Instructions to use CortexLM/Teutonic-1-Chat-Preview with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use CortexLM/Teutonic-1-Chat-Preview with Transformers:

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

pipe = pipeline("text-generation", model="CortexLM/Teutonic-1-Chat-Preview")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# pip install -U transformers accelerate
# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("CortexLM/Teutonic-1-Chat-Preview")
model = AutoModelForCausalLM.from_pretrained("CortexLM/Teutonic-1-Chat-Preview", 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=256)
print(tokenizer.decode(outputs[0][inputs["input_ids"].shape[-1]:]))

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use CortexLM/Teutonic-1-Chat-Preview with vLLM:

Install from pip and serve model

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

Use Docker

docker model run hf.co/CortexLM/Teutonic-1-Chat-Preview

SGLang

How to use CortexLM/Teutonic-1-Chat-Preview 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 "CortexLM/Teutonic-1-Chat-Preview" \
    --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": "CortexLM/Teutonic-1-Chat-Preview",
		"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 "CortexLM/Teutonic-1-Chat-Preview" \
        --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": "CortexLM/Teutonic-1-Chat-Preview",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use CortexLM/Teutonic-1-Chat-Preview with Docker Model Runner:
```
docker model run hf.co/CortexLM/Teutonic-1-Chat-Preview
```

echoctx commited on Sep 1

Commit

a809dcd

verified ·

1 Parent(s): 7c949bb

Add fair GPQA-Diamond + LiveCodeBench vs Qwen3.5-9B (thinking protocol)

Browse files

Files changed (1) hide show

benchmarks.md +8 -1

benchmarks.md CHANGED Viewed

@@ -2,7 +2,8 @@
 Evaluated with harness `teutonic-eval-v1.2.0` **after** the thinking-budget / chat-template fix.
 **Protocol:** chat · thinking **on** · seed `20260830` · MATH-500 max_new=16k · HE+/MBPP+ max_new=8k · lm-eval `enable_thinking` + strip `</think>` · Qwen uses native `apply_chat_template`.
-**Honesty:** Prior HF Qwen column (MMLU-Pro 21 / MATH 20 / MBPP 8) was an **invalid harness artifact** — do not cite. Details: training-repo `reports/bench-harness-bug.md`.
 Head-to-head vs `Qwen/Qwen3.5-9B` on the **same fixed harness**. Vendor **card** = Qwen’s published stack (not identical protocol).
@@ -15,6 +16,8 @@ Head-to-head vs `Qwen/Qwen3.5-9B` on the **same fixed harness**. Vendor **card**
 | HumanEval+ | **82.9** | 81.1 | **+1.8** |
 | MBPP+ | **75.1** | 64.6 | **+10.5** |
 | IFEval (prompt strict) | **68.6** | 65.4 | **+3.2** |
 | RULER (4k–64k) | **89.5** | 88.4 | **+1.1** |
 | Needle | 99.5 | **100.0** | −0.5 |
@@ -33,6 +36,9 @@ Head-to-head vs `Qwen/Qwen3.5-9B` on the **same fixed harness**. Vendor **card**
 | Benchmark | Teutonic | Qwen3.5-9B (fixed harness) | Qwen3.5-9B (card) |
 |---|---:|---:|---:|
 | MMLU-Pro | 69.8 | **73.2** | 82.5 |
 ## Math & code
@@ -41,6 +47,7 @@ Head-to-head vs `Qwen/Qwen3.5-9B` on the **same fixed harness**. Vendor **card**
 | MATH-500 | **88.6** | 68.2 |
 | HumanEval+ | **82.9** | 81.1 |
 | MBPP+ | **75.1** | 64.6 |
 ## Long context

 Evaluated with harness `teutonic-eval-v1.2.0` **after** the thinking-budget / chat-template fix.
 **Protocol:** chat · thinking **on** · seed `20260830` · MATH-500 max_new=16k · HE+/MBPP+ max_new=8k · lm-eval `enable_thinking` + strip `</think>` · Qwen uses native `apply_chat_template`.
+**Honesty:** Prior HF Qwen column (MMLU-Pro 21 / MATH 20 / MBPP 8) was an **invalid harness artifact** — do not cite. Details: training-repo `reports/bench-harness-bug.md`.
+**Residual risk:** Teutonic MATH-500 still had **7.6%** `finish=length` @16k; a generous re-check @64k (`20260901T133412Z`) scored **88.0** (vs **88.6**) with **7.2%** still truncated mid-think — no material uplift, scores below kept.
 Head-to-head vs `Qwen/Qwen3.5-9B` on the **same fixed harness**. Vendor **card** = Qwen’s published stack (not identical protocol).
 | HumanEval+ | **82.9** | 81.1 | **+1.8** |
 | MBPP+ | **75.1** | 64.6 | **+10.5** |
 | IFEval (prompt strict) | **68.6** | 65.4 | **+3.2** |
+| GPQA-Diamond (fair boxed) | 46.0 | **55.1** | −9.1 |
+| LiveCodeBench (v5+v6) | 26.9 | 26.9 | 0.0 |
 | RULER (4k–64k) | **89.5** | 88.4 | **+1.1** |
 | Needle | 99.5 | **100.0** | −0.5 |
 | Benchmark | Teutonic | Qwen3.5-9B (fixed harness) | Qwen3.5-9B (card) |
 |---|---:|---:|---:|
 | MMLU-Pro | 69.8 | **73.2** | 82.5 |
+| GPQA-Diamond | 46.0† | **55.1**† | 81.7 |
+† Fair generative boxed-letter (thinking ON, 32k, think-strip, fixed option order; n=198). Not card-comparable (no official shuffle). Truncation still high (~43–52% length) — absolute likely underestimated.
 ## Math & code
 | MATH-500 | **88.6** | 68.2 |
 | HumanEval+ | **82.9** | 81.1 |
 | MBPP+ | **75.1** | 64.6 |
+| LiveCodeBench (v5+v6) | 26.9 | 26.9 |
 ## Long context