|
2025-02-09:12:49:52,916 INFO [__main__.py:284] Verbosity set to INFO |
|
2025-02-09:12:50:05,514 INFO [__main__.py:381] Selected Tasks: ['gsm8k_cot_llama'] |
|
2025-02-09:12:50:05,516 INFO [evaluator.py:165] Setting random seed to 0 | Setting numpy seed to 1234 | Setting torch manual seed to 1234 | Setting fewshot manual seed to 1234 |
|
2025-02-09:12:50:05,516 WARNING [evaluator.py:176] generation_kwargs specified through cli, these settings will update set parameters in yaml tasks. Ensure 'do_sample=True' for non-greedy decoding! |
|
2025-02-09:12:50:05,516 INFO [evaluator.py:202] Initializing hf model, with arguments: {'pretrained': 'meta-llama/Llama-3.2-3B-Instruct', 'peft': 'JaqueMate/new_config_3b_only_dpo_checkpoint-1600'} |
|
2025-02-09:12:50:05,563 INFO [huggingface.py:135] Using device 'cuda:0' |
|
2025-02-09:12:50:06,334 INFO [huggingface.py:375] Model parallel was set to False, max memory was not set, and device map was set to {'': 'cuda:0'} |
|
Loading checkpoint shards: 0%| | 0/2 [00:00<?, ?it/s]
Loading checkpoint shards: 50%|█████ | 1/2 [00:01<00:01, 1.82s/it]
Loading checkpoint shards: 100%|██████████| 2/2 [00:02<00:00, 1.12s/it]
Loading checkpoint shards: 100%|██████████| 2/2 [00:02<00:00, 1.22s/it] |
|
2025-02-09:12:50:34,647 WARNING [evaluator.py:271] Overwriting default num_fewshot of gsm8k_cot_llama from 8 to 8 |
|
2025-02-09:12:50:34,648 WARNING [evaluator.py:417] Chat template formatting change affects loglikelihood and multiple-choice tasks. See docs/chat-template-readme.md for details. |
|
2025-02-09:12:50:34,648 INFO [task.py:420] Building contexts for gsm8k_cot_llama on rank 0... |
|
0%| | 0/1319 [00:00<?, ?it/s]
1%| | 9/1319 [00:00<00:15, 85.79it/s]
1%|▏ | 19/1319 [00:00<00:14, 92.61it/s]
2%|▏ | 30/1319 [00:00<00:13, 94.52it/s]
3%|▎ | 40/1319 [00:00<00:13, 94.11it/s]
4%|▍ | 51/1319 [00:00<00:12, 98.08it/s]
5%|▍ | 63/1319 [00:00<00:12, 102.29it/s]
6%|▌ | 75/1319 [00:00<00:11, 105.09it/s]
7%|▋ | 87/1319 [00:00<00:11, 107.44it/s]
8%|▊ | 99/1319 [00:00<00:11, 109.00it/s]
8%|▊ | 111/1319 [00:01<00:11, 109.74it/s]
9%|▉ | 123/1319 [00:01<00:10, 109.99it/s]
10%|█ | 135/1319 [00:01<00:10, 110.68it/s]
11%|█ | 147/1319 [00:01<00:10, 110.76it/s]
12%|█▏ | 159/1319 [00:01<00:10, 111.11it/s]
13%|█▎ | 171/1319 [00:01<00:10, 111.28it/s]
14%|█▍ | 183/1319 [00:01<00:10, 110.62it/s]
15%|█▍ | 195/1319 [00:01<00:10, 111.03it/s]
16%|█▌ | 207/1319 [00:01<00:10, 110.77it/s]
17%|█▋ | 219/1319 [00:02<00:09, 110.47it/s]
18%|█▊ | 231/1319 [00:02<00:09, 111.32it/s]
18%|█▊ | 243/1319 [00:02<00:09, 111.55it/s]
19%|█▉ | 255/1319 [00:02<00:09, 112.07it/s]
20%|██ | 267/1319 [00:02<00:09, 111.95it/s]
21%|██ | 279/1319 [00:02<00:09, 112.14it/s]
22%|██▏ | 291/1319 [00:02<00:09, 111.99it/s]
23%|██▎ | 303/1319 [00:02<00:09, 111.87it/s]
24%|██▍ | 315/1319 [00:02<00:08, 111.75it/s]
25%|██▍ | 327/1319 [00:03<00:08, 111.25it/s]
26%|██▌ | 339/1319 [00:03<00:08, 111.41it/s]
27%|██▋ | 351/1319 [00:03<00:08, 110.97it/s]
