Upload 8 files
Browse files- config.json +69 -0
- model.safetensors +3 -0
- optimizer.pt +3 -0
- preprocessor_config.json +27 -0
- rng_state.pth +3 -0
- scheduler.pt +3 -0
- trainer_state.json +2218 -0
- training_args.bin +3 -0
config.json
ADDED
|
@@ -0,0 +1,69 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "facebook/dinov2-large",
|
| 3 |
+
"apply_layernorm": true,
|
| 4 |
+
"architectures": [
|
| 5 |
+
"Dinov2ForImageClassification"
|
| 6 |
+
],
|
| 7 |
+
"attention_probs_dropout_prob": 0.0,
|
| 8 |
+
"drop_path_rate": 0.0,
|
| 9 |
+
"hidden_act": "gelu",
|
| 10 |
+
"hidden_dropout_prob": 0.0,
|
| 11 |
+
"hidden_size": 1024,
|
| 12 |
+
"id2label": {
|
| 13 |
+
"0": "correct",
|
| 14 |
+
"1": "incorrect"
|
| 15 |
+
},
|
| 16 |
+
"image_size": 518,
|
| 17 |
+
"initializer_range": 0.02,
|
| 18 |
+
"label2id": {
|
| 19 |
+
"correct": 0,
|
| 20 |
+
"incorrect": 1
|
| 21 |
+
},
|
| 22 |
+
"layer_norm_eps": 1e-06,
|
| 23 |
+
"layerscale_value": 1.0,
|
| 24 |
+
"mlp_ratio": 4,
|
| 25 |
+
"model_type": "dinov2",
|
| 26 |
+
"num_attention_heads": 16,
|
| 27 |
+
"num_channels": 3,
|
| 28 |
+
"num_hidden_layers": 24,
|
| 29 |
+
"out_features": [
|
| 30 |
+
"stage24"
|
| 31 |
+
],
|
| 32 |
+
"out_indices": [
|
| 33 |
+
24
|
| 34 |
+
],
|
| 35 |
+
"patch_size": 14,
|
| 36 |
+
"problem_type": "single_label_classification",
|
| 37 |
+
"qkv_bias": true,
|
| 38 |
+
"reshape_hidden_states": true,
|
| 39 |
+
"stage_names": [
|
| 40 |
+
"stem",
|
| 41 |
+
"stage1",
|
| 42 |
+
"stage2",
|
| 43 |
+
"stage3",
|
| 44 |
+
"stage4",
|
| 45 |
+
"stage5",
|
| 46 |
+
"stage6",
|
| 47 |
+
"stage7",
|
| 48 |
+
"stage8",
|
| 49 |
+
"stage9",
|
| 50 |
+
"stage10",
|
| 51 |
+
"stage11",
|
| 52 |
+
"stage12",
|
| 53 |
+
"stage13",
|
| 54 |
+
"stage14",
|
| 55 |
+
"stage15",
|
| 56 |
+
"stage16",
|
| 57 |
+
"stage17",
|
| 58 |
+
"stage18",
|
| 59 |
+
"stage19",
|
| 60 |
+
"stage20",
|
| 61 |
+
"stage21",
|
| 62 |
+
"stage22",
|
| 63 |
+
"stage23",
|
| 64 |
+
"stage24"
|
| 65 |
+
],
|
| 66 |
+
"torch_dtype": "float32",
|
| 67 |
+
"transformers_version": "4.47.1",
|
| 68 |
+
"use_swiglu_ffn": false
|
| 69 |
+
}
|
model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:a6959497928ab28564b34d956a334baee012c6a52af5ed46fd6e33dd6999246a
|
| 3 |
+
size 1217542512
|
optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:068f0ef49469943ce4e61797f2c8dd1435b87edb8111a3f8f05d754152d12ca0
|
| 3 |
+
size 2435341946
|
preprocessor_config.json
ADDED
|
@@ -0,0 +1,27 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"crop_size": {
|
| 3 |
+
"height": 224,
|
| 4 |
+
"width": 224
|
| 5 |
+
},
|
| 6 |
+
"do_center_crop": true,
|
| 7 |
+
"do_convert_rgb": true,
|
| 8 |
+
"do_normalize": true,
|
| 9 |
+
"do_rescale": true,
|
| 10 |
+
"do_resize": true,
|
| 11 |
+
"image_mean": [
|
| 12 |
+
0.485,
|
| 13 |
+
0.456,
|
| 14 |
+
0.406
|
| 15 |
+
],
|
| 16 |
+
"image_processor_type": "BitImageProcessor",
|
| 17 |
+
"image_std": [
|
| 18 |
+
0.229,
|
| 19 |
+
0.224,
|
| 20 |
+
0.225
|
| 21 |
+
],
|
| 22 |
+
"resample": 3,
|
| 23 |
+
"rescale_factor": 0.00392156862745098,
|
| 24 |
+
"size": {
|
| 25 |
+
"shortest_edge": 256
|
| 26 |
+
}
|
| 27 |
+
}
|
rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:7e89cbde481fa889fb095dccd0e27796b9df4d9d6e2b39417c22fb2c0ec6285b
|
| 3 |
+
size 14244
|
scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:e26293a4ac99ad99306eb23f6a8d8c5e13b975b27036f16e65252fbc3ffff2ff
|
| 3 |
+
size 1064
|
trainer_state.json
ADDED
|
@@ -0,0 +1,2218 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 0.8017572519758047,
|
| 3 |
+
"best_model_checkpoint": "Obstacle/dinov2/checkpoint-2800",
|
| 4 |
+
"epoch": 24.783964365256125,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 2800,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.08908685968819599,
|
| 13 |
+
"grad_norm": 105.50035095214844,
|
| 14 |
+
"learning_rate": 3.5714285714285716e-07,
|
| 15 |
+
"loss": 3.6297,
|
| 16 |
+
"step": 10
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.17817371937639198,
|
| 20 |
+
"grad_norm": 120.46121978759766,
|
| 21 |
+
"learning_rate": 7.142857142857143e-07,
|
| 22 |
+
"loss": 3.5465,
|
| 23 |
+
"step": 20
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.267260579064588,
|
| 27 |
+
"grad_norm": 201.12693786621094,
|
| 28 |
+
"learning_rate": 1.0714285714285714e-06,
|
| 29 |
+
"loss": 3.1817,
|
| 30 |
+
"step": 30
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.35634743875278396,
|
| 34 |
+
"grad_norm": 125.14132690429688,
|
| 35 |
+
"learning_rate": 1.4285714285714286e-06,
|
| 36 |
+
"loss": 3.1466,
|
| 37 |
+
"step": 40
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.44543429844098,
|
| 41 |
+
"grad_norm": 125.6826400756836,
|
| 42 |
+
"learning_rate": 1.7857142857142859e-06,
|
| 43 |
+
"loss": 2.6557,
|
| 44 |
+
"step": 50
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.534521158129176,
|
| 48 |
+
"grad_norm": 85.71980285644531,
|
| 49 |
+
"learning_rate": 2.1428571428571427e-06,
|
| 50 |
+
"loss": 2.5632,
|
| 51 |
+
"step": 60
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.623608017817372,
|
| 55 |
+
"grad_norm": 109.28083801269531,
|
| 56 |
+
"learning_rate": 2.5e-06,
|
| 57 |
+
"loss": 2.8634,
|
| 58 |
+
"step": 70
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.7126948775055679,
|
| 62 |
+
"grad_norm": 151.45330810546875,
|
| 63 |
+
"learning_rate": 2.8571428571428573e-06,
|
| 64 |
+
"loss": 2.6854,
|
| 65 |
+
"step": 80
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.8017817371937639,
|
| 69 |
+
"grad_norm": 85.33807373046875,
|
| 70 |
+
"learning_rate": 3.2142857142857147e-06,
|
| 71 |
+
"loss": 2.6114,
|
| 72 |
+
"step": 90
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.89086859688196,
|
| 76 |
+
"grad_norm": 186.34913635253906,
|
| 77 |
+
"learning_rate": 3.5714285714285718e-06,
|
| 78 |
+
"loss": 2.5097,
|
| 79 |
+
"step": 100
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.9799554565701559,
|
| 83 |
+
"grad_norm": 64.78919219970703,
|
| 84 |
+
"learning_rate": 3.928571428571429e-06,
|
| 85 |
+
"loss": 2.5137,
|
| 86 |
+
"step": 110
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 1.0,
|
| 90 |
+
"eval_loss": 0.5945847630500793,
|
| 91 |
+
"eval_macro_f1": 0.6702382389633386,
|
| 92 |
+
"eval_runtime": 38.7104,
|
| 93 |
+
"eval_samples_per_second": 23.198,
|
| 94 |
+
"eval_steps_per_second": 2.919,
|
| 95 |
+
"step": 113
|
| 96 |
+
},
|
| 97 |
+
{
|
| 98 |
+
"epoch": 1.0623608017817372,
|
| 99 |
+
"grad_norm": 70.03704071044922,
|
| 100 |
+
"learning_rate": 4.2857142857142855e-06,
|
| 101 |
+
"loss": 2.0547,
|
| 102 |
+
"step": 120
|
| 103 |
+
},
|
| 104 |
+
{
|
| 105 |
+
"epoch": 1.1514476614699332,
|
| 106 |
+
"grad_norm": 60.801490783691406,
|
| 107 |
+
"learning_rate": 4.642857142857144e-06,
|
| 108 |
+
"loss": 2.2826,
|
| 109 |
+
"step": 130
|
| 110 |
+
},
|
| 111 |
+
{
|
| 112 |
+
"epoch": 1.2405345211581291,
|
| 113 |
+
"grad_norm": 316.7510070800781,
|
| 114 |
+
"learning_rate": 5e-06,
|
| 115 |
+
"loss": 2.4794,
|
| 116 |
+
"step": 140
|
| 117 |
+
},
|
| 118 |
+
{
|
| 119 |
+
"epoch": 1.329621380846325,
|
| 120 |
+
"grad_norm": 58.09334945678711,
|
| 121 |
+
"learning_rate": 5.357142857142857e-06,
|
| 122 |
+
"loss": 2.1821,
|
| 123 |
+
"step": 150
|
| 124 |
+
},
|
| 125 |
+
{
|
| 126 |
+
"epoch": 1.4187082405345213,
|
| 127 |
+
"grad_norm": 189.68980407714844,
|
| 128 |
+
"learning_rate": 5.7142857142857145e-06,
|
| 129 |
+
"loss": 2.4679,
|
| 130 |
+
"step": 160
|
| 131 |
+
},
|
| 132 |
+
{
|
| 133 |
+
"epoch": 1.507795100222717,
|
| 134 |
+
"grad_norm": 3325.044189453125,
|
| 135 |
+
"learning_rate": 6.071428571428571e-06,
|
| 136 |
+
"loss": 2.1725,
|
| 137 |
+
"step": 170
|
| 138 |
+
},
|
| 139 |
+
{
|
| 140 |
+
"epoch": 1.5968819599109132,
|
| 141 |
+
"grad_norm": 189.6019287109375,
|
| 142 |
+
"learning_rate": 6.4285714285714295e-06,
|
| 143 |
+
"loss": 2.2086,
|
| 144 |
+
"step": 180
|
| 145 |
+
},
|
| 146 |
+
{
|
| 147 |
+
"epoch": 1.6859688195991092,
|
| 148 |
+
"grad_norm": 46.3530387878418,
|
| 149 |
+
"learning_rate": 6.785714285714287e-06,
|
| 150 |
+
"loss": 2.27,
|
| 151 |
+
"step": 190
|
| 152 |
+
},
|
| 153 |
+
{
|
| 154 |
+
"epoch": 1.7750556792873051,
|
| 155 |
+
"grad_norm": 311.6956787109375,
|
| 156 |
+
"learning_rate": 7.1428571428571436e-06,
|
| 157 |
+
"loss": 2.2687,
|
| 158 |
+
"step": 200
|
| 159 |
+
},
|
| 160 |
+
{
|
| 161 |
+
"epoch": 1.864142538975501,
|
| 162 |
+
"grad_norm": 59.799678802490234,
|
| 163 |
+
"learning_rate": 7.500000000000001e-06,
|
| 164 |
+
"loss": 2.219,
|
| 165 |
+
"step": 210
|
| 166 |
+
},
|
| 167 |
+
{
|
| 168 |
+
"epoch": 1.953229398663697,
|
| 169 |
+
"grad_norm": 110.08448791503906,
|
| 170 |
+
"learning_rate": 7.857142857142858e-06,
|
| 171 |
+
"loss": 2.6144,
|
| 172 |
+
"step": 220
|
| 173 |
+
},
|
| 174 |
+
{
|
| 175 |
+
"epoch": 2.0,
|
| 176 |
+
"eval_loss": 0.5521021485328674,
|
| 177 |
+
"eval_macro_f1": 0.7138090898678897,
|
| 178 |
+
"eval_runtime": 25.7637,
|
| 179 |
+
"eval_samples_per_second": 34.855,
|
| 180 |
+
"eval_steps_per_second": 4.386,
|
| 181 |
+
"step": 226
|
| 182 |
+
},
|
| 183 |
+
{
|
| 184 |
+
"epoch": 2.035634743875278,
|
| 185 |
+
"grad_norm": 102.7503433227539,
|
| 186 |
+
"learning_rate": 8.214285714285714e-06,
|
| 187 |
+
"loss": 2.4624,
|
| 188 |
+
"step": 230
|
| 189 |
+
},
|
