Hugging Face's logo

Datasets:
ParrotLabs
/
Preprocessed

Dataset card Files
xet
 Community
thaferbe commited on
Commit
7060d69
·
verified ·
1 Parent(s): 7b6b1ee

Upload training run run_20260409_192244

Browse files

ParrotLLM end-of-training sync for local run runs/big_run/exp_c/run_20260409_192244.

Files changed (25) hide show
  1. .gitattributes +1 -0
  2. runs/big_run/exp_c/run_20260409_192244/checkpoints/best_loss_3p5130_epoch_0000_step_0003000.pt +3 -0
  3. runs/big_run/exp_c/run_20260409_192244/checkpoints/best_loss_3p5190_epoch_0000_step_0002900.pt +3 -0
  4. runs/big_run/exp_c/run_20260409_192244/checkpoints/best_loss_3p5242_epoch_0000_step_0002800.pt +3 -0
  5. runs/big_run/exp_c/run_20260409_192244/checkpoints/best_loss_3p5320_epoch_0000_step_0002700.pt +3 -0
  6. runs/big_run/exp_c/run_20260409_192244/checkpoints/best_loss_3p5420_epoch_0000_step_0002600.pt +3 -0
  7. runs/big_run/exp_c/run_20260409_192244/checkpoints/best_loss_3p5520_epoch_0000_step_0002500.pt +3 -0
  8. runs/big_run/exp_c/run_20260409_192244/checkpoints/best_loss_3p5645_epoch_0000_step_0002400.pt +3 -0
  9. runs/big_run/exp_c/run_20260409_192244/checkpoints/best_loss_3p5771_epoch_0000_step_0002300.pt +3 -0
  10. runs/big_run/exp_c/run_20260409_192244/checkpoints/best_loss_3p5905_epoch_0000_step_0002200.pt +3 -0
  11. runs/big_run/exp_c/run_20260409_192244/checkpoints/best_loss_3p6072_epoch_0000_step_0002100.pt +3 -0
  12. runs/big_run/exp_c/run_20260409_192244/checkpoints/last_epoch_0000_step_0001000.pt +3 -0
  13. runs/big_run/exp_c/run_20260409_192244/checkpoints/last_epoch_0000_step_0001250.pt +3 -0
  14. runs/big_run/exp_c/run_20260409_192244/checkpoints/last_epoch_0000_step_0001500.pt +3 -0
  15. runs/big_run/exp_c/run_20260409_192244/checkpoints/last_epoch_0000_step_0001750.pt +3 -0
  16. runs/big_run/exp_c/run_20260409_192244/checkpoints/last_epoch_0000_step_0002000.pt +3 -0
  17. runs/big_run/exp_c/run_20260409_192244/checkpoints/last_epoch_0000_step_0002250.pt +3 -0
  18. runs/big_run/exp_c/run_20260409_192244/checkpoints/last_epoch_0000_step_0002500.pt +3 -0
  19. runs/big_run/exp_c/run_20260409_192244/checkpoints/last_epoch_0000_step_0002750.pt +3 -0
  20. runs/big_run/exp_c/run_20260409_192244/checkpoints/last_epoch_0000_step_0003000.pt +3 -0
  21. runs/big_run/exp_c/run_20260409_192244/checkpoints/last_epoch_0000_step_0003052.pt +3 -0
  22. runs/big_run/exp_c/run_20260409_192244/config.json +91 -0
  23. runs/big_run/exp_c/run_20260409_192244/metrics.jsonl +0 -0
  24. runs/big_run/exp_c/run_20260409_192244/train.log +713 -0
  25. runs/big_run/exp_c/run_20260409_192244/training_plots.pdf +3 -0
.gitattributes CHANGED
@@ -61,3 +61,4 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
61
  runs/big_run/exp_c/run_20260408_124138/training_plots.pdf filter=lfs diff=lfs merge=lfs -text
62
  runs/big_run/exp_c/run_20260408_135803/training_plots.pdf filter=lfs diff=lfs merge=lfs -text
63
  runs/big_run/exp_c/run_20260408_151821/training_plots.pdf filter=lfs diff=lfs merge=lfs -text
 
 
61
  runs/big_run/exp_c/run_20260408_124138/training_plots.pdf filter=lfs diff=lfs merge=lfs -text
62
  runs/big_run/exp_c/run_20260408_135803/training_plots.pdf filter=lfs diff=lfs merge=lfs -text
63
  runs/big_run/exp_c/run_20260408_151821/training_plots.pdf filter=lfs diff=lfs merge=lfs -text
64
+ runs/big_run/exp_c/run_20260409_192244/training_plots.pdf filter=lfs diff=lfs merge=lfs -text
runs/big_run/exp_c/run_20260409_192244/checkpoints/best_loss_3p5130_epoch_0000_step_0003000.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:22ad7db83ccefbe429d12730963f88a60a4444b3cac843175f14fb39dc33e093
3
+ size 479869564
runs/big_run/exp_c/run_20260409_192244/checkpoints/best_loss_3p5190_epoch_0000_step_0002900.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:abd05531be8ee9393f350bc51fe13a2ee409f0a1d3b773cf06a85791eb4d31e9
3
+ size 479869564
runs/big_run/exp_c/run_20260409_192244/checkpoints/best_loss_3p5242_epoch_0000_step_0002800.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:cbccf35547887b713fe373b6a9d5a5ad6750a6a0dcbf1bc1b9b89f962e304f80
3
+ size 479869564
runs/big_run/exp_c/run_20260409_192244/checkpoints/best_loss_3p5320_epoch_0000_step_0002700.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f826a4e313cbe484bd0baca8117f6ed52cf0ea8eb0062ead85aeda9bc0912325
3
+ size 479869564
runs/big_run/exp_c/run_20260409_192244/checkpoints/best_loss_3p5420_epoch_0000_step_0002600.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:34968c6037acab036169f87f197a2b5d4b14511b48dd2197a7561176e6b8f82b
3
+ size 479869564
runs/big_run/exp_c/run_20260409_192244/checkpoints/best_loss_3p5520_epoch_0000_step_0002500.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d424d4b12fe85544d9d3f142976359cc1896dc2b737d4560ee5ea8c9e582ef97
3
+ size 479869564
runs/big_run/exp_c/run_20260409_192244/checkpoints/best_loss_3p5645_epoch_0000_step_0002400.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5793a6387c7c54acb180b64170e6f02e1fbf493e8b54ad428b07c58314d7e762
3
+ size 479869564
runs/big_run/exp_c/run_20260409_192244/checkpoints/best_loss_3p5771_epoch_0000_step_0002300.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c34f13b2dd46e63fb33bd061d340ed21ddb0fd42cf3289a449d8af54a5920754
3
+ size 479869564
runs/big_run/exp_c/run_20260409_192244/checkpoints/best_loss_3p5905_epoch_0000_step_0002200.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:41c46b29c7995b8ee02fe0d5a45cb7cb9aeb3734febc9180d60e9b504c3379f0
3
+ size 479869564
runs/big_run/exp_c/run_20260409_192244/checkpoints/best_loss_3p6072_epoch_0000_step_0002100.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b6cef302c8d1b1fd443d554b6db2340cd44ee69f91deaff4be600b12fb4eabf9
3
+ size 479869564
runs/big_run/exp_c/run_20260409_192244/checkpoints/last_epoch_0000_step_0001000.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:68f4ef930770e9c012b82b2a71644f733419cc3ce862e6f1c76bffd88d7e09e1
3
+ size 479860552
runs/big_run/exp_c/run_20260409_192244/checkpoints/last_epoch_0000_step_0001250.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:bafe1475d6892d9451e17b5425117f0080fbb53ffc57ec7b47561fcd83e8dca9
3
+ size 479860552
runs/big_run/exp_c/run_20260409_192244/checkpoints/last_epoch_0000_step_0001500.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:2af677e21ee4735d9cb40f4a79bf4b2550bd9eff5a1d7102912b035d9b362050
3
+ size 479860552
runs/big_run/exp_c/run_20260409_192244/checkpoints/last_epoch_0000_step_0001750.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:6e9863ea4e5623553e591a3a10f052a2228c95ec8f9df4aab85fbbd79e05e7ac
3
+ size 479860552
runs/big_run/exp_c/run_20260409_192244/checkpoints/last_epoch_0000_step_0002000.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:549f464b99a68c1026ee905567c883ee5899ed5be5f25765ed13c8470131868e
3
+ size 479860552
runs/big_run/exp_c/run_20260409_192244/checkpoints/last_epoch_0000_step_0002250.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:959e5ced86e7d8cbc9e0289a7532b5c0dccb19c2431676ba1897bed1845970ca
3
+ size 479860552
runs/big_run/exp_c/run_20260409_192244/checkpoints/last_epoch_0000_step_0002500.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:072f9f0f722c0760620d7fca39e07093a07c7fa044ea8f4b1e556cfa2ba5cbc1
3
+ size 479860552
runs/big_run/exp_c/run_20260409_192244/checkpoints/last_epoch_0000_step_0002750.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:0494be31b8e08fd5af9ff58df169f8d9e121eeba5e8bd74e64726db2260af2f9
3
+ size 479860552
runs/big_run/exp_c/run_20260409_192244/checkpoints/last_epoch_0000_step_0003000.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d9258ac3ddd2f7b71ec0c7713fc16cfa3292e6c6ef12a6a32344a9a76acfd298
3
+ size 479860552
runs/big_run/exp_c/run_20260409_192244/checkpoints/last_epoch_0000_step_0003052.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5cdbf4ad1d19b721b65af77bbc5744ec244528b56292f54e21bbcece82fe35f0
3
+ size 479860552
runs/big_run/exp_c/run_20260409_192244/config.json ADDED
@@ -0,0 +1,91 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "preprocess": null,
3
+ "streaming_preprocess": null,
4
+ "model": {
5
+ "vocab_size": 50258,
6
+ "pad_token_id": 50257,
7
+ "bos_token_id": 50256,
8
+ "eos_token_id": 50256,
9
+ "d_model": 384,
10
+ "n_layers": 14,
11
+ "n_heads": 6,
12
+ "d_ff": 768,
13
+ "context_length": 1024,
14
+ "bias": false,
15
+ "dropout": 0.01287921140461293,
16
+ "rope_theta": 10000.0,
17
+ "gradient_checkpointing": false
18
+ },
19
+ "training": {
20
+ "device": "auto",
21
+ "train_bin": "data/exp_c/train.bin",
22
+ "val_bin": "data/exp_c/val.bin",
23
+ "num_workers": 2,
24
+ "pin_memory": true,
25
+ "batch_size": 32,
26
+ "gradient_accumulation_steps": 1,
27
+ "train_samples_per_epoch": null,
28
+ "val_sequence_stride": null,
29
+ "learning_rate": 0.0026291804804649314,
30
+ "min_lr": 0.0002629180480464931,
31
+ "weight_decay": 0.13113982116099054,
32
+ "beta1": 0.9,
33
+ "beta2": 0.9606057712410854,
34
+ "grad_clip": 0.9147944011374611,
35
+ "warmup_steps": 400,
36
+ "max_steps": 3052,
37
+ "lr_schedule": "cosine",
38
+ "lr_decay_ratio": 0.25826046407623926,
39
+ "z_loss_coeff": 0.00013887413811122497,
40
+ "save_every": 250,
41
+ "eval_every": 100,
42
+ "checkpoint_dir": "checkpoints",
43
+ "keep_last_checkpoints": 10,
44
+ "keep_best_checkpoints": 10,
45
+ "early_stopping_patience": 0,
46
+ "early_stopping_min_delta": 0.001,
47
+ "runs_dir": "runs/big_run/exp_c",
48
+ "log_every": 10,
49
+ "hf_upload": {
50
+ "repo_id": "ParrotLabs/Preprocessed",
51
+ "repo_type": "dataset",
52
+ "path_in_repo": "",
53
+ "private": null
54
+ },
55
+ "compile": true
56
+ },
57
+ "tune": null,
58
+ "logging": {
59
+ "console_level": "INFO",
60
+ "file_level": "DEBUG",
61
+ "components": {
62
+ "data_preprocessing": "DEBUG",
63
+ "model_initialization": "INFO",
64
+ "training": "INFO"
65
+ },
66
+ "profiler": null
67
+ },
68
+ "eval": {
69
+ "device": "auto",
70
+ "batch_size": 32,
71
+ "max_sequences": 512,
72
+ "datasets": [
73
+ {
74
+ "name": "wikitext",
75
+ "path": "wikitext",
76
+ "subset": "wikitext-103-raw-v1",
77
+ "split": "test",
78
+ "stride": null
79
+ },
80
+ {
81
+ "name": "exp_c_val",
82
+ "path": "data/exp_c/val.bin",
83
+ "subset": null,
84
+ "split": null,
85
+ "stride": null
86
+ }
87
+ ]
88
+ },
89
+ "inference": null,
90
+ "chat": null
91
+ }
runs/big_run/exp_c/run_20260409_192244/metrics.jsonl ADDED
The diff for this file is too large to render. See raw diff
 
runs/big_run/exp_c/run_20260409_192244/train.log ADDED
@@ -0,0 +1,713 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ 2026-04-09 19:22:45 - INFO - parrotllm - Logging initialised -> runs/big_run/exp_c/run_20260409_192244/train.log
2
+ 2026-04-09 19:22:45 - INFO - parrotllm.training - device=cuda:0 | rank=0 | world_size=8 | distributed=yes
3
+ 2026-04-09 19:22:45 - INFO - parrotllm.training - checkpoint directory=/home/user/code/ParrotLLM/runs/big_run/exp_c/run_20260409_192244/checkpoints | keep_last=10 | keep_best=10
4
+ 2026-04-09 19:22:45 - INFO - parrotllm.training - train windows=random-overlapping | tokens=508793600 | valid_starts=508792576 | sampled_windows_per_epoch=508792576
5
+ 2026-04-09 19:22:45 - INFO - parrotllm.training - val windows=sequential | tokens=5138325 | stride=512 | windows=10034
6
+ 2026-04-09 19:22:46 - INFO - parrotllm.training - compiling model with torch.compile...