28%|██▊ | 363/1319 [00:03<00:08, 110.78it/s]
28%|██▊ | 375/1319 [00:03<00:08, 110.92it/s]
29%|██▉ | 387/1319 [00:03<00:08, 109.70it/s]
30%|███ | 399/1319 [00:03<00:08, 110.19it/s]
31%|███ | 411/1319 [00:03<00:08, 109.33it/s]
32%|███▏ | 422/1319 [00:03<00:08, 109.17it/s]
33%|███▎ | 433/1319 [00:03<00:08, 105.49it/s]
34%|███▎ | 444/1319 [00:04<00:08, 105.22it/s]
35%|███▍ | 456/1319 [00:04<00:08, 107.25it/s]
35%|███▌ | 468/1319 [00:04<00:07, 108.68it/s]
36%|███▋ | 479/1319 [00:04<00:07, 108.02it/s]
37%|███▋ | 490/1319 [00:04<00:07, 107.47it/s]
38%|███▊ | 501/1319 [00:04<00:07, 106.79it/s]
39%|███▉ | 512/1319 [00:04<00:07, 107.17it/s]
40%|███▉ | 523/1319 [00:04<00:07, 107.68it/s]
40%|████ | 534/1319 [00:04<00:07, 108.30it/s]
41%|████▏ | 545/1319 [00:05<00:07, 107.72it/s]
42%|████▏ | 556/1319 [00:05<00:07, 106.23it/s]
43%|████▎ | 567/1319 [00:05<00:07, 105.88it/s]
44%|████▍ | 578/1319 [00:05<00:06, 106.39it/s]
45%|████▍ | 589/1319 [00:05<00:06, 105.76it/s]
45%|████▌ | 600/1319 [00:05<00:06, 105.65it/s]
46%|████▋ | 611/1319 [00:05<00:06, 106.64it/s]
47%|████▋ | 623/1319 [00:05<00:06, 108.65it/s]
48%|████▊ | 635/1319 [00:05<00:06, 109.95it/s]
49%|████▉ | 647/1319 [00:05<00:06, 110.21it/s]
50%|████▉ | 659/1319 [00:06<00:05, 110.77it/s]
51%|█████ | 671/1319 [00:06<00:05, 111.55it/s]
52%|█████▏ | 683/1319 [00:06<00:05, 111.58it/s]
53%|█████▎ | 695/1319 [00:06<00:05, 109.52it/s]
54%|█████▎ | 707/1319 [00:06<00:05, 110.47it/s]
55%|█████▍ | 719/1319 [00:06<00:05, 111.57it/s]
55%|█████▌ | 731/1319 [00:06<00:05, 110.51it/s]
56%|█████▋ | 743/1319 [00:06<00:05, 110.70it/s]
57%|█████▋ | 755/1319 [00:06<00:05, 111.54it/s]
58%|█████▊ | 767/1319 [00:07<00:04, 112.99it/s]
59%|█████▉ | 779/1319 [00:07<00:04, 113.11it/s]
60%|█████▉ | 791/1319 [00:07<00:04, 114.19it/s]
61%|██████ | 803/1319 [00:07<00:04, 115.39it/s]
62%|██████▏ | 815/1319 [00:07<00:04, 116.28it/s]
63%|██████▎ | 827/1319 [00:07<00:04, 116.51it/s]
64%|██████▎ | 839/1319 [00:07<00:04, 116.40it/s]
65%|██████▍ | 851/1319 [00:07<00:04, 115.94it/s]
65%|██████▌ | 863/1319 [00:07<00:03, 116.62it/s]
66%|██████▋ | 875/1319 [00:07<00:03, 117.14it/s]
67%|██████▋ | 887/1319 [00:08<00:03, 117.49it/s]
68%|██████▊ | 899/1319 [00:08<00:03, 117.51it/s]
69%|██████▉ | 911/1319 [00:08<00:03, 117.73it/s]
70%|██████▉ | 923/1319 [00:08<00:03, 117.85it/s]
71%|███████ | 935/1319 [00:08<00:03, 117.70it/s]
72%|███████▏ | 947/1319 [00:08<00:03, 117.68it/s]
73%|███████▎ | 959/1319 [00:08<00:03, 117.88it/s]
74%|███████▎ | 971/1319 [00:08<00:02, 117.98it/s]
75%|███████▍ | 983/1319 [00:08<00:02, 118.17it/s]
75%|███████▌ | 995/1319 [00:08<00:02, 117.62it/s]
76%|███████▋ | 1007/1319 [00:09<00:02, 117.14it/s]
77%|███████▋ | 1019/1319 [00:09<00:02, 117.54it/s]
78%|███████▊ | 1031/1319 [00:09<00:02, 118.02it/s]
79%|███████▉ | 1043/1319 [00:09<00:02, 118.27it/s]
80%|███████▉ | 1055/1319 [00:09<00:02, 118.49it/s]
81%|████████ | 1067/1319 [00:09<00:02, 118.31it/s]
82%|████████▏ | 1079/1319 [00:09<00:02, 117.67it/s]