| 190 |
+
{
|
| 191 |
+
"epoch": 2.1247216035634744,
|
| 192 |
+
"grad_norm": 145.72938537597656,
|
| 193 |
+
"learning_rate": 8.571428571428571e-06,
|
| 194 |
+
"loss": 2.0923,
|
| 195 |
+
"step": 240
|
| 196 |
+
},
|
| 197 |
+
{
|
| 198 |
+
"epoch": 2.2138084632516706,
|
| 199 |
+
"grad_norm": 55.424922943115234,
|
| 200 |
+
"learning_rate": 8.92857142857143e-06,
|
| 201 |
+
"loss": 1.8513,
|
| 202 |
+
"step": 250
|
| 203 |
+
},
|
| 204 |
+
{
|
| 205 |
+
"epoch": 2.3028953229398663,
|
| 206 |
+
"grad_norm": 60.43095779418945,
|
| 207 |
+
"learning_rate": 9.285714285714288e-06,
|
| 208 |
+
"loss": 2.278,
|
| 209 |
+
"step": 260
|
| 210 |
+
},
|
| 211 |
+
{
|
| 212 |
+
"epoch": 2.3919821826280625,
|
| 213 |
+
"grad_norm": 51.914947509765625,
|
| 214 |
+
"learning_rate": 9.642857142857144e-06,
|
| 215 |
+
"loss": 2.2194,
|
| 216 |
+
"step": 270
|
| 217 |
+
},
|
| 218 |
+
{
|
| 219 |
+
"epoch": 2.4810690423162582,
|
| 220 |
+
"grad_norm": 72.88397979736328,
|
| 221 |
+
"learning_rate": 1e-05,
|
| 222 |
+
"loss": 2.1676,
|
| 223 |
+
"step": 280
|
| 224 |
+
},
|
| 225 |
+
{
|
| 226 |
+
"epoch": 2.5701559020044544,
|
| 227 |
+
"grad_norm": 70.16185760498047,
|
| 228 |
+
"learning_rate": 9.960317460317462e-06,
|
| 229 |
+
"loss": 2.0797,
|
| 230 |
+
"step": 290
|
| 231 |
+
},
|
| 232 |
+
{
|
| 233 |
+
"epoch": 2.65924276169265,
|
| 234 |
+
"grad_norm": 42.27340316772461,
|
| 235 |
+
"learning_rate": 9.920634920634922e-06,
|
| 236 |
+
"loss": 2.0556,
|
| 237 |
+
"step": 300
|
| 238 |
+
},
|
| 239 |
+
{
|
| 240 |
+
"epoch": 2.7483296213808464,
|
| 241 |
+
"grad_norm": 147.33399963378906,
|
| 242 |
+
"learning_rate": 9.880952380952381e-06,
|
| 243 |
+
"loss": 1.7408,
|
| 244 |
+
"step": 310
|
| 245 |
+
},
|
| 246 |
+
{
|
| 247 |
+
"epoch": 2.8374164810690425,
|
| 248 |
+
"grad_norm": 101.6998519897461,
|
| 249 |
+
"learning_rate": 9.841269841269842e-06,
|
| 250 |
+
"loss": 2.1287,
|
| 251 |
+
"step": 320
|
| 252 |
+
},
|
| 253 |
+
{
|
| 254 |
+
"epoch": 2.9265033407572383,
|
| 255 |
+
"grad_norm": 62.05525207519531,
|
| 256 |
+
"learning_rate": 9.801587301587301e-06,
|
| 257 |
+
"loss": 2.2084,
|
| 258 |
+
"step": 330
|
| 259 |
+
},
|
| 260 |
+
{
|
| 261 |
+
"epoch": 3.0,
|
| 262 |
+
"eval_loss": 0.9312511682510376,
|
| 263 |
+
"eval_macro_f1": 0.5517047201798773,
|
| 264 |
+
"eval_runtime": 30.6876,
|
| 265 |
+
"eval_samples_per_second": 29.263,
|
| 266 |
+
"eval_steps_per_second": 3.682,
|
| 267 |
+
"step": 339
|
| 268 |
+
},
|
| 269 |
+
{
|
| 270 |
+
"epoch": 3.0089086859688194,
|
| 271 |
+
"grad_norm": 223.84640502929688,
|
| 272 |
+
"learning_rate": 9.761904761904762e-06,
|
| 273 |
+
"loss": 1.9799,
|
| 274 |
+
"step": 340
|
| 275 |
+
},
|
| 276 |
+
{
|
| 277 |
+
"epoch": 3.0979955456570156,
|
| 278 |
+
"grad_norm": 96.64312744140625,
|
| 279 |
+
"learning_rate": 9.722222222222223e-06,
|
| 280 |
+
"loss": 2.769,
|
| 281 |
+
"step": 350
|
| 282 |
+
},
|
| 283 |
+
{
|
| 284 |
+
"epoch": 3.187082405345212,
|
| 285 |
+
"grad_norm": 45.61930465698242,
|
| 286 |
+
"learning_rate": 9.682539682539683e-06,
|
| 287 |
+
"loss": 1.853,
|
| 288 |
+
"step": 360
|
| 289 |
+
},
|
| 290 |
+
{
|
| 291 |
+
"epoch": 3.2761692650334076,
|
| 292 |
+
"grad_norm": 66.43023681640625,
|
| 293 |
+
"learning_rate": 9.642857142857144e-06,
|
| 294 |
+
"loss": 1.4603,
|
| 295 |
+
"step": 370
|
| 296 |
+
},
|
| 297 |
+
{
|
| 298 |
+
"epoch": 3.3652561247216037,
|
| 299 |
+
"grad_norm": 123.35740661621094,
|
| 300 |
+
"learning_rate": 9.603174603174605e-06,
|
| 301 |
+
"loss": 2.124,
|
| 302 |
+
"step": 380
|
| 303 |
+
},
|
| 304 |
+
{
|
| 305 |
+
"epoch": 3.4543429844097995,
|
| 306 |
+
"grad_norm": 46.054222106933594,
|
| 307 |
+
"learning_rate": 9.563492063492064e-06,
|
| 308 |
+
"loss": 1.8356,
|
| 309 |
+
"step": 390
|
| 310 |
+
},
|
| 311 |
+
{
|
| 312 |
+
"epoch": 3.5434298440979957,
|
| 313 |
+
"grad_norm": 202.58151245117188,
|
| 314 |
+
"learning_rate": 9.523809523809525e-06,
|
| 315 |
+
"loss": 1.6048,
|
| 316 |
+
"step": 400
|
| 317 |
+
},
|
| 318 |
+
{
|
| 319 |
+
"epoch": 3.6325167037861914,
|
| 320 |
+
"grad_norm": 80.48709869384766,
|
| 321 |
+
"learning_rate": 9.484126984126984e-06,
|
| 322 |
+
"loss": 2.1942,
|
| 323 |
+
"step": 410
|
| 324 |
+
},
|
| 325 |
+
{
|
| 326 |
+
"epoch": 3.7216035634743876,
|
| 327 |
+
"grad_norm": 73.6802978515625,
|
| 328 |
+
"learning_rate": 9.444444444444445e-06,
|
| 329 |
+
"loss": 1.7107,
|
| 330 |
+
"step": 420
|
| 331 |
+
},
|
| 332 |
+
{
|
| 333 |
+
"epoch": 3.8106904231625833,
|
| 334 |
+
"grad_norm": 80.23323822021484,
|
| 335 |
+
"learning_rate": 9.404761904761905e-06,
|
| 336 |
+
"loss": 1.5238,
|
| 337 |
+
"step": 430
|
| 338 |
+
},
|
| 339 |
+
{
|
| 340 |
+
"epoch": 3.8997772828507795,
|
| 341 |
+
"grad_norm": 62.258705139160156,
|
| 342 |
+
"learning_rate": 9.365079365079366e-06,
|
| 343 |
+
"loss": 1.7122,
|
| 344 |
+
"step": 440
|
| 345 |
+
},
|
| 346 |
+
{
|
| 347 |
+
"epoch": 3.9888641425389757,
|
| 348 |
+
"grad_norm": 76.88375854492188,
|
| 349 |
+
"learning_rate": 9.325396825396827e-06,
|
| 350 |
+
"loss": 1.6925,
|
| 351 |
+
"step": 450
|
| 352 |
+
},
|
| 353 |
+
{
|
| 354 |
+
"epoch": 4.0,
|
| 355 |
+
"eval_loss": 0.49176380038261414,
|
| 356 |
+
"eval_macro_f1": 0.7682737403294986,
|
| 357 |
+
"eval_runtime": 27.8235,
|
| 358 |
+
"eval_samples_per_second": 32.275,
|
| 359 |
+
"eval_steps_per_second": 4.061,
|
| 360 |
+
"step": 452
|
| 361 |
+
},
|
| 362 |
+
{
|
| 363 |
+
"epoch": 4.071269487750556,
|
| 364 |
+
"grad_norm": 46.56773376464844,
|
| 365 |
+
"learning_rate": 9.285714285714288e-06,
|
| 366 |
+
"loss": 1.1574,
|
| 367 |
+
"step": 460
|
| 368 |
+
},
|
| 369 |
+
{
|
| 370 |
+
"epoch": 4.160356347438753,
|
| 371 |
+
"grad_norm": 53.96033477783203,
|
| 372 |
+
"learning_rate": 9.246031746031747e-06,
|
| 373 |
+
"loss": 1.5659,
|
| 374 |
+
"step": 470
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 4.249443207126949,
|
| 378 |
+
"grad_norm": 61.506927490234375,
|
| 379 |
+
"learning_rate": 9.206349206349207e-06,
|
| 380 |
+
"loss": 1.3966,
|
| 381 |
+
"step": 480
|
| 382 |
+
},
|
| 383 |
+
{
|
| 384 |
+
"epoch": 4.338530066815145,
|
| 385 |
+
"grad_norm": 60.22172546386719,
|
| 386 |
+
"learning_rate": 9.166666666666666e-06,
|
| 387 |
+
"loss": 1.1754,
|
| 388 |
+
"step": 490
|
| 389 |
+
},
|
| 390 |
+
{
|
| 391 |
+
"epoch": 4.427616926503341,
|
| 392 |
+
"grad_norm": 56.39080810546875,
|
| 393 |
+
"learning_rate": 9.126984126984127e-06,
|
| 394 |
+
"loss": 1.6733,
|
| 395 |
+
"step": 500
|
| 396 |
+
},
|
| 397 |
+
{
|
| 398 |
+
"epoch": 4.5167037861915365,
|
| 399 |
+
"grad_norm": 66.48773193359375,
|
| 400 |
+
"learning_rate": 9.087301587301588e-06,
|
| 401 |
+
"loss": 1.1759,
|
| 402 |
+
"step": 510
|
| 403 |
+
},
|
| 404 |
+
{
|
| 405 |
+
"epoch": 4.605790645879733,
|
| 406 |
+
"grad_norm": 75.4937973022461,
|
| 407 |
+
"learning_rate": 9.047619047619049e-06,
|
| 408 |
+
"loss": 1.3468,
|
| 409 |
+
"step": 520
|
| 410 |
+
},
|
| 411 |
+
{
|
| 412 |
+
"epoch": 4.694877505567929,
|
| 413 |
+
"grad_norm": 58.68571090698242,
|
| 414 |
+
"learning_rate": 9.00793650793651e-06,
|
| 415 |
+
"loss": 1.3929,
|
| 416 |
+
"step": 530
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 4.783964365256125,
|
| 420 |
+
"grad_norm": 52.5469970703125,
|
| 421 |
+
"learning_rate": 8.968253968253968e-06,
|
| 422 |
+
"loss": 1.3248,
|
| 423 |
+
"step": 540
|
| 424 |
+
},
|
| 425 |
+
{
|
| 426 |
+
"epoch": 4.873051224944321,
|
| 427 |
+
"grad_norm": 138.5091094970703,
|
| 428 |
+
"learning_rate": 8.92857142857143e-06,
|
| 429 |
+
"loss": 1.5388,
|
| 430 |
+
"step": 550
|
| 431 |
+
},
|
| 432 |
+
{
|
| 433 |
+
"epoch": 4.9621380846325165,
|
| 434 |
+
"grad_norm": 66.50778198242188,
|
| 435 |
+
"learning_rate": 8.888888888888888e-06,
|
| 436 |
+
"loss": 1.4344,
|
| 437 |
+
"step": 560
|
| 438 |
+
},
|
| 439 |
+
{
|
| 440 |
+
"epoch": 5.0,
|
| 441 |
+
"eval_loss": 0.49091583490371704,
|
| 442 |
+
"eval_macro_f1": 0.7592655412453324,
|
| 443 |
+
"eval_runtime": 27.4798,
|
| 444 |
+
"eval_samples_per_second": 32.679,
|
| 445 |
+
"eval_steps_per_second": 4.112,
|
| 446 |
+
"step": 565
|
| 447 |
+
},
|
| 448 |
+
{
|
| 449 |
+
"epoch": 5.044543429844098,
|
| 450 |
+
"grad_norm": 99.18921661376953,
|
| 451 |
+
"learning_rate": 8.849206349206349e-06,
|
| 452 |
+
"loss": 1.3262,
|
| 453 |
+
"step": 570
|
| 454 |
+
},
|
| 455 |
+
{
|
| 456 |
+
"epoch": 5.133630289532294,
|
| 457 |
+
"grad_norm": 73.59988403320312,
|
| 458 |
+
"learning_rate": 8.80952380952381e-06,
|
| 459 |
+
"loss": 1.2284,
|
| 460 |
+
"step": 580
|
| 461 |
+
},
|
| 462 |
+
{
|
| 463 |
+
"epoch": 5.22271714922049,
|
| 464 |
+
"grad_norm": 258.84918212890625,
|
| 465 |
+
"learning_rate": 8.76984126984127e-06,
|
| 466 |
+
"loss": 1.4667,
|
| 467 |
+
"step": 590
|
| 468 |
+
},
|
| 469 |
+
{
|
| 470 |
+
"epoch": 5.311804008908686,
|
| 471 |
+
"grad_norm": 69.11897277832031,
|
| 472 |
+
"learning_rate": 8.730158730158731e-06,
|
| 473 |
+
"loss": 1.3145,
|
| 474 |
+
"step": 600
|
| 475 |
+
},
|
| 476 |
+
{
|
| 477 |
+
"epoch": 5.400890868596882,
|
| 478 |
+
"grad_norm": 33.039669036865234,
|
| 479 |
+
"learning_rate": 8.690476190476192e-06,
|
| 480 |
+
"loss": 0.9991,
|
| 481 |
+
"step": 610
|
| 482 |
+
},
|
| 483 |
+
{
|
| 484 |
+
"epoch": 5.489977728285078,
|
| 485 |
+
"grad_norm": 78.3748779296875,
|
| 486 |
+
"learning_rate": 8.650793650793651e-06,
