[2025-11-16 23:09:34,342] [INFO] [axolotl.utils.data.sft._load_raw_datasets:320] [PID:33975] Loading raw datasets...
[2025-11-16 23:09:35,399] [WARNING] [huggingface_hub.repocard.content:108] [PID:33975] Repo card metadata block was not found. Setting CardData to empty.
[2025-11-16 23:09:36,684] [INFO] [axolotl.utils.data.wrappers.get_dataset_wrapper:87] [PID:33975] Loading dataset: allura-org/the-anarchist-library with base_type: completion and prompt_style: None

Tokenizing Prompts (num_proc=24):   0%|                                               | 0/7770 [00:00<?, ? examples/s]
Tokenizing Prompts (num_proc=24):   4%|█▌                                   | 324/7770 [00:10<03:52, 32.09 examples/s]
Tokenizing Prompts (num_proc=24):   8%|███                                  | 648/7770 [00:14<02:27, 48.21 examples/s]
Tokenizing Prompts (num_proc=24):  17%|██████                              | 1296/7770 [00:17<01:10, 91.95 examples/s]
Tokenizing Prompts (num_proc=24):  21%|███████▎                           | 1620/7770 [00:17<00:47, 128.75 examples/s]
Tokenizing Prompts (num_proc=24):  25%|████████▊                          | 1944/7770 [00:18<00:32, 180.84 examples/s]
Tokenizing Prompts (num_proc=24):  29%|██████████▏                        | 2268/7770 [00:18<00:22, 246.23 examples/s]
Tokenizing Prompts (num_proc=24):  33%|███████████▋                       | 2592/7770 [00:19<00:20, 257.06 examples/s]
Tokenizing Prompts (num_proc=24):  42%|██████████████▌                    | 3239/7770 [00:20<00:11, 399.17 examples/s]
Tokenizing Prompts (num_proc=24):  46%|████████████████                   | 3563/7770 [00:20<00:09, 433.08 examples/s]
Tokenizing Prompts (num_proc=24):  50%|█████████████████▌                 | 3887/7770 [00:21<00:08, 469.75 examples/s]
Tokenizing Prompts (num_proc=24):  54%|██████████████████▉                | 4210/7770 [00:22<00:08, 432.84 examples/s]
Tokenizing Prompts (num_proc=24):  58%|████████████████████▍              | 4534/7770 [00:22<00:07, 417.73 examples/s]
Tokenizing Prompts (num_proc=24):  63%|█████████████████████▉             | 4858/7770 [00:23<00:06, 438.43 examples/s]
Tokenizing Prompts (num_proc=24):  67%|███████████████████████▎           | 5182/7770 [00:23<00:04, 571.54 examples/s]
Tokenizing Prompts (num_proc=24):  71%|████████████████████████▊          | 5506/7770 [00:24<00:03, 632.83 examples/s]
Tokenizing Prompts (num_proc=24):  75%|██████████████████████████▎        | 5830/7770 [00:24<00:03, 589.22 examples/s]
Tokenizing Prompts (num_proc=24):  79%|███████████████████████████▋       | 6154/7770 [00:27<00:05, 289.62 examples/s]
Tokenizing Prompts (num_proc=24):  83%|█████████████████████████████▏     | 6477/7770 [00:27<00:03, 367.92 examples/s]
Tokenizing Prompts (num_proc=24):  96%|█████████████████████████████████▌ | 7447/7770 [00:27<00:00, 742.90 examples/s]
Tokenizing Prompts (num_proc=24): 100%|███████████████████████████████████| 7770/7770 [00:29<00:00, 406.44 examples/s]
Tokenizing Prompts (num_proc=24): 100%|███████████████████████████████████| 7770/7770 [00:30<00:00, 255.97 examples/s]