7
+ 2026-04-09 19:22:47 - INFO - parrotllm.training -
8
+ ============================================================
9
+ MODEL ARCHITECTURE SUMMARY
10
+ ============================================================
11
+
12
+ Configuration:
13
+ Vocab size: 50258
14
+ Block size (context): 1024
15
+ Layers: 14
16
+ Heads: 6
17
+ Embedding dim: 384
18
+ FFN hidden dim: 768
19
+ Dropout: 0.01287921140461293
20
+ Bias: False
21
+
22
+ Parameters (unique, weight-tied layers counted once):
23
+ Total: 39,966,592 (39.97M)
24
+ Trainable: 39,966,592 (39.97M)
25
+ Non-trainable: 0
26
+ Non-embedding: 39,966,592 (39.97M)
27
+ Position embeddings: 0 (RoPE: 0 learned params)
28
+ Size (MB): 159.87
29
+
30
+ Layer-wise breakdown (torchinfo):
31
+ Note: torchinfo double-counts weight-tied layers (tok_emb/lm_head).
32
+ ============================================================================================================================================
33
+ Layer (type:depth-idx) Input Shape Output Shape Param # Trainable
34
+ ============================================================================================================================================
35
+ ParrotLLM [32, 1024] [32, 1024, 50258] -- True
36
+ ├─Embedding: 1-1 [32, 1024] [32, 1024, 384] 19,299,072 True
37
+ ├─Dropout: 1-2 [32, 1024, 384] [32, 1024, 384] -- --
38
+ ├─ModuleList: 1-3 -- -- -- True
39
+ │ └─TransformerBlock: 2-1 [32, 1024, 384] [32, 1024, 384] -- True
40
+ │ │ └─RMSNorm: 3-1 [32, 1024, 384] [32, 1024, 384] 384 True
41
+ │ │ └─MultiHeadAttention: 3-2 [32, 1024, 384] [32, 1024, 384] 589,952 True
42
+ │ │ └─RMSNorm: 3-3 [32, 1024, 384] [32, 1024, 384] 384 True
43
+ │ │ └─RMSNorm: 3-4 [32, 1024, 384] [32, 1024, 384] 384 True
44
+ │ │ └─SwiGLUMLP: 3-5 [32, 1024, 384] [32, 1024, 384] 884,736 True
45
+ │ │ └─RMSNorm: 3-6 [32, 1024, 384] [32, 1024, 384] 384 True
46
+ │ └─TransformerBlock: 2-2 [32, 1024, 384] [32, 1024, 384] -- True
47
+ │ │ └─RMSNorm: 3-7 [32, 1024, 384] [32, 1024, 384] 384 True
48
+ │ │ └─MultiHeadAttention: 3-8 [32, 1024, 384] [32, 1024, 384] 589,952 True
49
+ │ │ └─RMSNorm: 3-9 [32, 1024, 384] [32, 1024, 384] 384 True
50
+ │ │ └─RMSNorm: 3-10 [32, 1024, 384] [32, 1024, 384] 384 True
51
+ │ │ └─SwiGLUMLP: 3-11 [32, 1024, 384] [32, 1024, 384] 884,736 True
52
+ │ │ └─RMSNorm: 3-12 [32, 1024, 384] [32, 1024, 384] 384 True
53
+ │ └─TransformerBlock: 2-3 [32, 1024, 384] [32, 1024, 384] -- True
54
+ │ │ └─RMSNorm: 3-13 [32, 1024, 384] [32, 1024, 384] 384 True
55
+ │ │ └─MultiHeadAttention: 3-14 [32, 1024, 384] [32, 1024, 384] 589,952 True
56
+ │ │ └─RMSNorm: 3-15 [32, 1024, 384] [32, 1024, 384] 384 True
57
+ │ │ └─RMSNorm: 3-16 [32, 1024, 384] [32, 1024, 384] 384 True
58
+ │ │ └─SwiGLUMLP: 3-17 [32, 1024, 384] [32, 1024, 384] 884,736 True
59
+ │ │ └─RMSNorm: 3-18 [32, 1024, 384] [32, 1024, 384] 384 True
60
+ │ └─TransformerBlock: 2-4 [32, 1024, 384] [32, 1024, 384] -- True
61
+ │ │ └─RMSNorm: 3-19 [32, 1024, 384] [32, 1024, 384] 384 True
62
+ │ │ └─MultiHeadAttention: 3-20 [32, 1024, 384] [32, 1024, 384] 589,952 True
63
+ │ │ └─RMSNorm: 3-21 [32, 1024, 384] [32, 1024, 384] 384 True
64
+ │ │ └─RMSNorm: 3-22 [32, 1024, 384] [32, 1024, 384] 384 True
65
+ │ │ └─SwiGLUMLP: 3-23 [32, 1024, 384] [32, 1024, 384] 884,736 True
66
+ │ │ └─RMSNorm: 3-24 [32, 1024, 384] [32, 1024, 384] 384 True
67
+ │ └─TransformerBlock: 2-5 [32, 1024, 384] [32, 1024, 384] -- True
68
+ │ │ └─RMSNorm: 3-25 [32, 1024, 384] [32, 1024, 384] 384 True
69
+ │ │ └─MultiHeadAttention: 3-26 [32, 1024, 384] [32, 1024, 384] 589,952 True
70
+ │ │ └─RMSNorm: 3-27 [32, 1024, 384] [32, 1024, 384] 384 True
71
+ │ │ └─RMSNorm: 3-28 [32, 1024, 384] [32, 1024, 384] 384 True
72
+ │ │ └─SwiGLUMLP: 3-29 [32, 1024, 384] [32, 1024, 384] 884,736 True
73
+ │ │ └─RMSNorm: 3-30 [32, 1024, 384] [32, 1024, 384] 384 True
74
+ │ └─TransformerBlock: 2-6 [32, 1024, 384] [32, 1024, 384] -- True
75
+ │ │ └─RMSNorm: 3-31 [32, 1024, 384] [32, 1024, 384] 384 True
76
+ │ │ └─MultiHeadAttention: 3-32 [32, 1024, 384] [32, 1024, 384] 589,952 True
77
+ │ │ └─RMSNorm: 3-33 [32, 1024, 384] [32, 1024, 384] 384 True
78
+ │ │ └─RMSNorm: 3-34 [32, 1024, 384] [32, 1024, 384] 384 True
79
+ │ │ └─SwiGLUMLP: 3-35 [32, 1024, 384] [32, 1024, 384] 884,736 True
80
+ │ │ └─RMSNorm: 3-36 [32, 1024, 384] [32, 1024, 384] 384 True
81
+ │ └─TransformerBlock: 2-7 [32, 1024, 384] [32, 1024, 384] -- True
82
+ │ │ └─RMSNorm: 3-37 [32, 1024, 384] [32, 1024, 384] 384 True
83
+ │ │ └─MultiHeadAttention: 3-38 [32, 1024, 384] [32, 1024, 384] 589,952 True
84
+ │ │ └─RMSNorm: 3-39 [32, 1024, 384] [32, 1024, 384] 384 True
85
+ │ │ └─RMSNorm: 3-40 [32, 1024, 384] [32, 1024, 384] 384 True
86
+ │ │ └─SwiGLUMLP: 3-41 [32, 1024, 384] [32, 1024, 384] 884,736 True
87
+ │ │ └─RMSNorm: 3-42 [32, 1024, 384] [32, 1024, 384] 384 True
88
+ │ └─TransformerBlock: 2-8 [32, 1024, 384] [32, 1024, 384] -- True
89
+ │ │ └─RMSNorm: 3-43 [32, 1024, 384] [32, 1024, 384] 384 True
90
+ │ │ └─MultiHeadAttention: 3-44 [32, 1024, 384] [32, 1024, 384] 589,952 True
91
+ │ │ └─RMSNorm: 3-45 [32, 1024, 384] [32, 1024, 384] 384 True
92
+ │ │ └─RMSNorm: 3-46 [32, 1024, 384] [32, 1024, 384] 384 True
93
+ │ │ └─SwiGLUMLP: 3-47 [32, 1024, 384] [32, 1024, 384] 884,736 True
94
+ │ │ └─RMSNorm: 3-48 [32, 1024, 384] [32, 1024, 384] 384 True
95
+ │ └─TransformerBlock: 2-9 [32, 1024, 384] [32, 1024, 384] -- True
96
+ │ │ └─RMSNorm: 3-49 [32, 1024, 384] [32, 1024, 384] 384 True
97
+ │ │ └─MultiHeadAttention: 3-50 [32, 1024, 384] [32, 1024, 384] 589,952 True
98
+ │ │ └─RMSNorm: 3-51 [32, 1024, 384] [32, 1024, 384] 384 True
99
+ │ │ └─RMSNorm: 3-52 [32, 1024, 384] [32, 1024, 384] 384 True
100
+ │ │ └─SwiGLUMLP: 3-53 [32, 1024, 384] [32, 1024, 384] 884,736 True
101
+ │ │ └─RMSNorm: 3-54 [32, 1024, 384] [32, 1024, 384] 384 True
102
+ │ └─TransformerBlock: 2-10 [32, 1024, 384] [32, 1024, 384] -- True
103
+ │ │ └─RMSNorm: 3-55 [32, 1024, 384] [32, 1024, 384] 384 True
104
+ │ │ └─MultiHeadAttention: 3-56 [32, 1024, 384] [32, 1024, 384] 589,952 True
105
+ │ │ └─RMSNorm: 3-57 [32, 1024, 384] [32, 1024, 384] 384 True
106
+ │ │ └─RMSNorm: 3-58 [32, 1024, 384] [32, 1024, 384] 384 True
107
+ │ │ └─SwiGLUMLP: 3-59 [32, 1024, 384] [32, 1024, 384] 884,736 True
108
+ │ │ └─RMSNorm: 3-60 [32, 1024, 384] [32, 1024, 384] 384 True
109
+ │ └─TransformerBlock: 2-11 [32, 1024, 384] [32, 1024, 384] -- True
110
+ │ │ └─RMSNorm: 3-61 [32, 1024, 384] [32, 1024, 384] 384 True
111
+ │ │ └─MultiHeadAttention: 3-62 [32, 1024, 384] [32, 1024, 384] 589,952 True
112
+ │ │ └─RMSNorm: 3-63 [32, 1024, 384] [32, 1024, 384] 384 True
113
+ │ │ └─RMSNorm: 3-64 [32, 1024, 384] [32, 1024, 384] 384 True
114
+ │ │ └─SwiGLUMLP: 3-65 [32, 1024, 384] [32, 1024, 384] 884,736 True
115
+ │ │ └─RMSNorm: 3-66 [32, 1024, 384] [32, 1024, 384] 384 True
116
+ │ └─TransformerBlock: 2-12 [32, 1024, 384] [32, 1024, 384] -- True
117
+ │ │ └─RMSNorm: 3-67 [32, 1024, 384] [32, 1024, 384] 384 True
118
+ │ │ └─MultiHeadAttention: 3-68 [32, 1024, 384] [32, 1024, 384] 589,952 True
119
+ │ │ └─RMSNorm: 3-69 [32, 1024, 384] [32, 1024, 384] 384 True
120
+ │ │ └─RMSNorm: 3-70 [32, 1024, 384] [32, 1024, 384] 384 True
121
+ │ │ └─SwiGLUMLP: 3-71 [32, 1024, 384] [32, 1024, 384] 884,736 True
122
+ │ │ └─RMSNorm: 3-72 [32, 1024, 384] [32, 1024, 384] 384 True
123
+ │ └─TransformerBlock: 2-13 [32, 1024, 384] [32, 1024, 384] -- True
124
+ │ │ └─RMSNorm: 3-73 [32, 1024, 384] [32, 1024, 384] 384 True
125
+ │ │ └─MultiHeadAttention: 3-74 [32, 1024, 384] [32, 1024, 384] 589,952 True
126
+ │ │ └─RMSNorm: 3-75 [32, 1024, 384] [32, 1024, 384] 384 True
127
+ │ │ └─RMSNorm: 3-76 [32, 1024, 384] [32, 1024, 384] 384 True
128
+ │ │ └─SwiGLUMLP: 3-77 [32, 1024, 384] [32, 1024, 384] 884,736 True
129
+ │ │ └─RMSNorm: 3-78 [32, 1024, 384] [32, 1024, 384] 384 True
130
+ │ └─TransformerBlock: 2-14 [32, 1024, 384] [32, 1024, 384] -- True
131
+ │ │ └─RMSNorm: 3-79 [32, 1024, 384] [32, 1024, 384] 384 True
132
+ │ │ └─MultiHeadAttention: 3-80 [32, 1024, 384] [32, 1024, 384] 589,952 True
133
+ │ │ └─RMSNorm: 3-81 [32, 1024, 384] [32, 1024, 384] 384 True
134
+ │ │ └─RMSNorm: 3-82 [32, 1024, 384] [32, 1024, 384] 384 True
135
+ │ │ └─SwiGLUMLP: 3-83 [32, 1024, 384] [32, 1024, 384] 884,736 True
136
+ │ │ └��RMSNorm: 3-84 [32, 1024, 384] [32, 1024, 384] 384 True
137
+ ├─RMSNorm: 1-4 [32, 1024, 384] [32, 1024, 384] 384 True
138
+ ├─Linear: 1-5 [32, 1024, 384] [32, 1024, 50258] 19,299,072 True
139
+ ============================================================================================================================================
140
+ Total params: 59,265,664
141
+ Trainable params: 59,265,664
142
+ Non-trainable params: 0
143
+ Total mult-adds (Units.GIGABYTES): 1.90
144
+ ============================================================================================================================================
145
+ Input size (MB): 0.26
146
+ Forward/backward pass size (MB): 34515.45
147
+ Params size (MB): 237.06
148
+ Estimated Total Size (MB): 34752.78
149
+ ============================================================================================================================================
150
+ ============================================================
151
+ 2026-04-09 19:22:47 - INFO - parrotllm.training - Starting evaluation...
152
+ 2026-04-09 19:22:54 - INFO - parrotllm.training - Initial val: loss=10.8930, ppl=53798.44
153
+ 2026-04-09 19:23:01 - INFO - parrotllm.training -
154
+ ============================================================
155
+ Starting training...
