Hugging Face's logo

Datasets:
ParrotLabs
/
Preprocessed

Dataset card Files
xet
 Community
thaferbe commited on
Commit
04d6525
·
verified ·
1 Parent(s): 9dc1bbc

Upload training run run_20260408_124138

Browse files

ParrotLLM end-of-training sync for local run runs/big_run/exp_c/run_20260408_124138.

Files changed (27) hide show
  1. .gitattributes +1 -0
  2. runs/big_run/exp_c/run_20260408_124138/.ipynb_checkpoints/train-checkpoint.log +491 -0
  3. runs/big_run/exp_c/run_20260408_124138/.ipynb_checkpoints/training_plots-checkpoint.pdf +0 -0
  4. runs/big_run/exp_c/run_20260408_124138/checkpoints/best_loss_3p5437_epoch_0001_step_0003000.pt +3 -0
  5. runs/big_run/exp_c/run_20260408_124138/checkpoints/best_loss_3p5634_epoch_0001_step_0002900.pt +3 -0
  6. runs/big_run/exp_c/run_20260408_124138/checkpoints/best_loss_3p5885_epoch_0001_step_0002800.pt +3 -0
  7. runs/big_run/exp_c/run_20260408_124138/checkpoints/best_loss_3p6128_epoch_0001_step_0002700.pt +3 -0
  8. runs/big_run/exp_c/run_20260408_124138/checkpoints/best_loss_3p6383_epoch_0001_step_0002600.pt +3 -0
  9. runs/big_run/exp_c/run_20260408_124138/checkpoints/best_loss_3p6614_epoch_0001_step_0002500.pt +3 -0
  10. runs/big_run/exp_c/run_20260408_124138/checkpoints/best_loss_3p6698_epoch_0001_step_0002400.pt +3 -0
  11. runs/big_run/exp_c/run_20260408_124138/checkpoints/best_loss_3p6838_epoch_0001_step_0002300.pt +3 -0
  12. runs/big_run/exp_c/run_20260408_124138/checkpoints/best_loss_3p6926_epoch_0001_step_0002200.pt +3 -0
  13. runs/big_run/exp_c/run_20260408_124138/checkpoints/best_loss_3p7036_epoch_0001_step_0002100.pt +3 -0
  14. runs/big_run/exp_c/run_20260408_124138/checkpoints/last_epoch_0000_step_0001000.pt +3 -0
  15. runs/big_run/exp_c/run_20260408_124138/checkpoints/last_epoch_0000_step_0001250.pt +3 -0
  16. runs/big_run/exp_c/run_20260408_124138/checkpoints/last_epoch_0000_step_0001500.pt +3 -0
  17. runs/big_run/exp_c/run_20260408_124138/checkpoints/last_epoch_0000_step_0001750.pt +3 -0
  18. runs/big_run/exp_c/run_20260408_124138/checkpoints/last_epoch_0001_step_0002000.pt +3 -0
  19. runs/big_run/exp_c/run_20260408_124138/checkpoints/last_epoch_0001_step_0002250.pt +3 -0
  20. runs/big_run/exp_c/run_20260408_124138/checkpoints/last_epoch_0001_step_0002500.pt +3 -0
  21. runs/big_run/exp_c/run_20260408_124138/checkpoints/last_epoch_0001_step_0002750.pt +3 -0
  22. runs/big_run/exp_c/run_20260408_124138/checkpoints/last_epoch_0001_step_0003000.pt +3 -0
  23. runs/big_run/exp_c/run_20260408_124138/checkpoints/last_epoch_0001_step_0003052.pt +3 -0
  24. runs/big_run/exp_c/run_20260408_124138/config.json +86 -0
  25. runs/big_run/exp_c/run_20260408_124138/metrics.jsonl +0 -0
  26. runs/big_run/exp_c/run_20260408_124138/train.log +719 -0
  27. runs/big_run/exp_c/run_20260408_124138/training_plots.pdf +3 -0
.gitattributes CHANGED
@@ -58,3 +58,4 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
58
  # Video files - compressed
59
  *.mp4 filter=lfs diff=lfs merge=lfs -text
60
  *.webm filter=lfs diff=lfs merge=lfs -text
 
 
58
  # Video files - compressed
59
  *.mp4 filter=lfs diff=lfs merge=lfs -text
60
  *.webm 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
runs/big_run/exp_c/run_20260408_124138/.ipynb_checkpoints/train-checkpoint.log ADDED
@@ -0,0 +1,491 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ 2026-04-08 12:41:39 - INFO - parrotllm - Logging initialised -> runs/big_run/exp_c/run_20260408_124138/train.log
2
+ 2026-04-08 12:41:39 - INFO - parrotllm.training - device=cuda:0 | rank=0 | world_size=8 | distributed=yes
3
+ 2026-04-08 12:41:39 - INFO - parrotllm.training - checkpoint directory=/home/user/code/ParrotLLM/runs/big_run/exp_c/run_20260408_124138/checkpoints | keep_last=10 | keep_best=10
4
+ 2026-04-08 12:41:40 - INFO - parrotllm.training - compiling model with torch.compile...
5
+ 2026-04-08 12:41:41 - INFO - parrotllm.training -
6
+ ============================================================
7
+ MODEL ARCHITECTURE SUMMARY
8
+ ============================================================
9
+
10
+ Configuration:
11
+ Vocab size: 50258
12
+ Block size (context): 1024
13
+ Layers: 14
14
+ Heads: 6
15
+ Embedding dim: 384
16
+ FFN hidden dim: 768
17
+ Dropout: 0.005
18
+ Bias: False
19
+
20
+ Parameters (unique, weight-tied layers counted once):
21
+ Total: 39,966,592 (39.97M)
22
+ Trainable: 39,966,592 (39.97M)
23
+ Non-trainable: 0
24
+ Non-embedding: 39,966,592 (39.97M)
25
+ Position embeddings: 0 (RoPE: 0 learned params)
26
+ Size (MB): 159.87
27
+
28
+ Layer-wise breakdown (torchinfo):
29
+ Note: torchinfo double-counts weight-tied layers (tok_emb/lm_head).
30
+ ============================================================================================================================================
31
+ Layer (type:depth-idx) Input Shape Output Shape Param # Trainable
32
+ ============================================================================================================================================
33
+ ParrotLLM [32, 1024] [32, 1024, 50258] -- True
34
+ ├─Embedding: 1-1 [32, 1024] [32, 1024, 384] 19,299,072 True
35
+ ├─Dropout: 1-2 [32, 1024, 384] [32, 1024, 384] -- --
36
+ ├─ModuleList: 1-3 -- -- -- True
37
+ │ └─TransformerBlock: 2-1 [32, 1024, 384] [32, 1024, 384] -- True
38
+ │ │ └─RMSNorm: 3-1 [32, 1024, 384] [32, 1024, 384] 384 True
39
+ │ │ └─MultiHeadAttention: 3-2 [32, 1024, 384] [32, 1024, 384] 589,952 True
40
+ │ │ └─RMSNorm: 3-3 [32, 1024, 384] [32, 1024, 384] 384 True
41
+ │ │ └─RMSNorm: 3-4 [32, 1024, 384] [32, 1024, 384] 384 True
42
+ │ │ └─SwiGLUMLP: 3-5 [32, 1024, 384] [32, 1024, 384] 884,736 True
43
+ │ │ └─RMSNorm: 3-6 [32, 1024, 384] [32, 1024, 384] 384 True
44
+ │ └─TransformerBlock: 2-2 [32, 1024, 384] [32, 1024, 384] -- True
45
+ │ │ └─RMSNorm: 3-7 [32, 1024, 384] [32, 1024, 384] 384 True
46
+ │ │ └─MultiHeadAttention: 3-8 [32, 1024, 384] [32, 1024, 384] 589,952 True
47
+ │ │ └─RMSNorm: 3-9 [32, 1024, 384] [32, 1024, 384] 384 True
48
+ │ │ └─RMSNorm: 3-10 [32, 1024, 384] [32, 1024, 384] 384 True
49
+ │ │ └─SwiGLUMLP: 3-11 [32, 1024, 384] [32, 1024, 384] 884,736 True
50
+ │ │ └─RMSNorm: 3-12 [32, 1024, 384] [32, 1024, 384] 384 True
51
+ │ └─TransformerBlock: 2-3 [32, 1024, 384] [32, 1024, 384] -- True
52
+ │ │ └─RMSNorm: 3-13 [32, 1024, 384] [32, 1024, 384] 384 True
53
+ │ │ └─MultiHeadAttention: 3-14 [32, 1024, 384] [32, 1024, 384] 589,952 True
54
+ │ │ └─RMSNorm: 3-15 [32, 1024, 384] [32, 1024, 384] 384 True
55
+ │ │ └─RMSNorm: 3-16 [32, 1024, 384] [32, 1024, 384] 384 True
56
+ │ │ └─SwiGLUMLP: 3-17 [32, 1024, 384] [32, 1024, 384] 884,736 True
57
+ │ │ └─RMSNorm: 3-18 [32, 1024, 384] [32, 1024, 384] 384 True
58
+ │ └─TransformerBlock: 2-4 [32, 1024, 384] [32, 1024, 384] -- True
59
+ │ │ └─RMSNorm: 3-19 [32, 1024, 384] [32, 1024, 384] 384 True
60
+ │ │ └─MultiHeadAttention: 3-20 [32, 1024, 384] [32, 1024, 384] 589,952 True
61
+ │ │ └─RMSNorm: 3-21 [32, 1024, 384] [32, 1024, 384] 384 True
62
+ │ │ └─RMSNorm: 3-22 [32, 1024, 384] [32, 1024, 384] 384 True
63
+ │ │ └─SwiGLUMLP: 3-23 [32, 1024, 384] [32, 1024, 384] 884,736 True
64
+ │ │ └─RMSNorm: 3-24 [32, 1024, 384] [32, 1024, 384] 384 True
65
+ │ └─TransformerBlock: 2-5 [32, 1024, 384] [32, 1024, 384] -- True
66
+ │ │ └─RMSNorm: 3-25 [32, 1024, 384] [32, 1024, 384] 384 True
67
+ │ │ └─MultiHeadAttention: 3-26 [32, 1024, 384] [32, 1024, 384] 589,952 True
68
+ │ │ └─RMSNorm: 3-27 [32, 1024, 384] [32, 1024, 384] 384 True
69
+ │ │ └─RMSNorm: 3-28 [32, 1024, 384] [32, 1024, 384] 384 True
70
+ │ │ └─SwiGLUMLP: 3-29 [32, 1024, 384] [32, 1024, 384] 884,736 True
71
+ │ │ └─RMSNorm: 3-30 [32, 1024, 384] [32, 1024, 384] 384 True
72
+ │ └─TransformerBlock: 2-6 [32, 1024, 384] [32, 1024, 384] -- True
73
+ │ │ └─RMSNorm: 3-31 [32, 1024, 384] [32, 1024, 384] 384 True
74
+ │ │ └─MultiHeadAttention: 3-32 [32, 1024, 384] [32, 1024, 384] 589,952 True
75
+ │ │ └─RMSNorm: 3-33 [32, 1024, 384] [32, 1024, 384] 384 True
76
+ │ │ └─RMSNorm: 3-34 [32, 1024, 384] [32, 1024, 384] 384 True
77
+ │ │ └─SwiGLUMLP: 3-35 [32, 1024, 384] [32, 1024, 384] 884,736 True
78
+ │ │ └─RMSNorm: 3-36 [32, 1024, 384] [32, 1024, 384] 384 True
79
+ │ └─TransformerBlock: 2-7 [32, 1024, 384] [32, 1024, 384] -- True
80
+ │ │ └─RMSNorm: 3-37 [32, 1024, 384] [32, 1024, 384] 384 True
81
+ │ │ └─MultiHeadAttention: 3-38 [32, 1024, 384] [32, 1024, 384] 589,952 True
82
+ │ │ └─RMSNorm: 3-39 [32, 1024, 384] [32, 1024, 384] 384 True
83
+ │ │ └─RMSNorm: 3-40 [32, 1024, 384] [32, 1024, 384] 384 True
84
+ │ │ └─SwiGLUMLP: 3-41 [32, 1024, 384] [32, 1024, 384] 884,736 True
85
+ │ │ └─RMSNorm: 3-42 [32, 1024, 384] [32, 1024, 384] 384 True
86
+ │ └─TransformerBlock: 2-8 [32, 1024, 384] [32, 1024, 384] -- True
87
+ │ │ └─RMSNorm: 3-43 [32, 1024, 384] [32, 1024, 384] 384 True
88
+ │ │ └─MultiHeadAttention: 3-44 [32, 1024, 384] [32, 1024, 384] 589,952 True
89
+ │ │ └─RMSNorm: 3-45 [32, 1024, 384] [32, 1024, 384] 384 True
90
+ │ │ └─RMSNorm: 3-46 [32, 1024, 384] [32, 1024, 384] 384 True
91
+ │ │ └─SwiGLUMLP: 3-47 [32, 1024, 384] [32, 1024, 384] 884,736 True
92
+ │ │ └─RMSNorm: 3-48 [32, 1024, 384] [32, 1024, 384] 384 True
93
+ │ └─TransformerBlock: 2-9 [32, 1024, 384] [32, 1024, 384] -- True
94
+ │ │ └─RMSNorm: 3-49 [32, 1024, 384] [32, 1024, 384] 384 True
95
+ │ │ └─MultiHeadAttention: 3-50 [32, 1024, 384] [32, 1024, 384] 589,952 True
96
+ │ │ └─RMSNorm: 3-51 [32, 1024, 384] [32, 1024, 384] 384 True
97
+ │ │ └─RMSNorm: 3-52 [32, 1024, 384] [32, 1024, 384] 384 True
98
+ │ │ └─SwiGLUMLP: 3-53 [32, 1024, 384] [32, 1024, 384] 884,736 True
99
+ │ │ └─RMSNorm: 3-54 [32, 1024, 384] [32, 1024, 384] 384 True
100
+ │ └─TransformerBlock: 2-10 [32, 1024, 384] [32, 1024, 384] -- True
101
+ │ │ └─RMSNorm: 3-55 [32, 1024, 384] [32, 1024, 384] 384 True
102
+ │ │ └─MultiHeadAttention: 3-56 [32, 1024, 384] [32, 1024, 384] 589,952 True
103
+ │ │ └─RMSNorm: 3-57 [32, 1024, 384] [32, 1024, 384] 384 True
104
+ │ │ └─RMSNorm: 3-58 [32, 1024, 384] [32, 1024, 384] 384 True
105
+ │ │ └─SwiGLUMLP: 3-59 [32, 1024, 384] [32, 1024, 384] 884,736 True
106
+ │ │ └─RMSNorm: 3-60 [32, 1024, 384] [32, 1024, 384] 384 True
107
+ │ └─TransformerBlock: 2-11 [32, 1024, 384] [32, 1024, 384] -- True
108
+ │ │ └─RMSNorm: 3-61 [32, 1024, 384] [32, 1024, 384] 384 True
109
+ │ │ └─MultiHeadAttention: 3-62 [32, 1024, 384] [32, 1024, 384] 589,952 True
110
+ │ │ └─RMSNorm: 3-63 [32, 1024, 384] [32, 1024, 384] 384 True
111
+ │ │ └─RMSNorm: 3-64 [32, 1024, 384] [32, 1024, 384] 384 True
112
+ │ │ └─SwiGLUMLP: 3-65 [32, 1024, 384] [32, 1024, 384] 884,736 True
113
+ │ │ └─RMSNorm: 3-66 [32, 1024, 384] [32, 1024, 384] 384 True
114
+ │ └─TransformerBlock: 2-12 [32, 1024, 384] [32, 1024, 384] -- True
115
+ │ │ └─RMSNorm: 3-67 [32, 1024, 384] [32, 1024, 384] 384 True
116
+ │ │ └─MultiHeadAttention: 3-68 [32, 1024, 384] [32, 1024, 384] 589,952 True
117
+ │ │ └─RMSNorm: 3-69 [32, 1024, 384] [32, 1024, 384] 384 True
118
+ │ │ └─RMSNorm: 3-70 [32, 1024, 384] [32, 1024, 384] 384 True
119
+ │ │ └─SwiGLUMLP: 3-71 [32, 1024, 384] [32, 1024, 384] 884,736 True
120
+ │ │ └─RMSNorm: 3-72 [32, 1024, 384] [32, 1024, 384] 384 True
121
+ │ └─TransformerBlock: 2-13 [32, 1024, 384] [32, 1024, 384] -- True
122
+ │ │ └─RMSNorm: 3-73 [32, 1024, 384] [32, 1024, 384] 384 True
123
+ │ │ └─MultiHeadAttention: 3-74 [32, 1024, 384] [32, 1024, 384] 589,952 True
124
+ │ │ └─RMSNorm: 3-75 [32, 1024, 384] [32, 1024, 384] 384 True
125
+ │ │ └─RMSNorm: 3-76 [32, 1024, 384] [32, 1024, 384] 384 True
126
+ │ │ └─SwiGLUMLP: 3-77 [32, 1024, 384] [32, 1024, 384] 884,736 True
127
+ │ │ └─RMSNorm: 3-78 [32, 1024, 384] [32, 1024, 384] 384 True
128
+ │ └─TransformerBlock: 2-14 [32, 1024, 384] [32, 1024, 384] -- True
129
+ │ │ └─RMSNorm: 3-79 [32, 1024, 384] [32, 1024, 384] 384 True
130
+ │ │ └─MultiHeadAttention: 3-80 [32, 1024, 384] [32, 1024, 384] 589,952 True
131
+ │ │ └─RMSNorm: 3-81 [32, 1024, 384] [32, 1024, 384] 384 True
132
+ │ │ └─RMSNorm: 3-82 [32, 1024, 384] [32, 1024, 384] 384 True
133
+ │ │ └─SwiGLUMLP: 3-83 [32, 1024, 384] [32, 1024, 384] 884,736 True
134
+ │ │ └─RMSNorm: 3-84 [32, 1024, 384] [32, 1024, 384] 384 True
135
+ ├─RMSNorm: 1-4 [32, 1024, 384] [32, 1024, 384] 384 True
136
+ ├─Linear: 1-5 [32, 1024, 384] [32, 1024, 50258] 19,299,072 True
137
+ ============================================================================================================================================
138
+ Total params: 59,265,664
139
+ Trainable params: 59,265,664
140
+ Non-trainable params: 0
141
+ Total mult-adds (Units.GIGABYTES): 1.90
142
+ ============================================================================================================================================
143
+ Input size (MB): 0.26
144
+ Forward/backward pass size (MB): 34515.45
145
+ Params size (MB): 237.06
146
+ Estimated Total Size (MB): 34752.78
147
+ ============================================================================================================================================
148
+ ============================================================
149
+ 2026-04-08 12:41:41 - INFO - parrotllm.training - Starting evaluation...
150
+ 2026-04-08 12:41:49 - INFO - parrotllm.training - Initial val: loss=10.8939, ppl=53846.40
151
+ 2026-04-08 12:42:05 - INFO - parrotllm.training -
152
+ ============================================================
153
+ Starting training...
154
+ Steps per epoch (approx): 1939
155
+ Max steps: 3052
156
+ ============================================================
157
+ 2026-04-08 12:42:05 - INFO - parrotllm.training - LR schedule=wsd | decay_ratio=0.179 | z_loss_coeff=0.000275
158
+ 2026-04-08 12:42:27 - INFO - parrotllm.training - step 1 | epoch 0 | loss 10.9283 | lr 4.28e-06 | grad 6.8122
159
+ 2026-04-08 12:42:30 - INFO - parrotllm.training - step 10 | epoch 0 | loss 10.2616 | lr 4.28e-05 | grad 1.8270
160
+ 2026-04-08 12:42:33 - INFO - parrotllm.training - step 20 | epoch 0 | loss 9.8953 | lr 8.56e-05 | grad 1.6945
161
+ 2026-04-08 12:42:36 - INFO - parrotllm.training - step 30 | epoch 0 | loss 9.2911 | lr 1.28e-04 | grad 2.0493
162
+ 2026-04-08 12:42:39 - INFO - parrotllm.training - step 40 | epoch 0 | loss 8.6841 | lr 1.71e-04 | grad 1.5425
163
+ 2026-04-08 12:42:42 - INFO - parrotllm.training - step 50 | epoch 0 | loss 7.9912 | lr 2.14e-04 | grad 1.3772
164
+ 2026-04-08 12:42:45 - INFO - parrotllm.training - step 60 | epoch 0 | loss 7.5500 | lr 2.57e-04 | grad 0.9932
165
+ 2026-04-08 12:42:48 - INFO - parrotllm.training - step 70 | epoch 0 | loss 7.2957 | lr 3.00e-04 | grad 0.8042
166
+ 2026-04-08 12:42:52 - INFO - parrotllm.training - step 80 | epoch 0 | loss 7.0405 | lr 3.42e-04 | grad 0.6572
167
+ 2026-04-08 12:42:55 - INFO - parrotllm.training - step 90 | epoch 0 | loss 6.9090 | lr 3.85e-04 | grad 0.7028
168
+ 2026-04-08 12:42:58 - INFO - parrotllm.training - step 100 | epoch 0 | loss 6.8356 | lr 4.28e-04 | grad 0.4364
169
+ 2026-04-08 12:42:58 - INFO - parrotllm.training - Starting evaluation...
170
+ 2026-04-08 12:42:58 - INFO - parrotllm.training - ------------------------------------------------------------
171
+ 2026-04-08 12:43:42 - INFO - parrotllm.training - Train: loss=6.8356, ppl=930.42
172
+ 2026-04-08 12:43:42 - INFO - parrotllm.training - Val: loss=6.7630, ppl=865.27
173
+ 2026-04-08 12:43:42 - INFO - parrotllm.training - ** New best validation loss! **
174
+ 2026-04-08 12:43:43 - INFO - parrotllm.training - Saved best checkpoint: /home/user/code/ParrotLLM/runs/big_run/exp_c/run_20260408_124138/checkpoints/best_loss_6p7630_epoch_0000_step_0000100.pt
175
+ 2026-04-08 12:43:44 - INFO - parrotllm.training - ------------------------------------------------------------
176
+ 2026-04-08 12:43:47 - INFO - parrotllm.training - step 110 | epoch 0 | loss 6.7452 | lr 4.71e-04 | grad 0.9823
177
+ 2026-04-08 12:43:50 - INFO - parrotllm.training - step 120 | epoch 0 | loss 6.5647 | lr 5.13e-04 | grad 0.8019
178
+ 2026-04-08 12:43:54 - INFO - parrotllm.training - step 130 | epoch 0 | loss 6.4961 | lr 5.56e-04 | grad 0.6392
179
+ 2026-04-08 12:43:57 - INFO - parrotllm.training - step 140 | epoch 0 | loss 6.2378 | lr 5.99e-04 | grad 0.8438
180
+ 2026-04-08 12:44:00 - INFO - parrotllm.training - step 150 | epoch 0 | loss 6.2534 | lr 6.42e-04 | grad 0.5381
181
+ 2026-04-08 12:44:03 - INFO - parrotllm.training - step 160 | epoch 0 | loss 6.2198 | lr 6.85e-04 | grad 0.6169
182
+ 2026-04-08 12:44:06 - INFO - parrotllm.training - step 170 | epoch 0 | loss 6.2544 | lr 7.27e-04 | grad 1.2573
183
+ 2026-04-08 12:44:09 - INFO - parrotllm.training - step 180 | epoch 0 | loss 6.0693 | lr 7.70e-04 | grad 1.0304
184
+ 2026-04-08 12:44:12 - INFO - parrotllm.training - step 190 | epoch 0 | loss 5.9096 | lr 8.13e-04 | grad 0.4648
185
+ 2026-04-08 12:44:15 - INFO - parrotllm.training - step 200 | epoch 0 | loss 5.9346 | lr 8.56e-04 | grad 1.1069
186
+ 2026-04-08 12:44:15 - INFO - parrotllm.training - Starting evaluation...
