Hugging Face's logo

c14kevincardenas
/
limbxy_seq_t2_heads2_layers1

Transformers
TensorBoard
Safetensors
custom_model
image-sequence-classification
vision
Generated from Trainer
Model card Files Metrics Community
c14kevincardenas commited on
Commit
7f9c2bd
·
verified ·
1 Parent(s): 29a637f

Training in progress, epoch 1

Browse files
Files changed (28) hide show
  1. README.md +82 -0
  2. all_results.json +13 -0
  3. config.json +92 -0
  4. eval_results.json +8 -0
  5. model.safetensors +3 -0
  6. preprocessor_config.json +28 -0
  7. runs/Feb11_18-11-05_galactica.ad.cirange.net/events.out.tfevents.1739297534.galactica.ad.cirange.net.2167895.0 +3 -0
  8. runs/Feb11_18-15-34_galactica.ad.cirange.net/events.out.tfevents.1739297739.galactica.ad.cirange.net.2168831.0 +3 -0
  9. runs/Feb11_18-20-27_galactica.ad.cirange.net/events.out.tfevents.1739298032.galactica.ad.cirange.net.2169870.0 +3 -0
  10. runs/Feb11_18-22-10_galactica.ad.cirange.net/events.out.tfevents.1739298135.galactica.ad.cirange.net.2170704.0 +3 -0
  11. runs/Feb11_18-22-10_galactica.ad.cirange.net/events.out.tfevents.1739305048.galactica.ad.cirange.net.2170704.1 +3 -0
  12. runs/Feb11_21-00-55_galactica.ad.cirange.net/events.out.tfevents.1739307661.galactica.ad.cirange.net.2201888.0 +3 -0
  13. runs/Feb11_21-02-58_galactica.ad.cirange.net/events.out.tfevents.1739307784.galactica.ad.cirange.net.2202655.0 +3 -0
  14. runs/Feb11_21-05-23_galactica.ad.cirange.net/events.out.tfevents.1739307928.galactica.ad.cirange.net.2203425.0 +3 -0
  15. runs/Feb11_21-05-23_galactica.ad.cirange.net/events.out.tfevents.1739314802.galactica.ad.cirange.net.2203425.1 +3 -0
  16. runs/Feb11_23-09-01_galactica.ad.cirange.net/events.out.tfevents.1739315347.galactica.ad.cirange.net.2234313.0 +3 -0
  17. runs/Feb12_16-23-39_galactica.ad.cirange.net/events.out.tfevents.1739377425.galactica.ad.cirange.net.2273555.0 +3 -0
  18. runs/Feb12_16-23-39_galactica.ad.cirange.net/events.out.tfevents.1739384332.galactica.ad.cirange.net.2273555.1 +3 -0
  19. runs/Feb12_18-37-50_galactica.ad.cirange.net/events.out.tfevents.1739385477.galactica.ad.cirange.net.2304333.0 +3 -0
  20. runs/Feb12_18-37-50_galactica.ad.cirange.net/events.out.tfevents.1739392366.galactica.ad.cirange.net.2304333.1 +3 -0
  21. runs/Feb19_23-30-29_galactica.ad.cirange.net/events.out.tfevents.1740007834.galactica.ad.cirange.net.3085606.0 +3 -0
  22. runs/Feb19_23-34-16_galactica.ad.cirange.net/events.out.tfevents.1740008060.galactica.ad.cirange.net.3086972.0 +3 -0
  23. runs/Feb19_23-37-01_galactica.ad.cirange.net/events.out.tfevents.1740008226.galactica.ad.cirange.net.3088755.0 +3 -0
  24. runs/Feb19_23-42-22_galactica.ad.cirange.net/events.out.tfevents.1740008546.galactica.ad.cirange.net.3090432.0 +3 -0
  25. runs/Feb19_23-43-10_galactica.ad.cirange.net/events.out.tfevents.1740008595.galactica.ad.cirange.net.3091725.0 +3 -0
  26. train_results.json +8 -0
  27. trainer_state.json +1062 -0
  28. training_args.bin +3 -0
README.md ADDED
@@ -0,0 +1,82 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ library_name: transformers
3
+ license: apache-2.0
4
+ base_model: c14kevincardenas/beit-large-patch16-384-limb
5
+ tags:
6
+ - image-sequence-classification
7
+ - vision
8
+ - generated_from_trainer
9
+ model-index:
10
+ - name: limbxy_seq_t2
11
+ results: []
12
+ ---
13
+
14
+ <!-- This model card has been generated automatically according to the information the Trainer had access to. You
15
+ should probably proofread and complete it, then remove this comment. -->
16
+
17
+ # limbxy_seq_t2
18
+
19
+ This model is a fine-tuned version of [c14kevincardenas/beit-large-patch16-384-limb](https://huggingface.co/c14kevincardenas/beit-large-patch16-384-limb) on the c14kevincardenas/beta_caller_284_limbxy_seq_2 dataset.
20
+ It achieves the following results on the evaluation set:
21
+ - Loss: 0.0055
22
+ - Rmse: 0.0740
23
+
24
+ ## Model description
25
+
26
+ More information needed
27
+
28
+ ## Intended uses & limitations
29
+
30
+ More information needed
31
+
32
+ ## Training and evaluation data
33
+
34
+ More information needed
35
+
36
+ ## Training procedure
37
+
38
+ ### Training hyperparameters
39
+
40
+ The following hyperparameters were used during training:
41
+ - learning_rate: 5e-05
42
+ - train_batch_size: 32
43
+ - eval_batch_size: 32
44
+ - seed: 2014
45
+ - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
46
+ - lr_scheduler_type: linear
47
+ - lr_scheduler_warmup_steps: 250
48
+ - num_epochs: 20.0
49
+ - mixed_precision_training: Native AMP
50
+
51
+ ### Training results
52
+
53
+ | Training Loss | Epoch | Step | Validation Loss | Rmse |
54
+ |:-------------:|:-----:|:----:|:---------------:|:------:|
55
+ | 0.0137 | 1.0 | 150 | 0.0119 | 0.1090 |
56
+ | 0.0112 | 2.0 | 300 | 0.0118 | 0.1088 |
57
+ | 0.0074 | 3.0 | 450 | 0.0101 | 0.1003 |
58
+ | 0.0064 | 4.0 | 600 | 0.0070 | 0.0835 |
59
+ | 0.0039 | 5.0 | 750 | 0.0069 | 0.0832 |
60
+ | 0.0033 | 6.0 | 900 | 0.0074 | 0.0859 |
61
+ | 0.0026 | 7.0 | 1050 | 0.0061 | 0.0784 |
62
+ | 0.002 | 8.0 | 1200 | 0.0069 | 0.0831 |
63
