Instructions to use CloudGoat/Baphomet-1.1-4B with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use CloudGoat/Baphomet-1.1-4B with Transformers:

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

pipe = pipeline("image-text-to-text", model="CloudGoat/Baphomet-1.1-4B")
messages = [
    {
        "role": "user",
        "content": [
            {"type": "image", "url": "https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/p-blog/candy.JPG"},
            {"type": "text", "text": "What animal is on the candy?"}
        ]
    },
]
pipe(text=messages)

# pip install -U transformers accelerate
# Load model directly
from transformers import AutoProcessor, AutoModelForMultimodalLM

processor = AutoProcessor.from_pretrained("CloudGoat/Baphomet-1.1-4B")
model = AutoModelForMultimodalLM.from_pretrained("CloudGoat/Baphomet-1.1-4B", device_map="auto")
messages = [
    {
        "role": "user",
        "content": [
            {"type": "image", "url": "https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/p-blog/candy.JPG"},
            {"type": "text", "text": "What animal is on the candy?"}
        ]
    },
]
inputs = processor.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=256)
print(processor.decode(outputs[0][inputs["input_ids"].shape[-1]:]))

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use CloudGoat/Baphomet-1.1-4B with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "CloudGoat/Baphomet-1.1-4B"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "CloudGoat/Baphomet-1.1-4B",
		"messages": [
			{
				"role": "user",
				"content": [
					{
						"type": "text",
						"text": "Describe this image in one sentence."
					},
					{
						"type": "image_url",
						"image_url": {
							"url": "https://cdn.britannica.com/61/93061-050-99147DCE/Statue-of-Liberty-Island-New-York-Bay.jpg"
						}
					}
				]
			}
		]
	}'

Use Docker

docker model run hf.co/CloudGoat/Baphomet-1.1-4B

SGLang

How to use CloudGoat/Baphomet-1.1-4B 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 "CloudGoat/Baphomet-1.1-4B" \
    --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": "CloudGoat/Baphomet-1.1-4B",
		"messages": [
			{
				"role": "user",
				"content": [
					{
						"type": "text",
						"text": "Describe this image in one sentence."
					},
					{
						"type": "image_url",
						"image_url": {
							"url": "https://cdn.britannica.com/61/93061-050-99147DCE/Statue-of-Liberty-Island-New-York-Bay.jpg"
						}
					}
				]
			}
		]
	}'

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 "CloudGoat/Baphomet-1.1-4B" \
        --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": "CloudGoat/Baphomet-1.1-4B",
		"messages": [
			{
				"role": "user",
				"content": [
					{
						"type": "text",
						"text": "Describe this image in one sentence."
					},
					{
						"type": "image_url",
						"image_url": {
							"url": "https://cdn.britannica.com/61/93061-050-99147DCE/Statue-of-Liberty-Island-New-York-Bay.jpg"
						}
					}
				]
			}
		]
	}'

Docker Model Runner
How to use CloudGoat/Baphomet-1.1-4B with Docker Model Runner:
```
docker model run hf.co/CloudGoat/Baphomet-1.1-4B
```
Browse Quantizations to use this model in llama.cpp, Ollama, LM Studio, or any compatible app.

Baphomet-1.1-4B

Linear merge of 8 task-vectors on Qwen3.5-4B, with per-donor coefficients found by an evolutionary search (EvoGM)

Base	Method	Task-Vectors	Output
`Qwen/Qwen3.5-4B`	`linear`, λ searched with EvoGM	8 vectors, searched λ (Σλ ≈ 1.139)	`safetensors ×2` (bf16)

日本語: Baphomet-1.0 と同じ 8 個のタスクベクトルを、EvoGM の進化的探索で求めた係数 λ で線形マージしたモデルです。

1. Overview

Baphomet-1.1-4B is the successor of Baphomet-1.0-4B. It uses the same base model and the same 8 donor task-vectors (T = W_donor − W_base), but instead of one uniform weight (0.1375 for every donor in 1.0), each donor gets its own coefficient λᵢ found by an evolutionary search.

W_merged = W_base + Σ λᵢ · Tᵢ

On a 180-item paired test (MMLU-Pro / GPQA-Diamond / BFCL non-live, 8192-token thinking cap), Baphomet-1.1 scored 79.4%, against 77.8% for Baphomet-1.0 and 70.0% for Qwen3.5-4B. The gain over Qwen3.5-4B is statistically significant. The gain over Baphomet-1.0 is not (see Section 6). It also thinks shorter: mean output 1,834 tokens vs 3,180 for 1.0.

