Upload folder using huggingface_hub
Browse files- bert/config.json +4 -1
- bert/model.safetensors +1 -1
- checkpoints/checkpoint-1774/model.safetensors +3 -0
- checkpoints/checkpoint-1774/optimizer.pt +3 -0
- checkpoints/checkpoint-1774/rng_state.pth +0 -0
- checkpoints/checkpoint-1774/scheduler.pt +3 -0
- checkpoints/checkpoint-1774/trainer_state.json +1299 -0
- checkpoints/checkpoint-1774/training_args.bin +3 -0
- checkpoints/checkpoint-2661/model.safetensors +3 -0
- checkpoints/checkpoint-2661/optimizer.pt +3 -0
- checkpoints/checkpoint-2661/rng_state.pth +0 -0
- checkpoints/checkpoint-2661/scheduler.pt +3 -0
- checkpoints/checkpoint-2661/trainer_state.json +1935 -0
- checkpoints/checkpoint-2661/training_args.bin +3 -0
- classifier_heads.pt +1 -1
- tokenizer/tokenizer_config.json +3 -45
bert/config.json
CHANGED
|
@@ -4,9 +4,11 @@
|
|
| 4 |
"DistilBertModel"
|
| 5 |
],
|
| 6 |
"attention_dropout": 0.1,
|
|
|
|
| 7 |
"dim": 768,
|
| 8 |
"dropout": 0.1,
|
| 9 |
"dtype": "float32",
|
|
|
|
| 10 |
"hidden_dim": 3072,
|
| 11 |
"initializer_range": 0.02,
|
| 12 |
"max_position_embeddings": 512,
|
|
@@ -18,6 +20,7 @@
|
|
| 18 |
"seq_classif_dropout": 0.2,
|
| 19 |
"sinusoidal_pos_embds": false,
|
| 20 |
"tie_weights_": true,
|
| 21 |
-
"
|
|
|
|
| 22 |
"vocab_size": 30522
|
| 23 |
}
|
|
|
|
| 4 |
"DistilBertModel"
|
| 5 |
],
|
| 6 |
"attention_dropout": 0.1,
|
| 7 |
+
"bos_token_id": null,
|
| 8 |
"dim": 768,
|
| 9 |
"dropout": 0.1,
|
| 10 |
"dtype": "float32",
|
| 11 |
+
"eos_token_id": null,
|
| 12 |
"hidden_dim": 3072,
|
| 13 |
"initializer_range": 0.02,
|
| 14 |
"max_position_embeddings": 512,
|
|
|
|
| 20 |
"seq_classif_dropout": 0.2,
|
| 21 |
"sinusoidal_pos_embds": false,
|
| 22 |
"tie_weights_": true,
|
| 23 |
+
"tie_word_embeddings": true,
|
| 24 |
+
"transformers_version": "5.0.0",
|
| 25 |
"vocab_size": 30522
|
| 26 |
}
|
bert/model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 265462608
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:86ceb9966f29819738922ed9d59b773374fb31f2743f8592838bbe28d161dbe8
|
| 3 |
size 265462608
|
checkpoints/checkpoint-1774/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:94ba8ae64fa8947f62f18edac97c1b3bf9920d8b54d611e2970c5f22ca914336
|
| 3 |
+
size 265475800
|
checkpoints/checkpoint-1774/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:0b022d4d31f613db94c0d2d137d1a573815c4e8406441d80e39d5548be7d9b60
|
| 3 |
+
size 531012363
|
checkpoints/checkpoint-1774/rng_state.pth
ADDED
|
Binary file (14.5 kB). View file
|
|
|
checkpoints/checkpoint-1774/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ff32ab76ce831d2a432ef5023b44ccaba68ccca07b93466400e486ebcd620a96
|
| 3 |
+
size 1465
|
checkpoints/checkpoint-1774/trainer_state.json
ADDED
|
@@ -0,0 +1,1299 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": 887,
|
| 3 |
+
"best_metric": 0.9984485190409027,
|
| 4 |
+
"best_model_checkpoint": "models/classifier_model/checkpoints/checkpoint-887",
|
| 5 |
+
"epoch": 2.0,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 1774,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.011273957158962795,
|
| 14 |
+
"grad_norm": 6.372345447540283,
|
| 15 |
+
"learning_rate": 3.6000000000000003e-06,
|
| 16 |
+
"loss": 1.4694642066955566,
|
| 17 |
+
"step": 10
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.02254791431792559,
|
| 21 |
+
"grad_norm": 3.5241456031799316,
|
| 22 |
+
"learning_rate": 7.600000000000001e-06,
|
| 23 |
+
"loss": 1.3229194641113282,
|
| 24 |
+
"step": 20
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.033821871476888386,
|
| 28 |
+
"grad_norm": 4.042177677154541,
|
| 29 |
+
"learning_rate": 1.16e-05,
|
| 30 |
+
"loss": 1.0544840812683105,
|
| 31 |
+
"step": 30
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.04509582863585118,
|
| 35 |
+
"grad_norm": 3.2139792442321777,
|
| 36 |
+
"learning_rate": 1.5600000000000003e-05,
|
| 37 |
+
"loss": 0.7191516399383545,
|
| 38 |
+
"step": 40
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.05636978579481398,
|
| 42 |
+
"grad_norm": 4.926011085510254,
|
| 43 |
+
"learning_rate": 1.9600000000000002e-05,
|
| 44 |
+
"loss": 0.40511221885681153,
|
| 45 |
+
"step": 50
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.06764374295377677,
|
| 49 |
+
"grad_norm": 4.614091873168945,
|
| 50 |
+
"learning_rate": 1.9931060896208353e-05,
|
| 51 |
+
"loss": 0.17188742160797119,
|
| 52 |
+
"step": 60
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 0.07891770011273957,
|
| 56 |
+
"grad_norm": 0.4526596665382385,
|
| 57 |
+
"learning_rate": 1.9854461891995405e-05,
|
| 58 |
+
"loss": 0.08630315661430359,
|
| 59 |
+
"step": 70
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 0.09019165727170236,
|
| 63 |
+
"grad_norm": 0.4451746940612793,
|
| 64 |
+
"learning_rate": 1.977786288778246e-05,
|
| 65 |
+
"loss": 0.03049740493297577,
|
| 66 |
