nohup: ignoring input
================================================================
Starting generation of multihop reasoning traces using Llama 3.1
Environment: irl_torch2
Dataset: /home/debarpanb1/TREA_2.0/pipeline/dataset_v5
================================================================
Loading meta-llama/Llama-3.1-8B-Instruct...

Loading checkpoint shards:   0%|          | 0/4 [00:00<?, ?it/s]
Loading checkpoint shards:  25%|██▌       | 1/4 [00:04<00:12,  4.24s/it]
Loading checkpoint shards:  50%|█████     | 2/4 [00:08<00:08,  4.16s/it]
Loading checkpoint shards:  75%|███████▌  | 3/4 [00:12<00:03,  3.95s/it]
Loading checkpoint shards: 100%|██████████| 4/4 [00:15<00:00,  3.72s/it]
Loading checkpoint shards: 100%|██████████| 4/4 [00:15<00:00,  3.85s/it]
The attention mask is not set and cannot be inferred from input because pad token is same as eos token. As a consequence, you may observe unexpected behavior. Please pass your input's `attention_mask` to obtain reliable results.
Model loaded on cuda:0

============================================================
Task: duration_gap
============================================================
  Processing duration_gap: 92 samples
    [RETRY] sample 4 attempt 1
    [RETRY] sample 7 attempt 1
    10/92 done
    20/92 done
    [RETRY] sample 21 attempt 1
    [RETRY] sample 27 attempt 1
    [RETRY] sample 27 attempt 2
    30/92 done
    40/92 done
    50/92 done
    60/92 done
    70/92 done
    [RETRY] sample 78 attempt 1
    80/92 done
    [RETRY] sample 80 attempt 1
    [RETRY] sample 80 attempt 2
    [RETRY] sample 86 attempt 1
    [RETRY] sample 88 attempt 1
    90/92 done
    [RETRY] sample 91 attempt 1
    [RETRY] sample 91 attempt 2
    Saved /home/debarpanb1/TREA_2.0/pipeline/dataset_v5/duration_gap/duration_gap_open_text.csv
    Saved /home/debarpanb1/TREA_2.0/pipeline/dataset_v5/duration_gap/duration_gap_mcq.csv
    Saved /home/debarpanb1/TREA_2.0/pipeline/dataset_v5/duration_gap/duration_gap_metadata.csv

============================================================
Task: temporal_arithmetic
============================================================
  Processing temporal_arithmetic: 91 samples
    10/91 done
    [RETRY] sample 12 attempt 1
    20/91 done
    [RETRY] sample 20 attempt 1
    [RETRY] sample 25 attempt 1
    30/91 done
    40/91 done
    50/91 done
    60/91 done
    70/91 done
    80/91 done
    90/91 done
    Saved /home/debarpanb1/TREA_2.0/pipeline/dataset_v5/temporal_arithmetic/temporal_arithmetic_open_text.csv
    Saved /home/debarpanb1/TREA_2.0/pipeline/dataset_v5/temporal_arithmetic/temporal_arithmetic_mcq.csv
    Saved /home/debarpanb1/TREA_2.0/pipeline/dataset_v5/temporal_arithmetic/temporal_arithmetic_metadata.csv

============================================================
Task: temporal_loudness
============================================================
  Processing temporal_loudness: 91 samples
    [RETRY] sample 3 attempt 1
    10/91 done
    [RETRY] sample 15 attempt 1
    [RETRY] sample 15 attempt 2
    [RETRY] sample 18 attempt 1
    [RETRY] sample 18 attempt 2
    20/91 done
    [RETRY] sample 28 attempt 1
    [RETRY] sample 28 attempt 2
    [RETRY] sample 29 attempt 1
    30/91 done
    [RETRY] sample 30 attempt 1
    [RETRY] sample 35 attempt 1
    40/91 done
    [RETRY] sample 43 attempt 1
    50/91 done
    [RETRY] sample 55 attempt 1
    [RETRY] sample 55 attempt 2
    [RETRY] sample 57 attempt 1
    60/91 done
    [RETRY] sample 61 attempt 1
    70/91 done
    80/91 done
    90/91 done
    Saved /home/debarpanb1/TREA_2.0/pipeline/dataset_v5/temporal_loudness/temporal_loudness_open_text.csv
    Saved /home/debarpanb1/TREA_2.0/pipeline/dataset_v5/temporal_loudness/temporal_loudness_mcq.csv
    Saved /home/debarpanb1/TREA_2.0/pipeline/dataset_v5/temporal_loudness/temporal_loudness_metadata.csv

============================================================
Task: multi_hop
============================================================
  Processing multi_hop: 78 samples
    [RETRY] sample 2 attempt 1
    [RETRY] sample 2 attempt 2
    [RETRY] sample 10 attempt 1
    10/78 done
    [RETRY] sample 17 attempt 1
    [RETRY] sample 19 attempt 1
    20/78 done
    30/78 done
    40/78 done
    [RETRY] sample 46 attempt 1
    50/78 done
    60/78 done
    70/78 done
    Saved /home/debarpanb1/TREA_2.0/pipeline/dataset_v5/multi_hop/multi_hop_open_text.csv
    Saved /home/debarpanb1/TREA_2.0/pipeline/dataset_v5/multi_hop/multi_hop_mcq.csv
    Saved /home/debarpanb1/TREA_2.0/pipeline/dataset_v5/multi_hop/multi_hop_metadata.csv

✓ All done!
================================================================
Completed generation of multihop reasoning traces.
================================================================