Downloads · 30 days
27
0% of all-time downloads
umarigan/Trendyol-LLM-7b-chat-v0.1-DPO
Trendyol-LLM-7b-chat-v0.1-DPO is a text generation model from umarigan. Use it when you need the model to write or continue text. It is set up for transformers. The card lists the license as apache-2.0.
This is the model card of a 🤗 transformers model that has been pushed on the Hub. This model card has been automatically generated.
Downloads · 30 days
27
0% of all-time downloads
All-time downloads
62.2K
Public
Parameters
6.8B
13.7 GB on disk
Likes
2
Public
Click a slice to open those files.
.safetensors13.7 GB · 100%
From the Hugging Face model README
This is the model card of a 🤗 transformers model that has been pushed on the Hub. This model card has been automatically generated.
# Use a pipeline as a high-level helper
from transformers import pipeline
pipe = pipeline("text-generation", model="umarigan/Trendyol-LLM-7b-chat-v0.1-DPO")
# Generate text
sequences = pipe(
"büyük dil modellerinin finans alanındaki kullanımları nelerdir",
do_sample=True,
temperature=0.7,
top_p=0.9,
num_return_sequences=1,
max_length=200,
)
print(sequences[0]['generated_text'])
Question: büyük dil modellerinin finans alanındaki kullanımları nelerdir?
Answer: Çok büyük dil modelleri, özellikle de Transformer gibi, karmaşık dil görevlerinin üstesinden gelmek için tasarlanmışlardır. Bu, finansal piyasalardaki veri işleme, fiyat tahmini ve analizleri, finansal haberler ve raporlama gibi süreçleri içerir. Ayrıca, büyük dil modelleri, doğal dil işleme, metin sınıflandırma ve soru cevaplama gibi görevlerin yanı sıra, müşteri hizmetleri gibi insan etkileşimi gerektiren finansal hizmetlerde de kullanılmaktadır.
This model trained on falcon instruction dataset that translated to Turkis language Dataset: https://huggingface.co/datasets/umarigan/falcon_feedback_instraction_Turkish
Some training arguments are as follow:
max_prompt_length=1024,
max_length=1536,
per_device_train_batch_size=4,
gradient_accumulation_steps=4,
gradient_checkpointing=True,
learning_rate=5e-5,
lr_scheduler_type="cosine",
max_steps=200,
save_strategy="no",
logging_steps=1,
output_dir=new_model,
optim="paged_adamw_32bit",
warmup_steps=100,
fp16=True,
wandb results: https://api.wandb.ai/links/umar-i-gan/0hnrvrdq