Training in progress, step 4800, checkpoint
Browse files- checkpoint-4800/config.json +62 -0
- checkpoint-4800/model.safetensors +3 -0
- checkpoint-4800/optimizer.pt +3 -0
- checkpoint-4800/preprocessor_config.json +26 -0
- checkpoint-4800/rng_state.pth +3 -0
- checkpoint-4800/scheduler.pt +3 -0
- checkpoint-4800/trainer_state.json +705 -0
- checkpoint-4800/training_args.bin +3 -0
checkpoint-4800/config.json
ADDED
|
@@ -0,0 +1,62 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "facebook/detr-resnet-50-dc5",
|
| 3 |
+
"activation_dropout": 0.0,
|
| 4 |
+
"activation_function": "relu",
|
| 5 |
+
"architectures": [
|
| 6 |
+
"DetrForObjectDetection"
|
| 7 |
+
],
|
| 8 |
+
"attention_dropout": 0.0,
|
| 9 |
+
"auxiliary_loss": false,
|
| 10 |
+
"backbone": "resnet50",
|
| 11 |
+
"backbone_config": null,
|
| 12 |
+
"backbone_kwargs": {
|
| 13 |
+
"in_chans": 3,
|
| 14 |
+
"out_indices": [
|
| 15 |
+
1,
|
| 16 |
+
2,
|
| 17 |
+
3,
|
| 18 |
+
4
|
| 19 |
+
],
|
| 20 |
+
"output_stride": 16
|
| 21 |
+
},
|
| 22 |
+
"bbox_cost": 5,
|
| 23 |
+
"bbox_loss_coefficient": 5,
|
| 24 |
+
"class_cost": 1,
|
| 25 |
+
"classifier_dropout": 0.0,
|
| 26 |
+
"d_model": 256,
|
| 27 |
+
"decoder_attention_heads": 8,
|
| 28 |
+
"decoder_ffn_dim": 2048,
|
| 29 |
+
"decoder_layerdrop": 0.0,
|
| 30 |
+
"decoder_layers": 6,
|
| 31 |
+
"dice_loss_coefficient": 1,
|
| 32 |
+
"dilation": true,
|
| 33 |
+
"dropout": 0.1,
|
| 34 |
+
"encoder_attention_heads": 8,
|
| 35 |
+
"encoder_ffn_dim": 2048,
|
| 36 |
+
"encoder_layerdrop": 0.0,
|
| 37 |
+
"encoder_layers": 6,
|
| 38 |
+
"eos_coefficient": 0.1,
|
| 39 |
+
"giou_cost": 2,
|
| 40 |
+
"giou_loss_coefficient": 2,
|
| 41 |
+
"id2label": {
|
| 42 |
+
"0": "segment"
|
| 43 |
+
},
|
| 44 |
+
"init_std": 0.02,
|
| 45 |
+
"init_xavier_std": 1.0,
|
| 46 |
+
"is_encoder_decoder": true,
|
| 47 |
+
"label2id": {
|
| 48 |
+
"segment": 0
|
| 49 |
+
},
|
| 50 |
+
"mask_loss_coefficient": 1,
|
| 51 |
+
"max_position_embeddings": 1024,
|
| 52 |
+
"model_type": "detr",
|
| 53 |
+
"num_channels": 3,
|
| 54 |
+
"num_hidden_layers": 6,
|
| 55 |
+
"num_queries": 500,
|
| 56 |
+
"position_embedding_type": "sine",
|
| 57 |
+
"scale_embedding": false,
|
| 58 |
+
"torch_dtype": "float32",
|
| 59 |
+
"transformers_version": "4.44.1",
|
| 60 |
+
"use_pretrained_backbone": true,
|
| 61 |
+
"use_timm_backbone": true
|
| 62 |
+
}
|
checkpoint-4800/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:4c39fc25a0d1bc3335c30d1ef64da9919134ecf02a3e991f7b49d8b13bfe44d3
|
| 3 |
+
size 166904424
|
checkpoint-4800/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:82c483e422cd5555010dcd4857b25d2264272e0c729d82fe6f07fa1a5ba19d16
|
| 3 |
+
size 331310970
|
checkpoint-4800/preprocessor_config.json
ADDED
|
@@ -0,0 +1,26 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"do_convert_annotations": true,
|
| 3 |
+
"do_normalize": true,
|
| 4 |
+
"do_pad": true,
|
| 5 |
+
"do_rescale": true,
|
| 6 |
+
"do_resize": true,
|
| 7 |
+
"format": "coco_detection",
|
| 8 |
+
"image_mean": [
|
| 9 |
+
0.485,
|
| 10 |
+
0.456,
|
| 11 |
+
0.406
|
| 12 |
+
],
|
| 13 |
+
"image_processor_type": "DetrImageProcessor",
|
| 14 |
+
"image_std": [
|
| 15 |
+
0.229,
|
| 16 |