+ | 0.0018 | 9.0 | 1350 | 0.0057 | 0.0757 |
64
+ | 0.002 | 10.0 | 1500 | 0.0059 | 0.0769 |
65
+ | 0.0012 | 11.0 | 1650 | 0.0057 | 0.0756 |
66
+ | 0.0011 | 12.0 | 1800 | 0.0059 | 0.0770 |
67
+ | 0.0009 | 13.0 | 1950 | 0.0058 | 0.0761 |
68
+ | 0.0008 | 14.0 | 2100 | 0.0056 | 0.0751 |
69
+ | 0.0008 | 15.0 | 2250 | 0.0057 | 0.0756 |
70
+ | 0.0006 | 16.0 | 2400 | 0.0056 | 0.0749 |
71
+ | 0.0006 | 17.0 | 2550 | 0.0055 | 0.0740 |
72
+ | 0.0004 | 18.0 | 2700 | 0.0057 | 0.0755 |
73
+ | 0.0004 | 19.0 | 2850 | 0.0055 | 0.0743 |
74
+ | 0.0002 | 20.0 | 3000 | 0.0055 | 0.0740 |
75
+
76
+
77
+ ### Framework versions
78
+
79
+ - Transformers 4.45.2
80
+ - Pytorch 2.5.0+cu124
81
+ - Datasets 3.0.1
82
+ - Tokenizers 0.20.1
all_results.json ADDED
@@ -0,0 +1,13 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 20.0,
3
+ "eval_loss": 0.005479985848069191,
4
+ "eval_rmse": 0.0740269273519516,
5
+ "eval_runtime": 24.539,
6
+ "eval_samples_per_second": 34.476,
7
+ "eval_steps_per_second": 1.1,
8
+ "total_flos": 0.0,
9
+ "train_loss": 0.00369019165293624,
10
+ "train_runtime": 6610.8402,
11
+ "train_samples_per_second": 14.485,
12
+ "train_steps_per_second": 0.454
13
+ }
config.json ADDED
@@ -0,0 +1,92 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "c14kevincardenas/beit-large-patch16-384-limb",
3
+ "add_fpn": false,
4
+ "architectures": [
5
+ "LimbXYSequenceModel"
6
+ ],
7
+ "attention_probs_dropout_prob": 0.0,
8
+ "auxiliary_channels": 256,
9
+ "auxiliary_concat_input": false,
10
+ "auxiliary_loss_weight": 0.4,
11
+ "auxiliary_num_convs": 1,
12
+ "d_model": 1024,
13
+ "drop_path_rate": 0.1,
14
+ "finetuning_task": "-sequence-regression",
15
+ "hidden_act": "gelu",
16
+ "hidden_dropout_prob": 0.0,
17
+ "hidden_size": 1024,
18
+ "id2label": {
19
+ "0": "left_foot",
20
+ "1": "left_hand",
21
+ "2": "right_foot",
22
+ "3": "right_hand"
23
+ },
24
+ "image_size": 384,
25
+ "initializer_range": 0.02,
26
+ "intermediate_size": 4096,
27
+ "label2id": {
28
+ "left_foot": "0",
29
+ "left_hand": "1",
30
+ "right_foot": "2",
31
+ "right_hand": "3"
32
+ },
33
+ "layer_norm_eps": 1e-12,
34
+ "layer_scale_init_value": 0.1,
35
+ "model_type": "custom_model",
36
+ "nhead": 2,
37
+ "num_attention_heads": 16,
38
+ "num_channels": 3,
39
+ "num_hidden_layers": 24,
40
+ "num_layers": 1,
41
+ "out_features": [
42
+ "stage24"
43
+ ],
44
+ "out_indices": [
45
+ 24
46
+ ],
47
+ "patch_size": 16,
48
+ "pool_scales": [
49
+ 1,
50
+ 2,
51
+ 3,
52
+ 6
53
+ ],
54
+ "reshape_hidden_states": true,
55
+ "semantic_loss_ignore_index": 255,
56
+ "stage_names": [
57
+ "stem",
58
+ "stage1",
59
+ "stage2",
60
+ "stage3",
61
+ "stage4",
62
+ "stage5",
63
+ "stage6",
64
+ "stage7",
65
+ "stage8",
66
+ "stage9",
67
+ "stage10",
68
+ "stage11",
69
+ "stage12",
70
+ "stage13",
71
+ "stage14",
72
+ "stage15",
73
+ "stage16",
74
+ "stage17",
75
+ "stage18",
76
+ "stage19",
77
+ "stage20",
78
+ "stage21",
79
+ "stage22",
80
+ "stage23",
81
+ "stage24"
82
+ ],
83
+ "torch_dtype": "float32",
84
+ "transformers_version": "4.45.2",
85
+ "use_absolute_position_embeddings": false,
86
+ "use_auxiliary_head": true,
87
+ "use_mask_token": false,
88
+ "use_mean_pooling": true,
89
+ "use_relative_position_bias": true,
90
+ "use_shared_relative_position_bias": false,
91
+ "vocab_size": 8192
92
+ }
eval_results.json ADDED
@@ -0,0 +1,8 @@
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 20.0,
3
+ "eval_loss": 0.005479985848069191,
4
+ "eval_rmse": 0.0740269273519516,
5
+ "eval_runtime": 24.539,
6
+ "eval_samples_per_second": 34.476,
7
+ "eval_steps_per_second": 1.1
8
+ }
model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ba27f4c3ecf783e6e46fad20a88f8444a5b86f84b6d0567d3f13d1f2bf70eda0
3
+ size 1250174024
preprocessor_config.json ADDED
@@ -0,0 +1,28 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "crop_size": {
3
+ "height": 224,
4
+ "width": 224
5
+ },
6
+ "do_center_crop": false,
7
+ "do_normalize": true,
8
+ "do_reduce_labels": false,
9
+ "do_rescale": true,
10
+ "do_resize": true,
11
+ "image_mean": [
12
+ 0.5,
13
+ 0.5,
14
+ 0.5
15
+ ],
16
+ "image_processor_type": "BeitImageProcessor",
17
+ "image_std": [
18
+ 0.5,
19
+ 0.5,
20
+ 0.5
21
+ ],
22
+ "resample": 2,
23
+ "rescale_factor": 0.00392156862745098,
24
+ "size": {
25
+ "height": 384,
26
+ "width": 384
27
+ }
28
+ }
runs/Feb11_18-11-05_galactica.ad.cirange.net/events.out.tfevents.1739297534.galactica.ad.cirange.net.2167895.0 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:4ed3d95ab74b195163cdd8474b1ea50184e8187be1d74e56268acf0a453576ab
3
+ size 6180
runs/Feb11_18-15-34_galactica.ad.cirange.net/events.out.tfevents.1739297739.galactica.ad.cirange.net.2168831.0 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e403194ed6bb73ef4078d2edb74035ba71d8010288d42203f99779974035f8d8
3
+ size 6180
runs/Feb11_18-20-27_galactica.ad.cirange.net/events.out.tfevents.1739298032.galactica.ad.cirange.net.2169870.0 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:0bba554beac14839fe8413ffc498d60aceb5a179a47e807aabf44f09679c0591
3
+ size 6180