83%|████████▎ | 1091/1319 [00:09<00:01, 117.96it/s]
84%|████████▎ | 1103/1319 [00:09<00:01, 117.69it/s]
85%|████████▍ | 1115/1319 [00:10<00:01, 117.63it/s]
85%|████████▌ | 1127/1319 [00:10<00:01, 117.53it/s]
86%|████████▋ | 1139/1319 [00:10<00:01, 117.96it/s]
87%|████████▋ | 1151/1319 [00:10<00:01, 118.26it/s]
88%|████████▊ | 1163/1319 [00:10<00:01, 118.41it/s]
89%|████████▉ | 1175/1319 [00:10<00:01, 118.61it/s]
90%|████████▉ | 1187/1319 [00:10<00:01, 114.98it/s]
91%|█████████ | 1199/1319 [00:10<00:01, 114.69it/s]
92%|█████████▏| 1211/1319 [00:10<00:00, 114.94it/s]
93%|█████████▎| 1223/1319 [00:10<00:00, 114.81it/s]
94%|█████████▎| 1235/1319 [00:11<00:00, 114.63it/s]
95%|█████████▍| 1247/1319 [00:11<00:00, 114.84it/s]
95%|█████████▌| 1259/1319 [00:11<00:00, 113.97it/s]
96%|█████████▋| 1271/1319 [00:11<00:00, 113.39it/s]
97%|█████████▋| 1283/1319 [00:11<00:00, 113.04it/s]
98%|█████████▊| 1295/1319 [00:11<00:00, 112.71it/s]
99%|█████████▉| 1307/1319 [00:11<00:00, 112.61it/s]
100%|██████████| 1319/1319 [00:11<00:00, 113.02it/s]
100%|██████████| 1319/1319 [00:11<00:00, 111.91it/s] |
|
2025-02-09:12:50:46,474 INFO [evaluator.py:513] Running generate_until requests |
|
Running generate_until requests: 0%| | 0/1319 [00:00<?, ?it/s]/home/luciano/Documents/Tesis Ezequiel/Tesis/env/lib/python3.10/site-packages/transformers/generation/configuration_utils.py:567: UserWarning: `do_sample` is set to `False`. However, `temperature` is set to `0.01` |
|
warnings.warn( |
|
/home/luciano/Documents/Tesis Ezequiel/Tesis/env/lib/python3.10/site-packages/transformers/generation/configuration_utils.py:589: UserWarning: `do_sample` is set to `False`. However, `top_k` is set to `0` |
|
warnings.warn( |
|
Running generate_until requests: 0%| | 1/1319 [00:20<7:36:42, 20.79s/it]
Running generate_until requests: 1%|▏ | 17/1319 [00:38<42:07, 1.94s/it]
Running generate_until requests: 3%|▎ | 33/1319 [00:56<31:23, 1.46s/it]
Running generate_until requests: 4%|▎ | 49/1319 [01:14<27:36, 1.30s/it]
Running generate_until requests: 5%|▍ | 65/1319 [01:31<25:38, 1.23s/it]
Running generate_until requests: 6%|▌ | 81/1319 [01:49<24:23, 1.18s/it]
Running generate_until requests: 7%|▋ | 97/1319 [02:07<23:31, 1.15s/it]
Running generate_until requests: 9%|▊ | 113/1319 [02:24<22:50, 1.14s/it]
Running generate_until requests: 10%|▉ | 129/1319 [02:42<22:18, 1.12s/it]
Running generate_until requests: 11%|█ | 145/1319 [02:59<21:50, 1.12s/it]
Running generate_until requests: 12%|█▏ | 161/1319 [03:17<21:24, 1.11s/it]
Running generate_until requests: 13%|█▎ | 177/1319 [03:35<21:01, 1.10s/it]
Running generate_until requests: 15%|█▍ | 193/1319 [03:52<20:39, 1.10s/it]
Running generate_until requests: 16%|█▌ | 209/1319 [04:09<20:19, 1.10s/it]
Running generate_until requests: 17%|█▋ | 225/1319 [04:27<19:59, 1.10s/it]
Running generate_until requests: 18%|█▊ | 241/1319 [04:43<19:09, 1.07s/it]
Running generate_until requests: 19%|█▉ | 257/1319 [05:00<18:58, 1.07s/it]