+
0.224,
|
| 17 |
+
0.225
|
| 18 |
+
],
|
| 19 |
+
"pad_size": null,
|
| 20 |
+
"resample": 2,
|
| 21 |
+
"rescale_factor": 0.00392156862745098,
|
| 22 |
+
"size": {
|
| 23 |
+
"longest_edge": 1333,
|
| 24 |
+
"shortest_edge": 800
|
| 25 |
+
}
|
| 26 |
+
}
|
checkpoint-4800/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:33689201c2f95566b82ac71b4c0510ad8ae017a0d38270028e01a468548fa22d
|
| 3 |
+
size 14244
|
checkpoint-4800/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:367d0a07234182d670ad5bdcc4099dadf05a93f2ce27151814f9ddcb3ed191e6
|
| 3 |
+
size 1064
|
checkpoint-4800/trainer_state.json
ADDED
|
@@ -0,0 +1,705 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 59.25925925925926,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 4800,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.6172839506172839,
|
| 13 |
+
"grad_norm": 62.88630294799805,
|
| 14 |
+
"learning_rate": 9.916000000000001e-06,
|
| 15 |
+
"loss": 6.6856,
|
| 16 |
+
"step": 50
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 1.2345679012345678,
|
| 20 |
+
"grad_norm": 387.6339111328125,
|
| 21 |
+
"learning_rate": 9.818000000000002e-06,
|
| 22 |
+
"loss": 5.5869,
|
| 23 |
+
"step": 100
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 1.8518518518518519,
|
| 27 |
+
"grad_norm": 245.18975830078125,
|
| 28 |
+
"learning_rate": 9.718e-06,
|
| 29 |
+
"loss": 5.159,
|
| 30 |
+
"step": 150
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 2.4691358024691357,
|
| 34 |
+
"grad_norm": 330.3158264160156,
|
| 35 |
+
"learning_rate": 9.618e-06,
|
| 36 |
+
"loss": 4.0494,
|
| 37 |
+
"step": 200
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 3.0864197530864197,
|
| 41 |
+
"grad_norm": 138.14215087890625,
|
| 42 |
+
"learning_rate": 9.518000000000001e-06,
|
| 43 |
+
"loss": 3.6313,
|
| 44 |
+
"step": 250
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 3.7037037037037037,
|
| 48 |
+
"grad_norm": 166.69479370117188,
|
| 49 |
+
"learning_rate": 9.418e-06,
|
| 50 |
+
"loss": 3.3391,
|
| 51 |
+
"step": 300
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 4.320987654320987,
|
| 55 |
+
"grad_norm": 107.54407501220703,
|
| 56 |
+
"learning_rate": 9.318e-06,
|
| 57 |
+
"loss": 3.3295,
|
| 58 |
+
"step": 350
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 4.938271604938271,
|
| 62 |
+
"grad_norm": 293.8923034667969,
|
| 63 |
+
"learning_rate": 9.218e-06,
|
| 64 |
+
"loss": 3.0956,
|
| 65 |
+
"step": 400
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 5.555555555555555,
|
| 69 |
+
"grad_norm": 226.05360412597656,
|
| 70 |
+
"learning_rate": 9.118000000000001e-06,
|
| 71 |
+
"loss": 3.0271,
|
| 72 |
+
"step": 450
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 6.172839506172839,
|
| 76 |
+
"grad_norm": 352.8487854003906,
|
| 77 |
+
"learning_rate": 9.018e-06,
|
| 78 |
+
"loss": 3.0627,
|
| 79 |
+
"step": 500
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 6.790123456790123,
|
| 83 |
+
"grad_norm": 97.6641616821289,
|
| 84 |
+
"learning_rate": 8.918000000000002e-06,
|
| 85 |
+
"loss": 3.0312,
|
| 86 |
+
"step": 550
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 7.407407407407407,
|
| 90 |
+
"grad_norm": 90.92312622070312,
