|
|
---
|
|
|
license: mit
|
|
|
pipeline_tag: text-generation
|
|
|
library_name: transformers
|
|
|
language: [
|
|
|
'en', 'am', 'ar', 'as', 'az', 'be', 'bg', 'bn', 'br', 'bs', 'ca', 'cs', 'cy', 'da', 'de', 'el',
|
|
|
'eo', 'es', 'et', 'eu', 'fa', 'ff', 'fi', 'fr', 'fy', 'ga', 'gd', 'gl', 'gn', 'gu', 'ha', 'he',
|
|
|
'hi', 'hr', 'ht', 'hu', 'hy', 'id', 'ig', 'is', 'it', 'ja', 'jv', 'ka', 'kk', 'km', 'kn', 'ko',
|
|
|
'ku', 'ky', 'la', 'lg', 'li', 'ln', 'lo', 'lt', 'lv', 'mg', 'mk', 'ml', 'mn', 'mr', 'ms', 'my',
|
|
|
'ne', 'nl', 'no', 'ns', 'om', 'or', 'pa', 'pl', 'ps', 'pt', 'qu', 'rm', 'ro', 'ru', 'sa', 'si',
|
|
|
'sc', 'sd', 'sk', 'sl', 'so', 'sq', 'sr', 'ss', 'su', 'sv', 'sw', 'ta', 'te', 'th', 'tl', 'tn',
|
|
|
'tr', 'ug', 'uk', 'ur', 'uz', 'vi', 'wo', 'xh', 'yi', 'yo', 'zu',
|
|
|
]
|
|
|
datasets:
|
|
|
|
|
|
- ontocord/fineweb-permissive-multilingual-2m
|
|
|
- distily/c4_multilingual_1M
|
|
|
- data-silence/sumnews
|
|
|
- xu-song/cc100-samples
|
|
|
- badrex/llm-emoji-dataset
|
|
|
- fblgit/simple-math
|
|
|
- Gusarich/math-expressions-1m
|
|
|
- neuralwork/arxiver
|
|
|
- christopher/rosetta-code
|
|
|
- nampdn-ai/tiny-codes
|
|
|
- JeanKaddour/minipile
|
|
|
|
|
|
- NousResearch/hermes-function-calling-v1
|
|
|
- simplescaling/s1K-1.1
|
|
|
|
|
|
- mlabonne/open-perfectblend
|
|
|
- allenai/tulu-3-sft-mixture
|
|
|
- rombodawg/Everything_Instruct_Multilingual
|
|
|
|
|
|
- open-r1/OpenR1-Math-220k
|
|
|
- open-thoughts/OpenThoughts-114k
|
|
|
- cognitivecomputations/dolphin-r1
|
|
|
- simplescaling/s1K-1.1
|
|
|
tags:
|
|
|
- chat
|
|
|
- core
|
|
|
- base
|
|
|
- instruct
|
|
|
- reason
|
|
|
---
|
|
|
|
|
|
# tangled-alpha-0.9-core
|
|
|
|
|
|
|
|
|
|
|
|
```bash
|
|
|
time python -B prepare_core_datasets.py
|
|
|
```
|
|
|
|
|
|
```
|
|
|
i=0, min_len=0, max_len=1073741824, block_size=1025, chunk_size=16400000, len(dataset)=5146620, len(dataset) * block_size=5275285500
|
|
|
Total number of tokens in the optimized dataset '../core-data-0-0-1073741824-1025-16000' is 5275285500
|
|
|
|
|
|
i=1, min_len=1025, max_len=2049, block_size=2049, chunk_size=16392000, len(dataset)=309838, len(dataset) * block_size=634858062
|
|
|
Total number of tokens in the optimized dataset '../core-data-1-1025-2049-2049-8000' is 634858062
|
|
|
|
|
|
i=2, min_len=2049, max_len=4097, block_size=4097, chunk_size=16388000, len(dataset)=113843, len(dataset) * block_size=466414771
|
|
|
Total number of tokens in the optimized dataset '../core-data-2-2049-4097-4097-4000' is 466414771
|
|
|
|
|
|
i=3, min_len=4097, max_len=8193, block_size=8193, chunk_size=16386000, len(dataset)=56713, len(dataset) * block_size=464649609
|
|
|
Total number of tokens in the optimized dataset '../core-data-3-4097-8193-8193-2000' is 464649609
|
|
|
|
|
|
i=4, min_len=8193, max_len=16385, block_size=16385, chunk_size=16385000, len(dataset)=37406, len(dataset) * block_size=612897310
|
|
|
Total number of tokens in the optimized dataset '../core-data-4-8193-16385-16385-1000' is 612897310
