File size: 4,737 Bytes
f905581
 
 
e14d359
f905581
 
 
e14d359
f905581
 
 
 
 
 
18683e4
f905581
 
e14d359
f905581
 
 
3f4651b
 
 
 
 
 
 
 
78aa666
 
273355d
78aa666
273355d
78aa666
3f4651b
f905581
 
 
 
 
 
 
 
e14d359
f905581
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
05dd50d
 
 
 
 
 
 
 
 
 
 
 
 
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
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
---
library_name: transformers
license: apache-2.0
license_link: https://huggingface.co/huihui-ai/Qwen2.5-Coder-7B-Instruct-abliterated/blob/main/LICENSE
language:
- en
pipeline_tag: text-generation
base_model: Qwen/Qwen2.5-Coder-7B-Instruct
tags:
- chat
- abliterated
- uncensored
---

# huihui-ai/Qwen2.5-Coder-7B-Instruct-abliterated


This is an uncensored version of [Qwen2.5-Coder-7B-Instruct](https://huggingface.co/Qwen/Qwen2.5-Coder-7B-Instruct) created with abliteration (see [this article](https://huggingface.co/blog/mlabonne/abliteration) to know more about it).

Special thanks to [@FailSpy](https://huggingface.co/failspy) for the original code and technique. Please follow him if you're interested in abliterated models.

Qwen2.5-Coder uncensored version has covered six mainstream model sizes, 
[0.5](https://huggingface.co/huihui-ai/Qwen2.5-Coder-0.5B-Instruct-abliterated), 
[1.5](https://huggingface.co/huihui-ai/Qwen2.5-Coder-1.5B-Instruct-abliterated), 
[3](https://huggingface.co/huihui-ai/Qwen2.5-Coder-3B-Instruct-abliterated), 
[7](https://huggingface.co/huihui-ai/Qwen2.5-Coder-7B-Instruct-abliterated), 
[14](https://huggingface.co/huihui-ai/Qwen2.5-Coder-14B-Instruct-abliterated), 
[32](https://huggingface.co/huihui-ai/Qwen2.5-Coder-32B-Instruct-abliterated) billion parameters.    

## ollama

You can use [huihui_ai/qwen2.5-coder-abliterate](https://ollama.com/huihui_ai/qwen2.5-coder-abliterate) directly, 
```
ollama run huihui_ai/qwen2.5-coder-abliterate
```

## Usage
You can use this model in your applications by loading it with Hugging Face's `transformers` library:


```python
from transformers import AutoModelForCausalLM, AutoTokenizer

# Load the model and tokenizer
model_name = "huihui-ai/Qwen2.5-Coder-7B-Instruct-abliterated"
model = AutoModelForCausalLM.from_pretrained(
    model_name,
    torch_dtype="auto",
    device_map="auto"
)
tokenizer = AutoTokenizer.from_pretrained(model_name)

# Initialize conversation context
initial_messages = [
    {"role": "system", "content": "You are Qwen, created by Alibaba Cloud. You are a helpful assistant."}
]
messages = initial_messages.copy()  # Copy the initial conversation context

# Enter conversation loop
while True:
    # Get user input
    user_input = input("User: ").strip()  # Strip leading and trailing spaces

    # If the user types '/exit', end the conversation
    if user_input.lower() == "/exit":
        print("Exiting chat.")
        break

    # If the user types '/clean', reset the conversation context
    if user_input.lower() == "/clean":
        messages = initial_messages.copy()  # Reset conversation context
        print("Chat history cleared. Starting a new conversation.")
        continue

    # If input is empty, prompt the user and continue
    if not user_input:
        print("Input cannot be empty. Please enter something.")
        continue

    # Add user input to the conversation
    messages.append({"role": "user", "content": user_input})

    # Build the chat template
    text = tokenizer.apply_chat_template(
        messages,
        tokenize=False,
        add_generation_prompt=True
    )

    # Tokenize input and prepare it for the model
    model_inputs = tokenizer([text], return_tensors="pt").to(model.device)

    # Generate a response from the model
    generated_ids = model.generate(
        **model_inputs,
        max_new_tokens=8192
    )

    # Extract model output, removing special tokens
    generated_ids = [
        output_ids[len(input_ids):] for input_ids, output_ids in zip(model_inputs.input_ids, generated_ids)
    ]
    response = tokenizer.batch_decode(generated_ids, skip_special_tokens=True)[0]

    # Add the model's response to the conversation
    messages.append({"role": "assistant", "content": response})

    # Print the model's response
    print(f"Qwen: {response}")

```

## Evaluations
The following data has been re-evaluated and calculated as the average for each test.

| Benchmark   | Qwen2.5-Coder-7B-Instruct | Qwen2.5-Coder-7B-Instruct-abliterated |
|-------------|---------------------------|---------------------------------------|
| IF_Eval     | **63.14**                 | 61.90                                 |
| MMLU Pro    | 33.54                     | **33.56**                             |
| TruthfulQA  | **51.804**                | 48.8                                  |
| BBH         | 46.98                     | **47.17**                             |
| GPQA        | **32.85**                 | 32.63                                 |

The script used for evaluation can be found inside this repository under /eval.sh, or click [here](https://huggingface.co/huihui-ai/Qwen2.5-Coder-7B-Instruct-abliterated/blob/main/eval.sh)