|
| 487 |
+
"loss": 1.1708,
|
| 488 |
+
"step": 620
|
| 489 |
+
},
|
| 490 |
+
{
|
| 491 |
+
"epoch": 5.579064587973274,
|
| 492 |
+
"grad_norm": 61.205711364746094,
|
| 493 |
+
"learning_rate": 8.611111111111112e-06,
|
| 494 |
+
"loss": 1.0329,
|
| 495 |
+
"step": 630
|
| 496 |
+
},
|
| 497 |
+
{
|
| 498 |
+
"epoch": 5.66815144766147,
|
| 499 |
+
"grad_norm": 62.874446868896484,
|
| 500 |
+
"learning_rate": 8.571428571428571e-06,
|
| 501 |
+
"loss": 1.1144,
|
| 502 |
+
"step": 640
|
| 503 |
+
},
|
| 504 |
+
{
|
| 505 |
+
"epoch": 5.757238307349666,
|
| 506 |
+
"grad_norm": 86.9574203491211,
|
| 507 |
+
"learning_rate": 8.531746031746032e-06,
|
| 508 |
+
"loss": 0.9814,
|
| 509 |
+
"step": 650
|
| 510 |
+
},
|
| 511 |
+
{
|
| 512 |
+
"epoch": 5.846325167037862,
|
| 513 |
+
"grad_norm": 55.24412536621094,
|
| 514 |
+
"learning_rate": 8.492063492063492e-06,
|
| 515 |
+
"loss": 1.0457,
|
| 516 |
+
"step": 660
|
| 517 |
+
},
|
| 518 |
+
{
|
| 519 |
+
"epoch": 5.935412026726058,
|
| 520 |
+
"grad_norm": 96.79296112060547,
|
| 521 |
+
"learning_rate": 8.452380952380953e-06,
|
| 522 |
+
"loss": 1.2971,
|
| 523 |
+
"step": 670
|
| 524 |
+
},
|
| 525 |
+
{
|
| 526 |
+
"epoch": 6.0,
|
| 527 |
+
"eval_loss": 0.6005875468254089,
|
| 528 |
+
"eval_macro_f1": 0.7454914303614315,
|
| 529 |
+
"eval_runtime": 29.9288,
|
| 530 |
+
"eval_samples_per_second": 30.004,
|
| 531 |
+
"eval_steps_per_second": 3.776,
|
| 532 |
+
"step": 678
|
| 533 |
+
},
|
| 534 |
+
{
|
| 535 |
+
"epoch": 6.017817371937639,
|
| 536 |
+
"grad_norm": 124.40799713134766,
|
| 537 |
+
"learning_rate": 8.412698412698414e-06,
|
| 538 |
+
"loss": 1.3457,
|
| 539 |
+
"step": 680
|
| 540 |
+
},
|
| 541 |
+
{
|
| 542 |
+
"epoch": 6.106904231625835,
|
| 543 |
+
"grad_norm": 75.6956787109375,
|
| 544 |
+
"learning_rate": 8.373015873015875e-06,
|
| 545 |
+
"loss": 1.1006,
|
| 546 |
+
"step": 690
|
| 547 |
+
},
|
| 548 |
+
{
|
| 549 |
+
"epoch": 6.195991091314031,
|
| 550 |
+
"grad_norm": 94.5655746459961,
|
| 551 |
+
"learning_rate": 8.333333333333334e-06,
|
| 552 |
+
"loss": 1.0031,
|
| 553 |
+
"step": 700
|
| 554 |
+
},
|
| 555 |
+
{
|
| 556 |
+
"epoch": 6.285077951002227,
|
| 557 |
+
"grad_norm": 84.76609802246094,
|
| 558 |
+
"learning_rate": 8.293650793650794e-06,
|
| 559 |
+
"loss": 0.9654,
|
| 560 |
+
"step": 710
|
| 561 |
+
},
|
| 562 |
+
{
|
| 563 |
+
"epoch": 6.374164810690424,
|
| 564 |
+
"grad_norm": 44.62492752075195,
|
| 565 |
+
"learning_rate": 8.253968253968254e-06,
|
| 566 |
+
"loss": 1.0131,
|
| 567 |
+
"step": 720
|
| 568 |
+
},
|
| 569 |
+
{
|
| 570 |
+
"epoch": 6.463251670378619,
|
| 571 |
+
"grad_norm": 65.90361022949219,
|
| 572 |
+
"learning_rate": 8.214285714285714e-06,
|
| 573 |
+
"loss": 0.8061,
|
| 574 |
+
"step": 730
|
| 575 |
+
},
|
| 576 |
+
{
|
| 577 |
+
"epoch": 6.552338530066815,
|
| 578 |
+
"grad_norm": 81.76494598388672,
|
| 579 |
+
"learning_rate": 8.174603174603175e-06,
|
| 580 |
+
"loss": 0.9876,
|
| 581 |
+
"step": 740
|
| 582 |
+
},
|
| 583 |
+
{
|
| 584 |
+
"epoch": 6.641425389755011,
|
| 585 |
+
"grad_norm": 116.49346923828125,
|
| 586 |
+
"learning_rate": 8.134920634920636e-06,
|
| 587 |
+
"loss": 1.208,
|
| 588 |
+
"step": 750
|
| 589 |
+
},
|
| 590 |
+
{
|
| 591 |
+
"epoch": 6.7305122494432075,
|
| 592 |
+
"grad_norm": 50.356597900390625,
|
| 593 |
+
"learning_rate": 8.095238095238097e-06,
|
| 594 |
+
"loss": 0.9099,
|
| 595 |
+
"step": 760
|
| 596 |
+
},
|
| 597 |
+
{
|
| 598 |
+
"epoch": 6.819599109131403,
|
| 599 |
+
"grad_norm": 98.01029205322266,
|
| 600 |
+
"learning_rate": 8.055555555555557e-06,
|
| 601 |
+
"loss": 0.7744,
|
| 602 |
+
"step": 770
|
| 603 |
+
},
|
| 604 |
+
{
|
| 605 |
+
"epoch": 6.908685968819599,
|
| 606 |
+
"grad_norm": 72.4462661743164,
|
| 607 |
+
"learning_rate": 8.015873015873016e-06,
|
| 608 |
+
"loss": 1.1695,
|
| 609 |
+
"step": 780
|
| 610 |
+
},
|
| 611 |
+
{
|
| 612 |
+
"epoch": 6.997772828507795,
|
| 613 |
+
"grad_norm": 67.33780670166016,
|
| 614 |
+
"learning_rate": 7.976190476190477e-06,
|
| 615 |
+
"loss": 0.8807,
|
| 616 |
+
"step": 790
|
| 617 |
+
},
|
| 618 |
+
{
|
| 619 |
+
"epoch": 7.0,
|
| 620 |
+
"eval_loss": 0.5682073831558228,
|
| 621 |
+
"eval_macro_f1": 0.7816147859922179,
|
| 622 |
+
"eval_runtime": 24.4148,
|
| 623 |
+
"eval_samples_per_second": 36.781,
|
| 624 |
+
"eval_steps_per_second": 4.628,
|
| 625 |
+
"step": 791
|
| 626 |
+
},
|
| 627 |
+
{
|
| 628 |
+
"epoch": 7.080178173719377,
|
| 629 |
+
"grad_norm": 44.45954513549805,
|
| 630 |
+
"learning_rate": 7.936507936507936e-06,
|
| 631 |
+
"loss": 0.591,
|
| 632 |
+
"step": 800
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 7.169265033407572,
|
| 636 |
+
"grad_norm": 36.26639175415039,
|
| 637 |
+
"learning_rate": 7.896825396825397e-06,
|
| 638 |
+
"loss": 0.6219,
|
| 639 |
+
"step": 810
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 7.258351893095768,
|
| 643 |
+
"grad_norm": 40.69953155517578,
|
| 644 |
+
"learning_rate": 7.857142857142858e-06,
|
| 645 |
+
"loss": 0.6987,
|
| 646 |
+
"step": 820
|
| 647 |
+
},
|
| 648 |
+
{
|
| 649 |
+
"epoch": 7.347438752783964,
|
| 650 |
+
"grad_norm": 64.46514129638672,
|
| 651 |
+
"learning_rate": 7.817460317460318e-06,
|
| 652 |
+
"loss": 0.7194,
|
| 653 |
+
"step": 830
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 7.436525612472161,
|
| 657 |
+
"grad_norm": 63.40292739868164,
|
| 658 |
+
"learning_rate": 7.77777777777778e-06,
|
| 659 |
+
"loss": 0.7534,
|
| 660 |
+
"step": 840
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"epoch": 7.525612472160356,
|
| 664 |
+
"grad_norm": 42.322147369384766,
|
| 665 |
+
"learning_rate": 7.738095238095238e-06,
|
| 666 |
+
"loss": 0.8983,
|
| 667 |
+
"step": 850
|
| 668 |
+
},
|
| 669 |
+
{
|
| 670 |
+
"epoch": 7.614699331848552,
|
| 671 |
+
"grad_norm": 70.19657135009766,
|
| 672 |
+
"learning_rate": 7.698412698412699e-06,
|
| 673 |
+
"loss": 0.9567,
|
| 674 |
+
"step": 860
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 7.703786191536748,
|
| 678 |
+
"grad_norm": 122.25581359863281,
|
| 679 |
+
"learning_rate": 7.65873015873016e-06,
|
| 680 |
+
"loss": 0.8068,
|
| 681 |
+
"step": 870
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 7.7928730512249444,
|
| 685 |
+
"grad_norm": 71.20338439941406,
|
| 686 |
+
"learning_rate": 7.61904761904762e-06,
|
| 687 |
+
"loss": 0.8394,
|
| 688 |
+
"step": 880
|
| 689 |
+
},
|
| 690 |
+
{
|
| 691 |
+
"epoch": 7.881959910913141,
|
| 692 |
+
"grad_norm": 79.9009017944336,
|
| 693 |
+
"learning_rate": 7.5793650793650795e-06,
|
| 694 |
+
"loss": 0.8389,
|
| 695 |
+
"step": 890
|
| 696 |
+
},
|
| 697 |
+
{
|
| 698 |
+
"epoch": 7.971046770601336,
|
| 699 |
+
"grad_norm": 113.75267791748047,
|
| 700 |
+
"learning_rate": 7.53968253968254e-06,
|
| 701 |
+
"loss": 0.9274,
|
| 702 |
+
"step": 900
|
| 703 |
+
},
|
| 704 |
+
{
|
| 705 |
+
"epoch": 8.0,
|
| 706 |
+
"eval_loss": 0.6042653322219849,
|
| 707 |
+
"eval_macro_f1": 0.7506470612211624,
|
| 708 |
+
"eval_runtime": 23.2579,
|
| 709 |
+
"eval_samples_per_second": 38.611,
|
| 710 |
+
"eval_steps_per_second": 4.859,
|
| 711 |
+
"step": 904
|
| 712 |
+
},
|
| 713 |
+
{
|
| 714 |
+
"epoch": 8.053452115812918,
|
| 715 |
+
"grad_norm": 67.79409790039062,
|
| 716 |
+
"learning_rate": 7.500000000000001e-06,
|
| 717 |
+
"loss": 0.7653,
|
| 718 |
+
"step": 910
|
| 719 |
+
},
|
| 720 |
+
{
|
| 721 |
+
"epoch": 8.142538975501113,
|
| 722 |
+
"grad_norm": 45.01826095581055,
|
| 723 |
+
"learning_rate": 7.460317460317461e-06,
|
| 724 |
+
"loss": 0.4532,
|
| 725 |
+
"step": 920
|
| 726 |
+
},
|
| 727 |
+
{
|
| 728 |
+
"epoch": 8.231625835189309,
|
| 729 |
+
"grad_norm": 66.2216796875,
|
| 730 |
+
"learning_rate": 7.420634920634922e-06,
|
| 731 |
+
"loss": 0.4336,
|
| 732 |
+
"step": 930
|
| 733 |
+
},
|
| 734 |
+
{
|
| 735 |
+
"epoch": 8.320712694877505,
|
| 736 |
+
"grad_norm": 82.00172424316406,
|
| 737 |
+
"learning_rate": 7.380952380952382e-06,
|
| 738 |
+
"loss": 0.771,
|
| 739 |
+
"step": 940
|
| 740 |
+
},
|
| 741 |
+
{
|
| 742 |
+
"epoch": 8.409799554565701,
|
| 743 |
+
"grad_norm": 37.49010467529297,
|
| 744 |
+
"learning_rate": 7.3412698412698415e-06,
|
| 745 |
+
"loss": 0.7434,
|
| 746 |
+
"step": 950
|
| 747 |
+
},
|
| 748 |
+
{
|
| 749 |
+
"epoch": 8.498886414253898,
|
| 750 |
+
"grad_norm": 52.00114440917969,
|
| 751 |
+
"learning_rate": 7.301587301587301e-06,
|
| 752 |
+
"loss": 0.5471,
|
| 753 |
+
"step": 960
|
| 754 |
+
},
|
| 755 |
+
{
|
| 756 |
+
"epoch": 8.587973273942094,
|
| 757 |
+
"grad_norm": 58.92430114746094,
|
| 758 |
+
"learning_rate": 7.261904761904762e-06,
|
| 759 |
+
"loss": 0.95,
|
| 760 |
+
"step": 970
|
| 761 |
+
},
|
| 762 |
+
{
|
| 763 |
+
"epoch": 8.67706013363029,
|
| 764 |
+
"grad_norm": 48.648902893066406,
|
| 765 |
+
"learning_rate": 7.222222222222223e-06,
|
| 766 |
+
"loss": 0.9863,
|
| 767 |
+
"step": 980
|
| 768 |
+
},
|
| 769 |
+
{
|
| 770 |
+
"epoch": 8.766146993318486,
|
| 771 |
+
"grad_norm": 50.49428176879883,
|
| 772 |
+
"learning_rate": 7.182539682539683e-06,
|
| 773 |
+
"loss": 0.7554,
|
| 774 |
+
"step": 990
|
| 775 |
+
},
|
| 776 |
+
{
|
| 777 |
+
"epoch": 8.855233853006682,
|
| 778 |
+
"grad_norm": 24.385663986206055,
|
| 779 |
+
"learning_rate": 7.1428571428571436e-06,
|
| 780 |
+
"loss": 0.5002,
|
| 781 |
+
"step": 1000
|
| 782 |
+
},
|
| 783 |
+
{
|
| 784 |
+
"epoch": 8.944320712694877,