|
|
|
|
|
|
i=5, min_len=16385, max_len=32769, block_size=32769, chunk_size=16384500, len(dataset)=12737, len(dataset) * block_size=417378753
|
|
|
Total number of tokens in the optimized dataset '../core-data-5-16385-32769-32769-500' is 417378753
|
|
|
|
|
|
i=6, min_len=32769, max_len=65537, block_size=65537, chunk_size=16384250, len(dataset)=2824, len(dataset) * block_size=185076488
|
|
|
Total number of tokens in the optimized dataset '../core-data-6-32769-65537-65537-250' is 185076488
|
|
|
|
|
|
i=7, min_len=65537, max_len=131073, block_size=131073, chunk_size=16384125, len(dataset)=634, len(dataset) * block_size=83100282
|
|
|
Total number of tokens in the optimized dataset '../core-data-7-65537-131073-131073-125' is 83100282
|
|
|
|
|
|
real 292m54.341s
|
|
|
user 2118m1.154s
|
|
|
sys 12m2.746s
|
|
|
|
|
|
20G tangled-alpha-0.9-core/core-data-0-0-1073741824-1025-16000
|
|
|
2.4G tangled-alpha-0.9-core/core-data-1-1025-2049-2049-8000
|
|
|
1.8G tangled-alpha-0.9-core/core-data-2-2049-4097-4097-4000
|
|
|
1.8G tangled-alpha-0.9-core/core-data-3-4097-8193-8193-2000
|
|
|
2.3G tangled-alpha-0.9-core/core-data-4-8193-16385-16385-1000
|
|
|
1.6G tangled-alpha-0.9-core/core-data-5-16385-32769-32769-500
|
|
|
709M tangled-alpha-0.9-core/core-data-6-32769-65537-65537-250
|
|
|
321M tangled-alpha-0.9-core/core-data-7-65537-131073-131073-125
|
|
|
```
|
|
|
|
|
|
```bash
|
|
|
CUDA_VISIBLE_DEVICES=0 CUDA_LAUNCH_BLOCKING=0 PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True litgpt pretrain --config pretrain_core_model_0.yaml
|
|
|
```
|
|
|
|
|
|
```
|
|
|
Seed set to 23
|
|
|
Time to instantiate model: 0.44 seconds.
|
|
|
Total parameters: 234,914,304
|
|
|
Verifying settings ...
|
|
|
Measured TFLOPs: 55520.94
|
|
|
Epoch 1 | iter 64 step 1 | loss train: 11.977, val: n/a | iter time: 490.27 ms (step) remaining time: 6 days, 22:47:04
|
|
|
Epoch 1 | iter 128 step 2 | loss train: 11.970, val: n/a | iter time: 351.11 ms (step) remaining time: 4 days, 16:53:01
|
|
|
Epoch 1 | iter 192 step 3 | loss train: 11.971, val: n/a | iter time: 353.74 ms (step) remaining time: 3 days, 23:43:23
|
|
|
Epoch 1 | iter 256 step 4 | loss train: 11.974, val: n/a | iter time: 355.03 ms (step) remaining time: 3 days, 14:41:57
|
|
|
Epoch 1 | iter 320 step 5 | loss train: 11.964, val: n/a | iter time: 357.36 ms (step) remaining time: 3 days, 9:21:54
|
|
|
Epoch 1 | iter 384 step 6 | loss train: 11.957, val: n/a | iter time: 362.27 ms (step) remaining time: 3 days, 5:53:20
|
|
|
Epoch 1 | iter 448 step 7 | loss train: 11.948, val: n/a | iter time: 359.89 ms (step) remaining time: 3 days, 3:26:34
|
|
|
Epoch 1 | iter 512 step 8 | loss train: 11.938, val: n/a | iter time: 363.84 ms (step) remaining time: 3 days, 1:37:54
|
|
|
Epoch 1 | iter 576 step 9 | loss train: 11.920, val: n/a | iter time: 362.75 ms (step) remaining time: 3 days, 0:13:59
|
|
|
Epoch 1 | iter 640 step 10 | loss train: 11.900, val: n/a | iter time: 363.46 ms (step) remaining time: 2 days, 23:07:06
