File size: 17,005 Bytes
c3ecbbe |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 |
2024-07-17 06:14:19,185 - INFO: Calling run..
2024-07-17 06:14:19,186 - INFO: Environment configuration: ConfigNLPCausalClassificationEnvironment(gpus=['0'], mixed_precision=False, compile_model=False, use_deepspeed=False, deepspeed_reduce_bucket_size=10000000.0, deepspeed_stage3_prefetch_bucket_size=10000000.0, deepspeed_stage3_param_persistence_threshold=10000000.0, deepspeed_offload_optimizer=False, deepspeed_stage3_max_live_parameters=10000000.0, deepspeed_stage3_max_reuse_distance=10000000.0, find_unused_parameters=False, trust_remote_code=False, huggingface_branch='main', number_of_workers=8, seed=-1, _seed=0, _distributed=False, _distributed_inference=True, _local_rank=0, _world_size=1, _curr_step=0, _curr_val_step=0, _rank=0, _device='cuda', _cpu_comm=None, _model_card_template='text_causal_classification_model_card_template.md', _summary_card_template='text_causal_classification_experiment_summary_card_template.md')
2024-07-17 06:14:19,186 - INFO: cfg.environment._distributed set to False
2024-07-17 06:14:19,186 - INFO: Problem Type: text_causal_classification_modeling
2024-07-17 06:14:19,186 - INFO: Global random seed: 419783
2024-07-17 06:14:19,186 - INFO: Preparing the data...
2024-07-17 06:14:19,186 - INFO: Setting up automatic validation split...
2024-07-17 06:14:19,192 - INFO: The dataframe has following columns: Index(['Description', 'category', 'sub_category', 'label'], dtype='object')
2024-07-17 06:14:19,195 - INFO: Preparing train and validation data, dataset config to be used: ConfigNLPCausalClassificationDataset(dataset_class=<class 'llm_studio.src.datasets.text_causal_classification_ds.CustomDataset'>, personalize=False, chatbot_name='OI_AI', chatbot_author='openinnovation.ai', train_dataframe='/app/train_df.csv', validation_strategy='automatic', validation_dataframe='/app/validation_df.csv', validation_size=0.0099999998, data_sample=1.0, data_sample_choice=('Train', 'Validation'), system_column='None', prompt_column=(), answer_column='category', parent_id_column='None', text_system_start='', text_prompt_start='', text_answer_separator='', limit_chained_samples=False, add_eos_token_to_system=False, add_eos_token_to_prompt=False, add_eos_token_to_answer=False, mask_prompt_labels=True, _allowed_file_extensions=('csv', 'pq', 'parquet'), num_classes=2)
2024-07-17 06:14:19,195 - INFO: Loading train dataset...
2024-07-17 06:14:19,195 - INFO: Columns found: Index(['Description', 'category', 'sub_category', 'label'], dtype='object')
2024-07-17 06:14:20,210 - INFO: Loading validation dataset...
2024-07-17 06:14:20,791 - INFO: Number of observations in train dataset: 494
2024-07-17 06:14:20,791 - INFO: Number of observations in validation dataset: 5
2024-07-17 06:14:21,246 - WARNING: PAD token id not matching between config and tokenizer. Overwriting with tokenizer id.
2024-07-17 06:14:21,246 - INFO: Setting pretraining_tp of model config to 1.
2024-07-17 06:14:21,251 - INFO: Using int4 for backbone
2024-07-17 06:14:21,251 - INFO: Loading TinyLlama/TinyLlama_v1.1. This may take a while.
2024-07-17 06:14:35,909 - INFO: Loaded TinyLlama/TinyLlama_v1.1.
2024-07-17 06:14:35,916 - INFO: Lora module names: ['q_proj', 'k_proj', 'v_proj', 'o_proj', 'gate_proj', 'up_proj', 'down_proj']
2024-07-17 06:14:36,191 - INFO: Enough space available for saving model weights.Required space: 1003.87MB, Available space: 993953.90MB.