+
"step": 80
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 0.10146561443066517,
|
| 70 |
+
"grad_norm": 0.13958168029785156,
|
| 71 |
+
"learning_rate": 1.9701263883569516e-05,
|
| 72 |
+
"loss": 0.012567885220050812,
|
| 73 |
+
"step": 90
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 0.11273957158962795,
|
| 77 |
+
"grad_norm": 0.06256264448165894,
|
| 78 |
+
"learning_rate": 1.962466487935657e-05,
|
| 79 |
+
"loss": 0.06832309961318969,
|
| 80 |
+
"step": 100
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.12401352874859076,
|
| 84 |
+
"grad_norm": 0.048310115933418274,
|
| 85 |
+
"learning_rate": 1.9548065875143624e-05,
|
| 86 |
+
"loss": 0.011482110619544983,
|
| 87 |
+
"step": 110
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 0.13528748590755355,
|
| 91 |
+
"grad_norm": 0.12978595495224,
|
| 92 |
+
"learning_rate": 1.947146687093068e-05,
|
| 93 |
+
"loss": 0.04678139984607697,
|
| 94 |
+
"step": 120
|
| 95 |
+
},
|
| 96 |
+
{
|
| 97 |
+
"epoch": 0.14656144306651633,
|
| 98 |
+
"grad_norm": 2.551255226135254,
|
| 99 |
+
"learning_rate": 1.9394867866717735e-05,
|
| 100 |
+
"loss": 0.11264426708221435,
|
| 101 |
+
"step": 130
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"epoch": 0.15783540022547915,
|
| 105 |
+
"grad_norm": 0.3405976891517639,
|
| 106 |
+
"learning_rate": 1.931826886250479e-05,
|
| 107 |
+
"loss": 0.031431844830513,
|
| 108 |
+
"step": 140
|
| 109 |
+
},
|
| 110 |
+
{
|
| 111 |
+
"epoch": 0.16910935738444194,
|
| 112 |
+
"grad_norm": 0.030345242470502853,
|
| 113 |
+
"learning_rate": 1.9241669858291846e-05,
|
| 114 |
+
"loss": 0.002859099581837654,
|
| 115 |
+
"step": 150
|
| 116 |
+
},
|
| 117 |
+
{
|
| 118 |
+
"epoch": 0.18038331454340473,
|
| 119 |
+
"grad_norm": 0.11610831320285797,
|
| 120 |
+
"learning_rate": 1.9165070854078898e-05,
|
| 121 |
+
"loss": 0.0036297302693128588,
|
| 122 |
+
"step": 160
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 0.19165727170236754,
|
| 126 |
+
"grad_norm": 0.0447811484336853,
|
| 127 |
+
"learning_rate": 1.9088471849865953e-05,
|
| 128 |
+
"loss": 0.003958044201135635,
|
| 129 |
+
"step": 170
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 0.20293122886133033,
|
| 133 |
+
"grad_norm": 0.02229328267276287,
|
| 134 |
+
"learning_rate": 1.901187284565301e-05,
|
| 135 |
+
"loss": 0.002498266473412514,
|
| 136 |
+
"step": 180
|
| 137 |
+
},
|
| 138 |
+
{
|
| 139 |
+
"epoch": 0.21420518602029312,
|
| 140 |
+
"grad_norm": 0.016443341970443726,
|
| 141 |
+
"learning_rate": 1.8935273841440064e-05,
|
| 142 |
+
"loss": 0.0019413232803344727,
|
| 143 |
+
"step": 190
|
| 144 |
+
},
|
| 145 |
+
{
|
| 146 |
+
"epoch": 0.2254791431792559,
|
| 147 |
+
"grad_norm": 0.025817960500717163,
|
| 148 |
+
"learning_rate": 1.8858674837227116e-05,
|
| 149 |
+
"loss": 0.10970286130905152,
|
| 150 |
+
"step": 200
|
| 151 |
+
},
|
| 152 |
+
{
|
| 153 |
+
"epoch": 0.23675310033821872,
|
| 154 |
+
"grad_norm": 0.030361849814653397,
|
| 155 |
+
"learning_rate": 1.8782075833014172e-05,
|
| 156 |
+
"loss": 0.034755557775497437,
|
| 157 |
+
"step": 210
|
| 158 |
+
},
|
| 159 |
+
{
|
| 160 |
+
"epoch": 0.2480270574971815,
|
| 161 |
+
"grad_norm": 0.6030261516571045,
|
| 162 |
+
"learning_rate": 1.8705476828801227e-05,
|
| 163 |
+
"loss": 0.0343275398015976,
|
| 164 |
+
"step": 220
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 0.2593010146561443,
|
| 168 |
+
"grad_norm": 0.016219334676861763,
|
| 169 |
+
"learning_rate": 1.8628877824588283e-05,
|
| 170 |
+
"loss": 0.020017309486865996,
|
| 171 |
+
"step": 230
|
| 172 |
+
},
|
| 173 |
+
{
|
| 174 |
+
"epoch": 0.2705749718151071,
|
| 175 |
+
"grad_norm": 0.013941165059804916,
|
| 176 |
+
"learning_rate": 1.8552278820375335e-05,
|
| 177 |
+
"loss": 0.0028075359761714934,
|
| 178 |
+
"step": 240
|
| 179 |
+
},
|
| 180 |
+
{
|
| 181 |
+
"epoch": 0.2818489289740699,
|
| 182 |
+
"grad_norm": 1.2171980142593384,
|
| 183 |
+
"learning_rate": 1.847567981616239e-05,
|
| 184 |
+
"loss": 0.0019562991335988046,
|
| 185 |
+
"step": 250
|
| 186 |
+
},
|
| 187 |
+
{
|
| 188 |
+
"epoch": 0.29312288613303267,
|
| 189 |
+
"grad_norm": 0.02682333253324032,
|
| 190 |
+
"learning_rate": 1.8399080811949446e-05,
|
| 191 |
+
"loss": 0.0012471863068640233,
|
| 192 |
+
"step": 260
|
| 193 |
+
},
|
| 194 |
+
{
|
| 195 |
+
"epoch": 0.3043968432919955,
|
| 196 |
+
"grad_norm": 0.013817372731864452,
|
| 197 |
+
"learning_rate": 1.83224818077365e-05,
|
| 198 |
+
"loss": 0.060174965858459474,
|
| 199 |
+
"step": 270
|
| 200 |
+
},
|
| 201 |
+
{
|
| 202 |
+
"epoch": 0.3156708004509583,
|
| 203 |
+
"grad_norm": 0.023286759853363037,
|
| 204 |
+
"learning_rate": 1.8245882803523557e-05,
|
| 205 |
+
"loss": 0.0016331907361745834,
|
| 206 |
+
"step": 280
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 0.3269447576099211,
|
| 210 |
+
"grad_norm": 13.508488655090332,
