Upload 7 files
Browse files- config.json +48 -0
- model.safetensors +3 -0
- optimizer.pt +3 -0
- rng_state.pth +3 -0
- scheduler.pt +3 -0
- trainer_state.json +1706 -0
- training_args.bin +3 -0
config.json
ADDED
|
@@ -0,0 +1,48 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"activation": "gelu",
|
| 3 |
+
"architectures": [
|
| 4 |
+
"DistilBertForSequenceClassification"
|
| 5 |
+
],
|
| 6 |
+
"attention_dropout": 0.1,
|
| 7 |
+
"dim": 768,
|
| 8 |
+
"dropout": 0.1,
|
| 9 |
+
"hidden_dim": 3072,
|
| 10 |
+
"id2label": {
|
| 11 |
+
"0": "LABEL_0",
|
| 12 |
+
"1": "LABEL_1",
|
| 13 |
+
"2": "LABEL_2",
|
| 14 |
+
"3": "LABEL_3",
|
| 15 |
+
"4": "LABEL_4",
|
| 16 |
+
"5": "LABEL_5",
|
| 17 |
+
"6": "LABEL_6",
|
| 18 |
+
"7": "LABEL_7",
|
| 19 |
+
"8": "LABEL_8",
|
| 20 |
+
"9": "LABEL_9"
|
| 21 |
+
},
|
| 22 |
+
"initializer_range": 0.02,
|
| 23 |
+
"label2id": {
|
| 24 |
+
"LABEL_0": 0,
|
| 25 |
+
"LABEL_1": 1,
|
| 26 |
+
"LABEL_2": 2,
|
| 27 |
+
"LABEL_3": 3,
|
| 28 |
+
"LABEL_4": 4,
|
| 29 |
+
"LABEL_5": 5,
|
| 30 |
+
"LABEL_6": 6,
|
| 31 |
+
"LABEL_7": 7,
|
| 32 |
+
"LABEL_8": 8,
|
| 33 |
+
"LABEL_9": 9
|
| 34 |
+
},
|
| 35 |
+
"max_position_embeddings": 512,
|
| 36 |
+
"model_type": "distilbert",
|
| 37 |
+
"n_heads": 12,
|
| 38 |
+
"n_layers": 6,
|
| 39 |
+
"pad_token_id": 0,
|
| 40 |
+
"problem_type": "single_label_classification",
|
| 41 |
+
"qa_dropout": 0.1,
|
| 42 |
+
"seq_classif_dropout": 0.2,
|
| 43 |
+
"sinusoidal_pos_embds": false,
|
| 44 |
+
"tie_weights_": true,
|
| 45 |
+
"torch_dtype": "float32",
|
| 46 |
+
"transformers_version": "4.54.1",
|
| 47 |
+
"vocab_size": 30522
|
| 48 |
+
}
|
model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:86a085b705b5b2ff88683d30b97e5178eb10e9ea686cb98c59fae70ef0a23c93
|
| 3 |
+
size 267857176
|
optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:b7189beebcd45da3c166eec5a01989f12fbdfd3971a519058de96824ddeb2705
|
| 3 |
+
size 535774027
|
rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:e25ac78d752ae6f1b0d38c573074ffd0b857c15289e95f8842c7cd38bbf2676e
|
| 3 |
+
size 14455
|
scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:34b4735aea3a9846304ee3b81cccc8ad3d5595d8cabad14c57723aca911f56ad
|
| 3 |
+
size 1465
|
trainer_state.json
ADDED
|
@@ -0,0 +1,1706 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": 8800,
|
| 3 |
+
"best_metric": 0.8558670305135202,
|
| 4 |
+
"best_model_checkpoint": "./distilbert-feedback/checkpoint-8800",
|
| 5 |
+
"epoch": 2.0,
|
| 6 |
+
"eval_steps": 200,
|
| 7 |
+
"global_step": 9070,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.011025358324145534,
|
| 14 |
+
"grad_norm": 6.280033111572266,
|
| 15 |
+
"learning_rate": 4.972987872105844e-05,
|
| 16 |
+
"loss": 1.7001,
|
| 17 |
+
"step": 50
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.022050716648291068,
|
| 21 |
+
"grad_norm": 6.537757873535156,
|
| 22 |
+
"learning_rate": 4.94542447629548e-05,
|
| 23 |
+
"loss": 1.3668,
|
| 24 |
+
"step": 100
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.03307607497243661,
|
| 28 |
+
"grad_norm": 6.568482875823975,
|
| 29 |
+
"learning_rate": 4.917861080485116e-05,
|
| 30 |
+
"loss": 1.2012,
|
| 31 |
+
"step": 150
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.044101433296582136,
|
| 35 |
+
"grad_norm": 7.553868293762207,
|
| 36 |
+
"learning_rate": 4.890297684674752e-05,
|
| 37 |
+
"loss": 1.2455,
|
| 38 |
+
"step": 200
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.044101433296582136,
|
| 42 |
+
"eval_accuracy": 0.5194740759116845,
|
| 43 |
+
"eval_loss": 1.2424163818359375,
|
| 44 |
+
"eval_runtime": 44.855,
|
| 45 |
+
"eval_samples_per_second": 89.867,
|
| 46 |
+
"eval_steps_per_second": 11.236,
|
| 47 |
+
"step": 200
|
| 48 |
+
},
|
| 49 |
+
{
|
| 50 |
+
"epoch": 0.05512679162072767,
|
| 51 |
+
"grad_norm": 9.827457427978516,
|
| 52 |
+
"learning_rate": 4.8627342888643884e-05,
|
| 53 |
+
"loss": 1.1816,
|
| 54 |
+
"step": 250
|
| 55 |
+
},
|
| 56 |
+
{
|
| 57 |
+
"epoch": 0.06615214994487321,
|
| 58 |
+
"grad_norm": 9.297719955444336,
|
| 59 |
+
"learning_rate": 4.835170893054025e-05,
|
| 60 |
+
"loss": 1.1326,
|
| 61 |
+
"step": 300
|
| 62 |
+
},
|
| 63 |
+
{
|
| 64 |
+
"epoch": 0.07717750826901874,
|
| 65 |
+
"grad_norm": 7.382741451263428,
|
| 66 |
+
"learning_rate": 4.807607497243661e-05,
|
| 67 |
+
"loss": 1.061,
|
| 68 |
+
"step": 350
|
| 69 |
+
},
|
| 70 |
+
{
|
| 71 |
+
"epoch": 0.08820286659316427,
|
| 72 |
+
"grad_norm": 7.543405532836914,
|
| 73 |
+
"learning_rate": 4.780044101433297e-05,
|
| 74 |
+
"loss": 1.0238,
|
| 75 |
+
"step": 400
|
| 76 |
+
},
|
| 77 |
+
{
|
| 78 |
+
"epoch": 0.08820286659316427,
|
| 79 |
+
"eval_accuracy": 0.6350781443810469,
|
| 80 |
+
"eval_loss": 0.9708635210990906,
|
| 81 |
+
"eval_runtime": 45.4434,
|
| 82 |
+
"eval_samples_per_second": 88.704,
|
| 83 |
+
"eval_steps_per_second": 11.091,
|
| 84 |
+
"step": 400
|
| 85 |
+
},
|
| 86 |
+
{
|
| 87 |
+
"epoch": 0.09922822491730982,
|
| 88 |
+
"grad_norm": 21.24835968017578,
|
| 89 |
+
"learning_rate": 4.752480705622933e-05,
|
| 90 |
+
"loss": 0.9726,
|
| 91 |
+
"step": 450
|
| 92 |
+
},
|
| 93 |
+
{
|
| 94 |
+
"epoch": 0.11025358324145534,
|
| 95 |
+
"grad_norm": 11.24039077758789,
|
| 96 |
+
"learning_rate": 4.724917309812569e-05,
|
| 97 |
+
"loss": 1.0019,
|
| 98 |
+
"step": 500
|
| 99 |
+
},
|
| 100 |
+
{
|
| 101 |
+
"epoch": 0.12127894156560089,
|
| 102 |
+
"grad_norm": 14.267502784729004,
|
| 103 |
+
"learning_rate": 4.6973539140022054e-05,
|
| 104 |
+
"loss": 0.9928,
|
| 105 |
+
"step": 550
|
| 106 |
+
},
|
| 107 |
+
{
|
| 108 |
+
"epoch": 0.13230429988974643,
|
| 109 |
+
"grad_norm": 11.773629188537598,
|
| 110 |
+
"learning_rate": 4.669790518191842e-05,
|
| 111 |
+
"loss": 0.8424,
|
| 112 |
+
"step": 600
|
| 113 |
+
},
|
| 114 |
+
{
|
| 115 |
+
"epoch": 0.13230429988974643,
|
| 116 |
+
"eval_accuracy": 0.6789878442073927,
|
| 117 |
+
"eval_loss": 0.8846617937088013,
|
| 118 |
+
"eval_runtime": 45.2393,
|
| 119 |
+
"eval_samples_per_second": 89.104,
|
| 120 |
+
"eval_steps_per_second": 11.141,
|
| 121 |
+
"step": 600
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.14332965821389196,
|
| 125 |
+
"grad_norm": 4.492119312286377,
|
| 126 |
+
"learning_rate": 4.642227122381478e-05,
|
| 127 |
+
"loss": 0.9987,
|
| 128 |
+
"step": 650
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.1543550165380375,
|
| 132 |
+
"grad_norm": 8.530481338500977,
|
| 133 |
+
"learning_rate": 4.614663726571114e-05,
|
| 134 |
+
"loss": 0.8802,
|
| 135 |
+
"step": 700
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.16538037486218302,
|
| 139 |
+
"grad_norm": 15.022302627563477,
|
| 140 |
+
"learning_rate": 4.58710033076075e-05,
|
| 141 |
+
"loss": 0.8428,
|
| 142 |
+
"step": 750
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.17640573318632854,
|
| 146 |
+
"grad_norm": 11.992727279663086,
|
| 147 |
+
"learning_rate": 4.559536934950386e-05,
|
| 148 |
+
"loss": 0.9022,
|
| 149 |
+
"step": 800
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.17640573318632854,
|
| 153 |
+
"eval_accuracy": 0.6566608781939965,
|
| 154 |
+
"eval_loss": 0.9172375202178955,
|
| 155 |
+
"eval_runtime": 46.9407,
|
| 156 |
+
"eval_samples_per_second": 85.874,
|
| 157 |
+
"eval_steps_per_second": 10.737,
|
| 158 |
+
"step": 800