2024-07-17 06:14:36,200 - INFO: Optimizer AdamW has been provided with parameters {'weight_decay': 0.0, 'eps': 1e-08, 'betas': (0.8999999762, 0.9990000129), 'lr': 0.0001}
2024-07-17 06:14:36,637 - INFO: started process: 0, can_track: True, tracking_mode: TrackingMode.AFTER_EPOCH
2024-07-17 06:14:36,638 - INFO: Training Epoch: 1 / 1
2024-07-17 06:14:36,638 - INFO: train loss: 0%| | 0/247 [00:00<?, ?it/s]
2024-07-17 06:14:36,787 - INFO: Evaluation step: 247
2024-07-17 06:14:40,478 - INFO: train loss: 0.69: 1%| | 2/247 [00:03<07:50, 1.92s/it]
2024-07-17 06:14:44,085 - INFO: train loss: 0.69: 2%|1 | 4/247 [00:07<07:29, 1.85s/it]
2024-07-17 06:14:47,702 - INFO: train loss: 0.69: 2%|2 | 6/247 [00:11<07:21, 1.83s/it]
2024-07-17 06:14:51,319 - INFO: train loss: 0.69: 3%|3 | 8/247 [00:14<07:15, 1.82s/it]
2024-07-17 06:14:54,943 - INFO: train loss: 0.69: 4%|4 | 10/247 [00:18<07:11, 1.82s/it]
2024-07-17 06:14:58,568 - INFO: train loss: 0.69: 5%|4 | 12/247 [00:21<07:06, 1.82s/it]
2024-07-17 06:15:02,198 - INFO: train loss: 0.69: 6%|5 | 14/247 [00:25<07:03, 1.82s/it]
2024-07-17 06:15:05,836 - INFO: train loss: 0.69: 6%|6 | 16/247 [00:29<06:59, 1.82s/it]
2024-07-17 06:15:09,477 - INFO: train loss: 0.69: 7%|7 | 18/247 [00:32<06:56, 1.82s/it]
2024-07-17 06:15:13,118 - INFO: train loss: 0.69: 8%|8 | 20/247 [00:36<06:52, 1.82s/it]
2024-07-17 06:15:16,760 - INFO: train loss: 0.69: 9%|8 | 22/247 [00:40<06:49, 1.82s/it]
2024-07-17 06:15:20,405 - INFO: train loss: 0.69: 10%|9 | 24/247 [00:43<06:45, 1.82s/it]
2024-07-17 06:15:24,053 - INFO: train loss: 0.69: 11%|# | 26/247 [00:47<06:42, 1.82s/it]
2024-07-17 06:15:27,704 - INFO: train loss: 0.69: 11%|#1 | 28/247 [00:51<06:39, 1.82s/it]
2024-07-17 06:15:31,359 - INFO: train loss: 0.69: 12%|#2 | 30/247 [00:54<06:35, 1.82s/it]
2024-07-17 06:15:35,012 - INFO: train loss: 0.69: 13%|#2 | 32/247 [00:58<06:32, 1.82s/it]
2024-07-17 06:15:38,664 - INFO: train loss: 0.69: 14%|#3 | 34/247 [01:02<06:28, 1.83s/it]
2024-07-17 06:15:42,324 - INFO: train loss: 0.69: 15%|#4 | 36/247 [01:05<06:25, 1.83s/it]
2024-07-17 06:15:45,991 - INFO: train loss: 0.69: 15%|#5 | 38/247 [01:09<06:22, 1.83s/it]
2024-07-17 06:15:49,653 - INFO: train loss: 0.69: 16%|#6 | 40/247 [01:13<06:18, 1.83s/it]
2024-07-17 06:15:53,319 - INFO: train loss: 0.69: 17%|#7 | 42/247 [01:16<06:15, 1.83s/it]
2024-07-17 06:15:56,988 - INFO: train loss: 0.69: 18%|#7 | 44/247 [01:20<06:11, 1.83s/it]
2024-07-17 06:16:00,655 - INFO: train loss: 0.69: 19%|#8 | 46/247 [01:24<06:08, 1.83s/it]
2024-07-17 06:16:04,324 - INFO: train loss: 0.69: 19%|#9 | 48/247 [01:27<06:04, 1.83s/it]
