Instructions to use HachiML/BitLlama2-jp-127M-optim-4 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use HachiML/BitLlama2-jp-127M-optim-4 with Transformers:

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

pipe = pipeline("text-generation", model="HachiML/BitLlama2-jp-127M-optim-4", trust_remote_code=True)

# Load model directly
from transformers import AutoModelForCausalLM
model = AutoModelForCausalLM.from_pretrained("HachiML/BitLlama2-jp-127M-optim-4", trust_remote_code=True, dtype="auto")

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use HachiML/BitLlama2-jp-127M-optim-4 with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "HachiML/BitLlama2-jp-127M-optim-4"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "HachiML/BitLlama2-jp-127M-optim-4",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker

docker model run hf.co/HachiML/BitLlama2-jp-127M-optim-4

SGLang

How to use HachiML/BitLlama2-jp-127M-optim-4 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 "HachiML/BitLlama2-jp-127M-optim-4" \
    --host 0.0.0.0 \
    --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "HachiML/BitLlama2-jp-127M-optim-4",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

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 "HachiML/BitLlama2-jp-127M-optim-4" \
        --host 0.0.0.0 \
        --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "HachiML/BitLlama2-jp-127M-optim-4",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Docker Model Runner
How to use HachiML/BitLlama2-jp-127M-optim-4 with Docker Model Runner:
```
docker model run hf.co/HachiML/BitLlama2-jp-127M-optim-4
```

BitLlama2-jp-127M-optim-4

This model is a fine-tuned version of on an unknown dataset. It achieves the following results on the evaluation set:

Loss: 3.4021

Model description

More information needed

Intended uses & limitations

More information needed

Training and evaluation data

More information needed

Training procedure

Training hyperparameters

The following hyperparameters were used during training:

learning_rate: 0.0024
train_batch_size: 96
eval_batch_size: 96
seed: 42
optimizer: Adam with betas=(0.9,0.95) and epsilon=1e-08
lr_scheduler_type: linear
lr_scheduler_warmup_steps: 500
num_epochs: 1

Training results

Training Loss	Epoch	Step	Validation Loss
6.8073	0.07	200	4.8769
4.5389	0.15	400	4.3762
4.2297	0.22	600	4.1527
4.0242	0.29	800	3.9881
3.8902	0.36	1000	3.8885
3.7927	0.44	1200	3.8047
3.7141	0.51	1400	3.7333
3.6597	0.58	1600	3.6681
3.579	0.66	1800	3.6041
3.5141	0.73	2000	3.5424
3.4606	0.8	2200	3.4941
3.4116	0.88	2400	3.4467
3.361	0.95	2600	3.4021

Framework versions

Transformers 4.38.2
Pytorch 2.2.1+cu121
Datasets 2.18.0
Tokenizers 0.15.2

Downloads last month: 3

Safetensors

Model size

0.1B params

Tensor type

F32