Dropping Long Sequences (>2048) (num_proc=24):   0%|                                 | 0/53143 [00:00<?, ? examples/s]
Dropping Long Sequences (>2048) (num_proc=24):   2%|▍                    | 1000/53143 [00:01<01:00, 857.68 examples/s]
Dropping Long Sequences (>2048) (num_proc=24):  21%|███▋              | 11000/53143 [00:01<00:03, 11670.65 examples/s]
Dropping Long Sequences (>2048) (num_proc=24):  36%|██████▍           | 19000/53143 [00:01<00:01, 20204.76 examples/s]
Dropping Long Sequences (>2048) (num_proc=24):  49%|████████▊         | 26000/53143 [00:01<00:01, 18256.01 examples/s]
Dropping Long Sequences (>2048) (num_proc=24):  70%|████████████▌     | 37000/53143 [00:01<00:00, 29833.50 examples/s]
Dropping Long Sequences (>2048) (num_proc=24):  83%|██████████████▉   | 44143/53143 [00:02<00:00, 36188.95 examples/s]
Dropping Long Sequences (>2048) (num_proc=24):  97%|█████████████████▍| 51500/53143 [00:02<00:00, 35504.31 examples/s]
Dropping Long Sequences (>2048) (num_proc=24): 100%|██████████████████| 53143/53143 [00:02<00:00, 20670.58 examples/s]

Drop Samples with Zero Trainable Tokens (num_proc=24):   0%|                         | 0/53142 [00:00<?, ? examples/s]
Drop Samples with Zero Trainable Tokens (num_proc=24):   2%|▏            | 1000/53142 [00:01<01:01, 843.03 examples/s]
Drop Samples with Zero Trainable Tokens (num_proc=24):  40%|███▉      | 21000/53142 [00:01<00:01, 21873.84 examples/s]
Drop Samples with Zero Trainable Tokens (num_proc=24):  60%|██████    | 32000/53142 [00:01<00:01, 19954.86 examples/s]
Drop Samples with Zero Trainable Tokens (num_proc=24):  84%|████████▎ | 44430/53142 [00:02<00:00, 30297.23 examples/s]
Drop Samples with Zero Trainable Tokens (num_proc=24): 100%|██████████| 53142/53142 [00:02<00:00, 27727.04 examples/s]
Drop Samples with Zero Trainable Tokens (num_proc=24): 100%|██████████| 53142/53142 [00:02<00:00, 20503.23 examples/s]

Add position_id column (Sample Packing) (num_proc=24):   0%|                         | 0/53142 [00:00<?, ? examples/s]
Add position_id column (Sample Packing) (num_proc=24):   2%|▏            | 1000/53142 [00:01<01:30, 577.59 examples/s]
Add position_id column (Sample Packing) (num_proc=24):  41%|████▏     | 22000/53142 [00:01<00:01, 16323.83 examples/s]
Add position_id column (Sample Packing) (num_proc=24):  62%|██████▏   | 33000/53142 [00:03<00:01, 11604.22 examples/s]
Add position_id column (Sample Packing) (num_proc=24):  89%|████████▉ | 47215/53142 [00:03<00:00, 18676.16 examples/s]
Add position_id column (Sample Packing) (num_proc=24): 100%|██████████| 53142/53142 [00:04<00:00, 12413.83 examples/s]