|
| 785 |
+
"grad_norm": 54.94889450073242,
|
| 786 |
+
"learning_rate": 7.103174603174604e-06,
|
| 787 |
+
"loss": 0.6093,
|
| 788 |
+
"step": 1010
|
| 789 |
+
},
|
| 790 |
+
{
|
| 791 |
+
"epoch": 9.0,
|
| 792 |
+
"eval_loss": 0.6689905524253845,
|
| 793 |
+
"eval_macro_f1": 0.7741357801118758,
|
| 794 |
+
"eval_runtime": 19.7981,
|
| 795 |
+
"eval_samples_per_second": 45.358,
|
| 796 |
+
"eval_steps_per_second": 5.708,
|
| 797 |
+
"step": 1017
|
| 798 |
+
},
|
| 799 |
+
{
|
| 800 |
+
"epoch": 9.026726057906458,
|
| 801 |
+
"grad_norm": 60.106876373291016,
|
| 802 |
+
"learning_rate": 7.063492063492064e-06,
|
| 803 |
+
"loss": 0.6911,
|
| 804 |
+
"step": 1020
|
| 805 |
+
},
|
| 806 |
+
{
|
| 807 |
+
"epoch": 9.115812917594655,
|
| 808 |
+
"grad_norm": 39.45969772338867,
|
| 809 |
+
"learning_rate": 7.023809523809524e-06,
|
| 810 |
+
"loss": 0.5881,
|
| 811 |
+
"step": 1030
|
| 812 |
+
},
|
| 813 |
+
{
|
| 814 |
+
"epoch": 9.20489977728285,
|
| 815 |
+
"grad_norm": 60.09040069580078,
|
| 816 |
+
"learning_rate": 6.984126984126984e-06,
|
| 817 |
+
"loss": 0.4936,
|
| 818 |
+
"step": 1040
|
| 819 |
+
},
|
| 820 |
+
{
|
| 821 |
+
"epoch": 9.293986636971047,
|
| 822 |
+
"grad_norm": 53.53305435180664,
|
| 823 |
+
"learning_rate": 6.944444444444445e-06,
|
| 824 |
+
"loss": 0.6178,
|
| 825 |
+
"step": 1050
|
| 826 |
+
},
|
| 827 |
+
{
|
| 828 |
+
"epoch": 9.383073496659243,
|
| 829 |
+
"grad_norm": 23.2546443939209,
|
| 830 |
+
"learning_rate": 6.9047619047619055e-06,
|
| 831 |
+
"loss": 0.6101,
|
| 832 |
+
"step": 1060
|
| 833 |
+
},
|
| 834 |
+
{
|
| 835 |
+
"epoch": 9.47216035634744,
|
| 836 |
+
"grad_norm": 53.65586471557617,
|
| 837 |
+
"learning_rate": 6.8650793650793654e-06,
|
| 838 |
+
"loss": 0.5035,
|
| 839 |
+
"step": 1070
|
| 840 |
+
},
|
| 841 |
+
{
|
| 842 |
+
"epoch": 9.561247216035635,
|
| 843 |
+
"grad_norm": 48.904598236083984,
|
| 844 |
+
"learning_rate": 6.825396825396826e-06,
|
| 845 |
+
"loss": 0.5098,
|
| 846 |
+
"step": 1080
|
| 847 |
+
},
|
| 848 |
+
{
|
| 849 |
+
"epoch": 9.65033407572383,
|
| 850 |
+
"grad_norm": 30.416839599609375,
|
| 851 |
+
"learning_rate": 6.785714285714287e-06,
|
| 852 |
+
"loss": 0.5424,
|
| 853 |
+
"step": 1090
|
| 854 |
+
},
|
| 855 |
+
{
|
| 856 |
+
"epoch": 9.739420935412026,
|
| 857 |
+
"grad_norm": 66.94727325439453,
|
| 858 |
+
"learning_rate": 6.746031746031747e-06,
|
| 859 |
+
"loss": 0.5678,
|
| 860 |
+
"step": 1100
|
| 861 |
+
},
|
| 862 |
+
{
|
| 863 |
+
"epoch": 9.828507795100222,
|
| 864 |
+
"grad_norm": 46.05035400390625,
|
| 865 |
+
"learning_rate": 6.706349206349207e-06,
|
| 866 |
+
"loss": 0.4964,
|
| 867 |
+
"step": 1110
|
| 868 |
+
},
|
| 869 |
+
{
|
| 870 |
+
"epoch": 9.917594654788418,
|
| 871 |
+
"grad_norm": 137.4243927001953,
|
| 872 |
+
"learning_rate": 6.666666666666667e-06,
|
| 873 |
+
"loss": 0.6045,
|
| 874 |
+
"step": 1120
|
| 875 |
+
},
|
| 876 |
+
{
|
| 877 |
+
"epoch": 10.0,
|
| 878 |
+
"grad_norm": 7.9022393226623535,
|
| 879 |
+
"learning_rate": 6.626984126984127e-06,
|
| 880 |
+
"loss": 0.7078,
|
| 881 |
+
"step": 1130
|
| 882 |
+
},
|
| 883 |
+
{
|
| 884 |
+
"epoch": 10.0,
|
| 885 |
+
"eval_loss": 0.6592049598693848,
|
| 886 |
+
"eval_macro_f1": 0.7537628898597082,
|
| 887 |
+
"eval_runtime": 22.1013,
|
| 888 |
+
"eval_samples_per_second": 40.631,
|
| 889 |
+
"eval_steps_per_second": 5.113,
|
| 890 |
+
"step": 1130
|
| 891 |
+
},
|
| 892 |
+
{
|
| 893 |
+
"epoch": 10.089086859688196,
|
| 894 |
+
"grad_norm": 52.17204284667969,
|
| 895 |
+
"learning_rate": 6.587301587301588e-06,
|
| 896 |
+
"loss": 0.6119,
|
| 897 |
+
"step": 1140
|
| 898 |
+
},
|
| 899 |
+
{
|
| 900 |
+
"epoch": 10.178173719376392,
|
| 901 |
+
"grad_norm": 21.41937255859375,
|
| 902 |
+
"learning_rate": 6.547619047619048e-06,
|
| 903 |
+
"loss": 0.6984,
|
| 904 |
+
"step": 1150
|
| 905 |
+
},
|
| 906 |
+
{
|
| 907 |
+
"epoch": 10.267260579064589,
|
| 908 |
+
"grad_norm": 68.02864837646484,
|
| 909 |
+
"learning_rate": 6.507936507936509e-06,
|
| 910 |
+
"loss": 0.6225,
|
| 911 |
+
"step": 1160
|
| 912 |
+
},
|
| 913 |
+
{
|
| 914 |
+
"epoch": 10.356347438752785,
|
| 915 |
+
"grad_norm": 125.40614318847656,
|
| 916 |
+
"learning_rate": 6.468253968253969e-06,
|
| 917 |
+
"loss": 0.6566,
|
| 918 |
+
"step": 1170
|
| 919 |
+
},
|
| 920 |
+
{
|
| 921 |
+
"epoch": 10.44543429844098,
|
| 922 |
+
"grad_norm": 37.92251205444336,
|
| 923 |
+
"learning_rate": 6.4285714285714295e-06,
|
| 924 |
+
"loss": 0.5528,
|
| 925 |
+
"step": 1180
|
| 926 |
+
},
|
| 927 |
+
{
|
| 928 |
+
"epoch": 10.534521158129175,
|
| 929 |
+
"grad_norm": 51.56312561035156,
|
| 930 |
+
"learning_rate": 6.3888888888888885e-06,
|
| 931 |
+
"loss": 0.3961,
|
| 932 |
+
"step": 1190
|
| 933 |
+
},
|
| 934 |
+
{
|
| 935 |
+
"epoch": 10.623608017817372,
|
| 936 |
+
"grad_norm": 56.06890869140625,
|
| 937 |
+
"learning_rate": 6.349206349206349e-06,
|
| 938 |
+
"loss": 0.4413,
|
| 939 |
+
"step": 1200
|
| 940 |
+
},
|
| 941 |
+
{
|
| 942 |
+
"epoch": 10.712694877505568,
|
| 943 |
+
"grad_norm": 396.573974609375,
|
| 944 |
+
"learning_rate": 6.30952380952381e-06,
|
| 945 |
+
"loss": 0.625,
|
| 946 |
+
"step": 1210
|
| 947 |
+
},
|
| 948 |
+
{
|
| 949 |
+
"epoch": 10.801781737193764,
|
| 950 |
+
"grad_norm": 28.54755401611328,
|
| 951 |
+
"learning_rate": 6.26984126984127e-06,
|
| 952 |
+
"loss": 0.2236,
|
| 953 |
+
"step": 1220
|
| 954 |
+
},
|
| 955 |
+
{
|
| 956 |
+
"epoch": 10.89086859688196,
|
| 957 |
+
"grad_norm": 90.92909240722656,
|
| 958 |
+
"learning_rate": 6.230158730158731e-06,
|
| 959 |
+
"loss": 0.6644,
|
| 960 |
+
"step": 1230
|
| 961 |
+
},
|
| 962 |
+
{
|
| 963 |
+
"epoch": 10.979955456570156,
|
| 964 |
+
"grad_norm": 42.810672760009766,
|
| 965 |
+
"learning_rate": 6.1904761904761914e-06,
|
| 966 |
+
"loss": 0.5941,
|
| 967 |
+
"step": 1240
|
| 968 |
+
},
|
| 969 |
+
{
|
| 970 |
+
"epoch": 11.0,
|
| 971 |
+
"eval_loss": 0.8170925974845886,
|
| 972 |
+
"eval_macro_f1": 0.7553111396076517,
|
| 973 |
+
"eval_runtime": 21.8437,
|
| 974 |
+
"eval_samples_per_second": 41.11,
|
| 975 |
+
"eval_steps_per_second": 5.173,
|
| 976 |
+
"step": 1243
|
| 977 |
+
},
|
| 978 |
+
{
|
| 979 |
+
"epoch": 11.062360801781738,
|
| 980 |
+
"grad_norm": 45.25354766845703,
|
| 981 |
+
"learning_rate": 6.150793650793651e-06,
|
| 982 |
+
"loss": 0.6691,
|
| 983 |
+
"step": 1250
|
| 984 |
+
},
|
| 985 |
+
{
|
| 986 |
+
"epoch": 11.151447661469934,
|
| 987 |
+
"grad_norm": 27.80849266052246,
|
| 988 |
+
"learning_rate": 6.111111111111112e-06,
|
| 989 |
+
"loss": 0.5407,
|
| 990 |
+
"step": 1260
|
| 991 |
+
},
|
| 992 |
+
{
|
| 993 |
+
"epoch": 11.240534521158128,
|
| 994 |
+
"grad_norm": 43.23250961303711,
|
| 995 |
+
"learning_rate": 6.071428571428571e-06,
|
| 996 |
+
"loss": 0.4531,
|
| 997 |
+
"step": 1270
|
| 998 |
+
},
|
| 999 |
+
{
|
| 1000 |
+
"epoch": 11.329621380846325,
|
| 1001 |
+
"grad_norm": 91.41048431396484,
|
| 1002 |
+
"learning_rate": 6.031746031746032e-06,
|
| 1003 |
+
"loss": 0.3653,
|
| 1004 |
+
"step": 1280
|
| 1005 |
+
},
|
| 1006 |
+
{
|
| 1007 |
+
"epoch": 11.41870824053452,
|
| 1008 |
+
"grad_norm": 62.16683578491211,
|
| 1009 |
+
"learning_rate": 5.992063492063493e-06,
|
| 1010 |
+
"loss": 0.6402,
|
| 1011 |
+
"step": 1290
|
| 1012 |
+
},
|
| 1013 |
+
{
|
| 1014 |
+
"epoch": 11.507795100222717,
|
| 1015 |
+
"grad_norm": 29.885025024414062,
|
| 1016 |
+
"learning_rate": 5.9523809523809525e-06,
|
| 1017 |
+
"loss": 0.3978,
|
| 1018 |
+
"step": 1300
|
| 1019 |
+
},
|
| 1020 |
+
{
|
| 1021 |
+
"epoch": 11.596881959910913,
|
| 1022 |
+
"grad_norm": 35.54779815673828,
|
| 1023 |
+
"learning_rate": 5.912698412698413e-06,
|
| 1024 |
+
"loss": 0.5595,
|
| 1025 |
+
"step": 1310
|
| 1026 |
+
},
|
| 1027 |
+
{
|
| 1028 |
+
"epoch": 11.68596881959911,
|
| 1029 |
+
"grad_norm": 111.557373046875,
|
| 1030 |
+
"learning_rate": 5.873015873015874e-06,
|
| 1031 |
+
"loss": 0.4358,
|
| 1032 |
+
"step": 1320
|
| 1033 |
+
},
|
| 1034 |
+
{
|
| 1035 |
+
"epoch": 11.775055679287306,
|
| 1036 |
+
"grad_norm": 63.33297348022461,
|
| 1037 |
+
"learning_rate": 5.833333333333334e-06,
|
| 1038 |
+
"loss": 0.5095,
|
| 1039 |
+
"step": 1330
|
| 1040 |
+
},
|
| 1041 |
+
{
|
| 1042 |
+
"epoch": 11.864142538975502,
|
| 1043 |
+
"grad_norm": 96.7313232421875,
|
| 1044 |
+
"learning_rate": 5.793650793650795e-06,
|
| 1045 |
+
"loss": 0.6317,
|
| 1046 |
+
"step": 1340
|
| 1047 |
+
},
|
| 1048 |
+
{
|
| 1049 |
+
"epoch": 11.953229398663698,
|
| 1050 |
+
"grad_norm": 92.20706176757812,
|
| 1051 |
+
"learning_rate": 5.753968253968254e-06,
|
| 1052 |
+
"loss": 0.6447,
|
| 1053 |
+
"step": 1350
|
| 1054 |
+
},
|
| 1055 |
+
{
|
| 1056 |
+
"epoch": 12.0,
|
| 1057 |
+
"eval_loss": 0.6678270101547241,
|
| 1058 |
+
"eval_macro_f1": 0.7712242502141149,
|
| 1059 |
+
"eval_runtime": 21.8974,
|
| 1060 |
+
"eval_samples_per_second": 41.009,
|
| 1061 |
+
"eval_steps_per_second": 5.16,
|
| 1062 |
+
"step": 1356
|
| 1063 |
+
},
|
| 1064 |
+
{
|
| 1065 |
+
"epoch": 12.035634743875278,
|
| 1066 |
+
"grad_norm": 42.24013900756836,
|
| 1067 |
+
"learning_rate": 5.7142857142857145e-06,
|
| 1068 |
+
"loss": 0.2802,
|
| 1069 |
+
"step": 1360
|
| 1070 |
+
},
|
| 1071 |
+
{
|
| 1072 |
+
"epoch": 12.124721603563474,
|
| 1073 |
+
"grad_norm": 64.52838134765625,
|
| 1074 |
+
"learning_rate": 5.674603174603175e-06,
|
| 1075 |
+
"loss": 0.5235,
|
| 1076 |
+
"step": 1370
|
| 1077 |