|
|
|
# ...
|
|
|
Epoch 1 | iter 643264 step 10051 | loss train: 2.834, val: 2.669 | iter time: 360.50 ms (step) remaining time: 0:03:59
|
|
|
Epoch 2 | iter 643328 step 10052 | loss train: 2.837, val: 2.669 | iter time: 359.53 ms (step) remaining time: 0:03:37
|
|
|
Epoch 2 | iter 643392 step 10053 | loss train: 2.768, val: 2.669 | iter time: 362.83 ms (step) remaining time: 0:03:15
|
|
|
Epoch 2 | iter 643456 step 10054 | loss train: 2.695, val: 2.669 | iter time: 363.85 ms (step) remaining time: 0:02:53
|
|
|
Epoch 2 | iter 643520 step 10055 | loss train: 2.768, val: 2.669 | iter time: 365.40 ms (step) remaining time: 0:02:30
|
|
|
Epoch 2 | iter 643584 step 10056 | loss train: 2.710, val: 2.669 | iter time: 364.72 ms (step) remaining time: 0:02:08
|
|
|
Epoch 2 | iter 643648 step 10057 | loss train: 2.749, val: 2.669 | iter time: 365.00 ms (step) remaining time: 0:01:46
|
|
|
Epoch 2 | iter 643712 step 10058 | loss train: 2.748, val: 2.669 | iter time: 363.42 ms (step) remaining time: 0:01:24
|
|
|
Epoch 2 | iter 643776 step 10059 | loss train: 2.710, val: 2.669 | iter time: 364.49 ms (step) remaining time: 0:01:02
|
|
|
Epoch 2 | iter 643840 step 10060 | loss train: 2.738, val: 2.669 | iter time: 364.43 ms (step) remaining time: 0:00:39
|
|
|
Epoch 2 | iter 643904 step 10061 | loss train: 2.734, val: 2.669 | iter time: 364.94 ms (step) remaining time: 0:00:17
|
|
|
Validating ...
|
|
|
Final evaluation | val loss: 2.669 | val ppl: 14.422
|
|
|
Saving checkpoint to '../out/pretrain-core-0/final/lit_model.pth'
|
|
|
----------------------------------------
|
|
|
| Performance
|
|
|
| - Total tokens : 5,275,279,360
|
|
|
| - Training Time : 223314.37 s
|
|
|
| - Tok/sec : 5541.09 tok/s
|
|
|
| ----------------------------------------
|
|
|
| Memory Usage
|
|
|
| - Memory Used : 22.33 GB
|
|
|
----------------------------------------
|
|
|
```
|
|
|
|
|
|
Backup `wandb`:
|
|
|
|
|
|
```bash
|
|
|
mv wandb wandb-pretrain-core-0
|
|
|
```
|
|
|
|
|
|
Copy config:
|
|
|
|
|
|
```bash
|
|
|
cp ../config-0.json ../out/pretrain-core-0/final/config.json
|
|
|
```
|
|
|
|
|
|
Chat with model:
|
|
|
|
|
|
```bash
|
|
|
CUDA_VISIBLE_DEVICES=0 CUDA_LAUNCH_BLOCKING=0 PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True litgpt chat ../out/pretrain-core-0/final
|
|
|
```
|
|
|
|
|
|
```bash
|
|
|
CUDA_VISIBLE_DEVICES=0 CUDA_LAUNCH_BLOCKING=0 PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True time litgpt evaluate --tasks 'leaderboard' --out_dir '../evaluate/pretrain-core-0/leaderboard/' --batch_size '4' --dtype 'bfloat16' '../out/pretrain-core-0/final'
|
|
|
```
|
|
|
|
|
|
```
|
|
|
| Tasks |Version|Filter|n-shot| Metric | |Value | |Stderr|
|
|
|
|-----------------------------------------------------------|-------|------|-----:|-----------------------|---|-----:|---|------|
|
|
|
|leaderboard | N/A| | | | | | | |
|
|
|
| - leaderboard_bbh | N/A| | | | | | | |
|
|
|
| - leaderboard_bbh_boolean_expressions | 1|none | 3|acc_norm |↑ |0.4600|± |0.0316|
|
|
|
| - leaderboard_bbh_causal_judgement | 1|none | 3|acc_norm |↑ |0.5134|± |0.0366|
|
|
|
| - leaderboard_bbh_date_understanding | 1|none | 3|acc_norm |↑ |0.1960|± |0.0252|
|
|
|
| - leaderboard_bbh_disambiguation_qa | 1|none | 3|acc_norm |↑ |0.3320|± |0.0298|
|
|
|
| - leaderboard_bbh_formal_fallacies | 1|none | 3|acc_norm |↑ |0.4680|± |0.0316|
|
|
|
| - leaderboard_bbh_geometric_shapes | 1|none | 3|acc_norm |↑ |0.2400|± |0.0271|
|
|
|
| - leaderboard_bbh_hyperbaton | 1|none | 3|acc_norm |↑ |0.5160|± |0.0317|
|
|
|
| - leaderboard_bbh_logical_deduction_five_objects | 1|none | 3|acc_norm |↑ |0.2040|± |0.0255|