|
| 91 |
+
"learning_rate": 8.818000000000001e-06,
|
| 92 |
+
"loss": 3.0228,
|
| 93 |
+
"step": 600
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 8.024691358024691,
|
| 97 |
+
"grad_norm": 241.7625732421875,
|
| 98 |
+
"learning_rate": 8.718e-06,
|
| 99 |
+
"loss": 3.0873,
|
| 100 |
+
"step": 650
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 8.641975308641975,
|
| 104 |
+
"grad_norm": 162.14071655273438,
|
| 105 |
+
"learning_rate": 8.618000000000001e-06,
|
| 106 |
+
"loss": 2.9582,
|
| 107 |
+
"step": 700
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 9.25925925925926,
|
| 111 |
+
"grad_norm": 128.74766540527344,
|
| 112 |
+
"learning_rate": 8.518e-06,
|
| 113 |
+
"loss": 2.9787,
|
| 114 |
+
"step": 750
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 9.876543209876543,
|
| 118 |
+
"grad_norm": 164.0232391357422,
|
| 119 |
+
"learning_rate": 8.418000000000001e-06,
|
| 120 |
+
"loss": 2.969,
|
| 121 |
+
"step": 800
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 10.493827160493828,
|
| 125 |
+
"grad_norm": 110.58142852783203,
|
| 126 |
+
"learning_rate": 8.318e-06,
|
| 127 |
+
"loss": 2.884,
|
| 128 |
+
"step": 850
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 11.11111111111111,
|
| 132 |
+
"grad_norm": 111.70948791503906,
|
| 133 |
+
"learning_rate": 8.218e-06,
|
| 134 |
+
"loss": 2.8216,
|
| 135 |
+
"step": 900
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 11.728395061728396,
|
| 139 |
+
"grad_norm": 111.31181335449219,
|
| 140 |
+
"learning_rate": 8.118000000000001e-06,
|
| 141 |
+
"loss": 2.8621,
|
| 142 |
+
"step": 950
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 12.345679012345679,
|
| 146 |
+
"grad_norm": 84.1104507446289,
|
| 147 |
+
"learning_rate": 8.018e-06,
|
| 148 |
+
"loss": 2.7653,
|
| 149 |
+
"step": 1000
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 12.962962962962964,
|
| 153 |
+
"grad_norm": 85.67730712890625,
|
| 154 |
+
"learning_rate": 7.918e-06,
|
| 155 |
+
"loss": 2.8418,
|
| 156 |
+
"step": 1050
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 13.580246913580247,
|
| 160 |
+
"grad_norm": 567.8911743164062,
|
| 161 |
+
"learning_rate": 7.818e-06,
|
| 162 |
+
"loss": 2.789,
|
| 163 |
+
"step": 1100
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 14.197530864197532,
|
| 167 |
+
"grad_norm": 98.47197723388672,
|
| 168 |
+
"learning_rate": 7.718000000000001e-06,
|
| 169 |
+
"loss": 2.7996,
|
| 170 |
+
"step": 1150
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 14.814814814814815,
|
| 174 |
+
"grad_norm": 86.87311553955078,
|
| 175 |
+
"learning_rate": 7.618000000000001e-06,
|
| 176 |
+
"loss": 2.7087,
|
| 177 |
+
"step": 1200
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 15.432098765432098,
|
| 181 |
+
"grad_norm": 115.806640625,
|
| 182 |
+
"learning_rate": 7.518000000000001e-06,
|
| 183 |
+
"loss": 2.6898,
|
| 184 |
+
"step": 1250
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 16.049382716049383,
|
| 188 |
+
"grad_norm": 108.76448059082031,
|
| 189 |
+
"learning_rate": 7.418000000000001e-06,
|
| 190 |
+
"loss": 2.7533,
|
| 191 |
+
"step": 1300
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 16.666666666666668,