|
| 211 |
+
"learning_rate": 1.8169283799310612e-05,
|
| 212 |
+
"loss": 0.010956646502017975,
|
| 213 |
+
"step": 290
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"epoch": 0.3382187147688839,
|
| 217 |
+
"grad_norm": 0.012220031581819057,
|
| 218 |
+
"learning_rate": 1.8092684795097664e-05,
|
| 219 |
+
"loss": 0.04829314053058624,
|
| 220 |
+
"step": 300
|
| 221 |
+
},
|
| 222 |
+
{
|
| 223 |
+
"epoch": 0.34949267192784667,
|
| 224 |
+
"grad_norm": 21.441791534423828,
|
| 225 |
+
"learning_rate": 1.801608579088472e-05,
|
| 226 |
+
"loss": 0.007926396280527114,
|
| 227 |
+
"step": 310
|
| 228 |
+
},
|
| 229 |
+
{
|
| 230 |
+
"epoch": 0.36076662908680945,
|
| 231 |
+
"grad_norm": 0.8508244156837463,
|
| 232 |
+
"learning_rate": 1.7939486786671775e-05,
|
| 233 |
+
"loss": 0.0012166524305939675,
|
| 234 |
+
"step": 320
|
| 235 |
+
},
|
| 236 |
+
{
|
| 237 |
+
"epoch": 0.37204058624577224,
|
| 238 |
+
"grad_norm": 0.01957777701318264,
|
| 239 |
+
"learning_rate": 1.7862887782458827e-05,
|
| 240 |
+
"loss": 0.0012863848358392715,
|
| 241 |
+
"step": 330
|
| 242 |
+
},
|
| 243 |
+
{
|
| 244 |
+
"epoch": 0.3833145434047351,
|
| 245 |
+
"grad_norm": 0.013677588663995266,
|
| 246 |
+
"learning_rate": 1.7786288778245883e-05,
|
| 247 |
+
"loss": 0.0008751305751502514,
|
| 248 |
+
"step": 340
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 0.3945885005636979,
|
| 252 |
+
"grad_norm": 0.02119549550116062,
|
| 253 |
+
"learning_rate": 1.7709689774032938e-05,
|
| 254 |
+
"loss": 0.05260458588600159,
|
| 255 |
+
"step": 350
|
| 256 |
+
},
|
| 257 |
+
{
|
| 258 |
+
"epoch": 0.40586245772266066,
|
| 259 |
+
"grad_norm": 0.02179444581270218,
|
| 260 |
+
"learning_rate": 1.7633090769819994e-05,
|
| 261 |
+
"loss": 0.01002354547381401,
|
| 262 |
+
"step": 360
|
| 263 |
+
},
|
| 264 |
+
{
|
| 265 |
+
"epoch": 0.41713641488162345,
|
| 266 |
+
"grad_norm": 0.01586312986910343,
|
| 267 |
+
"learning_rate": 1.755649176560705e-05,
|
| 268 |
+
"loss": 0.0009520040825009346,
|
| 269 |
+
"step": 370
|
| 270 |
+
},
|
| 271 |
+
{
|
| 272 |
+
"epoch": 0.42841037204058624,
|
| 273 |
+
"grad_norm": 0.00843344908207655,
|
| 274 |
+
"learning_rate": 1.7479892761394105e-05,
|
| 275 |
+
"loss": 0.0007581338752061129,
|
| 276 |
+
"step": 380
|
| 277 |
+
},
|
| 278 |
+
{
|
| 279 |
+
"epoch": 0.43968432919954903,
|
| 280 |
+
"grad_norm": 0.0289614200592041,
|
| 281 |
+
"learning_rate": 1.740329375718116e-05,
|
| 282 |
+
"loss": 0.08814518451690674,
|
| 283 |
+
"step": 390
|
| 284 |
+
},
|
| 285 |
+
{
|
| 286 |
+
"epoch": 0.4509582863585118,
|
| 287 |
+
"grad_norm": 0.20195287466049194,
|
| 288 |
+
"learning_rate": 1.7326694752968215e-05,
|
| 289 |
+
"loss": 0.02897307574748993,
|
| 290 |
+
"step": 400
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 0.46223224351747466,
|
| 294 |
+
"grad_norm": 0.2169051319360733,
|
| 295 |
+
"learning_rate": 1.7250095748755268e-05,
|
| 296 |
+
"loss": 0.01624799370765686,
|
| 297 |
+
"step": 410
|
| 298 |
+
},
|
| 299 |
+
{
|
| 300 |
+
"epoch": 0.47350620067643745,
|
| 301 |
+
"grad_norm": 0.009392499923706055,
|
| 302 |
+
"learning_rate": 1.7173496744542323e-05,
|
| 303 |
+
"loss": 0.036326673626899716,
|
| 304 |
+
"step": 420
|
| 305 |
+
},
|
| 306 |
+
{
|
| 307 |
+
"epoch": 0.48478015783540024,
|
| 308 |
+
"grad_norm": 0.008052267134189606,
|
| 309 |
+
"learning_rate": 1.7096897740329375e-05,
|
| 310 |
+
"loss": 0.0008220113813877106,
|
| 311 |
+
"step": 430
|
| 312 |
+
},
|
| 313 |
+
{
|
| 314 |
+
"epoch": 0.496054114994363,
|
| 315 |
+
"grad_norm": 0.09780994057655334,
|
| 316 |
+
"learning_rate": 1.702029873611643e-05,
|
| 317 |
+
"loss": 0.0006682470440864563,
|
| 318 |
+
"step": 440
|
| 319 |
+
},
|
| 320 |
+
{
|
| 321 |
+
"epoch": 0.5073280721533259,
|
| 322 |
+
"grad_norm": 0.09860614687204361,
|
| 323 |
+
"learning_rate": 1.6943699731903486e-05,
|
| 324 |
+
"loss": 0.0006374984513968229,
|
| 325 |
+
"step": 450
|
| 326 |
+
},
|
| 327 |
+
{
|
| 328 |
+
"epoch": 0.5186020293122886,
|
| 329 |
+
"grad_norm": 0.007464884780347347,
|
| 330 |
+
"learning_rate": 1.686710072769054e-05,
|
| 331 |
+
"loss": 0.0005574076902121305,
|
| 332 |
+
"step": 460
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 0.5298759864712514,
|
| 336 |
+
"grad_norm": 0.0052289823070168495,
|
| 337 |
+
"learning_rate": 1.6790501723477597e-05,
|
| 338 |
+
"loss": 0.0011871119029819965,
|
| 339 |
+
"step": 470
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 0.5411499436302142,
|
| 343 |
+
"grad_norm": 0.03991198539733887,
|
| 344 |
+
"learning_rate": 1.6713902719264652e-05,
|
| 345 |
+
"loss": 0.004352982714772224,
|
| 346 |
+
"step": 480
|
| 347 |
+
},
|
| 348 |
+
{
|
| 349 |
+
"epoch": 0.552423900789177,
|
| 350 |
+
"grad_norm": 0.005847420543432236,
|
| 351 |
+
"learning_rate": 1.6637303715051705e-05,
|
| 352 |
+
"loss": 0.023742355406284332,
|
| 353 |
+