156
+ Steps per epoch (approx): 1987471
157
+ Max steps: 3052
158
+ ============================================================
159
+ 2026-04-09 19:23:01 - INFO - parrotllm.training - LR schedule=cosine | decay_ratio=0.25826046407623926 | z_loss_coeff=0.00013887413811122497
160
+ 2026-04-09 19:23:42 - INFO - parrotllm.training - step 1 | epoch 0 | loss 10.9091 | lr 6.57e-06 | grad 7.1365
161
+ 2026-04-09 19:23:45 - INFO - parrotllm.training - step 10 | epoch 0 | loss 10.1779 | lr 6.57e-05 | grad 1.7449
162
+ 2026-04-09 19:23:48 - INFO - parrotllm.training - step 20 | epoch 0 | loss 9.6632 | lr 1.31e-04 | grad 1.7042
163
+ 2026-04-09 19:23:51 - INFO - parrotllm.training - step 30 | epoch 0 | loss 8.9701 | lr 1.97e-04 | grad 1.6798
164
+ 2026-04-09 19:23:54 - INFO - parrotllm.training - step 40 | epoch 0 | loss 8.1648 | lr 2.63e-04 | grad 1.4100
165
+ 2026-04-09 19:23:57 - INFO - parrotllm.training - step 50 | epoch 0 | loss 7.5419 | lr 3.29e-04 | grad 0.9384
166
+ 2026-04-09 19:24:00 - INFO - parrotllm.training - step 60 | epoch 0 | loss 7.2730 | lr 3.94e-04 | grad 0.7149
167
+ 2026-04-09 19:24:03 - INFO - parrotllm.training - step 70 | epoch 0 | loss 7.0623 | lr 4.60e-04 | grad 0.4622
168
+ 2026-04-09 19:24:06 - INFO - parrotllm.training - step 80 | epoch 0 | loss 6.7960 | lr 5.26e-04 | grad 1.0549
169
+ 2026-04-09 19:24:09 - INFO - parrotllm.training - step 90 | epoch 0 | loss 6.6086 | lr 5.92e-04 | grad 1.0717
170
+ 2026-04-09 19:24:12 - INFO - parrotllm.training - step 100 | epoch 0 | loss 6.4815 | lr 6.57e-04 | grad 1.0519
171
+ 2026-04-09 19:24:12 - INFO - parrotllm.training - Starting evaluation...
172
+ 2026-04-09 19:24:12 - INFO - parrotllm.training - ------------------------------------------------------------
173
+ 2026-04-09 19:24:16 - INFO - parrotllm.training - Train: loss=6.4815, ppl=652.94
174
+ 2026-04-09 19:24:16 - INFO - parrotllm.training - Val: loss=6.5388, ppl=691.42
175
+ 2026-04-09 19:24:16 - INFO - parrotllm.training - ** New best validation loss! **
176
+ 2026-04-09 19:24:17 - INFO - parrotllm.training - Saved best checkpoint: /home/user/code/ParrotLLM/runs/big_run/exp_c/run_20260409_192244/checkpoints/best_loss_6p5388_epoch_0000_step_0000100.pt
177
+ 2026-04-09 19:24:18 - INFO - parrotllm.training - ------------------------------------------------------------
178
+ 2026-04-09 19:24:21 - INFO - parrotllm.training - step 110 | epoch 0 | loss 6.4356 | lr 7.23e-04 | grad 0.7613
179
+ 2026-04-09 19:24:24 - INFO - parrotllm.training - step 120 | epoch 0 | loss 6.3023 | lr 7.89e-04 | grad 0.6637
180
+ 2026-04-09 19:24:27 - INFO - parrotllm.training - step 130 | epoch 0 | loss 6.1590 | lr 8.54e-04 | grad 0.7438
181
+ 2026-04-09 19:24:30 - INFO - parrotllm.training - step 140 | epoch 0 | loss 6.1493 | lr 9.20e-04 | grad 0.6014
182
+ 2026-04-09 19:24:33 - INFO - parrotllm.training - step 150 | epoch 0 | loss 6.1315 | lr 9.86e-04 | grad 0.7622
183
+ 2026-04-09 19:24:36 - INFO - parrotllm.training - step 160 | epoch 0 | loss 5.8643 | lr 1.05e-03 | grad 0.7324
184
+ 2026-04-09 19:24:39 - INFO - parrotllm.training - step 170 | epoch 0 | loss 5.9140 | lr 1.12e-03 | grad 0.4627
185
+ 2026-04-09 19:24:42 - INFO - parrotllm.training - step 180 | epoch 0 | loss 5.7912 | lr 1.18e-03 | grad 0.8399
186
+ 2026-04-09 19:24:45 - INFO - parrotllm.training - step 190 | epoch 0 | loss 5.7677 | lr 1.25e-03 | grad 0.4284
187
+ 2026-04-09 19:24:48 - INFO - parrotllm.training - step 200 | epoch 0 | loss 5.7406 | lr 1.31e-03 | grad 0.6489
188
+ 2026-04-09 19:24:48 - INFO - parrotllm.training - Starting evaluation...
189
+ 2026-04-09 19:24:48 - INFO - parrotllm.training - ------------------------------------------------------------
190
+ 2026-04-09 19:24:53 - INFO - parrotllm.training - Train: loss=5.7406, ppl=311.24
191
+ 2026-04-09 19:24:53 - INFO - parrotllm.training - Val: loss=5.7229, ppl=305.81
192
+ 2026-04-09 19:24:53 - INFO - parrotllm.training - ** New best validation loss! **
193
+ 2026-04-09 19:24:54 - INFO - parrotllm.training - Saved best checkpoint: /home/user/code/ParrotLLM/runs/big_run/exp_c/run_20260409_192244/checkpoints/best_loss_5p7229_epoch_0000_step_0000200.pt
194
+ 2026-04-09 19:24:54 - INFO - parrotllm.training - ------------------------------------------------------------
195
+ 2026-04-09 19:24:57 - INFO - parrotllm.training - step 210 | epoch 0 | loss 5.6598 | lr 1.38e-03 | grad 0.4494
196
+ 2026-04-09 19:25:00 - INFO - parrotllm.training - step 220 | epoch 0 | loss 5.6268 | lr 1.45e-03 | grad 0.7713
197
+ 2026-04-09 19:25:03 - INFO - parrotllm.training - step 230 | epoch 0 | loss 5.5324 | lr 1.51e-03 | grad 0.4183
198
+ 2026-04-09 19:25:06 - INFO - parrotllm.training - step 240 | epoch 0 | loss 5.5348 | lr 1.58e-03 | grad 0.4157
199
+ 2026-04-09 19:25:09 - INFO - parrotllm.training - step 250 | epoch 0 | loss 5.4719 | lr 1.64e-03 | grad 0.4365
200
+ 2026-04-09 19:25:10 - INFO - parrotllm.training - Saved checkpoint: /home/user/code/ParrotLLM/runs/big_run/exp_c/run_20260409_192244/checkpoints/last_epoch_0000_step_0000250.pt
201
+ 2026-04-09 19:25:14 - INFO - parrotllm.training - step 260 | epoch 0 | loss 5.3748 | lr 1.71e-03 | grad 0.4167
202
+ 2026-04-09 19:25:17 - INFO - parrotllm.training - step 270 | epoch 0 | loss 5.3018 | lr 1.77e-03 | grad 0.4486
203
+ 2026-04-09 19:25:20 - INFO - parrotllm.training - step 280 | epoch 0 | loss 5.3781 | lr 1.84e-03 | grad 0.6678
204
+ 2026-04-09 19:25:23 - INFO - parrotllm.training - step 290 | epoch 0 | loss 5.2077 | lr 1.91e-03 | grad 0.3669
205
+ 2026-04-09 19:25:26 - INFO - parrotllm.training - step 300 | epoch 0 | loss 5.1543 | lr 1.97e-03 | grad 0.2915
206
+ 2026-04-09 19:25:26 - INFO - parrotllm.training - Starting evaluation...
207
+ 2026-04-09 19:25:26 - INFO - parrotllm.training - ------------------------------------------------------------
208
+ 2026-04-09 19:25:31 - INFO - parrotllm.training - Train: loss=5.1543, ppl=173.18
209
+ 2026-04-09 19:25:31 - INFO - parrotllm.training - Val: loss=5.1434, ppl=171.30
210
+ 2026-04-09 19:25:31 - INFO - parrotllm.training - ** New best validation loss! **
211
+ 2026-04-09 19:25:31 - INFO - parrotllm.training - Saved best checkpoint: /home/user/code/ParrotLLM/runs/big_run/exp_c/run_20260409_192244/checkpoints/best_loss_5p1434_epoch_0000_step_0000300.pt
212
+ 2026-04-09 19:25:32 - INFO - parrotllm.training - ------------------------------------------------------------
213
+ 2026-04-09 19:25:35 - INFO - parrotllm.training - step 310 | epoch 0 | loss 5.1923 | lr 2.04e-03 | grad 0.6414
214
+ 2026-04-09 19:25:38 - INFO - parrotllm.training - step 320 | epoch 0 | loss 5.2201 | lr 2.10e-03 | grad 0.3256
215
+ 2026-04-09 19:25:41 - INFO - parrotllm.training - step 330 | epoch 0 | loss 5.0706 | lr 2.17e-03 | grad 0.2946
216
+ 2026-04-09 19:25:44 - INFO - parrotllm.training - step 340 | epoch 0 | loss 4.9316 | lr 2.23e-03 | grad 0.3419
217
+ 2026-04-09 19:25:47 - INFO - parrotllm.training - step 350 | epoch 0 | loss 4.9997 | lr 2.30e-03 | grad 0.3016
218
+ 2026-04-09 19:25:50 - INFO - parrotllm.training - step 360 | epoch 0 | loss 5.0225 | lr 2.37e-03 | grad 0.3246
219
+ 2026-04-09 19:25:53 - INFO - parrotllm.training - step 370 | epoch 0 | loss 5.0837 | lr 2.43e-03 | grad 0.4280
220
+ 2026-04-09 19:25:56 - INFO - parrotllm.training - step 380 | epoch 0 | loss 4.8680 | lr 2.50e-03 | grad 0.2313
221
+ 2026-04-09 19:25:59 - INFO - parrotllm.training - step 390 | epoch 0 | loss 4.8151 | lr 2.56e-03 | grad 0.3728
222
+ 2026-04-09 19:26:02 - INFO - parrotllm.training - step 400 | epoch 0 | loss 4.8596 | lr 2.63e-03 | grad 0.2835
223
+ 2026-04-09 19:26:02 - INFO - parrotllm.training - Starting evaluation...
224
+ 2026-04-09 19:26:02 - INFO - parrotllm.training - ------------------------------------------------------------
225
+ 2026-04-09 19:26:07 - INFO - parrotllm.training - Train: loss=4.8596, ppl=128.98
226
+ 2026-04-09 19:26:07 - INFO - parrotllm.training - Val: loss=4.7841, ppl=119.59
227
+ 2026-04-09 19:26:07 - INFO - parrotllm.training - ** New best validation loss! **
228
+ 2026-04-09 19:26:08 - INFO - parrotllm.training - Saved best checkpoint: /home/user/code/ParrotLLM/runs/big_run/exp_c/run_20260409_192244/checkpoints/best_loss_4p7841_epoch_0000_step_0000400.pt
229
+ 2026-04-09 19:26:08 - INFO - parrotllm.training - ------------------------------------------------------------
230
+ 2026-04-09 19:26:12 - INFO - parrotllm.training - step 410 | epoch 0 | loss 4.8199 | lr 2.63e-03 | grad 0.4502
231
+ 2026-04-09 19:26:15 - INFO - parrotllm.training - step 420 | epoch 0 | loss 4.8968 | lr 2.63e-03 | grad 0.2516
232
+ 2026-04-09 19:26:18 - INFO - parrotllm.training - step 430 | epoch 0 | loss 4.6891 | lr 2.63e-03 | grad 0.3294
233
+ 2026-04-09 19:26:21 - INFO - parrotllm.training - step 440 | epoch 0 | loss 4.7090 | lr 2.63e-03 | grad 0.2523
234
+ 2026-04-09 19:26:24 - INFO - parrotllm.training - step 450 | epoch 0 | loss 4.6402 | lr 2.63e-03 | grad 0.3459
235
+ 2026-04-09 19:26:27 - INFO - parrotllm.training - step 460 | epoch 0 | loss 4.6079 | lr 2.63e-03 | grad 0.2878
236
+ 2026-04-09 19:26:30 - INFO - parrotllm.training - step 470 | epoch 0 | loss 4.5183 | lr 2.63e-03 | grad 0.2449
237
+ 2026-04-09 19:26:33 - INFO - parrotllm.training - step 480 | epoch 0 | loss 4.5264 | lr 2.62e-03 | grad 0.2478
238
+ 2026-04-09 19:26:36 - INFO - parrotllm.training - step 490 | epoch 0 | loss 4.5953 | lr 2.62e-03 | grad 0.1814
239
+ 2026-04-09 19:26:39 - INFO - parrotllm.training - step 500 | epoch 0 | loss 4.4539 | lr 2.62e-03 | grad 0.2546
240
+ 2026-04-09 19:26:39 - INFO - parrotllm.training - Starting evaluation...
