Instructions to use T145/KRONOS-8B-V1-P3 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use T145/KRONOS-8B-V1-P3 with Transformers:

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

pipe = pipeline("text-generation", model="T145/KRONOS-8B-V1-P3")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("T145/KRONOS-8B-V1-P3")
model = AutoModelForCausalLM.from_pretrained("T145/KRONOS-8B-V1-P3")
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=40)
print(tokenizer.decode(outputs[0][inputs["input_ids"].shape[-1]:]))

Notebooks
Google Colab
Kaggle
Local Apps

vLLM

How to use T145/KRONOS-8B-V1-P3 with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "T145/KRONOS-8B-V1-P3"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "T145/KRONOS-8B-V1-P3",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/T145/KRONOS-8B-V1-P3

SGLang

How to use T145/KRONOS-8B-V1-P3 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 "T145/KRONOS-8B-V1-P3" \
    --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": "T145/KRONOS-8B-V1-P3",
		"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 "T145/KRONOS-8B-V1-P3" \
        --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": "T145/KRONOS-8B-V1-P3",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use T145/KRONOS-8B-V1-P3 with Docker Model Runner:
```
docker model run hf.co/T145/KRONOS-8B-V1-P3
```
Browse Quantizations to use this model in llama.cpp, Ollama, LM Studio, or any compatible app.

KRONOS V1 P3

This is a merge of Meta Llama 3.1 Instruct and REILIX's "750MB" LORA, created using llm-tools.

The primary purpose of this model is to be merged into other models in the same family using the TIES merge method.

Creating quants for this is entirely unnecessary.

Merge Details

Configuration

The following Bash command was used to produce this model:

python /llm-tools/merge-lora.py -m unsloth/Meta-Llama-3.1-8B-Instruct -l REILX/Llama-3-8B-Instruct-750Mb-lora

Open LLM Leaderboard Evaluation Results

Detailed results can be found here! Summarized results can be found here!

Metric	Value (%)
Average	25.67
IFEval (0-Shot)	71.37
BBH (3-Shot)	30.27
MATH Lvl 5 (4-Shot)	18.35
GPQA (0-shot)	1.34
MuSR (0-shot)	5.96
MMLU-PRO (5-shot)	26.72

Downloads last month: 7

Safetensors

Model size

8B params

Tensor type

F16

Model tree for T145/KRONOS-8B-V1-P3

REILX/Llama-3-8B-Instruct-750Mb-lora

unsloth/Meta-Llama-3.1-8B-Instruct

Merge model

this model

Merges

7 models

Quantizations

1 model

Datasets used to train T145/KRONOS-8B-V1-P3

Evaluation results

averaged accuracy on IFEval (0-Shot)
Open LLM Leaderboard

71.370
normalized accuracy on BBH (3-Shot)
test set Open LLM Leaderboard

30.270
exact match on MATH Lvl 5 (4-Shot)
test set Open LLM Leaderboard

18.350
acc_norm on GPQA (0-shot)
Open LLM Leaderboard

1.340
acc_norm on MuSR (0-shot)
Open LLM Leaderboard

5.960
accuracy on MMLU-PRO (5-shot)
test set Open LLM Leaderboard

26.720