Hugging Face's logo

NorGLM
/
NorLlama-3B-conversation-peft

Text Generation
Norwegian
Model card Files Community
NorLlama-3B-conversation-peft
File size: 2,737 Bytes 
c60ebce
 
cc7bc77
 
 
 
 
c60ebce
cc7bc77
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
 
---
license: cc-by-nc-sa-4.0
datasets:
- NorGLM/NO-ConvAI2
language:
- 'no'
pipeline_tag: text-generation
---

# Model Card

NorLlama-3B-conversation-peft is trained on top of [NorLlama-3B](https://huggingface.co/NorGLM/NorLlama-3B) model on [NO-ConvAI2](https://huggingface.co/datasets/NorGLM/NO-ConvAI2) dataset.

Prompt format:
```
Human: {prompt} Robot: |||\n {answer}
```

Inference prompt:
```
Human: {prompt} Robot: |||\n
```

## Run the Model
```python
from peft import PeftModel, PeftConfig
from transformers import AutoModelForCausalLM, AutoTokenizer
import torch
from tqdm.auto import tqdm

source_model_id = "NorGLM/NorLlama-3B"
peft_model_id = "NorGLM/NorLlama-3B-conversation-peft"

config = PeftConfig.from_pretrained(peft_model_id)
model = AutoModelForCausalLM.from_pretrained(source_model_id, device_map='balanced')

tokenizer_max_len = 2048
tokenizer_config = {'pretrained_model_name_or_path': source_model_id,
                            'max_len': tokenizer_max_len}
tokenizer = tokenizer = AutoTokenizer.from_pretrained(**tokenizer_config)
tokenizer.pad_token = tokenizer.eos_token

model = PeftModel.from_pretrained(model, peft_model_id)
```

## Inference Example
Load the model to evaluate on the test set of NO-CNN/DailyMail dataset:
```python
def load_and_prepare_data_last_prompt(df):
    """ Load and spearates last prompt from prompt """
     # id, turn_id, prompt, answer
    last_prompt = ["Human: " + df['prompt']
                   [i].split("Human:")[-1] for i in range(len(df))]
    df['last_prompt'] = last_prompt
    return df

def generate_text(text, max_length=200):
    # generate with greedy search
    model_inputs = tokenizer(text, return_attention_mask=True, return_tensors="pt",
              padding=True, truncation=True,  max_length=tokenizer_max_len)

    with torch.no_grad():
        output_tokens = model.generate(
                  **model_inputs, max_new_tokens=50, pad_token_id=tokenizer.eos_token_id)

    text_outputs = [tokenizer.decode(
                x, skip_special_tokens=True) for x in output_tokens]

    return text_outputs

print("--LOADING EVAL DATAS---")
eval_data = load_dataset("NorGLM/NO-ConvAI2", data_files="test_PersonaChat_prompt.json")
prompts = eval_data['train']['prompt']
positive_samples = eval_data['train']['answer']

print("--MAKING PREDICTIONS---")
model.eval()

output_file = <output file name>
generated_text = []

for prompt in tqdm(prompts):
    generated_text.append(generate_text(prompt, max_length=tokenizer_max_len))

df = pd.DataFrame({'prompts':prompts, 'generated_text':generated_text, 'positive_sample':positive_samples})

print("Save results to csv file...")
df.to_csv(output_file)

```

## Note
More training details will be released soon!