Instructions to use SZLHOLDINGS/chaski with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use SZLHOLDINGS/chaski with Transformers:

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

pipe = pipeline("text-generation", model="SZLHOLDINGS/chaski")
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)

# Load model directly
from transformers import AutoProcessor, AutoModelForMultimodalLM

processor = AutoProcessor.from_pretrained("SZLHOLDINGS/chaski")
model = AutoModelForMultimodalLM.from_pretrained("SZLHOLDINGS/chaski", 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=40)
print(processor.decode(outputs[0][inputs["input_ids"].shape[-1]:]))

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use SZLHOLDINGS/chaski with vLLM:

Install from pip and serve model

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

Use Docker

docker model run hf.co/SZLHOLDINGS/chaski

SGLang

How to use SZLHOLDINGS/chaski 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 "SZLHOLDINGS/chaski" \
    --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": "SZLHOLDINGS/chaski",
		"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 "SZLHOLDINGS/chaski" \
        --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": "SZLHOLDINGS/chaski",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use SZLHOLDINGS/chaski with Docker Model Runner:
```
docker model run hf.co/SZLHOLDINGS/chaski
```

betterwithage commited on 1 day ago

Commit

bef153f

verified ·

1 Parent(s): 5be5fac

Remove Khipu MEASURED pin from Chaski; Jobs RUNNING only

Browse files

Files changed (1) hide show

README.md +11 -26

README.md CHANGED Viewed

@@ -20,22 +20,11 @@ szl:
   artifact_class: ADAPTER
   originality: FINETUNE_DISCLOSED_BASE
   collection: "SZL Fall 2026 — Original Cuts"
-  demo_space: https://huggingface.co/spaces/SZLHOLDINGS/szl-model-inference-lab
   jobs: RUNNING
   job_id: 6a91b8ba984507d9db4ea071
   job_namespace: SZLHOLDINGS
 ---
-<!-- SZL-ESTATE-CARD:v2:START -->
-<p align="center"><a href="https://a-11-oy.com/"><img src="https://huggingface.co/spaces/SZLHOLDINGS/README/resolve/main/assets/estate-banner-v2.svg" alt="SZL Holdings — governed, receipted, verifiable" width="100%"></a></p>
-<p align="center">
-  <a href="https://github.com/szl-holdings/.github/tree/main/doctrine"><img src="https://img.shields.io/badge/doctrine-v11%20LOCKED-0B1F3A?style=flat-square" alt="doctrine v11"></a>
-  <a href="https://huggingface.co/spaces/SZLHOLDINGS/szl-model-inference-lab"><img src="https://img.shields.io/badge/demo-lab%20LIVE%20%C2%B7%20Chaski%20not%20loaded-3AF4C8?style=flat-square" alt="demo Space"></a>
-  <a href="https://a-11-oy.com/"><img src="https://img.shields.io/badge/evidence%20wall-verify%20in%20browser-C9B787?style=flat-square" alt="evidence wall"></a>
-</p>
-<p align="center"><sub>Part of the <a href="https://huggingface.co/SZLHOLDINGS">SZL Holdings</a> governed estate. Verification proves integrity and origin, never accuracy or performance. Λ = Conjecture 1.</sub></p>
-<!-- SZL-ESTATE-CARD:v2:END -->
 # Chaski
 **One line.** Messenger LLM. Proposal-only drafts and honest refusals for the SZL controller.
@@ -46,13 +35,15 @@ szl:
 | **Originality** | SZL fine-tune of a disclosed Apache Qwen instruct base |
 | **Base** | `Qwen/Qwen3.5-0.8B` (Apache-2.0, 0.6B–2B lock) |
 | **License** | `apache-2.0` |
-| **House lab** | [szl-model-inference-lab](https://huggingface.co/spaces/SZLHOLDINGS/szl-model-inference-lab) — LIVE on Khipu GGUF. **This repo is not loaded there.** |
-| **HF Jobs** | RUNNING `6a91b8ba984507d9db4ea071` under `SZLHOLDINGS`. Not MEASURED until the job finishes and a receipt exists. |
 | **Status** | CUTTING · train in flight · weights ROADMAP |
 | **Sibling** | [`szl-receiptagent-qwen35-0.8b-v2`](https://huggingface.co/SZLHOLDINGS/szl-receiptagent-qwen35-0.8b-v2) |
 > **Fashion rule.** Silhouette from Qwen3 / Qwen3.5 instruct. Cut is original SZL. We do not republish someone else's tensors.
 ## Intended use
 - **Who:** a11oy / Alloy controller, not an end-user chatbot
@@ -62,25 +53,19 @@ szl:
 ## What it is NOT
 - Not an autonomous agent, executor, factual oracle, or weapon.
-- Not a Qwen rehost. Tensors here, when they exist, are SZL-trained adapters or a merged SZL checkpoint.
-- Not a source of ground-truth numbers.
-- Not the model running in the house lab today (that pin is `SZL-Khipu-1.5B-GGUF` Q4_K_M).
-## Honesty labels
-**Forbidden:** a score with no method, a copied leaderboard number, SOTA without a dated run, or Λ called a theorem.
 ## Evaluation
-**Status: ROADMAP.** Protocol after train: held-out schema-valid drafts, adversarial refusal. Report k/n only. No MMLU copy-paste. Quality is UNKNOWN until that run.
 ## Training
-- **Recipe:** Unsloth QLoRA SFT, receiptagent knobs (seed 11, LoRA r=16 α=32, max_steps 64, lr 2e-4, response-only CE). Script: `train_chaski.py`.
-- **Data:** `SZLHOLDINGS/szl-1-doctrine-sft` (41 MEASURED rows) plus ouroboros identity turns. Brain corpus excluded until rights admission.
-- **Job:** `6a91b8ba984507d9db4ea071` namespace `SZLHOLDINGS`. Label: RUNNING. Not a finished receipt.
-- **Train receipt:** absent until the job finishes. Absence = do not claim signed provenance.
 ## Limitations
-- Narrow identity/doctrine curriculum. Controller required. Λ remains Conjecture 1. Trust ceiling 0.97.