2024-07-17 06:16:07,995 - INFO: train loss: 0.69: 20%|## | 50/247 [01:31<06:01, 1.83s/it]
2024-07-17 06:16:11,667 - INFO: train loss: 0.69: 21%|##1 | 52/247 [01:35<05:57, 1.83s/it]
2024-07-17 06:16:15,344 - INFO: train loss: 0.69: 22%|##1 | 54/247 [01:38<05:54, 1.84s/it]
2024-07-17 06:16:19,020 - INFO: train loss: 0.69: 23%|##2 | 56/247 [01:42<05:50, 1.84s/it]
2024-07-17 06:16:22,696 - INFO: train loss: 0.69: 23%|##3 | 58/247 [01:46<05:47, 1.84s/it]
2024-07-17 06:16:26,371 - INFO: train loss: 0.69: 24%|##4 | 60/247 [01:49<05:43, 1.84s/it]
2024-07-17 06:16:30,047 - INFO: train loss: 0.69: 25%|##5 | 62/247 [01:53<05:39, 1.84s/it]
2024-07-17 06:16:33,723 - INFO: train loss: 0.69: 26%|##5 | 64/247 [01:57<05:36, 1.84s/it]
2024-07-17 06:16:37,401 - INFO: train loss: 0.69: 27%|##6 | 66/247 [02:00<05:32, 1.84s/it]
2024-07-17 06:16:41,082 - INFO: train loss: 0.69: 28%|##7 | 68/247 [02:04<05:29, 1.84s/it]
2024-07-17 06:16:44,762 - INFO: train loss: 0.69: 28%|##8 | 70/247 [02:08<05:25, 1.84s/it]
2024-07-17 06:16:48,448 - INFO: train loss: 0.69: 29%|##9 | 72/247 [02:11<05:22, 1.84s/it]
2024-07-17 06:16:52,124 - INFO: train loss: 0.69: 30%|##9 | 74/247 [02:15<05:18, 1.84s/it]
2024-07-17 06:16:55,803 - INFO: train loss: 0.69: 31%|### | 76/247 [02:19<05:14, 1.84s/it]
2024-07-17 06:16:59,486 - INFO: train loss: 0.69: 32%|###1 | 78/247 [02:22<05:10, 1.84s/it]
2024-07-17 06:17:03,165 - INFO: train loss: 0.69: 32%|###2 | 80/247 [02:26<05:07, 1.84s/it]
2024-07-17 06:17:06,841 - INFO: train loss: 0.69: 33%|###3 | 82/247 [02:30<05:03, 1.84s/it]
2024-07-17 06:17:10,530 - INFO: train loss: 0.69: 34%|###4 | 84/247 [02:33<05:00, 1.84s/it]
2024-07-17 06:17:14,215 - INFO: train loss: 0.69: 35%|###4 | 86/247 [02:37<04:56, 1.84s/it]
2024-07-17 06:17:17,898 - INFO: train loss: 0.69: 36%|###5 | 88/247 [02:41<04:52, 1.84s/it]
2024-07-17 06:17:21,582 - INFO: train loss: 0.69: 36%|###6 | 90/247 [02:44<04:49, 1.84s/it]
2024-07-17 06:17:25,270 - INFO: train loss: 0.69: 37%|###7 | 92/247 [02:48<04:45, 1.84s/it]
2024-07-17 06:17:28,955 - INFO: train loss: 0.69: 38%|###8 | 94/247 [02:52<04:41, 1.84s/it]
2024-07-17 06:17:32,638 - INFO: train loss: 0.69: 39%|###8 | 96/247 [02:55<04:38, 1.84s/it]
2024-07-17 06:17:36,324 - INFO: train loss: 0.69: 40%|###9 | 98/247 [02:59<04:34, 1.84s/it]
2024-07-17 06:17:40,011 - INFO: train loss: 0.69: 40%|#### | 100/247 [03:03<04:30, 1.84s/it]
2024-07-17 06:17:43,698 - INFO: train loss: 0.69: 41%|####1 | 102/247 [03:07<04:27, 1.84s/it]
2024-07-17 06:17:47,382 - INFO: train loss: 0.69: 42%|####2 | 104/247 [03:10<04:23, 1.84s/it]
