Instructions to use myjellylabs/Krypton-Fly-1.0 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use myjellylabs/Krypton-Fly-1.0 with Transformers:

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

pipe = pipeline("text-generation", model="myjellylabs/Krypton-Fly-1.0")
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("myjellylabs/Krypton-Fly-1.0")
model = AutoModelForCausalLM.from_pretrained("myjellylabs/Krypton-Fly-1.0", 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 myjellylabs/Krypton-Fly-1.0 with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "myjellylabs/Krypton-Fly-1.0"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "myjellylabs/Krypton-Fly-1.0",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/myjellylabs/Krypton-Fly-1.0

SGLang

How to use myjellylabs/Krypton-Fly-1.0 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 "myjellylabs/Krypton-Fly-1.0" \
    --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": "myjellylabs/Krypton-Fly-1.0",
		"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 "myjellylabs/Krypton-Fly-1.0" \
        --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": "myjellylabs/Krypton-Fly-1.0",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use myjellylabs/Krypton-Fly-1.0 with Docker Model Runner:
```
docker model run hf.co/myjellylabs/Krypton-Fly-1.0
```
Browse Quantizations to use this model in llama.cpp, Ollama, LM Studio, or any compatible app.

Krypton Fly 1.0 (`Krypton`)

Krypton Fly 1.0 (main name: Krypton) is a standalone ~2.5B-parameter open-source language model trained and released by Jelly Lab Game (myjellylabs) under the Apache-2.0 license.

It features standalone BF16 weights (LlamaForCausalLM architecture, 42 layers, Grouped-Query Attention, 131,072-token native context window), dual-mode thinking (<think>...</think>), and native XML tool calling.

Highlights

Identity: Main name Krypton, series Krypton Fly 1.0, trained by Jelly Lab Game (myjellylabs).
Agent & Terminal Coding: Trained on round-trip verified XML tool calls and coding agent trajectories.
OPD-Safe Training: Fine-tuned with LoRA (r=16, alpha=32) at lr=1e-5 (1 epoch, BF16, completion-only loss) paired with zero-drift self-replay to preserve reasoning and coding alignment.

Internal Quality Gate Checks

check	baseline	Krypton Fly 1.0
identity prompts mention "Krypton" / "Krypton Fly 1.0"	0.00	0.85
well-formed XML tool call (held-out agent prompts)	0.92	0.92
normal chat ends cleanly (no endless generation)	0.60	0.47
basic sanity prompts	1.00	1.00

Usage

import torch
from transformers import AutoModelForCausalLM, AutoTokenizer
repo = "myjellylabs/Krypton-Fly-1.0"
tok = AutoTokenizer.from_pretrained(repo)
model = AutoModelForCausalLM.from_pretrained(repo, dtype=torch.bfloat16, device_map="auto")
msgs = [{"role": "user", "content": "Who are you and who trained you?"}]
text = tok.apply_chat_template(msgs, tokenize=False, add_generation_prompt=True, enable_thinking=False)
ids = tok(text, return_tensors="pt", add_special_tokens=False).to(model.device)
out = model.generate(**ids, max_new_tokens=256, do_sample=True, temperature=1.0, top_p=0.95)
print(tok.decode(out[0][ids["input_ids"].shape[1]:], skip_special_tokens=True))

Recommended sampling: temperature=1.0, top_p=0.95. Requires transformers>=5.6.

License and Attribution (Apache-2.0)

Licensed under the Apache-2.0 License (see LICENSE and NOTICE files included in this repository). In compliance with Apache-2.0 Section 4, this repository distributes modified, merged weight tensors trained by Jelly Lab Game (myjellylabs).

Downloads last month: 209

Safetensors

Model size

3B params

Tensor type

BF16

Model tree for myjellylabs/Krypton-Fly-1.0

Quantizations