|
| 159 |
+
},
|
| 160 |
+
{
|
| 161 |
+
"epoch": 0.1874310915104741,
|
| 162 |
+
"grad_norm": 10.07384204864502,
|
| 163 |
+
"learning_rate": 4.5319735391400225e-05,
|
| 164 |
+
"loss": 0.8949,
|
| 165 |
+
"step": 850
|
| 166 |
+
},
|
| 167 |
+
{
|
| 168 |
+
"epoch": 0.19845644983461963,
|
| 169 |
+
"grad_norm": 7.739764213562012,
|
| 170 |
+
"learning_rate": 4.504410143329658e-05,
|
| 171 |
+
"loss": 0.8179,
|
| 172 |
+
"step": 900
|
| 173 |
+
},
|
| 174 |
+
{
|
| 175 |
+
"epoch": 0.20948180815876516,
|
| 176 |
+
"grad_norm": 13.591370582580566,
|
| 177 |
+
"learning_rate": 4.4768467475192944e-05,
|
| 178 |
+
"loss": 0.8201,
|
| 179 |
+
"step": 950
|
| 180 |
+
},
|
| 181 |
+
{
|
| 182 |
+
"epoch": 0.2205071664829107,
|
| 183 |
+
"grad_norm": 10.648216247558594,
|
| 184 |
+
"learning_rate": 4.449283351708931e-05,
|
| 185 |
+
"loss": 0.8128,
|
| 186 |
+
"step": 1000
|
| 187 |
+
},
|
| 188 |
+
{
|
| 189 |
+
"epoch": 0.2205071664829107,
|
| 190 |
+
"eval_accuracy": 0.7142148350285289,
|
| 191 |
+
"eval_loss": 0.7965238690376282,
|
| 192 |
+
"eval_runtime": 48.533,
|
| 193 |
+
"eval_samples_per_second": 83.057,
|
| 194 |
+
"eval_steps_per_second": 10.385,
|
| 195 |
+
"step": 1000
|
| 196 |
+
},
|
| 197 |
+
{
|
| 198 |
+
"epoch": 0.23153252480705622,
|
| 199 |
+
"grad_norm": 8.068305015563965,
|
| 200 |
+
"learning_rate": 4.421719955898567e-05,
|
| 201 |
+
"loss": 0.8249,
|
| 202 |
+
"step": 1050
|
| 203 |
+
},
|
| 204 |
+
{
|
| 205 |
+
"epoch": 0.24255788313120177,
|
| 206 |
+
"grad_norm": 6.480292797088623,
|
| 207 |
+
"learning_rate": 4.3941565600882026e-05,
|
| 208 |
+
"loss": 0.7703,
|
| 209 |
+
"step": 1100
|
| 210 |
+
},
|
| 211 |
+
{
|
| 212 |
+
"epoch": 0.2535832414553473,
|
| 213 |
+
"grad_norm": 7.40410041809082,
|
| 214 |
+
"learning_rate": 4.366593164277839e-05,
|
| 215 |
+
"loss": 0.8165,
|
| 216 |
+
"step": 1150
|
| 217 |
+
},
|
| 218 |
+
{
|
| 219 |
+
"epoch": 0.26460859977949286,
|
| 220 |
+
"grad_norm": 13.612632751464844,
|
| 221 |
+
"learning_rate": 4.339029768467475e-05,
|
| 222 |
+
"loss": 0.8101,
|
| 223 |
+
"step": 1200
|
| 224 |
+
},
|
| 225 |
+
{
|
| 226 |
+
"epoch": 0.26460859977949286,
|
| 227 |
+
"eval_accuracy": 0.6998263458198958,
|
| 228 |
+
"eval_loss": 0.7793557643890381,
|
| 229 |
+
"eval_runtime": 47.2524,
|
| 230 |
+
"eval_samples_per_second": 85.308,
|
| 231 |
+
"eval_steps_per_second": 10.666,
|
| 232 |
+
"step": 1200
|
| 233 |
+
},
|
| 234 |
+
{
|
| 235 |
+
"epoch": 0.2756339581036384,
|
| 236 |
+
"grad_norm": 11.233166694641113,
|
| 237 |
+
"learning_rate": 4.3114663726571114e-05,
|
| 238 |
+
"loss": 0.7948,
|
| 239 |
+
"step": 1250
|
| 240 |
+
},
|
| 241 |
+
{
|
| 242 |
+
"epoch": 0.2866593164277839,
|
| 243 |
+
"grad_norm": 12.69255256652832,
|
| 244 |
+
"learning_rate": 4.283902976846748e-05,
|
| 245 |
+
"loss": 0.7261,
|
| 246 |
+
"step": 1300
|
| 247 |
+
},
|
| 248 |
+
{
|
| 249 |
+
"epoch": 0.29768467475192945,
|
| 250 |
+
"grad_norm": 8.976383209228516,
|
| 251 |
+
"learning_rate": 4.256339581036384e-05,
|
| 252 |
+
"loss": 0.7946,
|
| 253 |
+
"step": 1350
|
| 254 |
+
},
|
| 255 |
+
{
|
| 256 |
+
"epoch": 0.308710033076075,
|
| 257 |
+
"grad_norm": 8.328288078308105,
|
| 258 |
+
"learning_rate": 4.2287761852260196e-05,
|
| 259 |
+
"loss": 0.7378,
|
| 260 |
+
"step": 1400
|
| 261 |
+
},
|
| 262 |
+
{
|
| 263 |
+
"epoch": 0.308710033076075,
|
| 264 |
+
"eval_accuracy": 0.7050359712230215,
|
| 265 |
+
"eval_loss": 0.7756134867668152,
|
| 266 |
+
"eval_runtime": 48.2156,
|
| 267 |
+
"eval_samples_per_second": 83.604,
|
| 268 |
+
"eval_steps_per_second": 10.453,
|
| 269 |
+
"step": 1400
|
| 270 |
+
},
|
| 271 |
+
{
|
| 272 |
+
"epoch": 0.3197353914002205,
|
| 273 |
+
"grad_norm": 9.693521499633789,
|
| 274 |
+
"learning_rate": 4.201212789415656e-05,
|
| 275 |
+
"loss": 0.7439,
|
| 276 |
+
"step": 1450
|
| 277 |
+
},
|
| 278 |
+
{
|
| 279 |
+
"epoch": 0.33076074972436603,
|
| 280 |
+
"grad_norm": 9.462974548339844,
|
| 281 |
+
"learning_rate": 4.173649393605292e-05,
|
| 282 |
+
"loss": 0.7106,
|
| 283 |
+
"step": 1500
|
| 284 |
+
},
|
| 285 |
+
{
|
| 286 |
+
"epoch": 0.34178610804851156,
|
| 287 |
+
"grad_norm": 11.551827430725098,
|
| 288 |
+
"learning_rate": 4.1460859977949285e-05,
|
| 289 |
+
"loss": 0.6444,
|
| 290 |
+
"step": 1550
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 0.3528114663726571,
|
| 294 |
+
"grad_norm": 8.0066499710083,
|
| 295 |
+
"learning_rate": 4.118522601984565e-05,
|
| 296 |
+
"loss": 0.7226,
|
| 297 |
+
"step": 1600
|
| 298 |
+
},
|
| 299 |
+
{
|
| 300 |
+
"epoch": 0.3528114663726571,
|
| 301 |
+
"eval_accuracy": 0.7271148598362689,
|
| 302 |
+
"eval_loss": 0.7424366474151611,
|
| 303 |
+
"eval_runtime": 49.1199,
|
| 304 |
+
"eval_samples_per_second": 82.064,
|
| 305 |
+
"eval_steps_per_second": 10.261,
|
| 306 |
+
"step": 1600
|
| 307 |
+
},
|
| 308 |
+
{
|
| 309 |
+
"epoch": 0.3638368246968026,
|
| 310 |
+
"grad_norm": 11.164033889770508,
|
| 311 |
+
"learning_rate": 4.090959206174201e-05,
|
| 312 |
+
"loss": 0.6636,
|
| 313 |
+
"step": 1650
|
| 314 |
+
},
|
| 315 |
+
{
|
| 316 |
+
"epoch": 0.3748621830209482,
|
| 317 |
+
"grad_norm": 9.30005168914795,
|
| 318 |
+
"learning_rate": 4.063395810363837e-05,
|
| 319 |
+
"loss": 0.7728,
|
| 320 |
+
"step": 1700
|
| 321 |
+
},
|
| 322 |
+
{
|
| 323 |
+
"epoch": 0.38588754134509373,
|
| 324 |
+
"grad_norm": 8.798018455505371,
|
| 325 |
+
"learning_rate": 4.035832414553473e-05,
|
| 326 |
+
"loss": 0.755,
|
| 327 |
+
"step": 1750
|
| 328 |
+
},
|
| 329 |
+
{
|
| 330 |
+
"epoch": 0.39691289966923926,
|
| 331 |
+
"grad_norm": 17.514118194580078,
|
| 332 |
+
"learning_rate": 4.008269018743109e-05,
|
| 333 |
+
"loss": 0.7246,
|
| 334 |
+
"step": 1800
|
| 335 |
+
},
|
| 336 |
+
{
|
| 337 |
+
"epoch": 0.39691289966923926,
|
| 338 |
+
"eval_accuracy": 0.7472091292483255,
|
| 339 |
+
"eval_loss": 0.6997360587120056,
|
| 340 |
+
"eval_runtime": 50.8367,
|
| 341 |
+
"eval_samples_per_second": 79.293,
|
| 342 |
+
"eval_steps_per_second": 9.914,
|
| 343 |
+
"step": 1800
|
| 344 |
+
},
|
| 345 |
+
{
|
| 346 |
+
"epoch": 0.4079382579933848,
|
| 347 |
+
"grad_norm": 19.774938583374023,
|
| 348 |
+
"learning_rate": 3.9807056229327455e-05,
|
| 349 |
+
"loss": 0.6325,
|
| 350 |
+
"step": 1850
|
| 351 |
+
},
|
| 352 |
+
{
|
| 353 |
+
"epoch": 0.4189636163175303,
|
| 354 |
+
"grad_norm": 11.262466430664062,
|
| 355 |
+
"learning_rate": 3.953142227122382e-05,
|
| 356 |
+
"loss": 0.7494,
|
| 357 |
+
"step": 1900
|
| 358 |
+
},
|
| 359 |
+
{
|
| 360 |
+
"epoch": 0.42998897464167585,
|
| 361 |
+
"grad_norm": 6.449489593505859,
|
| 362 |
+
"learning_rate": 3.925578831312018e-05,
|
| 363 |
+
"loss": 0.6959,
|
| 364 |
+
"step": 1950
|
| 365 |
+
},
|
| 366 |
+
{
|
| 367 |
+
"epoch": 0.4410143329658214,
|
| 368 |
+
"grad_norm": 11.40916633605957,
|
| 369 |
+
"learning_rate": 3.898015435501654e-05,
|
| 370 |
+
"loss": 0.7348,
|
| 371 |
+
"step": 2000
|
| 372 |
+
},
|
| 373 |
+
{
|
| 374 |
+
"epoch": 0.4410143329658214,
|
| 375 |
+
"eval_accuracy": 0.7591168444554701,
|
| 376 |
+
"eval_loss": 0.6443637013435364,
|
| 377 |
+
"eval_runtime": 49.1885,
|
| 378 |
+
"eval_samples_per_second": 81.95,
|
| 379 |
+
"eval_steps_per_second": 10.246,
|
| 380 |
+
"step": 2000
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 0.4520396912899669,