   artifact_class: ADAPTER
   originality: FINETUNE_DISCLOSED_BASE
   collection: "SZL Fall 2026 — Original Cuts"
   jobs: RUNNING
   job_id: 6a91b8ba984507d9db4ea071
   job_namespace: SZLHOLDINGS
 ---
 # Chaski
 **One line.** Messenger LLM. Proposal-only drafts and honest refusals for the SZL controller.
 | **Originality** | SZL fine-tune of a disclosed Apache Qwen instruct base |
 | **Base** | `Qwen/Qwen3.5-0.8B` (Apache-2.0, 0.6B–2B lock) |
 | **License** | `apache-2.0` |
+| **HF Jobs** | RUNNING `6a91b8ba984507d9db4ea071` under `SZLHOLDINGS`. Not MEASURED. No serve pin. |
 | **Status** | CUTTING · train in flight · weights ROADMAP |
+| **Later SKU** | `A11OY-MINI` will be a GGUF of **this** model after train. ROADMAP. Not Khipu. |
 | **Sibling** | [`szl-receiptagent-qwen35-0.8b-v2`](https://huggingface.co/SZLHOLDINGS/szl-receiptagent-qwen35-0.8b-v2) |
 > **Fashion rule.** Silhouette from Qwen3 / Qwen3.5 instruct. Cut is original SZL. We do not republish someone else's tensors.
+The house CPU lab serves **Khipu GGUF**, not Chaski. Cite that pin only on Khipu cards.
 ## Intended use
 - **Who:** a11oy / Alloy controller, not an end-user chatbot
 ## What it is NOT
 - Not an autonomous agent, executor, factual oracle, or weapon.
+- Not a Qwen rehost.
+- Not the live lab model. Not a tokens/s claim.
 ## Evaluation
+**Status: ROADMAP.** Quality is UNKNOWN until a held-out k/n run exists.
 ## Training
+- **Recipe:** Unsloth QLoRA SFT. Script: `train_chaski.py`.
+- **Data:** `SZLHOLDINGS/szl-1-doctrine-sft` (41 MEASURED rows) plus ouroboros identity turns. Brain corpus excluded.
+- **Job:** `6a91b8ba984507d9db4ea071` namespace `SZLHOLDINGS`. RUNNING. Train receipt absent until finish.
 ## Limitations
+- Narrow curriculum. Controller required. Λ = Conjecture 1. Trust ceiling 0.97.