README.md

metadata

language:
  - en
  - hi
license: apache-2.0
tags:
  - text-generation-inference
  - transformers
  - unsloth
  - llama
  - trl
base_model: TinyLlama/TinyLlama-1.1B-Chat-v1.0
datasets:
  - yahma/alpaca-cleaned
  - ravithejads/samvaad-hi-filtered
  - HydraIndicLM/hindi_alpaca_dolly_67k

TinyLlama-1.1B-Hinglish-LORA-v1.0 model

• Developed by: Kiran Kunapuli
• License: apache-2.0
• Finetuned from model : unsloth/gemma-2b-bnb-4bit
• • Model config:

    model = FastLanguageModel.get_peft_model(
    model,
    r = 16, 
    target_modules = ["q_proj", "k_proj", "v_proj", "o_proj",
                      "gate_proj", "up_proj", "down_proj",],
    lora_alpha = 32,
    lora_dropout = 0, 
    bias = "none",   
    use_gradient_checkpointing = True, 
    random_state = 42,
    use_rslora = True,  
    loftq_config = None, 
    )
  

• Training parameters:

    trainer = SFTTrainer(
    model = model,
    tokenizer = tokenizer,
    train_dataset = dataset,
    dataset_text_field = "text",
    max_seq_length = max_seq_length,
    dataset_num_proc = 2,
    packing = True,
    args = TrainingArguments(
        per_device_train_batch_size = 2,
        gradient_accumulation_steps = 4,
        warmup_steps = 5,
        max_steps = 120,
        learning_rate = 2e-4,
        fp16 = not torch.cuda.is_bf16_supported(),
        bf16 = torch.cuda.is_bf16_supported(),
        logging_steps = 1,
        optim = "adamw_8bit",
        weight_decay = 0.01,
        lr_scheduler_type = "linear",
        seed = 42,
        output_dir = "outputs",
        report_to = "wandb",
      ),
    )
  

• Training details:

  ==((====))==  Unsloth - 2x faster free finetuning | Num GPUs = 1
     \\   /|    Num examples = 14,343 | Num Epochs = 1
  O^O/ \_/ \    Batch size per device = 2 | Gradient Accumulation steps = 4
  \        /    Total batch size = 8 | Total steps = 120
   "-____-"     Number of trainable parameters = 19,611,648
  
  GPU = Tesla T4. Max memory = 14.748 GB.
  2118.7553 seconds used for training.
  35.31 minutes used for training.
  Peak reserved memory = 9.172 GB.
  Peak reserved memory for training = 6.758 GB.
  Peak reserved memory % of max memory = 62.191 %.
  Peak reserved memory for training % of max memory = 45.823 %.
  

This gemma model was trained 2x faster with Unsloth and Huggingface's TRL library.