|
| 384 |
+
"grad_norm": 8.649637222290039,
|
| 385 |
+
"learning_rate": 3.87045203969129e-05,
|
| 386 |
+
"loss": 0.7338,
|
| 387 |
+
"step": 2050
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 0.46306504961411243,
|
| 391 |
+
"grad_norm": 10.41679573059082,
|
| 392 |
+
"learning_rate": 3.842888643880926e-05,
|
| 393 |
+
"loss": 0.6482,
|
| 394 |
+
"step": 2100
|
| 395 |
+
},
|
| 396 |
+
{
|
| 397 |
+
"epoch": 0.474090407938258,
|
| 398 |
+
"grad_norm": 12.902926445007324,
|
| 399 |
+
"learning_rate": 3.8153252480705626e-05,
|
| 400 |
+
"loss": 0.6849,
|
| 401 |
+
"step": 2150
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 0.48511576626240355,
|
| 405 |
+
"grad_norm": 10.325972557067871,
|
| 406 |
+
"learning_rate": 3.787761852260199e-05,
|
| 407 |
+
"loss": 0.7373,
|
| 408 |
+
"step": 2200
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"epoch": 0.48511576626240355,
|
| 412 |
+
"eval_accuracy": 0.7578764574547259,
|
| 413 |
+
"eval_loss": 0.6530741453170776,
|
| 414 |
+
"eval_runtime": 50.2869,
|
| 415 |
+
"eval_samples_per_second": 80.16,
|
| 416 |
+
"eval_steps_per_second": 10.022,
|
| 417 |
+
"step": 2200
|
| 418 |
+
},
|
| 419 |
+
{
|
| 420 |
+
"epoch": 0.4961411245865491,
|
| 421 |
+
"grad_norm": 7.638971328735352,
|
| 422 |
+
"learning_rate": 3.760198456449835e-05,
|
| 423 |
+
"loss": 0.5985,
|
| 424 |
+
"step": 2250
|
| 425 |
+
},
|
| 426 |
+
{
|
| 427 |
+
"epoch": 0.5071664829106945,
|
| 428 |
+
"grad_norm": 6.052102565765381,
|
| 429 |
+
"learning_rate": 3.7326350606394714e-05,
|
| 430 |
+
"loss": 0.6742,
|
| 431 |
+
"step": 2300
|
| 432 |
+
},
|
| 433 |
+
{
|
| 434 |
+
"epoch": 0.5181918412348401,
|
| 435 |
+
"grad_norm": 19.927108764648438,
|
| 436 |
+
"learning_rate": 3.705071664829107e-05,
|
| 437 |
+
"loss": 0.5846,
|
| 438 |
+
"step": 2350
|
| 439 |
+
},
|
| 440 |
+
{
|
| 441 |
+
"epoch": 0.5292171995589857,
|
| 442 |
+
"grad_norm": 11.63871955871582,
|
| 443 |
+
"learning_rate": 3.677508269018743e-05,
|
| 444 |
+
"loss": 0.6943,
|
| 445 |
+
"step": 2400
|
| 446 |
+
},
|
| 447 |
+
{
|
| 448 |
+
"epoch": 0.5292171995589857,
|
| 449 |
+
"eval_accuracy": 0.7568841478541305,
|
| 450 |
+
"eval_loss": 0.6658030152320862,
|
| 451 |
+
"eval_runtime": 51.4998,
|
| 452 |
+
"eval_samples_per_second": 78.272,
|
| 453 |
+
"eval_steps_per_second": 9.786,
|
| 454 |
+
"step": 2400
|
| 455 |
+
},
|
| 456 |
+
{
|
| 457 |
+
"epoch": 0.5402425578831312,
|
| 458 |
+
"grad_norm": 10.362872123718262,
|
| 459 |
+
"learning_rate": 3.6499448732083796e-05,
|
| 460 |
+
"loss": 0.6303,
|
| 461 |
+
"step": 2450
|
| 462 |
+
},
|
| 463 |
+
{
|
| 464 |
+
"epoch": 0.5512679162072768,
|
| 465 |
+
"grad_norm": 6.9453606605529785,
|
| 466 |
+
"learning_rate": 3.622381477398016e-05,
|
| 467 |
+
"loss": 0.6494,
|
| 468 |
+
"step": 2500
|
| 469 |
+
},
|
| 470 |
+
{
|
| 471 |
+
"epoch": 0.5622932745314223,
|
| 472 |
+
"grad_norm": 22.353281021118164,
|
| 473 |
+
"learning_rate": 3.594818081587652e-05,
|
| 474 |
+
"loss": 0.6271,
|
| 475 |
+
"step": 2550
|
| 476 |
+
},
|
| 477 |
+
{
|
| 478 |
+
"epoch": 0.5733186328555678,
|
| 479 |
+
"grad_norm": 4.380390644073486,
|
| 480 |
+
"learning_rate": 3.5672546857772885e-05,
|
| 481 |
+
"loss": 0.6887,
|
| 482 |
+
"step": 2600
|
| 483 |
+
},
|
| 484 |
+
{
|
| 485 |
+
"epoch": 0.5733186328555678,
|
| 486 |
+
"eval_accuracy": 0.7412552716447531,
|
| 487 |
+
"eval_loss": 0.7230509519577026,
|
| 488 |
+
"eval_runtime": 51.9788,
|
| 489 |
+
"eval_samples_per_second": 77.551,
|
| 490 |
+
"eval_steps_per_second": 9.696,
|
| 491 |
+
"step": 2600
|
| 492 |
+
},
|
| 493 |
+
{
|
| 494 |
+
"epoch": 0.5843439911797134,
|
| 495 |
+
"grad_norm": 15.384754180908203,
|
| 496 |
+
"learning_rate": 3.539691289966924e-05,
|
| 497 |
+
"loss": 0.71,
|
| 498 |
+
"step": 2650
|
| 499 |
+
},
|
| 500 |
+
{
|
| 501 |
+
"epoch": 0.5953693495038589,
|
| 502 |
+
"grad_norm": 7.5070390701293945,
|
| 503 |
+
"learning_rate": 3.5121278941565604e-05,
|
| 504 |
+
"loss": 0.6521,
|
| 505 |
+
"step": 2700
|
| 506 |
+
},
|
| 507 |
+
{
|
| 508 |
+
"epoch": 0.6063947078280044,
|
| 509 |
+
"grad_norm": 9.219125747680664,
|
| 510 |
+
"learning_rate": 3.484564498346197e-05,
|
| 511 |
+
"loss": 0.6781,
|
| 512 |
+
"step": 2750
|
| 513 |
+
},
|
| 514 |
+
{
|
| 515 |
+
"epoch": 0.61742006615215,
|
| 516 |
+
"grad_norm": 9.403863906860352,
|
| 517 |
+
"learning_rate": 3.457001102535832e-05,
|
| 518 |
+
"loss": 0.6017,
|
| 519 |
+
"step": 2800
|
| 520 |
+
},
|
| 521 |
+
{
|
| 522 |
+
"epoch": 0.61742006615215,
|
| 523 |
+
"eval_accuracy": 0.759364921855619,
|
| 524 |
+
"eval_loss": 0.6455864906311035,
|
| 525 |
+
"eval_runtime": 51.1726,
|
| 526 |
+
"eval_samples_per_second": 78.773,
|
| 527 |
+
"eval_steps_per_second": 9.849,
|
| 528 |
+
"step": 2800
|
| 529 |
+
},
|
| 530 |
+
{
|
| 531 |
+
"epoch": 0.6284454244762955,
|
| 532 |
+
"grad_norm": 12.217251777648926,
|
| 533 |
+
"learning_rate": 3.4294377067254686e-05,
|
| 534 |
+
"loss": 0.6596,
|
| 535 |
+
"step": 2850
|
| 536 |
+
},
|
| 537 |
+
{
|
| 538 |
+
"epoch": 0.639470782800441,
|
| 539 |
+
"grad_norm": 12.6351957321167,
|
| 540 |
+
"learning_rate": 3.401874310915105e-05,
|
| 541 |
+
"loss": 0.7351,
|
| 542 |
+
"step": 2900
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 0.6504961411245865,
|
| 546 |
+
"grad_norm": 16.1507568359375,
|
| 547 |
+
"learning_rate": 3.374310915104741e-05,
|
| 548 |
+
"loss": 0.6256,
|
| 549 |
+
"step": 2950
|
| 550 |
+
},
|
| 551 |
+
{
|
| 552 |
+
"epoch": 0.6615214994487321,
|
| 553 |
+
"grad_norm": 12.999460220336914,
|
| 554 |
+
"learning_rate": 3.346747519294377e-05,
|
| 555 |
+
"loss": 0.6082,
|
| 556 |
+
"step": 3000
|
| 557 |
+
},
|
| 558 |
+
{
|
| 559 |
+
"epoch": 0.6615214994487321,
|
| 560 |
+
"eval_accuracy": 0.7749937980649962,
|
| 561 |
+
"eval_loss": 0.6113554835319519,
|
| 562 |
+
"eval_runtime": 51.8574,
|
| 563 |
+
"eval_samples_per_second": 77.732,
|
| 564 |
+
"eval_steps_per_second": 9.719,
|
| 565 |
+
"step": 3000
|
| 566 |
+
},
|
| 567 |
+
{
|
| 568 |
+
"epoch": 0.6725468577728776,
|
| 569 |
+
"grad_norm": 12.123536109924316,
|
| 570 |
+
"learning_rate": 3.319184123484013e-05,
|
| 571 |
+
"loss": 0.5792,
|
| 572 |
+
"step": 3050
|
| 573 |
+
},
|
| 574 |
+
{
|
| 575 |
+
"epoch": 0.6835722160970231,
|
| 576 |
+
"grad_norm": 13.215846061706543,
|
| 577 |
+
"learning_rate": 3.291620727673649e-05,
|
| 578 |
+
"loss": 0.6383,
|
| 579 |
+
"step": 3100
|
| 580 |
+
},
|
| 581 |
+
{
|
| 582 |
+
"epoch": 0.6945975744211687,
|
| 583 |
+
"grad_norm": 18.97852325439453,
|
| 584 |
+
"learning_rate": 3.2640573318632856e-05,
|
| 585 |
+
"loss": 0.616,
|
| 586 |
+
"step": 3150
|
| 587 |
+
},
|
| 588 |
+
{
|
| 589 |
+
"epoch": 0.7056229327453142,
|
| 590 |
+
"grad_norm": 13.575655937194824,
|
| 591 |
+
"learning_rate": 3.236493936052922e-05,
|
| 592 |
+
"loss": 0.4946,
|
| 593 |
+
"step": 3200
|
| 594 |
+
},
|
| 595 |
+
{
|
| 596 |
+
"epoch": 0.7056229327453142,
|
| 597 |
+
"eval_accuracy": 0.7752418754651451,
|
| 598 |
+
"eval_loss": 0.6550247669219971,
|
| 599 |
+
"eval_runtime": 51.001,
|
| 600 |
+
"eval_samples_per_second": 79.038,
|
| 601 |
+
"eval_steps_per_second": 9.882,
|
| 602 |
+
"step": 3200
|
| 603 |
+
},
|
| 604 |
+
{
|
| 605 |
+
"epoch": 0.7166482910694597,
|
| 606 |
+
"grad_norm": 13.980043411254883,
|
| 607 |
+