241
+ 2026-04-09 19:26:39 - INFO - parrotllm.training - ------------------------------------------------------------
242
+ 2026-04-09 19:26:43 - INFO - parrotllm.training - Train: loss=4.4539, ppl=85.96
243
+ 2026-04-09 19:26:43 - INFO - parrotllm.training - Val: loss=4.4201, ppl=83.10
244
+ 2026-04-09 19:26:43 - INFO - parrotllm.training - ** New best validation loss! **
245
+ 2026-04-09 19:26:44 - INFO - parrotllm.training - Saved best checkpoint: /home/user/code/ParrotLLM/runs/big_run/exp_c/run_20260409_192244/checkpoints/best_loss_4p4201_epoch_0000_step_0000500.pt
246
+ 2026-04-09 19:26:45 - INFO - parrotllm.training - ------------------------------------------------------------
247
+ 2026-04-09 19:26:46 - INFO - parrotllm.training - Saved checkpoint: /home/user/code/ParrotLLM/runs/big_run/exp_c/run_20260409_192244/checkpoints/last_epoch_0000_step_0000500.pt
248
+ 2026-04-09 19:26:50 - INFO - parrotllm.training - step 510 | epoch 0 | loss 4.4126 | lr 2.62e-03 | grad 0.2475
249
+ 2026-04-09 19:26:53 - INFO - parrotllm.training - step 520 | epoch 0 | loss 4.5607 | lr 2.62e-03 | grad 0.2804
250
+ 2026-04-09 19:26:56 - INFO - parrotllm.training - step 530 | epoch 0 | loss 4.4480 | lr 2.62e-03 | grad 0.2286
251
+ 2026-04-09 19:26:59 - INFO - parrotllm.training - step 540 | epoch 0 | loss 4.4839 | lr 2.61e-03 | grad 0.2396
252
+ 2026-04-09 19:27:02 - INFO - parrotllm.training - step 550 | epoch 0 | loss 4.3834 | lr 2.61e-03 | grad 0.2426
253
+ 2026-04-09 19:27:05 - INFO - parrotllm.training - step 560 | epoch 0 | loss 4.3143 | lr 2.61e-03 | grad 0.2707
254
+ 2026-04-09 19:27:08 - INFO - parrotllm.training - step 570 | epoch 0 | loss 4.3453 | lr 2.61e-03 | grad 0.2171
255
+ 2026-04-09 19:27:11 - INFO - parrotllm.training - step 580 | epoch 0 | loss 4.1980 | lr 2.60e-03 | grad 0.2505
256
+ 2026-04-09 19:27:14 - INFO - parrotllm.training - step 590 | epoch 0 | loss 4.2947 | lr 2.60e-03 | grad 0.2174
257
+ 2026-04-09 19:27:17 - INFO - parrotllm.training - step 600 | epoch 0 | loss 4.3160 | lr 2.60e-03 | grad 0.2016
258
+ 2026-04-09 19:27:17 - INFO - parrotllm.training - Starting evaluation...
259
+ 2026-04-09 19:27:17 - INFO - parrotllm.training - ------------------------------------------------------------
260
+ 2026-04-09 19:27:22 - INFO - parrotllm.training - Train: loss=4.3160, ppl=74.89
261
+ 2026-04-09 19:27:22 - INFO - parrotllm.training - Val: loss=4.2285, ppl=68.61
262
+ 2026-04-09 19:27:22 - INFO - parrotllm.training - ** New best validation loss! **
263
+ 2026-04-09 19:27:22 - INFO - parrotllm.training - Saved best checkpoint: /home/user/code/ParrotLLM/runs/big_run/exp_c/run_20260409_192244/checkpoints/best_loss_4p2285_epoch_0000_step_0000600.pt
264
+ 2026-04-09 19:27:23 - INFO - parrotllm.training - ------------------------------------------------------------
265
+ 2026-04-09 19:27:26 - INFO - parrotllm.training - step 610 | epoch 0 | loss 4.2845 | lr 2.59e-03 | grad 0.2130
266
+ 2026-04-09 19:27:29 - INFO - parrotllm.training - step 620 | epoch 0 | loss 4.3043 | lr 2.59e-03 | grad 0.2123
267
+ 2026-04-09 19:27:32 - INFO - parrotllm.training - step 630 | epoch 0 | loss 4.2337 | lr 2.59e-03 | grad 0.2116
268
+ 2026-04-09 19:27:35 - INFO - parrotllm.training - step 640 | epoch 0 | loss 4.2180 | lr 2.58e-03 | grad 0.1879
269
+ 2026-04-09 19:27:38 - INFO - parrotllm.training - step 650 | epoch 0 | loss 4.2806 | lr 2.58e-03 | grad 0.2006
270
+ 2026-04-09 19:27:41 - INFO - parrotllm.training - step 660 | epoch 0 | loss 4.1075 | lr 2.57e-03 | grad 0.1932
271
+ 2026-04-09 19:27:45 - INFO - parrotllm.training - step 670 | epoch 0 | loss 4.2384 | lr 2.57e-03 | grad 0.2334
272
+ 2026-04-09 19:27:48 - INFO - parrotllm.training - step 680 | epoch 0 | loss 4.2468 | lr 2.56e-03 | grad 0.1772
273
+ 2026-04-09 19:27:51 - INFO - parrotllm.training - step 690 | epoch 0 | loss 4.0709 | lr 2.56e-03 | grad 0.2007
274
+ 2026-04-09 19:27:54 - INFO - parrotllm.training - step 700 | epoch 0 | loss 4.1832 | lr 2.56e-03 | grad 0.2443
275
+ 2026-04-09 19:27:54 - INFO - parrotllm.training - Starting evaluation...
276
+ 2026-04-09 19:27:54 - INFO - parrotllm.training - ------------------------------------------------------------
277
+ 2026-04-09 19:27:58 - INFO - parrotllm.training - Train: loss=4.1832, ppl=65.57
278
+ 2026-04-09 19:27:58 - INFO - parrotllm.training - Val: loss=4.1122, ppl=61.08
279
+ 2026-04-09 19:27:58 - INFO - parrotllm.training - ** New best validation loss! **
280
+ 2026-04-09 19:27:59 - INFO - parrotllm.training - Saved best checkpoint: /home/user/code/ParrotLLM/runs/big_run/exp_c/run_20260409_192244/checkpoints/best_loss_4p1122_epoch_0000_step_0000700.pt
281
+ 2026-04-09 19:27:59 - INFO - parrotllm.training - ------------------------------------------------------------
282
+ 2026-04-09 19:28:03 - INFO - parrotllm.training - step 710 | epoch 0 | loss 4.1110 | lr 2.55e-03 | grad 0.1995
283
+ 2026-04-09 19:28:06 - INFO - parrotllm.training - step 720 | epoch 0 | loss 4.1813 | lr 2.55e-03 | grad 0.2107
284
+ 2026-04-09 19:28:09 - INFO - parrotllm.training - step 730 | epoch 0 | loss 4.0497 | lr 2.54e-03 | grad 0.1844
285
+ 2026-04-09 19:28:12 - INFO - parrotllm.training - step 740 | epoch 0 | loss 4.0766 | lr 2.53e-03 | grad 0.2249
286
+ 2026-04-09 19:28:15 - INFO - parrotllm.training - step 750 | epoch 0 | loss 4.0992 | lr 2.53e-03 | grad 0.2323
287
+ 2026-04-09 19:28:16 - INFO - parrotllm.training - Saved checkpoint: /home/user/code/ParrotLLM/runs/big_run/exp_c/run_20260409_192244/checkpoints/last_epoch_0000_step_0000750.pt
288
+ 2026-04-09 19:28:20 - INFO - parrotllm.training - step 760 | epoch 0 | loss 4.0914 | lr 2.52e-03 | grad 0.1809
289
+ 2026-04-09 19:28:23 - INFO - parrotllm.training - step 770 | epoch 0 | loss 4.0153 | lr 2.52e-03 | grad 0.2023
290
+ 2026-04-09 19:28:26 - INFO - parrotllm.training - step 780 | epoch 0 | loss 4.1063 | lr 2.51e-03 | grad 0.2175
291
+ 2026-04-09 19:28:29 - INFO - parrotllm.training - step 790 | epoch 0 | loss 4.1249 | lr 2.51e-03 | grad 0.1779
292
+ 2026-04-09 19:28:32 - INFO - parrotllm.training - step 800 | epoch 0 | loss 4.1547 | lr 2.50e-03 | grad 0.1781
293
+ 2026-04-09 19:28:32 - INFO - parrotllm.training - Starting evaluation...
294
+ 2026-04-09 19:28:32 - INFO - parrotllm.training - ------------------------------------------------------------
295
+ 2026-04-09 19:28:36 - INFO - parrotllm.training - Train: loss=4.1547, ppl=63.74
296
+ 2026-04-09 19:28:36 - INFO - parrotllm.training - Val: loss=4.0191, ppl=55.65
297
+ 2026-04-09 19:28:36 - INFO - parrotllm.training - ** New best validation loss! **
298
+ 2026-04-09 19:28:37 - INFO - parrotllm.training - Saved best checkpoint: /home/user/code/ParrotLLM/runs/big_run/exp_c/run_20260409_192244/checkpoints/best_loss_4p0191_epoch_0000_step_0000800.pt
299
+ 2026-04-09 19:28:38 - INFO - parrotllm.training - ------------------------------------------------------------
300
+ 2026-04-09 19:28:41 - INFO - parrotllm.training - step 810 | epoch 0 | loss 3.9622 | lr 2.49e-03 | grad 0.2000
301
+ 2026-04-09 19:28:44 - INFO - parrotllm.training - step 820 | epoch 0 | loss 4.0192 | lr 2.49e-03 | grad 0.1981
302
+ 2026-04-09 19:28:47 - INFO - parrotllm.training - step 830 | epoch 0 | loss 4.0988 | lr 2.48e-03 | grad 0.1624
303
+ 2026-04-09 19:28:50 - INFO - parrotllm.training - step 840 | epoch 0 | loss 4.0545 | lr 2.47e-03 | grad 0.1760
304
+ 2026-04-09 19:28:53 - INFO - parrotllm.training - step 850 | epoch 0 | loss 3.9883 | lr 2.47e-03 | grad 0.1823
305
+ 2026-04-09 19:28:56 - INFO - parrotllm.training - step 860 | epoch 0 | loss 3.9583 | lr 2.46e-03 | grad 0.1637
306
+ 2026-04-09 19:28:59 - INFO - parrotllm.training - step 870 | epoch 0 | loss 3.9888 | lr 2.45e-03 | grad 0.1650
307
+ 2026-04-09 19:29:02 - INFO - parrotllm.training - step 880 | epoch 0 | loss 4.0734 | lr 2.44e-03 | grad 0.1684
308
+ 2026-04-09 19:29:05 - INFO - parrotllm.training - step 890 | epoch 0 | loss 4.0500 | lr 2.44e-03 | grad 0.1879
309
+ 2026-04-09 19:29:08 - INFO - parrotllm.training - step 900 | epoch 0 | loss 3.9939 | lr 2.43e-03 | grad 0.1874
310
+ 2026-04-09 19:29:08 - INFO - parrotllm.training - Starting evaluation...
311
+ 2026-04-09 19:29:08 - INFO - parrotllm.training - ------------------------------------------------------------
312
+ 2026-04-09 19:29:12 - INFO - parrotllm.training - Train: loss=3.9939, ppl=54.27
313
+ 2026-04-09 19:29:12 - INFO - parrotllm.training - Val: loss=3.9550, ppl=52.20
314
+ 2026-04-09 19:29:12 - INFO - parrotllm.training - ** New best validation loss! **
315
+ 2026-04-09 19:29:13 - INFO - parrotllm.training - Saved best checkpoint: /home/user/code/ParrotLLM/runs/big_run/exp_c/run_20260409_192244/checkpoints/best_loss_3p9550_epoch_0000_step_0000900.pt
316
+ 2026-04-09 19:29:14 - INFO - parrotllm.training - ------------------------------------------------------------
317
+ 2026-04-09 19:29:17 - INFO - parrotllm.training - step 910 | epoch 0 | loss 4.0394 | lr 2.42e-03 | grad 0.1649
318
+ 2026-04-09 19:29:20 - INFO - parrotllm.training - step 920 | epoch 0 | loss 4.0690 | lr 2.41e-03 | grad 0.2173
319
+ 2026-04-09 19:29:23 - INFO - parrotllm.training - step 930 | epoch 0 | loss 4.0223 | lr 2.40e-03 | grad 0.1505
320
+ 2026-04-09 19:29:26 - INFO - parrotllm.training - step 940 | epoch 0 | loss 4.0437 | lr 2.40e-03 | grad 0.1540
321
+ 2026-04-09 19:29:30 - INFO - parrotllm.training - step 950 | epoch 0 | loss 4.0797 | lr 2.39e-03 | grad 0.1777
322
+ 2026-04-09 19:29:33 - INFO - parrotllm.training - step 960 | epoch 0 | loss 3.9175 | lr 2.38e-03 | grad 0.1780
323
+ 2026-04-09 19:29:36 - INFO - parrotllm.training - step 970 | epoch 0 | loss 4.1060 | lr 2.37e-03 | grad 0.2104
324
+ 2026-04-09 19:29:39 - INFO - parrotllm.training - step 980 | epoch 0 | loss 3.9461 | lr 2.36e-03 | grad 0.1551
325
+ 2026-04-09 19:29:42 - INFO - parrotllm.training - step 990 | epoch 0 | loss 3.8077 | lr 2.35e-03 | grad 0.1596
326
+ 2026-04-09 19:29:45 - INFO - parrotllm.training - step 1000 | epoch 0 | loss 4.0107 | lr 2.34e-03 | grad 0.1732
327
+ 2026-04-09 19:29:45 - INFO - parrotllm.training - Starting evaluation...
328
+ 2026-04-09 19:29:45 - INFO - parrotllm.training - ------------------------------------------------------------
329
+ 2026-04-09 19:29:49 - INFO - parrotllm.training - Train: loss=4.0107, ppl=55.19
330
+ 2026-04-09 19:29:49 - INFO - parrotllm.training - Val: loss=3.9101, ppl=49.91
331
+ 2026-04-09 19:29:49 - INFO - parrotllm.training - ** New best validation loss! **
332
+ 2026-04-09 19:29:50 - INFO - parrotllm.training - Saved best checkpoint: /home/user/code/ParrotLLM/runs/big_run/exp_c/run_20260409_192244/checkpoints/best_loss_3p9101_epoch_0000_step_0001000.pt
333
+ 2026-04-09 19:29:51 - INFO - parrotllm.training - ------------------------------------------------------------
334
+ 2026-04-09 19:29:52 - INFO - parrotllm.training - Saved checkpoint: /home/user/code/ParrotLLM/runs/big_run/exp_c/run_20260409_192244/checkpoints/last_epoch_0000_step_0001000.pt
335
+ 2026-04-09 19:29:55 - INFO - parrotllm.training - step 1010 | epoch 0 | loss 3.8678 | lr 2.33e-03 | grad 0.1643
336
+ 2026-04-09 19:29:59 - INFO - parrotllm.training - step 1020 | epoch 0 | loss 3.8991 | lr 2.32e-03 | grad 0.1999
337
+ 2026-04-09 19:30:02 - INFO - parrotllm.training - step 1030 | epoch 0 | loss 3.9746 | lr 2.31e-03 | grad 0.1807
338
+ 2026-04-09 19:30:05 - INFO - parrotllm.training - step 1040 | epoch 0 | loss 3.9925 | lr 2.31e-03 | grad 0.1795
339
+ 2026-04-09 19:30:08 - INFO - parrotllm.training - step 1050 | epoch 0 | loss 3.9166 | lr 2.30e-03 | grad 0.1554
340
+ 2026-04-09 19:30:11 - INFO - parrotllm.training - step 1060 | epoch 0 | loss 3.9566 | lr 2.29e-03 | grad 0.1591
341
+ 2026-04-09 19:30:14 - INFO - parrotllm.training - step 1070 | epoch 0 | loss 4.0184 | lr 2.28e-03 | grad 0.2120
342
+ 2026-04-09 19:30:17 - INFO - parrotllm.training - step 1080 | epoch 0 | loss 4.0702 | lr 2.27e-03 | grad 0.1746
343
+ 2026-04-09 19:30:20 - INFO - parrotllm.training - step 1090 | epoch 0 | loss 3.9541 | lr 2.26e-03 | grad 0.1647
344
+ 2026-04-09 19:30:23 - INFO - parrotllm.training - step 1100 | epoch 0 | loss 3.8897 | lr 2.25e-03 | grad 0.1644
345
+ 2026-04-09 19:30:23 - INFO - parrotllm.training - Starting evaluation...
