Instructions to use tsavage68/UTI_L3_625steps_1e7rate_SFT with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use tsavage68/UTI_L3_625steps_1e7rate_SFT with Transformers:

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

pipe = pipeline("text-generation", model="tsavage68/UTI_L3_625steps_1e7rate_SFT")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("tsavage68/UTI_L3_625steps_1e7rate_SFT")
model = AutoModelForCausalLM.from_pretrained("tsavage68/UTI_L3_625steps_1e7rate_SFT", 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=40)
print(tokenizer.decode(outputs[0][inputs["input_ids"].shape[-1]:]))

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use tsavage68/UTI_L3_625steps_1e7rate_SFT with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "tsavage68/UTI_L3_625steps_1e7rate_SFT"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "tsavage68/UTI_L3_625steps_1e7rate_SFT",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/tsavage68/UTI_L3_625steps_1e7rate_SFT

SGLang

How to use tsavage68/UTI_L3_625steps_1e7rate_SFT 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 "tsavage68/UTI_L3_625steps_1e7rate_SFT" \
    --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": "tsavage68/UTI_L3_625steps_1e7rate_SFT",
		"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 "tsavage68/UTI_L3_625steps_1e7rate_SFT" \
        --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": "tsavage68/UTI_L3_625steps_1e7rate_SFT",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use tsavage68/UTI_L3_625steps_1e7rate_SFT with Docker Model Runner:
```
docker model run hf.co/tsavage68/UTI_L3_625steps_1e7rate_SFT
```

UTI_L3_625steps_1e7rate_SFT

This model is a fine-tuned version of meta-llama/Meta-Llama-3-8B-Instruct on an unknown dataset. It achieves the following results on the evaluation set:

Loss: 1.9216

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: 1e-07
train_batch_size: 2
eval_batch_size: 1
seed: 42
gradient_accumulation_steps: 2
total_train_batch_size: 4
optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
lr_scheduler_type: cosine
lr_scheduler_warmup_steps: 100
training_steps: 625

Training results

Training Loss	Epoch	Step	Validation Loss
2.4485	0.3333	25	2.4666
2.4645	0.6667	50	2.4522
2.452	1.0	75	2.4164
2.391	1.3333	100	2.3529
2.2825	1.6667	125	2.2874
2.1758	2.0	150	2.2259
2.2189	2.3333	175	2.1723
2.1642	2.6667	200	2.1247
2.1223	3.0	225	2.0833
2.0341	3.3333	250	2.0461
1.9642	3.6667	275	2.0153
1.998	4.0	300	1.9903
2.0321	4.3333	325	1.9697
1.9443	4.6667	350	1.9542
1.9787	5.0	375	1.9421
1.9403	5.3333	400	1.9349
1.9464	5.6667	425	1.9279
1.9067	6.0	450	1.9245
1.927	6.3333	475	1.9228
1.8829	6.6667	500	1.9220
1.9613	7.0	525	1.9229
1.9701	7.3333	550	1.9218
1.9411	7.6667	575	1.9217
1.9127	8.0	600	1.9216
1.9497	8.3333	625	1.9216

Framework versions

Transformers 4.41.2
Pytorch 2.0.0+cu117
Datasets 2.19.2
Tokenizers 0.19.1

Downloads last month: 6

Safetensors

Model size

8B params

Tensor type

F16

Model tree for tsavage68/UTI_L3_625steps_1e7rate_SFT

Base model

meta-llama/Meta-Llama-3-8B-Instruct

Finetuned

(1076)

this model