Running generate_until requests: 21%|██ | 273/1319 [05:18<18:46, 1.08s/it]
Running generate_until requests: 22%|██▏ | 289/1319 [05:33<17:55, 1.04s/it]
Running generate_until requests: 23%|██▎ | 305/1319 [05:48<16:55, 1.00s/it]
Running generate_until requests: 24%|██▍ | 321/1319 [06:05<17:05, 1.03s/it]
Running generate_until requests: 26%|██▌ | 337/1319 [06:22<17:06, 1.05s/it]
Running generate_until requests: 27%|██▋ | 353/1319 [06:37<16:17, 1.01s/it]
Running generate_until requests: 28%|██▊ | 369/1319 [06:55<16:22, 1.03s/it]
Running generate_until requests: 29%|██▉ | 385/1319 [07:12<16:20, 1.05s/it]
Running generate_until requests: 30%|███ | 401/1319 [07:29<16:12, 1.06s/it]
Running generate_until requests: 32%|███▏ | 417/1319 [07:45<15:35, 1.04s/it]
Running generate_until requests: 33%|███▎ | 433/1319 [08:03<15:33, 1.05s/it]
Running generate_until requests: 34%|███▍ | 449/1319 [08:17<14:35, 1.01s/it]
Running generate_until requests: 35%|███▌ | 465/1319 [08:34<14:41, 1.03s/it]
Running generate_until requests: 36%|███▋ | 481/1319 [08:52<14:37, 1.05s/it]
Running generate_until requests: 38%|███▊ | 497/1319 [09:09<14:29, 1.06s/it]
Running generate_until requests: 39%|███▉ | 513/1319 [09:24<13:38, 1.02s/it]
Running generate_until requests: 40%|████ | 529/1319 [09:41<13:39, 1.04s/it]
Running generate_until requests: 41%|████▏ | 545/1319 [09:56<13:02, 1.01s/it]
Running generate_until requests: 43%|████▎ | 561/1319 [10:14<13:00, 1.03s/it]
Running generate_until requests: 44%|████▎ | 577/1319 [10:31<12:55, 1.04s/it]
Running generate_until requests: 45%|████▍ | 593/1319 [10:48<12:47, 1.06s/it]
Running generate_until requests: 46%|████▌ | 609/1319 [11:05<12:35, 1.06s/it]
Running generate_until requests: 47%|████▋ | 625/1319 [11:21<11:54, 1.03s/it]
Running generate_until requests: 49%|████▊ | 641/1319 [11:38<11:49, 1.05s/it]
Running generate_until requests: 50%|████▉ | 657/1319 [11:54<11:25, 1.04s/it]
Running generate_until requests: 51%|█████ | 673/1319 [12:11<11:11, 1.04s/it]
Running generate_until requests: 52%|█████▏ | 689/1319 [12:27<10:46, 1.03s/it]
Running generate_until requests: 53%|█████▎ | 705/1319 [12:44<10:40, 1.04s/it]
Running generate_until requests: 55%|█████▍ | 721/1319 [13:00<10:14, 1.03s/it]
Running generate_until requests: 56%|█████▌ | 737/1319 [13:17<10:07, 1.04s/it]
Running generate_until requests: 57%|█████▋ | 753/1319 [13:31<09:16, 1.02it/s]
Running generate_until requests: 58%|█████▊ | 769/1319 [13:48<09:15, 1.01s/it]
Running generate_until requests: 60%|█████▉ | 785/1319 [14:04<08:56, 1.00s/it]
Running generate_until requests: 61%|██████ | 801/1319 [14:18<08:22, 1.03it/s]
Running generate_until requests: 62%|██████▏ | 817/1319 [14:35<08:17, 1.01it/s]
Running generate_until requests: 63%|██████▎ | 833/1319 [14:52<08:13, 1.02s/it]
Running generate_until requests: 64%|██████▍ | 849/1319 [15:07<07:47, 1.01it/s]
Running generate_until requests: 66%|██████▌ | 865/1319 [15:24<07:43, 1.02s/it]