|
| 195 |
+
"grad_norm": 137.66285705566406,
|
| 196 |
+
"learning_rate": 7.318000000000001e-06,
|
| 197 |
+
"loss": 2.6817,
|
| 198 |
+
"step": 1350
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 17.28395061728395,
|
| 202 |
+
"grad_norm": 167.23983764648438,
|
| 203 |
+
"learning_rate": 7.218e-06,
|
| 204 |
+
"loss": 2.6612,
|
| 205 |
+
"step": 1400
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 17.901234567901234,
|
| 209 |
+
"grad_norm": 80.27250671386719,
|
| 210 |
+
"learning_rate": 7.118e-06,
|
| 211 |
+
"loss": 2.7053,
|
| 212 |
+
"step": 1450
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 18.51851851851852,
|
| 216 |
+
"grad_norm": 87.89417266845703,
|
| 217 |
+
"learning_rate": 7.018e-06,
|
| 218 |
+
"loss": 2.6016,
|
| 219 |
+
"step": 1500
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 19.135802469135804,
|
| 223 |
+
"grad_norm": 244.07081604003906,
|
| 224 |
+
"learning_rate": 6.9180000000000005e-06,
|
| 225 |
+
"loss": 2.6377,
|
| 226 |
+
"step": 1550
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 19.753086419753085,
|
| 230 |
+
"grad_norm": 66.56304931640625,
|
| 231 |
+
"learning_rate": 6.818e-06,
|
| 232 |
+
"loss": 2.5969,
|
| 233 |
+
"step": 1600
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 20.37037037037037,
|
| 237 |
+
"grad_norm": 146.2065887451172,
|
| 238 |
+
"learning_rate": 6.718e-06,
|
| 239 |
+
"loss": 2.5849,
|
| 240 |
+
"step": 1650
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 20.987654320987655,
|
| 244 |
+
"grad_norm": 105.38662719726562,
|
| 245 |
+
"learning_rate": 6.618000000000001e-06,
|
| 246 |
+
"loss": 2.588,
|
| 247 |
+
"step": 1700
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 21.604938271604937,
|
| 251 |
+
"grad_norm": 172.37232971191406,
|
| 252 |
+
"learning_rate": 6.518000000000001e-06,
|
| 253 |
+
"loss": 2.5201,
|
| 254 |
+
"step": 1750
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 22.22222222222222,
|
| 258 |
+
"grad_norm": 1186.192626953125,
|
| 259 |
+
"learning_rate": 6.418000000000001e-06,
|
| 260 |
+
"loss": 2.5618,
|
| 261 |
+
"step": 1800
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 22.839506172839506,
|
| 265 |
+
"grad_norm": 181.00384521484375,
|
| 266 |
+
"learning_rate": 6.318000000000001e-06,
|
| 267 |
+
"loss": 2.496,
|
| 268 |
+
"step": 1850
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 23.45679012345679,
|
| 272 |
+
"grad_norm": 90.32185363769531,
|
| 273 |
+
"learning_rate": 6.2180000000000004e-06,
|
| 274 |
+
"loss": 2.5254,
|
| 275 |
+
"step": 1900
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 24.074074074074073,
|
| 279 |
+
"grad_norm": 117.39266967773438,
|
| 280 |
+
"learning_rate": 6.1180000000000005e-06,
|
| 281 |
+
"loss": 2.4665,
|
| 282 |
+
"step": 1950
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 24.691358024691358,
|
| 286 |
+
"grad_norm": 117.4901123046875,
|
| 287 |
+
"learning_rate": 6.018000000000001e-06,
|
| 288 |
+
"loss": 2.4541,
|
| 289 |
+
"step": 2000
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 25.308641975308642,
|
| 293 |
+
"grad_norm": 112.75492095947266,
|
| 294 |
+
"learning_rate": 5.918000000000001e-06,
|
| 295 |
+
"loss": 2.4716,
