jmpers-lig-auto-1-dsr1-l8b

This is a fine-tuned version of the DeepSeek-R1-Distill-Llama-8B model using the sa16feb24/jmpers1 dataset via Lightning AI. The model is merged with LoRA adapters for efficient deployment .

Model Details

• Base Model: deepseek-ai/DeepSeek-R1-Distill-Llama-8B
• Fine-tuning Method: LoRA (merged)
• Dataset: sa16feb24/jmpers1
• Parameters:

  LoraConfig(
      r=16,
      lora_alpha=32,
      target_modules=["q_proj", "v_proj"],
      lora_dropout=0.1,
      bias="none",
      task_type="CAUSAL_LM"
  )
  

Usage

from transformers import AutoModelForCausalLM, AutoTokenizer

model = AutoModelForCausalLM.from_pretrained( "sa16feb24/jmpers-lig-auto-1-dsr1-l8b", torch_dtype="auto" ) tokenizer = AutoTokenizer.from_pretrained( "sa16feb24/jmpers-lig-auto-1-dsr1-l8b" )

Example inference

inputs = tokenizer("How did the AI chatbot project improve HR efficiency?", return_tensors="pt") outputs = model.generate(**inputs, max_new_tokens=100) print(tokenizer.decode(outputs[0]))

Training Configuration 7

Hardware: NVIDIA T4 GPU (Colab) Optimizations: 8-bit quantization Gradient checkpointing Mixed precision (FP16)