Saving the dataset (0/24 shards):   0%|                                              | 0/53142 [00:00<?, ? examples/s]
Saving the dataset (0/24 shards):   2%|▌                                | 1000/53142 [00:00<00:29, 1776.46 examples/s]
Saving the dataset (1/24 shards):  23%|███████▎                        | 12215/53142 [00:00<00:23, 1776.46 examples/s]
Saving the dataset (2/24 shards):  33%|██████████▍                     | 17430/53142 [00:00<00:20, 1776.46 examples/s]
Saving the dataset (3/24 shards):  37%|███████████▊                    | 19645/53142 [00:00<00:18, 1776.46 examples/s]
Saving the dataset (4/24 shards):  46%|██████████████▋                 | 24289/53142 [00:00<00:16, 1776.46 examples/s]
Saving the dataset (5/24 shards):  46%|██████████████▋                 | 24289/53142 [00:00<00:16, 1776.46 examples/s]
Saving the dataset (6/24 shards):  46%|██████████████▋                 | 24289/53142 [00:00<00:16, 1776.46 examples/s]
Saving the dataset (7/24 shards):  52%|████████████████▌               | 27504/53142 [00:00<00:14, 1776.46 examples/s]
Saving the dataset (8/24 shards):  58%|██████████████████▍             | 30718/53142 [00:00<00:12, 1776.46 examples/s]
Saving the dataset (9/24 shards):  68%|█████████████████████▋          | 35932/53142 [00:00<00:09, 1776.46 examples/s]
Saving the dataset (10/24 shards):  68%|█████████████████████          | 36146/53142 [00:00<00:09, 1776.46 examples/s]
Saving the dataset (11/24 shards):  68%|█████████████████████▏         | 36360/53142 [00:00<00:09, 1776.46 examples/s]
Saving the dataset (12/24 shards):  73%|██████████████████████▌        | 38574/53142 [00:00<00:08, 1776.46 examples/s]
Saving the dataset (13/24 shards):  73%|██████████████████████▋        | 38788/53142 [00:00<00:08, 1776.46 examples/s]
Saving the dataset (14/24 shards):  85%|██████████████████████████▎    | 45002/53142 [00:00<00:04, 1776.46 examples/s]
Saving the dataset (15/24 shards):  85%|██████████████████████████▍    | 45216/53142 [00:00<00:04, 1776.46 examples/s]
Saving the dataset (16/24 shards):  85%|██████████████████████████▌    | 45430/53142 [00:00<00:04, 1776.46 examples/s]
Saving the dataset (17/24 shards):  86%|██████████████████████████▊    | 45858/53142 [00:00<00:04, 1776.46 examples/s]
Saving the dataset (18/24 shards):  86%|██████████████████████████▊    | 45858/53142 [00:00<00:04, 1776.46 examples/s]
Saving the dataset (19/24 shards):  89%|███████████████████████████▍   | 47072/53142 [00:00<00:03, 1776.46 examples/s]
Saving the dataset (20/24 shards):  91%|████████████████████████████▏  | 48286/53142 [00:00<00:02, 1776.46 examples/s]
Saving the dataset (21/24 shards):  93%|████████████████████████████▉  | 49500/53142 [00:00<00:02, 1776.46 examples/s]
Saving the dataset (22/24 shards):  95%|█████████████████████████████▌ | 50714/53142 [00:00<00:01, 1776.46 examples/s]
Saving the dataset (22/24 shards):  98%|████████████████████████████▎| 51928/53142 [00:00<00:00, 104138.06 examples/s]
Saving the dataset (23/24 shards):  98%|████████████████████████████▎| 51928/53142 [00:00<00:00, 104138.06 examples/s]
Saving the dataset (24/24 shards): 100%|█████████████████████████████| 53142/53142 [00:00<00:00, 104138.06 examples/s]
Saving the dataset (24/24 shards): 100%|██████████████████████████████| 53142/53142 [00:00<00:00, 72003.81 examples/s]
Using Liger RMSNorm!

Loading checkpoint shards:   0%|                                                                | 0/5 [00:00<?, ?it/s]
Loading checkpoint shards: 100%|███████████████████████████████████████████████████████| 5/5 [00:00<00:00, 446.06it/s]

Loading checkpoint shards:   0%|                                                                | 0/5 [00:00<?, ?it/s]
Loading checkpoint shards: 100%|███████████████████████████████████████████████████████| 5/5 [00:00<00:00, 422.96it/s]
[2025-11-17 04:48:45,282] [WARNING] [py.warnings._showwarnmsg:110] [PID:33975] /root/miniconda3/envs/py3.11/lib/python3.11/site-packages/torch/distributed/fsdp/fully_sharded_data_parallel.py:680: FutureWarning: FSDP.state_dict_type() and FSDP.set_state_dict_type() are being deprecated. Please use APIs, get_state_dict() and set_state_dict(), which can support different parallelisms, FSDP1, FSDP2, DDP. API doc: https://pytorch.org/docs/stable/distributed.checkpoint.html#torch.distributed.checkpoint.state_dict.get_state_dict .Tutorial: https://pytorch.org/tutorials/recipes/distributed_checkpoint_recipe.html .
  warnings.warn(