+
},
|
| 1078 |
+
{
|
| 1079 |
+
"epoch": 12.21380846325167,
|
| 1080 |
+
"grad_norm": 26.648761749267578,
|
| 1081 |
+
"learning_rate": 5.634920634920635e-06,
|
| 1082 |
+
"loss": 0.2894,
|
| 1083 |
+
"step": 1380
|
| 1084 |
+
},
|
| 1085 |
+
{
|
| 1086 |
+
"epoch": 12.302895322939866,
|
| 1087 |
+
"grad_norm": 38.63190841674805,
|
| 1088 |
+
"learning_rate": 5.595238095238096e-06,
|
| 1089 |
+
"loss": 0.3347,
|
| 1090 |
+
"step": 1390
|
| 1091 |
+
},
|
| 1092 |
+
{
|
| 1093 |
+
"epoch": 12.391982182628063,
|
| 1094 |
+
"grad_norm": 38.316375732421875,
|
| 1095 |
+
"learning_rate": 5.555555555555557e-06,
|
| 1096 |
+
"loss": 0.408,
|
| 1097 |
+
"step": 1400
|
| 1098 |
+
},
|
| 1099 |
+
{
|
| 1100 |
+
"epoch": 12.481069042316259,
|
| 1101 |
+
"grad_norm": 9.889010429382324,
|
| 1102 |
+
"learning_rate": 5.5158730158730166e-06,
|
| 1103 |
+
"loss": 0.4411,
|
| 1104 |
+
"step": 1410
|
| 1105 |
+
},
|
| 1106 |
+
{
|
| 1107 |
+
"epoch": 12.570155902004455,
|
| 1108 |
+
"grad_norm": 53.189300537109375,
|
| 1109 |
+
"learning_rate": 5.476190476190477e-06,
|
| 1110 |
+
"loss": 0.6681,
|
| 1111 |
+
"step": 1420
|
| 1112 |
+
},
|
| 1113 |
+
{
|
| 1114 |
+
"epoch": 12.659242761692651,
|
| 1115 |
+
"grad_norm": 30.18885612487793,
|
| 1116 |
+
"learning_rate": 5.436507936507936e-06,
|
| 1117 |
+
"loss": 0.5419,
|
| 1118 |
+
"step": 1430
|
| 1119 |
+
},
|
| 1120 |
+
{
|
| 1121 |
+
"epoch": 12.748329621380847,
|
| 1122 |
+
"grad_norm": 40.49357604980469,
|
| 1123 |
+
"learning_rate": 5.396825396825397e-06,
|
| 1124 |
+
"loss": 0.5676,
|
| 1125 |
+
"step": 1440
|
| 1126 |
+
},
|
| 1127 |
+
{
|
| 1128 |
+
"epoch": 12.837416481069042,
|
| 1129 |
+
"grad_norm": 87.46019744873047,
|
| 1130 |
+
"learning_rate": 5.357142857142857e-06,
|
| 1131 |
+
"loss": 0.7195,
|
| 1132 |
+
"step": 1450
|
| 1133 |
+
},
|
| 1134 |
+
{
|
| 1135 |
+
"epoch": 12.926503340757238,
|
| 1136 |
+
"grad_norm": 58.786277770996094,
|
| 1137 |
+
"learning_rate": 5.317460317460318e-06,
|
| 1138 |
+
"loss": 0.3799,
|
| 1139 |
+
"step": 1460
|
| 1140 |
+
},
|
| 1141 |
+
{
|
| 1142 |
+
"epoch": 13.0,
|
| 1143 |
+
"eval_loss": 0.6582129001617432,
|
| 1144 |
+
"eval_macro_f1": 0.7994080527168146,
|
| 1145 |
+
"eval_runtime": 22.0321,
|
| 1146 |
+
"eval_samples_per_second": 40.759,
|
| 1147 |
+
"eval_steps_per_second": 5.129,
|
| 1148 |
+
"step": 1469
|
| 1149 |
+
},
|
| 1150 |
+
{
|
| 1151 |
+
"epoch": 13.00890868596882,
|
| 1152 |
+
"grad_norm": 23.018239974975586,
|
| 1153 |
+
"learning_rate": 5.2777777777777785e-06,
|
| 1154 |
+
"loss": 0.4031,
|
| 1155 |
+
"step": 1470
|
| 1156 |
+
},
|
| 1157 |
+
{
|
| 1158 |
+
"epoch": 13.097995545657016,
|
| 1159 |
+
"grad_norm": 47.901573181152344,
|
| 1160 |
+
"learning_rate": 5.2380952380952384e-06,
|
| 1161 |
+
"loss": 0.2202,
|
| 1162 |
+
"step": 1480
|
| 1163 |
+
},
|
| 1164 |
+
{
|
| 1165 |
+
"epoch": 13.187082405345212,
|
| 1166 |
+
"grad_norm": 95.25432586669922,
|
| 1167 |
+
"learning_rate": 5.198412698412699e-06,
|
| 1168 |
+
"loss": 0.58,
|
| 1169 |
+
"step": 1490
|
| 1170 |
+
},
|
| 1171 |
+
{
|
| 1172 |
+
"epoch": 13.276169265033408,
|
| 1173 |
+
"grad_norm": 52.78740692138672,
|
| 1174 |
+
"learning_rate": 5.15873015873016e-06,
|
| 1175 |
+
"loss": 0.399,
|
| 1176 |
+
"step": 1500
|
| 1177 |
+
},
|
| 1178 |
+
{
|
| 1179 |
+
"epoch": 13.365256124721604,
|
| 1180 |
+
"grad_norm": 45.429832458496094,
|
| 1181 |
+
"learning_rate": 5.119047619047619e-06,
|
| 1182 |
+
"loss": 0.4761,
|
| 1183 |
+
"step": 1510
|
| 1184 |
+
},
|
| 1185 |
+
{
|
| 1186 |
+
"epoch": 13.4543429844098,
|
| 1187 |
+
"grad_norm": 64.204833984375,
|
| 1188 |
+
"learning_rate": 5.07936507936508e-06,
|
| 1189 |
+
"loss": 0.3218,
|
| 1190 |
+
"step": 1520
|
| 1191 |
+
},
|
| 1192 |
+
{
|
| 1193 |
+
"epoch": 13.543429844097995,
|
| 1194 |
+
"grad_norm": 20.63060188293457,
|
| 1195 |
+
"learning_rate": 5.03968253968254e-06,
|
| 1196 |
+
"loss": 0.3994,
|
| 1197 |
+
"step": 1530
|
| 1198 |
+
},
|
| 1199 |
+
{
|
| 1200 |
+
"epoch": 13.632516703786191,
|
| 1201 |
+
"grad_norm": 35.755001068115234,
|
| 1202 |
+
"learning_rate": 5e-06,
|
| 1203 |
+
"loss": 0.5116,
|
| 1204 |
+
"step": 1540
|
| 1205 |
+
},
|
| 1206 |
+
{
|
| 1207 |
+
"epoch": 13.721603563474387,
|
| 1208 |
+
"grad_norm": 64.8443603515625,
|
| 1209 |
+
"learning_rate": 4.960317460317461e-06,
|
| 1210 |
+
"loss": 0.2796,
|
| 1211 |
+
"step": 1550
|
| 1212 |
+
},
|
| 1213 |
+
{
|
| 1214 |
+
"epoch": 13.810690423162583,
|
| 1215 |
+
"grad_norm": 66.51502990722656,
|
| 1216 |
+
"learning_rate": 4.920634920634921e-06,
|
| 1217 |
+
"loss": 0.4389,
|
| 1218 |
+
"step": 1560
|
| 1219 |
+
},
|
| 1220 |
+
{
|
| 1221 |
+
"epoch": 13.89977728285078,
|
| 1222 |
+
"grad_norm": 31.428354263305664,
|
| 1223 |
+
"learning_rate": 4.880952380952381e-06,
|
| 1224 |
+
"loss": 0.4831,
|
| 1225 |
+
"step": 1570
|
| 1226 |
+
},
|
| 1227 |
+
{
|
| 1228 |
+
"epoch": 13.988864142538976,
|
| 1229 |
+
"grad_norm": 17.072656631469727,
|
| 1230 |
+
"learning_rate": 4.841269841269842e-06,
|
| 1231 |
+
"loss": 0.353,
|
| 1232 |
+
"step": 1580
|
| 1233 |
+
},
|
| 1234 |
+
{
|
| 1235 |
+
"epoch": 14.0,
|
| 1236 |
+
"eval_loss": 0.7872561812400818,
|
| 1237 |
+
"eval_macro_f1": 0.7766969045913902,
|
| 1238 |
+
"eval_runtime": 22.1801,
|
| 1239 |
+
"eval_samples_per_second": 40.487,
|
| 1240 |
+
"eval_steps_per_second": 5.095,
|
| 1241 |
+
"step": 1582
|
| 1242 |
+
},
|
| 1243 |
+
{
|
| 1244 |
+
"epoch": 14.071269487750557,
|
| 1245 |
+
"grad_norm": 54.113372802734375,
|
| 1246 |
+
"learning_rate": 4.8015873015873025e-06,
|
| 1247 |
+
"loss": 0.553,
|
| 1248 |
+
"step": 1590
|
| 1249 |
+
},
|
| 1250 |
+
{
|
| 1251 |
+
"epoch": 14.160356347438753,
|
| 1252 |
+
"grad_norm": 47.31705093383789,
|
| 1253 |
+
"learning_rate": 4.761904761904762e-06,
|
| 1254 |
+
"loss": 0.3445,
|
| 1255 |
+
"step": 1600
|
| 1256 |
+
},
|
| 1257 |
+
{
|
| 1258 |
+
"epoch": 14.249443207126948,
|
| 1259 |
+
"grad_norm": 69.48124694824219,
|
| 1260 |
+
"learning_rate": 4.722222222222222e-06,
|
| 1261 |
+
"loss": 0.3443,
|
| 1262 |
+
"step": 1610
|
| 1263 |
+
},
|
| 1264 |
+
{
|
| 1265 |
+
"epoch": 14.338530066815144,
|
| 1266 |
+
"grad_norm": 46.73830032348633,
|
| 1267 |
+
"learning_rate": 4.682539682539683e-06,
|
| 1268 |
+
"loss": 0.5484,
|
| 1269 |
+
"step": 1620
|
| 1270 |
+
},
|
| 1271 |
+
{
|
| 1272 |
+
"epoch": 14.42761692650334,
|
| 1273 |
+
"grad_norm": 41.929805755615234,
|
| 1274 |
+
"learning_rate": 4.642857142857144e-06,
|
| 1275 |
+
"loss": 0.2983,
|
| 1276 |
+
"step": 1630
|
| 1277 |
+
},
|
| 1278 |
+
{
|
| 1279 |
+
"epoch": 14.516703786191536,
|
| 1280 |
+
"grad_norm": 25.568206787109375,
|
| 1281 |
+
"learning_rate": 4.603174603174604e-06,
|
| 1282 |
+
"loss": 0.3027,
|
| 1283 |
+
"step": 1640
|
| 1284 |
+
},
|
| 1285 |
+
{
|
| 1286 |
+
"epoch": 14.605790645879733,
|
| 1287 |
+
"grad_norm": 77.96745300292969,
|
| 1288 |
+
"learning_rate": 4.563492063492064e-06,
|
| 1289 |
+
"loss": 0.6002,
|
| 1290 |
+
"step": 1650
|
| 1291 |
+
},
|
| 1292 |
+
{
|
| 1293 |
+
"epoch": 14.694877505567929,
|
| 1294 |
+
"grad_norm": 32.5285758972168,
|
| 1295 |
+
"learning_rate": 4.523809523809524e-06,
|
| 1296 |
+
"loss": 0.3205,
|
| 1297 |
+
"step": 1660
|
| 1298 |
+
},
|
| 1299 |
+
{
|
| 1300 |
+
"epoch": 14.783964365256125,
|
| 1301 |
+
"grad_norm": 68.35356903076172,
|
| 1302 |
+
"learning_rate": 4.484126984126984e-06,
|
| 1303 |
+
"loss": 0.2666,
|
| 1304 |
+
"step": 1670
|
| 1305 |
+
},
|
| 1306 |
+
{
|
| 1307 |
+
"epoch": 14.873051224944321,
|
| 1308 |
+
"grad_norm": 2.4230611324310303,
|
| 1309 |
+
"learning_rate": 4.444444444444444e-06,
|
| 1310 |
+
"loss": 0.2947,
|
| 1311 |
+
"step": 1680
|
| 1312 |
+
},
|
| 1313 |
+
{
|
| 1314 |
+
"epoch": 14.962138084632517,
|
| 1315 |
+
"grad_norm": 18.08596420288086,
|
| 1316 |
+
"learning_rate": 4.404761904761905e-06,
|
| 1317 |
+
"loss": 0.4441,
|
| 1318 |
+
"step": 1690
|
| 1319 |
+
},
|
| 1320 |
+
{
|
| 1321 |
+
"epoch": 15.0,
|
| 1322 |
+
"eval_loss": 0.8163199424743652,
|
| 1323 |
+
"eval_macro_f1": 0.7852367846635664,
|
| 1324 |
+
"eval_runtime": 22.2418,
|
| 1325 |
+
"eval_samples_per_second": 40.374,
|
| 1326 |
+
"eval_steps_per_second": 5.081,
|
| 1327 |
+
"step": 1695
|
| 1328 |
+
},
|
| 1329 |
+
{
|
| 1330 |
+
"epoch": 15.044543429844097,
|
| 1331 |
+
"grad_norm": 11.731039047241211,
|
| 1332 |
+
"learning_rate": 4.365079365079366e-06,
|
| 1333 |
+
"loss": 0.239,
|
| 1334 |
+
"step": 1700
|
| 1335 |
+
},
|
| 1336 |
+
{
|
| 1337 |
+
"epoch": 15.133630289532293,
|
| 1338 |
+
"grad_norm": 44.82026672363281,
|
| 1339 |
+
"learning_rate": 4.3253968253968256e-06,
|
| 1340 |
+
"loss": 0.1949,
|
| 1341 |
+
"step": 1710
|
| 1342 |
+
},
|
| 1343 |
+
{
|
| 1344 |
+
"epoch": 15.22271714922049,
|
| 1345 |
+
"grad_norm": 24.14330291748047,
|
| 1346 |
+
"learning_rate": 4.2857142857142855e-06,
|
| 1347 |
+
"loss": 0.2407,
|
| 1348 |
+
"step": 1720
|
| 1349 |
+
},
|
| 1350 |
+
{
|
| 1351 |
+
"epoch": 15.311804008908686,
|
| 1352 |
+
"grad_norm": 71.19681549072266,
|
| 1353 |
+
"learning_rate": 4.246031746031746e-06,
|
| 1354 |
+
"loss": 0.6281,
|
| 1355 |
+
"step": 1730
|
| 1356 |
+
},
|
| 1357 |
+
{
|
| 1358 |
+
"epoch": 15.400890868596882,
|
| 1359 |
+
"grad_norm": 86.13356018066406,
|
| 1360 |
+
"learning_rate": 4.206349206349207e-06,
|
| 1361 |
+
"loss": 0.4672,
|
| 1362 |
+
"step": 1740
|
| 1363 |
+
},
|
| 1364 |
+