346
+ 2026-04-09 19:30:23 - INFO - parrotllm.training - ------------------------------------------------------------
347
+ 2026-04-09 19:30:27 - INFO - parrotllm.training - Train: loss=3.8897, ppl=48.90
348
+ 2026-04-09 19:30:27 - INFO - parrotllm.training - Val: loss=3.8621, ppl=47.56
349
+ 2026-04-09 19:30:28 - INFO - parrotllm.training - ** New best validation loss! **
350
+ 2026-04-09 19:30:28 - INFO - parrotllm.training - Saved best checkpoint: /home/user/code/ParrotLLM/runs/big_run/exp_c/run_20260409_192244/checkpoints/best_loss_3p8621_epoch_0000_step_0001100.pt
351
+ 2026-04-09 19:30:29 - INFO - parrotllm.training - ------------------------------------------------------------
352
+ 2026-04-09 19:30:32 - INFO - parrotllm.training - step 1110 | epoch 0 | loss 3.8817 | lr 2.23e-03 | grad 0.1680
353
+ 2026-04-09 19:30:35 - INFO - parrotllm.training - step 1120 | epoch 0 | loss 3.8925 | lr 2.22e-03 | grad 0.1684
354
+ 2026-04-09 19:30:38 - INFO - parrotllm.training - step 1130 | epoch 0 | loss 3.9406 | lr 2.21e-03 | grad 0.1782
355
+ 2026-04-09 19:30:41 - INFO - parrotllm.training - step 1140 | epoch 0 | loss 3.8975 | lr 2.20e-03 | grad 0.1835
356
+ 2026-04-09 19:30:44 - INFO - parrotllm.training - step 1150 | epoch 0 | loss 3.9416 | lr 2.19e-03 | grad 0.1377
357
+ 2026-04-09 19:30:47 - INFO - parrotllm.training - step 1160 | epoch 0 | loss 3.8793 | lr 2.18e-03 | grad 0.1616
358
+ 2026-04-09 19:30:51 - INFO - parrotllm.training - step 1170 | epoch 0 | loss 3.9490 | lr 2.17e-03 | grad 0.1472
359
+ 2026-04-09 19:30:54 - INFO - parrotllm.training - step 1180 | epoch 0 | loss 3.8381 | lr 2.16e-03 | grad 0.1818
360
+ 2026-04-09 19:30:57 - INFO - parrotllm.training - step 1190 | epoch 0 | loss 3.8718 | lr 2.15e-03 | grad 0.1754
361
+ 2026-04-09 19:31:00 - INFO - parrotllm.training - step 1200 | epoch 0 | loss 3.9290 | lr 2.14e-03 | grad 0.1397
362
+ 2026-04-09 19:31:00 - INFO - parrotllm.training - Starting evaluation...
363
+ 2026-04-09 19:31:00 - INFO - parrotllm.training - ------------------------------------------------------------
364
+ 2026-04-09 19:31:04 - INFO - parrotllm.training - Train: loss=3.9290, ppl=50.86
365
+ 2026-04-09 19:31:04 - INFO - parrotllm.training - Val: loss=3.8243, ppl=45.80
366
+ 2026-04-09 19:31:04 - INFO - parrotllm.training - ** New best validation loss! **
367
+ 2026-04-09 19:31:05 - INFO - parrotllm.training - Saved best checkpoint: /home/user/code/ParrotLLM/runs/big_run/exp_c/run_20260409_192244/checkpoints/best_loss_3p8243_epoch_0000_step_0001200.pt
368
+ 2026-04-09 19:31:06 - INFO - parrotllm.training - ------------------------------------------------------------
369
+ 2026-04-09 19:31:09 - INFO - parrotllm.training - step 1210 | epoch 0 | loss 3.8821 | lr 2.13e-03 | grad 0.1633
370
+ 2026-04-09 19:31:12 - INFO - parrotllm.training - step 1220 | epoch 0 | loss 3.8870 | lr 2.11e-03 | grad 0.1922
371
+ 2026-04-09 19:31:15 - INFO - parrotllm.training - step 1230 | epoch 0 | loss 3.8452 | lr 2.10e-03 | grad 0.1457
372
+ 2026-04-09 19:31:18 - INFO - parrotllm.training - step 1240 | epoch 0 | loss 3.7907 | lr 2.09e-03 | grad 0.1580
373
+ 2026-04-09 19:31:21 - INFO - parrotllm.training - step 1250 | epoch 0 | loss 3.8199 | lr 2.08e-03 | grad 0.1638
374
+ 2026-04-09 19:31:22 - INFO - parrotllm.training - Saved checkpoint: /home/user/code/ParrotLLM/runs/big_run/exp_c/run_20260409_192244/checkpoints/last_epoch_0000_step_0001250.pt
375
+ 2026-04-09 19:31:26 - INFO - parrotllm.training - step 1260 | epoch 0 | loss 3.6741 | lr 2.07e-03 | grad 0.1678
376
+ 2026-04-09 19:31:29 - INFO - parrotllm.training - step 1270 | epoch 0 | loss 3.8253 | lr 2.05e-03 | grad 0.1685
377
+ 2026-04-09 19:31:32 - INFO - parrotllm.training - step 1280 | epoch 0 | loss 3.7942 | lr 2.04e-03 | grad 0.1476
378
+ 2026-04-09 19:31:35 - INFO - parrotllm.training - step 1290 | epoch 0 | loss 3.8815 | lr 2.03e-03 | grad 0.1480
379
+ 2026-04-09 19:31:38 - INFO - parrotllm.training - step 1300 | epoch 0 | loss 3.7826 | lr 2.02e-03 | grad 0.1907
380
+ 2026-04-09 19:31:38 - INFO - parrotllm.training - Starting evaluation...
381
+ 2026-04-09 19:31:38 - INFO - parrotllm.training - ------------------------------------------------------------
382
+ 2026-04-09 19:31:42 - INFO - parrotllm.training - Train: loss=3.7826, ppl=43.93
383
+ 2026-04-09 19:31:42 - INFO - parrotllm.training - Val: loss=3.7944, ppl=44.45
384
+ 2026-04-09 19:31:42 - INFO - parrotllm.training - ** New best validation loss! **
385
+ 2026-04-09 19:31:43 - INFO - parrotllm.training - Saved best checkpoint: /home/user/code/ParrotLLM/runs/big_run/exp_c/run_20260409_192244/checkpoints/best_loss_3p7944_epoch_0000_step_0001300.pt
386
+ 2026-04-09 19:31:44 - INFO - parrotllm.training - ------------------------------------------------------------
387
+ 2026-04-09 19:31:47 - INFO - parrotllm.training - step 1310 | epoch 0 | loss 3.7988 | lr 2.01e-03 | grad 0.1539
388
+ 2026-04-09 19:31:50 - INFO - parrotllm.training - step 1320 | epoch 0 | loss 3.8407 | lr 1.99e-03 | grad 0.1622
389
+ 2026-04-09 19:31:53 - INFO - parrotllm.training - step 1330 | epoch 0 | loss 3.7445 | lr 1.98e-03 | grad 0.1511
390
+ 2026-04-09 19:31:56 - INFO - parrotllm.training - step 1340 | epoch 0 | loss 3.8838 | lr 1.97e-03 | grad 0.1756
391
+ 2026-04-09 19:31:59 - INFO - parrotllm.training - step 1350 | epoch 0 | loss 3.7474 | lr 1.96e-03 | grad 0.1575
392
+ 2026-04-09 19:32:02 - INFO - parrotllm.training - step 1360 | epoch 0 | loss 3.8234 | lr 1.94e-03 | grad 0.1641
393
+ 2026-04-09 19:32:05 - INFO - parrotllm.training - step 1370 | epoch 0 | loss 3.9062 | lr 1.93e-03 | grad 0.1524
394
+ 2026-04-09 19:32:08 - INFO - parrotllm.training - step 1380 | epoch 0 | loss 3.7031 | lr 1.92e-03 | grad 0.1499
395
+ 2026-04-09 19:32:11 - INFO - parrotllm.training - step 1390 | epoch 0 | loss 3.9524 | lr 1.90e-03 | grad 0.1539
396
+ 2026-04-09 19:32:14 - INFO - parrotllm.training - step 1400 | epoch 0 | loss 3.6965 | lr 1.89e-03 | grad 0.1777
397
+ 2026-04-09 19:32:14 - INFO - parrotllm.training - Starting evaluation...
398
+ 2026-04-09 19:32:14 - INFO - parrotllm.training - ------------------------------------------------------------
399
+ 2026-04-09 19:32:18 - INFO - parrotllm.training - Train: loss=3.6965, ppl=40.31
400
+ 2026-04-09 19:32:18 - INFO - parrotllm.training - Val: loss=3.7617, ppl=43.02
401
+ 2026-04-09 19:32:18 - INFO - parrotllm.training - ** New best validation loss! **
402
+ 2026-04-09 19:32:19 - INFO - parrotllm.training - Saved best checkpoint: /home/user/code/ParrotLLM/runs/big_run/exp_c/run_20260409_192244/checkpoints/best_loss_3p7617_epoch_0000_step_0001400.pt
403
+ 2026-04-09 19:32:20 - INFO - parrotllm.training - ------------------------------------------------------------
404
+ 2026-04-09 19:32:23 - INFO - parrotllm.training - step 1410 | epoch 0 | loss 3.7099 | lr 1.88e-03 | grad 0.1702
405
+ 2026-04-09 19:32:26 - INFO - parrotllm.training - step 1420 | epoch 0 | loss 3.8496 | lr 1.87e-03 | grad 0.1519
406
+ 2026-04-09 19:32:29 - INFO - parrotllm.training - step 1430 | epoch 0 | loss 3.7974 | lr 1.85e-03 | grad 0.1633
407
+ 2026-04-09 19:32:32 - INFO - parrotllm.training - step 1440 | epoch 0 | loss 3.8250 | lr 1.84e-03 | grad 0.1639
408
+ 2026-04-09 19:32:35 - INFO - parrotllm.training - step 1450 | epoch 0 | loss 3.7777 | lr 1.83e-03 | grad 0.1590
409
+ 2026-04-09 19:32:38 - INFO - parrotllm.training - step 1460 | epoch 0 | loss 3.8960 | lr 1.81e-03 | grad 0.1570
410
+ 2026-04-09 19:32:41 - INFO - parrotllm.training - step 1470 | epoch 0 | loss 3.8311 | lr 1.80e-03 | grad 0.1560
411
+ 2026-04-09 19:32:45 - INFO - parrotllm.training - step 1480 | epoch 0 | loss 3.7554 | lr 1.79e-03 | grad 0.1625
412
+ 2026-04-09 19:32:48 - INFO - parrotllm.training - step 1490 | epoch 0 | loss 3.7514 | lr 1.77e-03 | grad 0.1456
413
+ 2026-04-09 19:32:51 - INFO - parrotllm.training - step 1500 | epoch 0 | loss 3.7365 | lr 1.76e-03 | grad 0.1668
414
+ 2026-04-09 19:32:51 - INFO - parrotllm.training - Starting evaluation...
415
+ 2026-04-09 19:32:51 - INFO - parrotllm.training - ------------------------------------------------------------
416
+ 2026-04-09 19:32:55 - INFO - parrotllm.training - Train: loss=3.7365, ppl=41.95
417
+ 2026-04-09 19:32:55 - INFO - parrotllm.training - Val: loss=3.7332, ppl=41.81
418
+ 2026-04-09 19:32:55 - INFO - parrotllm.training - ** New best validation loss! **
419
+ 2026-04-09 19:32:56 - INFO - parrotllm.training - Saved best checkpoint: /home/user/code/ParrotLLM/runs/big_run/exp_c/run_20260409_192244/checkpoints/best_loss_3p7332_epoch_0000_step_0001500.pt
420
+ 2026-04-09 19:32:57 - INFO - parrotllm.training - ------------------------------------------------------------
421
+ 2026-04-09 19:32:58 - INFO - parrotllm.training - Saved checkpoint: /home/user/code/ParrotLLM/runs/big_run/exp_c/run_20260409_192244/checkpoints/last_epoch_0000_step_0001500.pt
422
+ 2026-04-09 19:33:02 - INFO - parrotllm.training - step 1510 | epoch 0 | loss 3.8477 | lr 1.75e-03 | grad 0.1684
423
+ 2026-04-09 19:33:05 - INFO - parrotllm.training - step 1520 | epoch 0 | loss 3.6948 | lr 1.73e-03 | grad 0.1714
424
+ 2026-04-09 19:33:08 - INFO - parrotllm.training - step 1530 | epoch 0 | loss 3.7956 | lr 1.72e-03 | grad 0.1781
425
+ 2026-04-09 19:33:11 - INFO - parrotllm.training - step 1540 | epoch 0 | loss 3.7642 | lr 1.70e-03 | grad 0.1696
426
+ 2026-04-09 19:33:14 - INFO - parrotllm.training - step 1550 | epoch 0 | loss 3.7717 | lr 1.69e-03 | grad 0.1462
427
+ 2026-04-09 19:33:17 - INFO - parrotllm.training - step 1560 | epoch 0 | loss 3.7977 | lr 1.68e-03 | grad 0.1627
428
+ 2026-04-09 19:33:20 - INFO - parrotllm.training - step 1570 | epoch 0 | loss 3.7921 | lr 1.66e-03 | grad 0.1573
429
+ 2026-04-09 19:33:23 - INFO - parrotllm.training - step 1580 | epoch 0 | loss 3.7405 | lr 1.65e-03 | grad 0.1530
430
+ 2026-04-09 19:33:26 - INFO - parrotllm.training - step 1590 | epoch 0 | loss 3.7340 | lr 1.64e-03 | grad 0.1744
431
+ 2026-04-09 19:33:29 - INFO - parrotllm.training - step 1600 | epoch 0 | loss 3.8452 | lr 1.62e-03 | grad 0.1447
432
+ 2026-04-09 19:33:29 - INFO - parrotllm.training - Starting evaluation...
