[16:44:43] longeval: T=16384 prompt_len=16360 max_new=24 n=100 greedy=1 gpus=0,1 [16:44:43] haystack: /root/Adhi/BluTrain/dist/Context_Parallelism/lendata/haystack/b2val_haystack.bin [16:44:43] fan-out: 2 GPU(s), NP=1 -> 4 concurrent job(s) [16:44:43] armA_c/niah_single already generated [16:44:43] l1/niah_single already generated [16:44:43] l2/niah_single already generated [16:44:43] l3/niah_single already generated [16:44:43] l3long/niah_single already generated [16:44:43] 1 job(s) to run [16:44:43] [gpu 0] l3long2k / niah_single [16:49:01] scoring l3long2k / niah_single === niah_single (prompt_len=16360, n=100) === accuracy: 43/100 = 43.0% depth acc (needle position in context) 0.00 0/20 0.0% 0.25 7/20 35.0% 0.50 7/20 35.0% 0.75 13/20 65.0% 1.00 16/20 80.0% --- 2 sample generations --- [MISS] expect=8056020 got= 8.\nQuestion: What is the magic number for yarrow?\n [HIT ] expect=5343902 got= 5343902.\n wrote decide_len_16360/scores/l3long2k_niah_single.csv [16:49:01] building summary === per-task accuracy % (secondary/descriptive: ~10 per depth cell) === task armA_c l1 l2 l3 l3long l3long2k niah_single 33.0 30.0 35.0 35.0 36.0 43.0 === PRIMARY: accuracy % pooled by depth across tasks (+-95% Wilson) === depth armA_c l1 l2 l3 l3long l3long2k 0.00 0.0+- 8.1 0.0+- 8.1 0.0+- 8.1 0.0+- 8.1 0.0+- 8.1 0.0+- 8.1 0.25 0.0+- 8.1 0.0+- 8.1 0.0+- 8.1 5.0+-11.4 15.0+-15.4 35.0+-19.3 0.50 20.0+-16.8 5.0+-11.4 25.0+-17.8 30.0+-18.7 35.0+-19.3 35.0+-19.3 0.75 55.0+-20.0 60.0+-19.7 60.0+-19.7 55.0+-20.0 50.0+-20.1 65.0+-19.3 1.00 90.0+-13.7 85.0+-15.4 90.0+-13.7 85.0+-15.4 80.0+-16.8 80.0+-16.8 wrote decide_len_16360/summary.csv and decide_len_16360/depth_summary.csv [16:49:01] DONE -> decide_len_16360/summary.csv