{
|
| 1365 |
+
"epoch": 15.489977728285078,
|
| 1366 |
+
"grad_norm": 6.331802845001221,
|
| 1367 |
+
"learning_rate": 4.166666666666667e-06,
|
| 1368 |
+
"loss": 0.2816,
|
| 1369 |
+
"step": 1750
|
| 1370 |
+
},
|
| 1371 |
+
{
|
| 1372 |
+
"epoch": 15.579064587973274,
|
| 1373 |
+
"grad_norm": 6.564174652099609,
|
| 1374 |
+
"learning_rate": 4.126984126984127e-06,
|
| 1375 |
+
"loss": 0.2649,
|
| 1376 |
+
"step": 1760
|
| 1377 |
+
},
|
| 1378 |
+
{
|
| 1379 |
+
"epoch": 15.66815144766147,
|
| 1380 |
+
"grad_norm": 43.143985748291016,
|
| 1381 |
+
"learning_rate": 4.0873015873015875e-06,
|
| 1382 |
+
"loss": 0.3499,
|
| 1383 |
+
"step": 1770
|
| 1384 |
+
},
|
| 1385 |
+
{
|
| 1386 |
+
"epoch": 15.757238307349667,
|
| 1387 |
+
"grad_norm": 11.393004417419434,
|
| 1388 |
+
"learning_rate": 4.047619047619048e-06,
|
| 1389 |
+
"loss": 0.2171,
|
| 1390 |
+
"step": 1780
|
| 1391 |
+
},
|
| 1392 |
+
{
|
| 1393 |
+
"epoch": 15.846325167037861,
|
| 1394 |
+
"grad_norm": 119.17877960205078,
|
| 1395 |
+
"learning_rate": 4.007936507936508e-06,
|
| 1396 |
+
"loss": 0.3459,
|
| 1397 |
+
"step": 1790
|
| 1398 |
+
},
|
| 1399 |
+
{
|
| 1400 |
+
"epoch": 15.935412026726057,
|
| 1401 |
+
"grad_norm": 135.7924041748047,
|
| 1402 |
+
"learning_rate": 3.968253968253968e-06,
|
| 1403 |
+
"loss": 0.3436,
|
| 1404 |
+
"step": 1800
|
| 1405 |
+
},
|
| 1406 |
+
{
|
| 1407 |
+
"epoch": 16.0,
|
| 1408 |
+
"eval_loss": 0.7593511343002319,
|
| 1409 |
+
"eval_macro_f1": 0.782747836115098,
|
| 1410 |
+
"eval_runtime": 22.0369,
|
| 1411 |
+
"eval_samples_per_second": 40.75,
|
| 1412 |
+
"eval_steps_per_second": 5.128,
|
| 1413 |
+
"step": 1808
|
| 1414 |
+
},
|
| 1415 |
+
{
|
| 1416 |
+
"epoch": 16.01781737193764,
|
| 1417 |
+
"grad_norm": 32.96919250488281,
|
| 1418 |
+
"learning_rate": 3.928571428571429e-06,
|
| 1419 |
+
"loss": 0.1912,
|
| 1420 |
+
"step": 1810
|
| 1421 |
+
},
|
| 1422 |
+
{
|
| 1423 |
+
"epoch": 16.106904231625837,
|
| 1424 |
+
"grad_norm": 54.140220642089844,
|
| 1425 |
+
"learning_rate": 3.88888888888889e-06,
|
| 1426 |
+
"loss": 0.3197,
|
| 1427 |
+
"step": 1820
|
| 1428 |
+
},
|
| 1429 |
+
{
|
| 1430 |
+
"epoch": 16.19599109131403,
|
| 1431 |
+
"grad_norm": 122.20573425292969,
|
| 1432 |
+
"learning_rate": 3.8492063492063495e-06,
|
| 1433 |
+
"loss": 0.4163,
|
| 1434 |
+
"step": 1830
|
| 1435 |
+
},
|
| 1436 |
+
{
|
| 1437 |
+
"epoch": 16.285077951002226,
|
| 1438 |
+
"grad_norm": 26.77570915222168,
|
| 1439 |
+
"learning_rate": 3.80952380952381e-06,
|
| 1440 |
+
"loss": 0.3443,
|
| 1441 |
+
"step": 1840
|
| 1442 |
+
},
|
| 1443 |
+
{
|
| 1444 |
+
"epoch": 16.374164810690424,
|
| 1445 |
+
"grad_norm": 41.13142395019531,
|
| 1446 |
+
"learning_rate": 3.76984126984127e-06,
|
| 1447 |
+
"loss": 0.3533,
|
| 1448 |
+
"step": 1850
|
| 1449 |
+
},
|
| 1450 |
+
{
|
| 1451 |
+
"epoch": 16.463251670378618,
|
| 1452 |
+
"grad_norm": 42.146873474121094,
|
| 1453 |
+
"learning_rate": 3.7301587301587305e-06,
|
| 1454 |
+
"loss": 0.4773,
|
| 1455 |
+
"step": 1860
|
| 1456 |
+
},
|
| 1457 |
+
{
|
| 1458 |
+
"epoch": 16.552338530066816,
|
| 1459 |
+
"grad_norm": 21.7828426361084,
|
| 1460 |
+
"learning_rate": 3.690476190476191e-06,
|
| 1461 |
+
"loss": 0.3173,
|
| 1462 |
+
"step": 1870
|
| 1463 |
+
},
|
| 1464 |
+
{
|
| 1465 |
+
"epoch": 16.64142538975501,
|
| 1466 |
+
"grad_norm": 43.677886962890625,
|
| 1467 |
+
"learning_rate": 3.6507936507936507e-06,
|
| 1468 |
+
"loss": 0.2824,
|
| 1469 |
+
"step": 1880
|
| 1470 |
+
},
|
| 1471 |
+
{
|
| 1472 |
+
"epoch": 16.73051224944321,
|
| 1473 |
+
"grad_norm": 24.393888473510742,
|
| 1474 |
+
"learning_rate": 3.6111111111111115e-06,
|
| 1475 |
+
"loss": 0.2087,
|
| 1476 |
+
"step": 1890
|
| 1477 |
+
},
|
| 1478 |
+
{
|
| 1479 |
+
"epoch": 16.819599109131403,
|
| 1480 |
+
"grad_norm": 281.832275390625,
|
| 1481 |
+
"learning_rate": 3.5714285714285718e-06,
|
| 1482 |
+
"loss": 0.4357,
|
| 1483 |
+
"step": 1900
|
| 1484 |
+
},
|
| 1485 |
+
{
|
| 1486 |
+
"epoch": 16.9086859688196,
|
| 1487 |
+
"grad_norm": 42.33871078491211,
|
| 1488 |
+
"learning_rate": 3.531746031746032e-06,
|
| 1489 |
+
"loss": 0.2472,
|
| 1490 |
+
"step": 1910
|
| 1491 |
+
},
|
| 1492 |
+
{
|
| 1493 |
+
"epoch": 16.997772828507795,
|
| 1494 |
+
"grad_norm": 65.45475769042969,
|
| 1495 |
+
"learning_rate": 3.492063492063492e-06,
|
| 1496 |
+
"loss": 0.478,
|
| 1497 |
+
"step": 1920
|
| 1498 |
+
},
|
| 1499 |
+
{
|
| 1500 |
+
"epoch": 17.0,
|
| 1501 |
+
"eval_loss": 0.8115331530570984,
|
| 1502 |
+
"eval_macro_f1": 0.7822360888195559,
|
| 1503 |
+
"eval_runtime": 21.7537,
|
| 1504 |
+
"eval_samples_per_second": 41.28,
|
| 1505 |
+
"eval_steps_per_second": 5.195,
|
| 1506 |
+
"step": 1921
|
| 1507 |
+
},
|
| 1508 |
+
{
|
| 1509 |
+
"epoch": 17.080178173719375,
|
| 1510 |
+
"grad_norm": 14.766924858093262,
|
| 1511 |
+
"learning_rate": 3.4523809523809528e-06,
|
| 1512 |
+
"loss": 0.3372,
|
| 1513 |
+
"step": 1930
|
| 1514 |
+
},
|
| 1515 |
+
{
|
| 1516 |
+
"epoch": 17.169265033407573,
|
| 1517 |
+
"grad_norm": 15.446084022521973,
|
| 1518 |
+
"learning_rate": 3.412698412698413e-06,
|
| 1519 |
+
"loss": 0.2074,
|
| 1520 |
+
"step": 1940
|
| 1521 |
+
},
|
| 1522 |
+
{
|
| 1523 |
+
"epoch": 17.258351893095767,
|
| 1524 |
+
"grad_norm": 61.16306686401367,
|
| 1525 |
+
"learning_rate": 3.3730158730158734e-06,
|
| 1526 |
+
"loss": 0.1885,
|
| 1527 |
+
"step": 1950
|
| 1528 |
+
},
|
| 1529 |
+
{
|
| 1530 |
+
"epoch": 17.347438752783965,
|
| 1531 |
+
"grad_norm": 23.0779972076416,
|
| 1532 |
+
"learning_rate": 3.3333333333333333e-06,
|
| 1533 |
+
"loss": 0.1597,
|
| 1534 |
+
"step": 1960
|
| 1535 |
+
},
|
| 1536 |
+
{
|
| 1537 |
+
"epoch": 17.43652561247216,
|
| 1538 |
+
"grad_norm": 76.70426940917969,
|
| 1539 |
+
"learning_rate": 3.293650793650794e-06,
|
| 1540 |
+
"loss": 0.2077,
|
| 1541 |
+
"step": 1970
|
| 1542 |
+
},
|
| 1543 |
+
{
|
| 1544 |
+
"epoch": 17.525612472160358,
|
| 1545 |
+
"grad_norm": 33.99974822998047,
|
| 1546 |
+
"learning_rate": 3.2539682539682544e-06,
|
| 1547 |
+
"loss": 0.353,
|
| 1548 |
+
"step": 1980
|
| 1549 |
+
},
|
| 1550 |
+
{
|
| 1551 |
+
"epoch": 17.614699331848552,
|
| 1552 |
+
"grad_norm": 4.826251029968262,
|
| 1553 |
+
"learning_rate": 3.2142857142857147e-06,
|
| 1554 |
+
"loss": 0.3525,
|
| 1555 |
+
"step": 1990
|
| 1556 |
+
},
|
| 1557 |
+
{
|
| 1558 |
+
"epoch": 17.70378619153675,
|
| 1559 |
+
"grad_norm": 13.412989616394043,
|
| 1560 |
+
"learning_rate": 3.1746031746031746e-06,
|
| 1561 |
+
"loss": 0.2875,
|
| 1562 |
+
"step": 2000
|
| 1563 |
+
},
|
| 1564 |
+
{
|
| 1565 |
+
"epoch": 17.792873051224944,
|
| 1566 |
+
"grad_norm": 18.342254638671875,
|
| 1567 |
+
"learning_rate": 3.134920634920635e-06,
|
| 1568 |
+
"loss": 0.4062,
|
| 1569 |
+
"step": 2010
|
| 1570 |
+
},
|
| 1571 |
+
{
|
| 1572 |
+
"epoch": 17.88195991091314,
|
| 1573 |
+
"grad_norm": 1.9873216152191162,
|
| 1574 |
+
"learning_rate": 3.0952380952380957e-06,
|
| 1575 |
+
"loss": 0.1457,
|
| 1576 |
+
"step": 2020
|
| 1577 |
+
},
|
| 1578 |
+
{
|
| 1579 |
+
"epoch": 17.971046770601337,
|
| 1580 |
+
"grad_norm": 68.14187622070312,
|
| 1581 |
+
"learning_rate": 3.055555555555556e-06,
|
| 1582 |
+
"loss": 0.3032,
|
| 1583 |
+
"step": 2030
|
| 1584 |
+
},
|
| 1585 |
+
{
|
| 1586 |
+
"epoch": 18.0,
|
| 1587 |
+
"eval_loss": 0.9094008803367615,
|
| 1588 |
+
"eval_macro_f1": 0.7945872168335913,
|
| 1589 |
+
"eval_runtime": 22.2168,
|
| 1590 |
+
"eval_samples_per_second": 40.42,
|
| 1591 |
+
"eval_steps_per_second": 5.086,
|
| 1592 |
+
"step": 2034
|
| 1593 |
+
},
|
| 1594 |
+
{
|
| 1595 |
+
"epoch": 18.053452115812917,
|
| 1596 |
+
"grad_norm": 36.161376953125,
|
| 1597 |
+
"learning_rate": 3.015873015873016e-06,
|
| 1598 |
+
"loss": 0.2413,
|
| 1599 |
+
"step": 2040
|
| 1600 |
+
},
|
| 1601 |
+
{
|
| 1602 |
+
"epoch": 18.142538975501115,
|
| 1603 |
+
"grad_norm": 23.837491989135742,
|
| 1604 |
+
"learning_rate": 2.9761904761904763e-06,
|
| 1605 |
+
"loss": 0.5601,
|
| 1606 |
+
"step": 2050
|
| 1607 |
+
},
|
| 1608 |
+
{
|
| 1609 |
+
"epoch": 18.23162583518931,
|
| 1610 |
+
"grad_norm": 338.7473449707031,
|
| 1611 |
+
"learning_rate": 2.936507936507937e-06,
|
| 1612 |
+
"loss": 0.4942,
|
| 1613 |
+
"step": 2060
|
| 1614 |
+
},
|
| 1615 |
+
{
|
| 1616 |
+
"epoch": 18.320712694877507,
|
| 1617 |
+
"grad_norm": 114.42424774169922,
|
| 1618 |
+
"learning_rate": 2.8968253968253974e-06,
|
| 1619 |
+
"loss": 0.4017,
|
| 1620 |
+
"step": 2070
|
| 1621 |
+
},
|
| 1622 |
+
{
|
| 1623 |
+
"epoch": 18.4097995545657,
|
| 1624 |
+
"grad_norm": 56.394432067871094,
|
| 1625 |
+
"learning_rate": 2.8571428571428573e-06,
|
| 1626 |
+
"loss": 0.2997,
|
| 1627 |
+
"step": 2080
|
| 1628 |
+
},
|
| 1629 |
+
{
|
| 1630 |
+
"epoch": 18.498886414253896,
|
| 1631 |
+
"grad_norm": 38.77593994140625,
|
| 1632 |
+
"learning_rate": 2.8174603174603176e-06,
|
| 1633 |
+
"loss": 0.4311,
|
| 1634 |
+
"step": 2090
|
| 1635 |
+
},
|
| 1636 |
+
{
|
| 1637 |
+
"epoch": 18.587973273942094,
|
| 1638 |
+
"grad_norm": 32.89288330078125,
|
| 1639 |
+
"learning_rate": 2.7777777777777783e-06,
|
| 1640 |
+
"loss": 0.3228,
|
| 1641 |
+
"step": 2100
|
| 1642 |
+
},
|
| 1643 |
+
{
|
| 1644 |
+
"epoch": 18.677060133630288,
|
| 1645 |
+
"grad_norm": 17.314373016357422,
|
| 1646 |
+
"learning_rate": 2.7380952380952387e-06,
|
| 1647 |
+
"loss": 0.2455,
|
| 1648 |
+
"step": 2110
|
| 1649 |
+
},
|
| 1650 |
