metadata

base_model:
  - huihui-ai/DeepSeek-R1-Distill-Qwen-32B-abliterated
library_name: transformers
tags:
  - Text Generation
  - text-generation-inference
  - Inference Endpoints
  - Transformers
  - Fusion
language:
  - en

DeepSeek-R1-Distill-Qwen-Coder-32B-Fusion-9010

Overview

DeepSeek-R1-Distill-Qwen-Coder-32B-Fusion-9010 is a mixed model that combines the strengths of two powerful DeepSeek-R1-Distill-Qwen-based models: huihui-ai/DeepSeek-R1-Distill-Qwen-32B-abliterated and huihui-ai/Qwen2.5-Coder-32B-Instruct-abliterated. The weights are blended in a 9:1 ratio, with 90% of the weights from DeepSeek-R1-Distill-Qwen-32B-abliterated and 10% from the Qwen2.5-Coder-32B-Instruct-abliterated model. Although it's a simple mix, the model is usable, and no gibberish has appeared. This is an experiment. I test the 9:1, 8:2, 7:3, 6:4 and 5:5 ratios separately to see how much impact they have on the model.

Improve thinking abilities in programming and code. If any of the models meet your expectations, please give a thumbs up. This will help us finalize which model best meets everyone's expectations.

Model Details

Base Models:
- huihui-ai/DeepSeek-R1-Distill-Qwen-32B-abliterated (90%)
- huihui-ai/Qwen2.5-Coder-32B-Instruct-abliterated (10%)
Model Size: 32B parameters
Architecture: Qwen2.5
Mixing Ratio: 9:1 (DeepSeek-R1-Distill-Qwen-32B-abliterated:Qwen2.5-Coder-32B-Instruct-abliterated)

Key Features

DeepSeek-R1-Distill-Qwen-32B-abliterated(90%): This is an uncensored version of DeepSeek-R1-Distill-Qwen-32B created with abliteration.
Qwen2.5-Coder-32B-Instruct-abliterated Contributions (10%): This is an uncensored version of Qwen2.5-Coder-32B-Instruct created with abliteration.

Usage

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

from transformers import AutoModelForCausalLM, AutoTokenizer, BitsAndBytesConfig
import torch

# Load the model and tokenizer
model_name = "huihui-ai/DeepSeek-R1-Distill-Qwen-Coder-32B-Fusion-9010"
quant_config_4 = BitsAndBytesConfig(
    load_in_4bit=True,
    bnb_4bit_compute_dtype=torch.bfloat16,
    bnb_4bit_use_double_quant=True,
    llm_int8_enable_fp32_cpu_offload=True,
)

model = AutoModelForCausalLM.from_pretrained(
    model_name,
    trust_remote_code=True,
        torch_dtype=torch.bfloat16,
    quantization_config=quant_config_4,
        device_map="auto",
)
tokenizer = AutoTokenizer.from_pretrained(model_name, trust_remote_code=True)

# Initialize conversation context
initial_messages = [
    {"role": "system", "content": "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"Response: {response}")