File size: 2,146 Bytes
11d8908
 
90614ef
 
11d8908
 
6aa5ec2
11d8908
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
bada9fb
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
6aa5ec2
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
---
base_model: unsloth/gemma-2-2b-bnb-4bit
datasets:
- myzens/alpaca-turkish-combined
language:
- en
- tr
license: apache-2.0
tags:
- text-generation-inference
- transformers
- unsloth
- gemma2
- trl
---

# Uploaded  model

- **Developed by:** atasoglu
- **License:** apache-2.0
- **Finetuned from model :** unsloth/gemma-2-2b-bnb-4bit

This gemma2 model was trained 2x faster with [Unsloth](https://github.com/unslothai/unsloth) and Huggingface's TRL library.

[<img src="https://raw.githubusercontent.com/unslothai/unsloth/main/images/unsloth%20made%20with%20love.png" width="200"/>](https://github.com/unslothai/unsloth)

# Usage

```py
from unsloth import FastLanguageModel
import torch
max_seq_length = 2048 # Choose any! We auto support RoPE Scaling internally!
dtype = None # None for auto detection. Float16 for Tesla T4, V100, Bfloat16 for Ampere+
load_in_4bit = True # Use 4bit quantization to reduce memory usage. Can be False.

model, tokenizer = FastLanguageModel.from_pretrained(
    model_name = "atasoglu/gemma-2-2b-tr",
    max_seq_length = max_seq_length,
    dtype = dtype,
    load_in_4bit = load_in_4bit,
)

FastLanguageModel.for_inference(model) # Enable native 2x faster inference

alpaca_prompt = """Aşağıda verilen talimat ve giriş ifadelerine uygun bir cevap yaz.

### Talimat:
Aşağıdaki programlama dillerinden hangisi yapay zeka çalışmak için daha uygundur?
Sebebini açıkla.

### Giriş:
Python, C++, Java, Rust

### Cevap:
"""
inputs = tokenizer(alpaca_prompt, return_tensors="pt").to("cuda")
outputs = model.generate(
    **inputs,
    max_new_tokens=128,
    do_sample=True,
    temperature=0.2,
    repetition_penalty=1.15,
    top_k=20,
    top_p=0.7,
)

generated_tokens = outputs[:, inputs.input_ids.shape[1]:]
response = tokenizer.batch_decode(generated_tokens, skip_special_tokens=True)
print(response)
# Output: ['C++ veya Python en iyi seçenektir çünkü bu iki dilde çok sayıda yapay zeka araçları vardır. Bu araçlar, veri analizi, öğrenme algoritmaları ve karar verme süreçlerini kolaylaştırır. Ayrıca, her ikisinin de güçlü bir kütüphanesi olması da önemlidir.']
```