2024-07-17 06:17:51,066 - INFO: train loss: 0.69: 43%|####2 | 106/247 [03:14<04:19, 1.84s/it]
2024-07-17 06:17:54,758 - INFO: train loss: 0.69: 44%|####3 | 108/247 [03:18<04:16, 1.84s/it]
2024-07-17 06:17:58,443 - INFO: train loss: 0.69: 45%|####4 | 110/247 [03:21<04:12, 1.84s/it]
2024-07-17 06:18:02,129 - INFO: train loss: 0.69: 45%|####5 | 112/247 [03:25<04:08, 1.84s/it]
2024-07-17 06:18:05,813 - INFO: train loss: 0.69: 46%|####6 | 114/247 [03:29<04:05, 1.84s/it]
2024-07-17 06:18:09,498 - INFO: train loss: 0.69: 47%|####6 | 116/247 [03:32<04:01, 1.84s/it]
2024-07-17 06:18:13,182 - INFO: train loss: 0.69: 48%|####7 | 118/247 [03:36<03:57, 1.84s/it]
2024-07-17 06:18:16,869 - INFO: train loss: 0.69: 49%|####8 | 120/247 [03:40<03:54, 1.84s/it]
2024-07-17 06:18:20,560 - INFO: train loss: 0.69: 49%|####9 | 122/247 [03:43<03:50, 1.84s/it]
2024-07-17 06:18:24,248 - INFO: train loss: 0.69: 50%|##### | 124/247 [03:47<03:46, 1.84s/it]
2024-07-17 06:18:27,937 - INFO: train loss: 0.69: 51%|#####1 | 126/247 [03:51<03:43, 1.84s/it]
2024-07-17 06:18:31,621 - INFO: train loss: 0.69: 52%|#####1 | 128/247 [03:54<03:39, 1.84s/it]
2024-07-17 06:18:35,310 - INFO: train loss: 0.69: 53%|#####2 | 130/247 [03:58<03:35, 1.84s/it]
2024-07-17 06:18:38,999 - INFO: train loss: 0.69: 53%|#####3 | 132/247 [04:02<03:32, 1.84s/it]
2024-07-17 06:18:42,685 - INFO: train loss: 0.69: 54%|#####4 | 134/247 [04:06<03:28, 1.84s/it]
2024-07-17 06:18:46,373 - INFO: train loss: 0.69: 55%|#####5 | 136/247 [04:09<03:24, 1.84s/it]
2024-07-17 06:18:50,063 - INFO: train loss: 0.69: 56%|#####5 | 138/247 [04:13<03:21, 1.84s/it]
2024-07-17 06:18:53,753 - INFO: train loss: 0.69: 57%|#####6 | 140/247 [04:17<03:17, 1.84s/it]
2024-07-17 06:18:57,439 - INFO: train loss: 0.69: 57%|#####7 | 142/247 [04:20<03:13, 1.84s/it]
2024-07-17 06:19:01,127 - INFO: train loss: 0.69: 58%|#####8 | 144/247 [04:24<03:09, 1.84s/it]
2024-07-17 06:19:04,818 - INFO: train loss: 0.69: 59%|#####9 | 146/247 [04:28<03:06, 1.84s/it]
2024-07-17 06:19:08,506 - INFO: train loss: 0.69: 60%|#####9 | 148/247 [04:31<03:02, 1.84s/it]
2024-07-17 06:19:12,194 - INFO: train loss: 0.69: 61%|###### | 150/247 [04:35<02:58, 1.84s/it]
2024-07-17 06:19:15,881 - INFO: train loss: 0.69: 62%|######1 | 152/247 [04:39<02:55, 1.84s/it]
2024-07-17 06:19:19,570 - INFO: train loss: 0.69: 62%|######2 | 154/247 [04:42<02:51, 1.84s/it]
2024-07-17 06:19:23,257 - INFO: train loss: 0.69: 63%|######3 | 156/247 [04:46<02:47, 1.84s/it]
2024-07-17 06:19:26,952 - INFO: train loss: 0.69: 64%|######3 | 158/247 [04:50<02:44, 1.84s/it]
