diff --git "a/logs/l40-4gpu-7.txt" "b/logs/l40-4gpu-7.txt" --- "a/logs/l40-4gpu-7.txt" +++ "b/logs/l40-4gpu-7.txt" @@ -185,4 +185,12 @@ Qwen/Qwen2-72B-Instruct/checkpoint-560/rpp-1.02 metrics: {'meteor': 0.4774721644 Batch output: ['Old Geng raised his gun, squinted with one of his triangular eyes, and pulled the trigger. The hail of golden sparrows fell to the ground in a flurry, and bits of shot flew through the air, making a rustling sound as they struck the willow branches.'] Qwen/Qwen2-72B-Instruct/checkpoint-560/rpp-1.04 metrics: {'meteor': 0.47514647664115106, 'sacrebleu': {'score': 18.555137061787658, 'counts': [14548, 6669, 3688, 2158], 'totals': [27912, 26779, 25646, 24517], 'precisions': [52.12095156205216, 24.90384256320251, 14.38041020042112, 8.802055716441652], 'bp': 0.9216279674480742, 'sys_len': 27912, 'ref_len': 30190}, 'bleu_scores': {'bleu': 0.18555137061787666, 'precisions': [0.5212095156205216, 0.2490384256320251, 0.14380410200421118, 0.08802055716441652], 'brevity_penalty': 0.9216279674480742, 'length_ratio': 0.9245445511758861, 'translation_length': 27912, 'reference_length': 30190}, 'rouge_scores': {'rouge1': 0.5037027818899031, 'rouge2': 0.26351726356609145, 'rougeL': 0.4477341548477711, 'rougeLsum': 0.4474439809226646}, 'accuracy': 0.01323918799646955, 'correct_ids': [41, 77, 133, 193, 199, 272, 348, 364, 413, 567, 658, 712, 778, 893, 1011]} *** Evaluating with repetition_penalty: 1.06 - 0%| | 0/1133 [00:00