"step": 490
|
| 354 |
+
},
|
| 355 |
+
{
|
| 356 |
+
"epoch": 0.5636978579481398,
|
| 357 |
+
"grad_norm": 0.007102746516466141,
|
| 358 |
+
"learning_rate": 1.656070471083876e-05,
|
| 359 |
+
"loss": 0.0033394105732440947,
|
| 360 |
+
"step": 500
|
| 361 |
+
},
|
| 362 |
+
{
|
| 363 |
+
"epoch": 0.5749718151071026,
|
| 364 |
+
"grad_norm": 0.00974489375948906,
|
| 365 |
+
"learning_rate": 1.6484105706625815e-05,
|
| 366 |
+
"loss": 0.0006546341348439455,
|
| 367 |
+
"step": 510
|
| 368 |
+
},
|
| 369 |
+
{
|
| 370 |
+
"epoch": 0.5862457722660653,
|
| 371 |
+
"grad_norm": 0.011177345179021358,
|
| 372 |
+
"learning_rate": 1.640750670241287e-05,
|
| 373 |
+
"loss": 0.0005123908165842294,
|
| 374 |
+
"step": 520
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 0.5975197294250282,
|
| 378 |
+
"grad_norm": 0.45112329721450806,
|
| 379 |
+
"learning_rate": 1.6330907698199926e-05,
|
| 380 |
+
"loss": 0.06648756265640259,
|
| 381 |
+
"step": 530
|
| 382 |
+
},
|
| 383 |
+
{
|
| 384 |
+
"epoch": 0.608793686583991,
|
| 385 |
+
"grad_norm": 0.060878559947013855,
|
| 386 |
+
"learning_rate": 1.625430869398698e-05,
|
| 387 |
+
"loss": 0.03919784724712372,
|
| 388 |
+
"step": 540
|
| 389 |
+
},
|
| 390 |
+
{
|
| 391 |
+
"epoch": 0.6200676437429538,
|
| 392 |
+
"grad_norm": 0.01804164983332157,
|
| 393 |
+
"learning_rate": 1.6177709689774034e-05,
|
| 394 |
+
"loss": 0.0012153918854892255,
|
| 395 |
+
"step": 550
|
| 396 |
+
},
|
| 397 |
+
{
|
| 398 |
+
"epoch": 0.6313416009019166,
|
| 399 |
+
"grad_norm": 0.0058420985005795956,
|
| 400 |
+
"learning_rate": 1.610111068556109e-05,
|
| 401 |
+
"loss": 0.0008219529874622822,
|
| 402 |
+
"step": 560
|
| 403 |
+
},
|
| 404 |
+
{
|
| 405 |
+
"epoch": 0.6426155580608793,
|
| 406 |
+
"grad_norm": 0.011106841266155243,
|
| 407 |
+
"learning_rate": 1.602451168134814e-05,
|
| 408 |
+
"loss": 0.0005111472215503454,
|
| 409 |
+
"step": 570
|
| 410 |
+
},
|
| 411 |
+
{
|
| 412 |
+
"epoch": 0.6538895152198422,
|
| 413 |
+
"grad_norm": 0.005702407564967871,
|
| 414 |
+
"learning_rate": 1.5947912677135197e-05,
|
| 415 |
+
"loss": 0.00047746491618454455,
|
| 416 |
+
"step": 580
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 0.6651634723788049,
|
| 420 |
+
"grad_norm": 0.006169776897877455,
|
| 421 |
+
"learning_rate": 1.5871313672922252e-05,
|
| 422 |
+
"loss": 0.0448866218328476,
|
| 423 |
+
"step": 590
|
| 424 |
+
},
|
| 425 |
+
{
|
| 426 |
+
"epoch": 0.6764374295377678,
|
| 427 |
+
"grad_norm": 15.104083061218262,
|
| 428 |
+
"learning_rate": 1.5794714668709308e-05,
|
| 429 |
+
"loss": 0.006607178598642349,
|
| 430 |
+
"step": 600
|
| 431 |
+
},
|
| 432 |
+
{
|
| 433 |
+
"epoch": 0.6877113866967306,
|
| 434 |
+
"grad_norm": 0.004718424286693335,
|
| 435 |
+
"learning_rate": 1.5718115664496363e-05,
|
| 436 |
+
"loss": 0.0009398681111633777,
|
| 437 |
+
"step": 610
|
| 438 |
+
},
|
| 439 |
+
{
|
| 440 |
+
"epoch": 0.6989853438556933,
|
| 441 |
+
"grad_norm": 0.033868566155433655,
|
| 442 |
+
"learning_rate": 1.564151666028342e-05,
|
| 443 |
+
"loss": 0.11185940504074096,
|
| 444 |
+
"step": 620
|
| 445 |
+
},
|
| 446 |
+
{
|
| 447 |
+
"epoch": 0.7102593010146562,
|
| 448 |
+
"grad_norm": 0.06708049029111862,
|
| 449 |
+
"learning_rate": 1.5564917656070474e-05,
|
| 450 |
+
"loss": 0.025441676378250122,
|
| 451 |
+
"step": 630
|
| 452 |
+
},
|
| 453 |
+
{
|
| 454 |
+
"epoch": 0.7215332581736189,
|
| 455 |
+
"grad_norm": 0.04639710858464241,
|
| 456 |
+
"learning_rate": 1.5488318651857526e-05,
|
| 457 |
+
"loss": 0.04628426432609558,
|
| 458 |
+
"step": 640
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 0.7328072153325818,
|
| 462 |
+
"grad_norm": 0.012839795090258121,
|
| 463 |
+
"learning_rate": 1.5411719647644582e-05,
|
| 464 |
+
"loss": 0.001783803105354309,
|
| 465 |
+
"step": 650
|
| 466 |
+
},
|
| 467 |
+
{
|
| 468 |
+
"epoch": 0.7440811724915445,
|
| 469 |
+
"grad_norm": 0.29861995577812195,
|
| 470 |
+
"learning_rate": 1.5335120643431637e-05,
|
| 471 |
+
"loss": 0.046310481429100034,
|
| 472 |
+
"step": 660
|
| 473 |
+
},
|
| 474 |
+
{
|
| 475 |
+
"epoch": 0.7553551296505073,
|
| 476 |
+
"grad_norm": 0.23833876848220825,
|
| 477 |
+
"learning_rate": 1.5258521639218691e-05,
|
| 478 |
+
"loss": 0.03593424260616303,
|
| 479 |
+
"step": 670
|
| 480 |
+
},
|
| 481 |
+
{
|
| 482 |
+
"epoch": 0.7666290868094702,
|
| 483 |
+
"grad_norm": 0.016054296866059303,
|
| 484 |
+
"learning_rate": 1.5181922635005747e-05,
|
| 485 |
+
"loss": 0.01490943431854248,
|
| 486 |
+
"step": 680
|
| 487 |
+
},
|
| 488 |
+
{
|
| 489 |
+
"epoch": 0.7779030439684329,
|
| 490 |
+
"grad_norm": 25.791549682617188,
|
| 491 |
+
"learning_rate": 1.51053236307928e-05,
|
| 492 |
+
"loss": 0.057176172733306885,
|
| 493 |
+
"step": 690
|
| 494 |
+
},
|
| 495 |
+
{
|
| 496 |
+
"epoch": 0.7891770011273957,
|
| 497 |
+
"grad_norm": 3.5461034774780273,
|
| 498 |