|
| 296 |
+
"step": 2050
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 25.925925925925927,
|
| 300 |
+
"grad_norm": 75.71508026123047,
|
| 301 |
+
"learning_rate": 5.818e-06,
|
| 302 |
+
"loss": 2.4643,
|
| 303 |
+
"step": 2100
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 26.54320987654321,
|
| 307 |
+
"grad_norm": 87.87818145751953,
|
| 308 |
+
"learning_rate": 5.718e-06,
|
| 309 |
+
"loss": 2.452,
|
| 310 |
+
"step": 2150
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 27.160493827160494,
|
| 314 |
+
"grad_norm": 119.02462005615234,
|
| 315 |
+
"learning_rate": 5.618e-06,
|
| 316 |
+
"loss": 2.4431,
|
| 317 |
+
"step": 2200
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 27.77777777777778,
|
| 321 |
+
"grad_norm": 95.6996841430664,
|
| 322 |
+
"learning_rate": 5.518e-06,
|
| 323 |
+
"loss": 2.396,
|
| 324 |
+
"step": 2250
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 28.395061728395063,
|
| 328 |
+
"grad_norm": 82.37759399414062,
|
| 329 |
+
"learning_rate": 5.420000000000001e-06,
|
| 330 |
+
"loss": 2.4554,
|
| 331 |
+
"step": 2300
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 29.012345679012345,
|
| 335 |
+
"grad_norm": 63.04928207397461,
|
| 336 |
+
"learning_rate": 5.320000000000001e-06,
|
| 337 |
+
"loss": 2.3868,
|
| 338 |
+
"step": 2350
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 29.62962962962963,
|
| 342 |
+
"grad_norm": 85.531494140625,
|
| 343 |
+
"learning_rate": 5.220000000000001e-06,
|
| 344 |
+
"loss": 2.4088,
|
| 345 |
+
"step": 2400
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 30.246913580246915,
|
| 349 |
+
"grad_norm": 148.06163024902344,
|
| 350 |
+
"learning_rate": 5.12e-06,
|
| 351 |
+
"loss": 2.4479,
|
| 352 |
+
"step": 2450
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 30.864197530864196,
|
| 356 |
+
"grad_norm": 120.42308807373047,
|
| 357 |
+
"learning_rate": 5.02e-06,
|
| 358 |
+
"loss": 2.4293,
|
| 359 |
+
"step": 2500
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 31.48148148148148,
|
| 363 |
+
"grad_norm": 47.70389938354492,
|
| 364 |
+
"learning_rate": 4.92e-06,
|
| 365 |
+
"loss": 2.3727,
|
| 366 |
+
"step": 2550
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 32.098765432098766,
|
| 370 |
+
"grad_norm": 67.6549301147461,
|
| 371 |
+
"learning_rate": 4.8200000000000004e-06,
|
| 372 |
+
"loss": 2.3938,
|
| 373 |
+
"step": 2600
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 32.71604938271605,
|
| 377 |
+
"grad_norm": 71.10252380371094,
|
| 378 |
+
"learning_rate": 4.7200000000000005e-06,
|
| 379 |
+
"loss": 2.3758,
|
| 380 |
+
"step": 2650
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 33.333333333333336,
|
| 384 |
+
"grad_norm": 69.4305419921875,
|
| 385 |
+
"learning_rate": 4.620000000000001e-06,
|
| 386 |
+
"loss": 2.3855,
|
| 387 |
+
"step": 2700
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 33.95061728395062,
|
| 391 |
+
"grad_norm": 58.54874038696289,
|
| 392 |
+
"learning_rate": 4.520000000000001e-06,
|
| 393 |
+
"loss": 2.3638,
|
| 394 |
+
"step": 2750
|
| 395 |
+
},
|
| 396 |
+
{
|
| 397 |
+
"epoch": 34.5679012345679,
|
| 398 |
+
"grad_norm": 135.54541015625,
|
| 399 |
+