+
{
|
| 1651 |
+
"epoch": 18.766146993318486,
|
| 1652 |
+
"grad_norm": 79.07820892333984,
|
| 1653 |
+
"learning_rate": 2.6984126984126986e-06,
|
| 1654 |
+
"loss": 0.2307,
|
| 1655 |
+
"step": 2120
|
| 1656 |
+
},
|
| 1657 |
+
{
|
| 1658 |
+
"epoch": 18.85523385300668,
|
| 1659 |
+
"grad_norm": 61.03816223144531,
|
| 1660 |
+
"learning_rate": 2.658730158730159e-06,
|
| 1661 |
+
"loss": 0.3474,
|
| 1662 |
+
"step": 2130
|
| 1663 |
+
},
|
| 1664 |
+
{
|
| 1665 |
+
"epoch": 18.94432071269488,
|
| 1666 |
+
"grad_norm": 14.016550064086914,
|
| 1667 |
+
"learning_rate": 2.6190476190476192e-06,
|
| 1668 |
+
"loss": 0.46,
|
| 1669 |
+
"step": 2140
|
| 1670 |
+
},
|
| 1671 |
+
{
|
| 1672 |
+
"epoch": 19.0,
|
| 1673 |
+
"eval_loss": 0.8602247834205627,
|
| 1674 |
+
"eval_macro_f1": 0.7783843611181741,
|
| 1675 |
+
"eval_runtime": 23.5346,
|
| 1676 |
+
"eval_samples_per_second": 38.157,
|
| 1677 |
+
"eval_steps_per_second": 4.801,
|
| 1678 |
+
"step": 2147
|
| 1679 |
+
},
|
| 1680 |
+
{
|
| 1681 |
+
"epoch": 19.02672605790646,
|
| 1682 |
+
"grad_norm": 6.988764762878418,
|
| 1683 |
+
"learning_rate": 2.57936507936508e-06,
|
| 1684 |
+
"loss": 0.2748,
|
| 1685 |
+
"step": 2150
|
| 1686 |
+
},
|
| 1687 |
+
{
|
| 1688 |
+
"epoch": 19.115812917594656,
|
| 1689 |
+
"grad_norm": 101.83765411376953,
|
| 1690 |
+
"learning_rate": 2.53968253968254e-06,
|
| 1691 |
+
"loss": 0.2884,
|
| 1692 |
+
"step": 2160
|
| 1693 |
+
},
|
| 1694 |
+
{
|
| 1695 |
+
"epoch": 19.20489977728285,
|
| 1696 |
+
"grad_norm": 43.04463577270508,
|
| 1697 |
+
"learning_rate": 2.5e-06,
|
| 1698 |
+
"loss": 0.248,
|
| 1699 |
+
"step": 2170
|
| 1700 |
+
},
|
| 1701 |
+
{
|
| 1702 |
+
"epoch": 19.293986636971045,
|
| 1703 |
+
"grad_norm": 37.474327087402344,
|
| 1704 |
+
"learning_rate": 2.4603174603174605e-06,
|
| 1705 |
+
"loss": 0.1679,
|
| 1706 |
+
"step": 2180
|
| 1707 |
+
},
|
| 1708 |
+
{
|
| 1709 |
+
"epoch": 19.383073496659243,
|
| 1710 |
+
"grad_norm": 33.84281539916992,
|
| 1711 |
+
"learning_rate": 2.420634920634921e-06,
|
| 1712 |
+
"loss": 0.1555,
|
| 1713 |
+
"step": 2190
|
| 1714 |
+
},
|
| 1715 |
+
{
|
| 1716 |
+
"epoch": 19.472160356347437,
|
| 1717 |
+
"grad_norm": 7.630841255187988,
|
| 1718 |
+
"learning_rate": 2.380952380952381e-06,
|
| 1719 |
+
"loss": 0.2762,
|
| 1720 |
+
"step": 2200
|
| 1721 |
+
},
|
| 1722 |
+
{
|
| 1723 |
+
"epoch": 19.561247216035635,
|
| 1724 |
+
"grad_norm": 34.77897262573242,
|
| 1725 |
+
"learning_rate": 2.3412698412698415e-06,
|
| 1726 |
+
"loss": 0.3175,
|
| 1727 |
+
"step": 2210
|
| 1728 |
+
},
|
| 1729 |
+
{
|
| 1730 |
+
"epoch": 19.65033407572383,
|
| 1731 |
+
"grad_norm": 24.479206085205078,
|
| 1732 |
+
"learning_rate": 2.301587301587302e-06,
|
| 1733 |
+
"loss": 0.2666,
|
| 1734 |
+
"step": 2220
|
| 1735 |
+
},
|
| 1736 |
+
{
|
| 1737 |
+
"epoch": 19.739420935412028,
|
| 1738 |
+
"grad_norm": 54.75190734863281,
|
| 1739 |
+
"learning_rate": 2.261904761904762e-06,
|
| 1740 |
+
"loss": 0.2699,
|
| 1741 |
+
"step": 2230
|
| 1742 |
+
},
|
| 1743 |
+
{
|
| 1744 |
+
"epoch": 19.828507795100222,
|
| 1745 |
+
"grad_norm": 27.636972427368164,
|
| 1746 |
+
"learning_rate": 2.222222222222222e-06,
|
| 1747 |
+
"loss": 0.4514,
|
| 1748 |
+
"step": 2240
|
| 1749 |
+
},
|
| 1750 |
+
{
|
| 1751 |
+
"epoch": 19.91759465478842,
|
| 1752 |
+
"grad_norm": 61.61141586303711,
|
| 1753 |
+
"learning_rate": 2.182539682539683e-06,
|
| 1754 |
+
"loss": 0.4043,
|
| 1755 |
+
"step": 2250
|
| 1756 |
+
},
|
| 1757 |
+
{
|
| 1758 |
+
"epoch": 20.0,
|
| 1759 |
+
"grad_norm": 48.8805046081543,
|
| 1760 |
+
"learning_rate": 2.1428571428571427e-06,
|
| 1761 |
+
"loss": 0.1551,
|
| 1762 |
+
"step": 2260
|
| 1763 |
+
},
|
| 1764 |
+
{
|
| 1765 |
+
"epoch": 20.0,
|
| 1766 |
+
"eval_loss": 0.9069338440895081,
|
| 1767 |
+
"eval_macro_f1": 0.7952638618677043,
|
| 1768 |
+
"eval_runtime": 22.5334,
|
| 1769 |
+
"eval_samples_per_second": 39.852,
|
| 1770 |
+
"eval_steps_per_second": 5.015,
|
| 1771 |
+
"step": 2260
|
| 1772 |
+
},
|
| 1773 |
+
{
|
| 1774 |
+
"epoch": 20.089086859688194,
|
| 1775 |
+
"grad_norm": 34.62713623046875,
|
| 1776 |
+
"learning_rate": 2.1031746031746035e-06,
|
| 1777 |
+
"loss": 0.2323,
|
| 1778 |
+
"step": 2270
|
| 1779 |
+
},
|
| 1780 |
+
{
|
| 1781 |
+
"epoch": 20.178173719376392,
|
| 1782 |
+
"grad_norm": 55.7051887512207,
|
| 1783 |
+
"learning_rate": 2.0634920634920634e-06,
|
| 1784 |
+
"loss": 0.239,
|
| 1785 |
+
"step": 2280
|
| 1786 |
+
},
|
| 1787 |
+
{
|
| 1788 |
+
"epoch": 20.267260579064587,
|
| 1789 |
+
"grad_norm": 11.973965644836426,
|
| 1790 |
+
"learning_rate": 2.023809523809524e-06,
|
| 1791 |
+
"loss": 0.221,
|
| 1792 |
+
"step": 2290
|
| 1793 |
+
},
|
| 1794 |
+
{
|
| 1795 |
+
"epoch": 20.356347438752785,
|
| 1796 |
+
"grad_norm": 18.674625396728516,
|
| 1797 |
+
"learning_rate": 1.984126984126984e-06,
|
| 1798 |
+
"loss": 0.1411,
|
| 1799 |
+
"step": 2300
|
| 1800 |
+
},
|
| 1801 |
+
{
|
| 1802 |
+
"epoch": 20.44543429844098,
|
| 1803 |
+
"grad_norm": 64.24128723144531,
|
| 1804 |
+
"learning_rate": 1.944444444444445e-06,
|
| 1805 |
+
"loss": 0.3944,
|
| 1806 |
+
"step": 2310
|
| 1807 |
+
},
|
| 1808 |
+
{
|
| 1809 |
+
"epoch": 20.534521158129177,
|
| 1810 |
+
"grad_norm": 19.434814453125,
|
| 1811 |
+
"learning_rate": 1.904761904761905e-06,
|
| 1812 |
+
"loss": 0.3017,
|
| 1813 |
+
"step": 2320
|
| 1814 |
+
},
|
| 1815 |
+
{
|
| 1816 |
+
"epoch": 20.62360801781737,
|
| 1817 |
+
"grad_norm": 76.87932586669922,
|
| 1818 |
+
"learning_rate": 1.8650793650793652e-06,
|
| 1819 |
+
"loss": 0.2937,
|
| 1820 |
+
"step": 2330
|
| 1821 |
+
},
|
| 1822 |
+
{
|
| 1823 |
+
"epoch": 20.71269487750557,
|
| 1824 |
+
"grad_norm": 79.9623794555664,
|
| 1825 |
+
"learning_rate": 1.8253968253968254e-06,
|
| 1826 |
+
"loss": 0.162,
|
| 1827 |
+
"step": 2340
|
| 1828 |
+
},
|
| 1829 |
+
{
|
| 1830 |
+
"epoch": 20.801781737193764,
|
| 1831 |
+
"grad_norm": 0.5564860105514526,
|
| 1832 |
+
"learning_rate": 1.7857142857142859e-06,
|
| 1833 |
+
"loss": 0.1174,
|
| 1834 |
+
"step": 2350
|
| 1835 |
+
},
|
| 1836 |
+
{
|
| 1837 |
+
"epoch": 20.89086859688196,
|
| 1838 |
+
"grad_norm": 63.84634017944336,
|
| 1839 |
+
"learning_rate": 1.746031746031746e-06,
|
| 1840 |
+
"loss": 0.3684,
|
| 1841 |
+
"step": 2360
|
| 1842 |
+
},
|
| 1843 |
+
{
|
| 1844 |
+
"epoch": 20.979955456570156,
|
| 1845 |
+
"grad_norm": 159.61268615722656,
|
| 1846 |
+
"learning_rate": 1.7063492063492065e-06,
|
| 1847 |
+
"loss": 0.3583,
|
| 1848 |
+
"step": 2370
|
| 1849 |
+
},
|
| 1850 |
+
{
|
| 1851 |
+
"epoch": 21.0,
|
| 1852 |
+
"eval_loss": 0.9667259454727173,
|
| 1853 |
+
"eval_macro_f1": 0.7853648934851267,
|
| 1854 |
+
"eval_runtime": 21.6042,
|
| 1855 |
+
"eval_samples_per_second": 41.566,
|
| 1856 |
+
"eval_steps_per_second": 5.23,
|
| 1857 |
+
"step": 2373
|
| 1858 |
+
},
|
| 1859 |
+
{
|
| 1860 |
+
"epoch": 21.062360801781736,
|
| 1861 |
+
"grad_norm": 3.4698734283447266,
|
| 1862 |
+
"learning_rate": 1.6666666666666667e-06,
|
| 1863 |
+
"loss": 0.0825,
|
| 1864 |
+
"step": 2380
|
| 1865 |
+
},
|
| 1866 |
+
{
|
| 1867 |
+
"epoch": 21.151447661469934,
|
| 1868 |
+
"grad_norm": 30.75046730041504,
|
| 1869 |
+
"learning_rate": 1.6269841269841272e-06,
|
| 1870 |
+
"loss": 0.2367,
|
| 1871 |
+
"step": 2390
|
| 1872 |
+
},
|
| 1873 |
+
{
|
| 1874 |
+
"epoch": 21.24053452115813,
|
| 1875 |
+
"grad_norm": 58.869632720947266,
|
| 1876 |
+
"learning_rate": 1.5873015873015873e-06,
|
| 1877 |
+
"loss": 0.2491,
|
| 1878 |
+
"step": 2400
|
| 1879 |
+
},
|
| 1880 |
+
{
|
| 1881 |
+
"epoch": 21.329621380846326,
|
| 1882 |
+
"grad_norm": 78.80653381347656,
|
| 1883 |
+
"learning_rate": 1.5476190476190479e-06,
|
| 1884 |
+
"loss": 0.3368,
|
| 1885 |
+
"step": 2410
|
| 1886 |
+
},
|
| 1887 |
+
{
|
| 1888 |
+
"epoch": 21.41870824053452,
|
| 1889 |
+
"grad_norm": 42.70677947998047,
|
| 1890 |
+
"learning_rate": 1.507936507936508e-06,
|
| 1891 |
+
"loss": 0.2251,
|
| 1892 |
+
"step": 2420
|
| 1893 |
+
},
|
| 1894 |
+
{
|
| 1895 |
+
"epoch": 21.50779510022272,
|
| 1896 |
+
"grad_norm": 24.363916397094727,
|
| 1897 |
+
"learning_rate": 1.4682539682539685e-06,
|
| 1898 |
+
"loss": 0.154,
|
| 1899 |
+
"step": 2430
|
| 1900 |
+
},
|
| 1901 |
+
{
|
| 1902 |
+
"epoch": 21.596881959910913,
|
| 1903 |
+
"grad_norm": 70.08970642089844,
|
| 1904 |
+
"learning_rate": 1.4285714285714286e-06,
|
| 1905 |
+
"loss": 0.3436,
|
| 1906 |
+
"step": 2440
|
| 1907 |
+
},
|
| 1908 |
+
{
|
| 1909 |
+
"epoch": 21.685968819599108,
|
| 1910 |
+
"grad_norm": 59.4857292175293,
|
| 1911 |
+
"learning_rate": 1.3888888888888892e-06,
|
| 1912 |
+
"loss": 0.2631,
|
| 1913 |
+
"step": 2450
|
| 1914 |
+
},
|
| 1915 |
+
{
|
| 1916 |
+
"epoch": 21.775055679287306,
|
| 1917 |
+
"grad_norm": 2.151820421218872,
|
| 1918 |
+
"learning_rate": 1.3492063492063493e-06,
|
| 1919 |
+
"loss": 0.2338,
|
| 1920 |
+
"step": 2460
|
| 1921 |
+
},
|
| 1922 |
+
{
|
| 1923 |
+
"epoch": 21.8641425389755,
|
| 1924 |
+
"grad_norm": 32.367698669433594,
|
| 1925 |
+
"learning_rate": 1.3095238095238096e-06,
|
| 1926 |
+
"loss": 0.3898,
|
| 1927 |
+
"step": 2470
|
| 1928 |
+
},
|
| 1929 |
+
{
|
| 1930 |
+
"epoch": 21.953229398663698,
|
| 1931 |
+
"grad_norm": 62.534358978271484,
|
| 1932 |
+
"learning_rate": 1.26984126984127e-06,
|
| 1933 |
+
"loss": 0.3501,
|
| 1934 |
+
"step": 2480
|
| 1935 |
+
},
|
| 1936 |
+
{
|
| 1937 |
+
"epoch": 22.0,