+
"learning_rate": 1.5028724626579856e-05,
|
| 499 |
+
"loss": 0.07008817791938782,
|
| 500 |
+
"step": 700
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 0.8004509582863585,
|
| 504 |
+
"grad_norm": 8.450429916381836,
|
| 505 |
+
"learning_rate": 1.4952125622366911e-05,
|
| 506 |
+
"loss": 0.081763756275177,
|
| 507 |
+
"step": 710
|
| 508 |
+
},
|
| 509 |
+
{
|
| 510 |
+
"epoch": 0.8117249154453213,
|
| 511 |
+
"grad_norm": 0.015128090977668762,
|
| 512 |
+
"learning_rate": 1.4875526618153965e-05,
|
| 513 |
+
"loss": 0.005710385367274284,
|
| 514 |
+
"step": 720
|
| 515 |
+
},
|
| 516 |
+
{
|
| 517 |
+
"epoch": 0.8229988726042841,
|
| 518 |
+
"grad_norm": 0.006377412937581539,
|
| 519 |
+
"learning_rate": 1.479892761394102e-05,
|
| 520 |
+
"loss": 0.0007186648901551962,
|
| 521 |
+
"step": 730
|
| 522 |
+
},
|
| 523 |
+
{
|
| 524 |
+
"epoch": 0.8342728297632469,
|
| 525 |
+
"grad_norm": 0.009410051628947258,
|
| 526 |
+
"learning_rate": 1.4722328609728074e-05,
|
| 527 |
+
"loss": 0.0129980206489563,
|
| 528 |
+
"step": 740
|
| 529 |
+
},
|
| 530 |
+
{
|
| 531 |
+
"epoch": 0.8455467869222097,
|
| 532 |
+
"grad_norm": 0.00972823053598404,
|
| 533 |
+
"learning_rate": 1.4645729605515128e-05,
|
| 534 |
+
"loss": 0.0006546798162162304,
|
| 535 |
+
"step": 750
|
| 536 |
+
},
|
| 537 |
+
{
|
| 538 |
+
"epoch": 0.8568207440811725,
|
| 539 |
+
"grad_norm": 0.013078595511615276,
|
| 540 |
+
"learning_rate": 1.4569130601302184e-05,
|
| 541 |
+
"loss": 0.058796089887619016,
|
| 542 |
+
"step": 760
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 0.8680947012401353,
|
| 546 |
+
"grad_norm": 0.005399172194302082,
|
| 547 |
+
"learning_rate": 1.4492531597089239e-05,
|
| 548 |
+
"loss": 0.004361886158585548,
|
| 549 |
+
"step": 770
|
| 550 |
+
},
|
| 551 |
+
{
|
| 552 |
+
"epoch": 0.8793686583990981,
|
| 553 |
+
"grad_norm": 0.005994090810418129,
|
| 554 |
+
"learning_rate": 1.4415932592876295e-05,
|
| 555 |
+
"loss": 0.0005545248743146658,
|
| 556 |
+
"step": 780
|
| 557 |
+
},
|
| 558 |
+
{
|
| 559 |
+
"epoch": 0.8906426155580609,
|
| 560 |
+
"grad_norm": 7.531558990478516,
|
| 561 |
+
"learning_rate": 1.4339333588663348e-05,
|
| 562 |
+
"loss": 0.03603121936321259,
|
| 563 |
+
"step": 790
|
| 564 |
+
},
|
| 565 |
+
{
|
| 566 |
+
"epoch": 0.9019165727170236,
|
| 567 |
+
"grad_norm": 0.026947202160954475,
|
| 568 |
+
"learning_rate": 1.4262734584450404e-05,
|
| 569 |
+
"loss": 0.0008070439100265503,
|
| 570 |
+
"step": 800
|
| 571 |
+
},
|
| 572 |
+
{
|
| 573 |
+
"epoch": 0.9131905298759865,
|
| 574 |
+
"grad_norm": 0.1804335117340088,
|
| 575 |
+
"learning_rate": 1.418613558023746e-05,
|
| 576 |
+
"loss": 0.0011755324900150298,
|
| 577 |
+
"step": 810
|
| 578 |
+
},
|
| 579 |
+
{
|
| 580 |
+
"epoch": 0.9244644870349493,
|
| 581 |
+
"grad_norm": 0.02349584922194481,
|
| 582 |
+
"learning_rate": 1.4109536576024511e-05,
|
| 583 |
+
"loss": 0.0017551105469465257,
|
| 584 |
+
"step": 820
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 0.9357384441939121,
|
| 588 |
+
"grad_norm": 0.006294341292232275,
|
| 589 |
+
"learning_rate": 1.4032937571811567e-05,
|
| 590 |
+
"loss": 0.0004298537038266659,
|
| 591 |
+
"step": 830
|
| 592 |
+
},
|
| 593 |
+
{
|
| 594 |
+
"epoch": 0.9470124013528749,
|
| 595 |
+
"grad_norm": 0.0048352451995015144,
|
| 596 |
+
"learning_rate": 1.3956338567598622e-05,
|
| 597 |
+
"loss": 0.0004683373495936394,
|
| 598 |
+
"step": 840
|
| 599 |
+
},
|
| 600 |
+
{
|
| 601 |
+
"epoch": 0.9582863585118376,
|
| 602 |
+
"grad_norm": 0.010727937333285809,
|
| 603 |
+
"learning_rate": 1.3879739563385676e-05,
|
| 604 |
+
"loss": 0.0005296945571899414,
|
| 605 |
+
"step": 850
|
| 606 |
+
},
|
| 607 |
+
{
|
| 608 |
+
"epoch": 0.9695603156708005,
|
| 609 |
+
"grad_norm": 0.006586082745343447,
|
| 610 |
+
"learning_rate": 1.3803140559172731e-05,
|
| 611 |
+
"loss": 0.03151362538337708,
|
| 612 |
+
"step": 860
|
| 613 |
+
},
|
| 614 |
+
{
|
| 615 |
+
"epoch": 0.9808342728297632,
|
| 616 |
+
"grad_norm": 0.004769907798618078,
|
| 617 |
+
"learning_rate": 1.3726541554959787e-05,
|
| 618 |
+
"loss": 0.0008974189870059491,
|
| 619 |
+
"step": 870
|
| 620 |
+
},
|
| 621 |
+
{
|
| 622 |
+
"epoch": 0.992108229988726,
|
| 623 |
+
"grad_norm": 0.036083221435546875,
|
| 624 |
+
"learning_rate": 1.3649942550746842e-05,
|
| 625 |
+
"loss": 0.03273430764675141,
|
| 626 |
+
"step": 880
|
| 627 |
+
},
|
| 628 |
+
{
|
| 629 |
+
"epoch": 1.0,
|
| 630 |
+
"eval_avg_accuracy": 0.9984485190409027,
|
| 631 |
+
"eval_explanation_accuracy": 0.9983074753173484,
|
| 632 |
+
"eval_explanation_f1": 0.9966814159292036,
|
| 633 |
+
"eval_loss": 0.014823827892541885,
|
| 634 |
+
"eval_refusal_accuracy": 0.998589562764457,
|
| 635 |
+
"eval_refusal_f1": 0.9985919459307238,
|
| 636 |
+
"eval_runtime": 7.0154,
|
| 637 |
+
"eval_samples_per_second": 505.316,
|
| 638 |
+
