akp4u/amit-gc-trained-model

Amit-GC Trained Model

This is a LoRA fine-tuned version of EleutherAI/pythia-410m, trained on Grey Chain content about AI and digital transformation.

Model Details

• Base Model: EleutherAI/pythia-410m (410M parameters)
• Training Method: LoRA (Low-Rank Adaptation)
• Training Data: Grey Chain content about AI & digital transformation
• Trainable Parameters: ~0.77% of base model

Usage

from transformers import AutoTokenizer, AutoModelForCausalLM
from peft import PeftModel
import torch

# Load tokenizer and base model
tokenizer = AutoTokenizer.from_pretrained("akp4u/amit-gc-trained-model")
base_model = AutoModelForCausalLM.from_pretrained(
    "EleutherAI/pythia-410m",
    device_map="cpu",
    torch_dtype=torch.float32
)

# Load LoRA adapters
model = PeftModel.from_pretrained(base_model, "akp4u/amit-gc-trained-model")

# Generate text
prompt = "What is Grey Chain?"
inputs = tokenizer(prompt, return_tensors="pt")
outputs = model.generate(**inputs, max_new_tokens=100, do_sample=True, temperature=0.7)
response = tokenizer.decode(outputs[0], skip_special_tokens=True)
print(response)

Training Details

• Trained for 3 epochs on Grey Chain content
• Uses LoRA with r=8, alpha=16, dropout=0.05
• Optimized for Mac/CPU inference (no quantization required)

Sample Outputs

The model can answer questions about:

• Grey Chain services and capabilities
• AI and digital transformation
• Machine learning concepts
• Prompt engineering

Limitations

• Small model (410M parameters) - responses may be limited
• Trained on specific domain content
• Best for Grey Chain and AI-related queries