"learning_rate": 4.42e-06,
|
| 400 |
+
"loss": 2.3785,
|
| 401 |
+
"step": 2800
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 35.18518518518518,
|
| 405 |
+
"grad_norm": 191.87705993652344,
|
| 406 |
+
"learning_rate": 4.32e-06,
|
| 407 |
+
"loss": 2.3247,
|
| 408 |
+
"step": 2850
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"epoch": 35.80246913580247,
|
| 412 |
+
"grad_norm": 101.19218444824219,
|
| 413 |
+
"learning_rate": 4.22e-06,
|
| 414 |
+
"loss": 2.3322,
|
| 415 |
+
"step": 2900
|
| 416 |
+
},
|
| 417 |
+
{
|
| 418 |
+
"epoch": 36.41975308641975,
|
| 419 |
+
"grad_norm": 97.59493255615234,
|
| 420 |
+
"learning_rate": 4.12e-06,
|
| 421 |
+
"loss": 2.3189,
|
| 422 |
+
"step": 2950
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 37.03703703703704,
|
| 426 |
+
"grad_norm": 50.98225784301758,
|
| 427 |
+
"learning_rate": 4.0200000000000005e-06,
|
| 428 |
+
"loss": 2.3337,
|
| 429 |
+
"step": 3000
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 37.65432098765432,
|
| 433 |
+
"grad_norm": 78.88179016113281,
|
| 434 |
+
"learning_rate": 3.920000000000001e-06,
|
| 435 |
+
"loss": 2.3707,
|
| 436 |
+
"step": 3050
|
| 437 |
+
},
|
| 438 |
+
{
|
| 439 |
+
"epoch": 38.27160493827161,
|
| 440 |
+
"grad_norm": 136.63319396972656,
|
| 441 |
+
"learning_rate": 3.820000000000001e-06,
|
| 442 |
+
"loss": 2.2602,
|
| 443 |
+
"step": 3100
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 38.888888888888886,
|
| 447 |
+
"grad_norm": 87.68150329589844,
|
| 448 |
+
"learning_rate": 3.7200000000000004e-06,
|
| 449 |
+
"loss": 2.3111,
|
| 450 |
+
"step": 3150
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"epoch": 39.50617283950617,
|
| 454 |
+
"grad_norm": 179.4459686279297,
|
| 455 |
+
"learning_rate": 3.62e-06,
|
| 456 |
+
"loss": 2.3092,
|
| 457 |
+
"step": 3200
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 40.123456790123456,
|
| 461 |
+
"grad_norm": 76.10836791992188,
|
| 462 |
+
"learning_rate": 3.52e-06,
|
| 463 |
+
"loss": 2.3029,
|
| 464 |
+
"step": 3250
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 40.74074074074074,
|
| 468 |
+
"grad_norm": 85.23873901367188,
|
| 469 |
+
"learning_rate": 3.4200000000000007e-06,
|
| 470 |
+
"loss": 2.2699,
|
| 471 |
+
"step": 3300
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 41.358024691358025,
|
| 475 |
+
"grad_norm": 156.77699279785156,
|
| 476 |
+
"learning_rate": 3.3200000000000004e-06,
|
| 477 |
+
"loss": 2.2795,
|
| 478 |
+
"step": 3350
|
| 479 |
+
},
|
| 480 |
+
{
|
| 481 |
+
"epoch": 41.97530864197531,
|
| 482 |
+
"grad_norm": 87.01202392578125,
|
| 483 |
+
"learning_rate": 3.2200000000000005e-06,
|
| 484 |
+
"loss": 2.2961,
|
| 485 |
+
"step": 3400
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 42.592592592592595,
|
| 489 |
+
"grad_norm": 208.42947387695312,
|
| 490 |
+
"learning_rate": 3.12e-06,
|
| 491 |
+
"loss": 2.2692,
|
| 492 |
+
"step": 3450
|
| 493 |
+
},
|
| 494 |
+
{
|
| 495 |
+
"epoch": 43.20987654320987,
|
| 496 |
+
"grad_norm": 82.78223419189453,
|
| 497 |
+
"learning_rate": 3.0200000000000003e-06,
|
| 498 |
+
"loss": 2.2482,
|
| 499 |
+
"step": 3500
|
| 500 |
+
},
|
| 501 |
+
{
|
| 502 |
+
"epoch": 43.82716049382716,
|
| 503 |
+