runs/Feb11_18-22-10_galactica.ad.cirange.net/events.out.tfevents.1739298135.galactica.ad.cirange.net.2170704.0 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:02a885ee6765d313992832683647e7baee37128ba9a3dde8ba5ea581c6944181
3
+ size 38214
runs/Feb11_18-22-10_galactica.ad.cirange.net/events.out.tfevents.1739305048.galactica.ad.cirange.net.2170704.1 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3e3095cab49abd15913722e448c5f2cf180cfb29aedb4a2ee3a03e1ed489b570
3
+ size 407
runs/Feb11_21-00-55_galactica.ad.cirange.net/events.out.tfevents.1739307661.galactica.ad.cirange.net.2201888.0 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d2b601e7e4de5d6db4df1c0711a16815896afd63ed3899ffad6e131b203f86d0
3
+ size 6180
runs/Feb11_21-02-58_galactica.ad.cirange.net/events.out.tfevents.1739307784.galactica.ad.cirange.net.2202655.0 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5c337f074fa1d898d8435d959c2fb395aebe1553183a51d20a3d2fb51b1a0ffe
3
+ size 6180
runs/Feb11_21-05-23_galactica.ad.cirange.net/events.out.tfevents.1739307928.galactica.ad.cirange.net.2203425.0 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:0eed997ca265e62dc25614a42d3e6aae1fe277d48a67d5b9450a0b78c61f37e9
3
+ size 38214
runs/Feb11_21-05-23_galactica.ad.cirange.net/events.out.tfevents.1739314802.galactica.ad.cirange.net.2203425.1 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:7ba3d195f38f9e99489ff0686b678cd4839ac6607579b69d5b36461160f18dc7
3
+ size 407
runs/Feb11_23-09-01_galactica.ad.cirange.net/events.out.tfevents.1739315347.galactica.ad.cirange.net.2234313.0 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:2b470f251e74b7fcf6d7431a5038b400485f19e7a1b03dfaa2a1d39a9e615524
3
+ size 38214
runs/Feb12_16-23-39_galactica.ad.cirange.net/events.out.tfevents.1739377425.galactica.ad.cirange.net.2273555.0 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b1a432f77758c4e7e4fdb7aa2225c8ade91095028847aece77ebd3d0fac5cb2e
3
+ size 38214
runs/Feb12_16-23-39_galactica.ad.cirange.net/events.out.tfevents.1739384332.galactica.ad.cirange.net.2273555.1 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:33f4454cbade6ad32689683169aca8f47d71f2e399b0cc57fab8ee9ce9ab7349
3
+ size 407
runs/Feb12_18-37-50_galactica.ad.cirange.net/events.out.tfevents.1739385477.galactica.ad.cirange.net.2304333.0 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a533c74aa25e0a110dd9adefe4646488145193dc5edc422abc5355b3d31c3e1f
3
+ size 38214
runs/Feb12_18-37-50_galactica.ad.cirange.net/events.out.tfevents.1739392366.galactica.ad.cirange.net.2304333.1 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d54496c0c5aa487277ea60702416ac4f837869146b18c058394208653e714114
3
+ size 407
runs/Feb19_23-30-29_galactica.ad.cirange.net/events.out.tfevents.1740007834.galactica.ad.cirange.net.3085606.0 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:587252ed7cca808a30b98d0a0ebb59fccbc3c45796b927d04c0beb116278111a
3
+ size 4184
runs/Feb19_23-34-16_galactica.ad.cirange.net/events.out.tfevents.1740008060.galactica.ad.cirange.net.3086972.0 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3e09e7a399512233ea45ca14b451056a8a7a95b182fe15aa738f75f1802c0f3f
3
+ size 6609
runs/Feb19_23-37-01_galactica.ad.cirange.net/events.out.tfevents.1740008226.galactica.ad.cirange.net.3088755.0 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:fc6632adc0f8119547e0edc7bc8686ca4e73873e6302f9b4f101949f2d971a53
3
+ size 4184
runs/Feb19_23-42-22_galactica.ad.cirange.net/events.out.tfevents.1740008546.galactica.ad.cirange.net.3090432.0 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:753d00161ba26911d2bd7e84b442fc7ee3a70c2c3487277036c3127a764f24cf
3
+ size 6195
runs/Feb19_23-43-10_galactica.ad.cirange.net/events.out.tfevents.1740008595.galactica.ad.cirange.net.3091725.0 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:92fefad565661bbfdd3e0a02cdc2ec81e62ea7c513518a55ab1613675ed1f81e
3
+ size 7760
train_results.json ADDED
@@ -0,0 +1,8 @@
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 20.0,
3
+ "total_flos": 0.0,
4
+ "train_loss": 0.00369019165293624,
5
+ "train_runtime": 6610.8402,
6
+ "train_samples_per_second": 14.485,
7
+ "train_steps_per_second": 0.454
8
+ }
trainer_state.json ADDED
@@ -0,0 +1,1062 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": 0.005479985848069191,
3
+ "best_model_checkpoint": "limbxy_seq_t2/checkpoint-2550",
4
+ "epoch": 20.0,
5
+ "eval_steps": 500,
6
+ "global_step": 3000,
7
+ "is_hyper_param_search": false,
8
+ "is_local_process_zero": true,
9
+ "is_world_process_zero": true,
10
+ "log_history": [
11
+ {
12
+ "epoch": 0.16666666666666666,
13
+ "grad_norm": 22046.798828125,
14
+ "learning_rate": 5e-06,
15
+ "loss": 0.0357,
16
+ "step": 25
17
+ },
18
+ {
19
+ "epoch": 0.3333333333333333,
20
+ "grad_norm": 42614.5546875,
21
+ "learning_rate": 1e-05,
22
+ "loss": 0.0278,
23
+ "step": 50
24
+ },
25
+ {
26
+ "epoch": 0.5,
27
+ "grad_norm": 17162.4296875,
28
+ "learning_rate": 1.5e-05,
29
+ "loss": 0.0243,
30
+ "step": 75
31
+ },
32
+ {
33
+ "epoch": 0.6666666666666666,
34
+ "grad_norm": 81025.53125,
35
+ "learning_rate": 2e-05,
36
+ "loss": 0.0188,