433
+ 2026-04-09 19:33:29 - INFO - parrotllm.training - ------------------------------------------------------------
434
+ 2026-04-09 19:33:34 - INFO - parrotllm.training - Train: loss=3.8452, ppl=46.77
435
+ 2026-04-09 19:33:34 - INFO - parrotllm.training - Val: loss=3.7085, ppl=40.79
436
+ 2026-04-09 19:33:34 - INFO - parrotllm.training - ** New best validation loss! **
437
+ 2026-04-09 19:33:35 - INFO - parrotllm.training - Saved best checkpoint: /home/user/code/ParrotLLM/runs/big_run/exp_c/run_20260409_192244/checkpoints/best_loss_3p7085_epoch_0000_step_0001600.pt
438
+ 2026-04-09 19:33:35 - INFO - parrotllm.training - ------------------------------------------------------------
439
+ 2026-04-09 19:33:38 - INFO - parrotllm.training - step 1610 | epoch 0 | loss 3.6809 | lr 1.61e-03 | grad 0.1392
440
+ 2026-04-09 19:33:42 - INFO - parrotllm.training - step 1620 | epoch 0 | loss 3.7141 | lr 1.59e-03 | grad 0.1385
441
+ 2026-04-09 19:33:45 - INFO - parrotllm.training - step 1630 | epoch 0 | loss 3.7681 | lr 1.58e-03 | grad 0.1424
442
+ 2026-04-09 19:33:48 - INFO - parrotllm.training - step 1640 | epoch 0 | loss 3.7280 | lr 1.57e-03 | grad 0.1671
443
+ 2026-04-09 19:33:51 - INFO - parrotllm.training - step 1650 | epoch 0 | loss 3.6658 | lr 1.55e-03 | grad 0.1553
444
+ 2026-04-09 19:33:54 - INFO - parrotllm.training - step 1660 | epoch 0 | loss 3.7679 | lr 1.54e-03 | grad 0.1440
445
+ 2026-04-09 19:33:57 - INFO - parrotllm.training - step 1670 | epoch 0 | loss 3.8576 | lr 1.52e-03 | grad 0.1582
446
+ 2026-04-09 19:34:00 - INFO - parrotllm.training - step 1680 | epoch 0 | loss 3.7788 | lr 1.51e-03 | grad 0.1434
447
+ 2026-04-09 19:34:03 - INFO - parrotllm.training - step 1690 | epoch 0 | loss 3.7171 | lr 1.50e-03 | grad 0.1471
448
+ 2026-04-09 19:34:06 - INFO - parrotllm.training - step 1700 | epoch 0 | loss 3.7558 | lr 1.48e-03 | grad 0.1509
449
+ 2026-04-09 19:34:06 - INFO - parrotllm.training - Starting evaluation...
450
+ 2026-04-09 19:34:06 - INFO - parrotllm.training - ------------------------------------------------------------
451
+ 2026-04-09 19:34:10 - INFO - parrotllm.training - Train: loss=3.7558, ppl=42.77
452
+ 2026-04-09 19:34:10 - INFO - parrotllm.training - Val: loss=3.6849, ppl=39.84
453
+ 2026-04-09 19:34:10 - INFO - parrotllm.training - ** New best validation loss! **
454
+ 2026-04-09 19:34:11 - INFO - parrotllm.training - Saved best checkpoint: /home/user/code/ParrotLLM/runs/big_run/exp_c/run_20260409_192244/checkpoints/best_loss_3p6849_epoch_0000_step_0001700.pt
455
+ 2026-04-09 19:34:12 - INFO - parrotllm.training - ------------------------------------------------------------
456
+ 2026-04-09 19:34:15 - INFO - parrotllm.training - step 1710 | epoch 0 | loss 3.7556 | lr 1.47e-03 | grad 0.1578
457
+ 2026-04-09 19:34:18 - INFO - parrotllm.training - step 1720 | epoch 0 | loss 3.6854 | lr 1.45e-03 | grad 0.1708
458
+ 2026-04-09 19:34:21 - INFO - parrotllm.training - step 1730 | epoch 0 | loss 3.7151 | lr 1.44e-03 | grad 0.1564
459
+ 2026-04-09 19:34:24 - INFO - parrotllm.training - step 1740 | epoch 0 | loss 3.6616 | lr 1.43e-03 | grad 0.1424
460
+ 2026-04-09 19:34:27 - INFO - parrotllm.training - step 1750 | epoch 0 | loss 3.6739 | lr 1.41e-03 | grad 0.1505
461
+ 2026-04-09 19:34:28 - INFO - parrotllm.training - Saved checkpoint: /home/user/code/ParrotLLM/runs/big_run/exp_c/run_20260409_192244/checkpoints/last_epoch_0000_step_0001750.pt
462
+ 2026-04-09 19:34:32 - INFO - parrotllm.training - step 1760 | epoch 0 | loss 3.8875 | lr 1.40e-03 | grad 0.1450
463
+ 2026-04-09 19:34:35 - INFO - parrotllm.training - step 1770 | epoch 0 | loss 3.7368 | lr 1.38e-03 | grad 0.1441
464
+ 2026-04-09 19:34:38 - INFO - parrotllm.training - step 1780 | epoch 0 | loss 3.6378 | lr 1.37e-03 | grad 0.1433
465
+ 2026-04-09 19:34:41 - INFO - parrotllm.training - step 1790 | epoch 0 | loss 3.7168 | lr 1.36e-03 | grad 0.1590
466
+ 2026-04-09 19:34:44 - INFO - parrotllm.training - step 1800 | epoch 0 | loss 3.6690 | lr 1.34e-03 | grad 0.1482
467
+ 2026-04-09 19:34:44 - INFO - parrotllm.training - Starting evaluation...
468
+ 2026-04-09 19:34:44 - INFO - parrotllm.training - ------------------------------------------------------------
469
+ 2026-04-09 19:34:49 - INFO - parrotllm.training - Train: loss=3.6690, ppl=39.21
470
+ 2026-04-09 19:34:49 - INFO - parrotllm.training - Val: loss=3.6632, ppl=38.98
471
+ 2026-04-09 19:34:49 - INFO - parrotllm.training - ** New best validation loss! **
472
+ 2026-04-09 19:34:50 - INFO - parrotllm.training - Saved best checkpoint: /home/user/code/ParrotLLM/runs/big_run/exp_c/run_20260409_192244/checkpoints/best_loss_3p6632_epoch_0000_step_0001800.pt
473
+ 2026-04-09 19:34:50 - INFO - parrotllm.training - ------------------------------------------------------------
474
+ 2026-04-09 19:34:54 - INFO - parrotllm.training - step 1810 | epoch 0 | loss 3.7849 | lr 1.33e-03 | grad 0.1552
475
+ 2026-04-09 19:34:57 - INFO - parrotllm.training - step 1820 | epoch 0 | loss 3.7080 | lr 1.31e-03 | grad 0.1540
476
+ 2026-04-09 19:35:00 - INFO - parrotllm.training - step 1830 | epoch 0 | loss 3.7286 | lr 1.30e-03 | grad 0.1545
477
+ 2026-04-09 19:35:03 - INFO - parrotllm.training - step 1840 | epoch 0 | loss 3.6558 | lr 1.29e-03 | grad 0.1678
478
+ 2026-04-09 19:35:06 - INFO - parrotllm.training - step 1850 | epoch 0 | loss 3.6477 | lr 1.27e-03 | grad 0.1455
479
+ 2026-04-09 19:35:09 - INFO - parrotllm.training - step 1860 | epoch 0 | loss 3.6229 | lr 1.26e-03 | grad 0.1489
480
+ 2026-04-09 19:35:12 - INFO - parrotllm.training - step 1870 | epoch 0 | loss 3.8349 | lr 1.25e-03 | grad 0.1573
481
+ 2026-04-09 19:35:15 - INFO - parrotllm.training - step 1880 | epoch 0 | loss 3.7567 | lr 1.23e-03 | grad 0.1689
482
+ 2026-04-09 19:35:18 - INFO - parrotllm.training - step 1890 | epoch 0 | loss 3.7272 | lr 1.22e-03 | grad 0.1624
483
+ 2026-04-09 19:35:21 - INFO - parrotllm.training - step 1900 | epoch 0 | loss 3.7089 | lr 1.20e-03 | grad 0.1574
484
+ 2026-04-09 19:35:21 - INFO - parrotllm.training - Starting evaluation...
485
+ 2026-04-09 19:35:21 - INFO - parrotllm.training - ------------------------------------------------------------
486
+ 2026-04-09 19:35:25 - INFO - parrotllm.training - Train: loss=3.7089, ppl=40.81
487
+ 2026-04-09 19:35:25 - INFO - parrotllm.training - Val: loss=3.6427, ppl=38.19
488
+ 2026-04-09 19:35:25 - INFO - parrotllm.training - ** New best validation loss! **
489
+ 2026-04-09 19:35:26 - INFO - parrotllm.training - Saved best checkpoint: /home/user/code/ParrotLLM/runs/big_run/exp_c/run_20260409_192244/checkpoints/best_loss_3p6427_epoch_0000_step_0001900.pt
490
+ 2026-04-09 19:35:27 - INFO - parrotllm.training - ------------------------------------------------------------
491
+ 2026-04-09 19:35:30 - INFO - parrotllm.training - step 1910 | epoch 0 | loss 3.6757 | lr 1.19e-03 | grad 0.1428
492
+ 2026-04-09 19:35:33 - INFO - parrotllm.training - step 1920 | epoch 0 | loss 3.7864 | lr 1.18e-03 | grad 0.1525
493
+ 2026-04-09 19:35:36 - INFO - parrotllm.training - step 1930 | epoch 0 | loss 3.7259 | lr 1.16e-03 | grad 0.1621
494
+ 2026-04-09 19:35:39 - INFO - parrotllm.training - step 1940 | epoch 0 | loss 3.7083 | lr 1.15e-03 | grad 0.1398
495
+ 2026-04-09 19:35:42 - INFO - parrotllm.training - step 1950 | epoch 0 | loss 3.6407 | lr 1.14e-03 | grad 0.1455
496
+ 2026-04-09 19:35:45 - INFO - parrotllm.training - step 1960 | epoch 0 | loss 3.7185 | lr 1.12e-03 | grad 0.1533
497
+ 2026-04-09 19:35:48 - INFO - parrotllm.training - step 1970 | epoch 0 | loss 3.5741 | lr 1.11e-03 | grad 0.1645
498
+ 2026-04-09 19:35:51 - INFO - parrotllm.training - step 1980 | epoch 0 | loss 3.6570 | lr 1.10e-03 | grad 0.1363
499
+ 2026-04-09 19:35:54 - INFO - parrotllm.training - step 1990 | epoch 0 | loss 3.6748 | lr 1.08e-03 | grad 0.1477
500
+ 2026-04-09 19:35:57 - INFO - parrotllm.training - step 2000 | epoch 0 | loss 3.7078 | lr 1.07e-03 | grad 0.1542
501
+ 2026-04-09 19:35:57 - INFO - parrotllm.training - Starting evaluation...
502
+ 2026-04-09 19:35:57 - INFO - parrotllm.training - ------------------------------------------------------------
503
+ 2026-04-09 19:36:02 - INFO - parrotllm.training - Train: loss=3.7078, ppl=40.76
504
+ 2026-04-09 19:36:02 - INFO - parrotllm.training - Val: loss=3.6249, ppl=37.52
505
+ 2026-04-09 19:36:02 - INFO - parrotllm.training - ** New best validation loss! **
506
+ 2026-04-09 19:36:03 - INFO - parrotllm.training - Saved best checkpoint: /home/user/code/ParrotLLM/runs/big_run/exp_c/run_20260409_192244/checkpoints/best_loss_3p6249_epoch_0000_step_0002000.pt
507
+ 2026-04-09 19:36:03 - INFO - parrotllm.training - ------------------------------------------------------------
508
+ 2026-04-09 19:36:04 - INFO - parrotllm.training - Saved checkpoint: /home/user/code/ParrotLLM/runs/big_run/exp_c/run_20260409_192244/checkpoints/last_epoch_0000_step_0002000.pt
509
+ 2026-04-09 19:36:08 - INFO - parrotllm.training - step 2010 | epoch 0 | loss 3.6597 | lr 1.06e-03 | grad 0.1367
510
+ 2026-04-09 19:36:12 - INFO - parrotllm.training - step 2020 | epoch 0 | loss 3.6953 | lr 1.04e-03 | grad 0.1411
511
+ 2026-04-09 19:36:15 - INFO - parrotllm.training - step 2030 | epoch 0 | loss 3.7515 | lr 1.03e-03 | grad 0.1533
512
+ 2026-04-09 19:36:18 - INFO - parrotllm.training - step 2040 | epoch 0 | loss 3.6105 | lr 1.02e-03 | grad 0.1512
513
+ 2026-04-09 19:36:21 - INFO - parrotllm.training - step 2050 | epoch 0 | loss 3.5427 | lr 1.00e-03 | grad 0.1503
514
+ 2026-04-09 19:36:24 - INFO - parrotllm.training - step 2060 | epoch 0 | loss 3.7112 | lr 9.90e-04 | grad 0.1319
515
+ 2026-04-09 19:36:27 - INFO - parrotllm.training - step 2070 | epoch 0 | loss 3.6418 | lr 9.77e-04 | grad 0.1432
516
+ 2026-04-09 19:36:30 - INFO - parrotllm.training - step 2080 | epoch 0 | loss 3.6107 | lr 9.64e-04 | grad 0.1359
517
+ 2026-04-09 19:36:33 - INFO - parrotllm.training - step 2090 | epoch 0 | loss 3.6785 | lr 9.52e-04 | grad 0.1453
518
+ 2026-04-09 19:36:36 - INFO - parrotllm.training - step 2100 | epoch 0 | loss 3.6705 | lr 9.39e-04 | grad 0.1484
519
+ 2026-04-09 19:36:36 - INFO - parrotllm.training - Starting evaluation...