|
|
|
| - leaderboard_bbh_logical_deduction_seven_objects | 1|none | 3|acc_norm |↑ |0.1320|± |0.0215|
|
|
|
| - leaderboard_bbh_logical_deduction_three_objects | 1|none | 3|acc_norm |↑ |0.3440|± |0.0301|
|
|
|
| - leaderboard_bbh_movie_recommendation | 1|none | 3|acc_norm |↑ |0.2680|± |0.0281|
|
|
|
| - leaderboard_bbh_navigate | 1|none | 3|acc_norm |↑ |0.5720|± |0.0314|
|
|
|
| - leaderboard_bbh_object_counting | 1|none | 3|acc_norm |↑ |0.0680|± |0.0160|
|
|
|
| - leaderboard_bbh_penguins_in_a_table | 1|none | 3|acc_norm |↑ |0.2055|± |0.0336|
|
|
|
| - leaderboard_bbh_reasoning_about_colored_objects | 1|none | 3|acc_norm |↑ |0.1760|± |0.0241|
|
|
|
| - leaderboard_bbh_ruin_names | 1|none | 3|acc_norm |↑ |0.2120|± |0.0259|
|
|
|
| - leaderboard_bbh_salient_translation_error_detection | 1|none | 3|acc_norm |↑ |0.2240|± |0.0264|
|
|
|
| - leaderboard_bbh_snarks | 1|none | 3|acc_norm |↑ |0.5393|± |0.0375|
|
|
|
| - leaderboard_bbh_sports_understanding | 1|none | 3|acc_norm |↑ |0.4600|± |0.0316|
|
|
|
| - leaderboard_bbh_temporal_sequences | 1|none | 3|acc_norm |↑ |0.2760|± |0.0283|
|
|
|
| - leaderboard_bbh_tracking_shuffled_objects_five_objects | 1|none | 3|acc_norm |↑ |0.1720|± |0.0239|
|
|
|
| - leaderboard_bbh_tracking_shuffled_objects_seven_objects| 1|none | 3|acc_norm |↑ |0.1360|± |0.0217|
|
|
|
| - leaderboard_bbh_tracking_shuffled_objects_three_objects| 1|none | 3|acc_norm |↑ |0.3320|± |0.0298|
|
|
|
| - leaderboard_bbh_web_of_lies | 1|none | 3|acc_norm |↑ |0.4880|± |0.0317|
|
|
|
| - leaderboard_gpqa | N/A| | | | | | | |
|
|
|
| - leaderboard_gpqa_diamond | 1|none | 0|acc_norm |↑ |0.2071|± |0.0289|
|
|
|
| - leaderboard_gpqa_extended | 1|none | 0|acc_norm |↑ |0.2637|± |0.0189|
|
|
|
| - leaderboard_gpqa_main | 1|none | 0|acc_norm |↑ |0.2612|± |0.0208|
|
|
|
| - leaderboard_ifeval | 3|none | 0|inst_level_loose_acc |↑ |0.2770|± | N/A|
|
|
|
| | |none | 0|inst_level_strict_acc |↑ |0.2710|± | N/A|
|
|
|
| | |none | 0|prompt_level_loose_acc |↑ |0.1534|± |0.0155|
|
|
|
| | |none | 0|prompt_level_strict_acc|↑ |0.1497|± |0.0154|
|
|
|
| - leaderboard_math_hard | N/A| | | | | | | |
|
|
|
| - leaderboard_math_algebra_hard | 2|none | 4|exact_match |↑ |0.0017|± |0.0012|
|
|
|
| - leaderboard_math_counting_and_prob_hard | 2|none | 4|exact_match |↑ |0.0000|± | 0|
|
|
|
| - leaderboard_math_geometry_hard | 2|none | 4|exact_match |↑ |0.0000|± | 0|
|
|
|
| - leaderboard_math_intermediate_algebra_hard | 2|none | 4|exact_match |↑ |0.0033|± |0.0019|
|
|
|
| - leaderboard_math_num_theory_hard | 2|none | 4|exact_match |↑ |0.0037|± |0.0026|
|
|
|
| - leaderboard_math_prealgebra_hard | 2|none | 4|exact_match |↑ |0.0046|± |0.0023|
|
|
|
| - leaderboard_math_precalculus_hard | 2|none | 4|exact_match |↑ |0.0000|± | 0|
|
|
|
| - leaderboard_mmlu_pro | 0.1|none | 5|acc |↑ |0.1068|± |0.0028|
|
|
|
| - leaderboard_musr | N/A| | | | | | | |
|
|
|
| - leaderboard_musr_murder_mysteries | 1|none | 0|acc_norm |↑ |0.5160|± |0.0317|
|
|
|
| - leaderboard_musr_object_placements | 1|none | 0|acc_norm |↑ |0.2344|± |0.0265|
|
|
|
| - leaderboard_musr_team_allocation | 1|none | 0|acc_norm |↑ |0.3200|± |0.0296|
|
|
|
```
|
|
|
|
|
|
```bash
|
|
|
litgpt convert_pretrained_checkpoint ../out/pretrain-core-0/final ../out/pretrain-core-0/checkpoint
|
|
|
```
|
|
|
|
|
|
```bash
|
|
|
CUDA_VISIBLE_DEVICES=0 CUDA_LAUNCH_BLOCKING=0 PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True litgpt pretrain --config pretrain_core_model_1.yaml
|
|
|
```
|
|
|
|
|
|
```
|
|
|
Seed set to 23
|
|
|
Time to instantiate model: 0.32 seconds.