37
+ "step": 100
38
+ },
39
+ {
40
+ "epoch": 0.8333333333333334,
41
+ "grad_norm": 52040.0,
42
+ "learning_rate": 2.5e-05,
43
+ "loss": 0.0201,
44
+ "step": 125
45
+ },
46
+ {
47
+ "epoch": 1.0,
48
+ "grad_norm": 16304.97265625,
49
+ "learning_rate": 3e-05,
50
+ "loss": 0.0137,
51
+ "step": 150
52
+ },
53
+ {
54
+ "epoch": 1.0,
55
+ "eval_loss": 0.011878792196512222,
56
+ "eval_rmse": 0.10898987948894501,
57
+ "eval_runtime": 23.3047,
58
+ "eval_samples_per_second": 36.302,
59
+ "eval_steps_per_second": 1.159,
60
+ "step": 150
61
+ },
62
+ {
63
+ "epoch": 1.1666666666666667,
64
+ "grad_norm": 11179.8837890625,
65
+ "learning_rate": 3.5e-05,
66
+ "loss": 0.0143,
67
+ "step": 175
68
+ },
69
+ {
70
+ "epoch": 1.3333333333333333,
71
+ "grad_norm": 40079.0546875,
72
+ "learning_rate": 4e-05,
73
+ "loss": 0.0188,
74
+ "step": 200
75
+ },
76
+ {
77
+ "epoch": 1.5,
78
+ "grad_norm": 22735.115234375,
79
+ "learning_rate": 4.5e-05,
80
+ "loss": 0.012,
81
+ "step": 225
82
+ },
83
+ {
84
+ "epoch": 1.6666666666666665,
85
+ "grad_norm": 54281.23046875,
86
+ "learning_rate": 5e-05,
87
+ "loss": 0.0117,
88
+ "step": 250
89
+ },
90
+ {
91
+ "epoch": 1.8333333333333335,
92
+ "grad_norm": 51424.09375,
93
+ "learning_rate": 4.9545454545454553e-05,
94
+ "loss": 0.0104,
95
+ "step": 275
96
+ },
97
+ {
98
+ "epoch": 2.0,
99
+ "grad_norm": 99929.515625,
100
+ "learning_rate": 4.909090909090909e-05,
101
+ "loss": 0.0112,
102
+ "step": 300
103
+ },
104
+ {
105
+ "epoch": 2.0,
106
+ "eval_loss": 0.011832492426037788,
107
+ "eval_rmse": 0.10877726227045059,
108
+ "eval_runtime": 24.0969,
109
+ "eval_samples_per_second": 35.108,
110
+ "eval_steps_per_second": 1.12,
111
+ "step": 300
112
+ },
113
+ {
114
+ "epoch": 2.1666666666666665,
115
+ "grad_norm": 41221.3046875,
116
+ "learning_rate": 4.863636363636364e-05,
117
+ "loss": 0.0093,
118
+ "step": 325
119
+ },
120
+ {
121
+ "epoch": 2.3333333333333335,
122
+ "grad_norm": 55105.00390625,
123
+ "learning_rate": 4.8181818181818186e-05,
124
+ "loss": 0.0074,
125
+ "step": 350
126
+ },
127
+ {
128
+ "epoch": 2.5,
129
+ "grad_norm": 54882.34765625,
130
+ "learning_rate": 4.772727272727273e-05,
131
+ "loss": 0.0073,
132
+ "step": 375
133
+ },
134
+ {
135
+ "epoch": 2.6666666666666665,
136
+ "grad_norm": 20329.99609375,
137
+ "learning_rate": 4.7272727272727275e-05,
138
+ "loss": 0.0079,
139
+ "step": 400
140
+ },
141
+ {
142
+ "epoch": 2.8333333333333335,
143
+ "grad_norm": 29952.67578125,
144
+ "learning_rate": 4.681818181818182e-05,
145
+ "loss": 0.0082,
146
+ "step": 425
147
+ },
148
+ {
149
+ "epoch": 3.0,
150
+ "grad_norm": 36055.50390625,
151
+ "learning_rate": 4.636363636363636e-05,
152
+ "loss": 0.0074,
153
+ "step": 450
154
+ },
155
+ {
156
+ "epoch": 3.0,
157
+ "eval_loss": 0.010063888505101204,
158
+ "eval_rmse": 0.10031893104314804,
159
+ "eval_runtime": 24.7801,
160
+ "eval_samples_per_second": 34.14,
161
+ "eval_steps_per_second": 1.09,
162
+ "step": 450
163
+ },
164
+ {
165
+ "epoch": 3.1666666666666665,
166
+ "grad_norm": 8092.63427734375,
167
+ "learning_rate": 4.5909090909090914e-05,
168
+ "loss": 0.0059,
169
+ "step": 475
170
+ },
171
+ {
172
+ "epoch": 3.3333333333333335,
173
+ "grad_norm": 39814.0,
174
+ "learning_rate": 4.545454545454546e-05,
175
+ "loss": 0.0065,
176
+ "step": 500
177
+ },
178
+ {
179
+ "epoch": 3.5,
180
+ "grad_norm": 7769.53125,
181
+ "learning_rate": 4.5e-05,
182
+ "loss": 0.0055,
183
+ "step": 525
184
+ },
185
+ {
186
+ "epoch": 3.6666666666666665,
187
+ "grad_norm": 27703.099609375,
188
+ "learning_rate": 4.454545454545455e-05,
189
+ "loss": 0.0057,
190
+ "step": 550
191
+ },
192
+ {
193
+ "epoch": 3.8333333333333335,
194
+ "grad_norm": 20657.283203125,
195
+ "learning_rate": 4.409090909090909e-05,
196
+ "loss": 0.0072,
197
+ "step": 575
198
+ },
199
+ {
200
+ "epoch": 4.0,
201
+ "grad_norm": 20261.841796875,
202
+ "learning_rate": 4.3636363636363636e-05,
203
+ "loss": 0.0064,
204
+ "step": 600
205
+ },
206
+ {
207
+ "epoch": 4.0,
208
+ "eval_loss": 0.006974302232265472,
209
+ "eval_rmse": 0.08351229131221771,
210
+ "eval_runtime": 23.547,
211
+ "eval_samples_per_second": 35.928,
212
+ "eval_steps_per_second": 1.147,
213
+ "step": 600
214
+ },
215
+ {
216
+ "epoch": 4.166666666666667,
217
+ "grad_norm": 18018.79296875,
218
+ "learning_rate": 4.318181818181819e-05,
219
+ "loss": 0.004,
220
+ "step": 625
221
+ },
222
+ {
223
+ "epoch": 4.333333333333333,
224
+ "grad_norm": 11542.9560546875,
225
+ "learning_rate": 4.2727272727272724e-05,
226
+ "loss": 0.0044,
227
+ "step": 650
228
+ },
229
+ {
230
+ "epoch": 4.5,
231
+ "grad_norm": 34041.24609375,
232
+ "learning_rate": 4.2272727272727275e-05,
233
+ "loss": 0.0056,
234
+ "step": 675
235
+ },
236
+ {
237
+ "epoch": 4.666666666666667,
238
+ "grad_norm": 11050.5517578125,
239
+ "learning_rate": 4.181818181818182e-05,
240
+ "loss": 0.0046,
241
+ "step": 700
242
+ },
243
+ {
244
+ "epoch": 4.833333333333333,
245
+ "grad_norm": 13684.1787109375,
246
+ "learning_rate": 4.1363636363636364e-05,
247
+ "loss": 0.0036,
248
+ "step": 725
249
+ },
250
+ {
251
+ "epoch": 5.0,
252