520
+ 2026-04-09 19:36:36 - INFO - parrotllm.training - ------------------------------------------------------------
521
+ 2026-04-09 19:36:40 - INFO - parrotllm.training - Train: loss=3.6705, ppl=39.27
522
+ 2026-04-09 19:36:40 - INFO - parrotllm.training - Val: loss=3.6072, ppl=36.86
523
+ 2026-04-09 19:36:40 - INFO - parrotllm.training - ** New best validation loss! **
524
+ 2026-04-09 19:36:41 - INFO - parrotllm.training - Saved best checkpoint: /home/user/code/ParrotLLM/runs/big_run/exp_c/run_20260409_192244/checkpoints/best_loss_3p6072_epoch_0000_step_0002100.pt
525
+ 2026-04-09 19:36:42 - INFO - parrotllm.training - ------------------------------------------------------------
526
+ 2026-04-09 19:36:45 - INFO - parrotllm.training - step 2110 | epoch 0 | loss 3.6450 | lr 9.26e-04 | grad 0.1534
527
+ 2026-04-09 19:36:48 - INFO - parrotllm.training - step 2120 | epoch 0 | loss 3.6801 | lr 9.14e-04 | grad 0.1471
528
+ 2026-04-09 19:36:51 - INFO - parrotllm.training - step 2130 | epoch 0 | loss 3.6407 | lr 9.01e-04 | grad 0.1397
529
+ 2026-04-09 19:36:54 - INFO - parrotllm.training - step 2140 | epoch 0 | loss 3.6711 | lr 8.89e-04 | grad 0.1434
530
+ 2026-04-09 19:36:57 - INFO - parrotllm.training - step 2150 | epoch 0 | loss 3.7211 | lr 8.76e-04 | grad 0.1421
531
+ 2026-04-09 19:37:00 - INFO - parrotllm.training - step 2160 | epoch 0 | loss 3.5758 | lr 8.64e-04 | grad 0.1594
532
+ 2026-04-09 19:37:03 - INFO - parrotllm.training - step 2170 | epoch 0 | loss 3.5716 | lr 8.52e-04 | grad 0.1354
533
+ 2026-04-09 19:37:06 - INFO - parrotllm.training - step 2180 | epoch 0 | loss 3.6347 | lr 8.40e-04 | grad 0.1495
534
+ 2026-04-09 19:37:09 - INFO - parrotllm.training - step 2190 | epoch 0 | loss 3.6442 | lr 8.28e-04 | grad 0.1412
535
+ 2026-04-09 19:37:12 - INFO - parrotllm.training - step 2200 | epoch 0 | loss 3.6278 | lr 8.16e-04 | grad 0.1514
536
+ 2026-04-09 19:37:12 - INFO - parrotllm.training - Starting evaluation...
537
+ 2026-04-09 19:37:12 - INFO - parrotllm.training - ------------------------------------------------------------
538
+ 2026-04-09 19:37:17 - INFO - parrotllm.training - Train: loss=3.6278, ppl=37.63
539
+ 2026-04-09 19:37:17 - INFO - parrotllm.training - Val: loss=3.5905, ppl=36.25
540
+ 2026-04-09 19:37:17 - INFO - parrotllm.training - ** New best validation loss! **
541
+ 2026-04-09 19:37:18 - INFO - parrotllm.training - Saved best checkpoint: /home/user/code/ParrotLLM/runs/big_run/exp_c/run_20260409_192244/checkpoints/best_loss_3p5905_epoch_0000_step_0002200.pt
542
+ 2026-04-09 19:37:18 - INFO - parrotllm.training - ------------------------------------------------------------
543
+ 2026-04-09 19:37:21 - INFO - parrotllm.training - step 2210 | epoch 0 | loss 3.6088 | lr 8.04e-04 | grad 0.1355
544
+ 2026-04-09 19:37:25 - INFO - parrotllm.training - step 2220 | epoch 0 | loss 3.6510 | lr 7.93e-04 | grad 0.1476
545
+ 2026-04-09 19:37:28 - INFO - parrotllm.training - step 2230 | epoch 0 | loss 3.5770 | lr 7.81e-04 | grad 0.1391
546
+ 2026-04-09 19:37:31 - INFO - parrotllm.training - step 2240 | epoch 0 | loss 3.5278 | lr 7.69e-04 | grad 0.1373
547
+ 2026-04-09 19:37:34 - INFO - parrotllm.training - step 2250 | epoch 0 | loss 3.5650 | lr 7.58e-04 | grad 0.1433
548
+ 2026-04-09 19:37:35 - INFO - parrotllm.training - Saved checkpoint: /home/user/code/ParrotLLM/runs/big_run/exp_c/run_20260409_192244/checkpoints/last_epoch_0000_step_0002250.pt
549
+ 2026-04-09 19:37:38 - INFO - parrotllm.training - step 2260 | epoch 0 | loss 3.6896 | lr 7.47e-04 | grad 0.1451
550
+ 2026-04-09 19:37:42 - INFO - parrotllm.training - step 2270 | epoch 0 | loss 3.5713 | lr 7.35e-04 | grad 0.1370
551
+ 2026-04-09 19:37:45 - INFO - parrotllm.training - step 2280 | epoch 0 | loss 3.6741 | lr 7.24e-04 | grad 0.1521
552
+ 2026-04-09 19:37:48 - INFO - parrotllm.training - step 2290 | epoch 0 | loss 3.7122 | lr 7.13e-04 | grad 0.1403
553
+ 2026-04-09 19:37:51 - INFO - parrotllm.training - step 2300 | epoch 0 | loss 3.6300 | lr 7.02e-04 | grad 0.1498
554
+ 2026-04-09 19:37:51 - INFO - parrotllm.training - Starting evaluation...
555
+ 2026-04-09 19:37:51 - INFO - parrotllm.training - ------------------------------------------------------------
556
+ 2026-04-09 19:37:55 - INFO - parrotllm.training - Train: loss=3.6300, ppl=37.71
557
+ 2026-04-09 19:37:55 - INFO - parrotllm.training - Val: loss=3.5771, ppl=35.77
558
+ 2026-04-09 19:37:55 - INFO - parrotllm.training - ** New best validation loss! **
559
+ 2026-04-09 19:37:56 - INFO - parrotllm.training - Saved best checkpoint: /home/user/code/ParrotLLM/runs/big_run/exp_c/run_20260409_192244/checkpoints/best_loss_3p5771_epoch_0000_step_0002300.pt
560
+ 2026-04-09 19:37:57 - INFO - parrotllm.training - ------------------------------------------------------------
561
+ 2026-04-09 19:38:00 - INFO - parrotllm.training - step 2310 | epoch 0 | loss 3.6523 | lr 6.91e-04 | grad 0.1646
562
+ 2026-04-09 19:38:03 - INFO - parrotllm.training - step 2320 | epoch 0 | loss 3.6099 | lr 6.81e-04 | grad 0.1446
563
+ 2026-04-09 19:38:06 - INFO - parrotllm.training - step 2330 | epoch 0 | loss 3.4824 | lr 6.70e-04 | grad 0.1476
564
+ 2026-04-09 19:38:09 - INFO - parrotllm.training - step 2340 | epoch 0 | loss 3.6276 | lr 6.59e-04 | grad 0.1464
565
+ 2026-04-09 19:38:12 - INFO - parrotllm.training - step 2350 | epoch 0 | loss 3.6549 | lr 6.49e-04 | grad 0.1371
566
+ 2026-04-09 19:38:15 - INFO - parrotllm.training - step 2360 | epoch 0 | loss 3.6301 | lr 6.39e-04 | grad 0.1446
567
+ 2026-04-09 19:38:18 - INFO - parrotllm.training - step 2370 | epoch 0 | loss 3.6084 | lr 6.28e-04 | grad 0.1419
568
+ 2026-04-09 19:38:21 - INFO - parrotllm.training - step 2380 | epoch 0 | loss 3.5490 | lr 6.18e-04 | grad 0.1553
569
+ 2026-04-09 19:38:24 - INFO - parrotllm.training - step 2390 | epoch 0 | loss 3.6327 | lr 6.08e-04 | grad 0.1450
570
+ 2026-04-09 19:38:27 - INFO - parrotllm.training - step 2400 | epoch 0 | loss 3.6931 | lr 5.99e-04 | grad 0.1572
571
+ 2026-04-09 19:38:27 - INFO - parrotllm.training - Starting evaluation...
572
+ 2026-04-09 19:38:27 - INFO - parrotllm.training - ------------------------------------------------------------
573
+ 2026-04-09 19:38:32 - INFO - parrotllm.training - Train: loss=3.6931, ppl=40.17
574
+ 2026-04-09 19:38:32 - INFO - parrotllm.training - Val: loss=3.5645, ppl=35.32
575
+ 2026-04-09 19:38:32 - INFO - parrotllm.training - ** New best validation loss! **
576
+ 2026-04-09 19:38:33 - INFO - parrotllm.training - Saved best checkpoint: /home/user/code/ParrotLLM/runs/big_run/exp_c/run_20260409_192244/checkpoints/best_loss_3p5645_epoch_0000_step_0002400.pt
577
+ 2026-04-09 19:38:34 - INFO - parrotllm.training - ------------------------------------------------------------
578
+ 2026-04-09 19:38:37 - INFO - parrotllm.training - step 2410 | epoch 0 | loss 3.6506 | lr 5.89e-04 | grad 0.1459
579
+ 2026-04-09 19:38:40 - INFO - parrotllm.training - step 2420 | epoch 0 | loss 3.5918 | lr 5.79e-04 | grad 0.1401
580
+ 2026-04-09 19:38:43 - INFO - parrotllm.training - step 2430 | epoch 0 | loss 3.4819 | lr 5.70e-04 | grad 0.1392
581
+ 2026-04-09 19:38:46 - INFO - parrotllm.training - step 2440 | epoch 0 | loss 3.5462 | lr 5.60e-04 | grad 0.1481
582
+ 2026-04-09 19:38:49 - INFO - parrotllm.training - step 2450 | epoch 0 | loss 3.6378 | lr 5.51e-04 | grad 0.1371
583
+ 2026-04-09 19:38:52 - INFO - parrotllm.training - step 2460 | epoch 0 | loss 3.5615 | lr 5.42e-04 | grad 0.1417
584
+ 2026-04-09 19:38:55 - INFO - parrotllm.training - step 2470 | epoch 0 | loss 3.5752 | lr 5.33e-04 | grad 0.1447
585
+ 2026-04-09 19:38:58 - INFO - parrotllm.training - step 2480 | epoch 0 | loss 3.5410 | lr 5.24e-04 | grad 0.1493
586
+ 2026-04-09 19:39:01 - INFO - parrotllm.training - step 2490 | epoch 0 | loss 3.5590 | lr 5.16e-04 | grad 0.1514
587
+ 2026-04-09 19:39:04 - INFO - parrotllm.training - step 2500 | epoch 0 | loss 3.5974 | lr 5.07e-04 | grad 0.1356
588
+ 2026-04-09 19:39:04 - INFO - parrotllm.training - Starting evaluation...
589
+ 2026-04-09 19:39:04 - INFO - parrotllm.training - ------------------------------------------------------------
590
+ 2026-04-09 19:39:08 - INFO - parrotllm.training - Train: loss=3.5974, ppl=36.50
591
+ 2026-04-09 19:39:08 - INFO - parrotllm.training - Val: loss=3.5520, ppl=34.88
592
+ 2026-04-09 19:39:08 - INFO - parrotllm.training - ** New best validation loss! **
593
+ 2026-04-09 19:39:09 - INFO - parrotllm.training - Saved best checkpoint: /home/user/code/ParrotLLM/runs/big_run/exp_c/run_20260409_192244/checkpoints/best_loss_3p5520_epoch_0000_step_0002500.pt
594
+ 2026-04-09 19:39:10 - INFO - parrotllm.training - ------------------------------------------------------------
595
+ 2026-04-09 19:39:11 - INFO - parrotllm.training - Saved checkpoint: /home/user/code/ParrotLLM/runs/big_run/exp_c/run_20260409_192244/checkpoints/last_epoch_0000_step_0002500.pt
596
+ 2026-04-09 19:39:15 - INFO - parrotllm.training - step 2510 | epoch 0 | loss 3.6468 | lr 4.99e-04 | grad 0.1362
597
+ 2026-04-09 19:39:18 - INFO - parrotllm.training - step 2520 | epoch 0 | loss 3.6082 | lr 4.90e-04 | grad 0.1547
598
+ 2026-04-09 19:39:21 - INFO - parrotllm.training - step 2530 | epoch 0 | loss 3.6388 | lr 4.82e-04 | grad 0.1436
599
+ 2026-04-09 19:39:24 - INFO - parrotllm.training - step 2540 | epoch 0 | loss 3.5896 | lr 4.74e-04 | grad 0.1529
600
+ 2026-04-09 19:39:27 - INFO - parrotllm.training - step 2550 | epoch 0 | loss 3.4892 | lr 4.66e-04 | grad 0.1363
601
+ 2026-04-09 19:39:30 - INFO - parrotllm.training - step 2560 | epoch 0 | loss 3.6016 | lr 4.58e-04 | grad 0.1442
602
+ 2026-04-09 19:39:33 - INFO - parrotllm.training - step 2570 | epoch 0 | loss 3.5068 | lr 4.51e-04 | grad 0.1346
603
+ 2026-04-09 19:39:36 - INFO - parrotllm.training - step 2580 | epoch 0 | loss 3.4655 | lr 4.43e-04 | grad 0.1403
604
+ 2026-04-09 19:39:39 - INFO - parrotllm.training - step 2590 | epoch 0 | loss 3.5318 | lr 4.36e-04 | grad 0.1326
605
+ 2026-04-09 19:39:42 - INFO - parrotllm.training - step 2600 | epoch 0 | loss 3.5394 | lr 4.29e-04 | grad 0.1379
606
+ 2026-04-09 19:39:42 - INFO - parrotllm.training - Starting evaluation...