|
|
|
Total parameters: 234,914,304
|
|
|
Validating ...
|
|
|
Measured TFLOPs: 27760.47
|
|
|
Epoch 1 | iter 128 step 1 | loss train: 3.205, val: 3.457 | iter time: 450.87 ms (step) remaining time: 10:13:33
|
|
|
Epoch 1 | iter 256 step 2 | loss train: 3.175, val: 3.457 | iter time: 386.03 ms (step) remaining time: 9:00:10
|
|
|
Epoch 1 | iter 384 step 3 | loss train: 3.144, val: 3.457 | iter time: 388.88 ms (step) remaining time: 8:35:53
|
|
|
Epoch 1 | iter 512 step 4 | loss train: 3.260, val: 3.457 | iter time: 390.66 ms (step) remaining time: 8:24:17
|
|
|
Epoch 1 | iter 640 step 5 | loss train: 3.247, val: 3.457 | iter time: 392.99 ms (step) remaining time: 8:17:54
|
|
|
Epoch 1 | iter 768 step 6 | loss train: 3.264, val: 3.457 | iter time: 397.01 ms (step) remaining time: 8:13:51
|
|
|
Epoch 1 | iter 896 step 7 | loss train: 3.232, val: 3.457 | iter time: 396.00 ms (step) remaining time: 8:11:00
|
|
|
Epoch 1 | iter 1024 step 8 | loss train: 3.287, val: 3.457 | iter time: 396.93 ms (step) remaining time: 8:08:43
|
|
|
Epoch 1 | iter 1152 step 9 | loss train: 3.236, val: 3.457 | iter time: 398.67 ms (step) remaining time: 8:06:49
|
|
|
Epoch 1 | iter 1280 step 10 | loss train: 3.274, val: 3.457 | iter time: 399.49 ms (step) remaining time: 8:05:09
|
|
|
# ...
|
|
|
Epoch 1 | iter 76928 step 601 | loss train: 3.177, val: 3.304 | iter time: 400.61 ms (step) remaining time: 0:03:35
|
|
|
Epoch 1 | iter 77056 step 602 | loss train: 3.191, val: 3.304 | iter time: 396.14 ms (step) remaining time: 0:02:46
|
|
|
Epoch 1 | iter 77184 step 603 | loss train: 3.173, val: 3.304 | iter time: 399.39 ms (step) remaining time: 0:01:58
|
|
|
Epoch 1 | iter 77312 step 604 | loss train: 3.211, val: 3.304 | iter time: 398.61 ms (step) remaining time: 0:01:09
|
|
|
Epoch 1 | iter 77440 step 605 | loss train: 3.203, val: 3.304 | iter time: 399.31 ms (step) remaining time: 0:00:21
|
|
|
Validating ...
|
|
|
Final evaluation | val loss: 3.304 | val ppl: 27.217
|
|
|
Saving checkpoint to '../out/pretrain-core-1/final/lit_model.pth'
|
|
|
----------------------------------------
|
|
|
| Performance
|
|
|
| - Total tokens : 634,855,424
|
|
|
| - Training Time : 29361.39 s
|
|
|
| - Tok/sec : 524.18 tok/s
|
|
|
| ----------------------------------------
|
|
|
| Memory Usage
|
|
|
| - Memory Used : 22.33 GB
|
|
|
----------------------------------------
|
|
|
```
|
|
|
|
|
|
```bash
|
|
|
mv wandb wandb-pretrain-core-1
|
|
|
```
|
|
|
|
|
|
```bash
|
|
|
cp ../config-1.json ../out/pretrain-core-1/final/config.json
|
|
|
```
|
|
|
|
|
|
```bash
|
|
|
CUDA_VISIBLE_DEVICES=0 CUDA_LAUNCH_BLOCKING=0 PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True litgpt chat ../out/pretrain-core-1/final
|
|
|
```
|
|
|
|
|
|
```bash
|
|
|
litgpt convert_pretrained_checkpoint ../out/pretrain-core-1/final ../out/pretrain-core-1/checkpoint
|
|
|
```
|
|
|
|
|
|
```bash
|
|
|
CUDA_VISIBLE_DEVICES=0 CUDA_LAUNCH_BLOCKING=0 PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True litgpt pretrain --config pretrain_core_model_2.yaml
|
|
|
```
|
|
|
|
|
|
```
|
|
|
Seed set to 23
|
|
|
Time to instantiate model: 0.30 seconds.
|
|
|
Total parameters: 234,914,304
|
|
|
Validating ...
