Instructions to use djuna/G2-GSHT with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use djuna/G2-GSHT with Transformers:

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

pipe = pipeline("text-generation", model="djuna/G2-GSHT")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("djuna/G2-GSHT")
model = AutoModelForCausalLM.from_pretrained("djuna/G2-GSHT")
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 djuna/G2-GSHT with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "djuna/G2-GSHT"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "djuna/G2-GSHT",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/djuna/G2-GSHT

SGLang

How to use djuna/G2-GSHT 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 "djuna/G2-GSHT" \
    --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": "djuna/G2-GSHT",
		"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 "djuna/G2-GSHT" \
        --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": "djuna/G2-GSHT",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use djuna/G2-GSHT with Docker Model Runner:
```
docker model run hf.co/djuna/G2-GSHT
```

djuna commited on Sep 9, 2024

Commit

afa34f8

verified ·

1 Parent(s): bf671ab

Allow system prompt

Browse files

Files changed (1) hide show

tokenizer_config.json +2 -2

tokenizer_config.json CHANGED Viewed

@@ -1744,10 +1744,10 @@
     "<end_of_turn>"
   ],
   "bos_token": "<bos>",
-  "chat_template": "{{ bos_token }}{% if messages[0]['role'] == 'system' %}{{ raise_exception('System role not supported') }}{% endif %}{% for message in messages %}{% if (message['role'] == 'user') != (loop.index0 % 2 == 0) %}{{ raise_exception('Conversation roles must alternate user/assistant/user/assistant/...') }}{% endif %}{% if (message['role'] == 'assistant') %}{% set role = 'model' %}{% else %}{% set role = message['role'] %}{% endif %}{{ '<start_of_turn>' + role + '\n' + message['content'] | trim + '<end_of_turn>\n' }}{% endfor %}{% if add_generation_prompt %}{{'<start_of_turn>model\n'}}{% endif %}",
   "clean_up_tokenization_spaces": false,
   "eos_token": "<eos>",
-  "model_max_length": 2048,
   "pad_token": "<pad>",
   "sp_model_kwargs": {},
   "spaces_between_special_tokens": false,

     "<end_of_turn>"
   ],
   "bos_token": "<bos>",
+  "chat_template": "{{ bos_token }}{% for message in messages %}{% set system_adjust = 0 if loop.index0 == 0 and message['role'] != 'system' else 0 %}{% if loop.index0 == 0 and message['role'] == 'system' %}{{ '<start_of_turn>user\n'+message['content'] | trim +'\n'}}{% if messages | length > 1 and messages[1]['role'] == 'user' %}{{ messages[1]['content'] | trim }}{% elif messages | length > 1 %}{{ raise_exception('Error: Expected user role after system role.') }}{% endif %}{{ '<end_of_turn>\n' }}{% elif loop.index0 == 1 and message['role'] == 'user' and messages[0]['role'] == 'system' %}{% else %}{% set role = 'model' if message['role'] == 'assistant' else message['role'] %}{% if loop.index0 <= 1 or messages[loop.index0 - 1]['role'] != role %}{{ '<start_of_turn>' }}{{ role }}{{ '\n'+message['content'] | trim }}{{ '<end_of_turn>'+'\n' }}{% elif loop.index0 > 1 %}{{ raise_exception('Error: Consecutive messages with the same role are not allowed.') }}{% endif %}{% endif %}{% endfor %}{% if add_generation_prompt %}{{'<start_of_turn>model'}}{% endif %}",
   "clean_up_tokenization_spaces": false,
   "eos_token": "<eos>",
+  "model_max_length": 1048576,
   "pad_token": "<pad>",
   "sp_model_kwargs": {},
   "spaces_between_special_tokens": false,