"eval_steps_per_second": 31.645,
|
| 639 |
+
"step": 887
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 1.003382187147689,
|
| 643 |
+
"grad_norm": 0.09646238386631012,
|
| 644 |
+
"learning_rate": 1.3573343546533896e-05,
|
| 645 |
+
"loss": 0.000399038614705205,
|
| 646 |
+
"step": 890
|
| 647 |
+
},
|
| 648 |
+
{
|
| 649 |
+
"epoch": 1.0146561443066517,
|
| 650 |
+
"grad_norm": 0.013944474048912525,
|
| 651 |
+
"learning_rate": 1.3496744542320952e-05,
|
| 652 |
+
"loss": 0.0006968233734369278,
|
| 653 |
+
"step": 900
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 1.0259301014656144,
|
| 657 |
+
"grad_norm": 0.005086906719952822,
|
| 658 |
+
"learning_rate": 1.3420145538108005e-05,
|
| 659 |
+
"loss": 0.0006555594969540835,
|
| 660 |
+
"step": 910
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"epoch": 1.0372040586245772,
|
| 664 |
+
"grad_norm": 0.004092383664101362,
|
| 665 |
+
"learning_rate": 1.334354653389506e-05,
|
| 666 |
+
"loss": 0.0004657072015106678,
|
| 667 |
+
"step": 920
|
| 668 |
+
},
|
| 669 |
+
{
|
| 670 |
+
"epoch": 1.04847801578354,
|
| 671 |
+
"grad_norm": 1.3944780826568604,
|
| 672 |
+
"learning_rate": 1.3266947529682115e-05,
|
| 673 |
+
"loss": 0.0009271042421460151,
|
| 674 |
+
"step": 930
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 1.059751972942503,
|
| 678 |
+
"grad_norm": 0.004638408776372671,
|
| 679 |
+
"learning_rate": 1.319034852546917e-05,
|
| 680 |
+
"loss": 0.00035289865918457507,
|
| 681 |
+
"step": 940
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 1.0710259301014655,
|
| 685 |
+
"grad_norm": 0.004089709371328354,
|
| 686 |
+
"learning_rate": 1.3113749521256226e-05,
|
| 687 |
+
"loss": 0.0005369900725781917,
|
| 688 |
+
"step": 950
|
| 689 |
+
},
|
| 690 |
+
{
|
| 691 |
+
"epoch": 1.0822998872604284,
|
| 692 |
+
"grad_norm": 0.003287773346528411,
|
| 693 |
+
"learning_rate": 1.303715051704328e-05,
|
| 694 |
+
"loss": 0.04581061005592346,
|
| 695 |
+
"step": 960
|
| 696 |
+
},
|
| 697 |
+
{
|
| 698 |
+
"epoch": 1.0935738444193912,
|
| 699 |
+
"grad_norm": 0.015667209401726723,
|
| 700 |
+
"learning_rate": 1.2960551512830335e-05,
|
| 701 |
+
"loss": 0.0005310946144163609,
|
| 702 |
+
"step": 970
|
| 703 |
+
},
|
| 704 |
+
{
|
| 705 |
+
"epoch": 1.104847801578354,
|
| 706 |
+
"grad_norm": 0.03467559069395065,
|
| 707 |
+
"learning_rate": 1.288395250861739e-05,
|
| 708 |
+
"loss": 0.0005203377455472946,
|
| 709 |
+
"step": 980
|
| 710 |
+
},
|
| 711 |
+
{
|
| 712 |
+
"epoch": 1.116121758737317,
|
| 713 |
+
"grad_norm": 0.00872899405658245,
|
| 714 |
+
"learning_rate": 1.2807353504404442e-05,
|
| 715 |
+
"loss": 0.0024556104093790055,
|
| 716 |
+
"step": 990
|
| 717 |
+
},
|
| 718 |
+
{
|
| 719 |
+
"epoch": 1.1273957158962795,
|
| 720 |
+
"grad_norm": 0.003907124977558851,
|
| 721 |
+
"learning_rate": 1.2730754500191498e-05,
|
| 722 |
+
"loss": 0.0003500781022012234,
|
| 723 |
+
"step": 1000
|
| 724 |
+
},
|
| 725 |
+
{
|
| 726 |
+
"epoch": 1.1386696730552424,
|
| 727 |
+
"grad_norm": 4.096004009246826,
|
| 728 |
+
"learning_rate": 1.2654155495978553e-05,
|
| 729 |
+
"loss": 0.04666096866130829,
|
| 730 |
+
"step": 1010
|
| 731 |
+
},
|
| 732 |
+
{
|
| 733 |
+
"epoch": 1.1499436302142052,
|
| 734 |
+
"grad_norm": 0.03219756856560707,
|
| 735 |
+
"learning_rate": 1.2577556491765607e-05,
|
| 736 |
+
"loss": 0.00044811274856328966,
|
| 737 |
+
"step": 1020
|
| 738 |
+
},
|
| 739 |
+
{
|
| 740 |
+
"epoch": 1.161217587373168,
|
| 741 |
+
"grad_norm": 0.007884475402534008,
|
| 742 |
+
"learning_rate": 1.2500957487552663e-05,
|
| 743 |
+
"loss": 0.004064217209815979,
|
| 744 |
+
"step": 1030
|
| 745 |
+
},
|
| 746 |
+
{
|
| 747 |
+
"epoch": 1.172491544532131,
|
| 748 |
+
"grad_norm": 0.01001141220331192,
|
| 749 |
+
"learning_rate": 1.2424358483339718e-05,
|
| 750 |
+
"loss": 0.08857936263084412,
|
| 751 |
+
"step": 1040
|
| 752 |
+
},
|
| 753 |
+
{
|
| 754 |
+
"epoch": 1.1837655016910935,
|
| 755 |
+
"grad_norm": 0.005555972456932068,
|
| 756 |
+
"learning_rate": 1.2347759479126774e-05,
|
| 757 |
+
"loss": 0.0006975364871323108,
|
| 758 |
+
"step": 1050
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"epoch": 1.1950394588500564,
|
| 762 |
+
"grad_norm": 0.14273758232593536,
|
| 763 |
+
"learning_rate": 1.2271160474913827e-05,
|
| 764 |
+
"loss": 0.0048938382416963576,
|
| 765 |
+
"step": 1060
|
| 766 |
+
},
|
| 767 |
+
{
|
| 768 |
+
"epoch": 1.2063134160090192,
|
| 769 |
+
"grad_norm": 0.005249196197837591,
|
| 770 |
+
"learning_rate": 1.2194561470700881e-05,
|
| 771 |
+
"loss": 0.00038248966448009016,
|
| 772 |
+
"step": 1070
|
| 773 |
+
},
|
| 774 |
+
{
|
| 775 |
+
"epoch": 1.217587373167982,
|
| 776 |
+
"grad_norm": 0.004775014240294695,
|
| 777 |
+
"learning_rate": 1.2117962466487937e-05,
|
| 778 |
+
"loss": 0.00037580323405563833,
|
| 779 |
+
"step": 1080
|
| 780 |
+
},
|
| 781 |
+
{
|
| 782 |
+