"learning_rate": 3.208930540242558e-05,
|
| 608 |
+
"loss": 0.7087,
|
| 609 |
+
"step": 3250
|
| 610 |
+
},
|
| 611 |
+
{
|
| 612 |
+
"epoch": 0.7276736493936052,
|
| 613 |
+
"grad_norm": 16.2875919342041,
|
| 614 |
+
"learning_rate": 3.181367144432194e-05,
|
| 615 |
+
"loss": 0.6054,
|
| 616 |
+
"step": 3300
|
| 617 |
+
},
|
| 618 |
+
{
|
| 619 |
+
"epoch": 0.7386990077177509,
|
| 620 |
+
"grad_norm": 4.533209323883057,
|
| 621 |
+
"learning_rate": 3.15380374862183e-05,
|
| 622 |
+
"loss": 0.5812,
|
| 623 |
+
"step": 3350
|
| 624 |
+
},
|
| 625 |
+
{
|
| 626 |
+
"epoch": 0.7497243660418964,
|
| 627 |
+
"grad_norm": 9.643322944641113,
|
| 628 |
+
"learning_rate": 3.1262403528114664e-05,
|
| 629 |
+
"loss": 0.6084,
|
| 630 |
+
"step": 3400
|
| 631 |
+
},
|
| 632 |
+
{
|
| 633 |
+
"epoch": 0.7497243660418964,
|
| 634 |
+
"eval_accuracy": 0.759364921855619,
|
| 635 |
+
"eval_loss": 0.6849791407585144,
|
| 636 |
+
"eval_runtime": 51.3096,
|
| 637 |
+
"eval_samples_per_second": 78.562,
|
| 638 |
+
"eval_steps_per_second": 9.823,
|
| 639 |
+
"step": 3400
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 0.7607497243660419,
|
| 643 |
+
"grad_norm": 12.94963550567627,
|
| 644 |
+
"learning_rate": 3.0986769570011027e-05,
|
| 645 |
+
"loss": 0.5558,
|
| 646 |
+
"step": 3450
|
| 647 |
+
},
|
| 648 |
+
{
|
| 649 |
+
"epoch": 0.7717750826901875,
|
| 650 |
+
"grad_norm": 11.617074966430664,
|
| 651 |
+
"learning_rate": 3.071113561190739e-05,
|
| 652 |
+
"loss": 0.6267,
|
| 653 |
+
"step": 3500
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 0.782800441014333,
|
| 657 |
+
"grad_norm": 11.406350135803223,
|
| 658 |
+
"learning_rate": 3.043550165380375e-05,
|
| 659 |
+
"loss": 0.5174,
|
| 660 |
+
"step": 3550
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"epoch": 0.7938257993384785,
|
| 664 |
+
"grad_norm": 6.776385307312012,
|
| 665 |
+
"learning_rate": 3.0159867695700112e-05,
|
| 666 |
+
"loss": 0.71,
|
| 667 |
+
"step": 3600
|
| 668 |
+
},
|
| 669 |
+
{
|
| 670 |
+
"epoch": 0.7938257993384785,
|
| 671 |
+
"eval_accuracy": 0.7928553708757132,
|
| 672 |
+
"eval_loss": 0.578441858291626,
|
| 673 |
+
"eval_runtime": 53.7733,
|
| 674 |
+
"eval_samples_per_second": 74.963,
|
| 675 |
+
"eval_steps_per_second": 9.373,
|
| 676 |
+
"step": 3600
|
| 677 |
+
},
|
| 678 |
+
{
|
| 679 |
+
"epoch": 0.804851157662624,
|
| 680 |
+
"grad_norm": 21.484195709228516,
|
| 681 |
+
"learning_rate": 2.9884233737596475e-05,
|
| 682 |
+
"loss": 0.5633,
|
| 683 |
+
"step": 3650
|
| 684 |
+
},
|
| 685 |
+
{
|
| 686 |
+
"epoch": 0.8158765159867696,
|
| 687 |
+
"grad_norm": 19.93082046508789,
|
| 688 |
+
"learning_rate": 2.9608599779492834e-05,
|
| 689 |
+
"loss": 0.5926,
|
| 690 |
+
"step": 3700
|
| 691 |
+
},
|
| 692 |
+
{
|
| 693 |
+
"epoch": 0.8269018743109151,
|
| 694 |
+
"grad_norm": 13.265941619873047,
|
| 695 |
+
"learning_rate": 2.9332965821389197e-05,
|
| 696 |
+
"loss": 0.6305,
|
| 697 |
+
"step": 3750
|
| 698 |
+
},
|
| 699 |
+
{
|
| 700 |
+
"epoch": 0.8379272326350606,
|
| 701 |
+
"grad_norm": 42.45393371582031,
|
| 702 |
+
"learning_rate": 2.905733186328556e-05,
|
| 703 |
+
"loss": 0.5575,
|
| 704 |
+
"step": 3800
|
| 705 |
+
},
|
| 706 |
+
{
|
| 707 |
+
"epoch": 0.8379272326350606,
|
| 708 |
+
"eval_accuracy": 0.8020342346812206,
|
| 709 |
+
"eval_loss": 0.5739887356758118,
|
| 710 |
+
"eval_runtime": 55.181,
|
| 711 |
+
"eval_samples_per_second": 73.051,
|
| 712 |
+
"eval_steps_per_second": 9.134,
|
| 713 |
+
"step": 3800
|
| 714 |
+
},
|
| 715 |
+
{
|
| 716 |
+
"epoch": 0.8489525909592062,
|
| 717 |
+
"grad_norm": 10.487258911132812,
|
| 718 |
+
"learning_rate": 2.878169790518192e-05,
|
| 719 |
+
"loss": 0.6158,
|
| 720 |
+
"step": 3850
|
| 721 |
+
},
|
| 722 |
+
{
|
| 723 |
+
"epoch": 0.8599779492833517,
|
| 724 |
+
"grad_norm": 8.130553245544434,
|
| 725 |
+
"learning_rate": 2.8506063947078282e-05,
|
| 726 |
+
"loss": 0.6652,
|
| 727 |
+
"step": 3900
|
| 728 |
+
},
|
| 729 |
+
{
|
| 730 |
+
"epoch": 0.8710033076074972,
|
| 731 |
+
"grad_norm": 13.992973327636719,
|
| 732 |
+
"learning_rate": 2.8230429988974645e-05,
|
| 733 |
+
"loss": 0.5247,
|
| 734 |
+
"step": 3950
|
| 735 |
+
},
|
| 736 |
+
{
|
| 737 |
+
"epoch": 0.8820286659316428,
|
| 738 |
+
"grad_norm": 15.441995620727539,
|
| 739 |
+
"learning_rate": 2.7954796030871005e-05,
|
| 740 |
+
"loss": 0.6478,
|
| 741 |
+
"step": 4000
|
| 742 |
+
},
|
| 743 |
+
{
|
| 744 |
+
"epoch": 0.8820286659316428,
|
| 745 |
+
"eval_accuracy": 0.8144381046886628,
|
| 746 |
+
"eval_loss": 0.524382472038269,
|
| 747 |
+
"eval_runtime": 52.1097,
|
| 748 |
+
"eval_samples_per_second": 77.356,
|
| 749 |
+
"eval_steps_per_second": 9.672,
|
| 750 |
+
"step": 4000
|
| 751 |
+
},
|
| 752 |
+
{
|
| 753 |
+
"epoch": 0.8930540242557883,
|
| 754 |
+
"grad_norm": 18.06987762451172,
|
| 755 |
+
"learning_rate": 2.7679162072767367e-05,
|
| 756 |
+
"loss": 0.5071,
|
| 757 |
+
"step": 4050
|
| 758 |
+
},
|
| 759 |
+
{
|
| 760 |
+
"epoch": 0.9040793825799338,
|
| 761 |
+
"grad_norm": 7.0000505447387695,
|
| 762 |
+
"learning_rate": 2.740352811466373e-05,
|
| 763 |
+
"loss": 0.5079,
|
| 764 |
+
"step": 4100
|
| 765 |
+
},
|
| 766 |
+
{
|
| 767 |
+
"epoch": 0.9151047409040793,
|
| 768 |
+
"grad_norm": 11.618234634399414,
|
| 769 |
+
"learning_rate": 2.712789415656009e-05,
|
| 770 |
+
"loss": 0.5333,
|
| 771 |
+
"step": 4150
|
| 772 |
+
},
|
| 773 |
+
{
|
| 774 |
+
"epoch": 0.9261300992282249,
|
| 775 |
+
"grad_norm": 29.748830795288086,
|
| 776 |
+
"learning_rate": 2.6852260198456453e-05,
|
| 777 |
+
"loss": 0.5765,
|
| 778 |
+
"step": 4200
|
| 779 |
+
},
|
| 780 |
+
{
|
| 781 |
+
"epoch": 0.9261300992282249,
|
| 782 |
+
"eval_accuracy": 0.8196477300917886,
|
| 783 |
+
"eval_loss": 0.537652850151062,
|
| 784 |
+
"eval_runtime": 53.3275,
|
| 785 |
+
"eval_samples_per_second": 75.59,
|
| 786 |
+
"eval_steps_per_second": 9.451,
|
| 787 |
+
"step": 4200
|
| 788 |
+
},
|
| 789 |
+
{
|
| 790 |
+
"epoch": 0.9371554575523704,
|
| 791 |
+
"grad_norm": 35.87691116333008,
|
| 792 |
+
"learning_rate": 2.6576626240352816e-05,
|
| 793 |
+
"loss": 0.6954,
|
| 794 |
+
"step": 4250
|
| 795 |
+
},
|
| 796 |
+
{
|
| 797 |
+
"epoch": 0.948180815876516,
|
| 798 |
+
"grad_norm": 9.353912353515625,
|
| 799 |
+
"learning_rate": 2.6300992282249175e-05,
|
| 800 |
+
"loss": 0.5576,
|
| 801 |
+
"step": 4300
|
| 802 |
+
},
|
| 803 |
+
{
|
| 804 |
+
"epoch": 0.9592061742006616,
|
| 805 |
+
"grad_norm": 26.52071189880371,
|
| 806 |
+
"learning_rate": 2.6025358324145538e-05,
|
| 807 |
+
"loss": 0.6409,
|
| 808 |
+
"step": 4350
|
| 809 |
+
},
|
| 810 |
+
{
|
| 811 |
+
"epoch": 0.9702315325248071,
|
| 812 |
+
"grad_norm": 15.028057098388672,
|
| 813 |
+
"learning_rate": 2.57497243660419e-05,
|
| 814 |
+
"loss": 0.5363,
|
| 815 |
+
"step": 4400
|
| 816 |
+
},
|
| 817 |
+
{
|
| 818 |
+
"epoch": 0.9702315325248071,
|
| 819 |
+
"eval_accuracy": 0.8022823120813694,
|
| 820 |
+
"eval_loss": 0.5548905730247498,
|
| 821 |
+
"eval_runtime": 51.9595,
|
| 822 |
+
"eval_samples_per_second": 77.58,
|
| 823 |
+
"eval_steps_per_second": 9.7,
|
| 824 |
+
"step": 4400
|
| 825 |
+
},
|
| 826 |
+
{
|
| 827 |
+
"epoch": 0.9812568908489526,
|
| 828 |
+
"grad_norm": 12.458972930908203,
|
| 829 |
+
"learning_rate": 2.547409040793826e-05,
|
| 830 |
+
"loss": 0.551,
|
| 831 |