187
+ 2026-04-08 12:44:15 - INFO - parrotllm.training - ------------------------------------------------------------
188
+ 2026-04-08 12:44:21 - INFO - parrotllm.training - Train: loss=5.9346, ppl=377.87
189
+ 2026-04-08 12:44:21 - INFO - parrotllm.training - Val: loss=5.9015, ppl=365.60
190
+ 2026-04-08 12:44:21 - INFO - parrotllm.training - ** New best validation loss! **
191
+ 2026-04-08 12:44:21 - INFO - parrotllm.training - Saved best checkpoint: /home/user/code/ParrotLLM/runs/big_run/exp_c/run_20260408_124138/checkpoints/best_loss_5p9015_epoch_0000_step_0000200.pt
192
+ 2026-04-08 12:44:22 - INFO - parrotllm.training - ------------------------------------------------------------
193
+ 2026-04-08 12:44:25 - INFO - parrotllm.training - step 210 | epoch 0 | loss 5.7735 | lr 8.99e-04 | grad 0.5488
194
+ 2026-04-08 12:44:28 - INFO - parrotllm.training - step 220 | epoch 0 | loss 5.8331 | lr 9.41e-04 | grad 0.9377
195
+ 2026-04-08 12:44:31 - INFO - parrotllm.training - step 230 | epoch 0 | loss 5.8403 | lr 9.84e-04 | grad 1.2072
196
+ 2026-04-08 12:44:34 - INFO - parrotllm.training - step 240 | epoch 0 | loss 5.6031 | lr 1.03e-03 | grad 0.3195
197
+ 2026-04-08 12:44:37 - INFO - parrotllm.training - step 250 | epoch 0 | loss 5.5036 | lr 1.07e-03 | grad 0.4908
198
+ 2026-04-08 12:44:38 - INFO - parrotllm.training - Saved checkpoint: /home/user/code/ParrotLLM/runs/big_run/exp_c/run_20260408_124138/checkpoints/last_epoch_0000_step_0000250.pt
199
+ 2026-04-08 12:44:42 - INFO - parrotllm.training - step 260 | epoch 0 | loss 5.5156 | lr 1.11e-03 | grad 0.5607
200
+ 2026-04-08 12:44:46 - INFO - parrotllm.training - step 270 | epoch 0 | loss 5.5554 | lr 1.16e-03 | grad 0.4806
201
+ 2026-04-08 12:44:49 - INFO - parrotllm.training - step 280 | epoch 0 | loss 5.4411 | lr 1.20e-03 | grad 0.4974
202
+ 2026-04-08 12:44:52 - INFO - parrotllm.training - step 290 | epoch 0 | loss 5.4694 | lr 1.24e-03 | grad 1.0310
203
+ 2026-04-08 12:44:55 - INFO - parrotllm.training - step 300 | epoch 0 | loss 5.4289 | lr 1.28e-03 | grad 0.4870
204
+ 2026-04-08 12:44:55 - INFO - parrotllm.training - Starting evaluation...
205
+ 2026-04-08 12:44:55 - INFO - parrotllm.training - ------------------------------------------------------------
206
+ 2026-04-08 12:45:00 - INFO - parrotllm.training - Train: loss=5.4289, ppl=227.90
207
+ 2026-04-08 12:45:00 - INFO - parrotllm.training - Val: loss=5.3999, ppl=221.38
208
+ 2026-04-08 12:45:00 - INFO - parrotllm.training - ** New best validation loss! **
209
+ 2026-04-08 12:45:01 - INFO - parrotllm.training - Saved best checkpoint: /home/user/code/ParrotLLM/runs/big_run/exp_c/run_20260408_124138/checkpoints/best_loss_5p3999_epoch_0000_step_0000300.pt
210
+ 2026-04-08 12:45:02 - INFO - parrotllm.training - ------------------------------------------------------------
211
+ 2026-04-08 12:45:05 - INFO - parrotllm.training - step 310 | epoch 0 | loss 5.3706 | lr 1.30e-03 | grad 0.3568
212
+ 2026-04-08 12:45:08 - INFO - parrotllm.training - step 320 | epoch 0 | loss 5.2355 | lr 1.30e-03 | grad 0.3656
213
+ 2026-04-08 12:45:11 - INFO - parrotllm.training - step 330 | epoch 0 | loss 5.1612 | lr 1.30e-03 | grad 0.4844
214
+ 2026-04-08 12:45:14 - INFO - parrotllm.training - step 340 | epoch 0 | loss 5.2862 | lr 1.30e-03 | grad 0.4443
215
+ 2026-04-08 12:45:17 - INFO - parrotllm.training - step 350 | epoch 0 | loss 4.9924 | lr 1.30e-03 | grad 0.2914
216
+ 2026-04-08 12:45:20 - INFO - parrotllm.training - step 360 | epoch 0 | loss 5.2429 | lr 1.30e-03 | grad 0.7404
217
+ 2026-04-08 12:45:24 - INFO - parrotllm.training - step 370 | epoch 0 | loss 5.0984 | lr 1.30e-03 | grad 0.4491
218
+ 2026-04-08 12:45:27 - INFO - parrotllm.training - step 380 | epoch 0 | loss 5.0499 | lr 1.30e-03 | grad 0.4170
219
+ 2026-04-08 12:45:30 - INFO - parrotllm.training - step 390 | epoch 0 | loss 4.9054 | lr 1.30e-03 | grad 0.2777
220
+ 2026-04-08 12:45:33 - INFO - parrotllm.training - step 400 | epoch 0 | loss 5.0143 | lr 1.30e-03 | grad 0.4620
221
+ 2026-04-08 12:45:33 - INFO - parrotllm.training - Starting evaluation...
222
+ 2026-04-08 12:45:33 - INFO - parrotllm.training - ------------------------------------------------------------
223
+ 2026-04-08 12:45:38 - INFO - parrotllm.training - Train: loss=5.0143, ppl=150.55
224
+ 2026-04-08 12:45:38 - INFO - parrotllm.training - Val: loss=4.9547, ppl=141.85
225
+ 2026-04-08 12:45:38 - INFO - parrotllm.training - ** New best validation loss! **
226
+ 2026-04-08 12:45:39 - INFO - parrotllm.training - Saved best checkpoint: /home/user/code/ParrotLLM/runs/big_run/exp_c/run_20260408_124138/checkpoints/best_loss_4p9547_epoch_0000_step_0000400.pt
227
+ 2026-04-08 12:45:40 - INFO - parrotllm.training - ------------------------------------------------------------
228
+ 2026-04-08 12:45:43 - INFO - parrotllm.training - step 410 | epoch 0 | loss 4.9213 | lr 1.30e-03 | grad 0.3111
229
+ 2026-04-08 12:45:46 - INFO - parrotllm.training - step 420 | epoch 0 | loss 4.9263 | lr 1.30e-03 | grad 0.2881
230
+ 2026-04-08 12:45:49 - INFO - parrotllm.training - step 430 | epoch 0 | loss 4.8843 | lr 1.30e-03 | grad 0.4029
231
+ 2026-04-08 12:45:52 - INFO - parrotllm.training - step 440 | epoch 0 | loss 4.7801 | lr 1.30e-03 | grad 0.3063
232
+ 2026-04-08 12:45:55 - INFO - parrotllm.training - step 450 | epoch 0 | loss 4.7493 | lr 1.30e-03 | grad 0.3415
233
+ 2026-04-08 12:45:58 - INFO - parrotllm.training - step 460 | epoch 0 | loss 4.7937 | lr 1.30e-03 | grad 0.3799
234
+ 2026-04-08 12:46:01 - INFO - parrotllm.training - step 470 | epoch 0 | loss 4.7403 | lr 1.30e-03 | grad 0.6610
235
+ 2026-04-08 12:46:04 - INFO - parrotllm.training - step 480 | epoch 0 | loss 4.7418 | lr 1.30e-03 | grad 0.3100
236
+ 2026-04-08 12:46:07 - INFO - parrotllm.training - step 490 | epoch 0 | loss 4.7975 | lr 1.30e-03 | grad 0.4635
237
+ 2026-04-08 12:46:11 - INFO - parrotllm.training - step 500 | epoch 0 | loss 4.6381 | lr 1.30e-03 | grad 0.3594
238
+ 2026-04-08 12:46:11 - INFO - parrotllm.training - Starting evaluation...
239
+ 2026-04-08 12:46:11 - INFO - parrotllm.training - ------------------------------------------------------------
240
+ 2026-04-08 12:46:16 - INFO - parrotllm.training - Train: loss=4.6381, ppl=103.34
241
+ 2026-04-08 12:46:16 - INFO - parrotllm.training - Val: loss=4.6275, ppl=102.26
242
+ 2026-04-08 12:46:16 - INFO - parrotllm.training - ** New best validation loss! **
243
+ 2026-04-08 12:46:17 - INFO - parrotllm.training - Saved best checkpoint: /home/user/code/ParrotLLM/runs/big_run/exp_c/run_20260408_124138/checkpoints/best_loss_4p6275_epoch_0000_step_0000500.pt
244
+ 2026-04-08 12:46:18 - INFO - parrotllm.training - ------------------------------------------------------------
245
+ 2026-04-08 12:46:19 - INFO - parrotllm.training - Saved checkpoint: /home/user/code/ParrotLLM/runs/big_run/exp_c/run_20260408_124138/checkpoints/last_epoch_0000_step_0000500.pt
246
+ 2026-04-08 12:46:23 - INFO - parrotllm.training - step 510 | epoch 0 | loss 4.6129 | lr 1.30e-03 | grad 0.3387
247
+ 2026-04-08 12:46:26 - INFO - parrotllm.training - step 520 | epoch 0 | loss 4.6002 | lr 1.30e-03 | grad 0.2965
248
+ 2026-04-08 12:46:29 - INFO - parrotllm.training - step 530 | epoch 0 | loss 4.6989 | lr 1.30e-03 | grad 0.4270
249
+ 2026-04-08 12:46:32 - INFO - parrotllm.training - step 540 | epoch 0 | loss 4.4593 | lr 1.30e-03 | grad 0.3768
250
+ 2026-04-08 12:46:35 - INFO - parrotllm.training - step 550 | epoch 0 | loss 4.4864 | lr 1.30e-03 | grad 0.3919
251
+ 2026-04-08 12:46:38 - INFO - parrotllm.training - step 560 | epoch 0 | loss 4.5896 | lr 1.30e-03 | grad 0.3289
252
+ 2026-04-08 12:46:41 - INFO - parrotllm.training - step 570 | epoch 0 | loss 4.5558 | lr 1.30e-03 | grad 0.5545
253
+ 2026-04-08 12:46:44 - INFO - parrotllm.training - step 580 | epoch 0 | loss 4.4624 | lr 1.30e-03 | grad 0.2772
254
+ 2026-04-08 12:46:47 - INFO - parrotllm.training - step 590 | epoch 0 | loss 4.4730 | lr 1.30e-03 | grad 0.2939
255
+ 2026-04-08 12:46:50 - INFO - parrotllm.training - step 600 | epoch 0 | loss 4.5336 | lr 1.30e-03 | grad 0.2799
256
+ 2026-04-08 12:46:50 - INFO - parrotllm.training - Starting evaluation...
257
+ 2026-04-08 12:46:50 - INFO - parrotllm.training - ------------------------------------------------------------
258
+ 2026-04-08 12:46:56 - INFO - parrotllm.training - Train: loss=4.5336, ppl=93.09
259
+ 2026-04-08 12:46:56 - INFO - parrotllm.training - Val: loss=4.3700, ppl=79.05
260
+ 2026-04-08 12:46:56 - INFO - parrotllm.training - ** New best validation loss! **
261
+ 2026-04-08 12:46:57 - INFO - parrotllm.training - Saved best checkpoint: /home/user/code/ParrotLLM/runs/big_run/exp_c/run_20260408_124138/checkpoints/best_loss_4p3700_epoch_0000_step_0000600.pt
262
+ 2026-04-08 12:46:58 - INFO - parrotllm.training - ------------------------------------------------------------
263
+ 2026-04-08 12:47:01 - INFO - parrotllm.training - step 610 | epoch 0 | loss 4.4782 | lr 1.30e-03 | grad 0.4029
264
+ 2026-04-08 12:47:04 - INFO - parrotllm.training - step 620 | epoch 0 | loss 4.4498 | lr 1.30e-03 | grad 0.3437
265
+ 2026-04-08 12:47:07 - INFO - parrotllm.training - step 630 | epoch 0 | loss 4.3908 | lr 1.30e-03 | grad 0.4935
266
+ 2026-04-08 12:47:10 - INFO - parrotllm.training - step 640 | epoch 0 | loss 4.3158 | lr 1.30e-03 | grad 0.2801
267
+ 2026-04-08 12:47:13 - INFO - parrotllm.training - step 650 | epoch 0 | loss 4.3310 | lr 1.30e-03 | grad 0.3586
268
+ 2026-04-08 12:47:16 - INFO - parrotllm.training - step 660 | epoch 0 | loss 4.1952 | lr 1.30e-03 | grad 0.3386
269
+ 2026-04-08 12:47:19 - INFO - parrotllm.training - step 670 | epoch 0 | loss 4.3220 | lr 1.30e-03 | grad 0.2795
270
+ 2026-04-08 12:47:22 - INFO - parrotllm.training - step 680 | epoch 0 | loss 4.3062 | lr 1.30e-03 | grad 0.3723
271
+ 2026-04-08 12:47:26 - INFO - parrotllm.training - step 690 | epoch 0 | loss 4.2446 | lr 1.30e-03 | grad 0.3151
272
+ 2026-04-08 12:47:29 - INFO - parrotllm.training - step 700 | epoch 0 | loss 4.2909 | lr 1.30e-03 | grad 0.4253
273
+ 2026-04-08 12:47:29 - INFO - parrotllm.training - Starting evaluation...
274
+ 2026-04-08 12:47:29 - INFO - parrotllm.training - ------------------------------------------------------------
275
+ 2026-04-08 12:47:34 - INFO - parrotllm.training - Train: loss=4.2909, ppl=73.03
276
+ 2026-04-08 12:47:34 - INFO - parrotllm.training - Val: loss=4.2423, ppl=69.57
277
+ 2026-04-08 12:47:34 - INFO - parrotllm.training - ** New best validation loss! **
278
+ 2026-04-08 12:47:35 - INFO - parrotllm.training - Saved best checkpoint: /home/user/code/ParrotLLM/runs/big_run/exp_c/run_20260408_124138/checkpoints/best_loss_4p2423_epoch_0000_step_0000700.pt
279
+ 2026-04-08 12:47:36 - INFO - parrotllm.training - ------------------------------------------------------------
280
+ 2026-04-08 12:47:39 - INFO - parrotllm.training - step 710 | epoch 0 | loss 4.2638 | lr 1.30e-03 | grad 0.2893
281
+ 2026-04-08 12:47:42 - INFO - parrotllm.training - step 720 | epoch 0 | loss 4.3053 | lr 1.30e-03 | grad 0.3133
282
+ 2026-04-08 12:47:45 - INFO - parrotllm.training - step 730 | epoch 0 | loss 4.2717 | lr 1.30e-03 | grad 0.3125
283
+ 2026-04-08 12:47:48 - INFO - parrotllm.training - step 740 | epoch 0 | loss 4.3364 | lr 1.30e-03 | grad 0.3499
284
+ 2026-04-08 12:47:51 - INFO - parrotllm.training - step 750 | epoch 0 | loss 4.2178 | lr 1.30e-03 | grad 0.4627
285
+ 2026-04-08 12:47:52 - INFO - parrotllm.training - Saved checkpoint: /home/user/code/ParrotLLM/runs/big_run/exp_c/run_20260408_124138/checkpoints/last_epoch_0000_step_0000750.pt
286
+ 2026-04-08 12:47:56 - INFO - parrotllm.training - step 760 | epoch 0 | loss 4.1247 | lr 1.30e-03 | grad 0.2749
287
+ 2026-04-08 12:48:00 - INFO - parrotllm.training - step 770 | epoch 0 | loss 4.3647 | lr 1.30e-03 | grad 0.2520
288
+ 2026-04-08 12:48:03 - INFO - parrotllm.training - step 780 | epoch 0 | loss 4.2694 | lr 1.30e-03 | grad 0.2739
289
+ 2026-04-08 12:48:06 - INFO - parrotllm.training - step 790 | epoch 0 | loss 4.2489 | lr 1.30e-03 | grad 0.3426
290
+ 2026-04-08 12:48:09 - INFO - parrotllm.training - step 800 | epoch 0 | loss 4.2317 | lr 1.30e-03 | grad 0.3093
291
+ 2026-04-08 12:48:09 - INFO - parrotllm.training - Starting evaluation...
292
+ 2026-04-08 12:48:09 - INFO - parrotllm.training - ------------------------------------------------------------
293
+ 2026-04-08 12:48:15 - INFO - parrotllm.training - Train: loss=4.2317, ppl=68.83
294
+ 2026-04-08 12:48:15 - INFO - parrotllm.training - Val: loss=4.1199, ppl=61.55
295
+ 2026-04-08 12:48:15 - INFO - parrotllm.training - ** New best validation loss! **
296
+ 2026-04-08 12:48:15 - INFO - parrotllm.training - Saved best checkpoint: /home/user/code/ParrotLLM/runs/big_run/exp_c/run_20260408_124138/checkpoints/best_loss_4p1199_epoch_0000_step_0000800.pt
297
+ 2026-04-08 12:48:16 - INFO - parrotllm.training - ------------------------------------------------------------
298
+ 2026-04-08 12:48:19 - INFO - parrotllm.training - step 810 | epoch 0 | loss 4.1372 | lr 1.30e-03 | grad 0.2648
299
+ 2026-04-08 12:48:22 - INFO - parrotllm.training - step 820 | epoch 0 | loss 4.1945 | lr 1.30e-03 | grad 0.5349
300
+ 2026-04-08 12:48:25 - INFO - parrotllm.training - step 830 | epoch 0 | loss 4.1505 | lr 1.30e-03 | grad 0.2663
301
+ 2026-04-08 12:48:28 - INFO - parrotllm.training - step 840 | epoch 0 | loss 4.2378 | lr 1.30e-03 | grad 0.3815
302
+ 2026-04-08 12:48:32 - INFO - parrotllm.training - step 850 | epoch 0 | loss 4.1529 | lr 1.30e-03 | grad 0.2902
303
+ 2026-04-08 12:48:35 - INFO - parrotllm.training - step 860 | epoch 0 | loss 4.1264 | lr 1.30e-03 | grad 0.3139
304
+ 2026-04-08 12:48:38 - INFO - parrotllm.training - step 870 | epoch 0 | loss 4.1175 | lr 1.30e-03 | grad 0.3062
305
+ 2026-04-08 12:48:41 - INFO - parrotllm.training - step 880 | epoch 0 | loss 4.0828 | lr 1.30e-03 | grad 0.2848
306
+ 2026-04-08 12:48:44 - INFO - parrotllm.training - step 890 | epoch 0 | loss 4.0311 | lr 1.30e-03 | grad 0.3449
307
+ 2026-04-08 12:48:47 - INFO - parrotllm.training - step 900 | epoch 0 | loss 4.0856 | lr 1.30e-03 | grad 0.2898
308
+ 2026-04-08 12:48:47 - INFO - parrotllm.training - Starting evaluation...
309
+ 2026-04-08 12:48:47 - INFO - parrotllm.training - ------------------------------------------------------------
310
+ 2026-04-08 12:48:53 - INFO - parrotllm.training - Train: loss=4.0856, ppl=59.47
311
+ 2026-04-08 12:48:53 - INFO - parrotllm.training - Val: loss=4.0408, ppl=56.87
312
+ 2026-04-08 12:48:53 - INFO - parrotllm.training - ** New best validation loss! **
313
+ 2026-04-08 12:48:53 - INFO - parrotllm.training - Saved best checkpoint: /home/user/code/ParrotLLM/runs/big_run/exp_c/run_20260408_124138/checkpoints/best_loss_4p0408_epoch_0000_step_0000900.pt
314
+ 2026-04-08 12:48:54 - INFO - parrotllm.training - ------------------------------------------------------------
315
+ 2026-04-08 12:48:57 - INFO - parrotllm.training - step 910 | epoch 0 | loss 4.0652 | lr 1.30e-03 | grad 0.2460
316
+ 2026-04-08 12:49:00 - INFO - parrotllm.training - step 920 | epoch 0 | loss 4.1638 | lr 1.30e-03 | grad 0.3769
317
+ 2026-04-08 12:49:03 - INFO - parrotllm.training - step 930 | epoch 0 | loss 4.0077 | lr 1.30e-03 | grad 0.2636
318
+ 2026-04-08 12:49:06 - INFO - parrotllm.training - step 940 | epoch 0 | loss 4.0963 | lr 1.30e-03 | grad 0.2925
319
+ 2026-04-08 12:49:09 - INFO - parrotllm.training - step 950 | epoch 0 | loss 4.0018 | lr 1.30e-03 | grad 0.2907
320
+ 2026-04-08 12:49:12 - INFO - parrotllm.training - step 960 | epoch 0 | loss 4.0079 | lr 1.30e-03 | grad 0.3901
321
+ 2026-04-08 12:49:15 - INFO - parrotllm.training - step 970 | epoch 0 | loss 4.0677 | lr 1.30e-03 | grad 0.2483
322
+ 2026-04-08 12:49:19 - INFO - parrotllm.training - step 980 | epoch 0 | loss 4.1012 | lr 1.30e-03 | grad 0.2528
323
+ 2026-04-08 12:49:22 - INFO - parrotllm.training - step 990 | epoch 0 | loss 4.0414 | lr 1.30e-03 | grad 0.2952
324
+ 2026-04-08 12:49:25 - INFO - parrotllm.training - step 1000 | epoch 0 | loss 4.0708 | lr 1.30e-03 | grad 0.3298
325
+ 2026-04-08 12:49:25 - INFO - parrotllm.training - Starting evaluation...
