README.md

metadata

license: mit
tags:
  - generated_from_trainer
model-index:
  - name: gpt-m-large
    results: []

gpt-m-large

This model is a fine-tuned version of augustocsc/gpt-m-large on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.0327

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: 5e-05
• train_batch_size: 64
• eval_batch_size: 64
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 1

Training results

Training Loss	Epoch	Step	Validation Loss
0.0343	0.03	1000	0.0343
0.0337	0.06	2000	0.0342
0.0338	0.09	3000	0.0338
0.0349	0.13	4000	0.0337
0.034	0.16	5000	0.0335
0.0342	0.19	6000	0.0334
0.0341	0.22	7000	0.0333
0.0339	0.25	8000	0.0333
0.0336	0.28	9000	0.0331
0.0335	0.31	10000	0.0330
0.0334	0.35	11000	0.0330
0.0331	0.38	12000	0.0328
0.0332	0.41	13000	0.0328
0.0327	0.44	14000	0.0327
0.0331	0.47	15000	0.0327
0.0335	0.5	16000	0.0327
0.0333	0.53	17000	0.0327
0.0333	0.57	18000	0.0327
0.0333	0.6	19000	0.0327
0.0332	0.63	20000	0.0327
0.0331	0.66	21000	0.0327
0.0328	0.69	22000	0.0327
0.033	0.72	23000	0.0327
0.0334	0.75	24000	0.0327
0.0334	0.79	25000	0.0327
0.0333	0.82	26000	0.0327
0.0332	0.85	27000	0.0327
0.033	0.88	28000	0.0327
0.033	0.91	29000	0.0327
0.0331	0.94	30000	0.0327
0.0335	0.97	31000	0.0327

Framework versions

• Transformers 4.27.3
• Pytorch 2.0.0+cu117
• Datasets 2.10.1
• Tokenizers 0.13.2