2024-07-17 06:19:30,641 - INFO: train loss: 0.69: 65%|######4 | 160/247 [04:54<02:40, 1.84s/it]
2024-07-17 06:19:34,331 - INFO: train loss: 0.69: 66%|######5 | 162/247 [04:57<02:36, 1.84s/it]
2024-07-17 06:19:38,020 - INFO: train loss: 0.69: 66%|######6 | 164/247 [05:01<02:33, 1.84s/it]
2024-07-17 06:19:41,710 - INFO: train loss: 0.69: 67%|######7 | 166/247 [05:05<02:29, 1.84s/it]
2024-07-17 06:19:45,399 - INFO: train loss: 0.69: 68%|######8 | 168/247 [05:08<02:25, 1.84s/it]
2024-07-17 06:19:49,092 - INFO: train loss: 0.69: 69%|######8 | 170/247 [05:12<02:22, 1.85s/it]
2024-07-17 06:19:52,778 - INFO: train loss: 0.69: 70%|######9 | 172/247 [05:16<02:18, 1.84s/it]
2024-07-17 06:19:56,471 - INFO: train loss: 0.69: 70%|####### | 174/247 [05:19<02:14, 1.85s/it]
2024-07-17 06:20:00,161 - INFO: train loss: 0.69: 71%|#######1 | 176/247 [05:23<02:11, 1.85s/it]
2024-07-17 06:20:03,851 - INFO: train loss: 0.69: 72%|#######2 | 178/247 [05:27<02:07, 1.85s/it]
2024-07-17 06:20:07,541 - INFO: train loss: 0.69: 73%|#######2 | 180/247 [05:30<02:03, 1.85s/it]
2024-07-17 06:20:11,231 - INFO: train loss: 0.69: 74%|#######3 | 182/247 [05:34<01:59, 1.85s/it]
2024-07-17 06:20:14,920 - INFO: train loss: 0.69: 74%|#######4 | 184/247 [05:38<01:56, 1.84s/it]
2024-07-17 06:20:18,607 - INFO: train loss: 0.69: 75%|#######5 | 186/247 [05:41<01:52, 1.84s/it]
2024-07-17 06:20:22,292 - INFO: train loss: 0.69: 76%|#######6 | 188/247 [05:45<01:48, 1.84s/it]
2024-07-17 06:20:25,982 - INFO: train loss: 0.69: 77%|#######6 | 190/247 [05:49<01:45, 1.84s/it]
2024-07-17 06:20:29,678 - INFO: train loss: 0.69: 78%|#######7 | 192/247 [05:53<01:41, 1.85s/it]
2024-07-17 06:20:33,371 - INFO: train loss: 0.69: 79%|#######8 | 194/247 [05:56<01:37, 1.85s/it]
2024-07-17 06:20:37,064 - INFO: train loss: 0.69: 79%|#######9 | 196/247 [06:00<01:34, 1.85s/it]
2024-07-17 06:20:40,756 - INFO: train loss: 0.69: 80%|######## | 198/247 [06:04<01:30, 1.85s/it]
2024-07-17 06:20:44,444 - INFO: train loss: 0.69: 81%|######## | 200/247 [06:07<01:26, 1.85s/it]
2024-07-17 06:20:48,128 - INFO: train loss: 0.69: 82%|########1 | 202/247 [06:11<01:22, 1.84s/it]
2024-07-17 06:20:51,813 - INFO: train loss: 0.69: 83%|########2 | 204/247 [06:15<01:19, 1.84s/it]
2024-07-17 06:20:55,505 - INFO: train loss: 0.69: 83%|########3 | 206/247 [06:18<01:15, 1.84s/it]
2024-07-17 06:20:59,200 - INFO: train loss: 0.69: 84%|########4 | 208/247 [06:22<01:11, 1.85s/it]
2024-07-17 06:21:02,890 - INFO: train loss: 0.69: 85%|########5 | 210/247 [06:26<01:08, 1.85s/it]
2024-07-17 06:21:06,580 - INFO: train loss: 0.69: 86%|########5 | 212/247 [06:29<01:04, 1.85s/it]