+ "grad_norm": 20930.271484375,
253
+ "learning_rate": 4.0909090909090915e-05,
254
+ "loss": 0.0039,
255
+ "step": 750
256
+ },
257
+ {
258
+ "epoch": 5.0,
259
+ "eval_loss": 0.006916658021509647,
260
+ "eval_rmse": 0.0831664502620697,
261
+ "eval_runtime": 23.3555,
262
+ "eval_samples_per_second": 36.223,
263
+ "eval_steps_per_second": 1.156,
264
+ "step": 750
265
+ },
266
+ {
267
+ "epoch": 5.166666666666667,
268
+ "grad_norm": 6945.47314453125,
269
+ "learning_rate": 4.045454545454546e-05,
270
+ "loss": 0.0032,
271
+ "step": 775
272
+ },
273
+ {
274
+ "epoch": 5.333333333333333,
275
+ "grad_norm": 21823.470703125,
276
+ "learning_rate": 4e-05,
277
+ "loss": 0.004,
278
+ "step": 800
279
+ },
280
+ {
281
+ "epoch": 5.5,
282
+ "grad_norm": 6344.1767578125,
283
+ "learning_rate": 3.954545454545455e-05,
284
+ "loss": 0.0032,
285
+ "step": 825
286
+ },
287
+ {
288
+ "epoch": 5.666666666666667,
289
+ "grad_norm": 28923.7578125,
290
+ "learning_rate": 3.909090909090909e-05,
291
+ "loss": 0.0031,
292
+ "step": 850
293
+ },
294
+ {
295
+ "epoch": 5.833333333333333,
296
+ "grad_norm": 18844.93359375,
297
+ "learning_rate": 3.8636363636363636e-05,
298
+ "loss": 0.0037,
299
+ "step": 875
300
+ },
301
+ {
302
+ "epoch": 6.0,
303
+ "grad_norm": 14070.8798828125,
304
+ "learning_rate": 3.818181818181819e-05,
305
+ "loss": 0.0033,
306
+ "step": 900
307
+ },
308
+ {
309
+ "epoch": 6.0,
310
+ "eval_loss": 0.007384075783193111,
311
+ "eval_rmse": 0.08593064546585083,
312
+ "eval_runtime": 24.2639,
313
+ "eval_samples_per_second": 34.867,
314
+ "eval_steps_per_second": 1.113,
315
+ "step": 900
316
+ },
317
+ {
318
+ "epoch": 6.166666666666667,
319
+ "grad_norm": 7107.0751953125,
320
+ "learning_rate": 3.7727272727272725e-05,
321
+ "loss": 0.0025,
322
+ "step": 925
323
+ },
324
+ {
325
+ "epoch": 6.333333333333333,
326
+ "grad_norm": 26281.095703125,
327
+ "learning_rate": 3.7272727272727276e-05,
328
+ "loss": 0.0033,
329
+ "step": 950
330
+ },
331
+ {
332
+ "epoch": 6.5,
333
+ "grad_norm": 6169.37744140625,
334
+ "learning_rate": 3.681818181818182e-05,
335
+ "loss": 0.0026,
336
+ "step": 975
337
+ },
338
+ {
339
+ "epoch": 6.666666666666667,
340
+ "grad_norm": 15514.4677734375,
341
+ "learning_rate": 3.6363636363636364e-05,
342
+ "loss": 0.0027,
343
+ "step": 1000
344
+ },
345
+ {
346
+ "epoch": 6.833333333333333,
347
+ "grad_norm": 4963.125,
348
+ "learning_rate": 3.590909090909091e-05,
349
+ "loss": 0.0027,
350
+ "step": 1025
351
+ },
352
+ {
353
+ "epoch": 7.0,
354
+ "grad_norm": 20854.5859375,
355
+ "learning_rate": 3.545454545454546e-05,
356
+ "loss": 0.0026,
357
+ "step": 1050
358
+ },
359
+ {
360
+ "epoch": 7.0,
361
+ "eval_loss": 0.006149372085928917,
362
+ "eval_rmse": 0.07841793447732925,
363
+ "eval_runtime": 24.4157,
364
+ "eval_samples_per_second": 34.65,
365
+ "eval_steps_per_second": 1.106,
366
+ "step": 1050
367
+ },
368
+ {
369
+ "epoch": 7.166666666666667,
370
+ "grad_norm": 9573.7587890625,
371
+ "learning_rate": 3.5e-05,
372
+ "loss": 0.0027,
373
+ "step": 1075
374
+ },
375
+ {
376
+ "epoch": 7.333333333333333,
377
+ "grad_norm": 25957.48046875,
378
+ "learning_rate": 3.454545454545455e-05,
379
+ "loss": 0.0024,
380
+ "step": 1100
381
+ },
382
+ {
383
+ "epoch": 7.5,
384
+ "grad_norm": 14093.68359375,
385
+ "learning_rate": 3.409090909090909e-05,
386
+ "loss": 0.0026,
387
+ "step": 1125
388
+ },
389
+ {
390
+ "epoch": 7.666666666666667,
391
+ "grad_norm": 4025.608642578125,
392
+ "learning_rate": 3.3636363636363636e-05,
393
+ "loss": 0.002,
394
+ "step": 1150
395
+ },
396
+ {
397
+ "epoch": 7.833333333333333,
398
+ "grad_norm": 6692.55908203125,
399
+ "learning_rate": 3.318181818181819e-05,
400
+ "loss": 0.0029,
401
+ "step": 1175
402
+ },
403
+ {
404
+ "epoch": 8.0,
405
+ "grad_norm": 4868.02978515625,
406
+ "learning_rate": 3.272727272727273e-05,
407
+ "loss": 0.002,
408
+ "step": 1200
409
+ },
410
+ {
411
+ "epoch": 8.0,
412
+ "eval_loss": 0.006902625784277916,
413
+ "eval_rmse": 0.08308204263448715,
414
+ "eval_runtime": 23.2531,
415
+ "eval_samples_per_second": 36.382,
416
+ "eval_steps_per_second": 1.161,
417
+ "step": 1200
418
+ },
419
+ {
420
+ "epoch": 8.166666666666666,
421
+ "grad_norm": 16112.8388671875,
422
+ "learning_rate": 3.2272727272727276e-05,
423
+ "loss": 0.0019,
424
+ "step": 1225
425
+ },
426
+ {
427
+ "epoch": 8.333333333333334,
428
+ "grad_norm": 7358.75732421875,
429
+ "learning_rate": 3.181818181818182e-05,
430
+ "loss": 0.0021,
431
+ "step": 1250
432
+ },
433
+ {
434
+ "epoch": 8.5,
435
+ "grad_norm": 6156.6611328125,
436
+ "learning_rate": 3.1363636363636365e-05,
437
+ "loss": 0.0017,
438
+ "step": 1275
439
+ },
440
+ {
441
+ "epoch": 8.666666666666666,
442
+ "grad_norm": 10837.3037109375,
443
+ "learning_rate": 3.090909090909091e-05,
444
+ "loss": 0.0024,
445
+ "step": 1300
446
+ },
447
+ {
448
+ "epoch": 8.833333333333334,
449
+ "grad_norm": 12637.4150390625,
450
+ "learning_rate": 3.0454545454545456e-05,
451
+ "loss": 0.0018,
452
+ "step": 1325
453
+ },
454
+ {
455
+ "epoch": 9.0,
456
+ "grad_norm": 23124.64453125,
457
+ "learning_rate": 3e-05,