"epoch": 1.2288613303269447,
|
| 783 |
+
"grad_norm": 0.005466618109494448,
|
| 784 |
+
"learning_rate": 1.204136346227499e-05,
|
| 785 |
+
"loss": 0.00038032194133847953,
|
| 786 |
+
"step": 1090
|
| 787 |
+
},
|
| 788 |
+
{
|
| 789 |
+
"epoch": 1.2401352874859075,
|
| 790 |
+
"grad_norm": 0.009438618086278439,
|
| 791 |
+
"learning_rate": 1.1964764458062046e-05,
|
| 792 |
+
"loss": 0.0026585064828395844,
|
| 793 |
+
"step": 1100
|
| 794 |
+
},
|
| 795 |
+
{
|
| 796 |
+
"epoch": 1.2514092446448704,
|
| 797 |
+
"grad_norm": 0.0035781103651970625,
|
| 798 |
+
"learning_rate": 1.1888165453849101e-05,
|
| 799 |
+
"loss": 0.000338529609143734,
|
| 800 |
+
"step": 1110
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 1.2626832018038332,
|
| 804 |
+
"grad_norm": 0.008734864182770252,
|
| 805 |
+
"learning_rate": 1.1811566449636157e-05,
|
| 806 |
+
"loss": 0.0003517736680805683,
|
| 807 |
+
"step": 1120
|
| 808 |
+
},
|
| 809 |
+
{
|
| 810 |
+
"epoch": 1.273957158962796,
|
| 811 |
+
"grad_norm": 0.0038098106160759926,
|
| 812 |
+
"learning_rate": 1.173496744542321e-05,
|
| 813 |
+
"loss": 0.0003093461273238063,
|
| 814 |
+
"step": 1130
|
| 815 |
+
},
|
| 816 |
+
{
|
| 817 |
+
"epoch": 1.2852311161217587,
|
| 818 |
+
"grad_norm": 0.004756222013384104,
|
| 819 |
+
"learning_rate": 1.1658368441210266e-05,
|
| 820 |
+
"loss": 0.00035437943879514935,
|
| 821 |
+
"step": 1140
|
| 822 |
+
},
|
| 823 |
+
{
|
| 824 |
+
"epoch": 1.2965050732807215,
|
| 825 |
+
"grad_norm": 0.0054916017688810825,
|
| 826 |
+
"learning_rate": 1.1581769436997321e-05,
|
| 827 |
+
"loss": 0.03466772437095642,
|
| 828 |
+
"step": 1150
|
| 829 |
+
},
|
| 830 |
+
{
|
| 831 |
+
"epoch": 1.3077790304396844,
|
| 832 |
+
"grad_norm": 0.004026887938380241,
|
| 833 |
+
"learning_rate": 1.1505170432784374e-05,
|
| 834 |
+
"loss": 0.018098314106464387,
|
| 835 |
+
"step": 1160
|
| 836 |
+
},
|
| 837 |
+
{
|
| 838 |
+
"epoch": 1.3190529875986472,
|
| 839 |
+
"grad_norm": 0.010358154773712158,
|
| 840 |
+
"learning_rate": 1.1428571428571429e-05,
|
| 841 |
+
"loss": 0.0005080571398139,
|
| 842 |
+
"step": 1170
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"epoch": 1.3303269447576098,
|
| 846 |
+
"grad_norm": 0.006685700733214617,
|
| 847 |
+
"learning_rate": 1.1351972424358484e-05,
|
| 848 |
+
"loss": 0.0006039659958332777,
|
| 849 |
+
"step": 1180
|
| 850 |
+
},
|
| 851 |
+
{
|
| 852 |
+
"epoch": 1.3416009019165727,
|
| 853 |
+
"grad_norm": 0.041763920336961746,
|
| 854 |
+
"learning_rate": 1.1275373420145538e-05,
|
| 855 |
+
"loss": 0.0003080849768593907,
|
| 856 |
+
"step": 1190
|
| 857 |
+
},
|
| 858 |
+
{
|
| 859 |
+
"epoch": 1.3528748590755355,
|
| 860 |
+
"grad_norm": 0.0038634659722447395,
|
| 861 |
+
"learning_rate": 1.1198774415932594e-05,
|
| 862 |
+
"loss": 0.00025509917177259924,
|
| 863 |
+
"step": 1200
|
| 864 |
+
},
|
| 865 |
+
{
|
| 866 |
+
"epoch": 1.3641488162344984,
|
| 867 |
+
"grad_norm": 0.007722954265773296,
|
| 868 |
+
"learning_rate": 1.112217541171965e-05,
|
| 869 |
+
"loss": 0.0002860145643353462,
|
| 870 |
+
"step": 1210
|
| 871 |
+
},
|
| 872 |
+
{
|
| 873 |
+
"epoch": 1.3754227733934612,
|
| 874 |
+
"grad_norm": 0.003805127227678895,
|
| 875 |
+
"learning_rate": 1.1045576407506705e-05,
|
| 876 |
+
"loss": 0.006610937416553497,
|
| 877 |
+
"step": 1220
|
| 878 |
+
},
|
| 879 |
+
{
|
| 880 |
+
"epoch": 1.386696730552424,
|
| 881 |
+
"grad_norm": 0.003794404910877347,
|
| 882 |
+
"learning_rate": 1.0968977403293758e-05,
|
| 883 |
+
"loss": 0.0002713572233915329,
|
| 884 |
+
"step": 1230
|
| 885 |
+
},
|
| 886 |
+
{
|
| 887 |
+
"epoch": 1.3979706877113867,
|
| 888 |
+
"grad_norm": 0.0029455258045345545,
|
| 889 |
+
"learning_rate": 1.0892378399080812e-05,
|
| 890 |
+
"loss": 0.0002766825258731842,
|
| 891 |
+
"step": 1240
|
| 892 |
+
},
|
| 893 |
+
{
|
| 894 |
+
"epoch": 1.4092446448703495,
|
| 895 |
+
"grad_norm": 0.0034676918294280767,
|
| 896 |
+
"learning_rate": 1.0815779394867868e-05,
|
| 897 |
+
"loss": 0.0002525749383494258,
|
| 898 |
+
"step": 1250
|
| 899 |
+
},
|
| 900 |
+
{
|
| 901 |
+
"epoch": 1.4205186020293123,
|
| 902 |
+
"grad_norm": 0.0027564624324440956,
|
| 903 |
+
"learning_rate": 1.0739180390654921e-05,
|
| 904 |
+
"loss": 0.024026018381118775,
|
| 905 |
+
"step": 1260
|
| 906 |
+
},
|
| 907 |
+
{
|
| 908 |
+
"epoch": 1.431792559188275,
|
| 909 |
+
"grad_norm": 0.004176211077719927,
|
| 910 |
+
"learning_rate": 1.0662581386441977e-05,
|
| 911 |
+
"loss": 0.04461250603199005,
|
| 912 |
+
"step": 1270
|
| 913 |
+
},
|
| 914 |
+
{
|
| 915 |
+
"epoch": 1.4430665163472378,
|
| 916 |
+
"grad_norm": 0.7301909327507019,
|
| 917 |
+
"learning_rate": 1.0585982382229032e-05,
|
| 918 |
+
"loss": 0.0013874154537916183,
|
| 919 |
+
"step": 1280
|
| 920 |
+
},
|
| 921 |
+
{
|
| 922 |
+
"epoch": 1.4543404735062007,
|
| 923 |
+
"grad_norm": 0.006072559859603643,
|
| 924 |
+