+
"step": 4450
|
| 832 |
+
},
|
| 833 |
+
{
|
| 834 |
+
"epoch": 0.9922822491730982,
|
| 835 |
+
"grad_norm": 13.601612091064453,
|
| 836 |
+
"learning_rate": 2.5198456449834623e-05,
|
| 837 |
+
"loss": 0.4871,
|
| 838 |
+
"step": 4500
|
| 839 |
+
},
|
| 840 |
+
{
|
| 841 |
+
"epoch": 1.0033076074972436,
|
| 842 |
+
"grad_norm": 4.1361870765686035,
|
| 843 |
+
"learning_rate": 2.4922822491730983e-05,
|
| 844 |
+
"loss": 0.4908,
|
| 845 |
+
"step": 4550
|
| 846 |
+
},
|
| 847 |
+
{
|
| 848 |
+
"epoch": 1.014332965821389,
|
| 849 |
+
"grad_norm": 7.244948863983154,
|
| 850 |
+
"learning_rate": 2.4647188533627346e-05,
|
| 851 |
+
"loss": 0.4169,
|
| 852 |
+
"step": 4600
|
| 853 |
+
},
|
| 854 |
+
{
|
| 855 |
+
"epoch": 1.014332965821389,
|
| 856 |
+
"eval_accuracy": 0.8005457702803275,
|
| 857 |
+
"eval_loss": 0.5810744166374207,
|
| 858 |
+
"eval_runtime": 44.6894,
|
| 859 |
+
"eval_samples_per_second": 90.2,
|
| 860 |
+
"eval_steps_per_second": 11.278,
|
| 861 |
+
"step": 4600
|
| 862 |
+
},
|
| 863 |
+
{
|
| 864 |
+
"epoch": 1.0253583241455346,
|
| 865 |
+
"grad_norm": 1.200239896774292,
|
| 866 |
+
"learning_rate": 2.4371554575523705e-05,
|
| 867 |
+
"loss": 0.4883,
|
| 868 |
+
"step": 4650
|
| 869 |
+
},
|
| 870 |
+
{
|
| 871 |
+
"epoch": 1.0363836824696802,
|
| 872 |
+
"grad_norm": 6.8026251792907715,
|
| 873 |
+
"learning_rate": 2.4095920617420068e-05,
|
| 874 |
+
"loss": 0.443,
|
| 875 |
+
"step": 4700
|
| 876 |
+
},
|
| 877 |
+
{
|
| 878 |
+
"epoch": 1.0474090407938257,
|
| 879 |
+
"grad_norm": 12.317420959472656,
|
| 880 |
+
"learning_rate": 2.382028665931643e-05,
|
| 881 |
+
"loss": 0.5633,
|
| 882 |
+
"step": 4750
|
| 883 |
+
},
|
| 884 |
+
{
|
| 885 |
+
"epoch": 1.0584343991179714,
|
| 886 |
+
"grad_norm": 12.111879348754883,
|
| 887 |
+
"learning_rate": 2.354465270121279e-05,
|
| 888 |
+
"loss": 0.4637,
|
| 889 |
+
"step": 4800
|
| 890 |
+
},
|
| 891 |
+
{
|
| 892 |
+
"epoch": 1.0584343991179714,
|
| 893 |
+
"eval_accuracy": 0.820640039692384,
|
| 894 |
+
"eval_loss": 0.530491292476654,
|
| 895 |
+
"eval_runtime": 44.799,
|
| 896 |
+
"eval_samples_per_second": 89.98,
|
| 897 |
+
"eval_steps_per_second": 11.25,
|
| 898 |
+
"step": 4800
|
| 899 |
+
},
|
| 900 |
+
{
|
| 901 |
+
"epoch": 1.069459757442117,
|
| 902 |
+
"grad_norm": 7.548985004425049,
|
| 903 |
+
"learning_rate": 2.3269018743109153e-05,
|
| 904 |
+
"loss": 0.4133,
|
| 905 |
+
"step": 4850
|
| 906 |
+
},
|
| 907 |
+
{
|
| 908 |
+
"epoch": 1.0804851157662625,
|
| 909 |
+
"grad_norm": 7.3979997634887695,
|
| 910 |
+
"learning_rate": 2.2993384785005516e-05,
|
| 911 |
+
"loss": 0.4151,
|
| 912 |
+
"step": 4900
|
| 913 |
+
},
|
| 914 |
+
{
|
| 915 |
+
"epoch": 1.091510474090408,
|
| 916 |
+
"grad_norm": 11.04232120513916,
|
| 917 |
+
"learning_rate": 2.2717750826901875e-05,
|
| 918 |
+
"loss": 0.4044,
|
| 919 |
+
"step": 4950
|
| 920 |
+
},
|
| 921 |
+
{
|
| 922 |
+
"epoch": 1.1025358324145536,
|
| 923 |
+
"grad_norm": 9.042499542236328,
|
| 924 |
+
"learning_rate": 2.244211686879824e-05,
|
| 925 |
+
"loss": 0.5072,
|
| 926 |
+
"step": 5000
|
| 927 |
+
},
|
| 928 |
+
{
|
| 929 |
+
"epoch": 1.1025358324145536,
|
| 930 |
+
"eval_accuracy": 0.7878938228727363,
|
| 931 |
+
"eval_loss": 0.6346498727798462,
|
| 932 |
+
"eval_runtime": 46.0611,
|
| 933 |
+
"eval_samples_per_second": 87.514,
|
| 934 |
+
"eval_steps_per_second": 10.942,
|
| 935 |
+
"step": 5000
|
| 936 |
+
},
|
| 937 |
+
{
|
| 938 |
+
"epoch": 1.113561190738699,
|
| 939 |
+
"grad_norm": 11.084900856018066,
|
| 940 |
+
"learning_rate": 2.21664829106946e-05,
|
| 941 |
+
"loss": 0.415,
|
| 942 |
+
"step": 5050
|
| 943 |
+
},
|
| 944 |
+
{
|
| 945 |
+
"epoch": 1.1245865490628446,
|
| 946 |
+
"grad_norm": 12.728482246398926,
|
| 947 |
+
"learning_rate": 2.189084895259096e-05,
|
| 948 |
+
"loss": 0.4814,
|
| 949 |
+
"step": 5100
|
| 950 |
+
},
|
| 951 |
+
{
|
| 952 |
+
"epoch": 1.1356119073869901,
|
| 953 |
+
"grad_norm": 4.183909893035889,
|
| 954 |
+
"learning_rate": 2.161521499448732e-05,
|
| 955 |
+
"loss": 0.417,
|
| 956 |
+
"step": 5150
|
| 957 |
+
},
|
| 958 |
+
{
|
| 959 |
+
"epoch": 1.1466372657111357,
|
| 960 |
+
"grad_norm": 4.626537322998047,
|
| 961 |
+
"learning_rate": 2.1339581036383683e-05,
|
| 962 |
+
"loss": 0.5391,
|
| 963 |
+
"step": 5200
|
| 964 |
+
},
|
| 965 |
+
{
|
| 966 |
+
"epoch": 1.1466372657111357,
|
| 967 |
+
"eval_accuracy": 0.8196477300917886,
|
| 968 |
+
"eval_loss": 0.5200665593147278,
|
| 969 |
+
"eval_runtime": 51.4326,
|
| 970 |
+
"eval_samples_per_second": 78.374,
|
| 971 |
+
"eval_steps_per_second": 9.799,
|
| 972 |
+
"step": 5200
|
| 973 |
+
},
|
| 974 |
+
{
|
| 975 |
+
"epoch": 1.1576626240352812,
|
| 976 |
+
"grad_norm": 2.050447463989258,
|
| 977 |
+
"learning_rate": 2.1063947078280043e-05,
|
| 978 |
+
"loss": 0.4601,
|
| 979 |
+
"step": 5250
|
| 980 |
+
},
|
| 981 |
+
{
|
| 982 |
+
"epoch": 1.1686879823594267,
|
| 983 |
+
"grad_norm": 4.204873561859131,
|
| 984 |
+
"learning_rate": 2.0788313120176405e-05,
|
| 985 |
+
"loss": 0.3647,
|
| 986 |
+
"step": 5300
|
| 987 |
+
},
|
| 988 |
+
{
|
| 989 |
+
"epoch": 1.1797133406835723,
|
| 990 |
+
"grad_norm": 4.201039791107178,
|
| 991 |
+
"learning_rate": 2.051267916207277e-05,
|
| 992 |
+
"loss": 0.4608,
|
| 993 |
+
"step": 5350
|
| 994 |
+
},
|
| 995 |
+
{
|
| 996 |
+
"epoch": 1.1907386990077178,
|
| 997 |
+
"grad_norm": 26.340166091918945,
|
| 998 |
+
"learning_rate": 2.0237045203969128e-05,
|
| 999 |
+
"loss": 0.4807,
|
| 1000 |
+
"step": 5400
|
| 1001 |
+
},
|
| 1002 |
+
{
|
| 1003 |
+
"epoch": 1.1907386990077178,
|
| 1004 |
+
"eval_accuracy": 0.8007938476804763,
|
| 1005 |
+
"eval_loss": 0.6092058420181274,
|
| 1006 |
+
"eval_runtime": 48.0907,
|
| 1007 |
+
"eval_samples_per_second": 83.821,
|
| 1008 |
+
"eval_steps_per_second": 10.48,
|
| 1009 |
+
"step": 5400
|
| 1010 |
+
},
|
| 1011 |
+
{
|
| 1012 |
+
"epoch": 1.2017640573318633,
|
| 1013 |
+
"grad_norm": 3.7308127880096436,
|
| 1014 |
+
"learning_rate": 1.996141124586549e-05,
|
| 1015 |
+
"loss": 0.4879,
|
| 1016 |
+
"step": 5450
|
| 1017 |
+
},
|
| 1018 |
+
{
|
| 1019 |
+
"epoch": 1.2127894156560088,
|
| 1020 |
+
"grad_norm": 2.054511785507202,
|
| 1021 |
+
"learning_rate": 1.9685777287761854e-05,
|
| 1022 |
+
"loss": 0.4749,
|
| 1023 |
+
"step": 5500
|
| 1024 |
+
},
|
| 1025 |
+
{
|
| 1026 |
+
"epoch": 1.2238147739801544,
|
| 1027 |
+
"grad_norm": 21.566608428955078,
|
| 1028 |
+
"learning_rate": 1.9410143329658213e-05,
|
| 1029 |
+
"loss": 0.4921,
|
| 1030 |
+
"step": 5550
|
| 1031 |
+
},
|
| 1032 |
+
{
|
| 1033 |
+
"epoch": 1.2348401323043,
|
| 1034 |
+
"grad_norm": 12.240718841552734,
|
| 1035 |
+
"learning_rate": 1.9134509371554576e-05,
|
| 1036 |
+
"loss": 0.3721,
|
| 1037 |
+
"step": 5600
|
| 1038 |
+
},
|
| 1039 |
+
{
|
| 1040 |
+
"epoch": 1.2348401323043,
|
| 1041 |
+
"eval_accuracy": 0.8248573554949145,
|
| 1042 |
+
"eval_loss": 0.5377492904663086,
|
| 1043 |
+
"eval_runtime": 50.1111,
|
| 1044 |
+
"eval_samples_per_second": 80.441,
|
| 1045 |
+
"eval_steps_per_second": 10.058,
|
| 1046 |
+
"step": 5600
|
| 1047 |
+
},
|
| 1048 |
+
{
|
| 1049 |
+
"epoch": 1.2458654906284454,
|
| 1050 |
+
"grad_norm": 25.768829345703125,
|
| 1051 |
+
"learning_rate": 1.885887541345094e-05,
|
| 1052 |
+
"loss": 0.4781,
|
| 1053 |
+
"step": 5650
|
| 1054 |
+
},
|