458
+ "loss": 0.0018,
459
+ "step": 1350
460
+ },
461
+ {
462
+ "epoch": 9.0,
463
+ "eval_loss": 0.005730382166802883,
464
+ "eval_rmse": 0.07569928467273712,
465
+ "eval_runtime": 23.3861,
466
+ "eval_samples_per_second": 36.175,
467
+ "eval_steps_per_second": 1.155,
468
+ "step": 1350
469
+ },
470
+ {
471
+ "epoch": 9.166666666666666,
472
+ "grad_norm": 6718.92919921875,
473
+ "learning_rate": 2.954545454545455e-05,
474
+ "loss": 0.0015,
475
+ "step": 1375
476
+ },
477
+ {
478
+ "epoch": 9.333333333333334,
479
+ "grad_norm": 10810.6953125,
480
+ "learning_rate": 2.909090909090909e-05,
481
+ "loss": 0.0013,
482
+ "step": 1400
483
+ },
484
+ {
485
+ "epoch": 9.5,
486
+ "grad_norm": 12909.9033203125,
487
+ "learning_rate": 2.863636363636364e-05,
488
+ "loss": 0.0015,
489
+ "step": 1425
490
+ },
491
+ {
492
+ "epoch": 9.666666666666666,
493
+ "grad_norm": 5796.60693359375,
494
+ "learning_rate": 2.818181818181818e-05,
495
+ "loss": 0.0013,
496
+ "step": 1450
497
+ },
498
+ {
499
+ "epoch": 9.833333333333334,
500
+ "grad_norm": 23500.55078125,
501
+ "learning_rate": 2.772727272727273e-05,
502
+ "loss": 0.0017,
503
+ "step": 1475
504
+ },
505
+ {
506
+ "epoch": 10.0,
507
+ "grad_norm": 2772.444580078125,
508
+ "learning_rate": 2.7272727272727273e-05,
509
+ "loss": 0.002,
510
+ "step": 1500
511
+ },
512
+ {
513
+ "epoch": 10.0,
514
+ "eval_loss": 0.005920641124248505,
515
+ "eval_rmse": 0.0769456997513771,
516
+ "eval_runtime": 24.5095,
517
+ "eval_samples_per_second": 34.517,
518
+ "eval_steps_per_second": 1.102,
519
+ "step": 1500
520
+ },
521
+ {
522
+ "epoch": 10.166666666666666,
523
+ "grad_norm": 15935.796875,
524
+ "learning_rate": 2.681818181818182e-05,
525
+ "loss": 0.0011,
526
+ "step": 1525
527
+ },
528
+ {
529
+ "epoch": 10.333333333333334,
530
+ "grad_norm": 13381.3623046875,
531
+ "learning_rate": 2.636363636363636e-05,
532
+ "loss": 0.0014,
533
+ "step": 1550
534
+ },
535
+ {
536
+ "epoch": 10.5,
537
+ "grad_norm": 8558.2841796875,
538
+ "learning_rate": 2.590909090909091e-05,
539
+ "loss": 0.0012,
540
+ "step": 1575
541
+ },
542
+ {
543
+ "epoch": 10.666666666666666,
544
+ "grad_norm": 13277.8525390625,
545
+ "learning_rate": 2.5454545454545454e-05,
546
+ "loss": 0.0013,
547
+ "step": 1600
548
+ },
549
+ {
550
+ "epoch": 10.833333333333334,
551
+ "grad_norm": 11197.5478515625,
552
+ "learning_rate": 2.5e-05,
553
+ "loss": 0.0012,
554
+ "step": 1625
555
+ },
556
+ {
557
+ "epoch": 11.0,
558
+ "grad_norm": 3204.55810546875,
559
+ "learning_rate": 2.4545454545454545e-05,
560
+ "loss": 0.0012,
561
+ "step": 1650
562
+ },
563
+ {
564
+ "epoch": 11.0,
565
+ "eval_loss": 0.005719946697354317,
566
+ "eval_rmse": 0.07563033699989319,
567
+ "eval_runtime": 25.7743,
568
+ "eval_samples_per_second": 32.823,
569
+ "eval_steps_per_second": 1.048,
570
+ "step": 1650
571
+ },
572
+ {
573
+ "epoch": 11.166666666666666,
574
+ "grad_norm": 7068.5830078125,
575
+ "learning_rate": 2.4090909090909093e-05,
576
+ "loss": 0.001,
577
+ "step": 1675
578
+ },
579
+ {
580
+ "epoch": 11.333333333333334,
581
+ "grad_norm": 6493.97265625,
582
+ "learning_rate": 2.3636363636363637e-05,
583
+ "loss": 0.0011,
584
+ "step": 1700
585
+ },
586
+ {
587
+ "epoch": 11.5,
588
+ "grad_norm": 6777.013671875,
589
+ "learning_rate": 2.318181818181818e-05,
590
+ "loss": 0.0011,
591
+ "step": 1725
592
+ },
593
+ {
594
+ "epoch": 11.666666666666666,
595
+ "grad_norm": 6261.11767578125,
596
+ "learning_rate": 2.272727272727273e-05,
597
+ "loss": 0.0011,
598
+ "step": 1750
599
+ },
600
+ {
601
+ "epoch": 11.833333333333334,
602
+ "grad_norm": 8973.4599609375,
603
+ "learning_rate": 2.2272727272727274e-05,
604
+ "loss": 0.0012,
605
+ "step": 1775
606
+ },
607
+ {
608
+ "epoch": 12.0,
609
+ "grad_norm": 11911.302734375,
610
+ "learning_rate": 2.1818181818181818e-05,
611
+ "loss": 0.0011,
612
+ "step": 1800
613
+ },
614
+ {
615
+ "epoch": 12.0,
616
+ "eval_loss": 0.005933709908276796,
617
+ "eval_rmse": 0.07703057676553726,
618
+ "eval_runtime": 23.8679,
619
+ "eval_samples_per_second": 35.445,
620
+ "eval_steps_per_second": 1.131,
621
+ "step": 1800
622
+ },
623
+ {
624
+ "epoch": 12.166666666666666,
625
+ "grad_norm": 7263.52197265625,
626
+ "learning_rate": 2.1363636363636362e-05,
627
+ "loss": 0.001,
628
+ "step": 1825
629
+ },
630
+ {
631
+ "epoch": 12.333333333333334,
632
+ "grad_norm": 2190.18115234375,
633
+ "learning_rate": 2.090909090909091e-05,
634
+ "loss": 0.0008,
635
+ "step": 1850
636
+ },
637
+ {
638
+ "epoch": 12.5,
639
+ "grad_norm": 17004.29296875,
640
+ "learning_rate": 2.0454545454545457e-05,
641
+ "loss": 0.0009,
642
+ "step": 1875
643
+ },
644
+ {
645
+ "epoch": 12.666666666666666,
646
+ "grad_norm": 3601.951416015625,
647
+ "learning_rate": 2e-05,
648
+ "loss": 0.0008,
649
+ "step": 1900
650
+ },
651
+ {
652
+ "epoch": 12.833333333333334,
653
+ "grad_norm": 16977.19140625,
654
+ "learning_rate": 1.9545454545454546e-05,
655
+ "loss": 0.0009,
656
+ "step": 1925
657
+ },
658
+ {
659
+ "epoch": 13.0,
660
+ "grad_norm": 9962.5048828125,
661
+ "learning_rate": 1.9090909090909094e-05,
662