326
+ 2026-04-08 12:49:25 - INFO - parrotllm.training - ------------------------------------------------------------
327
+ 2026-04-08 12:49:30 - INFO - parrotllm.training - Train: loss=4.0708, ppl=58.61
328
+ 2026-04-08 12:49:30 - INFO - parrotllm.training - Val: loss=3.9807, ppl=53.56
329
+ 2026-04-08 12:49:30 - INFO - parrotllm.training - ** New best validation loss! **
330
+ 2026-04-08 12:49:31 - INFO - parrotllm.training - Saved best checkpoint: /home/user/code/ParrotLLM/runs/big_run/exp_c/run_20260408_124138/checkpoints/best_loss_3p9807_epoch_0000_step_0001000.pt
331
+ 2026-04-08 12:49:32 - INFO - parrotllm.training - ------------------------------------------------------------
332
+ 2026-04-08 12:49:33 - INFO - parrotllm.training - Saved checkpoint: /home/user/code/ParrotLLM/runs/big_run/exp_c/run_20260408_124138/checkpoints/last_epoch_0000_step_0001000.pt
333
+ 2026-04-08 12:49:37 - INFO - parrotllm.training - step 1010 | epoch 0 | loss 4.0479 | lr 1.30e-03 | grad 0.2914
334
+ 2026-04-08 12:49:40 - INFO - parrotllm.training - step 1020 | epoch 0 | loss 3.9430 | lr 1.30e-03 | grad 0.2406
335
+ 2026-04-08 12:49:43 - INFO - parrotllm.training - step 1030 | epoch 0 | loss 3.8865 | lr 1.30e-03 | grad 0.3132
336
+ 2026-04-08 12:49:46 - INFO - parrotllm.training - step 1040 | epoch 0 | loss 4.0676 | lr 1.30e-03 | grad 0.2750
337
+ 2026-04-08 12:49:49 - INFO - parrotllm.training - step 1050 | epoch 0 | loss 3.9435 | lr 1.30e-03 | grad 0.3234
338
+ 2026-04-08 12:49:52 - INFO - parrotllm.training - step 1060 | epoch 0 | loss 4.0356 | lr 1.30e-03 | grad 0.2918
339
+ 2026-04-08 12:49:55 - INFO - parrotllm.training - step 1070 | epoch 0 | loss 3.9681 | lr 1.30e-03 | grad 0.3066
340
+ 2026-04-08 12:49:58 - INFO - parrotllm.training - step 1080 | epoch 0 | loss 4.0446 | lr 1.30e-03 | grad 0.2816
341
+ 2026-04-08 12:50:01 - INFO - parrotllm.training - step 1090 | epoch 0 | loss 3.9623 | lr 1.30e-03 | grad 0.2621
342
+ 2026-04-08 12:50:04 - INFO - parrotllm.training - step 1100 | epoch 0 | loss 3.9303 | lr 1.30e-03 | grad 0.3440
343
+ 2026-04-08 12:50:04 - INFO - parrotllm.training - Starting evaluation...
344
+ 2026-04-08 12:50:04 - INFO - parrotllm.training - ------------------------------------------------------------
345
+ 2026-04-08 12:50:10 - INFO - parrotllm.training - Train: loss=3.9303, ppl=50.92
346
+ 2026-04-08 12:50:10 - INFO - parrotllm.training - Val: loss=3.9380, ppl=51.32
347
+ 2026-04-08 12:50:10 - INFO - parrotllm.training - ** New best validation loss! **
348
+ 2026-04-08 12:50:11 - INFO - parrotllm.training - Saved best checkpoint: /home/user/code/ParrotLLM/runs/big_run/exp_c/run_20260408_124138/checkpoints/best_loss_3p9380_epoch_0000_step_0001100.pt
349
+ 2026-04-08 12:50:12 - INFO - parrotllm.training - ------------------------------------------------------------
350
+ 2026-04-08 12:50:15 - INFO - parrotllm.training - step 1110 | epoch 0 | loss 3.8959 | lr 1.30e-03 | grad 0.2816
351
+ 2026-04-08 12:50:18 - INFO - parrotllm.training - step 1120 | epoch 0 | loss 4.0434 | lr 1.30e-03 | grad 0.2867
352
+ 2026-04-08 12:50:21 - INFO - parrotllm.training - step 1130 | epoch 0 | loss 3.9461 | lr 1.30e-03 | grad 0.2710
353
+ 2026-04-08 12:50:24 - INFO - parrotllm.training - step 1140 | epoch 0 | loss 4.0211 | lr 1.30e-03 | grad 0.3127
354
+ 2026-04-08 12:50:27 - INFO - parrotllm.training - step 1150 | epoch 0 | loss 3.8917 | lr 1.30e-03 | grad 0.2831
355
+ 2026-04-08 12:50:30 - INFO - parrotllm.training - step 1160 | epoch 0 | loss 4.0191 | lr 1.30e-03 | grad 0.2494
356
+ 2026-04-08 12:50:33 - INFO - parrotllm.training - step 1170 | epoch 0 | loss 3.9068 | lr 1.30e-03 | grad 0.2486
357
+ 2026-04-08 12:50:36 - INFO - parrotllm.training - step 1180 | epoch 0 | loss 3.8646 | lr 1.30e-03 | grad 0.2616
358
+ 2026-04-08 12:50:39 - INFO - parrotllm.training - step 1190 | epoch 0 | loss 3.9071 | lr 1.30e-03 | grad 0.2977
359
+ 2026-04-08 12:50:42 - INFO - parrotllm.training - step 1200 | epoch 0 | loss 3.9299 | lr 1.30e-03 | grad 0.3024
360
+ 2026-04-08 12:50:42 - INFO - parrotllm.training - Starting evaluation...
361
+ 2026-04-08 12:50:42 - INFO - parrotllm.training - ------------------------------------------------------------
362
+ 2026-04-08 12:50:48 - INFO - parrotllm.training - Train: loss=3.9299, ppl=50.90
363
+ 2026-04-08 12:50:48 - INFO - parrotllm.training - Val: loss=3.8945, ppl=49.13
364
+ 2026-04-08 12:50:48 - INFO - parrotllm.training - ** New best validation loss! **
365
+ 2026-04-08 12:50:49 - INFO - parrotllm.training - Saved best checkpoint: /home/user/code/ParrotLLM/runs/big_run/exp_c/run_20260408_124138/checkpoints/best_loss_3p8945_epoch_0000_step_0001200.pt
366
+ 2026-04-08 12:50:50 - INFO - parrotllm.training - ------------------------------------------------------------
367
+ 2026-04-08 12:50:53 - INFO - parrotllm.training - step 1210 | epoch 0 | loss 3.8974 | lr 1.30e-03 | grad 0.2758
368
+ 2026-04-08 12:50:56 - INFO - parrotllm.training - step 1220 | epoch 0 | loss 3.8784 | lr 1.30e-03 | grad 0.2543
369
+ 2026-04-08 12:50:59 - INFO - parrotllm.training - step 1230 | epoch 0 | loss 3.9110 | lr 1.30e-03 | grad 0.2539
370
+ 2026-04-08 12:51:02 - INFO - parrotllm.training - step 1240 | epoch 0 | loss 3.9338 | lr 1.30e-03 | grad 0.2678
371
+ 2026-04-08 12:51:05 - INFO - parrotllm.training - step 1250 | epoch 0 | loss 3.8986 | lr 1.30e-03 | grad 0.3379
372
+ 2026-04-08 12:51:06 - INFO - parrotllm.training - Saved checkpoint: /home/user/code/ParrotLLM/runs/big_run/exp_c/run_20260408_124138/checkpoints/last_epoch_0000_step_0001250.pt
373
+ 2026-04-08 12:51:10 - INFO - parrotllm.training - step 1260 | epoch 0 | loss 3.8480 | lr 1.30e-03 | grad 0.3182
374
+ 2026-04-08 12:51:14 - INFO - parrotllm.training - step 1270 | epoch 0 | loss 3.9598 | lr 1.30e-03 | grad 0.2720
375
+ 2026-04-08 12:51:17 - INFO - parrotllm.training - step 1280 | epoch 0 | loss 3.9995 | lr 1.30e-03 | grad 0.2284
376
+ 2026-04-08 12:51:20 - INFO - parrotllm.training - step 1290 | epoch 0 | loss 3.8092 | lr 1.30e-03 | grad 0.2447
377
+ 2026-04-08 12:51:23 - INFO - parrotllm.training - step 1300 | epoch 0 | loss 3.8669 | lr 1.30e-03 | grad 0.3316
378
+ 2026-04-08 12:51:23 - INFO - parrotllm.training - Starting evaluation...
379
+ 2026-04-08 12:51:23 - INFO - parrotllm.training - ------------------------------------------------------------
380
+ 2026-04-08 12:51:29 - INFO - parrotllm.training - Train: loss=3.8669, ppl=47.79
381
+ 2026-04-08 12:51:29 - INFO - parrotllm.training - Val: loss=3.8603, ppl=47.48
382
+ 2026-04-08 12:51:29 - INFO - parrotllm.training - ** New best validation loss! **
383
+ 2026-04-08 12:51:29 - INFO - parrotllm.training - Saved best checkpoint: /home/user/code/ParrotLLM/runs/big_run/exp_c/run_20260408_124138/checkpoints/best_loss_3p8603_epoch_0000_step_0001300.pt
384
+ 2026-04-08 12:51:30 - INFO - parrotllm.training - ------------------------------------------------------------
385
+ 2026-04-08 12:51:33 - INFO - parrotllm.training - step 1310 | epoch 0 | loss 3.8566 | lr 1.30e-03 | grad 0.3084
386
+ 2026-04-08 12:51:36 - INFO - parrotllm.training - step 1320 | epoch 0 | loss 3.9365 | lr 1.30e-03 | grad 0.2760
387
+ 2026-04-08 12:51:39 - INFO - parrotllm.training - step 1330 | epoch 0 | loss 3.9222 | lr 1.30e-03 | grad 0.3011
388
+ 2026-04-08 12:51:42 - INFO - parrotllm.training - step 1340 | epoch 0 | loss 3.9364 | lr 1.30e-03 | grad 0.2259
389
+ 2026-04-08 12:51:45 - INFO - parrotllm.training - step 1350 | epoch 0 | loss 3.8879 | lr 1.30e-03 | grad 0.2769
390
+ 2026-04-08 12:51:49 - INFO - parrotllm.training - step 1360 | epoch 0 | loss 3.9090 | lr 1.30e-03 | grad 0.2422
391
+ 2026-04-08 12:51:52 - INFO - parrotllm.training - step 1370 | epoch 0 | loss 3.9005 | lr 1.30e-03 | grad 0.2232
392
+ 2026-04-08 12:51:55 - INFO - parrotllm.training - step 1380 | epoch 0 | loss 3.8989 | lr 1.30e-03 | grad 0.2461
393
+ 2026-04-08 12:51:58 - INFO - parrotllm.training - step 1390 | epoch 0 | loss 3.8862 | lr 1.30e-03 | grad 0.2708
394
+ 2026-04-08 12:52:01 - INFO - parrotllm.training - step 1400 | epoch 0 | loss 3.9330 | lr 1.30e-03 | grad 0.2284
395
+ 2026-04-08 12:52:01 - INFO - parrotllm.training - Starting evaluation...
396
+ 2026-04-08 12:52:01 - INFO - parrotllm.training - ------------------------------------------------------------
397
+ 2026-04-08 12:52:07 - INFO - parrotllm.training - Train: loss=3.9330, ppl=51.06
398
+ 2026-04-08 12:52:07 - INFO - parrotllm.training - Val: loss=3.8274, ppl=45.94
399
+ 2026-04-08 12:52:07 - INFO - parrotllm.training - ** New best validation loss! **
400
+ 2026-04-08 12:52:08 - INFO - parrotllm.training - Saved best checkpoint: /home/user/code/ParrotLLM/runs/big_run/exp_c/run_20260408_124138/checkpoints/best_loss_3p8274_epoch_0000_step_0001400.pt
401
+ 2026-04-08 12:52:08 - INFO - parrotllm.training - ------------------------------------------------------------
402
+ 2026-04-08 12:52:12 - INFO - parrotllm.training - step 1410 | epoch 0 | loss 3.8773 | lr 1.30e-03 | grad 0.2665
403
+ 2026-04-08 12:52:15 - INFO - parrotllm.training - step 1420 | epoch 0 | loss 3.9158 | lr 1.30e-03 | grad 0.2710
404
+ 2026-04-08 12:52:18 - INFO - parrotllm.training - step 1430 | epoch 0 | loss 3.8370 | lr 1.30e-03 | grad 0.2496
405
+ 2026-04-08 12:52:21 - INFO - parrotllm.training - step 1440 | epoch 0 | loss 3.8494 | lr 1.30e-03 | grad 0.2908
406
+ 2026-04-08 12:52:24 - INFO - parrotllm.training - step 1450 | epoch 0 | loss 3.8176 | lr 1.30e-03 | grad 0.2381
407
+ 2026-04-08 12:52:27 - INFO - parrotllm.training - step 1460 | epoch 0 | loss 3.9117 | lr 1.30e-03 | grad 0.2631
408
+ 2026-04-08 12:52:30 - INFO - parrotllm.training - step 1470 | epoch 0 | loss 3.8423 | lr 1.30e-03 | grad 0.3623
409
+ 2026-04-08 12:52:33 - INFO - parrotllm.training - step 1480 | epoch 0 | loss 3.8607 | lr 1.30e-03 | grad 0.2508
410
+ 2026-04-08 12:52:36 - INFO - parrotllm.training - step 1490 | epoch 0 | loss 3.7899 | lr 1.30e-03 | grad 0.2387
411
+ 2026-04-08 12:52:39 - INFO - parrotllm.training - step 1500 | epoch 0 | loss 3.7814 | lr 1.30e-03 | grad 0.2772
412
+ 2026-04-08 12:52:39 - INFO - parrotllm.training - Starting evaluation...
413
+ 2026-04-08 12:52:39 - INFO - parrotllm.training - ------------------------------------------------------------
414
+ 2026-04-08 12:52:45 - INFO - parrotllm.training - Train: loss=3.7814, ppl=43.88
415
+ 2026-04-08 12:52:45 - INFO - parrotllm.training - Val: loss=3.8074, ppl=45.03
416
+ 2026-04-08 12:52:45 - INFO - parrotllm.training - ** New best validation loss! **
417
+ 2026-04-08 12:52:46 - INFO - parrotllm.training - Saved best checkpoint: /home/user/code/ParrotLLM/runs/big_run/exp_c/run_20260408_124138/checkpoints/best_loss_3p8074_epoch_0000_step_0001500.pt
418
+ 2026-04-08 12:52:47 - INFO - parrotllm.training - ------------------------------------------------------------
419
+ 2026-04-08 12:52:48 - INFO - parrotllm.training - Saved checkpoint: /home/user/code/ParrotLLM/runs/big_run/exp_c/run_20260408_124138/checkpoints/last_epoch_0000_step_0001500.pt
420
+ 2026-04-08 12:52:52 - INFO - parrotllm.training - step 1510 | epoch 0 | loss 3.8355 | lr 1.30e-03 | grad 0.2602
421
+ 2026-04-08 12:52:55 - INFO - parrotllm.training - step 1520 | epoch 0 | loss 3.8465 | lr 1.30e-03 | grad 0.2622
422
+ 2026-04-08 12:52:58 - INFO - parrotllm.training - step 1530 | epoch 0 | loss 3.9231 | lr 1.30e-03 | grad 0.2776
423
+ 2026-04-08 12:53:01 - INFO - parrotllm.training - step 1540 | epoch 0 | loss 3.7939 | lr 1.30e-03 | grad 0.2384
424
+ 2026-04-08 12:53:04 - INFO - parrotllm.training - step 1550 | epoch 0 | loss 3.7998 | lr 1.30e-03 | grad 0.2477
425
+ 2026-04-08 12:53:07 - INFO - parrotllm.training - step 1560 | epoch 0 | loss 3.8265 | lr 1.30e-03 | grad 0.2496
426
+ 2026-04-08 12:53:10 - INFO - parrotllm.training - step 1570 | epoch 0 | loss 3.8365 | lr 1.30e-03 | grad 0.2401
427
+ 2026-04-08 12:53:13 - INFO - parrotllm.training - step 1580 | epoch 0 | loss 3.8086 | lr 1.30e-03 | grad 0.2788
428
+ 2026-04-08 12:53:16 - INFO - parrotllm.training - step 1590 | epoch 0 | loss 3.7861 | lr 1.30e-03 | grad 0.2817
429
+ 2026-04-08 12:53:19 - INFO - parrotllm.training - step 1600 | epoch 0 | loss 3.7310 | lr 1.30e-03 | grad 0.2718
430
+ 2026-04-08 12:53:19 - INFO - parrotllm.training - Starting evaluation...
431
+ 2026-04-08 12:53:19 - INFO - parrotllm.training - ------------------------------------------------------------
432
+ 2026-04-08 12:53:25 - INFO - parrotllm.training - Train: loss=3.7310, ppl=41.72
433
+ 2026-04-08 12:53:25 - INFO - parrotllm.training - Val: loss=3.7825, ppl=43.93
434
+ 2026-04-08 12:53:25 - INFO - parrotllm.training - ** New best validation loss! **
435
+ 2026-04-08 12:53:26 - INFO - parrotllm.training - Saved best checkpoint: /home/user/code/ParrotLLM/runs/big_run/exp_c/run_20260408_124138/checkpoints/best_loss_3p7825_epoch_0000_step_0001600.pt
436
+ 2026-04-08 12:53:27 - INFO - parrotllm.training - ------------------------------------------------------------
437
+ 2026-04-08 12:53:30 - INFO - parrotllm.training - step 1610 | epoch 0 | loss 3.7867 | lr 1.30e-03 | grad 0.2506
438
+ 2026-04-08 12:53:33 - INFO - parrotllm.training - step 1620 | epoch 0 | loss 3.7436 | lr 1.30e-03 | grad 0.2619
439
+ 2026-04-08 12:53:36 - INFO - parrotllm.training - step 1630 | epoch 0 | loss 3.9659 | lr 1.30e-03 | grad 0.2411
440
+ 2026-04-08 12:53:39 - INFO - parrotllm.training - step 1640 | epoch 0 | loss 3.7774 | lr 1.30e-03 | grad 0.2585
441
+ 2026-04-08 12:53:42 - INFO - parrotllm.training - step 1650 | epoch 0 | loss 3.8461 | lr 1.30e-03 | grad 0.3146
442
+ 2026-04-08 12:53:45 - INFO - parrotllm.training - step 1660 | epoch 0 | loss 3.8304 | lr 1.30e-03 | grad 0.2715
443
+ 2026-04-08 12:53:48 - INFO - parrotllm.training - step 1670 | epoch 0 | loss 3.8136 | lr 1.30e-03 | grad 0.2449
444
+ 2026-04-08 12:53:51 - INFO - parrotllm.training - step 1680 | epoch 0 | loss 3.8394 | lr 1.30e-03 | grad 0.2479
445
+ 2026-04-08 12:53:54 - INFO - parrotllm.training - step 1690 | epoch 0 | loss 3.7196 | lr 1.30e-03 | grad 0.2394
446
+ 2026-04-08 12:53:57 - INFO - parrotllm.training - step 1700 | epoch 0 | loss 3.7920 | lr 1.30e-03 | grad 0.2520
447
+ 2026-04-08 12:53:57 - INFO - parrotllm.training - Starting evaluation...
448
+ 2026-04-08 12:53:57 - INFO - parrotllm.training - ------------------------------------------------------------
449
+ 2026-04-08 12:54:03 - INFO - parrotllm.training - Train: loss=3.7920, ppl=44.34
450
+ 2026-04-08 12:54:03 - INFO - parrotllm.training - Val: loss=3.7612, ppl=43.00
451
+ 2026-04-08 12:54:03 - INFO - parrotllm.training - ** New best validation loss! **
452
+ 2026-04-08 12:54:04 - INFO - parrotllm.training - Saved best checkpoint: /home/user/code/ParrotLLM/runs/big_run/exp_c/run_20260408_124138/checkpoints/best_loss_3p7612_epoch_0000_step_0001700.pt
453
+ 2026-04-08 12:54:05 - INFO - parrotllm.training - ------------------------------------------------------------
454
+ 2026-04-08 12:54:08 - INFO - parrotllm.training - step 1710 | epoch 0 | loss 3.7862 | lr 1.30e-03 | grad 0.2962
455
+ 2026-04-08 12:54:11 - INFO - parrotllm.training - step 1720 | epoch 0 | loss 3.8174 | lr 1.30e-03 | grad 0.2350
456
+ 2026-04-08 12:54:14 - INFO - parrotllm.training - step 1730 | epoch 0 | loss 3.9581 | lr 1.30e-03 | grad 0.2243
457
+ 2026-04-08 12:54:17 - INFO - parrotllm.training - step 1740 | epoch 0 | loss 3.8328 | lr 1.30e-03 | grad 0.2576
458
+ 2026-04-08 12:54:20 - INFO - parrotllm.training - step 1750 | epoch 0 | loss 3.7767 | lr 1.30e-03 | grad 0.2573
459
+ 2026-04-08 12:54:21 - INFO - parrotllm.training - Saved checkpoint: /home/user/code/ParrotLLM/runs/big_run/exp_c/run_20260408_124138/checkpoints/last_epoch_0000_step_0001750.pt
460
+ 2026-04-08 12:54:25 - INFO - parrotllm.training - step 1760 | epoch 0 | loss 3.8260 | lr 1.30e-03 | grad 0.2822
461
+ 2026-04-08 12:54:29 - INFO - parrotllm.training - step 1770 | epoch 0 | loss 3.9070 | lr 1.30e-03 | grad 0.2378
462
+ 2026-04-08 12:54:32 - INFO - parrotllm.training - step 1780 | epoch 0 | loss 3.7750 | lr 1.30e-03 | grad 0.2820
463
+ 2026-04-08 12:54:35 - INFO - parrotllm.training - step 1790 | epoch 0 | loss 3.8340 | lr 1.30e-03 | grad 0.2993
464
+ 2026-04-08 12:54:38 - INFO - parrotllm.training - step 1800 | epoch 0 | loss 3.8334 | lr 1.30e-03 | grad 0.2901
465
+ 2026-04-08 12:54:38 - INFO - parrotllm.training - Starting evaluation...