2024-07-17 06:21:10,264 - INFO: train loss: 0.69: 87%|########6 | 214/247 [06:33<01:00, 1.84s/it]
2024-07-17 06:21:13,970 - INFO: train loss: 0.69: 87%|########7 | 216/247 [06:37<00:57, 1.85s/it]
2024-07-17 06:21:17,646 - INFO: train loss: 0.69: 88%|########8 | 218/247 [06:41<00:53, 1.84s/it]
2024-07-17 06:21:21,336 - INFO: train loss: 0.69: 89%|########9 | 220/247 [06:44<00:49, 1.84s/it]
2024-07-17 06:21:25,029 - INFO: train loss: 0.69: 90%|########9 | 222/247 [06:48<00:46, 1.85s/it]
2024-07-17 06:21:28,717 - INFO: train loss: 0.69: 91%|######### | 224/247 [06:52<00:42, 1.84s/it]
2024-07-17 06:21:32,402 - INFO: train loss: 0.69: 91%|#########1| 226/247 [06:55<00:38, 1.84s/it]
2024-07-17 06:21:36,093 - INFO: train loss: 0.69: 92%|#########2| 228/247 [06:59<00:35, 1.84s/it]
2024-07-17 06:21:39,782 - INFO: train loss: 0.69: 93%|#########3| 230/247 [07:03<00:31, 1.84s/it]
2024-07-17 06:21:43,472 - INFO: train loss: 0.69: 94%|#########3| 232/247 [07:06<00:27, 1.84s/it]
2024-07-17 06:21:47,159 - INFO: train loss: 0.69: 95%|#########4| 234/247 [07:10<00:23, 1.84s/it]
2024-07-17 06:21:50,847 - INFO: train loss: 0.69: 96%|#########5| 236/247 [07:14<00:20, 1.84s/it]
2024-07-17 06:21:54,541 - INFO: train loss: 0.69: 96%|#########6| 238/247 [07:17<00:16, 1.84s/it]
2024-07-17 06:21:58,231 - INFO: train loss: 0.69: 97%|#########7| 240/247 [07:21<00:12, 1.85s/it]
2024-07-17 06:22:01,922 - INFO: train loss: 0.69: 98%|#########7| 242/247 [07:25<00:09, 1.85s/it]
2024-07-17 06:22:05,605 - INFO: train loss: 0.69: 99%|#########8| 244/247 [07:28<00:05, 1.84s/it]
2024-07-17 06:22:09,294 - INFO: train loss: 0.69: 100%|#########9| 246/247 [07:32<00:01, 1.84s/it]
2024-07-17 06:22:11,136 - INFO: train loss: 0.69: 100%|##########| 247/247 [07:34<00:00, 1.84s/it]
2024-07-17 06:22:11,136 - INFO: Saving last model checkpoint to /app/output
2024-07-17 06:22:11,136 - INFO: Saving checkpoint..
2024-07-17 06:22:12,661 - INFO: Starting validation inference
2024-07-17 06:22:12,662 - INFO: validation progress: 0%| | 0/3 [00:00<?, ?it/s]
2024-07-17 06:22:13,336 - INFO: validation progress: 33%|###3 | 1/3 [00:00<00:01, 1.48it/s]
2024-07-17 06:22:13,896 - INFO: validation progress: 67%|######6 | 2/3 [00:01<00:00, 1.65it/s]
2024-07-17 06:22:14,207 - INFO: validation progress: 100%|##########| 3/3 [00:01<00:00, 2.12it/s]
2024-07-17 06:22:14,209 - INFO: validation progress: 100%|##########| 3/3 [00:01<00:00, 1.94it/s]
2024-07-17 06:22:14,247 - INFO: Validation Perplexity: 0.69315
2024-07-17 06:22:14,247 - INFO: Mean validation loss: 0.69315
|