+ "loss": 0.0009,
663
+ "step": 1950
664
+ },
665
+ {
666
+ "epoch": 13.0,
667
+ "eval_loss": 0.005788400769233704,
668
+ "eval_rmse": 0.0760815367102623,
669
+ "eval_runtime": 23.6915,
670
+ "eval_samples_per_second": 35.709,
671
+ "eval_steps_per_second": 1.14,
672
+ "step": 1950
673
+ },
674
+ {
675
+ "epoch": 13.166666666666666,
676
+ "grad_norm": 6503.86474609375,
677
+ "learning_rate": 1.8636363636363638e-05,
678
+ "loss": 0.0009,
679
+ "step": 1975
680
+ },
681
+ {
682
+ "epoch": 13.333333333333334,
683
+ "grad_norm": 3315.51904296875,
684
+ "learning_rate": 1.8181818181818182e-05,
685
+ "loss": 0.0008,
686
+ "step": 2000
687
+ },
688
+ {
689
+ "epoch": 13.5,
690
+ "grad_norm": 6505.07373046875,
691
+ "learning_rate": 1.772727272727273e-05,
692
+ "loss": 0.0007,
693
+ "step": 2025
694
+ },
695
+ {
696
+ "epoch": 13.666666666666666,
697
+ "grad_norm": 8015.05712890625,
698
+ "learning_rate": 1.7272727272727274e-05,
699
+ "loss": 0.0007,
700
+ "step": 2050
701
+ },
702
+ {
703
+ "epoch": 13.833333333333334,
704
+ "grad_norm": 19194.6015625,
705
+ "learning_rate": 1.6818181818181818e-05,
706
+ "loss": 0.0008,
707
+ "step": 2075
708
+ },
709
+ {
710
+ "epoch": 14.0,
711
+ "grad_norm": 9773.720703125,
712
+ "learning_rate": 1.6363636363636366e-05,
713
+ "loss": 0.0008,
714
+ "step": 2100
715
+ },
716
+ {
717
+ "epoch": 14.0,
718
+ "eval_loss": 0.005640079732984304,
719
+ "eval_rmse": 0.07510046660900116,
720
+ "eval_runtime": 25.0673,
721
+ "eval_samples_per_second": 33.749,
722
+ "eval_steps_per_second": 1.077,
723
+ "step": 2100
724
+ },
725
+ {
726
+ "epoch": 14.166666666666666,
727
+ "grad_norm": 5335.89013671875,
728
+ "learning_rate": 1.590909090909091e-05,
729
+ "loss": 0.0007,
730
+ "step": 2125
731
+ },
732
+ {
733
+ "epoch": 14.333333333333334,
734
+ "grad_norm": 5937.32421875,
735
+ "learning_rate": 1.5454545454545454e-05,
736
+ "loss": 0.0007,
737
+ "step": 2150
738
+ },
739
+ {
740
+ "epoch": 14.5,
741
+ "grad_norm": 5748.50048828125,
742
+ "learning_rate": 1.5e-05,
743
+ "loss": 0.0007,
744
+ "step": 2175
745
+ },
746
+ {
747
+ "epoch": 14.666666666666666,
748
+ "grad_norm": 7561.5009765625,
749
+ "learning_rate": 1.4545454545454545e-05,
750
+ "loss": 0.0007,
751
+ "step": 2200
752
+ },
753
+ {
754
+ "epoch": 14.833333333333334,
755
+ "grad_norm": 2272.05419921875,
756
+ "learning_rate": 1.409090909090909e-05,
757
+ "loss": 0.0007,
758
+ "step": 2225
759
+ },
760
+ {
761
+ "epoch": 15.0,
762
+ "grad_norm": 4282.1005859375,
763
+ "learning_rate": 1.3636363636363637e-05,
764
+ "loss": 0.0008,
765
+ "step": 2250
766
+ },
767
+ {
768
+ "epoch": 15.0,
769
+ "eval_loss": 0.0057137333787977695,
770
+ "eval_rmse": 0.07558923959732056,
771
+ "eval_runtime": 24.0972,
772
+ "eval_samples_per_second": 35.108,
773
+ "eval_steps_per_second": 1.12,
774
+ "step": 2250
775
+ },
776
+ {
777
+ "epoch": 15.166666666666666,
778
+ "grad_norm": 4606.4609375,
779
+ "learning_rate": 1.318181818181818e-05,
780
+ "loss": 0.0006,
781
+ "step": 2275
782
+ },
783
+ {
784
+ "epoch": 15.333333333333334,
785
+ "grad_norm": 9886.646484375,
786
+ "learning_rate": 1.2727272727272727e-05,
787
+ "loss": 0.0006,
788
+ "step": 2300
789
+ },
790
+ {
791
+ "epoch": 15.5,
792
+ "grad_norm": 4568.2802734375,
793
+ "learning_rate": 1.2272727272727273e-05,
794
+ "loss": 0.0005,
795
+ "step": 2325
796
+ },
797
+ {
798
+ "epoch": 15.666666666666666,
799
+ "grad_norm": 8178.26025390625,
800
+ "learning_rate": 1.1818181818181819e-05,
801
+ "loss": 0.0006,
802
+ "step": 2350
803
+ },
804
+ {
805
+ "epoch": 15.833333333333334,
806
+ "grad_norm": 3068.3349609375,
807
+ "learning_rate": 1.1363636363636365e-05,
808
+ "loss": 0.0006,
809
+ "step": 2375
810
+ },
811
+ {
812
+ "epoch": 16.0,
813
+ "grad_norm": 7865.693359375,
814
+ "learning_rate": 1.0909090909090909e-05,
815
+ "loss": 0.0006,
816
+ "step": 2400
817
+ },
818
+ {
819
+ "epoch": 16.0,
820
+ "eval_loss": 0.0056036608293652534,
821
+ "eval_rmse": 0.07485760003328323,
822
+ "eval_runtime": 24.1386,
823
+ "eval_samples_per_second": 35.048,
824
+ "eval_steps_per_second": 1.119,
825
+ "step": 2400
826
+ },
827
+ {
828
+ "epoch": 16.166666666666668,
829
+ "grad_norm": 12935.966796875,
830
+ "learning_rate": 1.0454545454545455e-05,
831
+ "loss": 0.0005,
832
+ "step": 2425
833
+ },
834
+ {
835
+ "epoch": 16.333333333333332,
836
+ "grad_norm": 2572.318359375,
837
+ "learning_rate": 1e-05,
838
+ "loss": 0.0004,
839
+ "step": 2450
840
+ },
841
+ {
842
+ "epoch": 16.5,
843
+ "grad_norm": 3913.558837890625,
844
+ "learning_rate": 9.545454545454547e-06,
845
+ "loss": 0.0005,
846
+ "step": 2475
847
+ },
848
+ {
849
+ "epoch": 16.666666666666668,
850
+ "grad_norm": 7128.30126953125,
851
+ "learning_rate": 9.090909090909091e-06,
852
+ "loss": 0.0005,
853
+ "step": 2500
854
+ },
855
+ {
856
+ "epoch": 16.833333333333332,
857
+ "grad_norm": 11092.6220703125,
858
+ "learning_rate": 8.636363636363637e-06,
859
+ "loss": 0.0005,
860
+ "step": 2525
861
+ },
862
+ {
863
+ "epoch": 17.0,
864
+ "grad_norm": 7797.11767578125,
865
+ "learning_rate": 8.181818181818183e-06,
866
