progress / artifacts /logs /qat.log
eyad-silx's picture
push artifacts/logs/qat.log
86db490 verified
Raw History Blame Contribute Delete
6.73 kB
[transformers] `torch_dtype` is deprecated! Use `dtype` instead!
samples: 16
Loading weights: 0%| | 0/851 [00:00<?, ?it/s] Loading weights: 0%| | 1/851 [00:00<11:14, 1.26it/s] Loading weights: 2%|▏ | 14/851 [00:00<00:40, 20.54it/s] Loading weights: 3%|β–Ž | 27/851 [00:01<00:21, 39.11it/s] Loading weights: 4%|▍ | 38/851 [00:01<00:15, 52.16it/s] Loading weights: 6%|β–Œ | 48/851 [00:01<00:17, 47.03it/s] Loading weights: 8%|β–Š | 65/851 [00:01<00:11, 69.41it/s] Loading weights: 9%|β–‰ | 76/851 [00:01<00:10, 70.69it/s] Loading weights: 10%|β–ˆ | 86/851 [00:01<00:11, 67.64it/s] Loading weights: 11%|β–ˆ | 95/851 [00:01<00:10, 69.47it/s] Loading weights: 12%|β–ˆβ– | 104/851 [00:02<00:11, 67.04it/s] Loading weights: 14%|β–ˆβ– | 119/851 [00:02<00:08, 81.77it/s] Loading weights: 15%|β–ˆβ–Œ | 129/851 [00:02<00:08, 84.80it/s] Loading weights: 16%|β–ˆβ–‹ | 139/851 [00:02<00:08, 86.34it/s] Loading weights: 18%|β–ˆβ–Š | 149/851 [00:02<00:08, 79.42it/s] Loading weights: 19%|β–ˆβ–Š | 158/851 [00:02<00:08, 79.87it/s] Loading weights: 20%|β–ˆβ–ˆ | 172/851 [00:02<00:07, 95.14it/s] Loading weights: 22%|β–ˆβ–ˆβ– | 183/851 [00:02<00:06, 99.02it/s] Loading weights: 23%|β–ˆβ–ˆβ–Ž | 194/851 [00:02<00:06, 94.31it/s] Loading weights: 24%|β–ˆβ–ˆβ– | 204/851 [00:03<00:07, 89.65it/s] Loading weights: 25%|β–ˆβ–ˆβ–Œ | 214/851 [00:03<00:07, 84.90it/s] Loading weights: 27%|β–ˆβ–ˆβ–‹ | 226/851 [00:03<00:07, 88.43it/s] Loading weights: 28%|β–ˆβ–ˆβ–Š | 239/851 [00:03<00:06, 94.41it/s] Loading weights: 30%|β–ˆβ–ˆβ–‰ | 252/851 [00:03<00:05, 101.22it/s] Loading weights: 31%|β–ˆβ–ˆβ–ˆ | 263/851 [00:03<00:07, 79.06it/s] Loading weights: 33%|β–ˆβ–ˆβ–ˆβ–Ž | 278/851 [00:03<00:06, 92.02it/s] Loading weights: 34%|β–ˆβ–ˆβ–ˆβ– | 290/851 [00:04<00:06, 88.76it/s] Loading weights: 36%|β–ˆβ–ˆβ–ˆβ–Œ | 303/851 [00:04<00:05, 95.23it/s] Loading weights: 37%|β–ˆβ–ˆβ–ˆβ–‹ | 314/851 [00:04<00:06, 80.20it/s] Loading weights: 39%|β–ˆβ–ˆβ–ˆβ–‰ | 331/851 [00:04<00:05, 97.77it/s] Loading weights: 40%|β–ˆβ–ˆβ–ˆβ–ˆ | 342/851 [00:04<00:05, 98.90it/s] Loading weights: 42%|β–ˆβ–ˆβ–ˆβ–ˆβ– | 355/851 [00:04<00:04, 100.67it/s] Loading weights: 43%|β–ˆβ–ˆβ–ˆβ–ˆβ–Ž | 366/851 [00:04<00:05, 82.39it/s] Loading weights: 45%|β–ˆβ–ˆβ–ˆβ–ˆβ–Œ | 383/851 [00:05<00:05, 92.94it/s] Loading weights: 47%|β–ˆβ–ˆβ–ˆβ–ˆβ–‹ | 398/851 [00:05<00:04, 103.68it/s] Loading weights: 48%|β–ˆβ–ˆβ–ˆβ–ˆβ–Š | 411/851 [00:05<00:04, 109.21it/s] Loading weights: 50%|β–ˆβ–ˆβ–ˆβ–ˆβ–‰ | 423/851 [00:05<00:04, 88.40it/s] Loading weights: 51%|β–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ– | 438/851 [00:05<00:04, 90.05it/s] Loading weights: 53%|β–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–Ž | 452/851 [00:05<00:03, 100.85it/s] Loading weights: 55%|β–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ– | 465/851 [00:05<00:03, 106.65it/s] Loading weights: 56%|β–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–Œ | 477/851 [00:06<00:04, 89.09it/s] Loading weights: 58%|β–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–Š | 491/851 [00:06<00:03, 95.04it/s] Loading weights: 59%|β–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–‰ | 504/851 [00:06<00:03, 100.35it/s] Loading weights: 61%|β–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆ | 515/851 [00:06<00:03, 100.63it/s] Loading weights: 62%|β–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ– | 526/851 [00:06<00:03, 84.89it/s] Loading weights: 64%|β–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–Ž | 542/851 [00:06<00:03, 92.76it/s] Loading weights: 65%|β–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–Œ | 555/851 [00:06<00:02, 100.77it/s] Loading weights: 67%|β–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–‹ | 571/851 [00:06<00:02, 107.71it/s] Loading weights: 69%|β–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–Š | 583/851 [00:07<00:02, 98.34it/s] Loading weights: 70%|β–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆ | 597/851 [00:07<00:02, 103.24it/s] Loading weights: 71%|β–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ– | 608/851 [00:07<00:02, 104.36it/s] Loading weights: 73%|β–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–Ž | 623/851 [00:07<00:02, 106.41it/s] Loading weights: 75%|β–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ– | 634/851 [00:07<00:02, 95.28it/s] Loading weights: 76%|β–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–‹ | 649/851 [00:07<00:01, 108.08it/s] Loading weights: 78%|β–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–Š | 661/851 [00:07<00:01, 110.98it/s] Loading weights: 79%|β–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–‰ | 673/851 [00:07<00:01, 112.51it/s] Loading weights: 80%|β–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆ | 685/851 [00:08<00:01, 91.60it/s] Loading weights: 82%|β–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ– | 702/851 [00:08<00:01, 108.21it/s] Loading weights: 84%|β–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ– | 714/851 [00:08<00:01, 110.44it/s] Loading weights: 85%|β–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–Œ | 726/851 [00:08<00:01, 104.46it/s] Loading weights: 87%|β–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–‹ | 737/851 [00:08<00:01, 87.23it/s] Loading weights: 89%|β–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–Š | 754/851 [00:08<00:00, 103.75it/s] Loading weights: 90%|β–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆ | 767/851 [00:08<00:00, 95.53it/s] Loading weights: 92%|β–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–| 783/851 [00:09<00:00, 110.05it/s] Loading weights: 93%|β–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–Ž| 795/851 [00:09<00:00, 98.54it/s] Loading weights: 95%|β–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–Œ| 809/851 [00:09<00:00, 104.18it/s] Loading weights: 96%|β–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–‹| 821/851 [00:09<00:00, 106.94it/s] Loading weights: 98%|β–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–Š| 835/851 [00:09<00:00, 115.44it/s] Loading weights: 100%|β–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–‰| 848/851 [00:09<00:00, 95.64it/s] Loading weights: 100%|β–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆ| 851/851 [00:09<00:00, 87.66it/s]
[transformers] `causal_conv1d_fn` is falling back to its reference PyTorch implementation because `causal_conv1d` is not installed. This is correct but much slower; install `causal_conv1d` for the optimized kernel.
trainable norms: 209
step 0: loss 0.15632 over 15 samples
step 1: loss 0.10217 over 15 samples
step 2: loss 0.08343 over 15 samples
step 3: loss 0.07402 over 15 samples
step 4: loss 0.06890 over 15 samples
step 5: loss 0.06615 over 15 samples
step 6: loss 0.06435 over 15 samples
step 7: loss 0.06303 over 15 samples
step 8: loss 0.06203 over 15 samples
step 9: loss 0.06120 over 15 samples
step 10: loss 0.06050 over 15 samples
step 11: loss 0.05988 over 15 samples
step 12: loss 0.05942 over 15 samples
step 13: loss 0.05874 over 15 samples
step 14: loss 0.05826 over 15 samples
step 15: loss 0.05782 over 15 samples
step 16: loss 0.05748 over 15 samples
step 17: loss 0.05716 over 15 samples
step 18: loss 0.05687 over 15 samples
step 19: loss 0.05663 over 15 samples
step 20: loss 0.05639 over 15 samples
step 21: loss 0.05615 over 15 samples
step 22: loss 0.05597 over 15 samples
step 23: loss 0.05579 over 15 samples
step 24: loss 0.05589 over 15 samples
saved /root/models/chal-n2/chal-norms.gguf