|
|
|
Measured TFLOPs: 13880.23
|
|
|
Epoch 1 | iter 256 step 1 | loss train: 2.857, val: 3.078 | iter time: 518.53 ms (step) remaining time: 8:49:46
|
|
|
Epoch 1 | iter 512 step 2 | loss train: 2.839, val: 3.078 | iter time: 461.04 ms (step) remaining time: 7:47:05
|
|
|
Epoch 1 | iter 768 step 3 | loss train: 2.835, val: 3.078 | iter time: 462.82 ms (step) remaining time: 7:27:16
|
|
|
Epoch 1 | iter 1024 step 4 | loss train: 2.872, val: 3.078 | iter time: 464.55 ms (step) remaining time: 7:16:22
|
|
|
Epoch 1 | iter 1280 step 5 | loss train: 2.867, val: 3.078 | iter time: 462.06 ms (step) remaining time: 7:09:02
|
|
|
Epoch 1 | iter 1536 step 6 | loss train: 2.899, val: 3.078 | iter time: 465.26 ms (step) remaining time: 7:03:30
|
|
|
Epoch 1 | iter 1792 step 7 | loss train: 2.878, val: 3.078 | iter time: 465.57 ms (step) remaining time: 6:59:00
|
|
|
Epoch 1 | iter 2048 step 8 | loss train: 2.919, val: 3.078 | iter time: 464.37 ms (step) remaining time: 6:55:10
|
|
|
Epoch 1 | iter 2304 step 9 | loss train: 2.922, val: 3.078 | iter time: 464.24 ms (step) remaining time: 6:51:45
|
|
|
Epoch 1 | iter 2560 step 10 | loss train: 2.924, val: 3.078 | iter time: 464.71 ms (step) remaining time: 6:48:39
|
|
|
# ...
|
|
|
Epoch 1 | iter 53760 step 210 | loss train: 2.904, val: 3.013 | iter time: 468.66 ms (step) remaining time: 0:23:26
|
|
|
Epoch 1 | iter 54016 step 211 | loss train: 2.903, val: 3.013 | iter time: 468.81 ms (step) remaining time: 0:21:32
|
|
|
Epoch 1 | iter 54272 step 212 | loss train: 2.951, val: 3.013 | iter time: 463.52 ms (step) remaining time: 0:19:39
|
|
|
Epoch 1 | iter 54528 step 213 | loss train: 2.941, val: 3.013 | iter time: 466.12 ms (step) remaining time: 0:17:45
|
|
|
Epoch 1 | iter 54784 step 214 | loss train: 2.950, val: 3.013 | iter time: 468.24 ms (step) remaining time: 0:15:52
|
|
|
Epoch 1 | iter 55040 step 215 | loss train: 2.943, val: 3.013 | iter time: 466.65 ms (step) remaining time: 0:13:59
|
|
|
Epoch 1 | iter 55296 step 216 | loss train: 2.903, val: 3.013 | iter time: 464.44 ms (step) remaining time: 0:12:05
|
|
|
Epoch 1 | iter 55552 step 217 | loss train: 2.954, val: 3.013 | iter time: 465.12 ms (step) remaining time: 0:10:12
|
|
|
Epoch 1 | iter 55808 step 218 | loss train: 2.907, val: 3.013 | iter time: 464.96 ms (step) remaining time: 0:08:19
|
|
|
Epoch 1 | iter 56064 step 219 | loss train: 2.909, val: 3.013 | iter time: 467.12 ms (step) remaining time: 0:06:25
|
|
|
Epoch 1 | iter 56320 step 220 | loss train: 2.908, val: 3.013 | iter time: 466.43 ms (step) remaining time: 0:04:32
|
|
|
Epoch 1 | iter 56576 step 221 | loss train: 2.894, val: 3.013 | iter time: 469.70 ms (step) remaining time: 0:02:38
|
|
|
Epoch 1 | iter 56832 step 222 | loss train: 2.809, val: 3.013 | iter time: 463.88 ms (step) remaining time: 0:00:45
|
|
|
Validating ...
|
|
|
Final evaluation | val loss: 3.011 | val ppl: 20.306
|
|
|
Saving checkpoint to '../out/pretrain-core-1/final/lit_model.pth'
|
|
|
----------------------------------------
|
|
|
| Performance
|
|
|
| - Total tokens : 466,411,520
|
|
|
| - Training Time : 25263.31 s
|
|
|
| - Tok/sec : 371.33 tok/s
|
|
|
| ----------------------------------------
|
|
|
| Memory Usage
|
|
|
| - Memory Used : 22.33 GB
|
|
|
----------------------------------------
|
|
|
```
|
|
|
|
|
|
```bash
|
|
|
cp ../config-2.json ../out/pretrain-core-2/final/config.json
|
|
|
```
|
|
|
|
|
|
```bash
|
|
|
mv wandb wandb-pretrain-core-2
|
|
|
```
|
|
|
|
|
|
```bash
|
|
|
CUDA_VISIBLE_DEVICES=0 CUDA_LAUNCH_BLOCKING=0 PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True litgpt chat ../out/pretrain-core-2/final
|
|
|
```
|
|
|
|
|
|
```bash
|
|
|
litgpt convert_pretrained_checkpoint ../out/pretrain-core-2/final ../out/pretrain-core-2/checkpoint
|
|
|
```
|
|
|
|
|
|
```bash
|
|
|
CUDA_VISIBLE_DEVICES=0 CUDA_LAUNCH_BLOCKING=0 PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True litgpt pretrain --config pretrain_core_model_3.yaml
|
|
|
```
|
|
|
|
|
|
```
|
|
|
Seed set to 23
|
|
|
Time to instantiate model: 0.30 seconds.