466
+ 2026-04-08 12:54:38 - INFO - parrotllm.training - ------------------------------------------------------------
467
+ 2026-04-08 12:54:43 - INFO - parrotllm.training - Train: loss=3.8334, ppl=46.22
468
+ 2026-04-08 12:54:43 - INFO - parrotllm.training - Val: loss=3.7424, ppl=42.20
469
+ 2026-04-08 12:54:43 - INFO - parrotllm.training - ** New best validation loss! **
470
+ 2026-04-08 12:54:44 - INFO - parrotllm.training - Saved best checkpoint: /home/user/code/ParrotLLM/runs/big_run/exp_c/run_20260408_124138/checkpoints/best_loss_3p7424_epoch_0000_step_0001800.pt
471
+ 2026-04-08 12:54:45 - INFO - parrotllm.training - ------------------------------------------------------------
472
+ 2026-04-08 12:54:48 - INFO - parrotllm.training - step 1810 | epoch 0 | loss 3.8063 | lr 1.30e-03 | grad 0.2505
473
+ 2026-04-08 12:54:51 - INFO - parrotllm.training - step 1820 | epoch 0 | loss 3.7296 | lr 1.30e-03 | grad 0.3443
474
+ 2026-04-08 12:54:54 - INFO - parrotllm.training - step 1830 | epoch 0 | loss 3.7140 | lr 1.30e-03 | grad 0.2645
475
+ 2026-04-08 12:54:57 - INFO - parrotllm.training - step 1840 | epoch 0 | loss 3.8231 | lr 1.30e-03 | grad 0.2163
476
+ 2026-04-08 12:55:00 - INFO - parrotllm.training - step 1850 | epoch 0 | loss 3.6996 | lr 1.30e-03 | grad 0.3037
477
+ 2026-04-08 12:55:03 - INFO - parrotllm.training - step 1860 | epoch 0 | loss 3.7551 | lr 1.30e-03 | grad 0.2463
478
+ 2026-04-08 12:55:07 - INFO - parrotllm.training - step 1870 | epoch 0 | loss 3.7731 | lr 1.30e-03 | grad 0.2317
479
+ 2026-04-08 12:55:10 - INFO - parrotllm.training - step 1880 | epoch 0 | loss 3.7692 | lr 1.30e-03 | grad 0.2434
480
+ 2026-04-08 12:55:13 - INFO - parrotllm.training - step 1890 | epoch 0 | loss 3.7588 | lr 1.30e-03 | grad 0.2451
481
+ 2026-04-08 12:55:16 - INFO - parrotllm.training - step 1900 | epoch 0 | loss 3.6820 | lr 1.30e-03 | grad 0.2575
482
+ 2026-04-08 12:55:16 - INFO - parrotllm.training - Starting evaluation...
483
+ 2026-04-08 12:55:16 - INFO - parrotllm.training - ------------------------------------------------------------
484
+ 2026-04-08 12:55:21 - INFO - parrotllm.training - Train: loss=3.6820, ppl=39.72
485
+ 2026-04-08 12:55:21 - INFO - parrotllm.training - Val: loss=3.7291, ppl=41.64
486
+ 2026-04-08 12:55:21 - INFO - parrotllm.training - ** New best validation loss! **
487
+ 2026-04-08 12:55:22 - INFO - parrotllm.training - Saved best checkpoint: /home/user/code/ParrotLLM/runs/big_run/exp_c/run_20260408_124138/checkpoints/best_loss_3p7291_epoch_0000_step_0001900.pt
488
+ 2026-04-08 12:55:23 - INFO - parrotllm.training - ------------------------------------------------------------
489
+ 2026-04-08 12:55:26 - INFO - parrotllm.training - step 1910 | epoch 0 | loss 3.7672 | lr 1.30e-03 | grad 0.2429
490
+ 2026-04-08 12:55:29 - INFO - parrotllm.training - step 1920 | epoch 0 | loss 3.6872 | lr 1.30e-03 | grad 0.2719
491
+ 2026-04-08 12:55:32 - INFO - parrotllm.training - step 1930 | epoch 0 | loss 3.6756 | lr 1.30e-03 | grad 0.2532
runs/big_run/exp_c/run_20260408_124138/.ipynb_checkpoints/training_plots-checkpoint.pdf ADDED
Binary file (82.2 kB). View file
 
runs/big_run/exp_c/run_20260408_124138/checkpoints/best_loss_3p5437_epoch_0001_step_0003000.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:0bc17499c02ac9f879aefc3f0d4538c1d7b78a6053733de6cc222a1252102467
3
+ size 479869436
runs/big_run/exp_c/run_20260408_124138/checkpoints/best_loss_3p5634_epoch_0001_step_0002900.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:4653a7f47a53b56b69dbe924cdef7bf3acd2b356a0a39b64d9a1fbd16ae32d0d
3
+ size 479869436
runs/big_run/exp_c/run_20260408_124138/checkpoints/best_loss_3p5885_epoch_0001_step_0002800.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:97b2edd3d94eb11d318f536c35e4c7cf64cac9c74a9ead333d966cea75e2e46d
3
+ size 479869436
runs/big_run/exp_c/run_20260408_124138/checkpoints/best_loss_3p6128_epoch_0001_step_0002700.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5ead087c1e9dc0001d0a793d0c1e9d4eff953bbf04a038370ea803ebf6ec2c07
3
+ size 479869436
runs/big_run/exp_c/run_20260408_124138/checkpoints/best_loss_3p6383_epoch_0001_step_0002600.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:1871e92fe42d9386f64226660007a5c675b625f9d596955e7f3574143e789790
3
+ size 479869436
runs/big_run/exp_c/run_20260408_124138/checkpoints/best_loss_3p6614_epoch_0001_step_0002500.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:68832a597637943f8a4a4ad27b1189b28278a152940e7840b3880705cbbe2bbd
3
+ size 479869436
runs/big_run/exp_c/run_20260408_124138/checkpoints/best_loss_3p6698_epoch_0001_step_0002400.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3c6a395766b8479d43e7994bda742076f30a93ee470083e73c5c7fa06934c6a1
3
+ size 479869436
runs/big_run/exp_c/run_20260408_124138/checkpoints/best_loss_3p6838_epoch_0001_step_0002300.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5ffb13d568d2c0ce19ea09e1d178abf15c54f0f97fb074b8d9f0f9ef0af84a1f
3
+ size 479869436
runs/big_run/exp_c/run_20260408_124138/checkpoints/best_loss_3p6926_epoch_0001_step_0002200.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:bfbc7ab10b9349eb2c0d058a34188ec684fe2a125747152ca8339d31fd2b361b
3
+ size 479869436
runs/big_run/exp_c/run_20260408_124138/checkpoints/best_loss_3p7036_epoch_0001_step_0002100.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:aa98f1799d4ae0c089b2e3aa93782d67beb631fc8a994b31e42ab65f6c4677e5
3
+ size 479869436
runs/big_run/exp_c/run_20260408_124138/checkpoints/last_epoch_0000_step_0001000.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:30d9e7a14ed9005cce3e010643d0f5796a919d41b000b72ae027c4ff6488e230
3
+ size 479860424
runs/big_run/exp_c/run_20260408_124138/checkpoints/last_epoch_0000_step_0001250.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f5ba15ef4593e4c0fefe28cd4ad029ca2bbcd6f3d02efcb610da76db212c516b
3
+ size 479860424
runs/big_run/exp_c/run_20260408_124138/checkpoints/last_epoch_0000_step_0001500.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3460efb58af027f77a4330bb0231f3c32fbe831c816156fe17421311b99a2292
3
+ size 479860424
runs/big_run/exp_c/run_20260408_124138/checkpoints/last_epoch_0000_step_0001750.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:15e326bdd2add94245e2a8bab37db48810b544bfb71b532aa8131fb202a99ebf
3
+ size 479860424
runs/big_run/exp_c/run_20260408_124138/checkpoints/last_epoch_0001_step_0002000.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e4d33382da9beb2f952f5afd7d8f249b77e40002665f63c43573da922a0b9ec5
3
+ size 479860424
runs/big_run/exp_c/run_20260408_124138/checkpoints/last_epoch_0001_step_0002250.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b42744b208cf323c8324c268a669968e3fa27ac59137bc0b04fbf3405f56c495
3
+ size 479860424
runs/big_run/exp_c/run_20260408_124138/checkpoints/last_epoch_0001_step_0002500.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3a056268861fddb846e3817d80746dc0ce9cb65971f7b2f0a2c12a32b325585a
3
+ size 479860424
runs/big_run/exp_c/run_20260408_124138/checkpoints/last_epoch_0001_step_0002750.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d3e385dc6a18fce600c3fd626e6ec558d66229c33e150432b990ec585744d94a
3
+ size 479860424
runs/big_run/exp_c/run_20260408_124138/checkpoints/last_epoch_0001_step_0003000.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a97b82abb1d5b2a4161036fe49b890dd03d299c918bc65a1d7611eb2523a199a
3
+ size 479860424
runs/big_run/exp_c/run_20260408_124138/checkpoints/last_epoch_0001_step_0003052.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:263f7db44ad406670e226b7ae096490df7bd4a0149c7908dcf8638e9fc0ef1b6
3
+ size 479860424
runs/big_run/exp_c/run_20260408_124138/config.json ADDED
@@ -0,0 +1,86 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "preprocess": null,
3
+ "model": {
4
+ "vocab_size": 50258,
5
+ "pad_token_id": 50257,
6
+ "bos_token_id": 50256,
7
+ "eos_token_id": 50256,
8
+ "d_model": 384,
9
+ "n_layers": 14,
10
+ "n_heads": 6,
11
+ "d_ff": 768,
12
+ "context_length": 1024,
13
+ "bias": false,
14
+ "dropout": 0.005,
15
+ "rope_theta": 10000.0,
16
+ "gradient_checkpointing": false
17
+ },
18
+ "training": {
19
+ "device": "auto",
20
+ "train_bin": "data/exp_c/train.bin",
21
+ "val_bin": "data/exp_c/val.bin",
22
+ "num_workers": 4,
23
+ "pin_memory": true,
24
+ "batch_size": 32,
25
+ "gradient_accumulation_steps": 1,
26
+ "learning_rate": 0.001305,
27
+ "min_lr": 0.0001305,
28
+ "weight_decay": 0.063,
29
+ "beta1": 0.9,
30
+ "beta2": 0.9978,
31
+ "grad_clip": 1.786,
32
+ "warmup_steps": 305,
33
+ "max_steps": 3052,
34
+ "lr_schedule": "wsd",
35
+ "lr_decay_ratio": 0.179,
36
+ "z_loss_coeff": 0.000275,
37
+ "save_every": 250,
38
+ "eval_every": 100,
39
+ "checkpoint_dir": "checkpoints",
40
+ "keep_last_checkpoints": 10,
41
+ "keep_best_checkpoints": 10,
42
+ "early_stopping_patience": 0,
43
+ "early_stopping_min_delta": 0.001,
44
+ "runs_dir": "runs/big_run/exp_c",
45
+ "log_every": 10,
46
+ "hf_upload": {
47
+ "repo_id": "ParrotLabs/Preprocessed",
48
+ "repo_type": "dataset",
49
+ "path_in_repo": "",
50
+ "private": null
51
+ },
52
+ "compile": true
53
+ },
54
+ "tune": null,
55
+ "logging": {
56
+ "console_level": "INFO",
57
+ "file_level": "DEBUG",
58
+ "components": {
59
+ "data_preprocessing": "DEBUG",
60
+ "model_initialization": "INFO",
61
+ "training": "INFO"
62
+ },
63
+ "profiler": null
64
+ },
65
+ "eval": {
66
+ "device": "auto",
67
+ "batch_size": 32,
68
+ "max_sequences": 512,
69
+ "datasets": [
70
+ {
71
+ "name": "wikitext",
72
+ "path": "wikitext",
73
+ "subset": "wikitext-103-raw-v1",
74
+ "split": "test"
75
+ },
76
+ {
77
+ "name": "exp_c_val",
78
+ "path": "data/exp_c/val.bin",
79
+ "subset": null,
80
+ "split": null
81
+ }
82
+ ]
83
+ },
84
+ "inference": null,
85
+ "chat": null
86
+ }
runs/big_run/exp_c/run_20260408_124138/metrics.jsonl ADDED
The diff for this file is too large to render. See raw diff
 
runs/big_run/exp_c/run_20260408_124138/train.log ADDED
@@ -0,0 +1,719 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ 2026-04-08 12:41:39 - INFO - parrotllm - Logging initialised -> runs/big_run/exp_c/run_20260408_124138/train.log
2
+ 2026-04-08 12:41:39 - INFO - parrotllm.training - device=cuda:0 | rank=0 | world_size=8 | distributed=yes
3
+ 2026-04-08 12:41:39 - INFO - parrotllm.training - checkpoint directory=/home/user/code/ParrotLLM/runs/big_run/exp_c/run_20260408_124138/checkpoints | keep_last=10 | keep_best=10
4
+ 2026-04-08 12:41:40 - INFO - parrotllm.training - compiling model with torch.compile...
5
+ 2026-04-08 12:41:41 - INFO - parrotllm.training -
6
+ ============================================================
7
+ MODEL ARCHITECTURE SUMMARY
8
+ ============================================================
9
+
10
+ Configuration:
11
+ Vocab size: 50258
12
+ Block size (context): 1024
13
+ Layers: 14
14
+ Heads: 6
15
+ Embedding dim: 384
16
+ FFN hidden dim: 768
17
+ Dropout: 0.005
18
+ Bias: False
19
+
20
+ Parameters (unique, weight-tied layers counted once):
21
+ Total: 39,966,592 (39.97M)
22
+ Trainable: 39,966,592 (39.97M)
23
+ Non-trainable: 0
24
+ Non-embedding: 39,966,592 (39.97M)
25
+ Position embeddings: 0 (RoPE: 0 learned params)
26
+ Size (MB): 159.87
27
+
28
+ Layer-wise breakdown (torchinfo):
29
+ Note: torchinfo double-counts weight-tied layers (tok_emb/lm_head).
30
+ ============================================================================================================================================
31
+ Layer (type:depth-idx) Input Shape Output Shape Param # Trainable
32
+ ============================================================================================================================================
33
+ ParrotLLM [32, 1024] [32, 1024, 50258] -- True
34
+ ├─Embedding: 1-1 [32, 1024] [32, 1024, 384] 19,299,072 True
35
+ ├─Dropout: 1-2 [32, 1024, 384] [32, 1024, 384] -- --
36
+ ├─ModuleList: 1-3 -- -- -- True
37
+ │ └─TransformerBlock: 2-1 [32, 1024, 384] [32, 1024, 384] -- True
38
+ │ │ └─RMSNorm: 3-1 [32, 1024, 384] [32, 1024, 384] 384 True
39
+ │ │ └─MultiHeadAttention: 3-2 [32, 1024, 384] [32, 1024, 384] 589,952 True
40
+ │ │ └─RMSNorm: 3-3 [32, 1024, 384] [32, 1024, 384] 384 True
41
+ │ │ └─RMSNorm: 3-4 [32, 1024, 384] [32, 1024, 384] 384 True
42
+ │ │ └─SwiGLUMLP: 3-5 [32, 1024, 384] [32, 1024, 384] 884,736 True
43
+ │ │ └─RMSNorm: 3-6 [32, 1024, 384] [32, 1024, 384] 384 True
44
+ │ └─TransformerBlock: 2-2 [32, 1024, 384] [32, 1024, 384] -- True
45
+ │ │ └─RMSNorm: 3-7 [32, 1024, 384] [32, 1024, 384] 384 True
46
+ │ │ └─MultiHeadAttention: 3-8 [32, 1024, 384] [32, 1024, 384] 589,952 True
47
+ │ │ └─RMSNorm: 3-9 [32, 1024, 384] [32, 1024, 384] 384 True
48
+ │ │ └─RMSNorm: 3-10 [32, 1024, 384] [32, 1024, 384] 384 True
49
+ │ │ └─SwiGLUMLP: 3-11 [32, 1024, 384] [32, 1024, 384] 884,736 True
50
+ │ │ └─RMSNorm: 3-12 [32, 1024, 384] [32, 1024, 384] 384 True
51
+ │ └─TransformerBlock: 2-3 [32, 1024, 384] [32, 1024, 384] -- True
52
+ │ │ └─RMSNorm: 3-13 [32, 1024, 384] [32, 1024, 384] 384 True
53
+ │ │ └─MultiHeadAttention: 3-14 [32, 1024, 384] [32, 1024, 384] 589,952 True
54
+ │ │ └─RMSNorm: 3-15 [32, 1024, 384] [32, 1024, 384] 384 True
55
+ │ │ └─RMSNorm: 3-16 [32, 1024, 384] [32, 1024, 384] 384 True
56
+ │ │ └─SwiGLUMLP: 3-17 [32, 1024, 384] [32, 1024, 384] 884,736 True
57
+ │ │ └─RMSNorm: 3-18 [32, 1024, 384] [32, 1024, 384] 384 True
58
+ │ └─TransformerBlock: 2-4 [32, 1024, 384] [32, 1024, 384] -- True
59
+ │ │ └─RMSNorm: 3-19 [32, 1024, 384] [32, 1024, 384] 384 True
60
+ │ │ └─MultiHeadAttention: 3-20 [32, 1024, 384] [32, 1024, 384] 589,952 True
61
+ │ │ └─RMSNorm: 3-21 [32, 1024, 384] [32, 1024, 384] 384 True
62
+ │ │ └─RMSNorm: 3-22 [32, 1024, 384] [32, 1024, 384] 384 True
63
+ │ │ └─SwiGLUMLP: 3-23 [32, 1024, 384] [32, 1024, 384] 884,736 True
64
+ │ │ └─RMSNorm: 3-24 [32, 1024, 384] [32, 1024, 384] 384 True
65
+ │ └─TransformerBlock: 2-5 [32, 1024, 384] [32, 1024, 384] -- True
66
+ │ │ └─RMSNorm: 3-25 [32, 1024, 384] [32, 1024, 384] 384 True
67
+ │ │ └─MultiHeadAttention: 3-26 [32, 1024, 384] [32, 1024, 384] 589,952 True
68
+ │ │ └─RMSNorm: 3-27 [32, 1024, 384] [32, 1024, 384] 384 True
69
+ │ │ └─RMSNorm: 3-28 [32, 1024, 384] [32, 1024, 384] 384 True
70
+ │ │ └─SwiGLUMLP: 3-29 [32, 1024, 384] [32, 1024, 384] 884,736 True
71
+ │ │ └─RMSNorm: 3-30 [32, 1024, 384] [32, 1024, 384] 384 True
72
+ │ └─TransformerBlock: 2-6 [32, 1024, 384] [32, 1024, 384] -- True
73
+ │ │ └─RMSNorm: 3-31 [32, 1024, 384] [32, 1024, 384] 384 True
74
+ │ │ └─MultiHeadAttention: 3-32 [32, 1024, 384] [32, 1024, 384] 589,952 True
75
+ │ │ └─RMSNorm: 3-33 [32, 1024, 384] [32, 1024, 384] 384 True
76
+ │ │ └─RMSNorm: 3-34 [32, 1024, 384] [32, 1024, 384] 384 True
77
+ │ │ └─SwiGLUMLP: 3-35 [32, 1024, 384] [32, 1024, 384] 884,736 True
78
+ │ │ └─RMSNorm: 3-36 [32, 1024, 384] [32, 1024, 384] 384 True
79
+ │ └─TransformerBlock: 2-7 [32, 1024, 384] [32, 1024, 384] -- True
80
+ │ │ └─RMSNorm: 3-37 [32, 1024, 384] [32, 1024, 384] 384 True
81
+ │ │ └─MultiHeadAttention: 3-38 [32, 1024, 384] [32, 1024, 384] 589,952 True
82
+ │ │ └─RMSNorm: 3-39 [32, 1024, 384] [32, 1024, 384] 384 True
83
+ │ │ └─RMSNorm: 3-40 [32, 1024, 384] [32, 1024, 384] 384 True
84
+ │ │ └─SwiGLUMLP: 3-41 [32, 1024, 384] [32, 1024, 384] 884,736 True
85
+ │ │ └─RMSNorm: 3-42 [32, 1024, 384] [32, 1024, 384] 384 True
86
+ │ └─TransformerBlock: 2-8 [32, 1024, 384] [32, 1024, 384] -- True
87
+ │ │ └─RMSNorm: 3-43 [32, 1024, 384] [32, 1024, 384] 384 True
88
+ │ │ └─MultiHeadAttention: 3-44 [32, 1024, 384] [32, 1024, 384] 589,952 True
89
+ │ │ └─RMSNorm: 3-45 [32, 1024, 384] [32, 1024, 384] 384 True
90
+ │ │ └─RMSNorm: 3-46 [32, 1024, 384] [32, 1024, 384] 384 True
91
+ │ │ └─SwiGLUMLP: 3-47 [32, 1024, 384] [32, 1024, 384] 884,736 True
92
+ │ │ └─RMSNorm: 3-48 [32, 1024, 384] [32, 1024, 384] 384 True
93
+ │ └─TransformerBlock: 2-9 [32, 1024, 384] [32, 1024, 384] -- True
94
+ │ │ └─RMSNorm: 3-49 [32, 1024, 384] [32, 1024, 384] 384 True
95
+ │ │ └─MultiHeadAttention: 3-50 [32, 1024, 384] [32, 1024, 384] 589,952 True
96
+ │ │ └─RMSNorm: 3-51 [32, 1024, 384] [32, 1024, 384] 384 True
97
+ │ │ └─RMSNorm: 3-52 [32, 1024, 384] [32, 1024, 384] 384 True
98
+ │ │ └─SwiGLUMLP: 3-53 [32, 1024, 384] [32, 1024, 384] 884,736 True
99
+ │ │ └─RMSNorm: 3-54 [32, 1024, 384] [32, 1024, 384] 384 True
100
+ │ └─TransformerBlock: 2-10 [32, 1024, 384] [32, 1024, 384] -- True
101
+ │ │ └─RMSNorm: 3-55 [32, 1024, 384] [32, 1024, 384] 384 True
102
+ │ │ └─MultiHeadAttention: 3-56 [32, 1024, 384] [32, 1024, 384] 589,952 True
103
+ │ │ └─RMSNorm: 3-57 [32, 1024, 384] [32, 1024, 384] 384 True
104
+ │ │ └─RMSNorm: 3-58 [32, 1024, 384] [32, 1024, 384] 384 True
105
+ │ │ └─SwiGLUMLP: 3-59 [32, 1024, 384] [32, 1024, 384] 884,736 True
106
+ │ │ └─RMSNorm: 3-60 [32, 1024, 384] [32, 1024, 384] 384 True
107
+ │ └─TransformerBlock: 2-11 [32, 1024, 384] [32, 1024, 384] -- True
108
+ │ │ └─RMSNorm: 3-61 [32, 1024, 384] [32, 1024, 384] 384 True
109
+ │ │ └─MultiHeadAttention: 3-62 [32, 1024, 384] [32, 1024, 384] 589,952 True
110
+ │ │ └─RMSNorm: 3-63 [32, 1024, 384] [32, 1024, 384] 384 True
111
+ │ │ └─RMSNorm: 3-64 [32, 1024, 384] [32, 1024, 384] 384 True
112
+ │ │ └─SwiGLUMLP: 3-65 [32, 1024, 384] [32, 1024, 384] 884,736 True
113
+ │ │ └─RMSNorm: 3-66 [32, 1024, 384] [32, 1024, 384] 384 True
114
+ │ └─TransformerBlock: 2-12 [32, 1024, 384] [32, 1024, 384] -- True
115
+ │ │ └─RMSNorm: 3-67 [32, 1024, 384] [32, 1024, 384] 384 True
116
+ │ │ └─MultiHeadAttention: 3-68 [32, 1024, 384] [32, 1024, 384] 589,952 True
117
+ │ │ └─RMSNorm: 3-69 [32, 1024, 384] [32, 1024, 384] 384 True
118
+ │ │ └─RMSNorm: 3-70 [32, 1024, 384] [32, 1024, 384] 384 True
119
+ │ │ └─SwiGLUMLP: 3-71 [32, 1024, 384] [32, 1024, 384] 884,736 True
120
+ │ │ └─RMSNorm: 3-72 [32, 1024, 384] [32, 1024, 384] 384 True
121
+ │ └─TransformerBlock: 2-13 [32, 1024, 384] [32, 1024, 384] -- True
122
+ │ │ └─RMSNorm: 3-73 [32, 1024, 384] [32, 1024, 384] 384 True
123
+ │ │ └─MultiHeadAttention: 3-74 [32, 1024, 384] [32, 1024, 384] 589,952 True
124
+ │ │ └─RMSNorm: 3-75 [32, 1024, 384] [32, 1024, 384] 384 True
125
+ │ │ └─RMSNorm: 3-76 [32, 1024, 384] [32, 1024, 384] 384 True
126
+ │ │ └─SwiGLUMLP: 3-77 [32, 1024, 384] [32, 1024, 384] 884,736 True
127
+ │ │ └─RMSNorm: 3-78 [32, 1024, 384] [32, 1024, 384] 384 True
128
+ │ └─TransformerBlock: 2-14 [32, 1024, 384] [32, 1024, 384] -- True
129
+ │ │ └─RMSNorm: 3-79 [32, 1024, 384] [32, 1024, 384] 384 True
130
+ │ │ └─MultiHeadAttention: 3-80 [32, 1024, 384] [32, 1024, 384] 589,952 True
131
+ │ │ └─RMSNorm: 3-81 [32, 1024, 384] [32, 1024, 384] 384 True
132
+ │ │ └─RMSNorm: 3-82 [32, 1024, 384] [32, 1024, 384] 384 True
133
+ │ │ └─SwiGLUMLP: 3-83 [32, 1024, 384] [32, 1024, 384] 884,736 True
134
+ │ │ └─RMSNorm: 3-84 [32, 1024, 384] [32, 1024, 384] 384 True
135
+ ├─RMSNorm: 1-4 [32, 1024, 384] [32, 1024, 384] 384 True
136
+ ├─Linear: 1-5 [32, 1024, 384] [32, 1024, 50258] 19,299,072 True
137
+ ============================================================================================================================================
138
+ Total params: 59,265,664
139
+ Trainable params: 59,265,664
140
+ Non-trainable params: 0
141
+ Total mult-adds (Units.GIGABYTES): 1.90
142
+ ============================================================================================================================================
143
+ Input size (MB): 0.26
144
+ Forward/backward pass size (MB): 34515.45
145
+ Params size (MB): 237.06
146
+ Estimated Total Size (MB): 34752.78
147
+ ============================================================================================================================================
148
+ ============================================================
149
+ 2026-04-08 12:41:41 - INFO - parrotllm.training - Starting evaluation...
