davelotito
/

donut_experiment_6

Image-Text-to-Text

vision-encoder-decoder

Model card Files Files and versions

Metrics Training metrics Community

donut_experiment_6 / metrics.jsonl

davelotito's picture

Training in progress, epoch 2

c732297 verified almost 2 years ago

history blame contribute delete

873 Bytes

	{"eval_loss": 1.865431785583496, "eval_bleu": 0.06996470821651866, "eval_precisions": [0.625, 0.5714285714285714, 0.5, 0.4], "eval_brevity_penalty": 0.1353352832366127, "eval_length_ratio": 0.3333333333333333, "eval_translation_length": 8, "eval_reference_length": 24, "eval_cer": 0.8220338983050848, "eval_wer": 0.9411764705882353, "eval_runtime": 1.2941, "eval_samples_per_second": 0.773, "eval_steps_per_second": 0.773, "epoch": 1.0}
	{"eval_loss": 1.8447256088256836, "eval_bleu": 0.0, "eval_precisions": [0.5, 0.2857142857142857, 0.16666666666666666, 0.0], "eval_brevity_penalty": 0.1353352832366127, "eval_length_ratio": 0.3333333333333333, "eval_translation_length": 8, "eval_reference_length": 24, "eval_cer": 0.8135593220338984, "eval_wer": 0.9411764705882353, "eval_runtime": 1.2838, "eval_samples_per_second": 0.779, "eval_steps_per_second": 0.779, "epoch": 2.0}