer. Access is granted automatically once you accept below — there is no waiting on a manual review.

This model writes brand-strategy deliverables. It is a small (1.7B) model and its output is a starting draft for a human strategist, not finished client work. Please do not present its output as professional brand consultancy without review.

BrandForge — Qwen3-1.7B

Qwen3-1.7B fine-tuned to write the individual sections of a brand strategy document, then merged back into standalone weights (no adapter needed).

This is the model behind Brand Studio in Mr. AI Marketer. It is published so that anyone running the app can serve it on their own GPU rather than a shared backend.

What it does

Given a structured brand intake (name, audience, archetype, competitors, personality sliders, channels, …) it writes one deliverable at a time from a fixed set of 12:

Phase	Sections
Strategy	Brand Core, Positioning Statement, Audience Personas, Competitive Frame, Archetype Expression
Verbal Identity	Brand Personality & Voice, Messaging Pillars, Tagline Territory
Visual Identity	Visual Direction Brief
Experience	Channel Voice Adaptation, Brand Guardrails
Growth	90-Day Activation Checklist

Prompt format

It was trained on one exact prompt shape, and it degrades noticeably if you depart from it. Use the chat template with thinking disabled.

System message:

You are a senior brand strategist writing one deliverable for a client's Brand Document.
Be concrete, specific, and free of marketing filler. Output clean markdown for a single
deliverable (do not repeat the whole document). Never invent brand facts beyond what the
intake provides — if information needed for a claim is missing, write 'TBD — needs founder
input' instead of inventing it.

User message:

DELIVERABLE: {section name}

INSTRUCTIONS:
{section instructions}

BRAND INTAKE (ground truth — do not contradict, do not invent beyond this):
{intake as JSON, indent=2}

Write only the "{section name}" deliverable now, in markdown.

import torch
from transformers import AutoModelForCausalLM, AutoTokenizer

model_id = "vivekchakraverty/qwen3-1.7b-brandforge"
tok = AutoTokenizer.from_pretrained(model_id)
model = AutoModelForCausalLM.from_pretrained(model_id, dtype=torch.float16).to("cuda")

prompt = tok.apply_chat_template(
    messages, tokenize=False, add_generation_prompt=True, enable_thinking=False
)
inputs = tok(prompt, return_tensors="pt").to("cuda")
out = model.generate(**inputs, max_new_tokens=1500, do_sample=True, temperature=0.7, top_p=0.9)

Qwen3 emits an empty <think></think> block even with thinking disabled — strip it before showing the text to a user.

Training

Supervised fine-tuning followed by DPO, on brand-document sections produced by a larger teacher model and then put through a critique-and-revise pass and a judge filter. Retrieval was deliberately discarded at training time: the branding knowledge is meant to live in the weights so that inference needs nothing but the prompt above.

Limitations

It is 1.7B. It writes to a consistent house structure, not with the judgement of a senior strategist. Treat every section as a first draft.
It only knows these 12 sections in this prompt format. It is not a general marketing assistant.
It was trained to write TBD — needs founder input rather than invent facts, but it can still overreach — check any claim that reads like a statistic, certification, or customer result.
English only.

License

Apache-2.0, inherited from Qwen3-1.7B.

Downloads last month: 10

Safetensors

Model size

2B params

Tensor type

F16

Model tree for vivekchakraverty/qwen3-1.7b-brandforge

Base model

Qwen/Qwen3-1.7B-Base

Finetuned

Qwen/Qwen3-1.7B

Finetuned

(1041)

this model