Running generate_until requests: 67%|██████▋ | 881/1319 [15:41<07:28, 1.02s/it]
Running generate_until requests: 68%|██████▊ | 897/1319 [15:58<07:19, 1.04s/it]
Running generate_until requests: 69%|██████▉ | 913/1319 [16:12<06:37, 1.02it/s]
Running generate_until requests: 70%|███████ | 929/1319 [16:26<06:10, 1.05it/s]
Running generate_until requests: 72%|███████▏ | 945/1319 [16:41<05:58, 1.04it/s]
Running generate_until requests: 73%|███████▎ | 961/1319 [16:54<05:24, 1.10it/s]
Running generate_until requests: 74%|███████▍ | 977/1319 [17:10<05:22, 1.06it/s]
Running generate_until requests: 75%|███████▌ | 993/1319 [17:24<05:00, 1.08it/s]
Running generate_until requests: 76%|███████▋ | 1009/1319 [17:41<04:57, 1.04it/s]
Running generate_until requests: 78%|███████▊ | 1025/1319 [17:55<04:33, 1.08it/s]
Running generate_until requests: 79%|███████▉ | 1041/1319 [18:09<04:14, 1.09it/s]
Running generate_until requests: 80%|████████ | 1057/1319 [18:23<03:54, 1.12it/s]
Running generate_until requests: 81%|████████▏ | 1073/1319 [18:40<03:53, 1.06it/s]
Running generate_until requests: 83%|████████▎ | 1089/1319 [18:54<03:36, 1.06it/s]
Running generate_until requests: 84%|████████▍ | 1105/1319 [19:05<03:05, 1.15it/s]
Running generate_until requests: 85%|████████▍ | 1121/1319 [19:21<02:59, 1.10it/s]
Running generate_until requests: 86%|████████▌ | 1137/1319 [19:35<02:41, 1.13it/s]
Running generate_until requests: 87%|████████▋ | 1153/1319 [19:52<02:36, 1.06it/s]
Running generate_until requests: 89%|████████▊ | 1169/1319 [20:05<02:14, 1.12it/s]
Running generate_until requests: 90%|████████▉ | 1185/1319 [20:22<02:06, 1.06it/s]
Running generate_until requests: 91%|█████████ | 1201/1319 [20:34<01:46, 1.11it/s]
Running generate_until requests: 92%|█████████▏| 1217/1319 [20:47<01:28, 1.15it/s]
Running generate_until requests: 93%|█████████▎| 1233/1319 [20:59<01:11, 1.21it/s]
Running generate_until requests: 95%|█████████▍| 1249/1319 [21:11<00:56, 1.24it/s]
Running generate_until requests: 96%|█████████▌| 1265/1319 [21:25<00:44, 1.22it/s]
Running generate_until requests: 97%|█████████▋| 1281/1319 [21:37<00:30, 1.23it/s]
Running generate_until requests: 98%|█████████▊| 1297/1319 [21:49<00:17, 1.26it/s]
Running generate_until requests: 100%|█████████▉| 1313/1319 [21:59<00:04, 1.37it/s]
Running generate_until requests: 100%|██████████| 1319/1319 [21:59<00:00, 1.00s/it] |
|
2025-02-09:13:12:49,053 INFO [evaluation_tracker.py:269] Output path not provided, skipping saving results aggregated |
|
hf (pretrained=meta-llama/Llama-3.2-3B-Instruct,peft=JaqueMate/new_config_3b_only_dpo_checkpoint-1600), gen_kwargs: (do_sample=False,temperature=0.01,top_k=0,top_p=1), limit: None, num_fewshot: 8, batch_size: 16 |
|
| Tasks |Version| Filter |n-shot| Metric | |Value | |Stderr| |
|
| |
|
|gsm8k_cot_llama| 3|flexible-extract| 8|exact_match|↑ |0.7407|± |0.0121| |
|
| | |strict-match | 8|exact_match|↑ |0.7225|± |0.0123| |
|
|
|
|