150
+ 2026-04-08 12:41:49 - INFO - parrotllm.training - Initial val: loss=10.8939, ppl=53846.40
151
+ 2026-04-08 12:42:05 - INFO - parrotllm.training -
152
+ ============================================================
153
+ Starting training...
154
+ Steps per epoch (approx): 1939
155
+ Max steps: 3052
156
+ ============================================================
157
+ 2026-04-08 12:42:05 - INFO - parrotllm.training - LR schedule=wsd | decay_ratio=0.179 | z_loss_coeff=0.000275
158
+ 2026-04-08 12:42:27 - INFO - parrotllm.training - step 1 | epoch 0 | loss 10.9283 | lr 4.28e-06 | grad 6.8122
159
+ 2026-04-08 12:42:30 - INFO - parrotllm.training - step 10 | epoch 0 | loss 10.2616 | lr 4.28e-05 | grad 1.8270
160
+ 2026-04-08 12:42:33 - INFO - parrotllm.training - step 20 | epoch 0 | loss 9.8953 | lr 8.56e-05 | grad 1.6945
161
+ 2026-04-08 12:42:36 - INFO - parrotllm.training - step 30 | epoch 0 | loss 9.2911 | lr 1.28e-04 | grad 2.0493
162
+ 2026-04-08 12:42:39 - INFO - parrotllm.training - step 40 | epoch 0 | loss 8.6841 | lr 1.71e-04 | grad 1.5425
163
+ 2026-04-08 12:42:42 - INFO - parrotllm.training - step 50 | epoch 0 | loss 7.9912 | lr 2.14e-04 | grad 1.3772
164
+ 2026-04-08 12:42:45 - INFO - parrotllm.training - step 60 | epoch 0 | loss 7.5500 | lr 2.57e-04 | grad 0.9932
165
+ 2026-04-08 12:42:48 - INFO - parrotllm.training - step 70 | epoch 0 | loss 7.2957 | lr 3.00e-04 | grad 0.8042
166
+ 2026-04-08 12:42:52 - INFO - parrotllm.training - step 80 | epoch 0 | loss 7.0405 | lr 3.42e-04 | grad 0.6572
167
+ 2026-04-08 12:42:55 - INFO - parrotllm.training - step 90 | epoch 0 | loss 6.9090 | lr 3.85e-04 | grad 0.7028
168
+ 2026-04-08 12:42:58 - INFO - parrotllm.training - step 100 | epoch 0 | loss 6.8356 | lr 4.28e-04 | grad 0.4364
169
+ 2026-04-08 12:42:58 - INFO - parrotllm.training - Starting evaluation...
170
+ 2026-04-08 12:42:58 - INFO - parrotllm.training - ------------------------------------------------------------
171
+ 2026-04-08 12:43:42 - INFO - parrotllm.training - Train: loss=6.8356, ppl=930.42
172
+ 2026-04-08 12:43:42 - INFO - parrotllm.training - Val: loss=6.7630, ppl=865.27
173
+ 2026-04-08 12:43:42 - INFO - parrotllm.training - ** New best validation loss! **
174
+ 2026-04-08 12:43:43 - INFO - parrotllm.training - Saved best checkpoint: /home/user/code/ParrotLLM/runs/big_run/exp_c/run_20260408_124138/checkpoints/best_loss_6p7630_epoch_0000_step_0000100.pt
175
+ 2026-04-08 12:43:44 - INFO - parrotllm.training - ------------------------------------------------------------
176
+ 2026-04-08 12:43:47 - INFO - parrotllm.training - step 110 | epoch 0 | loss 6.7452 | lr 4.71e-04 | grad 0.9823
177
+ 2026-04-08 12:43:50 - INFO - parrotllm.training - step 120 | epoch 0 | loss 6.5647 | lr 5.13e-04 | grad 0.8019
178
+ 2026-04-08 12:43:54 - INFO - parrotllm.training - step 130 | epoch 0 | loss 6.4961 | lr 5.56e-04 | grad 0.6392
179
+ 2026-04-08 12:43:57 - INFO - parrotllm.training - step 140 | epoch 0 | loss 6.2378 | lr 5.99e-04 | grad 0.8438
180
+ 2026-04-08 12:44:00 - INFO - parrotllm.training - step 150 | epoch 0 | loss 6.2534 | lr 6.42e-04 | grad 0.5381
181
+ 2026-04-08 12:44:03 - INFO - parrotllm.training - step 160 | epoch 0 | loss 6.2198 | lr 6.85e-04 | grad 0.6169
182
+ 2026-04-08 12:44:06 - INFO - parrotllm.training - step 170 | epoch 0 | loss 6.2544 | lr 7.27e-04 | grad 1.2573
183
+ 2026-04-08 12:44:09 - INFO - parrotllm.training - step 180 | epoch 0 | loss 6.0693 | lr 7.70e-04 | grad 1.0304
184
+ 2026-04-08 12:44:12 - INFO - parrotllm.training - step 190 | epoch 0 | loss 5.9096 | lr 8.13e-04 | grad 0.4648
185
+ 2026-04-08 12:44:15 - INFO - parrotllm.training - step 200 | epoch 0 | loss 5.9346 | lr 8.56e-04 | grad 1.1069
186
+ 2026-04-08 12:44:15 - INFO - parrotllm.training - Starting evaluation...
187
+ 2026-04-08 12:44:15 - INFO - parrotllm.training - ------------------------------------------------------------
188
+ 2026-04-08 12:44:21 - INFO - parrotllm.training - Train: loss=5.9346, ppl=377.87
189
+ 2026-04-08 12:44:21 - INFO - parrotllm.training - Val: loss=5.9015, ppl=365.60
190
+ 2026-04-08 12:44:21 - INFO - parrotllm.training - ** New best validation loss! **
191
+ 2026-04-08 12:44:21 - INFO - parrotllm.training - Saved best checkpoint: /home/user/code/ParrotLLM/runs/big_run/exp_c/run_20260408_124138/checkpoints/best_loss_5p9015_epoch_0000_step_0000200.pt
192
+ 2026-04-08 12:44:22 - INFO - parrotllm.training - ------------------------------------------------------------
193
+ 2026-04-08 12:44:25 - INFO - parrotllm.training - step 210 | epoch 0 | loss 5.7735 | lr 8.99e-04 | grad 0.5488
194
+ 2026-04-08 12:44:28 - INFO - parrotllm.training - step 220 | epoch 0 | loss 5.8331 | lr 9.41e-04 | grad 0.9377
195
+ 2026-04-08 12:44:31 - INFO - parrotllm.training - step 230 | epoch 0 | loss 5.8403 | lr 9.84e-04 | grad 1.2072
196
+ 2026-04-08 12:44:34 - INFO - parrotllm.training - step 240 | epoch 0 | loss 5.6031 | lr 1.03e-03 | grad 0.3195
197
+ 2026-04-08 12:44:37 - INFO - parrotllm.training - step 250 | epoch 0 | loss 5.5036 | lr 1.07e-03 | grad 0.4908
198
+ 2026-04-08 12:44:38 - INFO - parrotllm.training - Saved checkpoint: /home/user/code/ParrotLLM/runs/big_run/exp_c/run_20260408_124138/checkpoints/last_epoch_0000_step_0000250.pt
199
+ 2026-04-08 12:44:42 - INFO - parrotllm.training - step 260 | epoch 0 | loss 5.5156 | lr 1.11e-03 | grad 0.5607
200
+ 2026-04-08 12:44:46 - INFO - parrotllm.training - step 270 | epoch 0 | loss 5.5554 | lr 1.16e-03 | grad 0.4806
201
+ 2026-04-08 12:44:49 - INFO - parrotllm.training - step 280 | epoch 0 | loss 5.4411 | lr 1.20e-03 | grad 0.4974
202
+ 2026-04-08 12:44:52 - INFO - parrotllm.training - step 290 | epoch 0 | loss 5.4694 | lr 1.24e-03 | grad 1.0310
203
+ 2026-04-08 12:44:55 - INFO - parrotllm.training - step 300 | epoch 0 | loss 5.4289 | lr 1.28e-03 | grad 0.4870
204
+ 2026-04-08 12:44:55 - INFO - parrotllm.training - Starting evaluation...
205
+ 2026-04-08 12:44:55 - INFO - parrotllm.training - ------------------------------------------------------------
206
+ 2026-04-08 12:45:00 - INFO - parrotllm.training - Train: loss=5.4289, ppl=227.90
207
+ 2026-04-08 12:45:00 - INFO - parrotllm.training - Val: loss=5.3999, ppl=221.38
208
+ 2026-04-08 12:45:00 - INFO - parrotllm.training - ** New best validation loss! **
209
+ 2026-04-08 12:45:01 - INFO - parrotllm.training - Saved best checkpoint: /home/user/code/ParrotLLM/runs/big_run/exp_c/run_20260408_124138/checkpoints/best_loss_5p3999_epoch_0000_step_0000300.pt
210
+ 2026-04-08 12:45:02 - INFO - parrotllm.training - ------------------------------------------------------------
211
+ 2026-04-08 12:45:05 - INFO - parrotllm.training - step 310 | epoch 0 | loss 5.3706 | lr 1.30e-03 | grad 0.3568
212
+ 2026-04-08 12:45:08 - INFO - parrotllm.training - step 320 | epoch 0 | loss 5.2355 | lr 1.30e-03 | grad 0.3656
213
+ 2026-04-08 12:45:11 - INFO - parrotllm.training - step 330 | epoch 0 | loss 5.1612 | lr 1.30e-03 | grad 0.4844
214
+ 2026-04-08 12:45:14 - INFO - parrotllm.training - step 340 | epoch 0 | loss 5.2862 | lr 1.30e-03 | grad 0.4443
215
+ 2026-04-08 12:45:17 - INFO - parrotllm.training - step 350 | epoch 0 | loss 4.9924 | lr 1.30e-03 | grad 0.2914
216
+ 2026-04-08 12:45:20 - INFO - parrotllm.training - step 360 | epoch 0 | loss 5.2429 | lr 1.30e-03 | grad 0.7404
217
+ 2026-04-08 12:45:24 - INFO - parrotllm.training - step 370 | epoch 0 | loss 5.0984 | lr 1.30e-03 | grad 0.4491
218
+ 2026-04-08 12:45:27 - INFO - parrotllm.training - step 380 | epoch 0 | loss 5.0499 | lr 1.30e-03 | grad 0.4170
219
+ 2026-04-08 12:45:30 - INFO - parrotllm.training - step 390 | epoch 0 | loss 4.9054 | lr 1.30e-03 | grad 0.2777
220
+ 2026-04-08 12:45:33 - INFO - parrotllm.training - step 400 | epoch 0 | loss 5.0143 | lr 1.30e-03 | grad 0.4620
221
+ 2026-04-08 12:45:33 - INFO - parrotllm.training - Starting evaluation...
222
+ 2026-04-08 12:45:33 - INFO - parrotllm.training - ------------------------------------------------------------
223
+ 2026-04-08 12:45:38 - INFO - parrotllm.training - Train: loss=5.0143, ppl=150.55
224
+ 2026-04-08 12:45:38 - INFO - parrotllm.training - Val: loss=4.9547, ppl=141.85
225
+ 2026-04-08 12:45:38 - INFO - parrotllm.training - ** New best validation loss! **
226
+ 2026-04-08 12:45:39 - INFO - parrotllm.training - Saved best checkpoint: /home/user/code/ParrotLLM/runs/big_run/exp_c/run_20260408_124138/checkpoints/best_loss_4p9547_epoch_0000_step_0000400.pt
227
+ 2026-04-08 12:45:40 - INFO - parrotllm.training - ------------------------------------------------------------
228
+ 2026-04-08 12:45:43 - INFO - parrotllm.training - step 410 | epoch 0 | loss 4.9213 | lr 1.30e-03 | grad 0.3111
229
+ 2026-04-08 12:45:46 - INFO - parrotllm.training - step 420 | epoch 0 | loss 4.9263 | lr 1.30e-03 | grad 0.2881
230
+ 2026-04-08 12:45:49 - INFO - parrotllm.training - step 430 | epoch 0 | loss 4.8843 | lr 1.30e-03 | grad 0.4029
231
+ 2026-04-08 12:45:52 - INFO - parrotllm.training - step 440 | epoch 0 | loss 4.7801 | lr 1.30e-03 | grad 0.3063
232
+ 2026-04-08 12:45:55 - INFO - parrotllm.training - step 450 | epoch 0 | loss 4.7493 | lr 1.30e-03 | grad 0.3415
233
+ 2026-04-08 12:45:58 - INFO - parrotllm.training - step 460 | epoch 0 | loss 4.7937 | lr 1.30e-03 | grad 0.3799
234
+ 2026-04-08 12:46:01 - INFO - parrotllm.training - step 470 | epoch 0 | loss 4.7403 | lr 1.30e-03 | grad 0.6610
235
+ 2026-04-08 12:46:04 - INFO - parrotllm.training - step 480 | epoch 0 | loss 4.7418 | lr 1.30e-03 | grad 0.3100
236
+ 2026-04-08 12:46:07 - INFO - parrotllm.training - step 490 | epoch 0 | loss 4.7975 | lr 1.30e-03 | grad 0.4635
237
+ 2026-04-08 12:46:11 - INFO - parrotllm.training - step 500 | epoch 0 | loss 4.6381 | lr 1.30e-03 | grad 0.3594
238
+ 2026-04-08 12:46:11 - INFO - parrotllm.training - Starting evaluation...
239
+ 2026-04-08 12:46:11 - INFO - parrotllm.training - ------------------------------------------------------------
240
+ 2026-04-08 12:46:16 - INFO - parrotllm.training - Train: loss=4.6381, ppl=103.34
241
+ 2026-04-08 12:46:16 - INFO - parrotllm.training - Val: loss=4.6275, ppl=102.26
242
+ 2026-04-08 12:46:16 - INFO - parrotllm.training - ** New best validation loss! **
243
+ 2026-04-08 12:46:17 - INFO - parrotllm.training - Saved best checkpoint: /home/user/code/ParrotLLM/runs/big_run/exp_c/run_20260408_124138/checkpoints/best_loss_4p6275_epoch_0000_step_0000500.pt
244
+ 2026-04-08 12:46:18 - INFO - parrotllm.training - ------------------------------------------------------------
245
+ 2026-04-08 12:46:19 - INFO - parrotllm.training - Saved checkpoint: /home/user/code/ParrotLLM/runs/big_run/exp_c/run_20260408_124138/checkpoints/last_epoch_0000_step_0000500.pt
246
+ 2026-04-08 12:46:23 - INFO - parrotllm.training - step 510 | epoch 0 | loss 4.6129 | lr 1.30e-03 | grad 0.3387
247
+ 2026-04-08 12:46:26 - INFO - parrotllm.training - step 520 | epoch 0 | loss 4.6002 | lr 1.30e-03 | grad 0.2965
248
+ 2026-04-08 12:46:29 - INFO - parrotllm.training - step 530 | epoch 0 | loss 4.6989 | lr 1.30e-03 | grad 0.4270
249
+ 2026-04-08 12:46:32 - INFO - parrotllm.training - step 540 | epoch 0 | loss 4.4593 | lr 1.30e-03 | grad 0.3768
250
+ 2026-04-08 12:46:35 - INFO - parrotllm.training - step 550 | epoch 0 | loss 4.4864 | lr 1.30e-03 | grad 0.3919
251
+ 2026-04-08 12:46:38 - INFO - parrotllm.training - step 560 | epoch 0 | loss 4.5896 | lr 1.30e-03 | grad 0.3289
252
+ 2026-04-08 12:46:41 - INFO - parrotllm.training - step 570 | epoch 0 | loss 4.5558 | lr 1.30e-03 | grad 0.5545
253
+ 2026-04-08 12:46:44 - INFO - parrotllm.training - step 580 | epoch 0 | loss 4.4624 | lr 1.30e-03 | grad 0.2772
254
+ 2026-04-08 12:46:47 - INFO - parrotllm.training - step 590 | epoch 0 | loss 4.4730 | lr 1.30e-03 | grad 0.2939
255
+ 2026-04-08 12:46:50 - INFO - parrotllm.training - step 600 | epoch 0 | loss 4.5336 | lr 1.30e-03 | grad 0.2799
256
+ 2026-04-08 12:46:50 - INFO - parrotllm.training - Starting evaluation...
257
+ 2026-04-08 12:46:50 - INFO - parrotllm.training - ------------------------------------------------------------
258
+ 2026-04-08 12:46:56 - INFO - parrotllm.training - Train: loss=4.5336, ppl=93.09
259
+ 2026-04-08 12:46:56 - INFO - parrotllm.training - Val: loss=4.3700, ppl=79.05
260
+ 2026-04-08 12:46:56 - INFO - parrotllm.training - ** New best validation loss! **
261
+ 2026-04-08 12:46:57 - INFO - parrotllm.training - Saved best checkpoint: /home/user/code/ParrotLLM/runs/big_run/exp_c/run_20260408_124138/checkpoints/best_loss_4p3700_epoch_0000_step_0000600.pt
262
+ 2026-04-08 12:46:58 - INFO - parrotllm.training - ------------------------------------------------------------
263
+ 2026-04-08 12:47:01 - INFO - parrotllm.training - step 610 | epoch 0 | loss 4.4782 | lr 1.30e-03 | grad 0.4029
264
+ 2026-04-08 12:47:04 - INFO - parrotllm.training - step 620 | epoch 0 | loss 4.4498 | lr 1.30e-03 | grad 0.3437
265
+ 2026-04-08 12:47:07 - INFO - parrotllm.training - step 630 | epoch 0 | loss 4.3908 | lr 1.30e-03 | grad 0.4935
266
+ 2026-04-08 12:47:10 - INFO - parrotllm.training - step 640 | epoch 0 | loss 4.3158 | lr 1.30e-03 | grad 0.2801
267
+ 2026-04-08 12:47:13 - INFO - parrotllm.training - step 650 | epoch 0 | loss 4.3310 | lr 1.30e-03 | grad 0.3586
268
+ 2026-04-08 12:47:16 - INFO - parrotllm.training - step 660 | epoch 0 | loss 4.1952 | lr 1.30e-03 | grad 0.3386
269
+ 2026-04-08 12:47:19 - INFO - parrotllm.training - step 670 | epoch 0 | loss 4.3220 | lr 1.30e-03 | grad 0.2795
270
+ 2026-04-08 12:47:22 - INFO - parrotllm.training - step 680 | epoch 0 | loss 4.3062 | lr 1.30e-03 | grad 0.3723
271
+ 2026-04-08 12:47:26 - INFO - parrotllm.training - step 690 | epoch 0 | loss 4.2446 | lr 1.30e-03 | grad 0.3151
272
+ 2026-04-08 12:47:29 - INFO - parrotllm.training - step 700 | epoch 0 | loss 4.2909 | lr 1.30e-03 | grad 0.4253
273
+ 2026-04-08 12:47:29 - INFO - parrotllm.training - Starting evaluation...