| 1055 |
+
{
|
| 1056 |
+
"epoch": 1.256890848952591,
|
| 1057 |
+
"grad_norm": 34.470481872558594,
|
| 1058 |
+
"learning_rate": 1.85832414553473e-05,
|
| 1059 |
+
"loss": 0.5108,
|
| 1060 |
+
"step": 5700
|
| 1061 |
+
},
|
| 1062 |
+
{
|
| 1063 |
+
"epoch": 1.2679162072767365,
|
| 1064 |
+
"grad_norm": 12.565007209777832,
|
| 1065 |
+
"learning_rate": 1.830760749724366e-05,
|
| 1066 |
+
"loss": 0.4817,
|
| 1067 |
+
"step": 5750
|
| 1068 |
+
},
|
| 1069 |
+
{
|
| 1070 |
+
"epoch": 1.278941565600882,
|
| 1071 |
+
"grad_norm": 15.793746948242188,
|
| 1072 |
+
"learning_rate": 1.8031973539140024e-05,
|
| 1073 |
+
"loss": 0.4601,
|
| 1074 |
+
"step": 5800
|
| 1075 |
+
},
|
| 1076 |
+
{
|
| 1077 |
+
"epoch": 1.278941565600882,
|
| 1078 |
+
"eval_accuracy": 0.8226246588935748,
|
| 1079 |
+
"eval_loss": 0.5374011993408203,
|
| 1080 |
+
"eval_runtime": 55.1619,
|
| 1081 |
+
"eval_samples_per_second": 73.076,
|
| 1082 |
+
"eval_steps_per_second": 9.137,
|
| 1083 |
+
"step": 5800
|
| 1084 |
+
},
|
| 1085 |
+
{
|
| 1086 |
+
"epoch": 1.2899669239250275,
|
| 1087 |
+
"grad_norm": 7.197420597076416,
|
| 1088 |
+
"learning_rate": 1.7756339581036387e-05,
|
| 1089 |
+
"loss": 0.3642,
|
| 1090 |
+
"step": 5850
|
| 1091 |
+
},
|
| 1092 |
+
{
|
| 1093 |
+
"epoch": 1.300992282249173,
|
| 1094 |
+
"grad_norm": 16.694839477539062,
|
| 1095 |
+
"learning_rate": 1.7480705622932746e-05,
|
| 1096 |
+
"loss": 0.4722,
|
| 1097 |
+
"step": 5900
|
| 1098 |
+
},
|
| 1099 |
+
{
|
| 1100 |
+
"epoch": 1.3120176405733186,
|
| 1101 |
+
"grad_norm": 4.159826278686523,
|
| 1102 |
+
"learning_rate": 1.720507166482911e-05,
|
| 1103 |
+
"loss": 0.3647,
|
| 1104 |
+
"step": 5950
|
| 1105 |
+
},
|
| 1106 |
+
{
|
| 1107 |
+
"epoch": 1.3230429988974641,
|
| 1108 |
+
"grad_norm": 3.6812713146209717,
|
| 1109 |
+
"learning_rate": 1.6929437706725472e-05,
|
| 1110 |
+
"loss": 0.4906,
|
| 1111 |
+
"step": 6000
|
| 1112 |
+
},
|
| 1113 |
+
{
|
| 1114 |
+
"epoch": 1.3230429988974641,
|
| 1115 |
+
"eval_accuracy": 0.8322996774993798,
|
| 1116 |
+
"eval_loss": 0.5462606549263,
|
| 1117 |
+
"eval_runtime": 50.2107,
|
| 1118 |
+
"eval_samples_per_second": 80.282,
|
| 1119 |
+
"eval_steps_per_second": 10.038,
|
| 1120 |
+
"step": 6000
|
| 1121 |
+
},
|
| 1122 |
+
{
|
| 1123 |
+
"epoch": 1.3340683572216097,
|
| 1124 |
+
"grad_norm": 16.1009521484375,
|
| 1125 |
+
"learning_rate": 1.665380374862183e-05,
|
| 1126 |
+
"loss": 0.5013,
|
| 1127 |
+
"step": 6050
|
| 1128 |
+
},
|
| 1129 |
+
{
|
| 1130 |
+
"epoch": 1.3450937155457552,
|
| 1131 |
+
"grad_norm": 9.130171775817871,
|
| 1132 |
+
"learning_rate": 1.637816979051819e-05,
|
| 1133 |
+
"loss": 0.5111,
|
| 1134 |
+
"step": 6100
|
| 1135 |
+
},
|
| 1136 |
+
{
|
| 1137 |
+
"epoch": 1.3561190738699007,
|
| 1138 |
+
"grad_norm": 7.678464412689209,
|
| 1139 |
+
"learning_rate": 1.6102535832414554e-05,
|
| 1140 |
+
"loss": 0.4683,
|
| 1141 |
+
"step": 6150
|
| 1142 |
+
},
|
| 1143 |
+
{
|
| 1144 |
+
"epoch": 1.3671444321940462,
|
| 1145 |
+
"grad_norm": 14.63408088684082,
|
| 1146 |
+
"learning_rate": 1.5826901874310913e-05,
|
| 1147 |
+
"loss": 0.4328,
|
| 1148 |
+
"step": 6200
|
| 1149 |
+
},
|
| 1150 |
+
{
|
| 1151 |
+
"epoch": 1.3671444321940462,
|
| 1152 |
+
"eval_accuracy": 0.8322996774993798,
|
| 1153 |
+
"eval_loss": 0.5245828032493591,
|
| 1154 |
+
"eval_runtime": 51.9955,
|
| 1155 |
+
"eval_samples_per_second": 77.526,
|
| 1156 |
+
"eval_steps_per_second": 9.693,
|
| 1157 |
+
"step": 6200
|
| 1158 |
+
},
|
| 1159 |
+
{
|
| 1160 |
+
"epoch": 1.3781697905181918,
|
| 1161 |
+
"grad_norm": 35.199092864990234,
|
| 1162 |
+
"learning_rate": 1.5551267916207276e-05,
|
| 1163 |
+
"loss": 0.4871,
|
| 1164 |
+
"step": 6250
|
| 1165 |
+
},
|
| 1166 |
+
{
|
| 1167 |
+
"epoch": 1.3891951488423373,
|
| 1168 |
+
"grad_norm": 12.516600608825684,
|
| 1169 |
+
"learning_rate": 1.527563395810364e-05,
|
| 1170 |
+
"loss": 0.4812,
|
| 1171 |
+
"step": 6300
|
| 1172 |
+
},
|
| 1173 |
+
{
|
| 1174 |
+
"epoch": 1.4002205071664828,
|
| 1175 |
+
"grad_norm": 12.948488235473633,
|
| 1176 |
+
"learning_rate": 1.5e-05,
|
| 1177 |
+
"loss": 0.4149,
|
| 1178 |
+
"step": 6350
|
| 1179 |
+
},
|
| 1180 |
+
{
|
| 1181 |
+
"epoch": 1.4112458654906286,
|
| 1182 |
+
"grad_norm": 17.36246681213379,
|
| 1183 |
+
"learning_rate": 1.4724366041896362e-05,
|
| 1184 |
+
"loss": 0.3668,
|
| 1185 |
+
"step": 6400
|
| 1186 |
+
},
|
| 1187 |
+
{
|
| 1188 |
+
"epoch": 1.4112458654906286,
|
| 1189 |
+
"eval_accuracy": 0.8419746961051848,
|
| 1190 |
+
"eval_loss": 0.5143093466758728,
|
| 1191 |
+
"eval_runtime": 51.9669,
|
| 1192 |
+
"eval_samples_per_second": 77.569,
|
| 1193 |
+
"eval_steps_per_second": 9.698,
|
| 1194 |
+
"step": 6400
|
| 1195 |
+
},
|
| 1196 |
+
{
|
| 1197 |
+
"epoch": 1.422271223814774,
|
| 1198 |
+
"grad_norm": 22.98056411743164,
|
| 1199 |
+
"learning_rate": 1.4448732083792724e-05,
|
| 1200 |
+
"loss": 0.4214,
|
| 1201 |
+
"step": 6450
|
| 1202 |
+
},
|
| 1203 |
+
{
|
| 1204 |
+
"epoch": 1.4332965821389196,
|
| 1205 |
+
"grad_norm": 23.85386848449707,
|
| 1206 |
+
"learning_rate": 1.4173098125689086e-05,
|
| 1207 |
+
"loss": 0.5135,
|
| 1208 |
+
"step": 6500
|
| 1209 |
+
},
|
| 1210 |
+
{
|
| 1211 |
+
"epoch": 1.4443219404630652,
|
| 1212 |
+
"grad_norm": 15.967286109924316,
|
| 1213 |
+
"learning_rate": 1.3897464167585447e-05,
|
| 1214 |
+
"loss": 0.4906,
|
| 1215 |
+
"step": 6550
|
| 1216 |
+
},
|
| 1217 |
+
{
|
| 1218 |
+
"epoch": 1.4553472987872107,
|
| 1219 |
+
"grad_norm": 15.373555183410645,
|
| 1220 |
+
"learning_rate": 1.362183020948181e-05,
|
| 1221 |
+
"loss": 0.4426,
|
| 1222 |
+
"step": 6600
|
| 1223 |
+
},
|
| 1224 |
+
{
|
| 1225 |
+
"epoch": 1.4553472987872107,
|
| 1226 |
+
"eval_accuracy": 0.8347804515008683,
|
| 1227 |
+
"eval_loss": 0.5102077126502991,
|
| 1228 |
+
"eval_runtime": 56.3926,
|
| 1229 |
+
"eval_samples_per_second": 71.481,
|
| 1230 |
+
"eval_steps_per_second": 8.937,
|
| 1231 |
+
"step": 6600
|
| 1232 |
+
},
|
| 1233 |
+
{
|
| 1234 |
+
"epoch": 1.4663726571113562,
|
| 1235 |
+
"grad_norm": 14.563850402832031,
|
| 1236 |
+
"learning_rate": 1.334619625137817e-05,
|
| 1237 |
+
"loss": 0.3955,
|
| 1238 |
+
"step": 6650
|
| 1239 |
+
},
|
| 1240 |
+
{
|
| 1241 |
+
"epoch": 1.4773980154355018,
|
| 1242 |
+
"grad_norm": 15.834396362304688,
|
| 1243 |
+
"learning_rate": 1.3070562293274532e-05,
|
| 1244 |
+
"loss": 0.3435,
|
| 1245 |
+
"step": 6700
|
| 1246 |
+
},
|
| 1247 |
+
{
|
| 1248 |
+
"epoch": 1.4884233737596473,
|
| 1249 |
+
"grad_norm": 9.47835636138916,
|
| 1250 |
+
"learning_rate": 1.2794928335170895e-05,
|
| 1251 |
+
"loss": 0.3906,
|
| 1252 |
+
"step": 6750
|
| 1253 |
+
},
|
| 1254 |
+
{
|
| 1255 |
+
"epoch": 1.4994487320837928,
|
| 1256 |
+
"grad_norm": 20.611888885498047,
|
| 1257 |
+
"learning_rate": 1.2519294377067256e-05,
|
| 1258 |
+
"loss": 0.374,
|
| 1259 |
+
"step": 6800
|
| 1260 |
+
},
|
| 1261 |
+
{
|
| 1262 |
+
"epoch": 1.4994487320837928,
|
| 1263 |
+
"eval_accuracy": 0.8397419995038452,
|
| 1264 |
+
"eval_loss": 0.5277830958366394,
|
| 1265 |
+
"eval_runtime": 59.4632,
|
| 1266 |
+
"eval_samples_per_second": 67.79,
|
| 1267 |
+
"eval_steps_per_second": 8.476,
|
| 1268 |
+
"step": 6800
|
| 1269 |
+
},
|
| 1270 |
+
{
|
| 1271 |
+
"epoch": 1.5104740904079383,
|
| 1272 |
+
"grad_norm": 9.410921096801758,
|
