Trinoid/Deepseek_Chatbot

DeepSeek Chatbot

This is a fine-tuned version of DeepSeek-R1-Distill-Qwen-1.5B, optimized for conversational AI applications. The model maintains the base model's capabilities while being tuned for improved dialogue interactions.

Model Details

Model Description

• Developed by: Trinoid
• Model type: Conversational Language Model
• Language(s): English
• License: Same as base model (DeepSeek-R1-Distill-Qwen-1.5B)
• Finetuned from model: deepseek-ai/DeepSeek-R1-Distill-Qwen-1.5B

Uses

Direct Use

This model can be used for:

• General conversation
• Text generation
• Question answering
• Chat-based applications

Example usage:

from huggingface_hub import InferenceClient

client = InferenceClient("Trinoid/Deepseek_Chatbot")

messages = [
    {"role": "system", "content": "You are a helpful assistant."},
    {"role": "user", "content": "Hello, how are you?"}
]

response = client.chat_completion(
    messages,
    max_tokens=512,
    temperature=0.7,
    top_p=0.95
)

Out-of-Scope Use

This model should not be used for:

• Generation of harmful or malicious content
• Spreading misinformation
• Production of illegal content
• Making critical decisions without human oversight

Training Details

Training Procedure

Training Hyperparameters

• Training regime: fp16 mixed precision
• Framework: PEFT (Parameter-Efficient Fine-Tuning)
• PEFT Method: LoRA
• Version: PEFT 0.14.0

Technical Specifications

Model Architecture and Objective

• Base architecture: DeepSeek-R1-Distill-Qwen-1.5B
• Fine-tuning method: PEFT/LoRA
• Primary objective: Conversational AI

Compute Infrastructure

Software

• PEFT 0.14.0
• Transformers
• Python 3.x

Model Card Contact

For questions or issues about this model, please open an issue in the model repository.