Instructions to use danielgombas/llama_3b_step2_batch_v3 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use danielgombas/llama_3b_step2_batch_v3 with Transformers:

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

pipe = pipeline("text-generation", model="danielgombas/llama_3b_step2_batch_v3")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("danielgombas/llama_3b_step2_batch_v3")
model = AutoModelForCausalLM.from_pretrained("danielgombas/llama_3b_step2_batch_v3")
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 danielgombas/llama_3b_step2_batch_v3 with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "danielgombas/llama_3b_step2_batch_v3"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "danielgombas/llama_3b_step2_batch_v3",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/danielgombas/llama_3b_step2_batch_v3

SGLang

How to use danielgombas/llama_3b_step2_batch_v3 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 "danielgombas/llama_3b_step2_batch_v3" \
    --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": "danielgombas/llama_3b_step2_batch_v3",
		"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 "danielgombas/llama_3b_step2_batch_v3" \
        --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": "danielgombas/llama_3b_step2_batch_v3",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use danielgombas/llama_3b_step2_batch_v3 with Docker Model Runner:
```
docker model run hf.co/danielgombas/llama_3b_step2_batch_v3
```

llama_3b_step2_batch_v3

This model was trained from scratch on an unknown dataset. It achieves the following results on the evaluation set:

Loss: 0.3151

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: 3e-05
train_batch_size: 4
eval_batch_size: 40
seed: 42
gradient_accumulation_steps: 4
total_train_batch_size: 16
optimizer: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
lr_scheduler_type: linear
num_epochs: 2

Training results

Training Loss	Epoch	Step	Validation Loss
0.8424	0.0682	50	0.9705
0.8978	0.1363	100	0.7915
0.5886	0.2045	150	0.6863
0.7013	0.2727	200	0.6097
0.4795	0.3408	250	0.5510
0.5601	0.4090	300	0.5066
0.4968	0.4772	350	0.4675
0.3663	0.5453	400	0.4353
0.3754	0.6135	450	0.4074
0.3271	0.6817	500	0.3890
0.2731	0.7498	550	0.3744
0.3881	0.8180	600	0.3599
0.3295	0.8862	650	0.3489
0.3408	0.9543	700	0.3412
0.2966	1.0225	750	0.3418
0.2334	1.0907	800	0.3365
0.2457	1.1588	850	0.3321
0.2332	1.2270	900	0.3311
0.2279	1.2952	950	0.3285
0.2762	1.3633	1000	0.3245
0.1464	1.4315	1050	0.3233
0.2075	1.4997	1100	0.3211
0.2251	1.5678	1150	0.3184
0.1811	1.6360	1200	0.3167
0.2104	1.7042	1250	0.3164
0.224	1.7723	1300	0.3158
0.2483	1.8405	1350	0.3154
0.1773	1.9087	1400	0.3151
0.219	1.9768	1450	0.3151

Framework versions

Transformers 4.46.1
Pytorch 2.1.0+cu118
Datasets 3.0.2
Tokenizers 0.20.1

Downloads last month: 3

Safetensors

Model size

3B params

Tensor type

BF16