| 1273 |
+
"learning_rate": 1.2243660418963617e-05,
|
| 1274 |
+
"loss": 0.4374,
|
| 1275 |
+
"step": 6850
|
| 1276 |
+
},
|
| 1277 |
+
{
|
| 1278 |
+
"epoch": 1.5214994487320839,
|
| 1279 |
+
"grad_norm": 2.510164737701416,
|
| 1280 |
+
"learning_rate": 1.1968026460859978e-05,
|
| 1281 |
+
"loss": 0.4131,
|
| 1282 |
+
"step": 6900
|
| 1283 |
+
},
|
| 1284 |
+
{
|
| 1285 |
+
"epoch": 1.5325248070562294,
|
| 1286 |
+
"grad_norm": 7.07089376449585,
|
| 1287 |
+
"learning_rate": 1.169239250275634e-05,
|
| 1288 |
+
"loss": 0.4937,
|
| 1289 |
+
"step": 6950
|
| 1290 |
+
},
|
| 1291 |
+
{
|
| 1292 |
+
"epoch": 1.543550165380375,
|
| 1293 |
+
"grad_norm": 12.722711563110352,
|
| 1294 |
+
"learning_rate": 1.1416758544652702e-05,
|
| 1295 |
+
"loss": 0.391,
|
| 1296 |
+
"step": 7000
|
| 1297 |
+
},
|
| 1298 |
+
{
|
| 1299 |
+
"epoch": 1.543550165380375,
|
| 1300 |
+
"eval_accuracy": 0.8429670057057802,
|
| 1301 |
+
"eval_loss": 0.5098406076431274,
|
| 1302 |
+
"eval_runtime": 58.5831,
|
| 1303 |
+
"eval_samples_per_second": 68.808,
|
| 1304 |
+
"eval_steps_per_second": 8.603,
|
| 1305 |
+
"step": 7000
|
| 1306 |
+
},
|
| 1307 |
+
{
|
| 1308 |
+
"epoch": 1.5545755237045205,
|
| 1309 |
+
"grad_norm": 5.899824619293213,
|
| 1310 |
+
"learning_rate": 1.1141124586549064e-05,
|
| 1311 |
+
"loss": 0.4584,
|
| 1312 |
+
"step": 7050
|
| 1313 |
+
},
|
| 1314 |
+
{
|
| 1315 |
+
"epoch": 1.565600882028666,
|
| 1316 |
+
"grad_norm": 20.70318031311035,
|
| 1317 |
+
"learning_rate": 1.0865490628445425e-05,
|
| 1318 |
+
"loss": 0.4132,
|
| 1319 |
+
"step": 7100
|
| 1320 |
+
},
|
| 1321 |
+
{
|
| 1322 |
+
"epoch": 1.5766262403528115,
|
| 1323 |
+
"grad_norm": 16.140867233276367,
|
| 1324 |
+
"learning_rate": 1.0589856670341788e-05,
|
| 1325 |
+
"loss": 0.3984,
|
| 1326 |
+
"step": 7150
|
| 1327 |
+
},
|
| 1328 |
+
{
|
| 1329 |
+
"epoch": 1.587651598676957,
|
| 1330 |
+
"grad_norm": 20.221525192260742,
|
| 1331 |
+
"learning_rate": 1.0314222712238149e-05,
|
| 1332 |
+
"loss": 0.4062,
|
| 1333 |
+
"step": 7200
|
| 1334 |
+
},
|
| 1335 |
+
{
|
| 1336 |
+
"epoch": 1.587651598676957,
|
| 1337 |
+
"eval_accuracy": 0.8509054825105433,
|
| 1338 |
+
"eval_loss": 0.5056183934211731,
|
| 1339 |
+
"eval_runtime": 57.4166,
|
| 1340 |
+
"eval_samples_per_second": 70.206,
|
| 1341 |
+
"eval_steps_per_second": 8.778,
|
| 1342 |
+
"step": 7200
|
| 1343 |
+
},
|
| 1344 |
+
{
|
| 1345 |
+
"epoch": 1.5986769570011026,
|
| 1346 |
+
"grad_norm": 26.3918514251709,
|
| 1347 |
+
"learning_rate": 1.0038588754134508e-05,
|
| 1348 |
+
"loss": 0.4098,
|
| 1349 |
+
"step": 7250
|
| 1350 |
+
},
|
| 1351 |
+
{
|
| 1352 |
+
"epoch": 1.609702315325248,
|
| 1353 |
+
"grad_norm": 25.124940872192383,
|
| 1354 |
+
"learning_rate": 9.762954796030871e-06,
|
| 1355 |
+
"loss": 0.3674,
|
| 1356 |
+
"step": 7300
|
| 1357 |
+
},
|
| 1358 |
+
{
|
| 1359 |
+
"epoch": 1.6207276736493936,
|
| 1360 |
+
"grad_norm": 18.994909286499023,
|
| 1361 |
+
"learning_rate": 9.487320837927232e-06,
|
| 1362 |
+
"loss": 0.4137,
|
| 1363 |
+
"step": 7350
|
| 1364 |
+
},
|
| 1365 |
+
{
|
| 1366 |
+
"epoch": 1.6317530319735392,
|
| 1367 |
+
"grad_norm": 0.34515783190727234,
|
| 1368 |
+
"learning_rate": 9.211686879823594e-06,
|
| 1369 |
+
"loss": 0.3067,
|
| 1370 |
+
"step": 7400
|
| 1371 |
+
},
|
| 1372 |
+
{
|
| 1373 |
+
"epoch": 1.6317530319735392,
|
| 1374 |
+
"eval_accuracy": 0.8447035475068221,
|
| 1375 |
+
"eval_loss": 0.5277951955795288,
|
| 1376 |
+
"eval_runtime": 62.2601,
|
| 1377 |
+
"eval_samples_per_second": 64.745,
|
| 1378 |
+
"eval_steps_per_second": 8.095,
|
| 1379 |
+
"step": 7400
|
| 1380 |
+
},
|
| 1381 |
+
{
|
| 1382 |
+
"epoch": 1.6427783902976847,
|
| 1383 |
+
"grad_norm": 20.627864837646484,
|
| 1384 |
+
"learning_rate": 8.936052921719956e-06,
|
| 1385 |
+
"loss": 0.3772,
|
| 1386 |
+
"step": 7450
|
| 1387 |
+
},
|
| 1388 |
+
{
|
| 1389 |
+
"epoch": 1.6538037486218302,
|
| 1390 |
+
"grad_norm": 12.762669563293457,
|
| 1391 |
+
"learning_rate": 8.660418963616318e-06,
|
| 1392 |
+
"loss": 0.4413,
|
| 1393 |
+
"step": 7500
|
| 1394 |
+
},
|
| 1395 |
+
{
|
| 1396 |
+
"epoch": 1.6648291069459757,
|
| 1397 |
+
"grad_norm": 1.8872555494308472,
|
| 1398 |
+
"learning_rate": 8.38478500551268e-06,
|
| 1399 |
+
"loss": 0.414,
|
| 1400 |
+
"step": 7550
|
| 1401 |
+
},
|
| 1402 |
+
{
|
| 1403 |
+
"epoch": 1.6758544652701213,
|
| 1404 |
+
"grad_norm": 5.339422225952148,
|
| 1405 |
+
"learning_rate": 8.109151047409042e-06,
|
| 1406 |
+
"loss": 0.4419,
|
| 1407 |
+
"step": 7600
|
| 1408 |
+
},
|
| 1409 |
+
{
|
| 1410 |
+
"epoch": 1.6758544652701213,
|
| 1411 |
+
"eval_accuracy": 0.8486727859092037,
|
| 1412 |
+
"eval_loss": 0.508654773235321,
|
| 1413 |
+
"eval_runtime": 60.7174,
|
| 1414 |
+
"eval_samples_per_second": 66.39,
|
| 1415 |
+
"eval_steps_per_second": 8.301,
|
| 1416 |
+
"step": 7600
|
| 1417 |
+
},
|
| 1418 |
+
{
|
| 1419 |
+
"epoch": 1.6868798235942668,
|
| 1420 |
+
"grad_norm": 26.65837287902832,
|
| 1421 |
+
"learning_rate": 7.833517089305403e-06,
|
| 1422 |
+
"loss": 0.4067,
|
| 1423 |
+
"step": 7650
|
| 1424 |
+
},
|
| 1425 |
+
{
|
| 1426 |
+
"epoch": 1.6979051819184123,
|
| 1427 |
+
"grad_norm": 1.583815574645996,
|
| 1428 |
+
"learning_rate": 7.557883131201765e-06,
|
| 1429 |
+
"loss": 0.3626,
|
| 1430 |
+
"step": 7700
|
| 1431 |
+
},
|
| 1432 |
+
{
|
| 1433 |
+
"epoch": 1.7089305402425579,
|
| 1434 |
+
"grad_norm": 7.412702560424805,
|
| 1435 |
+
"learning_rate": 7.282249173098125e-06,
|
| 1436 |
+
"loss": 0.3608,
|
| 1437 |
+
"step": 7750
|
| 1438 |
+
},
|
| 1439 |
+
{
|
| 1440 |
+
"epoch": 1.7199558985667034,
|
| 1441 |
+
"grad_norm": 18.560007095336914,
|
| 1442 |
+
"learning_rate": 7.006615214994487e-06,
|
| 1443 |
+
"loss": 0.4403,
|
| 1444 |
+
"step": 7800
|
| 1445 |
+
},
|
| 1446 |
+
{
|
| 1447 |
+
"epoch": 1.7199558985667034,
|
| 1448 |
+
"eval_accuracy": 0.8521458695112876,
|
| 1449 |
+
"eval_loss": 0.498369038105011,
|
| 1450 |
+
"eval_runtime": 61.2014,
|
| 1451 |
+
"eval_samples_per_second": 65.865,
|
| 1452 |
+
"eval_steps_per_second": 8.235,
|
| 1453 |
+
"step": 7800
|
| 1454 |
+
},
|
| 1455 |
+
{
|
| 1456 |
+
"epoch": 1.730981256890849,
|
| 1457 |
+
"grad_norm": 0.4578012228012085,
|
| 1458 |
+
"learning_rate": 6.730981256890849e-06,
|
| 1459 |
+
"loss": 0.2849,
|
| 1460 |
+
"step": 7850
|
| 1461 |
+
},
|
| 1462 |
+
{
|
| 1463 |
+
"epoch": 1.7420066152149944,
|
| 1464 |
+
"grad_norm": 22.82770347595215,
|
| 1465 |
+
"learning_rate": 6.4553472987872105e-06,
|
| 1466 |
+
"loss": 0.4316,
|
| 1467 |
+
"step": 7900
|
| 1468 |
+
},
|
| 1469 |
+
{
|
| 1470 |
+
"epoch": 1.75303197353914,
|
| 1471 |
+
"grad_norm": 2.745412588119507,
|
| 1472 |
+
"learning_rate": 6.1797133406835725e-06,
|
| 1473 |
+
"loss": 0.4044,
|
| 1474 |
+
"step": 7950
|
| 1475 |
+
},
|
| 1476 |
+
{
|
| 1477 |
+
"epoch": 1.7640573318632855,
|
| 1478 |
+
"grad_norm": 2.724490165710449,
|
| 1479 |
+
"learning_rate": 5.9040793825799346e-06,
|
| 1480 |
+
"loss": 0.4577,
|
| 1481 |
+
"step": 8000
|
| 1482 |
+
},
|
| 1483 |
+
{
|
| 1484 |
+
"epoch": 1.7640573318632855,
|
| 1485 |
+
"eval_accuracy": 0.843463160506078,
|
| 1486 |
+
"eval_loss": 0.5229588150978088,
|
| 1487 |
+
"eval_runtime": 61.6171,
|
| 1488 |
+
"eval_samples_per_second": 65.42,
|
| 1489 |
+
"eval_steps_per_second": 8.18,
|
| 1490 |
+
"step": 8000
