metadata
language: en
tags:
- deepseek
- llama
- transformers
license: apache-2.0
DeepSeek Model
This is a converted version of the DeepSeek model.
Model Description
- Model Type: Causal Language Model
- Language: English
- Base Architecture: LLaMA
- Context Length: 2048 tokens
- Parameters: Custom implementation
Usage
from transformers import AutoModelForCausalLM, AutoTokenizer
model = AutoModelForCausalLM.from_pretrained("ashwinij2/deepseek-llama-converted")
tokenizer = AutoTokenizer.from_pretrained("ashwinij2/deepseek-llama-converted")
text = "Hello, how are you?"
inputs = tokenizer(text, return_tensors="pt")
outputs = model.generate(**inputs, max_length=50)
print(tokenizer.decode(outputs[0]))