274
+ 2026-04-08 12:47:29 - INFO - parrotllm.training - ------------------------------------------------------------
275
+ 2026-04-08 12:47:34 - INFO - parrotllm.training - Train: loss=4.2909, ppl=73.03
276
+ 2026-04-08 12:47:34 - INFO - parrotllm.training - Val: loss=4.2423, ppl=69.57
277
+ 2026-04-08 12:47:34 - INFO - parrotllm.training - ** New best validation loss! **
278
+ 2026-04-08 12:47:35 - INFO - parrotllm.training - Saved best checkpoint: /home/user/code/ParrotLLM/runs/big_run/exp_c/run_20260408_124138/checkpoints/best_loss_4p2423_epoch_0000_step_0000700.pt
279
+ 2026-04-08 12:47:36 - INFO - parrotllm.training - ------------------------------------------------------------
280
+ 2026-04-08 12:47:39 - INFO - parrotllm.training - step 710 | epoch 0 | loss 4.2638 | lr 1.30e-03 | grad 0.2893
281
+ 2026-04-08 12:47:42 - INFO - parrotllm.training - step 720 | epoch 0 | loss 4.3053 | lr 1.30e-03 | grad 0.3133
282
+ 2026-04-08 12:47:45 - INFO - parrotllm.training - step 730 | epoch 0 | loss 4.2717 | lr 1.30e-03 | grad 0.3125
283
+ 2026-04-08 12:47:48 - INFO - parrotllm.training - step 740 | epoch 0 | loss 4.3364 | lr 1.30e-03 | grad 0.3499
284
+ 2026-04-08 12:47:51 - INFO - parrotllm.training - step 750 | epoch 0 | loss 4.2178 | lr 1.30e-03 | grad 0.4627
285
+ 2026-04-08 12:47:52 - INFO - parrotllm.training - Saved checkpoint: /home/user/code/ParrotLLM/runs/big_run/exp_c/run_20260408_124138/checkpoints/last_epoch_0000_step_0000750.pt
286
+ 2026-04-08 12:47:56 - INFO - parrotllm.training - step 760 | epoch 0 | loss 4.1247 | lr 1.30e-03 | grad 0.2749
287
+ 2026-04-08 12:48:00 - INFO - parrotllm.training - step 770 | epoch 0 | loss 4.3647 | lr 1.30e-03 | grad 0.2520
288
+ 2026-04-08 12:48:03 - INFO - parrotllm.training - step 780 | epoch 0 | loss 4.2694 | lr 1.30e-03 | grad 0.2739
289
+ 2026-04-08 12:48:06 - INFO - parrotllm.training - step 790 | epoch 0 | loss 4.2489 | lr 1.30e-03 | grad 0.3426
290
+ 2026-04-08 12:48:09 - INFO - parrotllm.training - step 800 | epoch 0 | loss 4.2317 | lr 1.30e-03 | grad 0.3093
291
+ 2026-04-08 12:48:09 - INFO - parrotllm.training - Starting evaluation...
292
+ 2026-04-08 12:48:09 - INFO - parrotllm.training - ------------------------------------------------------------
293
+ 2026-04-08 12:48:15 - INFO - parrotllm.training - Train: loss=4.2317, ppl=68.83
294
+ 2026-04-08 12:48:15 - INFO - parrotllm.training - Val: loss=4.1199, ppl=61.55
295
+ 2026-04-08 12:48:15 - INFO - parrotllm.training - ** New best validation loss! **
296
+ 2026-04-08 12:48:15 - INFO - parrotllm.training - Saved best checkpoint: /home/user/code/ParrotLLM/runs/big_run/exp_c/run_20260408_124138/checkpoints/best_loss_4p1199_epoch_0000_step_0000800.pt
297
+ 2026-04-08 12:48:16 - INFO - parrotllm.training - ------------------------------------------------------------
298
+ 2026-04-08 12:48:19 - INFO - parrotllm.training - step 810 | epoch 0 | loss 4.1372 | lr 1.30e-03 | grad 0.2648
299
+ 2026-04-08 12:48:22 - INFO - parrotllm.training - step 820 | epoch 0 | loss 4.1945 | lr 1.30e-03 | grad 0.5349
300
+ 2026-04-08 12:48:25 - INFO - parrotllm.training - step 830 | epoch 0 | loss 4.1505 | lr 1.30e-03 | grad 0.2663
301
+ 2026-04-08 12:48:28 - INFO - parrotllm.training - step 840 | epoch 0 | loss 4.2378 | lr 1.30e-03 | grad 0.3815
302
+ 2026-04-08 12:48:32 - INFO - parrotllm.training - step 850 | epoch 0 | loss 4.1529 | lr 1.30e-03 | grad 0.2902
303
+ 2026-04-08 12:48:35 - INFO - parrotllm.training - step 860 | epoch 0 | loss 4.1264 | lr 1.30e-03 | grad 0.3139
304
+ 2026-04-08 12:48:38 - INFO - parrotllm.training - step 870 | epoch 0 | loss 4.1175 | lr 1.30e-03 | grad 0.3062
305
+ 2026-04-08 12:48:41 - INFO - parrotllm.training - step 880 | epoch 0 | loss 4.0828 | lr 1.30e-03 | grad 0.2848
306
+ 2026-04-08 12:48:44 - INFO - parrotllm.training - step 890 | epoch 0 | loss 4.0311 | lr 1.30e-03 | grad 0.3449
307
+ 2026-04-08 12:48:47 - INFO - parrotllm.training - step 900 | epoch 0 | loss 4.0856 | lr 1.30e-03 | grad 0.2898
308
+ 2026-04-08 12:48:47 - INFO - parrotllm.training - Starting evaluation...
309
+ 2026-04-08 12:48:47 - INFO - parrotllm.training - ------------------------------------------------------------
310
+ 2026-04-08 12:48:53 - INFO - parrotllm.training - Train: loss=4.0856, ppl=59.47
311
+ 2026-04-08 12:48:53 - INFO - parrotllm.training - Val: loss=4.0408, ppl=56.87
312
+ 2026-04-08 12:48:53 - INFO - parrotllm.training - ** New best validation loss! **
313
+ 2026-04-08 12:48:53 - INFO - parrotllm.training - Saved best checkpoint: /home/user/code/ParrotLLM/runs/big_run/exp_c/run_20260408_124138/checkpoints/best_loss_4p0408_epoch_0000_step_0000900.pt
314
+ 2026-04-08 12:48:54 - INFO - parrotllm.training - ------------------------------------------------------------
315
+ 2026-04-08 12:48:57 - INFO - parrotllm.training - step 910 | epoch 0 | loss 4.0652 | lr 1.30e-03 | grad 0.2460
316
+ 2026-04-08 12:49:00 - INFO - parrotllm.training - step 920 | epoch 0 | loss 4.1638 | lr 1.30e-03 | grad 0.3769
317
+ 2026-04-08 12:49:03 - INFO - parrotllm.training - step 930 | epoch 0 | loss 4.0077 | lr 1.30e-03 | grad 0.2636
318
+ 2026-04-08 12:49:06 - INFO - parrotllm.training - step 940 | epoch 0 | loss 4.0963 | lr 1.30e-03 | grad 0.2925
319
+ 2026-04-08 12:49:09 - INFO - parrotllm.training - step 950 | epoch 0 | loss 4.0018 | lr 1.30e-03 | grad 0.2907
320
+ 2026-04-08 12:49:12 - INFO - parrotllm.training - step 960 | epoch 0 | loss 4.0079 | lr 1.30e-03 | grad 0.3901
321
+ 2026-04-08 12:49:15 - INFO - parrotllm.training - step 970 | epoch 0 | loss 4.0677 | lr 1.30e-03 | grad 0.2483
322
+ 2026-04-08 12:49:19 - INFO - parrotllm.training - step 980 | epoch 0 | loss 4.1012 | lr 1.30e-03 | grad 0.2528
323
+ 2026-04-08 12:49:22 - INFO - parrotllm.training - step 990 | epoch 0 | loss 4.0414 | lr 1.30e-03 | grad 0.2952
324
+ 2026-04-08 12:49:25 - INFO - parrotllm.training - step 1000 | epoch 0 | loss 4.0708 | lr 1.30e-03 | grad 0.3298
325
+ 2026-04-08 12:49:25 - INFO - parrotllm.training - Starting evaluation...
326
+ 2026-04-08 12:49:25 - INFO - parrotllm.training - ------------------------------------------------------------
327
+ 2026-04-08 12:49:30 - INFO - parrotllm.training - Train: loss=4.0708, ppl=58.61
328
+ 2026-04-08 12:49:30 - INFO - parrotllm.training - Val: loss=3.9807, ppl=53.56
329
+ 2026-04-08 12:49:30 - INFO - parrotllm.training - ** New best validation loss! **
330
+ 2026-04-08 12:49:31 - INFO - parrotllm.training - Saved best checkpoint: /home/user/code/ParrotLLM/runs/big_run/exp_c/run_20260408_124138/checkpoints/best_loss_3p9807_epoch_0000_step_0001000.pt
331
+ 2026-04-08 12:49:32 - INFO - parrotllm.training - ------------------------------------------------------------
332
+ 2026-04-08 12:49:33 - INFO - parrotllm.training - Saved checkpoint: /home/user/code/ParrotLLM/runs/big_run/exp_c/run_20260408_124138/checkpoints/last_epoch_0000_step_0001000.pt
333
+ 2026-04-08 12:49:37 - INFO - parrotllm.training - step 1010 | epoch 0 | loss 4.0479 | lr 1.30e-03 | grad 0.2914
334
+ 2026-04-08 12:49:40 - INFO - parrotllm.training - step 1020 | epoch 0 | loss 3.9430 | lr 1.30e-03 | grad 0.2406
335
+ 2026-04-08 12:49:43 - INFO - parrotllm.training - step 1030 | epoch 0 | loss 3.8865 | lr 1.30e-03 | grad 0.3132
336
+ 2026-04-08 12:49:46 - INFO - parrotllm.training - step 1040 | epoch 0 | loss 4.0676 | lr 1.30e-03 | grad 0.2750
337
+ 2026-04-08 12:49:49 - INFO - parrotllm.training - step 1050 | epoch 0 | loss 3.9435 | lr 1.30e-03 | grad 0.3234
338
+ 2026-04-08 12:49:52 - INFO - parrotllm.training - step 1060 | epoch 0 | loss 4.0356 | lr 1.30e-03 | grad 0.2918
339
+ 2026-04-08 12:49:55 - INFO - parrotllm.training - step 1070 | epoch 0 | loss 3.9681 | lr 1.30e-03 | grad 0.3066
340
+ 2026-04-08 12:49:58 - INFO - parrotllm.training - step 1080 | epoch 0 | loss 4.0446 | lr 1.30e-03 | grad 0.2816
341
+ 2026-04-08 12:50:01 - INFO - parrotllm.training - step 1090 | epoch 0 | loss 3.9623 | lr 1.30e-03 | grad 0.2621
342
+ 2026-04-08 12:50:04 - INFO - parrotllm.training - step 1100 | epoch 0 | loss 3.9303 | lr 1.30e-03 | grad 0.3440
343
+ 2026-04-08 12:50:04 - INFO - parrotllm.training - Starting evaluation...
344
+ 2026-04-08 12:50:04 - INFO - parrotllm.training - ------------------------------------------------------------
345
+ 2026-04-08 12:50:10 - INFO - parrotllm.training - Train: loss=3.9303, ppl=50.92
346
+ 2026-04-08 12:50:10 - INFO - parrotllm.training - Val: loss=3.9380, ppl=51.32
347
+ 2026-04-08 12:50:10 - INFO - parrotllm.training - ** New best validation loss! **
348
+ 2026-04-08 12:50:11 - INFO - parrotllm.training - Saved best checkpoint: /home/user/code/ParrotLLM/runs/big_run/exp_c/run_20260408_124138/checkpoints/best_loss_3p9380_epoch_0000_step_0001100.pt
349
+ 2026-04-08 12:50:12 - INFO - parrotllm.training - ------------------------------------------------------------
350
+ 2026-04-08 12:50:15 - INFO - parrotllm.training - step 1110 | epoch 0 | loss 3.8959 | lr 1.30e-03 | grad 0.2816
351
+ 2026-04-08 12:50:18 - INFO - parrotllm.training - step 1120 | epoch 0 | loss 4.0434 | lr 1.30e-03 | grad 0.2867
352
+ 2026-04-08 12:50:21 - INFO - parrotllm.training - step 1130 | epoch 0 | loss 3.9461 | lr 1.30e-03 | grad 0.2710
353
+ 2026-04-08 12:50:24 - INFO - parrotllm.training - step 1140 | epoch 0 | loss 4.0211 | lr 1.30e-03 | grad 0.3127
354
+ 2026-04-08 12:50:27 - INFO - parrotllm.training - step 1150 | epoch 0 | loss 3.8917 | lr 1.30e-03 | grad 0.2831
355
+ 2026-04-08 12:50:30 - INFO - parrotllm.training - step 1160 | epoch 0 | loss 4.0191 | lr 1.30e-03 | grad 0.2494
356
+ 2026-04-08 12:50:33 - INFO - parrotllm.training - step 1170 | epoch 0 | loss 3.9068 | lr 1.30e-03 | grad 0.2486
357
+ 2026-04-08 12:50:36 - INFO - parrotllm.training - step 1180 | epoch 0 | loss 3.8646 | lr 1.30e-03 | grad 0.2616
358
+ 2026-04-08 12:50:39 - INFO - parrotllm.training - step 1190 | epoch 0 | loss 3.9071 | lr 1.30e-03 | grad 0.2977
359
+ 2026-04-08 12:50:42 - INFO - parrotllm.training - step 1200 | epoch 0 | loss 3.9299 | lr 1.30e-03 | grad 0.3024
360
+ 2026-04-08 12:50:42 - INFO - parrotllm.training - Starting evaluation...
361
+ 2026-04-08 12:50:42 - INFO - parrotllm.training - ------------------------------------------------------------
362
+ 2026-04-08 12:50:48 - INFO - parrotllm.training - Train: loss=3.9299, ppl=50.90
363
+ 2026-04-08 12:50:48 - INFO - parrotllm.training - Val: loss=3.8945, ppl=49.13
364
+ 2026-04-08 12:50:48 - INFO - parrotllm.training - ** New best validation loss! **
365
+ 2026-04-08 12:50:49 - INFO - parrotllm.training - Saved best checkpoint: /home/user/code/ParrotLLM/runs/big_run/exp_c/run_20260408_124138/checkpoints/best_loss_3p8945_epoch_0000_step_0001200.pt
366
+ 2026-04-08 12:50:50 - INFO - parrotllm.training - ------------------------------------------------------------
367
+ 2026-04-08 12:50:53 - INFO - parrotllm.training - step 1210 | epoch 0 | loss 3.8974 | lr 1.30e-03 | grad 0.2758
368
+ 2026-04-08 12:50:56 - INFO - parrotllm.training - step 1220 | epoch 0 | loss 3.8784 | lr 1.30e-03 | grad 0.2543
369
+ 2026-04-08 12:50:59 - INFO - parrotllm.training - step 1230 | epoch 0 | loss 3.9110 | lr 1.30e-03 | grad 0.2539
370
+ 2026-04-08 12:51:02 - INFO - parrotllm.training - step 1240 | epoch 0 | loss 3.9338 | lr 1.30e-03 | grad 0.2678
371
+ 2026-04-08 12:51:05 - INFO - parrotllm.training - step 1250 | epoch 0 | loss 3.8986 | lr 1.30e-03 | grad 0.3379
372
+ 2026-04-08 12:51:06 - INFO - parrotllm.training - Saved checkpoint: /home/user/code/ParrotLLM/runs/big_run/exp_c/run_20260408_124138/checkpoints/last_epoch_0000_step_0001250.pt
373
+ 2026-04-08 12:51:10 - INFO - parrotllm.training - step 1260 | epoch 0 | loss 3.8480 | lr 1.30e-03 | grad 0.3182
374
+ 2026-04-08 12:51:14 - INFO - parrotllm.training - step 1270 | epoch 0 | loss 3.9598 | lr 1.30e-03 | grad 0.2720
375
+ 2026-04-08 12:51:17 - INFO - parrotllm.training - step 1280 | epoch 0 | loss 3.9995 | lr 1.30e-03 | grad 0.2284
376
+ 2026-04-08 12:51:20 - INFO - parrotllm.training - step 1290 | epoch 0 | loss 3.8092 | lr 1.30e-03 | grad 0.2447
377
+ 2026-04-08 12:51:23 - INFO - parrotllm.training - step 1300 | epoch 0 | loss 3.8669 | lr 1.30e-03 | grad 0.3316
378
+ 2026-04-08 12:51:23 - INFO - parrotllm.training - Starting evaluation...
379
+ 2026-04-08 12:51:23 - INFO - parrotllm.training - ------------------------------------------------------------
380
+ 2026-04-08 12:51:29 - INFO - parrotllm.training - Train: loss=3.8669, ppl=47.79
381
+ 2026-04-08 12:51:29 - INFO - parrotllm.training - Val: loss=3.8603, ppl=47.48
382
+ 2026-04-08 12:51:29 - INFO - parrotllm.training - ** New best validation loss! **
383
+ 2026-04-08 12:51:29 - INFO - parrotllm.training - Saved best checkpoint: /home/user/code/ParrotLLM/runs/big_run/exp_c/run_20260408_124138/checkpoints/best_loss_3p8603_epoch_0000_step_0001300.pt
384
+ 2026-04-08 12:51:30 - INFO - parrotllm.training - ------------------------------------------------------------
385
+ 2026-04-08 12:51:33 - INFO - parrotllm.training - step 1310 | epoch 0 | loss 3.8566 | lr 1.30e-03 | grad 0.3084
386
+ 2026-04-08 12:51:36 - INFO - parrotllm.training - step 1320 | epoch 0 | loss 3.9365 | lr 1.30e-03 | grad 0.2760
387
+ 2026-04-08 12:51:39 - INFO - parrotllm.training - step 1330 | epoch 0 | loss 3.9222 | lr 1.30e-03 | grad 0.3011
388
+ 2026-04-08 12:51:42 - INFO - parrotllm.training - step 1340 | epoch 0 | loss 3.9364 | lr 1.30e-03 | grad 0.2259
389
+ 2026-04-08 12:51:45 - INFO - parrotllm.training - step 1350 | epoch 0 | loss 3.8879 | lr 1.30e-03 | grad 0.2769
390
+ 2026-04-08 12:51:49 - INFO - parrotllm.training - step 1360 | epoch 0 | loss 3.9090 | lr 1.30e-03 | grad 0.2422
391
+ 2026-04-08 12:51:52 - INFO - parrotllm.training - step 1370 | epoch 0 | loss 3.9005 | lr 1.30e-03 | grad 0.2232
392
+ 2026-04-08 12:51:55 - INFO - parrotllm.training - step 1380 | epoch 0 | loss 3.8989 | lr 1.30e-03 | grad 0.2461
393
+ 2026-04-08 12:51:58 - INFO - parrotllm.training - step 1390 | epoch 0 | loss 3.8862 | lr 1.30e-03 | grad 0.2708
394
+ 2026-04-08 12:52:01 - INFO - parrotllm.training - step 1400 | epoch 0 | loss 3.9330 | lr 1.30e-03 | grad 0.2284
395
+ 2026-04-08 12:52:01 - INFO - parrotllm.training - Starting evaluation...
396
+ 2026-04-08 12:52:01 - INFO - parrotllm.training - ------------------------------------------------------------
397
+ 2026-04-08 12:52:07 - INFO - parrotllm.training - Train: loss=3.9330, ppl=51.06
398
+ 2026-04-08 12:52:07 - INFO - parrotllm.training - Val: loss=3.8274, ppl=45.94
399
+ 2026-04-08 12:52:07 - INFO - parrotllm.training - ** New best validation loss! **
400
+ 2026-04-08 12:52:08 - INFO - parrotllm.training - Saved best checkpoint: /home/user/code/ParrotLLM/runs/big_run/exp_c/run_20260408_124138/checkpoints/best_loss_3p8274_epoch_0000_step_0001400.pt
401
+ 2026-04-08 12:52:08 - INFO - parrotllm.training - ------------------------------------------------------------
402
+ 2026-04-08 12:52:12 - INFO - parrotllm.training - step 1410 | epoch 0 | loss 3.8773 | lr 1.30e-03 | grad 0.2665
403
+ 2026-04-08 12:52:15 - INFO - parrotllm.training - step 1420 | epoch 0 | loss 3.9158 | lr 1.30e-03 | grad 0.2710
404
+ 2026-04-08 12:52:18 - INFO - parrotllm.training - step 1430 | epoch 0 | loss 3.8370 | lr 1.30e-03 | grad 0.2496
405
+ 2026-04-08 12:52:21 - INFO - parrotllm.training - step 1440 | epoch 0 | loss 3.8494 | lr 1.30e-03 | grad 0.2908
406
+ 2026-04-08 12:52:24 - INFO - parrotllm.training - step 1450 | epoch 0 | loss 3.8176 | lr 1.30e-03 | grad 0.2381
407
+ 2026-04-08 12:52:27 - INFO - parrotllm.training - step 1460 | epoch 0 | loss 3.9117 | lr 1.30e-03 | grad 0.2631
408
+ 2026-04-08 12:52:30 - INFO - parrotllm.training - step 1470 | epoch 0 | loss 3.8423 | lr 1.30e-03 | grad 0.3623
409
+ 2026-04-08 12:52:33 - INFO - parrotllm.training - step 1480 | epoch 0 | loss 3.8607 | lr 1.30e-03 | grad 0.2508
410
+ 2026-04-08 12:52:36 - INFO - parrotllm.training - step 1490 | epoch 0 | loss 3.7899 | lr 1.30e-03 | grad 0.2387
411
+ 2026-04-08 12:52:39 - INFO - parrotllm.training - step 1500 | epoch 0 | loss 3.7814 | lr 1.30e-03 | grad 0.2772
412
+ 2026-04-08 12:52:39 - INFO - parrotllm.training - Starting evaluation...
