Text Generation
Transformers
Safetensors
English
qwen3_5
image-text-to-text
szl-holdings
series-a
doctrine-v11
governed-ai
proposal-only
conversational
Instructions to use SZLHOLDINGS/chaski with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
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
Recut: 6a91b8ba FAILED CastError; retry 6a91b990 jsonl-only; no lab load
Browse files
README.md
CHANGED
|
@@ -20,12 +20,10 @@ szl:
|
|
| 20 |
artifact_class: ADAPTER
|
| 21 |
originality: FINETUNE_DISCLOSED_BASE
|
| 22 |
collection: "SZL Fall 2026 — Original Cuts"
|
| 23 |
-
jobs:
|
| 24 |
-
job_id:
|
|
|
|
| 25 |
job_namespace: SZLHOLDINGS
|
| 26 |
-
job_prior_failed:
|
| 27 |
-
- 6a91b8ba984507d9db4ea071
|
| 28 |
-
- 6a91b990984507d9db4ea077
|
| 29 |
---
|
| 30 |
|
| 31 |
# Chaski
|
|
@@ -34,18 +32,18 @@ szl:
|
|
| 34 |
|
| 35 |
| | |
|
| 36 |
|---|---|
|
| 37 |
-
| **Artifact** | adapter (none uploaded
|
| 38 |
| **Originality** | SZL fine-tune of a disclosed Apache Qwen instruct base |
|
| 39 |
| **Base** | `Qwen/Qwen3.5-0.8B` (Apache-2.0, 0.6B–2B lock) |
|
| 40 |
| **License** | `apache-2.0` |
|
| 41 |
-
| **HF Jobs** |
|
| 42 |
-
| **Status** | CUTTING ·
|
| 43 |
-
| **Later SKU** | `A11OY-MINI` will be a GGUF of **this** model after train. ROADMAP. Not Khipu. |
|
| 44 |
| **Sibling** | [`szl-receiptagent-qwen35-0.8b-v2`](https://huggingface.co/SZLHOLDINGS/szl-receiptagent-qwen35-0.8b-v2) |
|
| 45 |
|
| 46 |
> **Fashion rule.** Silhouette from Qwen3 / Qwen3.5 instruct. Cut is original SZL. We do not republish someone else's tensors.
|
| 47 |
|
| 48 |
-
The house CPU lab serves
|
| 49 |
|
| 50 |
## Intended use
|
| 51 |
|
|
@@ -57,19 +55,19 @@ The house CPU lab serves **Khipu GGUF**, not Chaski. Cite that pin only on Khipu
|
|
| 57 |
|
| 58 |
- Not an autonomous agent, executor, factual oracle, or weapon.
|
| 59 |
- Not a Qwen rehost.
|
| 60 |
-
- Not the live lab model. Not a tokens/s claim.
|
| 61 |
|
| 62 |
## Evaluation
|
| 63 |
|
| 64 |
-
**Status: ROADMAP.** Quality is UNKNOWN until a held-out k/n run exists.
|
| 65 |
|
| 66 |
## Training
|
| 67 |
|
| 68 |
-
- **Recipe:** Unsloth QLoRA SFT. Script: `train_chaski.py`.
|
| 69 |
-
- **Data:** `SZLHOLDINGS/szl-1-doctrine-sft` (
|
| 70 |
-
- **
|
| 71 |
-
- **
|
| 72 |
|
| 73 |
## Limitations
|
| 74 |
|
| 75 |
-
- Narrow curriculum. Controller required. Λ = Conjecture 1. Trust ceiling 0.97.
|
|
|
|
| 20 |
artifact_class: ADAPTER
|
| 21 |
originality: FINETUNE_DISCLOSED_BASE
|
| 22 |
collection: "SZL Fall 2026 — Original Cuts"
|
| 23 |
+
jobs: RETRY
|
| 24 |
+
job_id: 6a91b990984507d9db4ea077
|
| 25 |
+
job_failed: 6a91b8ba984507d9db4ea071
|
| 26 |
job_namespace: SZLHOLDINGS
|
|
|
|
|
|
|
|
|
|
| 27 |
---
|
| 28 |
|
| 29 |
# Chaski
|
|
|
|
| 32 |
|
| 33 |
| | |
|
| 34 |
|---|---|
|
| 35 |
+
| **Artifact** | adapter (none uploaded) |
|
| 36 |
| **Originality** | SZL fine-tune of a disclosed Apache Qwen instruct base |
|
| 37 |
| **Base** | `Qwen/Qwen3.5-0.8B` (Apache-2.0, 0.6B–2B lock) |
|
| 38 |
| **License** | `apache-2.0` |
|
| 39 |
+
| **HF Jobs** | `6a91b8ba984507d9db4ea071` **FAILED** (CastError, 79s, no weights). Retry `6a91b990984507d9db4ea077`, jsonl-only. Not MEASURED. |
|
| 40 |
+
| **Status** | CUTTING · weights ROADMAP · evals none this run |
|
| 41 |
+
| **Later SKU** | `A11OY-MINI` will be a GGUF of **this** model after a finished train. ROADMAP. Not Khipu. |
|
| 42 |
| **Sibling** | [`szl-receiptagent-qwen35-0.8b-v2`](https://huggingface.co/SZLHOLDINGS/szl-receiptagent-qwen35-0.8b-v2) |
|
| 43 |
|
| 44 |
> **Fashion rule.** Silhouette from Qwen3 / Qwen3.5 instruct. Cut is original SZL. We do not republish someone else's tensors.
|
| 45 |
|
| 46 |
+
**Do not load Chaski into the Khipu lab.** The house CPU lab serves Khipu GGUF only. Cite that pin only on Khipu cards.
|
| 47 |
|
| 48 |
## Intended use
|
| 49 |
|
|
|
|
| 55 |
|
| 56 |
- Not an autonomous agent, executor, factual oracle, or weapon.
|
| 57 |
- Not a Qwen rehost.
|
| 58 |
+
- Not the live lab model. Not a tokens/s claim. Not job `6a91b8ba` as RUNNING.
|
| 59 |
|
| 60 |
## Evaluation
|
| 61 |
|
| 62 |
+
**Status: ROADMAP.** Evals none this run. Quality is UNKNOWN until a held-out k/n run exists after a finished job with weights.
|
| 63 |
|
| 64 |
## Training
|
| 65 |
|
| 66 |
+
- **Recipe:** Unsloth QLoRA SFT. Script: `train_chaski.py`.
|
| 67 |
+
- **Data (retry):** jsonl-only from `SZLHOLDINGS/szl-1-doctrine-sft` (`szl_dataset.jsonl`). Do not pass `SZL_ESTATE_MANAGED.json` into the trainer.
|
| 68 |
+
- **Job `6a91b8ba984507d9db4ea071`:** FAILED. CastError on mixed files (`SZL_ESTATE_MANAGED.json` next to `szl_dataset.jsonl`). 79s. No weights. No train receipt.
|
| 69 |
+
- **Retry `6a91b990984507d9db4ea077`:** jsonl-only, namespace `SZLHOLDINGS`. Not MEASURED until finish + receipt. CUTTING until then.
|
| 70 |
|
| 71 |
## Limitations
|
| 72 |
|
| 73 |
+
- Narrow curriculum. Controller required. Λ = Conjecture 1. Trust ceiling 0.97. Failed first Jobs attempt is not a quality score.
|