Instructions to use k1h0/OpenCoder-8B-Instruct-query_nsx with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use k1h0/OpenCoder-8B-Instruct-query_nsx with Transformers:

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

pipe = pipeline("text-generation", model="k1h0/OpenCoder-8B-Instruct-query_nsx")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("k1h0/OpenCoder-8B-Instruct-query_nsx")
model = AutoModelForCausalLM.from_pretrained("k1h0/OpenCoder-8B-Instruct-query_nsx")
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 k1h0/OpenCoder-8B-Instruct-query_nsx with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "k1h0/OpenCoder-8B-Instruct-query_nsx"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "k1h0/OpenCoder-8B-Instruct-query_nsx",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/k1h0/OpenCoder-8B-Instruct-query_nsx

SGLang

How to use k1h0/OpenCoder-8B-Instruct-query_nsx 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 "k1h0/OpenCoder-8B-Instruct-query_nsx" \
    --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": "k1h0/OpenCoder-8B-Instruct-query_nsx",
		"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 "k1h0/OpenCoder-8B-Instruct-query_nsx" \
        --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": "k1h0/OpenCoder-8B-Instruct-query_nsx",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use k1h0/OpenCoder-8B-Instruct-query_nsx with Docker Model Runner:
```
docker model run hf.co/k1h0/OpenCoder-8B-Instruct-query_nsx
```

OpenCoder-8B-Instruct-query_nsx / training_args.yaml

k1h0

Upload folder using huggingface_hub

0208592 verified about 1 year ago

raw

history blame contribute delete

889 Bytes

	apollo_rank: 256
	apollo_scale: 1
	apollo_target: all
	apollo_update_interval: 200
	bf16: true
	cutoff_len: 4096
	dataset: codes_330k_nsx
	dataset_dir: data
	ddp_timeout: 180000000
	do_train: true
	enable_liger_kernel: true
	finetuning_type: freeze
	flash_attn: auto
	freeze_trainable_layers: 2
	freeze_trainable_modules: all
	gradient_accumulation_steps: 8
	include_num_input_tokens_seen: true
	learning_rate: 5.0e-05
	logging_steps: 1
	lr_scheduler_type: cosine
	max_grad_norm: 1.0
	max_samples: 50000000
	model_name_or_path: infly/OpenCoder-8B-Instruct
	neat_packing: true
	num_train_epochs: 1.0
	output_dir: saves/OpenCoder-8B-Instruct/freeze/opencoder_nsx
	packing: true
	per_device_train_batch_size: 16
	plot_loss: true
	preprocessing_num_workers: 16
	report_to: none
	rope_scaling: llama3
	save_steps: 500
	stage: sft
	template: opencoder
	trust_remote_code: true
	use_apollo: true
	use_llama_pro: true
	warmup_steps: 0