|
| 1938 |
+
"eval_loss": 1.012130618095398,
|
| 1939 |
+
"eval_macro_f1": 0.7793243771478484,
|
| 1940 |
+
"eval_runtime": 21.887,
|
| 1941 |
+
"eval_samples_per_second": 41.029,
|
| 1942 |
+
"eval_steps_per_second": 5.163,
|
| 1943 |
+
"step": 2486
|
| 1944 |
+
},
|
| 1945 |
+
{
|
| 1946 |
+
"epoch": 22.035634743875278,
|
| 1947 |
+
"grad_norm": 69.29458618164062,
|
| 1948 |
+
"learning_rate": 1.2301587301587303e-06,
|
| 1949 |
+
"loss": 0.2882,
|
| 1950 |
+
"step": 2490
|
| 1951 |
+
},
|
| 1952 |
+
{
|
| 1953 |
+
"epoch": 22.124721603563476,
|
| 1954 |
+
"grad_norm": 89.39969635009766,
|
| 1955 |
+
"learning_rate": 1.1904761904761906e-06,
|
| 1956 |
+
"loss": 0.2838,
|
| 1957 |
+
"step": 2500
|
| 1958 |
+
},
|
| 1959 |
+
{
|
| 1960 |
+
"epoch": 22.21380846325167,
|
| 1961 |
+
"grad_norm": 356.38787841796875,
|
| 1962 |
+
"learning_rate": 1.150793650793651e-06,
|
| 1963 |
+
"loss": 0.2001,
|
| 1964 |
+
"step": 2510
|
| 1965 |
+
},
|
| 1966 |
+
{
|
| 1967 |
+
"epoch": 22.302895322939868,
|
| 1968 |
+
"grad_norm": 4.26671028137207,
|
| 1969 |
+
"learning_rate": 1.111111111111111e-06,
|
| 1970 |
+
"loss": 0.1658,
|
| 1971 |
+
"step": 2520
|
| 1972 |
+
},
|
| 1973 |
+
{
|
| 1974 |
+
"epoch": 22.391982182628063,
|
| 1975 |
+
"grad_norm": 70.33140563964844,
|
| 1976 |
+
"learning_rate": 1.0714285714285714e-06,
|
| 1977 |
+
"loss": 0.2195,
|
| 1978 |
+
"step": 2530
|
| 1979 |
+
},
|
| 1980 |
+
{
|
| 1981 |
+
"epoch": 22.481069042316257,
|
| 1982 |
+
"grad_norm": 57.502559661865234,
|
| 1983 |
+
"learning_rate": 1.0317460317460317e-06,
|
| 1984 |
+
"loss": 0.1085,
|
| 1985 |
+
"step": 2540
|
| 1986 |
+
},
|
| 1987 |
+
{
|
| 1988 |
+
"epoch": 22.570155902004455,
|
| 1989 |
+
"grad_norm": 76.61205291748047,
|
| 1990 |
+
"learning_rate": 9.92063492063492e-07,
|
| 1991 |
+
"loss": 0.1872,
|
| 1992 |
+
"step": 2550
|
| 1993 |
+
},
|
| 1994 |
+
{
|
| 1995 |
+
"epoch": 22.65924276169265,
|
| 1996 |
+
"grad_norm": 72.69236755371094,
|
| 1997 |
+
"learning_rate": 9.523809523809525e-07,
|
| 1998 |
+
"loss": 0.2837,
|
| 1999 |
+
"step": 2560
|
| 2000 |
+
},
|
| 2001 |
+
{
|
| 2002 |
+
"epoch": 22.748329621380847,
|
| 2003 |
+
"grad_norm": 112.57559967041016,
|
| 2004 |
+
"learning_rate": 9.126984126984127e-07,
|
| 2005 |
+
"loss": 0.4063,
|
| 2006 |
+
"step": 2570
|
| 2007 |
+
},
|
| 2008 |
+
{
|
| 2009 |
+
"epoch": 22.83741648106904,
|
| 2010 |
+
"grad_norm": 102.09916687011719,
|
| 2011 |
+
"learning_rate": 8.73015873015873e-07,
|
| 2012 |
+
"loss": 0.3085,
|
| 2013 |
+
"step": 2580
|
| 2014 |
+
},
|
| 2015 |
+
{
|
| 2016 |
+
"epoch": 22.92650334075724,
|
| 2017 |
+
"grad_norm": 4.140219211578369,
|
| 2018 |
+
"learning_rate": 8.333333333333333e-07,
|
| 2019 |
+
"loss": 0.0432,
|
| 2020 |
+
"step": 2590
|
| 2021 |
+
},
|
| 2022 |
+
{
|
| 2023 |
+
"epoch": 23.0,
|
| 2024 |
+
"eval_loss": 0.9713873863220215,
|
| 2025 |
+
"eval_macro_f1": 0.7901658291457286,
|
| 2026 |
+
"eval_runtime": 21.8489,
|
| 2027 |
+
"eval_samples_per_second": 41.1,
|
| 2028 |
+
"eval_steps_per_second": 5.172,
|
| 2029 |
+
"step": 2599
|
| 2030 |
+
},
|
| 2031 |
+
{
|
| 2032 |
+
"epoch": 23.00890868596882,
|
| 2033 |
+
"grad_norm": 81.31112670898438,
|
| 2034 |
+
"learning_rate": 7.936507936507937e-07,
|
| 2035 |
+
"loss": 0.2301,
|
| 2036 |
+
"step": 2600
|
| 2037 |
+
},
|
| 2038 |
+
{
|
| 2039 |
+
"epoch": 23.097995545657014,
|
| 2040 |
+
"grad_norm": 133.33106994628906,
|
| 2041 |
+
"learning_rate": 7.53968253968254e-07,
|
| 2042 |
+
"loss": 0.3902,
|
| 2043 |
+
"step": 2610
|
| 2044 |
+
},
|
| 2045 |
+
{
|
| 2046 |
+
"epoch": 23.187082405345212,
|
| 2047 |
+
"grad_norm": 75.14102935791016,
|
| 2048 |
+
"learning_rate": 7.142857142857143e-07,
|
| 2049 |
+
"loss": 0.4124,
|
| 2050 |
+
"step": 2620
|
| 2051 |
+
},
|
| 2052 |
+
{
|
| 2053 |
+
"epoch": 23.276169265033406,
|
| 2054 |
+
"grad_norm": 56.319427490234375,
|
| 2055 |
+
"learning_rate": 6.746031746031746e-07,
|
| 2056 |
+
"loss": 0.2647,
|
| 2057 |
+
"step": 2630
|
| 2058 |
+
},
|
| 2059 |
+
{
|
| 2060 |
+
"epoch": 23.365256124721604,
|
| 2061 |
+
"grad_norm": 0.3156695067882538,
|
| 2062 |
+
"learning_rate": 6.34920634920635e-07,
|
| 2063 |
+
"loss": 0.2229,
|
| 2064 |
+
"step": 2640
|
| 2065 |
+
},
|
| 2066 |
+
{
|
| 2067 |
+
"epoch": 23.4543429844098,
|
| 2068 |
+
"grad_norm": 0.1939859390258789,
|
| 2069 |
+
"learning_rate": 5.952380952380953e-07,
|
| 2070 |
+
"loss": 0.11,
|
| 2071 |
+
"step": 2650
|
| 2072 |
+
},
|
| 2073 |
+
{
|
| 2074 |
+
"epoch": 23.543429844097997,
|
| 2075 |
+
"grad_norm": 27.532583236694336,
|
| 2076 |
+
"learning_rate": 5.555555555555555e-07,
|
| 2077 |
+
"loss": 0.269,
|
| 2078 |
+
"step": 2660
|
| 2079 |
+
},
|
| 2080 |
+
{
|
| 2081 |
+
"epoch": 23.63251670378619,
|
| 2082 |
+
"grad_norm": 5.016575336456299,
|
| 2083 |
+
"learning_rate": 5.158730158730158e-07,
|
| 2084 |
+
"loss": 0.2089,
|
| 2085 |
+
"step": 2670
|
| 2086 |
+
},
|
| 2087 |
+
{
|
| 2088 |
+
"epoch": 23.72160356347439,
|
| 2089 |
+
"grad_norm": 82.43734741210938,
|
| 2090 |
+
"learning_rate": 4.7619047619047623e-07,
|
| 2091 |
+
"loss": 0.3519,
|
| 2092 |
+
"step": 2680
|
| 2093 |
+
},
|
| 2094 |
+
{
|
| 2095 |
+
"epoch": 23.810690423162583,
|
| 2096 |
+
"grad_norm": 12.222328186035156,
|
| 2097 |
+
"learning_rate": 4.365079365079365e-07,
|
| 2098 |
+
"loss": 0.2363,
|
| 2099 |
+
"step": 2690
|
| 2100 |
+
},
|
| 2101 |
+
{
|
| 2102 |
+
"epoch": 23.899777282850778,
|
| 2103 |
+
"grad_norm": 56.74919509887695,
|
| 2104 |
+
"learning_rate": 3.9682539682539683e-07,
|
| 2105 |
+
"loss": 0.0908,
|
| 2106 |
+
"step": 2700
|
| 2107 |
+
},
|
| 2108 |
+
{
|
| 2109 |
+
"epoch": 23.988864142538976,
|
| 2110 |
+
"grad_norm": 77.4910659790039,
|
| 2111 |
+
"learning_rate": 3.5714285714285716e-07,
|
| 2112 |
+
"loss": 0.3777,
|
| 2113 |
+
"step": 2710
|
| 2114 |
+
},
|
| 2115 |
+
{
|
| 2116 |
+
"epoch": 24.0,
|
| 2117 |
+
"eval_loss": 0.9040295481681824,
|
| 2118 |
+
"eval_macro_f1": 0.7973446094765495,
|
| 2119 |
+
"eval_runtime": 21.8468,
|
| 2120 |
+
"eval_samples_per_second": 41.104,
|
| 2121 |
+
"eval_steps_per_second": 5.172,
|
| 2122 |
+
"step": 2712
|
| 2123 |
+
},
|
| 2124 |
+
{
|
| 2125 |
+
"epoch": 24.071269487750556,
|
| 2126 |
+
"grad_norm": 23.00518798828125,
|
| 2127 |
+
"learning_rate": 3.174603174603175e-07,
|
| 2128 |
+
"loss": 0.1797,
|
| 2129 |
+
"step": 2720
|
| 2130 |
+
},
|
| 2131 |
+
{
|
| 2132 |
+
"epoch": 24.160356347438753,
|
| 2133 |
+
"grad_norm": 81.12632751464844,
|
| 2134 |
+
"learning_rate": 2.7777777777777776e-07,
|
| 2135 |
+
"loss": 0.2499,
|
| 2136 |
+
"step": 2730
|
| 2137 |
+
},
|
| 2138 |
+
{
|
| 2139 |
+
"epoch": 24.249443207126948,
|
| 2140 |
+
"grad_norm": 124.60951232910156,
|
| 2141 |
+
"learning_rate": 2.3809523809523811e-07,
|
| 2142 |
+
"loss": 0.2462,
|
| 2143 |
+
"step": 2740
|
| 2144 |
+
},
|
| 2145 |
+
{
|
| 2146 |
+
"epoch": 24.338530066815146,
|
| 2147 |
+
"grad_norm": 90.49283599853516,
|
| 2148 |
+
"learning_rate": 1.9841269841269841e-07,
|
| 2149 |
+
"loss": 0.3619,
|
| 2150 |
+
"step": 2750
|
| 2151 |
+
},
|
| 2152 |
+
{
|
| 2153 |
+
"epoch": 24.42761692650334,
|
| 2154 |
+
"grad_norm": 4.63253927230835,
|
| 2155 |
+
"learning_rate": 1.5873015873015874e-07,
|
| 2156 |
+
"loss": 0.2901,
|
| 2157 |
+
"step": 2760
|
| 2158 |
+
},
|
| 2159 |
+
{
|
| 2160 |
+
"epoch": 24.51670378619154,
|
| 2161 |
+
"grad_norm": 8.05462646484375,
|
| 2162 |
+
"learning_rate": 1.1904761904761906e-07,
|
| 2163 |
+
"loss": 0.1431,
|
| 2164 |
+
"step": 2770
|
| 2165 |
+
},
|
| 2166 |
+
{
|
| 2167 |
+
"epoch": 24.605790645879733,
|
| 2168 |
+
"grad_norm": 10.225213050842285,
|
| 2169 |
+
"learning_rate": 7.936507936507937e-08,
|
| 2170 |
+
"loss": 0.1209,
|
| 2171 |
+
"step": 2780
|
| 2172 |
+
},
|
| 2173 |
+
{
|
| 2174 |
+
"epoch": 24.694877505567927,
|
| 2175 |
+
"grad_norm": 23.83371925354004,
|
| 2176 |
+
"learning_rate": 3.9682539682539686e-08,
|
| 2177 |
+
"loss": 0.2688,
|
| 2178 |
+
"step": 2790
|
| 2179 |
+
},
|
| 2180 |
+
{
|
| 2181 |
+
"epoch": 24.783964365256125,
|
| 2182 |
+
"grad_norm": 0.8638786673545837,
|
| 2183 |
+
"learning_rate": 0.0,
|
| 2184 |
+
"loss": 0.1488,
|
| 2185 |
+
"step": 2800
|
| 2186 |
+
},
|
| 2187 |
+
{
|
| 2188 |
+
"epoch": 24.783964365256125,
|
| 2189 |
+
"eval_loss": 0.9561133980751038,
|
| 2190 |
+
"eval_macro_f1": 0.8017572519758047,
|
| 2191 |
+
"eval_runtime": 22.04,
|
| 2192 |
+
"eval_samples_per_second": 40.744,
|
| 2193 |
+
"eval_steps_per_second": 5.127,
|
| 2194 |
+
"step": 2800
|
| 2195 |
+
}
|
| 2196 |
+
],
|
| 2197 |
+
"logging_steps": 10,
|
| 2198 |
+
"max_steps": 2800,
|
| 2199 |
+
"num_input_tokens_seen": 0,
|
| 2200 |
+
"num_train_epochs": 25,
|
| 2201 |
+
"save_steps": 500,
|
| 2202 |
+
"stateful_callbacks": {
|
| 2203 |
+
"TrainerControl": {
|
| 2204 |
+
"args": {
|
| 2205 |
+
"should_epoch_stop": false,
|
| 2206 |
+
"should_evaluate": false,
|
| 2207 |
+
"should_log": false,
|
| 2208 |
+
"should_save": true,
|
| 2209 |
+
"should_training_stop": true
|
| 2210 |
+
},
|
| 2211 |
+
"attributes": {}
|
| 2212 |
+
}
|
| 2213 |
+
},
|
| 2214 |
+
"total_flos": 3.1947072291024667e+19,
|
| 2215 |
+
"train_batch_size": 8,
|
| 2216 |
+
"trial_name": null,
|
| 2217 |
+
"trial_params": null
|
| 2218 |
+
}
|
training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:a2aa91e9fb5f6fb80530bd2f0be3b57b5dc44f26b993ccf79d89846501db4b5a
|
| 3 |
+
size 5304
|