|
|
|
Total parameters: 234,914,304
|
|
|
Validating ...
|
|
|
Measured TFLOPs: 6940.12
|
|
|
Epoch 1 | iter 512 step 1 | loss train: 2.698, val: 2.522 | iter time: 675.96 ms (step) remaining time: 9:49:31
|
|
|
Epoch 1 | iter 1024 step 2 | loss train: 2.627, val: 2.522 | iter time: 603.66 ms (step) remaining time: 9:19:41
|
|
|
Epoch 1 | iter 1536 step 3 | loss train: 2.653, val: 2.522 | iter time: 604.66 ms (step) remaining time: 9:06:15
|
|
|
Epoch 1 | iter 2048 step 4 | loss train: 2.608, val: 2.522 | iter time: 606.23 ms (step) remaining time: 8:57:08
|
|
|
Epoch 1 | iter 2560 step 5 | loss train: 2.604, val: 2.522 | iter time: 605.04 ms (step) remaining time: 8:49:43
|
|
|
Epoch 1 | iter 3072 step 6 | loss train: 2.578, val: 2.522 | iter time: 606.32 ms (step) remaining time: 8:43:08
|
|
|
Epoch 1 | iter 3584 step 7 | loss train: 2.692, val: 2.522 | iter time: 605.08 ms (step) remaining time: 8:37:01
|
|
|
Epoch 1 | iter 4096 step 8 | loss train: 2.570, val: 2.522 | iter time: 607.54 ms (step) remaining time: 8:31:20
|
|
|
Epoch 1 | iter 4608 step 9 | loss train: 2.646, val: 2.522 | iter time: 607.19 ms (step) remaining time: 8:25:47
|
|
|
Epoch 1 | iter 5120 step 10 | loss train: 2.565, val: 2.522 | iter time: 604.76 ms (step) remaining time: 8:20:23
|
|
|
# ...
|
|
|
Epoch 1 | iter 51712 step 101 | loss train: 2.562, val: 2.453 | iter time: 607.12 ms (step) remaining time: 0:48:29
|
|
|
Epoch 1 | iter 52224 step 102 | loss train: 2.637, val: 2.453 | iter time: 605.46 ms (step) remaining time: 0:43:31
|
|
|
Epoch 1 | iter 52736 step 103 | loss train: 2.629, val: 2.453 | iter time: 604.15 ms (step) remaining time: 0:38:34
|
|
|
Epoch 1 | iter 53248 step 104 | loss train: 2.629, val: 2.453 | iter time: 605.92 ms (step) remaining time: 0:33:36
|
|
|
Epoch 1 | iter 53760 step 105 | loss train: 2.606, val: 2.453 | iter time: 604.48 ms (step) remaining time: 0:28:38
|
|
|
Epoch 1 | iter 54272 step 106 | loss train: 2.581, val: 2.453 | iter time: 603.78 ms (step) remaining time: 0:23:41
|
|
|
Epoch 1 | iter 54784 step 107 | loss train: 2.580, val: 2.453 | iter time: 605.41 ms (step) remaining time: 0:18:43
|
|
|
Epoch 1 | iter 55296 step 108 | loss train: 2.602, val: 2.453 | iter time: 607.38 ms (step) remaining time: 0:13:46
|
|
|
Epoch 1 | iter 55808 step 109 | loss train: 2.633, val: 2.453 | iter time: 606.06 ms (step) remaining time: 0:08:49
|
|
|
Epoch 1 | iter 56320 step 110 | loss train: 2.631, val: 2.453 | iter time: 608.68 ms (step) remaining time: 0:03:51
|
|
|
Validating ...
|
|
|
iter 56320: val loss 2.4515, val time: 19303.40 ms
|
|
|
Saving checkpoint to '../out/pretrain-core-3/step-00000110/lit_model.pth'
|
|
|
Validating ...