|
| 1491 |
+
},
|
| 1492 |
+
{
|
| 1493 |
+
"epoch": 1.775082690187431,
|
| 1494 |
+
"grad_norm": 14.0263671875,
|
| 1495 |
+
"learning_rate": 5.628445424476296e-06,
|
| 1496 |
+
"loss": 0.4339,
|
| 1497 |
+
"step": 8050
|
| 1498 |
+
},
|
| 1499 |
+
{
|
| 1500 |
+
"epoch": 1.7861080485115766,
|
| 1501 |
+
"grad_norm": 10.231954574584961,
|
| 1502 |
+
"learning_rate": 5.352811466372657e-06,
|
| 1503 |
+
"loss": 0.366,
|
| 1504 |
+
"step": 8100
|
| 1505 |
+
},
|
| 1506 |
+
{
|
| 1507 |
+
"epoch": 1.797133406835722,
|
| 1508 |
+
"grad_norm": 7.47442102432251,
|
| 1509 |
+
"learning_rate": 5.077177508269019e-06,
|
| 1510 |
+
"loss": 0.3847,
|
| 1511 |
+
"step": 8150
|
| 1512 |
+
},
|
| 1513 |
+
{
|
| 1514 |
+
"epoch": 1.8081587651598676,
|
| 1515 |
+
"grad_norm": 34.1908073425293,
|
| 1516 |
+
"learning_rate": 4.801543550165381e-06,
|
| 1517 |
+
"loss": 0.3816,
|
| 1518 |
+
"step": 8200
|
| 1519 |
+
},
|
| 1520 |
+
{
|
| 1521 |
+
"epoch": 1.8081587651598676,
|
| 1522 |
+
"eval_accuracy": 0.8516497147109898,
|
| 1523 |
+
"eval_loss": 0.5057588219642639,
|
| 1524 |
+
"eval_runtime": 62.4381,
|
| 1525 |
+
"eval_samples_per_second": 64.56,
|
| 1526 |
+
"eval_steps_per_second": 8.072,
|
| 1527 |
+
"step": 8200
|
| 1528 |
+
},
|
| 1529 |
+
{
|
| 1530 |
+
"epoch": 1.8191841234840131,
|
| 1531 |
+
"grad_norm": 6.812930583953857,
|
| 1532 |
+
"learning_rate": 4.525909592061742e-06,
|
| 1533 |
+
"loss": 0.3876,
|
| 1534 |
+
"step": 8250
|
| 1535 |
+
},
|
| 1536 |
+
{
|
| 1537 |
+
"epoch": 1.8302094818081587,
|
| 1538 |
+
"grad_norm": 6.9469313621521,
|
| 1539 |
+
"learning_rate": 4.250275633958104e-06,
|
| 1540 |
+
"loss": 0.3791,
|
| 1541 |
+
"step": 8300
|
| 1542 |
+
},
|
| 1543 |
+
{
|
| 1544 |
+
"epoch": 1.8412348401323042,
|
| 1545 |
+
"grad_norm": 1.059395432472229,
|
| 1546 |
+
"learning_rate": 3.974641675854465e-06,
|
| 1547 |
+
"loss": 0.3403,
|
| 1548 |
+
"step": 8350
|
| 1549 |
+
},
|
| 1550 |
+
{
|
| 1551 |
+
"epoch": 1.8522601984564497,
|
| 1552 |
+
"grad_norm": 14.005857467651367,
|
| 1553 |
+
"learning_rate": 3.699007717750827e-06,
|
| 1554 |
+
"loss": 0.3351,
|
| 1555 |
+
"step": 8400
|
| 1556 |
+
},
|
| 1557 |
+
{
|
| 1558 |
+
"epoch": 1.8522601984564497,
|
| 1559 |
+
"eval_accuracy": 0.8536343339121806,
|
| 1560 |
+
"eval_loss": 0.49633243680000305,
|
| 1561 |
+
"eval_runtime": 60.6204,
|
| 1562 |
+
"eval_samples_per_second": 66.496,
|
| 1563 |
+
"eval_steps_per_second": 8.314,
|
| 1564 |
+
"step": 8400
|
| 1565 |
+
},
|
| 1566 |
+
{
|
| 1567 |
+
"epoch": 1.8632855567805953,
|
| 1568 |
+
"grad_norm": 10.774523735046387,
|
| 1569 |
+
"learning_rate": 3.4233737596471886e-06,
|
| 1570 |
+
"loss": 0.3543,
|
| 1571 |
+
"step": 8450
|
| 1572 |
+
},
|
| 1573 |
+
{
|
| 1574 |
+
"epoch": 1.8743109151047408,
|
| 1575 |
+
"grad_norm": 17.46559715270996,
|
| 1576 |
+
"learning_rate": 3.1477398015435506e-06,
|
| 1577 |
+
"loss": 0.4039,
|
| 1578 |
+
"step": 8500
|
| 1579 |
+
},
|
| 1580 |
+
{
|
| 1581 |
+
"epoch": 1.8853362734288863,
|
| 1582 |
+
"grad_norm": 8.823070526123047,
|
| 1583 |
+
"learning_rate": 2.8721058434399118e-06,
|
| 1584 |
+
"loss": 0.4881,
|
| 1585 |
+
"step": 8550
|
| 1586 |
+
},
|
| 1587 |
+
{
|
| 1588 |
+
"epoch": 1.8963616317530319,
|
| 1589 |
+
"grad_norm": 12.282500267028809,
|
| 1590 |
+
"learning_rate": 2.596471885336274e-06,
|
| 1591 |
+
"loss": 0.3894,
|
| 1592 |
+
"step": 8600
|
| 1593 |
+
},
|
| 1594 |
+
{
|
| 1595 |
+
"epoch": 1.8963616317530319,
|
| 1596 |
+
"eval_accuracy": 0.8551227983130737,
|
| 1597 |
+
"eval_loss": 0.48355239629745483,
|
| 1598 |
+
"eval_runtime": 61.0535,
|
| 1599 |
+
"eval_samples_per_second": 66.024,
|
| 1600 |
+
"eval_steps_per_second": 8.255,
|
| 1601 |
+
"step": 8600
|
| 1602 |
+
},
|
| 1603 |
+
{
|
| 1604 |
+
"epoch": 1.9073869900771774,
|
| 1605 |
+
"grad_norm": 8.131113052368164,
|
| 1606 |
+
"learning_rate": 2.320837927232635e-06,
|
| 1607 |
+
"loss": 0.4065,
|
| 1608 |
+
"step": 8650
|
| 1609 |
+
},
|
| 1610 |
+
{
|
| 1611 |
+
"epoch": 1.918412348401323,
|
| 1612 |
+
"grad_norm": 0.6533594727516174,
|
| 1613 |
+
"learning_rate": 2.045203969128997e-06,
|
| 1614 |
+
"loss": 0.3384,
|
| 1615 |
+
"step": 8700
|
| 1616 |
+
},
|
| 1617 |
+
{
|
| 1618 |
+
"epoch": 1.9294377067254684,
|
| 1619 |
+
"grad_norm": 8.164648056030273,
|
| 1620 |
+
"learning_rate": 1.7695700110253584e-06,
|
| 1621 |
+
"loss": 0.3509,
|
| 1622 |
+
"step": 8750
|
| 1623 |
+
},
|
| 1624 |
+
{
|
| 1625 |
+
"epoch": 1.940463065049614,
|
| 1626 |
+
"grad_norm": 4.368854522705078,
|
| 1627 |
+
"learning_rate": 1.49393605292172e-06,
|
| 1628 |
+
"loss": 0.3611,
|
| 1629 |
+
"step": 8800
|
| 1630 |
+
},
|
| 1631 |
+
{
|
| 1632 |
+
"epoch": 1.940463065049614,
|
| 1633 |
+
"eval_accuracy": 0.8558670305135202,
|
| 1634 |
+
"eval_loss": 0.47785690426826477,
|
| 1635 |
+
"eval_runtime": 61.2031,
|
| 1636 |
+
"eval_samples_per_second": 65.863,
|
| 1637 |
+
"eval_steps_per_second": 8.235,
|
| 1638 |
+
"step": 8800
|
| 1639 |
+
},
|
| 1640 |
+
{
|
| 1641 |
+
"epoch": 1.9514884233737595,
|
| 1642 |
+
"grad_norm": 7.265945911407471,
|
| 1643 |
+
"learning_rate": 1.2183020948180816e-06,
|
| 1644 |
+
"loss": 0.4892,
|
| 1645 |
+
"step": 8850
|
| 1646 |
+
},
|
| 1647 |
+
{
|
| 1648 |
+
"epoch": 1.962513781697905,
|
| 1649 |
+
"grad_norm": 2.180917978286743,
|
| 1650 |
+
"learning_rate": 9.426681367144433e-07,
|
| 1651 |
+
"loss": 0.4405,
|
| 1652 |
+
"step": 8900
|
| 1653 |
+
},
|
| 1654 |
+
{
|
| 1655 |
+
"epoch": 1.9735391400220506,
|
| 1656 |
+
"grad_norm": 11.077986717224121,
|
| 1657 |
+
"learning_rate": 6.670341786108048e-07,
|
| 1658 |
+
"loss": 0.3206,
|
| 1659 |
+
"step": 8950
|
| 1660 |
+
},
|
| 1661 |
+
{
|
| 1662 |
+
"epoch": 1.9845644983461963,
|
| 1663 |
+
"grad_norm": 20.9344482421875,
|
| 1664 |
+
"learning_rate": 3.914002205071665e-07,
|
| 1665 |
+
"loss": 0.3568,
|
| 1666 |
+
"step": 9000
|
| 1667 |
+
},
|
| 1668 |
+
{
|
| 1669 |
+
"epoch": 1.9845644983461963,
|
| 1670 |
+
"eval_accuracy": 0.8558670305135202,
|
| 1671 |
+
"eval_loss": 0.474208801984787,
|
| 1672 |
+
"eval_runtime": 64.6572,
|
| 1673 |
+
"eval_samples_per_second": 62.344,
|
| 1674 |
+
"eval_steps_per_second": 7.795,
|
| 1675 |
+
"step": 9000
|
| 1676 |
+
},
|
| 1677 |
+
{
|
| 1678 |
+
"epoch": 1.9955898566703418,
|
| 1679 |
+
"grad_norm": 11.30726432800293,
|
| 1680 |
+
"learning_rate": 1.1576626240352813e-07,
|
| 1681 |
+
"loss": 0.4508,
|
| 1682 |
+
"step": 9050
|
| 1683 |
+
}
|
| 1684 |
+
],
|
| 1685 |
+
"logging_steps": 50,
|
| 1686 |
+
"max_steps": 9070,
|
| 1687 |
+
"num_input_tokens_seen": 0,
|
| 1688 |
+
"num_train_epochs": 2,
|
| 1689 |
+
"save_steps": 200,
|
| 1690 |
+
"stateful_callbacks": {
|
| 1691 |
+
"TrainerControl": {
|
| 1692 |
+
"args": {
|
| 1693 |
+
"should_epoch_stop": false,
|
| 1694 |
+
"should_evaluate": false,
|
| 1695 |
+
"should_log": false,
|
| 1696 |
+
"should_save": true,
|
| 1697 |
+
"should_training_stop": true
|
| 1698 |
+
},
|
| 1699 |
+
"attributes": {}
|
| 1700 |
+
}
|
| 1701 |
+
},
|
| 1702 |
+
"total_flos": 2402970222720000.0,
|
| 1703 |
+
"train_batch_size": 8,
|
| 1704 |
+
"trial_name": null,
|
| 1705 |
+
"trial_params": null
|
| 1706 |
+
}
|
training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:4526d5003b0c4afbe7571e8f21db2f8331eb10f0d2c55963f08ea7a0630de03c
|
| 3 |
+
size 5713
|