"grad_norm": 110.6844482421875,
|
| 504 |
+
"learning_rate": 2.92e-06,
|
| 505 |
+
"loss": 2.2755,
|
| 506 |
+
"step": 3550
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 44.44444444444444,
|
| 510 |
+
"grad_norm": 57.944942474365234,
|
| 511 |
+
"learning_rate": 2.82e-06,
|
| 512 |
+
"loss": 2.302,
|
| 513 |
+
"step": 3600
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 45.06172839506173,
|
| 517 |
+
"grad_norm": 72.0767593383789,
|
| 518 |
+
"learning_rate": 2.7200000000000002e-06,
|
| 519 |
+
"loss": 2.2305,
|
| 520 |
+
"step": 3650
|
| 521 |
+
},
|
| 522 |
+
{
|
| 523 |
+
"epoch": 45.67901234567901,
|
| 524 |
+
"grad_norm": 154.40426635742188,
|
| 525 |
+
"learning_rate": 2.6200000000000003e-06,
|
| 526 |
+
"loss": 2.2532,
|
| 527 |
+
"step": 3700
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 46.2962962962963,
|
| 531 |
+
"grad_norm": 112.4802017211914,
|
| 532 |
+
"learning_rate": 2.52e-06,
|
| 533 |
+
"loss": 2.2596,
|
| 534 |
+
"step": 3750
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 46.91358024691358,
|
| 538 |
+
"grad_norm": 97.9261245727539,
|
| 539 |
+
"learning_rate": 2.42e-06,
|
| 540 |
+
"loss": 2.2296,
|
| 541 |
+
"step": 3800
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 47.53086419753087,
|
| 545 |
+
"grad_norm": 86.22404479980469,
|
| 546 |
+
"learning_rate": 2.3200000000000002e-06,
|
| 547 |
+
"loss": 2.2086,
|
| 548 |
+
"step": 3850
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 48.148148148148145,
|
| 552 |
+
"grad_norm": 118.13029479980469,
|
| 553 |
+
"learning_rate": 2.2200000000000003e-06,
|
| 554 |
+
"loss": 2.2208,
|
| 555 |
+
"step": 3900
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 48.76543209876543,
|
| 559 |
+
"grad_norm": 83.22029113769531,
|
| 560 |
+
"learning_rate": 2.12e-06,
|
| 561 |
+
"loss": 2.2228,
|
| 562 |
+
"step": 3950
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"epoch": 49.382716049382715,
|
| 566 |
+
"grad_norm": 46.90211868286133,
|
| 567 |
+
"learning_rate": 2.02e-06,
|
| 568 |
+
"loss": 2.2279,
|
| 569 |
+
"step": 4000
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 50.0,
|
| 573 |
+
"grad_norm": 666.5877685546875,
|
| 574 |
+
"learning_rate": 1.9200000000000003e-06,
|
| 575 |
+
"loss": 2.2356,
|
| 576 |
+
"step": 4050
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 50.617283950617285,
|
| 580 |
+
"grad_norm": 73.89546966552734,
|
| 581 |
+
"learning_rate": 1.8200000000000002e-06,
|
| 582 |
+
"loss": 2.2332,
|
| 583 |
+
"step": 4100
|
| 584 |
+
},
|
| 585 |
+
{
|
| 586 |
+
"epoch": 51.23456790123457,
|
| 587 |
+
"grad_norm": 205.4402313232422,
|
| 588 |
+
"learning_rate": 1.72e-06,
|
| 589 |
+
"loss": 2.1635,
|
| 590 |
+
"step": 4150
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 51.851851851851855,
|
| 594 |
+
"grad_norm": 46.229618072509766,
|
| 595 |
+
"learning_rate": 1.6200000000000002e-06,
|
| 596 |
+
"loss": 2.1759,
|
| 597 |
+
"step": 4200
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 52.46913580246913,
|
| 601 |
+
"grad_norm": 58.453006744384766,
|
| 602 |
+
"learning_rate": 1.52e-06,
|
| 603 |
+
"loss": 2.1893,
|
| 604 |
+
"step": 4250
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"epoch": 53.08641975308642,