2. Merge configuration

2.1 Coefficients (λ, donor order as searched)

#	Donor	λ
1	`InternScience/Agents-A1-4B`	0.011589
2	`Jackrong/DeepSeek-V4-Pro-Qwen3.5-4B`	0.203923
3	`empero-ai/Qwen3.8-4B-Distill`	0.590137
4	`Jackrong/Qwopus3.5-4B-Coder`	0.001337
5	`Jackrong/Qwopus3.5-4B-v3`	0.297193
6	`BAAI/AREX-Turbo`	0.032303
7	`TokenRhythm/NeoHorse-1-4B`	0.002979
8	`allenai/tmax-4b`	−0.000742

Most of the weight sits on three donors: Qwen3.8-4B-Distill (0.59), Qwopus3.5-4B-v3 (0.30) and DeepSeek-V4-Pro-Qwen3.5-4B (0.20). Full-precision values (float32):

[0.011588579043745995, 0.203922837972641, 0.5901365876197815, 0.0013365253107622266,
 0.2971934676170349, 0.032303329557180405, 0.002979394979774952, -0.0007424817304126918]

2.2 How the weights are computed

Tᵢ = W_donorᵢ − W_base in fp32, per tensor. Each task-vector is stored as fp8 e4m3 with a per-row absmax scale (scale = rowmax|T| / 448). Tensors where a donor equals the base exactly are skipped.
W = W_base + Σ λᵢ · dequant(Tᵢ), accumulated in fp32 in donor order, then cast back to each base tensor's dtype (bf16; the A_log / norm tensors the base stores in fp32 stay fp32).
A tensor a donor does not have contributes 0. Agents-A1-4B and AREX-Turbo have no MTP tensors. NeoHorse-1-4B and tmax-4b are text-only. NeoHorse's keys were remapped model.* → model.language_model.*.
Key names, shapes, dtypes and the 2-shard layout are identical to Qwen/Qwen3.5-4B. Tokenizer, chat template and processor configs are copied unchanged from the base.

The fp8 task-vector storage changes the merged bf16 weights by at most about 1 bf16 ulp in some elements, compared with an exact fp32 merge. These are the same weights that were evaluated. Note that vLLM loaded them as fp16 for evaluation.

2.3 Search (EvoGM)

The coefficients were searched with a port of EvoGM ("Learning to Merge LLMs via Evolutionary Generative Optimization", ICML 2026, github.com/JiangTao97/evogm). It uses dual MLP generators trained on winner/loser splits of the population, λᵢ ∈ [−1, 1], and basis-shift rounds.

Round 1: 40 candidates, each scored on a 110-item fitness set (40 MMLU-Pro + 30 GPQA + 40 BFCL non-live). Accuracy is computed per task and then averaged. Generation used thinking mode with a 4096-token cap, and truncated answers count as wrong.
Rounds 2–3 (IRT-accelerated): 64 more candidates were scored on a 40-item subset chosen for Fisher information, with the rest of the 110-item score estimated by a 2PL IRT model. Clearly weak candidates were stopped early. The best estimates were re-scored on all 110 items, and the winner is chosen by the exact 110-item score.
Fitness (110 items, 4096 cap): Baphomet-1.0 λ 0.603 → best of round 1 0.722 → round 2 0.733 → Baphomet-1.1 0.764.
The fitness items are disjoint from the test items in Section 6.
Compute: about 13.7 kernel-hours on Kaggle T4×2 (round 1: 7.1 h, rounds 2–3: 6.5 h), using vLLM 0.33.0.

3. Model specifications

Same architecture as the base (excerpt from config.json).

Item	Value
Architecture	`Qwen3_5ForConditionalGeneration`
Text layers	32 layers, hidden size `2560`, heads `16Q / 4KV (GQA)`, `head_dim 256`
Attention layer types	Repeating pattern of 3 `linear_attention` layers followed by 1 `full_attention` layer
FFN	`intermediate 9216`, `SiLU`, `tie_word_embeddings: true`
MTP	1 layer (`mtp_num_hidden_layers: 1`)
Vocabulary size	`248,320` (`Qwen2Tokenizer`, BPE)
Context length	`262,144`
Dtype	`bfloat16` (fp32 for `A_log` / norm tensors, as in the base)

4. Included files

File	Description
`model.safetensors-00001-of-00002.safetensors`, `model.safetensors-00002-of-00002.safetensors`	Model weights (bf16, 9.3 GB total), same layout as the base
`model.safetensors.index.json`	Shard index
`config.json`	Model configuration (base)
`generation_config.json`	Default sampling as used in evaluation
`tokenizer.json` / `vocab.json` / `merges.txt` / `tokenizer_config.json` / `chat_template.jinja`	Tokenizer and chat template (base)
`preprocessor_config.json` / `video_preprocessor_config.json`	Processor configuration (base)

5. Usage

5.1 vLLM

vllm serve CloudGoat/Baphomet-1.1-4B --max-model-len 32768

5.2 Transformers

from transformers import AutoModelForImageTextToText, AutoProcessor

model_id = "CloudGoat/Baphomet-1.1-4B"
model = AutoModelForImageTextToText.from_pretrained(model_id, dtype="auto", device_map="auto")
processor = AutoProcessor.from_pretrained(model_id)

messages = [{"role": "user", "content": [{"type": "text", "text": "Explain task arithmetic in two sentences."}]}]
inputs = processor.apply_chat_template(messages, add_generation_prompt=True, tokenize=True,
                                       return_dict=True, return_tensors="pt").to(model.device)
out = model.generate(**inputs, max_new_tokens=4096, do_sample=True, temperature=0.6, top_p=0.95, top_k=20)
print(processor.decode(out[0][inputs["input_ids"].shape[1]:], skip_special_tokens=False))

5.3 Recommended sampling

Every evaluation below used thinking mode (enable_thinking=True, the default) with temperature=0.6, top_p=0.95, top_k=20. This matches the Qwen3.5 recommendation for thinking mode on precise tasks. For general chat, the Qwen3.5 card also suggests temperature=1.0, top_p=0.95, top_k=20, presence_penalty=1.5. That setting was not evaluated for this merge.

