Instructions to use himalaya-ai/himalayagpt-0.5b-it with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use himalaya-ai/himalayagpt-0.5b-it with Transformers:

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

pipe = pipeline("text-generation", model="himalaya-ai/himalayagpt-0.5b-it", trust_remote_code=True)

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("himalaya-ai/himalayagpt-0.5b-it", trust_remote_code=True)
model = AutoModelForCausalLM.from_pretrained("himalaya-ai/himalayagpt-0.5b-it", trust_remote_code=True, device_map="auto")

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use himalaya-ai/himalayagpt-0.5b-it with vLLM:

Install from pip and serve model

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

Use Docker

docker model run hf.co/himalaya-ai/himalayagpt-0.5b-it

SGLang

How to use himalaya-ai/himalayagpt-0.5b-it 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 "himalaya-ai/himalayagpt-0.5b-it" \
    --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": "himalaya-ai/himalayagpt-0.5b-it",
		"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 "himalaya-ai/himalayagpt-0.5b-it" \
        --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": "himalaya-ai/himalayagpt-0.5b-it",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Docker Model Runner
How to use himalaya-ai/himalayagpt-0.5b-it with Docker Model Runner:
```
docker model run hf.co/himalaya-ai/himalayagpt-0.5b-it
```

himalayagpt-0.5b-it / config.json

HimalayaGPT

Export SFT checkpoint d15_harl_fulltokens_sdpa_bs32 step 49164 (instruct)

1613fb9 verified 3 months ago

Raw

History Blame

767 Bytes

	{
	"architectures": [
	"NanochatForCausalLM"
	],
	"model_type": "nanochat",
	"auto_map": {
	"AutoConfig": "configuration_nanochat.NanochatConfig",
	"AutoModelForCausalLM": "modeling_nanochat.NanochatForCausalLM",
	"AutoTokenizer": "tokenization_nanochat.NanochatTokenizer"
	},
	"vocab_size": 32768,
	"padded_vocab_size": 32768,
	"sequence_len": 2048,
	"n_layer": 15,
	"n_head": 8,
	"n_kv_head": 8,
	"n_embd": 1024,
	"hidden_size": 1024,
	"num_hidden_layers": 15,
	"num_attention_heads": 8,
	"num_key_value_heads": 8,
	"max_position_embeddings": 2048,
	"window_pattern": "L",
	"use_cache": false,
	"bos_token_id": 32759,
	"eos_token_id": 32759,
	"pad_token_id": 32759,
	"torch_dtype": "float32",
	"transformers_version": "4.57.0"
	}