Hugging Face's logo

Dans-DiscountModels
/
TinyLlama-1.1B-FFT-Test2

Text Generation
Transformers
PyTorch
llama
Generated from Trainer
text-generation-inference
Inference Endpoints
Model card Files Community
1
TinyLlama-1.1B-FFT-Test2 / README.md
leaderboard-pr-bot's picture
leaderboard-pr-bot
Adding Evaluation Results
a4d107b verified
preview code
|
raw
history blame
4.27 kB
metadata
license: apache-2.0
tags:
  - generated_from_trainer
base_model: TinyLlama/TinyLlama-1.1B-intermediate-step-1195k-token-2.5T
model-index:
  - name: out
    results: []

Built with Axolotl

out

This model is a fine-tuned version of TinyLlama/TinyLlama-1.1B-intermediate-step-1195k-token-2.5T on the None dataset. It achieves the following results on the evaluation set:

  • Loss: 1.9391

Model description

More information needed

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: 2.5e-05
  • train_batch_size: 6
  • eval_batch_size: 6
  • seed: 42
  • gradient_accumulation_steps: 6
  • total_train_batch_size: 36
  • optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
  • lr_scheduler_type: constant
  • lr_scheduler_warmup_steps: 300
  • num_epochs: 1

Training results

Training Loss Epoch Step Validation Loss
2.388 0.03 30 2.3495
2.3877 0.05 60 2.2353
2.1812 0.08 90 2.1786
2.1711 0.1 120 2.1453
2.098 0.13 150 2.1210
2.0884 0.15 180 2.1037
2.0129 0.18 210 2.0894
2.0921 0.2 240 2.0765
2.001 0.23 270 2.0648
2.0986 0.25 300 2.0550
2.0128 0.28 330 2.0458
2.1144 0.3 360 2.0385
2.0711 0.33 390 2.0308
1.9607 0.36 420 2.0239
1.9092 0.38 450 2.0179
2.1955 0.41 480 2.0121
2.0293 0.43 510 2.0063
2.0482 0.46 540 2.0017
2.0518 0.48 570 2.0036
1.9647 0.51 600 1.9984
2.0573 0.53 630 1.9938
1.9912 0.56 660 1.9879
1.9836 0.58 690 1.9866
1.9514 0.61 720 1.9834
1.9413 0.63 750 1.9792
1.9775 0.66 780 1.9726
1.9499 0.69 810 1.9690
2.0497 0.71 840 1.9663
2.0045 0.74 870 1.9636
2.0324 0.76 900 1.9605
1.9913 0.79 930 1.9574
1.9861 0.81 960 1.9553
1.8449 0.84 990 1.9527
1.9932 0.86 1020 1.9502
1.9845 0.89 1050 1.9485
2.0407 0.91 1080 1.9494
1.9496 0.94 1110 1.9448
1.9851 0.97 1140 1.9406
1.9856 0.99 1170 1.9391

Framework versions

  • Transformers 4.35.2
  • Pytorch 2.0.1+cu117
  • Datasets 2.15.0
  • Tokenizers 0.15.0

Open LLM Leaderboard Evaluation Results

Detailed results can be found here

Metric Value
Avg. 35.53
AI2 Reasoning Challenge (25-Shot) 34.22
HellaSwag (10-Shot) 57.96
MMLU (5-Shot) 25.54
TruthfulQA (0-shot) 36.32
Winogrande (5-shot) 58.80
GSM8k (5-shot) 0.38