413
+ 2026-04-08 12:52:39 - INFO - parrotllm.training - ------------------------------------------------------------
414
+ 2026-04-08 12:52:45 - INFO - parrotllm.training - Train: loss=3.7814, ppl=43.88
415
+ 2026-04-08 12:52:45 - INFO - parrotllm.training - Val: loss=3.8074, ppl=45.03
416
+ 2026-04-08 12:52:45 - INFO - parrotllm.training - ** New best validation loss! **
417
+ 2026-04-08 12:52:46 - INFO - parrotllm.training - Saved best checkpoint: /home/user/code/ParrotLLM/runs/big_run/exp_c/run_20260408_124138/checkpoints/best_loss_3p8074_epoch_0000_step_0001500.pt
418
+ 2026-04-08 12:52:47 - INFO - parrotllm.training - ------------------------------------------------------------
419
+ 2026-04-08 12:52:48 - INFO - parrotllm.training - Saved checkpoint: /home/user/code/ParrotLLM/runs/big_run/exp_c/run_20260408_124138/checkpoints/last_epoch_0000_step_0001500.pt
420
+ 2026-04-08 12:52:52 - INFO - parrotllm.training - step 1510 | epoch 0 | loss 3.8355 | lr 1.30e-03 | grad 0.2602
421
+ 2026-04-08 12:52:55 - INFO - parrotllm.training - step 1520 | epoch 0 | loss 3.8465 | lr 1.30e-03 | grad 0.2622
422
+ 2026-04-08 12:52:58 - INFO - parrotllm.training - step 1530 | epoch 0 | loss 3.9231 | lr 1.30e-03 | grad 0.2776
423
+ 2026-04-08 12:53:01 - INFO - parrotllm.training - step 1540 | epoch 0 | loss 3.7939 | lr 1.30e-03 | grad 0.2384
424
+ 2026-04-08 12:53:04 - INFO - parrotllm.training - step 1550 | epoch 0 | loss 3.7998 | lr 1.30e-03 | grad 0.2477
425
+ 2026-04-08 12:53:07 - INFO - parrotllm.training - step 1560 | epoch 0 | loss 3.8265 | lr 1.30e-03 | grad 0.2496
426
+ 2026-04-08 12:53:10 - INFO - parrotllm.training - step 1570 | epoch 0 | loss 3.8365 | lr 1.30e-03 | grad 0.2401
427
+ 2026-04-08 12:53:13 - INFO - parrotllm.training - step 1580 | epoch 0 | loss 3.8086 | lr 1.30e-03 | grad 0.2788
428
+ 2026-04-08 12:53:16 - INFO - parrotllm.training - step 1590 | epoch 0 | loss 3.7861 | lr 1.30e-03 | grad 0.2817
429
+ 2026-04-08 12:53:19 - INFO - parrotllm.training - step 1600 | epoch 0 | loss 3.7310 | lr 1.30e-03 | grad 0.2718
430
+ 2026-04-08 12:53:19 - INFO - parrotllm.training - Starting evaluation...
431
+ 2026-04-08 12:53:19 - INFO - parrotllm.training - ------------------------------------------------------------
432
+ 2026-04-08 12:53:25 - INFO - parrotllm.training - Train: loss=3.7310, ppl=41.72
433
+ 2026-04-08 12:53:25 - INFO - parrotllm.training - Val: loss=3.7825, ppl=43.93
434
+ 2026-04-08 12:53:25 - INFO - parrotllm.training - ** New best validation loss! **
435
+ 2026-04-08 12:53:26 - INFO - parrotllm.training - Saved best checkpoint: /home/user/code/ParrotLLM/runs/big_run/exp_c/run_20260408_124138/checkpoints/best_loss_3p7825_epoch_0000_step_0001600.pt
436
+ 2026-04-08 12:53:27 - INFO - parrotllm.training - ------------------------------------------------------------
437
+ 2026-04-08 12:53:30 - INFO - parrotllm.training - step 1610 | epoch 0 | loss 3.7867 | lr 1.30e-03 | grad 0.2506
438
+ 2026-04-08 12:53:33 - INFO - parrotllm.training - step 1620 | epoch 0 | loss 3.7436 | lr 1.30e-03 | grad 0.2619
439
+ 2026-04-08 12:53:36 - INFO - parrotllm.training - step 1630 | epoch 0 | loss 3.9659 | lr 1.30e-03 | grad 0.2411
440
+ 2026-04-08 12:53:39 - INFO - parrotllm.training - step 1640 | epoch 0 | loss 3.7774 | lr 1.30e-03 | grad 0.2585
441
+ 2026-04-08 12:53:42 - INFO - parrotllm.training - step 1650 | epoch 0 | loss 3.8461 | lr 1.30e-03 | grad 0.3146
442
+ 2026-04-08 12:53:45 - INFO - parrotllm.training - step 1660 | epoch 0 | loss 3.8304 | lr 1.30e-03 | grad 0.2715
443
+ 2026-04-08 12:53:48 - INFO - parrotllm.training - step 1670 | epoch 0 | loss 3.8136 | lr 1.30e-03 | grad 0.2449
444
+ 2026-04-08 12:53:51 - INFO - parrotllm.training - step 1680 | epoch 0 | loss 3.8394 | lr 1.30e-03 | grad 0.2479
445
+ 2026-04-08 12:53:54 - INFO - parrotllm.training - step 1690 | epoch 0 | loss 3.7196 | lr 1.30e-03 | grad 0.2394
446
+ 2026-04-08 12:53:57 - INFO - parrotllm.training - step 1700 | epoch 0 | loss 3.7920 | lr 1.30e-03 | grad 0.2520
447
+ 2026-04-08 12:53:57 - INFO - parrotllm.training - Starting evaluation...
448
+ 2026-04-08 12:53:57 - INFO - parrotllm.training - ------------------------------------------------------------
449
+ 2026-04-08 12:54:03 - INFO - parrotllm.training - Train: loss=3.7920, ppl=44.34
450
+ 2026-04-08 12:54:03 - INFO - parrotllm.training - Val: loss=3.7612, ppl=43.00
451
+ 2026-04-08 12:54:03 - INFO - parrotllm.training - ** New best validation loss! **
452
+ 2026-04-08 12:54:04 - INFO - parrotllm.training - Saved best checkpoint: /home/user/code/ParrotLLM/runs/big_run/exp_c/run_20260408_124138/checkpoints/best_loss_3p7612_epoch_0000_step_0001700.pt
453
+ 2026-04-08 12:54:05 - INFO - parrotllm.training - ------------------------------------------------------------
454
+ 2026-04-08 12:54:08 - INFO - parrotllm.training - step 1710 | epoch 0 | loss 3.7862 | lr 1.30e-03 | grad 0.2962
455
+ 2026-04-08 12:54:11 - INFO - parrotllm.training - step 1720 | epoch 0 | loss 3.8174 | lr 1.30e-03 | grad 0.2350
456
+ 2026-04-08 12:54:14 - INFO - parrotllm.training - step 1730 | epoch 0 | loss 3.9581 | lr 1.30e-03 | grad 0.2243
457
+ 2026-04-08 12:54:17 - INFO - parrotllm.training - step 1740 | epoch 0 | loss 3.8328 | lr 1.30e-03 | grad 0.2576
458
+ 2026-04-08 12:54:20 - INFO - parrotllm.training - step 1750 | epoch 0 | loss 3.7767 | lr 1.30e-03 | grad 0.2573
459
+ 2026-04-08 12:54:21 - INFO - parrotllm.training - Saved checkpoint: /home/user/code/ParrotLLM/runs/big_run/exp_c/run_20260408_124138/checkpoints/last_epoch_0000_step_0001750.pt
460
+ 2026-04-08 12:54:25 - INFO - parrotllm.training - step 1760 | epoch 0 | loss 3.8260 | lr 1.30e-03 | grad 0.2822
461
+ 2026-04-08 12:54:29 - INFO - parrotllm.training - step 1770 | epoch 0 | loss 3.9070 | lr 1.30e-03 | grad 0.2378
462
+ 2026-04-08 12:54:32 - INFO - parrotllm.training - step 1780 | epoch 0 | loss 3.7750 | lr 1.30e-03 | grad 0.2820
463
+ 2026-04-08 12:54:35 - INFO - parrotllm.training - step 1790 | epoch 0 | loss 3.8340 | lr 1.30e-03 | grad 0.2993
464
+ 2026-04-08 12:54:38 - INFO - parrotllm.training - step 1800 | epoch 0 | loss 3.8334 | lr 1.30e-03 | grad 0.2901
465
+ 2026-04-08 12:54:38 - INFO - parrotllm.training - Starting evaluation...
466
+ 2026-04-08 12:54:38 - INFO - parrotllm.training - ------------------------------------------------------------
467
+ 2026-04-08 12:54:43 - INFO - parrotllm.training - Train: loss=3.8334, ppl=46.22
468
+ 2026-04-08 12:54:43 - INFO - parrotllm.training - Val: loss=3.7424, ppl=42.20
469
+ 2026-04-08 12:54:43 - INFO - parrotllm.training - ** New best validation loss! **
470
+ 2026-04-08 12:54:44 - INFO - parrotllm.training - Saved best checkpoint: /home/user/code/ParrotLLM/runs/big_run/exp_c/run_20260408_124138/checkpoints/best_loss_3p7424_epoch_0000_step_0001800.pt
471
+ 2026-04-08 12:54:45 - INFO - parrotllm.training - ------------------------------------------------------------
472
+ 2026-04-08 12:54:48 - INFO - parrotllm.training - step 1810 | epoch 0 | loss 3.8063 | lr 1.30e-03 | grad 0.2505
473
+ 2026-04-08 12:54:51 - INFO - parrotllm.training - step 1820 | epoch 0 | loss 3.7296 | lr 1.30e-03 | grad 0.3443
474
+ 2026-04-08 12:54:54 - INFO - parrotllm.training - step 1830 | epoch 0 | loss 3.7140 | lr 1.30e-03 | grad 0.2645
475
+ 2026-04-08 12:54:57 - INFO - parrotllm.training - step 1840 | epoch 0 | loss 3.8231 | lr 1.30e-03 | grad 0.2163
476
+ 2026-04-08 12:55:00 - INFO - parrotllm.training - step 1850 | epoch 0 | loss 3.6996 | lr 1.30e-03 | grad 0.3037
477
+ 2026-04-08 12:55:03 - INFO - parrotllm.training - step 1860 | epoch 0 | loss 3.7551 | lr 1.30e-03 | grad 0.2463
478
+ 2026-04-08 12:55:07 - INFO - parrotllm.training - step 1870 | epoch 0 | loss 3.7731 | lr 1.30e-03 | grad 0.2317
479
+ 2026-04-08 12:55:10 - INFO - parrotllm.training - step 1880 | epoch 0 | loss 3.7692 | lr 1.30e-03 | grad 0.2434
480
+ 2026-04-08 12:55:13 - INFO - parrotllm.training - step 1890 | epoch 0 | loss 3.7588 | lr 1.30e-03 | grad 0.2451
481
+ 2026-04-08 12:55:16 - INFO - parrotllm.training - step 1900 | epoch 0 | loss 3.6820 | lr 1.30e-03 | grad 0.2575
482
+ 2026-04-08 12:55:16 - INFO - parrotllm.training - Starting evaluation...
483
+ 2026-04-08 12:55:16 - INFO - parrotllm.training - ------------------------------------------------------------
484
+ 2026-04-08 12:55:21 - INFO - parrotllm.training - Train: loss=3.6820, ppl=39.72
485
+ 2026-04-08 12:55:21 - INFO - parrotllm.training - Val: loss=3.7291, ppl=41.64
486
+ 2026-04-08 12:55:21 - INFO - parrotllm.training - ** New best validation loss! **
487
+ 2026-04-08 12:55:22 - INFO - parrotllm.training - Saved best checkpoint: /home/user/code/ParrotLLM/runs/big_run/exp_c/run_20260408_124138/checkpoints/best_loss_3p7291_epoch_0000_step_0001900.pt
488
+ 2026-04-08 12:55:23 - INFO - parrotllm.training - ------------------------------------------------------------
489
+ 2026-04-08 12:55:26 - INFO - parrotllm.training - step 1910 | epoch 0 | loss 3.7672 | lr 1.30e-03 | grad 0.2429
490
+ 2026-04-08 12:55:29 - INFO - parrotllm.training - step 1920 | epoch 0 | loss 3.6872 | lr 1.30e-03 | grad 0.2719
491
+ 2026-04-08 12:55:32 - INFO - parrotllm.training - step 1930 | epoch 0 | loss 3.6756 | lr 1.30e-03 | grad 0.2532
492
+ 2026-04-08 12:55:35 - INFO - parrotllm.training - Starting evaluation...
493
+ 2026-04-08 12:55:35 - INFO - parrotllm.training - ------------------------------------------------------------
494
+ 2026-04-08 12:55:41 - INFO - parrotllm.training - Epoch 1 complete:
495
+ 2026-04-08 12:55:41 - INFO - parrotllm.training - Train: loss=3.7445, ppl=42.29
496
+ 2026-04-08 12:55:41 - INFO - parrotllm.training - Val: loss=3.7228, ppl=41.38
497
+ 2026-04-08 12:55:41 - INFO - parrotllm.training - ** New best validation loss! **
498
+ 2026-04-08 12:55:42 - INFO - parrotllm.training - Saved best checkpoint: /home/user/code/ParrotLLM/runs/big_run/exp_c/run_20260408_124138/checkpoints/best_loss_3p7228_epoch_0001_step_0001939.pt
499
+ 2026-04-08 12:55:42 - INFO - parrotllm.training - ------------------------------------------------------------
500
+ 2026-04-08 12:55:43 - INFO - parrotllm.training - step 1940 | epoch 1 | loss 3.8277 | lr 1.30e-03 | grad 0.2471
501
+ 2026-04-08 12:55:46 - INFO - parrotllm.training - step 1950 | epoch 1 | loss 3.7919 | lr 1.30e-03 | grad 0.2724
502
+ 2026-04-08 12:55:49 - INFO - parrotllm.training - step 1960 | epoch 1 | loss 3.7537 | lr 1.30e-03 | grad 0.2530
503
+ 2026-04-08 12:55:52 - INFO - parrotllm.training - step 1970 | epoch 1 | loss 3.7633 | lr 1.30e-03 | grad 0.3022
504
+ 2026-04-08 12:55:55 - INFO - parrotllm.training - step 1980 | epoch 1 | loss 3.6716 | lr 1.30e-03 | grad 0.2998
505
+ 2026-04-08 12:55:58 - INFO - parrotllm.training - step 1990 | epoch 1 | loss 3.7742 | lr 1.30e-03 | grad 0.2340
506
+ 2026-04-08 12:56:01 - INFO - parrotllm.training - step 2000 | epoch 1 | loss 3.7044 | lr 1.30e-03 | grad 0.2280
507
+ 2026-04-08 12:56:01 - INFO - parrotllm.training - Starting evaluation...
508
+ 2026-04-08 12:56:01 - INFO - parrotllm.training - ------------------------------------------------------------
509
+ 2026-04-08 12:56:07 - INFO - parrotllm.training - Train: loss=3.7044, ppl=40.63
510
+ 2026-04-08 12:56:07 - INFO - parrotllm.training - Val: loss=3.7132, ppl=40.98
511
+ 2026-04-08 12:56:07 - INFO - parrotllm.training - ** New best validation loss! **
512
+ 2026-04-08 12:56:08 - INFO - parrotllm.training - Saved best checkpoint: /home/user/code/ParrotLLM/runs/big_run/exp_c/run_20260408_124138/checkpoints/best_loss_3p7132_epoch_0001_step_0002000.pt
513
+ 2026-04-08 12:56:08 - INFO - parrotllm.training - ------------------------------------------------------------
514
+ 2026-04-08 12:56:09 - INFO - parrotllm.training - Saved checkpoint: /home/user/code/ParrotLLM/runs/big_run/exp_c/run_20260408_124138/checkpoints/last_epoch_0001_step_0002000.pt
515
+ 2026-04-08 12:56:13 - INFO - parrotllm.training - step 2010 | epoch 1 | loss 3.7154 | lr 1.30e-03 | grad 0.2576
516
+ 2026-04-08 12:56:17 - INFO - parrotllm.training - step 2020 | epoch 1 | loss 3.8704 | lr 1.30e-03 | grad 0.2502
517
+ 2026-04-08 12:56:20 - INFO - parrotllm.training - step 2030 | epoch 1 | loss 3.6939 | lr 1.30e-03 | grad 0.3161
518
+ 2026-04-08 12:56:23 - INFO - parrotllm.training - step 2040 | epoch 1 | loss 3.7445 | lr 1.30e-03 | grad 0.2667
519
+ 2026-04-08 12:56:26 - INFO - parrotllm.training - step 2050 | epoch 1 | loss 3.7939 | lr 1.30e-03 | grad 0.2362
520
+ 2026-04-08 12:56:29 - INFO - parrotllm.training - step 2060 | epoch 1 | loss 3.7188 | lr 1.30e-03 | grad 0.2296
521
+ 2026-04-08 12:56:32 - INFO - parrotllm.training - step 2070 | epoch 1 | loss 3.7637 | lr 1.30e-03 | grad 0.2863
522
+ 2026-04-08 12:56:35 - INFO - parrotllm.training - step 2080 | epoch 1 | loss 3.7682 | lr 1.30e-03 | grad 0.2928
523
+ 2026-04-08 12:56:38 - INFO - parrotllm.training - step 2090 | epoch 1 | loss 3.7665 | lr 1.30e-03 | grad 0.2731
524
+ 2026-04-08 12:56:41 - INFO - parrotllm.training - step 2100 | epoch 1 | loss 3.7406 | lr 1.30e-03 | grad 0.2323
525
+ 2026-04-08 12:56:41 - INFO - parrotllm.training - Starting evaluation...
526
+ 2026-04-08 12:56:41 - INFO - parrotllm.training - ------------------------------------------------------------
527
+ 2026-04-08 12:56:47 - INFO - parrotllm.training - Train: loss=3.7406, ppl=42.12
528
+ 2026-04-08 12:56:47 - INFO - parrotllm.training - Val: loss=3.7036, ppl=40.59
529
+ 2026-04-08 12:56:47 - INFO - parrotllm.training - ** New best validation loss! **
530
+ 2026-04-08 12:56:48 - INFO - parrotllm.training - Saved best checkpoint: /home/user/code/ParrotLLM/runs/big_run/exp_c/run_20260408_124138/checkpoints/best_loss_3p7036_epoch_0001_step_0002100.pt
531
+ 2026-04-08 12:56:49 - INFO - parrotllm.training - ------------------------------------------------------------
532
+ 2026-04-08 12:56:52 - INFO - parrotllm.training - step 2110 | epoch 1 | loss 3.7433 | lr 1.30e-03 | grad 0.2996
533
+ 2026-04-08 12:56:55 - INFO - parrotllm.training - step 2120 | epoch 1 | loss 3.6569 | lr 1.30e-03 | grad 0.2546
534
+ 2026-04-08 12:56:58 - INFO - parrotllm.training - step 2130 | epoch 1 | loss 3.7969 | lr 1.30e-03 | grad 0.2329
535
+ 2026-04-08 12:57:01 - INFO - parrotllm.training - step 2140 | epoch 1 | loss 3.6794 | lr 1.30e-03 | grad 0.2936
536
+ 2026-04-08 12:57:04 - INFO - parrotllm.training - step 2150 | epoch 1 | loss 3.6947 | lr 1.30e-03 | grad 0.2479
537
+ 2026-04-08 12:57:07 - INFO - parrotllm.training - step 2160 | epoch 1 | loss 3.7705 | lr 1.30e-03 | grad 0.2700
538
+ 2026-04-08 12:57:10 - INFO - parrotllm.training - step 2170 | epoch 1 | loss 3.7438 | lr 1.30e-03 | grad 0.2319
539
+ 2026-04-08 12:57:13 - INFO - parrotllm.training - step 2180 | epoch 1 | loss 3.7602 | lr 1.30e-03 | grad 0.2488
540
+ 2026-04-08 12:57:16 - INFO - parrotllm.training - step 2190 | epoch 1 | loss 3.7940 | lr 1.30e-03 | grad 0.2288
541
+ 2026-04-08 12:57:19 - INFO - parrotllm.training - step 2200 | epoch 1 | loss 3.6871 | lr 1.30e-03 | grad 0.2307
542
+ 2026-04-08 12:57:19 - INFO - parrotllm.training - Starting evaluation...
543
+ 2026-04-08 12:57:19 - INFO - parrotllm.training - ------------------------------------------------------------
544
+ 2026-04-08 12:57:25 - INFO - parrotllm.training - Train: loss=3.6871, ppl=39.93
545
+ 2026-04-08 12:57:25 - INFO - parrotllm.training - Val: loss=3.6926, ppl=40.15
546
+ 2026-04-08 12:57:25 - INFO - parrotllm.training - ** New best validation loss! **
547
+ 2026-04-08 12:57:26 - INFO - parrotllm.training - Saved best checkpoint: /home/user/code/ParrotLLM/runs/big_run/exp_c/run_20260408_124138/checkpoints/best_loss_3p6926_epoch_0001_step_0002200.pt
548
+ 2026-04-08 12:57:27 - INFO - parrotllm.training - ------------------------------------------------------------
549
+ 2026-04-08 12:57:30 - INFO - parrotllm.training - step 2210 | epoch 1 | loss 3.6831 | lr 1.30e-03 | grad 0.2583
550
+ 2026-04-08 12:57:33 - INFO - parrotllm.training - step 2220 | epoch 1 | loss 3.6792 | lr 1.30e-03 | grad 0.3328
551
+ 2026-04-08 12:57:36 - INFO - parrotllm.training - step 2230 | epoch 1 | loss 3.6882 | lr 1.30e-03 | grad 0.2808
552
+ 2026-04-08 12:57:39 - INFO - parrotllm.training - step 2240 | epoch 1 | loss 3.6036 | lr 1.30e-03 | grad 0.2715
553
+ 2026-04-08 12:57:42 - INFO - parrotllm.training - step 2250 | epoch 1 | loss 3.8189 | lr 1.30e-03 | grad 0.2541
554
+ 2026-04-08 12:57:43 - INFO - parrotllm.training - Saved checkpoint: /home/user/code/ParrotLLM/runs/big_run/exp_c/run_20260408_124138/checkpoints/last_epoch_0001_step_0002250.pt
555
+ 2026-04-08 12:57:47 - INFO - parrotllm.training - step 2260 | epoch 1 | loss 3.7713 | lr 1.30e-03 | grad 0.2286
556
+ 2026-04-08 12:57:50 - INFO - parrotllm.training - step 2270 | epoch 1 | loss 3.7094 | lr 1.30e-03 | grad 0.2440
557
+ 2026-04-08 12:57:53 - INFO - parrotllm.training - step 2280 | epoch 1 | loss 3.8278 | lr 1.30e-03 | grad 0.3111
558
+ 2026-04-08 12:57:56 - INFO - parrotllm.training - step 2290 | epoch 1 | loss 3.7622 | lr 1.30e-03 | grad 0.2418
559
+ 2026-04-08 12:57:59 - INFO - parrotllm.training - step 2300 | epoch 1 | loss 3.7777 | lr 1.30e-03 | grad 0.2645
560
+ 2026-04-08 12:57:59 - INFO - parrotllm.training - Starting evaluation...
