Atomic-Ai/AtomicGPT2-data
Viewer • Updated • 13.3k • 21 • 1
How to use Atomic-Ai/AtomicGPT_2 with Transformers:
# Use a pipeline as a high-level helper
from transformers import pipeline
pipe = pipeline("text-generation", model="Atomic-Ai/AtomicGPT_2") # Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM
tokenizer = AutoTokenizer.from_pretrained("Atomic-Ai/AtomicGPT_2")
model = AutoModelForCausalLM.from_pretrained("Atomic-Ai/AtomicGPT_2")How to use Atomic-Ai/AtomicGPT_2 with vLLM:
# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "Atomic-Ai/AtomicGPT_2"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/completions" \
-H "Content-Type: application/json" \
--data '{
"model": "Atomic-Ai/AtomicGPT_2",
"prompt": "Once upon a time,",
"max_tokens": 512,
"temperature": 0.5
}'docker model run hf.co/Atomic-Ai/AtomicGPT_2
How to use Atomic-Ai/AtomicGPT_2 with SGLang:
# Install SGLang from pip:
pip install sglang
# Start the SGLang server:
python3 -m sglang.launch_server \
--model-path "Atomic-Ai/AtomicGPT_2" \
--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": "Atomic-Ai/AtomicGPT_2",
"prompt": "Once upon a time,",
"max_tokens": 512,
"temperature": 0.5
}'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 "Atomic-Ai/AtomicGPT_2" \
--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": "Atomic-Ai/AtomicGPT_2",
"prompt": "Once upon a time,",
"max_tokens": 512,
"temperature": 0.5
}'How to use Atomic-Ai/AtomicGPT_2 with Docker Model Runner:
docker model run hf.co/Atomic-Ai/AtomicGPT_2
AtomicGPT 2.0 ist die neueste Generation unseres KI-Modells und bietet eine deutlich bessere Leistung als unser Vorgänger, AtomicGPT 1.0. Dank eines erheblich erweiterten Datensatzes kann AtomicGPT 2.0 intelligentere Antworten liefern und ein besseres Verständnis der deutschen Sprache demonstrieren.
| Aufgaben | L-GPT_1 | L-GPT_1.1 | L-GPT_1.5 | L-GPT_1.5 mini | AtomicGPT 1.0< | AtomicGPT 2.0 | AtomicGPT 3.0 |
|---|---|---|---|---|---|---|---|
| Q&A | 7.5% | 44.17% | 73.33% | 64.17% | 58.33% | 59.17% | 90% |
kkirchheim/german-gpt2-medium, das bereits eine solide Grundlage für die deutsche Sprache bietet, konnten wir unser Modell effektiver trainieren.transformerstorchfrom transformers import pipeline, AutoTokenizer
import torch
# Modell und Tokenizer laden
MODEL_PATH = "Atomic-Ai/AtomicGPT_2"
tokenizer = AutoTokenizer.from_pretrained(MODEL_PATH)
chatbot = pipeline(
"text-generation",
model=MODEL_PATH,
tokenizer=MODEL_PATH,
device=0 if torch.cuda.is_available() else -1
)
def generate_response(prompt):
output = chatbot(
prompt,
max_new_tokens=256,
eos_token_id=tokenizer.eos_token_id, # Stoppt am <End>-Token
pad_token_id=tokenizer.pad_token_id,
do_sample=True,
temperature=0.7,
top_p=0.9
)
return output[0]['generated_text']
def format_chat(user_input):
return f"<user>{user_input}<End><AI Assistent>"
def extract_assistant_response(full_text):
parts = full_text.split("<AI Assistent>")
if len(parts) > 1:
return parts[1].split("<End>")[0].strip()
return "Error: Response format invalid"
def main():
print("Chat gestartet! Gib 'exit' ein, um zu beenden.")
print("----------------------------------")
while True:
user_input = input("\nDu: ")
if user_input.lower() == 'exit':
break
prompt = format_chat(user_input)
full_response = generate_response(prompt)
assistant_response = extract_assistant_response(full_response)
print(f"\nAI Assistent: {assistant_response}")
if __name__ == "__main__":
main()
Unterstütze uns und werde Teil unserer Community! Discord beitreten
Base model
kkirchheim/german-gpt2-medium