tinyllama-lora-fast

Trained primarily on Prothom Alo news data, this model naturally writes in that newspaper’s concise, reportorial Bangla style.

This model is a fine-tuned version of TinyLlama/TinyLlama-1.1B-Chat-v1.0 Bangla Newspaper Dataset. The single goal: enable fast training and low VRAM usage—even on small GPUs or TPUs.Use Dynamic Padding + Token Budget for low Gpu And Tpu It achieves the following results on the evaluation set:

• Loss: 0.8744

Model description

Base model: TinyLlama/TinyLlama-1.1B-Chat-v1.0

Adapter: PEFT-LoRA (r=16, alpha=32, dropout=0.05, bias=none)

Target modules: q_proj, k_proj, v_proj, o_proj, gate_proj, up_proj, down_proj

Context length (train max): MAX_LEN=768

Optimization: AdamW (lr≈3e-5, warmup≈0.03, weight_decay≈0.05)

Batching: length-based bucketing + dynamic padding

Precision: fp16 inference-ready (training setup Kaggle/Colab-friendly)

Decoding (low hallucination preset): temperature=0.0, do_sample=False, no_repeat_ngram_size=4, repetition_penalty≈1.1–1.15

Intended uses & limitations

More information needed

Training and evaluation data

More information needed

Training procedure

Training hyperparameters

The following hyperparameters were used during training:

• learning_rate: 3e-05
• train_batch_size: 2
• eval_batch_size: 2
• seed: 42
• gradient_accumulation_steps: 2
• total_train_batch_size: 4
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: cosine
• lr_scheduler_warmup_ratio: 0.03
• num_epochs: 2
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss
1.0752	0.1111	200	1.0674
1.0248	0.2222	400	1.0167
0.9839	0.3333	600	0.9843
0.9479	0.4444	800	0.9620
0.9478	0.5556	1000	0.9457
0.914	0.6667	1200	0.9318
0.9329	0.7778	1400	0.9182
0.8929	0.8889	1600	0.9104
0.911	1.0	1800	0.9005
0.869	1.1111	2000	0.8949
0.8873	1.2222	2200	0.8892
0.8551	1.3333	2400	0.8848
0.8543	1.4444	2600	0.8812
0.8741	1.5556	2800	0.8779
0.8652	1.6667	3000	0.8760
0.8534	1.7778	3200	0.8751
0.8378	1.8889	3400	0.8745
0.8514	2.0	3600	0.8744

Framework versions

• PEFT 0.12.0
• Transformers 4.45.2
• Pytorch 2.6.0+cu124
• Datasets 4.1.1
• Tokenizers 0.20.3