607
+ 2026-04-09 19:39:42 - INFO - parrotllm.training - ------------------------------------------------------------
608
+ 2026-04-09 19:39:47 - INFO - parrotllm.training - Train: loss=3.5394, ppl=34.45
609
+ 2026-04-09 19:39:47 - INFO - parrotllm.training - Val: loss=3.5420, ppl=34.53
610
+ 2026-04-09 19:39:47 - INFO - parrotllm.training - ** New best validation loss! **
611
+ 2026-04-09 19:39:48 - INFO - parrotllm.training - Saved best checkpoint: /home/user/code/ParrotLLM/runs/big_run/exp_c/run_20260409_192244/checkpoints/best_loss_3p5420_epoch_0000_step_0002600.pt
612
+ 2026-04-09 19:39:48 - INFO - parrotllm.training - ------------------------------------------------------------
613
+ 2026-04-09 19:39:51 - INFO - parrotllm.training - step 2610 | epoch 0 | loss 3.4649 | lr 4.21e-04 | grad 0.1430
614
+ 2026-04-09 19:39:55 - INFO - parrotllm.training - step 2620 | epoch 0 | loss 3.5399 | lr 4.14e-04 | grad 0.1355
615
+ 2026-04-09 19:39:58 - INFO - parrotllm.training - step 2630 | epoch 0 | loss 3.5766 | lr 4.08e-04 | grad 0.1434
616
+ 2026-04-09 19:40:01 - INFO - parrotllm.training - step 2640 | epoch 0 | loss 3.5205 | lr 4.01e-04 | grad 0.1273
617
+ 2026-04-09 19:40:04 - INFO - parrotllm.training - step 2650 | epoch 0 | loss 3.5356 | lr 3.95e-04 | grad 0.1390
618
+ 2026-04-09 19:40:07 - INFO - parrotllm.training - step 2660 | epoch 0 | loss 3.4929 | lr 3.88e-04 | grad 0.1335
619
+ 2026-04-09 19:40:10 - INFO - parrotllm.training - step 2670 | epoch 0 | loss 3.6162 | lr 3.82e-04 | grad 0.1337
620
+ 2026-04-09 19:40:13 - INFO - parrotllm.training - step 2680 | epoch 0 | loss 3.6167 | lr 3.76e-04 | grad 0.1389
621
+ 2026-04-09 19:40:16 - INFO - parrotllm.training - step 2690 | epoch 0 | loss 3.5075 | lr 3.70e-04 | grad 0.1497
622
+ 2026-04-09 19:40:19 - INFO - parrotllm.training - step 2700 | epoch 0 | loss 3.5553 | lr 3.64e-04 | grad 0.1404
623
+ 2026-04-09 19:40:19 - INFO - parrotllm.training - Starting evaluation...
624
+ 2026-04-09 19:40:19 - INFO - parrotllm.training - ------------------------------------------------------------
625
+ 2026-04-09 19:40:23 - INFO - parrotllm.training - Train: loss=3.5553, ppl=35.00
626
+ 2026-04-09 19:40:23 - INFO - parrotllm.training - Val: loss=3.5320, ppl=34.19
627
+ 2026-04-09 19:40:23 - INFO - parrotllm.training - ** New best validation loss! **
628
+ 2026-04-09 19:40:24 - INFO - parrotllm.training - Saved best checkpoint: /home/user/code/ParrotLLM/runs/big_run/exp_c/run_20260409_192244/checkpoints/best_loss_3p5320_epoch_0000_step_0002700.pt
629
+ 2026-04-09 19:40:25 - INFO - parrotllm.training - ------------------------------------------------------------
630
+ 2026-04-09 19:40:28 - INFO - parrotllm.training - step 2710 | epoch 0 | loss 3.5350 | lr 3.59e-04 | grad 0.1395
631
+ 2026-04-09 19:40:31 - INFO - parrotllm.training - step 2720 | epoch 0 | loss 3.4688 | lr 3.53e-04 | grad 0.1384
632
+ 2026-04-09 19:40:34 - INFO - parrotllm.training - step 2730 | epoch 0 | loss 3.5069 | lr 3.48e-04 | grad 0.1349
633
+ 2026-04-09 19:40:37 - INFO - parrotllm.training - step 2740 | epoch 0 | loss 3.5849 | lr 3.43e-04 | grad 0.1432
634
+ 2026-04-09 19:40:40 - INFO - parrotllm.training - step 2750 | epoch 0 | loss 3.5795 | lr 3.38e-04 | grad 0.1313
635
+ 2026-04-09 19:40:41 - INFO - parrotllm.training - Saved checkpoint: /home/user/code/ParrotLLM/runs/big_run/exp_c/run_20260409_192244/checkpoints/last_epoch_0000_step_0002750.pt
636
+ 2026-04-09 19:40:45 - INFO - parrotllm.training - step 2760 | epoch 0 | loss 3.5513 | lr 3.33e-04 | grad 0.1483
637
+ 2026-04-09 19:40:48 - INFO - parrotllm.training - step 2770 | epoch 0 | loss 3.5035 | lr 3.28e-04 | grad 0.1418
638
+ 2026-04-09 19:40:51 - INFO - parrotllm.training - step 2780 | epoch 0 | loss 3.6012 | lr 3.24e-04 | grad 0.1369
639
+ 2026-04-09 19:40:54 - INFO - parrotllm.training - step 2790 | epoch 0 | loss 3.5451 | lr 3.19e-04 | grad 0.1388
640
+ 2026-04-09 19:40:57 - INFO - parrotllm.training - step 2800 | epoch 0 | loss 3.5725 | lr 3.15e-04 | grad 0.1407
641
+ 2026-04-09 19:40:57 - INFO - parrotllm.training - Starting evaluation...
642
+ 2026-04-09 19:40:57 - INFO - parrotllm.training - ------------------------------------------------------------
643
+ 2026-04-09 19:41:01 - INFO - parrotllm.training - Train: loss=3.5725, ppl=35.61
644
+ 2026-04-09 19:41:01 - INFO - parrotllm.training - Val: loss=3.5242, ppl=33.93
645
+ 2026-04-09 19:41:01 - INFO - parrotllm.training - ** New best validation loss! **
646
+ 2026-04-09 19:41:02 - INFO - parrotllm.training - Saved best checkpoint: /home/user/code/ParrotLLM/runs/big_run/exp_c/run_20260409_192244/checkpoints/best_loss_3p5242_epoch_0000_step_0002800.pt
647
+ 2026-04-09 19:41:03 - INFO - parrotllm.training - ------------------------------------------------------------
648
+ 2026-04-09 19:41:06 - INFO - parrotllm.training - step 2810 | epoch 0 | loss 3.5289 | lr 3.11e-04 | grad 0.1398
649
+ 2026-04-09 19:41:09 - INFO - parrotllm.training - step 2820 | epoch 0 | loss 3.5736 | lr 3.07e-04 | grad 0.1376
650
+ 2026-04-09 19:41:12 - INFO - parrotllm.training - step 2830 | epoch 0 | loss 3.5175 | lr 3.04e-04 | grad 0.1417
651
+ 2026-04-09 19:41:15 - INFO - parrotllm.training - step 2840 | epoch 0 | loss 3.5799 | lr 3.00e-04 | grad 0.1371
652
+ 2026-04-09 19:41:18 - INFO - parrotllm.training - step 2850 | epoch 0 | loss 3.4911 | lr 2.97e-04 | grad 0.1369
653
+ 2026-04-09 19:41:21 - INFO - parrotllm.training - step 2860 | epoch 0 | loss 3.5844 | lr 2.93e-04 | grad 0.1362
654
+ 2026-04-09 19:41:24 - INFO - parrotllm.training - step 2870 | epoch 0 | loss 3.5515 | lr 2.90e-04 | grad 0.1411
655
+ 2026-04-09 19:41:27 - INFO - parrotllm.training - step 2880 | epoch 0 | loss 3.5206 | lr 2.87e-04 | grad 0.1356
656
+ 2026-04-09 19:41:30 - INFO - parrotllm.training - step 2890 | epoch 0 | loss 3.6623 | lr 2.85e-04 | grad 0.1302
657
+ 2026-04-09 19:41:33 - INFO - parrotllm.training - step 2900 | epoch 0 | loss 3.5918 | lr 2.82e-04 | grad 0.1431
658
+ 2026-04-09 19:41:33 - INFO - parrotllm.training - Starting evaluation...
659
+ 2026-04-09 19:41:33 - INFO - parrotllm.training - ------------------------------------------------------------
660
+ 2026-04-09 19:41:38 - INFO - parrotllm.training - Train: loss=3.5918, ppl=36.30
661
+ 2026-04-09 19:41:38 - INFO - parrotllm.training - Val: loss=3.5190, ppl=33.75
662
+ 2026-04-09 19:41:38 - INFO - parrotllm.training - ** New best validation loss! **
663
+ 2026-04-09 19:41:38 - INFO - parrotllm.training - Saved best checkpoint: /home/user/code/ParrotLLM/runs/big_run/exp_c/run_20260409_192244/checkpoints/best_loss_3p5190_epoch_0000_step_0002900.pt
664
+ 2026-04-09 19:41:39 - INFO - parrotllm.training - ------------------------------------------------------------
665
+ 2026-04-09 19:41:42 - INFO - parrotllm.training - step 2910 | epoch 0 | loss 3.5404 | lr 2.80e-04 | grad 0.1518
666
+ 2026-04-09 19:41:45 - INFO - parrotllm.training - step 2920 | epoch 0 | loss 3.5139 | lr 2.77e-04 | grad 0.1430
667
+ 2026-04-09 19:41:48 - INFO - parrotllm.training - step 2930 | epoch 0 | loss 3.5736 | lr 2.75e-04 | grad 0.1378
668
+ 2026-04-09 19:41:51 - INFO - parrotllm.training - step 2940 | epoch 0 | loss 3.3917 | lr 2.73e-04 | grad 0.1350
669
+ 2026-04-09 19:41:54 - INFO - parrotllm.training - step 2950 | epoch 0 | loss 3.4582 | lr 2.72e-04 | grad 0.1436
670
+ 2026-04-09 19:41:57 - INFO - parrotllm.training - step 2960 | epoch 0 | loss 3.5660 | lr 2.70e-04 | grad 0.1364
671
+ 2026-04-09 19:42:00 - INFO - parrotllm.training - step 2970 | epoch 0 | loss 3.5253 | lr 2.68e-04 | grad 0.1365
672
+ 2026-04-09 19:42:03 - INFO - parrotllm.training - step 2980 | epoch 0 | loss 3.4569 | lr 2.67e-04 | grad 0.1431
673
+ 2026-04-09 19:42:06 - INFO - parrotllm.training - step 2990 | epoch 0 | loss 3.5822 | lr 2.66e-04 | grad 0.1301
674
+ 2026-04-09 19:42:10 - INFO - parrotllm.training - step 3000 | epoch 0 | loss 3.5784 | lr 2.65e-04 | grad 0.1443
675
+ 2026-04-09 19:42:10 - INFO - parrotllm.training - Starting evaluation...
676
+ 2026-04-09 19:42:10 - INFO - parrotllm.training - ------------------------------------------------------------
677
+ 2026-04-09 19:42:14 - INFO - parrotllm.training - Train: loss=3.5784, ppl=35.82
678
+ 2026-04-09 19:42:14 - INFO - parrotllm.training - Val: loss=3.5130, ppl=33.55
679
+ 2026-04-09 19:42:14 - INFO - parrotllm.training - ** New best validation loss! **
680
+ 2026-04-09 19:42:15 - INFO - parrotllm.training - Saved best checkpoint: /home/user/code/ParrotLLM/runs/big_run/exp_c/run_20260409_192244/checkpoints/best_loss_3p5130_epoch_0000_step_0003000.pt
681
+ 2026-04-09 19:42:15 - INFO - parrotllm.training - ------------------------------------------------------------
682
+ 2026-04-09 19:42:16 - INFO - parrotllm.training - Saved checkpoint: /home/user/code/ParrotLLM/runs/big_run/exp_c/run_20260409_192244/checkpoints/last_epoch_0000_step_0003000.pt
683
+ 2026-04-09 19:42:20 - INFO - parrotllm.training - step 3010 | epoch 0 | loss 3.6020 | lr 2.64e-04 | grad 0.1366
684
+ 2026-04-09 19:42:24 - INFO - parrotllm.training - step 3020 | epoch 0 | loss 3.4855 | lr 2.64e-04 | grad 0.1366
685
+ 2026-04-09 19:42:27 - INFO - parrotllm.training - step 3030 | epoch 0 | loss 3.5683 | lr 2.63e-04 | grad 0.1481
686
+ 2026-04-09 19:42:30 - INFO - parrotllm.training - step 3040 | epoch 0 | loss 3.5218 | lr 2.63e-04 | grad 0.1450
687
+ 2026-04-09 19:42:33 - INFO - parrotllm.training - step 3050 | epoch 0 | loss 3.5467 | lr 2.63e-04 | grad 0.1403
688
+ 2026-04-09 19:42:34 - INFO - parrotllm.training - Saved final checkpoint: /home/user/code/ParrotLLM/runs/big_run/exp_c/run_20260409_192244/checkpoints/last_epoch_0000_step_0003052.pt
689
+ 2026-04-09 19:42:34 - INFO - parrotllm.training -
690
+ Loss Curve:
691
+ Loss
692
+ ^
693
+ 10.91 |*
694
+ |*
695
+ |*
696
+ |*
697
+ |**
698
+ | ****
699
+ | ******
700
+ 3.38 | *******************************************
701
+ +--------------------------------------------------> step
702
+ 1 3052
703
+ 2026-04-09 19:42:34 - INFO - parrotllm.training -
704
+ ============================================================
705
+ TRAINING COMPLETE
706
+ ============================================================
707
+ Epochs: 1
708
+ Total steps: 3052
709
+ Total time: 0.33 hours
710
+ Best validation loss: 3.5130
711
+ Run directory: runs/big_run/exp_c/run_20260409_192244
712
+ ============================================================
713
+ 2026-04-09 19:42:34 - INFO - parrotllm.training - Uploading completed run directory to Hugging Face: repo=ParrotLabs/Preprocessed
runs/big_run/exp_c/run_20260409_192244/training_plots.pdf ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9f43118167b534bebd2e2799ffbbf9d3ab57bdf2f2225c07f61401c34c98b17e
3
+ size 112948