|
|
|
Final evaluation | val loss: 2.451 | val ppl: 11.605
|
|
|
Saving checkpoint to '../out/pretrain-core-3/final/lit_model.pth'
|
|
|
----------------------------------------
|
|
|
| Performance
|
|
|
| - Total tokens : 464,642,048
|
|
|
| - Training Time : 33018.19 s
|
|
|
| - Tok/sec : 362.46 tok/s
|
|
|
| ----------------------------------------
|
|
|
| Memory Usage
|
|
|
| - Memory Used : 22.33 GB
|
|
|
----------------------------------------
|
|
|
```
|
|
|
|
|
|
```bash
|
|
|
cp ../config-3.json ../out/pretrain-core-3/final/config.json
|
|
|
```
|
|
|
|
|
|
```bash
|
|
|
mv wandb wandb-pretrain-core-3
|
|
|
```
|
|
|
|
|
|
```bash
|
|
|
CUDA_VISIBLE_DEVICES=0 CUDA_LAUNCH_BLOCKING=0 PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True litgpt chat ../out/pretrain-core-3/final
|
|
|
```
|
|
|
|
|
|
```bash
|
|
|
litgpt convert_pretrained_checkpoint ../out/pretrain-core-3/final ../out/pretrain-core-3/checkpoint
|
|
|
```
|
|
|
|
|
|
```bash
|
|
|
litgpt convert_from_litgpt ../out/pretrain-core-3/final ../out/cpt-core-pre-4
|
|
|
cp ../config-4.json ../out/cpt-core-pre-4/config.json
|
|
|
cp -rv ../tokenizer/* ../out/cpt-core-pre-4
|
|
|
python -B convert_pth_to_safetensors.py
|
|
|
```
|
|
|
|
|
|
```bash
|
|
|
CUDA_VISIBLE_DEVICES=0 python -B cpt_core_model_4.py
|
|
|
```
|
|
|
|
|
|
```
|
|
|
🦥 Unsloth: Will patch your computer to enable 2x faster free finetuning.
|
|
|
🦥 Unsloth Zoo will now patch everything to make training faster!
|
|
|
==((====))== Unsloth 2025.3.14: Fast Llama patching. Transformers: 4.49.0.
|
|
|
\\ /| NVIDIA GeForce RTX 3090. Num GPUs = 1. Max memory: 23.58 GB. Platform: Linux.
|
|
|
O^O/ \_/ \ Torch: 2.6.0+cu124. CUDA: 8.6. CUDA Toolkit: 12.4. Triton: 3.2.0
|
|
|
\ / Bfloat16 = TRUE. FA [Xformers = 0.0.29.post3. FA2 = False]
|
|
|
"-____-" Free license: http://github.com/unslothai/unsloth
|
|
|
Unsloth: Fast downloading is enabled - ignore downloading bars which are red colored!
|
|
|
Unsloth 2025.3.14 patched 32 layers with 32 QKV layers, 32 O layers and 32 MLP layers.
|
|
|
Unsloth: Training embed_tokens in mixed precision to save VRAM
|
|
|
Unsloth: Training lm_head in mixed precision to save VRAM
|
|
|
==((====))== Unsloth - 2x faster free finetuning | Num GPUs used = 1
|
|
|
\\ /| Num examples = 37,031 | Num Epochs = 2 | Total steps = 37,406
|
|
|
O^O/ \_/ \ Batch size per device = 1 | Gradient accumulation steps = 1
|
|
|
\ / Data Parallel GPUs = 1 | Total batch size (1 x 1 x 1) = 1
|
|
|
"-____-" Trainable parameters = 247,463,936/482,378,240 (51.30% trained)
|
|
|
wandb: Using wandb-core as the SDK backend. Please refer to https://wandb.me/wandb-core for more information.
|
|
|
wandb: Currently logged in as: mtasic85 to https://api.wandb.ai. Use `wandb login --relogin` to force relogin
|
|
|
wandb: Tracking run with wandb version 0.19.8
|
|
|
wandb: Run data is saved locally in /home/tangled/tangled-alpha-0.9-core/scripts/wandb/run-20250315_170030-4hx8rh26
|
|
|
wandb: Run `wandb offline` to turn off syncing.
|
|
|
wandb: Syncing run cpt-core-4
|
|
|
wandb: ⭐️ View project at https://wandb.ai/mtasic85/tangled-alpha-0.9-core
|
|
|
wandb: 🚀 View run at https://wandb.ai/mtasic85/tangled-alpha-0.9-core/runs/4hx8rh26
|
|
|
{'loss': 1.7227, 'grad_norm': 0.4360397756099701, 'learning_rate': 4.999999991182871e-05, 'epoch': 0.0}
|
|
|
0%| | 1/37406 [00:03<32:10:32, 3.10s/it]
|
|
|
Unsloth: Will smartly offload gradients to save VRAM!
|
|
|
{'loss': 3.296, 'grad_norm': 0.6972701549530029, 'learning_rate': 4.999999964731482e-05, 'epoch': 0.0}
|
|
|
{'loss': 1.7377, 'grad_norm': 0.5415557026863098, 'learning_rate': 4.9999999206458345e-05, 'epoch': 0.0}
|
|
|
{'loss': 1.774, 'grad_norm': 0.3695605993270874, 'learning_rate': 4.99999985892593e-05, 'epoch': 0.0}
|
|
|
0%| | 4/37406 [00:09<24:03:55, 2.32s/it]
|
|
|
# ...
|
|
|
```
|
|
|
|
