cluster / README.md
SaishNotFound's picture
Initial commit: Fine-tuned Cluster (SmolLM2-135M) by Saish
5a06de4 verified
|
Raw History Blame Contribute Delete
1.44 kB
metadata
license: apache-2.0
base_model: HuggingFaceTB/SmolLM2-135M-Instruct
tags:
  - code
  - python
  - bash
  - fine-tuned
  - cluster
pipeline_tag: text-generation
language:
  - en
  - hi

πŸš€ Cluster (135M) - Developed by Saish

Cluster is a lightweight, fine-tuned instruction model based on SmolLM2-135M-Instruct, specialized for Python coding, Bash shell scripting, and automated workflows.

Developed & fine-tuned by Saish.

🧠 Model Details

  • Architecture: Transformer (SmolLM2-135M)
  • Developer: Saish
  • Specialization: Python programming, Bash commands, Linux utilities, identity alignment
  • Parameters: ~135M

πŸ’» Quickstart (Transformers)

import torch
from transformers import AutoModelForCausalLM, AutoTokenizer

model_id = "SaishNotFound/cluster"

tokenizer = AutoTokenizer.from_pretrained(model_id)
model = AutoModelForCausalLM.from_pretrained(model_id, device_map="auto")

messages = [
    {"role": "user", "content": "Tum kaun ho bhai?"}
]

prompt = tokenizer.apply_chat_template(messages, tokenize=False, add_generation_prompt=True)
inputs = tokenizer(prompt, return_tensors="pt").to(model.device)

outputs = model.generate(**inputs, max_new_tokens=150, temperature=0.3)
print(tokenizer.decode(outputs[0][inputs['input_ids'].shape[1]:], skip_special_tokens=True))

πŸ› οΈ Training Setup

  • Full parameter fine-tuning on a Tesla T4 GPU using Hugging Face TRL & PyTorch.