+ "loss": 0.0006,
867
+ "step": 2550
868
+ },
869
+ {
870
+ "epoch": 17.0,
871
+ "eval_loss": 0.005479985848069191,
872
+ "eval_rmse": 0.0740269273519516,
873
+ "eval_runtime": 23.866,
874
+ "eval_samples_per_second": 35.448,
875
+ "eval_steps_per_second": 1.131,
876
+ "step": 2550
877
+ },
878
+ {
879
+ "epoch": 17.166666666666668,
880
+ "grad_norm": 3903.398193359375,
881
+ "learning_rate": 7.727272727272727e-06,
882
+ "loss": 0.0004,
883
+ "step": 2575
884
+ },
885
+ {
886
+ "epoch": 17.333333333333332,
887
+ "grad_norm": 9970.853515625,
888
+ "learning_rate": 7.272727272727272e-06,
889
+ "loss": 0.0004,
890
+ "step": 2600
891
+ },
892
+ {
893
+ "epoch": 17.5,
894
+ "grad_norm": 10561.7421875,
895
+ "learning_rate": 6.818181818181818e-06,
896
+ "loss": 0.0004,
897
+ "step": 2625
898
+ },
899
+ {
900
+ "epoch": 17.666666666666668,
901
+ "grad_norm": 4177.15771484375,
902
+ "learning_rate": 6.363636363636363e-06,
903
+ "loss": 0.0004,
904
+ "step": 2650
905
+ },
906
+ {
907
+ "epoch": 17.833333333333332,
908
+ "grad_norm": 7322.50634765625,
909
+ "learning_rate": 5.909090909090909e-06,
910
+ "loss": 0.0004,
911
+ "step": 2675
912
+ },
913
+ {
914
+ "epoch": 18.0,
915
+ "grad_norm": 13072.9580078125,
916
+ "learning_rate": 5.4545454545454545e-06,
917
+ "loss": 0.0004,
918
+ "step": 2700
919
+ },
920
+ {
921
+ "epoch": 18.0,
922
+ "eval_loss": 0.005703671369701624,
923
+ "eval_rmse": 0.07552265375852585,
924
+ "eval_runtime": 24.3429,
925
+ "eval_samples_per_second": 34.754,
926
+ "eval_steps_per_second": 1.109,
927
+ "step": 2700
928
+ },
929
+ {
930
+ "epoch": 18.166666666666668,
931
+ "grad_norm": 3103.575927734375,
932
+ "learning_rate": 5e-06,
933
+ "loss": 0.0004,
934
+ "step": 2725
935
+ },
936
+ {
937
+ "epoch": 18.333333333333332,
938
+ "grad_norm": 2713.65869140625,
939
+ "learning_rate": 4.5454545454545455e-06,
940
+ "loss": 0.0003,
941
+ "step": 2750
942
+ },
943
+ {
944
+ "epoch": 18.5,
945
+ "grad_norm": 2231.278564453125,
946
+ "learning_rate": 4.0909090909090915e-06,
947
+ "loss": 0.0003,
948
+ "step": 2775
949
+ },
950
+ {
951
+ "epoch": 18.666666666666668,
952
+ "grad_norm": 6196.794921875,
953
+ "learning_rate": 3.636363636363636e-06,
954
+ "loss": 0.0003,
955
+ "step": 2800
956
+ },
957
+ {
958
+ "epoch": 18.833333333333332,
959
+ "grad_norm": 1392.4776611328125,
960
+ "learning_rate": 3.1818181818181817e-06,
961
+ "loss": 0.0003,
962
+ "step": 2825
963
+ },
964
+ {
965
+ "epoch": 19.0,
966
+ "grad_norm": 2869.9462890625,
967
+ "learning_rate": 2.7272727272727272e-06,
968
+ "loss": 0.0004,
969
+ "step": 2850
970
+ },
971
+ {
972
+ "epoch": 19.0,
973
+ "eval_loss": 0.005522754043340683,
974
+ "eval_rmse": 0.07431523501873016,
975
+ "eval_runtime": 23.5254,
976
+ "eval_samples_per_second": 35.961,
977
+ "eval_steps_per_second": 1.148,
978
+ "step": 2850
979
+ },
980
+ {
981
+ "epoch": 19.166666666666668,
982
+ "grad_norm": 1801.104248046875,
983
+ "learning_rate": 2.2727272727272728e-06,
984
+ "loss": 0.0003,
985
+ "step": 2875
986
+ },
987
+ {
988
+ "epoch": 19.333333333333332,
989
+ "grad_norm": 2785.420654296875,
990
+ "learning_rate": 1.818181818181818e-06,
991
+ "loss": 0.0003,
992
+ "step": 2900
993
+ },
994
+ {
995
+ "epoch": 19.5,
996
+ "grad_norm": 3792.26416015625,
997
+ "learning_rate": 1.3636363636363636e-06,
998
+ "loss": 0.0003,
999
+ "step": 2925
1000
+ },
1001
+ {
1002
+ "epoch": 19.666666666666668,
1003
+ "grad_norm": 6763.56982421875,
1004
+ "learning_rate": 9.09090909090909e-07,
1005
+ "loss": 0.0003,
1006
+ "step": 2950
1007
+ },
1008
+ {
1009
+ "epoch": 19.833333333333332,
1010
+ "grad_norm": 6747.69091796875,
1011
+ "learning_rate": 4.545454545454545e-07,
1012
+ "loss": 0.0003,
1013
+ "step": 2975
1014
+ },
1015
+ {
1016
+ "epoch": 20.0,
1017
+ "grad_norm": 2657.044677734375,
1018
+ "learning_rate": 0.0,
1019
+ "loss": 0.0002,
1020
+ "step": 3000
1021
+ },
1022
+ {
1023
+ "epoch": 20.0,
1024
+ "eval_loss": 0.00548164127394557,
1025
+ "eval_rmse": 0.07403810322284698,
1026
+ "eval_runtime": 24.066,
1027
+ "eval_samples_per_second": 35.153,
1028
+ "eval_steps_per_second": 1.122,
1029
+ "step": 3000
1030
+ },
1031
+ {
1032
+ "epoch": 20.0,
1033
+ "step": 3000,
1034
+ "total_flos": 0.0,
1035
+ "train_loss": 0.00369019165293624,
1036
+ "train_runtime": 6610.8402,
1037
+ "train_samples_per_second": 14.485,
1038
+ "train_steps_per_second": 0.454
1039
+ }
1040
+ ],
1041
+ "logging_steps": 25,
1042
+ "max_steps": 3000,
1043
+ "num_input_tokens_seen": 0,
1044
+ "num_train_epochs": 20,
1045
+ "save_steps": 500,
1046
+ "stateful_callbacks": {
1047
+ "TrainerControl": {
1048
+ "args": {
1049
+ "should_epoch_stop": false,
1050
+ "should_evaluate": false,
1051
+ "should_log": false,
1052
+ "should_save": true,
1053
+ "should_training_stop": true
1054
+ },
1055
+ "attributes": {}
1056
+ }
1057
+ },
1058
+ "total_flos": 0.0,
1059
+ "train_batch_size": 32,
1060
+ "trial_name": null,
1061
+ "trial_params": null
1062
+ }
training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:db8e75a8e21456e88704217fbd86263bd0b55b648e6b17377db5ef5dc0e9f445
3
+ size 5240