|
| 608 |
+
"grad_norm": 102.3731689453125,
|
| 609 |
+
"learning_rate": 1.4220000000000001e-06,
|
| 610 |
+
"loss": 2.2028,
|
| 611 |
+
"step": 4300
|
| 612 |
+
},
|
| 613 |
+
{
|
| 614 |
+
"epoch": 53.7037037037037,
|
| 615 |
+
"grad_norm": 82.2330093383789,
|
| 616 |
+
"learning_rate": 1.3220000000000002e-06,
|
| 617 |
+
"loss": 2.2168,
|
| 618 |
+
"step": 4350
|
| 619 |
+
},
|
| 620 |
+
{
|
| 621 |
+
"epoch": 54.32098765432099,
|
| 622 |
+
"grad_norm": 197.8307342529297,
|
| 623 |
+
"learning_rate": 1.2220000000000001e-06,
|
| 624 |
+
"loss": 2.2205,
|
| 625 |
+
"step": 4400
|
| 626 |
+
},
|
| 627 |
+
{
|
| 628 |
+
"epoch": 54.93827160493827,
|
| 629 |
+
"grad_norm": 108.3754653930664,
|
| 630 |
+
"learning_rate": 1.122e-06,
|
| 631 |
+
"loss": 2.1846,
|
| 632 |
+
"step": 4450
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 55.55555555555556,
|
| 636 |
+
"grad_norm": 87.77450561523438,
|
| 637 |
+
"learning_rate": 1.0220000000000001e-06,
|
| 638 |
+
"loss": 2.1763,
|
| 639 |
+
"step": 4500
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 56.17283950617284,
|
| 643 |
+
"grad_norm": 93.21278381347656,
|
| 644 |
+
"learning_rate": 9.220000000000001e-07,
|
| 645 |
+
"loss": 2.2124,
|
| 646 |
+
"step": 4550
|
| 647 |
+
},
|
| 648 |
+
{
|
| 649 |
+
"epoch": 56.79012345679013,
|
| 650 |
+
"grad_norm": 65.71028137207031,
|
| 651 |
+
"learning_rate": 8.22e-07,
|
| 652 |
+
"loss": 2.1856,
|
| 653 |
+
"step": 4600
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 57.407407407407405,
|
| 657 |
+
"grad_norm": 84.39445495605469,
|
| 658 |
+
"learning_rate": 7.22e-07,
|
| 659 |
+
"loss": 2.1913,
|
| 660 |
+
"step": 4650
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"epoch": 58.02469135802469,
|
| 664 |
+
"grad_norm": 124.40509033203125,
|
| 665 |
+
"learning_rate": 6.22e-07,
|
| 666 |
+
"loss": 2.177,
|
| 667 |
+
"step": 4700
|
| 668 |
+
},
|
| 669 |
+
{
|
| 670 |
+
"epoch": 58.641975308641975,
|
| 671 |
+
"grad_norm": 64.60520935058594,
|
| 672 |
+
"learning_rate": 5.22e-07,
|
| 673 |
+
"loss": 2.1318,
|
| 674 |
+
"step": 4750
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 59.25925925925926,
|
| 678 |
+
"grad_norm": 94.89431762695312,
|
| 679 |
+
"learning_rate": 4.2200000000000005e-07,
|
| 680 |
+
"loss": 2.1938,
|
| 681 |
+
"step": 4800
|
| 682 |
+
}
|
| 683 |
+
],
|
| 684 |
+
"logging_steps": 50,
|
| 685 |
+
"max_steps": 5000,
|
| 686 |
+
"num_input_tokens_seen": 0,
|
| 687 |
+
"num_train_epochs": 62,
|
| 688 |
+
"save_steps": 100,
|
| 689 |
+
"stateful_callbacks": {
|
| 690 |
+
"TrainerControl": {
|
| 691 |
+
"args": {
|
| 692 |
+
"should_epoch_stop": false,
|
| 693 |
+
"should_evaluate": false,
|
| 694 |
+
"should_log": false,
|
| 695 |
+
"should_save": true,
|
| 696 |
+
"should_training_stop": false
|
| 697 |
+
},
|
| 698 |
+
"attributes": {}
|
| 699 |
+
}
|
| 700 |
+
},
|
| 701 |
+
"total_flos": 1.8347671977984e+19,
|
| 702 |
+
"train_batch_size": 8,
|
| 703 |
+
"trial_name": null,
|
| 704 |
+
"trial_params": null
|
| 705 |
+
}
|
checkpoint-4800/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:4dae0b4d732b15d565b589748a29a09a17e2ddd7d7ab9df017470bce43888a34
|
| 3 |
+
size 5240
|