561
+ 2026-04-08 12:57:59 - INFO - parrotllm.training - ------------------------------------------------------------
562
+ 2026-04-08 12:58:05 - INFO - parrotllm.training - Train: loss=3.7777, ppl=43.72
563
+ 2026-04-08 12:58:05 - INFO - parrotllm.training - Val: loss=3.6838, ppl=39.80
564
+ 2026-04-08 12:58:05 - INFO - parrotllm.training - ** New best validation loss! **
565
+ 2026-04-08 12:58:06 - INFO - parrotllm.training - Saved best checkpoint: /home/user/code/ParrotLLM/runs/big_run/exp_c/run_20260408_124138/checkpoints/best_loss_3p6838_epoch_0001_step_0002300.pt
566
+ 2026-04-08 12:58:07 - INFO - parrotllm.training - ------------------------------------------------------------
567
+ 2026-04-08 12:58:10 - INFO - parrotllm.training - step 2310 | epoch 1 | loss 3.6129 | lr 1.30e-03 | grad 0.2422
568
+ 2026-04-08 12:58:14 - INFO - parrotllm.training - step 2320 | epoch 1 | loss 3.6828 | lr 1.30e-03 | grad 0.2511
569
+ 2026-04-08 12:58:17 - INFO - parrotllm.training - step 2330 | epoch 1 | loss 3.7867 | lr 1.30e-03 | grad 0.2292
570
+ 2026-04-08 12:58:20 - INFO - parrotllm.training - step 2340 | epoch 1 | loss 3.6965 | lr 1.30e-03 | grad 0.2199
571
+ 2026-04-08 12:58:23 - INFO - parrotllm.training - step 2350 | epoch 1 | loss 3.7047 | lr 1.30e-03 | grad 0.2248
572
+ 2026-04-08 12:58:26 - INFO - parrotllm.training - step 2360 | epoch 1 | loss 3.6923 | lr 1.30e-03 | grad 0.2721
573
+ 2026-04-08 12:58:29 - INFO - parrotllm.training - step 2370 | epoch 1 | loss 3.7326 | lr 1.30e-03 | grad 0.2353
574
+ 2026-04-08 12:58:32 - INFO - parrotllm.training - step 2380 | epoch 1 | loss 3.7287 | lr 1.30e-03 | grad 0.2449
575
+ 2026-04-08 12:58:35 - INFO - parrotllm.training - step 2390 | epoch 1 | loss 3.6754 | lr 1.30e-03 | grad 0.2936
576
+ 2026-04-08 12:58:38 - INFO - parrotllm.training - step 2400 | epoch 1 | loss 3.6743 | lr 1.30e-03 | grad 0.2145
577
+ 2026-04-08 12:58:38 - INFO - parrotllm.training - Starting evaluation...
578
+ 2026-04-08 12:58:38 - INFO - parrotllm.training - ------------------------------------------------------------
579
+ 2026-04-08 12:58:44 - INFO - parrotllm.training - Train: loss=3.6743, ppl=39.42
580
+ 2026-04-08 12:58:44 - INFO - parrotllm.training - Val: loss=3.6698, ppl=39.25
581
+ 2026-04-08 12:58:44 - INFO - parrotllm.training - ** New best validation loss! **
582
+ 2026-04-08 12:58:45 - INFO - parrotllm.training - Saved best checkpoint: /home/user/code/ParrotLLM/runs/big_run/exp_c/run_20260408_124138/checkpoints/best_loss_3p6698_epoch_0001_step_0002400.pt
583
+ 2026-04-08 12:58:45 - INFO - parrotllm.training - ------------------------------------------------------------
584
+ 2026-04-08 12:58:49 - INFO - parrotllm.training - step 2410 | epoch 1 | loss 3.6792 | lr 1.30e-03 | grad 0.2501
585
+ 2026-04-08 12:58:52 - INFO - parrotllm.training - step 2420 | epoch 1 | loss 3.6492 | lr 1.30e-03 | grad 0.2364
586
+ 2026-04-08 12:58:55 - INFO - parrotllm.training - step 2430 | epoch 1 | loss 3.6545 | lr 1.30e-03 | grad 0.2830
587
+ 2026-04-08 12:58:58 - INFO - parrotllm.training - step 2440 | epoch 1 | loss 3.6382 | lr 1.30e-03 | grad 0.2191
588
+ 2026-04-08 12:59:01 - INFO - parrotllm.training - step 2450 | epoch 1 | loss 3.7305 | lr 1.30e-03 | grad 0.2267
589
+ 2026-04-08 12:59:04 - INFO - parrotllm.training - step 2460 | epoch 1 | loss 3.6532 | lr 1.30e-03 | grad 0.2310
590
+ 2026-04-08 12:59:07 - INFO - parrotllm.training - step 2470 | epoch 1 | loss 3.7181 | lr 1.30e-03 | grad 0.2645
591
+ 2026-04-08 12:59:10 - INFO - parrotllm.training - step 2480 | epoch 1 | loss 3.7002 | lr 1.30e-03 | grad 0.2352
592
+ 2026-04-08 12:59:13 - INFO - parrotllm.training - step 2490 | epoch 1 | loss 3.7501 | lr 1.30e-03 | grad 0.2171
593
+ 2026-04-08 12:59:16 - INFO - parrotllm.training - step 2500 | epoch 1 | loss 3.7505 | lr 1.30e-03 | grad 0.2759
594
+ 2026-04-08 12:59:16 - INFO - parrotllm.training - Starting evaluation...
595
+ 2026-04-08 12:59:16 - INFO - parrotllm.training - ------------------------------------------------------------
596
+ 2026-04-08 12:59:22 - INFO - parrotllm.training - Train: loss=3.7505, ppl=42.54
597
+ 2026-04-08 12:59:22 - INFO - parrotllm.training - Val: loss=3.6614, ppl=38.92
598
+ 2026-04-08 12:59:22 - INFO - parrotllm.training - ** New best validation loss! **
599
+ 2026-04-08 12:59:23 - INFO - parrotllm.training - Saved best checkpoint: /home/user/code/ParrotLLM/runs/big_run/exp_c/run_20260408_124138/checkpoints/best_loss_3p6614_epoch_0001_step_0002500.pt
600
+ 2026-04-08 12:59:24 - INFO - parrotllm.training - ------------------------------------------------------------
601
+ 2026-04-08 12:59:25 - INFO - parrotllm.training - Saved checkpoint: /home/user/code/ParrotLLM/runs/big_run/exp_c/run_20260408_124138/checkpoints/last_epoch_0001_step_0002500.pt
602
+ 2026-04-08 12:59:28 - INFO - parrotllm.training - step 2510 | epoch 1 | loss 3.6079 | lr 1.30e-03 | grad 0.2531
603
+ 2026-04-08 12:59:32 - INFO - parrotllm.training - step 2520 | epoch 1 | loss 3.7122 | lr 1.27e-03 | grad 0.2423
604
+ 2026-04-08 12:59:35 - INFO - parrotllm.training - step 2530 | epoch 1 | loss 3.6970 | lr 1.25e-03 | grad 0.2570
605
+ 2026-04-08 12:59:38 - INFO - parrotllm.training - step 2540 | epoch 1 | loss 3.7211 | lr 1.23e-03 | grad 0.2575
606
+ 2026-04-08 12:59:41 - INFO - parrotllm.training - step 2550 | epoch 1 | loss 3.7453 | lr 1.21e-03 | grad 0.2406
607
+ 2026-04-08 12:59:44 - INFO - parrotllm.training - step 2560 | epoch 1 | loss 3.6392 | lr 1.19e-03 | grad 0.2512
608
+ 2026-04-08 12:59:47 - INFO - parrotllm.training - step 2570 | epoch 1 | loss 3.5481 | lr 1.17e-03 | grad 0.2399
609
+ 2026-04-08 12:59:50 - INFO - parrotllm.training - step 2580 | epoch 1 | loss 3.7491 | lr 1.15e-03 | grad 0.2400
610
+ 2026-04-08 12:59:53 - INFO - parrotllm.training - step 2590 | epoch 1 | loss 3.6592 | lr 1.12e-03 | grad 0.1877
611
+ 2026-04-08 12:59:56 - INFO - parrotllm.training - step 2600 | epoch 1 | loss 3.6217 | lr 1.10e-03 | grad 0.2066
612
+ 2026-04-08 12:59:56 - INFO - parrotllm.training - Starting evaluation...
613
+ 2026-04-08 12:59:56 - INFO - parrotllm.training - ------------------------------------------------------------
614
+ 2026-04-08 13:00:02 - INFO - parrotllm.training - Train: loss=3.6217, ppl=37.40
615
+ 2026-04-08 13:00:02 - INFO - parrotllm.training - Val: loss=3.6383, ppl=38.03
616
+ 2026-04-08 13:00:02 - INFO - parrotllm.training - ** New best validation loss! **
617
+ 2026-04-08 13:00:03 - INFO - parrotllm.training - Saved best checkpoint: /home/user/code/ParrotLLM/runs/big_run/exp_c/run_20260408_124138/checkpoints/best_loss_3p6383_epoch_0001_step_0002600.pt
618
+ 2026-04-08 13:00:04 - INFO - parrotllm.training - ------------------------------------------------------------
619
+ 2026-04-08 13:00:07 - INFO - parrotllm.training - step 2610 | epoch 1 | loss 3.5967 | lr 1.08e-03 | grad 0.2223
620
+ 2026-04-08 13:00:10 - INFO - parrotllm.training - step 2620 | epoch 1 | loss 3.6822 | lr 1.06e-03 | grad 0.2510
621
+ 2026-04-08 13:00:13 - INFO - parrotllm.training - step 2630 | epoch 1 | loss 3.7362 | lr 1.04e-03 | grad 0.2201
622
+ 2026-04-08 13:00:16 - INFO - parrotllm.training - step 2640 | epoch 1 | loss 3.6199 | lr 1.02e-03 | grad 0.2164
623
+ 2026-04-08 13:00:19 - INFO - parrotllm.training - step 2650 | epoch 1 | loss 3.5362 | lr 9.95e-04 | grad 0.1980
624
+ 2026-04-08 13:00:22 - INFO - parrotllm.training - step 2660 | epoch 1 | loss 3.7150 | lr 9.74e-04 | grad 0.2174
625
+ 2026-04-08 13:00:25 - INFO - parrotllm.training - step 2670 | epoch 1 | loss 3.6617 | lr 9.52e-04 | grad 0.2100
626
+ 2026-04-08 13:00:29 - INFO - parrotllm.training - step 2680 | epoch 1 | loss 3.6115 | lr 9.31e-04 | grad 0.2233
627
+ 2026-04-08 13:00:32 - INFO - parrotllm.training - step 2690 | epoch 1 | loss 3.6299 | lr 9.09e-04 | grad 0.2142
628
+ 2026-04-08 13:00:35 - INFO - parrotllm.training - step 2700 | epoch 1 | loss 3.6080 | lr 8.88e-04 | grad 0.2194
629
+ 2026-04-08 13:00:35 - INFO - parrotllm.training - Starting evaluation...
630
+ 2026-04-08 13:00:35 - INFO - parrotllm.training - ------------------------------------------------------------
631
+ 2026-04-08 13:00:40 - INFO - parrotllm.training - Train: loss=3.6080, ppl=36.89
632
+ 2026-04-08 13:00:40 - INFO - parrotllm.training - Val: loss=3.6128, ppl=37.07
633
+ 2026-04-08 13:00:40 - INFO - parrotllm.training - ** New best validation loss! **
634
+ 2026-04-08 13:00:41 - INFO - parrotllm.training - Saved best checkpoint: /home/user/code/ParrotLLM/runs/big_run/exp_c/run_20260408_124138/checkpoints/best_loss_3p6128_epoch_0001_step_0002700.pt
635
+ 2026-04-08 13:00:42 - INFO - parrotllm.training - ------------------------------------------------------------
636
+ 2026-04-08 13:00:45 - INFO - parrotllm.training - step 2710 | epoch 1 | loss 3.6463 | lr 8.66e-04 | grad 0.1986
637
+ 2026-04-08 13:00:48 - INFO - parrotllm.training - step 2720 | epoch 1 | loss 3.5877 | lr 8.45e-04 | grad 0.1982
638
+ 2026-04-08 13:00:51 - INFO - parrotllm.training - step 2730 | epoch 1 | loss 3.6518 | lr 8.23e-04 | grad 0.2078
639
+ 2026-04-08 13:00:54 - INFO - parrotllm.training - step 2740 | epoch 1 | loss 3.5750 | lr 8.02e-04 | grad 0.1863
640
+ 2026-04-08 13:00:57 - INFO - parrotllm.training - step 2750 | epoch 1 | loss 3.6092 | lr 7.80e-04 | grad 0.1831
641
+ 2026-04-08 13:00:58 - INFO - parrotllm.training - Saved checkpoint: /home/user/code/ParrotLLM/runs/big_run/exp_c/run_20260408_124138/checkpoints/last_epoch_0001_step_0002750.pt
642
+ 2026-04-08 13:01:02 - INFO - parrotllm.training - step 2760 | epoch 1 | loss 3.6252 | lr 7.59e-04 | grad 0.2036
643
+ 2026-04-08 13:01:05 - INFO - parrotllm.training - step 2770 | epoch 1 | loss 3.6525 | lr 7.37e-04 | grad 0.1935
644
+ 2026-04-08 13:01:08 - INFO - parrotllm.training - step 2780 | epoch 1 | loss 3.7304 | lr 7.16e-04 | grad 0.1867
645
+ 2026-04-08 13:01:11 - INFO - parrotllm.training - step 2790 | epoch 1 | loss 3.6103 | lr 6.94e-04 | grad 0.1835
646
+ 2026-04-08 13:01:14 - INFO - parrotllm.training - step 2800 | epoch 1 | loss 3.5484 | lr 6.73e-04 | grad 0.1843
647
+ 2026-04-08 13:01:14 - INFO - parrotllm.training - Starting evaluation...
648
+ 2026-04-08 13:01:14 - INFO - parrotllm.training - ------------------------------------------------------------
649
+ 2026-04-08 13:01:20 - INFO - parrotllm.training - Train: loss=3.5484, ppl=34.76
650
+ 2026-04-08 13:01:20 - INFO - parrotllm.training - Val: loss=3.5885, ppl=36.18
651
+ 2026-04-08 13:01:20 - INFO - parrotllm.training - ** New best validation loss! **
652
+ 2026-04-08 13:01:21 - INFO - parrotllm.training - Saved best checkpoint: /home/user/code/ParrotLLM/runs/big_run/exp_c/run_20260408_124138/checkpoints/best_loss_3p5885_epoch_0001_step_0002800.pt
653
+ 2026-04-08 13:01:22 - INFO - parrotllm.training - ------------------------------------------------------------
654
+ 2026-04-08 13:01:25 - INFO - parrotllm.training - step 2810 | epoch 1 | loss 3.6053 | lr 6.51e-04 | grad 0.1741
655
+ 2026-04-08 13:01:28 - INFO - parrotllm.training - step 2820 | epoch 1 | loss 3.5340 | lr 6.30e-04 | grad 0.1929
656
+ 2026-04-08 13:01:31 - INFO - parrotllm.training - step 2830 | epoch 1 | loss 3.5621 | lr 6.08e-04 | grad 0.1737
657
+ 2026-04-08 13:01:34 - INFO - parrotllm.training - step 2840 | epoch 1 | loss 3.4720 | lr 5.87e-04 | grad 0.1846
658
+ 2026-04-08 13:01:37 - INFO - parrotllm.training - step 2850 | epoch 1 | loss 3.5570 | lr 5.65e-04 | grad 0.1913
659
+ 2026-04-08 13:01:40 - INFO - parrotllm.training - step 2860 | epoch 1 | loss 3.5383 | lr 5.44e-04 | grad 0.1892
660
+ 2026-04-08 13:01:43 - INFO - parrotllm.training - step 2870 | epoch 1 | loss 3.4660 | lr 5.22e-04 | grad 0.1735
661
+ 2026-04-08 13:01:46 - INFO - parrotllm.training - step 2880 | epoch 1 | loss 3.6119 | lr 5.00e-04 | grad 0.1769
662
+ 2026-04-08 13:01:49 - INFO - parrotllm.training - step 2890 | epoch 1 | loss 3.5662 | lr 4.79e-04 | grad 0.1743
663
+ 2026-04-08 13:01:53 - INFO - parrotllm.training - step 2900 | epoch 1 | loss 3.6884 | lr 4.57e-04 | grad 0.1812
664
+ 2026-04-08 13:01:53 - INFO - parrotllm.training - Starting evaluation...
665
+ 2026-04-08 13:01:53 - INFO - parrotllm.training - ------------------------------------------------------------
666
+ 2026-04-08 13:01:58 - INFO - parrotllm.training - Train: loss=3.6884, ppl=39.98
667
+ 2026-04-08 13:01:58 - INFO - parrotllm.training - Val: loss=3.5634, ppl=35.28
668
+ 2026-04-08 13:01:58 - INFO - parrotllm.training - ** New best validation loss! **
669
+ 2026-04-08 13:01:59 - INFO - parrotllm.training - Saved best checkpoint: /home/user/code/ParrotLLM/runs/big_run/exp_c/run_20260408_124138/checkpoints/best_loss_3p5634_epoch_0001_step_0002900.pt
670
+ 2026-04-08 13:02:00 - INFO - parrotllm.training - ------------------------------------------------------------
671
+ 2026-04-08 13:02:03 - INFO - parrotllm.training - step 2910 | epoch 1 | loss 3.5630 | lr 4.36e-04 | grad 0.1769
672
+ 2026-04-08 13:02:06 - INFO - parrotllm.training - step 2920 | epoch 1 | loss 3.6464 | lr 4.14e-04 | grad 0.1837
673
+ 2026-04-08 13:02:09 - INFO - parrotllm.training - step 2930 | epoch 1 | loss 3.6239 | lr 3.93e-04 | grad 0.1670
674
+ 2026-04-08 13:02:12 - INFO - parrotllm.training - step 2940 | epoch 1 | loss 3.5106 | lr 3.71e-04 | grad 0.1634
675
+ 2026-04-08 13:02:15 - INFO - parrotllm.training - step 2950 | epoch 1 | loss 3.5441 | lr 3.50e-04 | grad 0.1734
676
+ 2026-04-08 13:02:18 - INFO - parrotllm.training - step 2960 | epoch 1 | loss 3.5644 | lr 3.28e-04 | grad 0.1775
677
+ 2026-04-08 13:02:21 - INFO - parrotllm.training - step 2970 | epoch 1 | loss 3.5259 | lr 3.07e-04 | grad 0.1772
678
+ 2026-04-08 13:02:25 - INFO - parrotllm.training - step 2980 | epoch 1 | loss 3.6529 | lr 2.85e-04 | grad 0.1660
679
+ 2026-04-08 13:02:28 - INFO - parrotllm.training - step 2990 | epoch 1 | loss 3.6271 | lr 2.64e-04 | grad 0.1548
680
+ 2026-04-08 13:02:31 - INFO - parrotllm.training - step 3000 | epoch 1 | loss 3.6338 | lr 2.42e-04 | grad 0.1579
681
+ 2026-04-08 13:02:31 - INFO - parrotllm.training - Starting evaluation...
682
+ 2026-04-08 13:02:31 - INFO - parrotllm.training - ------------------------------------------------------------
683
+ 2026-04-08 13:02:36 - INFO - parrotllm.training - Train: loss=3.6338, ppl=37.86
684
+ 2026-04-08 13:02:36 - INFO - parrotllm.training - Val: loss=3.5437, ppl=34.60
685
+ 2026-04-08 13:02:36 - INFO - parrotllm.training - ** New best validation loss! **
686
+ 2026-04-08 13:02:37 - INFO - parrotllm.training - Saved best checkpoint: /home/user/code/ParrotLLM/runs/big_run/exp_c/run_20260408_124138/checkpoints/best_loss_3p5437_epoch_0001_step_0003000.pt
687
+ 2026-04-08 13:02:38 - INFO - parrotllm.training - ------------------------------------------------------------
688
+ 2026-04-08 13:02:39 - INFO - parrotllm.training - Saved checkpoint: /home/user/code/ParrotLLM/runs/big_run/exp_c/run_20260408_124138/checkpoints/last_epoch_0001_step_0003000.pt
689
+ 2026-04-08 13:02:43 - INFO - parrotllm.training - step 3010 | epoch 1 | loss 3.6042 | lr 2.21e-04 | grad 0.1526
690
+ 2026-04-08 13:02:46 - INFO - parrotllm.training - step 3020 | epoch 1 | loss 3.5415 | lr 1.99e-04 | grad 0.1572
691
+ 2026-04-08 13:02:49 - INFO - parrotllm.training - step 3030 | epoch 1 | loss 3.6448 | lr 1.78e-04 | grad 0.1602
692
+ 2026-04-08 13:02:52 - INFO - parrotllm.training - step 3040 | epoch 1 | loss 3.6179 | lr 1.56e-04 | grad 0.1556
693
+ 2026-04-08 13:02:55 - INFO - parrotllm.training - step 3050 | epoch 1 | loss 3.5033 | lr 1.35e-04 | grad 0.1501
694
+ 2026-04-08 13:02:57 - INFO - parrotllm.training - Saved final checkpoint: /home/user/code/ParrotLLM/runs/big_run/exp_c/run_20260408_124138/checkpoints/last_epoch_0001_step_0003052.pt
695
+ 2026-04-08 13:02:57 - INFO - parrotllm.training -
696
+ Loss Curve:
697
+ Loss
698
+ ^
699
+ 10.93 |*
700
+ |*
701
+ |*
702
+ |*
703
+ |**
704
+ | ****
705
+ | ********
706
+ 3.40 | ******************************************
707
+ +--------------------------------------------------> step
708
+ 1 3052
709
+ 2026-04-08 13:02:57 - INFO - parrotllm.training -
710
+ ============================================================
711
+ TRAINING COMPLETE
712
+ ============================================================
713
+ Epochs: 2
714
+ Total steps: 3052
715
+ Total time: 0.35 hours
716
+ Best validation loss: 3.5437
717
+ Run directory: runs/big_run/exp_c/run_20260408_124138
718
+ ============================================================
719
+ 2026-04-08 13:02:57 - INFO - parrotllm.training - Uploading completed run directory to Hugging Face: repo=ParrotLabs/Preprocessed
runs/big_run/exp_c/run_20260408_124138/training_plots.pdf ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:72c7bf527b8395dcd6fa666936af6e83756eece3d29bcaf9817ddfb08383be60
3
+ size 113825