6. Benchmark

Model	MMLU-Pro (n=80)	GPQA-Diamond (n=50)	BFCL non-live (n=50)	Average
CloudGoat/Baphomet-1.1-4B	81.2%	62.0%	94.0%	79.1%
CloudGoat/Baphomet-1.0-4B	77.5%	58.0%	98.0%	77.8%
Jackrong/Qwopus3.5-4B-v3	75.0%	60.0%	94.0%	76.3%
empero-ai/Qwen3.8-4B-Distill	75.0%	52.0%	96.0%	74.3%
allenai/tmax-4b	76.2%	50.0%	96.0%	74.1%
TokenRhythm/NeoHorse-1-4B	75.0%	42.0%	98.0%	71.7%
BAAI/AREX-Turbo ¹	71.2%	44.0%	98.0%	71.1%
Qwen/Qwen3.5-4B	71.2%	42.0%	96.0%	69.8%
Jackrong/DeepSeek-V4-Pro-Qwen3.5-4B	72.5%	50.0%	84.0%	68.8%
InternScience/Agents-A1-4B	63.8%	46.0%	90.0%	66.6%
Jackrong/Qwopus3.5-4B-Coder	66.2%	28.0%	96.0%	63.4%

Average = unweighted mean of the three benchmark accuracies. All models were run with the same 180 items, prompts and settings described below.

¹ AREX-Turbo does not emit <think> tags, so the standard scorer (which reads the answer after </think>) gives it 36/6/49. The numbers shown re-parse the final answer from the full output.

180-item paired test: the same items, prompts and settings for all three models. The items are MMLU-Pro (n=80), GPQA-Diamond (n=50) and BFCL non-live AST (n=50). Generation used thinking mode with an 8192-token cap and temperature=0.6, top_p=0.95, top_k=20, seed=0, with one sample per item. Truncated answers count as wrong. Results come from vLLM 0.33.0 on T4 (fp16).

Model	Overall	MMLU-Pro	GPQA-Diamond	BFCL non-live	Truncated	Mean output tokens
Qwen/Qwen3.5-4B	70.0%	71.3%	42.0%	96.0%	22.8%	3,983
Baphomet-1.0-4B	77.8%	77.5%	58.0%	98.0%	15.6%	3,180
Baphomet-1.1-4B	79.4%	81.2%	62.0%	94.0%	10.0%	1,834

Paired exact McNemar tests (overall, 180 items):

Comparison	1.1 right / other wrong	1.1 wrong / other right	p
vs Baphomet-1.0-4B	10	7	0.63 (not significant)
vs Qwen3.5-4B	20	3	< 0.001

Caveats

Small test. With 180 items, one item is worth 0.56 points. The +1.7-point difference vs Baphomet-1.0 is within noise. The only robust claim is the improvement over the base model.
Single sample at temperature 0.6. Scores from one seeded sample per item vary between runs.
Winner's curse. λ was chosen as the best of more than 100 candidates on a 110-item fitness set. The fitness score (0.764) is therefore optimistic. The held-out test items in this table were not used in the search.
BFCL dropped slightly (94% vs 98% for 1.0; 2 items).
None of the donors changes the vision tower or the MTP layer, so those tensors are identical to Qwen3.5-4B. Image input was not evaluated.

7. Limitations

The model is the result of a linear merge, and there is no guarantee that characteristics of individual donor models are preserved.
It may inherit the license terms and biases of each donor and the base model. Determine the scope of use after reviewing the license of each model.
Only the three benchmarks above were measured.
Use for final decisions in high-risk domains such as medical or legal matters is not intended.

8. Credits

Role	Model / work
Base	Qwen/Qwen3.5-4B
Donors	InternScience/Agents-A1-4B, Jackrong/DeepSeek-V4-Pro-Qwen3.5-4B, empero-ai/Qwen3.8-4B-Distill, Jackrong/Qwopus3.5-4B-Coder, Jackrong/Qwopus3.5-4B-v3, BAAI/AREX-Turbo, TokenRhythm/NeoHorse-1-4B, allenai/tmax-4b
Predecessor	CloudGoat/Baphomet-1.0-4B
Search method	EvoGM: Learning to Merge LLMs via Evolutionary Generative Optimization (ICML 2026), github.com/JiangTao97/evogm
References	O. K. Hitit, C. Sankar, Y. K. Kim, A. Risteski, and A. Elgohary, "A systematic study of in-the-wild model merging for large language models," arXiv:2511.21437, 2025.