Instructions to use W-61/llama-3-8b-base-sft-hh-harmless-8xh200 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use W-61/llama-3-8b-base-sft-hh-harmless-8xh200 with Transformers:

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

pipe = pipeline("text-generation", model="W-61/llama-3-8b-base-sft-hh-harmless-8xh200")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("W-61/llama-3-8b-base-sft-hh-harmless-8xh200")
model = AutoModelForCausalLM.from_pretrained("W-61/llama-3-8b-base-sft-hh-harmless-8xh200", 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 W-61/llama-3-8b-base-sft-hh-harmless-8xh200 with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "W-61/llama-3-8b-base-sft-hh-harmless-8xh200"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "W-61/llama-3-8b-base-sft-hh-harmless-8xh200",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/W-61/llama-3-8b-base-sft-hh-harmless-8xh200

SGLang

How to use W-61/llama-3-8b-base-sft-hh-harmless-8xh200 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 "W-61/llama-3-8b-base-sft-hh-harmless-8xh200" \
    --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": "W-61/llama-3-8b-base-sft-hh-harmless-8xh200",
		"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 "W-61/llama-3-8b-base-sft-hh-harmless-8xh200" \
        --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": "W-61/llama-3-8b-base-sft-hh-harmless-8xh200",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use W-61/llama-3-8b-base-sft-hh-harmless-8xh200 with Docker Model Runner:
```
docker model run hf.co/W-61/llama-3-8b-base-sft-hh-harmless-8xh200
```

llama-3-8b-base-sft-hh-harmless-8xh200 / trainer_state.json

W-61

Upload validated HH harmless SFT checkpoint

4b571a9 verified 6 months ago

Raw History Blame Contribute Delete

4.8 kB

Invalid JSON:Unexpected token 'I', ..."ad_norm": Infinity, "... is not valid JSON

	{
	"best_global_step": null,
	"best_metric": null,
	"best_model_checkpoint": null,
	"epoch": 1.0,
	"eval_steps": 100,
	"global_step": 104,
	"is_hyper_param_search": false,
	"is_local_process_zero": true,
	"is_world_process_zero": true,
	"log_history": [
	{
	"epoch": 0.009615384615384616,
	"grad_norm": Infinity,
	"learning_rate": 0.0,
	"loss": 3.4921,
	"step": 1
	},
	{
	"epoch": 0.04807692307692308,
	"grad_norm": Infinity,
	"learning_rate": 7.272727272727273e-06,
	"loss": 3.4372,
	"step": 5
	},
	{
	"epoch": 0.09615384615384616,
	"grad_norm": 1.3386897115800666e+19,
	"learning_rate": 1.6363636363636366e-05,
	"loss": 3.4356,
	"step": 10
	},
	{
	"epoch": 0.14423076923076922,
	"grad_norm": 49.75270462036133,
	"learning_rate": 1.994869323391895e-05,
	"loss": 3.1194,
	"step": 15
	},
	{
	"epoch": 0.19230769230769232,
	"grad_norm": 81.14952087402344,
	"learning_rate": 1.963705643889941e-05,
	"loss": 2.6138,
	"step": 20
	},
	{
	"epoch": 0.2403846153846154,
	"grad_norm": 11.374646186828613,
	"learning_rate": 1.9051145072503216e-05,
	"loss": 2.3041,
	"step": 25
	},
	{
	"epoch": 0.28846153846153844,
	"grad_norm": 7.008129119873047,
	"learning_rate": 1.8207634412072765e-05,
	"loss": 2.0668,
	"step": 30
	},
	{
	"epoch": 0.33653846153846156,
	"grad_norm": 16.298776626586914,
	"learning_rate": 1.7130531116312202e-05,
	"loss": 1.8889,
	"step": 35
	},
	{
	"epoch": 0.38461538461538464,
	"grad_norm": 5.356802940368652,
	"learning_rate": 1.5850489985953076e-05,
	"loss": 1.83,
	"step": 40
	},
	{
	"epoch": 0.4326923076923077,
	"grad_norm": 7.75869607925415,
	"learning_rate": 1.4403941515576344e-05,
	"loss": 1.7812,
	"step": 45
	},
	{
	"epoch": 0.4807692307692308,
	"grad_norm": 4.185631275177002,
	"learning_rate": 1.283205506682304e-05,
	"loss": 1.75,
	"step": 50
	},
	{
	"epoch": 0.5288461538461539,
	"grad_norm": 2.844418525695801,
	"learning_rate": 1.1179567171508463e-05,
	"loss": 1.6893,
	"step": 55
	},
	{
	"epoch": 0.5769230769230769,
	"grad_norm": 2.267258405685425,
	"learning_rate": 9.493508311612874e-06,
	"loss": 1.6535,
	"step": 60
	},
	{
	"epoch": 0.625,
	"grad_norm": 1.7939544916152954,
	"learning_rate": 7.821864412511485e-06,
	"loss": 1.6214,
	"step": 65
	},
	{
	"epoch": 0.6730769230769231,
	"grad_norm": 1.9716706275939941,
	"learning_rate": 6.21221114389424e-06,
	"loss": 1.6081,
	"step": 70
	},
	{
	"epoch": 0.7211538461538461,
	"grad_norm": 1.7848585844039917,
	"learning_rate": 4.710359896730379e-06,
	"loss": 1.5761,
	"step": 75
	},
	{
	"epoch": 0.7692307692307693,
	"grad_norm": 1.6977120637893677,
	"learning_rate": 3.3590539723276083e-06,
	"loss": 1.5688,
	"step": 80
	},
	{
	"epoch": 0.8173076923076923,
	"grad_norm": 1.863008737564087,
	"learning_rate": 2.196752090479083e-06,
	"loss": 1.5659,
	"step": 85
	},
	{
	"epoch": 0.8653846153846154,
	"grad_norm": 1.4840096235275269,
	"learning_rate": 1.2565338385541792e-06,
	"loss": 1.5474,
	"step": 90
	},
	{
	"epoch": 0.9134615384615384,
	"grad_norm": 1.3736684322357178,
	"learning_rate": 5.651582129001987e-07,
	"loss": 1.5542,
	"step": 95
	},
	{
	"epoch": 0.9615384615384616,
	"grad_norm": 1.390762448310852,
	"learning_rate": 1.4230204685196202e-07,
	"loss": 1.5418,
	"step": 100
	},
	{
	"epoch": 0.9615384615384616,
	"eval_loss": 1.5660021305084229,
	"eval_runtime": 1.8376,
	"eval_samples_per_second": 405.967,
	"eval_steps_per_second": 3.265,
	"step": 100
	},
	{
	"epoch": 1.0,
	"step": 104,
	"total_flos": 3.836373525279539e+16,
	"train_loss": 1.9909558915174925,
	"train_runtime": 382.237,
	"train_samples_per_second": 34.549,
	"train_steps_per_second": 0.272
	}
	],
	"logging_steps": 5,
	"max_steps": 104,
	"num_input_tokens_seen": 0,
	"num_train_epochs": 1,
	"save_steps": 200,
	"stateful_callbacks": {
	"TrainerControl": {
	"args": {
	"should_epoch_stop": false,
	"should_evaluate": false,
	"should_log": false,
	"should_save": true,
	"should_training_stop": true
	},
	"attributes": {}
	}
	},
	"total_flos": 3.836373525279539e+16,
	"train_batch_size": 16,
	"trial_name": null,
	"trial_params": null
	}