"learning_rate": 1.0509383378016086e-05,
|
| 925 |
+
"loss": 0.0008793852292001247,
|
| 926 |
+
"step": 1290
|
| 927 |
+
},
|
| 928 |
+
{
|
| 929 |
+
"epoch": 1.4656144306651635,
|
| 930 |
+
"grad_norm": 0.003349730744957924,
|
| 931 |
+
"learning_rate": 1.0432784373803142e-05,
|
| 932 |
+
"loss": 0.11177332401275634,
|
| 933 |
+
"step": 1300
|
| 934 |
+
},
|
| 935 |
+
{
|
| 936 |
+
"epoch": 1.4768883878241263,
|
| 937 |
+
"grad_norm": 0.0037300498224794865,
|
| 938 |
+
"learning_rate": 1.0356185369590197e-05,
|
| 939 |
+
"loss": 0.02746536731719971,
|
| 940 |
+
"step": 1310
|
| 941 |
+
},
|
| 942 |
+
{
|
| 943 |
+
"epoch": 1.4881623449830892,
|
| 944 |
+
"grad_norm": 0.0067192791029810905,
|
| 945 |
+
"learning_rate": 1.027958636537725e-05,
|
| 946 |
+
"loss": 0.0006490409374237061,
|
| 947 |
+
"step": 1320
|
| 948 |
+
},
|
| 949 |
+
{
|
| 950 |
+
"epoch": 1.4994363021420518,
|
| 951 |
+
"grad_norm": 0.007033471018075943,
|
| 952 |
+
"learning_rate": 1.0202987361164305e-05,
|
| 953 |
+
"loss": 0.00473451130092144,
|
| 954 |
+
"step": 1330
|
| 955 |
+
},
|
| 956 |
+
{
|
| 957 |
+
"epoch": 1.5107102593010147,
|
| 958 |
+
"grad_norm": 0.005484273657202721,
|
| 959 |
+
"learning_rate": 1.012638835695136e-05,
|
| 960 |
+
"loss": 0.0010361320339143277,
|
| 961 |
+
"step": 1340
|
| 962 |
+
},
|
| 963 |
+
{
|
| 964 |
+
"epoch": 1.5219842164599775,
|
| 965 |
+
"grad_norm": 0.003371752332895994,
|
| 966 |
+
"learning_rate": 1.0049789352738416e-05,
|
| 967 |
+
"loss": 0.00029721735045313836,
|
| 968 |
+
"step": 1350
|
| 969 |
+
},
|
| 970 |
+
{
|
| 971 |
+
"epoch": 1.5332581736189401,
|
| 972 |
+
"grad_norm": 0.003974182065576315,
|
| 973 |
+
"learning_rate": 9.97319034852547e-06,
|
| 974 |
+
"loss": 0.00026908733416348697,
|
| 975 |
+
"step": 1360
|
| 976 |
+
},
|
| 977 |
+
{
|
| 978 |
+
"epoch": 1.544532130777903,
|
| 979 |
+
"grad_norm": 0.003871662775054574,
|
| 980 |
+
"learning_rate": 9.896591344312525e-06,
|
| 981 |
+
"loss": 0.0018835226073861122,
|
| 982 |
+
"step": 1370
|
| 983 |
+
},
|
| 984 |
+
{
|
| 985 |
+
"epoch": 1.5558060879368658,
|
| 986 |
+
"grad_norm": 0.0033102212473750114,
|
| 987 |
+
"learning_rate": 9.81999234009958e-06,
|
| 988 |
+
"loss": 0.0002722503151744604,
|
| 989 |
+
"step": 1380
|
| 990 |
+
},
|
| 991 |
+
{
|
| 992 |
+
"epoch": 1.5670800450958287,
|
| 993 |
+
"grad_norm": 0.0036641808692365885,
|
| 994 |
+
"learning_rate": 9.743393335886634e-06,
|
| 995 |
+
"loss": 0.0002578678075224161,
|
| 996 |
+
"step": 1390
|
| 997 |
+
},
|
| 998 |
+
{
|
| 999 |
+
"epoch": 1.5783540022547915,
|
| 1000 |
+
"grad_norm": 0.0043280464597046375,
|
| 1001 |
+
"learning_rate": 9.66679433167369e-06,
|
| 1002 |
+
"loss": 0.000738756125792861,
|
| 1003 |
+
"step": 1400
|
| 1004 |
+
},
|
| 1005 |
+
{
|
| 1006 |
+
"epoch": 1.5896279594137543,
|
| 1007 |
+
"grad_norm": 0.004314345773309469,
|
| 1008 |
+
"learning_rate": 9.590195327460743e-06,
|
| 1009 |
+
"loss": 0.009816586226224899,
|
| 1010 |
+
"step": 1410
|
| 1011 |
+
},
|
| 1012 |
+
{
|
| 1013 |
+
"epoch": 1.6009019165727172,
|
| 1014 |
+
"grad_norm": 0.0035737520083785057,
|
| 1015 |
+
"learning_rate": 9.513596323247799e-06,
|
| 1016 |
+
"loss": 0.0002666026121005416,
|
| 1017 |
+
"step": 1420
|
| 1018 |
+
},
|
| 1019 |
+
{
|
| 1020 |
+
"epoch": 1.6121758737316798,
|
| 1021 |
+
"grad_norm": 0.006729898974299431,
|
| 1022 |
+
"learning_rate": 9.436997319034853e-06,
|
| 1023 |
+
"loss": 0.04153406620025635,
|
| 1024 |
+
"step": 1430
|
| 1025 |
+
},
|
| 1026 |
+
{
|
| 1027 |
+
"epoch": 1.6234498308906427,
|
| 1028 |
+
"grad_norm": 0.003635931294411421,
|
| 1029 |
+
"learning_rate": 9.360398314821908e-06,
|
| 1030 |
+
"loss": 0.0002528228098526597,
|
| 1031 |
+
"step": 1440
|
| 1032 |
+
},
|
| 1033 |
+
{
|
| 1034 |
+
"epoch": 1.6347237880496053,
|
| 1035 |
+
"grad_norm": 0.003705125069245696,
|
| 1036 |
+
"learning_rate": 9.283799310608964e-06,
|
| 1037 |
+
"loss": 0.0006835301406681538,
|
| 1038 |
+
"step": 1450
|
| 1039 |
+
},
|
| 1040 |
+
{
|
| 1041 |
+
"epoch": 1.6459977452085681,
|
| 1042 |
+
"grad_norm": 0.03155064582824707,
|
| 1043 |
+
"learning_rate": 9.207200306396017e-06,
|
| 1044 |
+
"loss": 0.0004594951402395964,
|
| 1045 |
+
"step": 1460
|
| 1046 |
+
},
|
| 1047 |
+
{
|
| 1048 |
+
"epoch": 1.657271702367531,
|
| 1049 |
+
"grad_norm": 0.005488164257258177,
|
| 1050 |
+
"learning_rate": 9.130601302183073e-06,
|
| 1051 |
+
"loss": 0.0002626030007377267,
|
| 1052 |
+
"step": 1470
|
| 1053 |
+
},
|
| 1054 |
+
{
|
| 1055 |
+
"epoch": 1.6685456595264938,
|
| 1056 |
+
"grad_norm": 0.004056842997670174,
|
| 1057 |
+
"learning_rate": 9.054002297970127e-06,
|
| 1058 |
+
"loss": 0.00027129012160003184,
|
| 1059 |
+
"step": 1480
|
| 1060 |
+
},
|
| 1061 |
+
{
|
| 1062 |
+
"epoch": 1.6798196166854567,
|
| 1063 |
+
"grad_norm": 0.004610530100762844,
|
| 1064 |
+
"learning_rate": 8.977403293757182e-06,
|
| 1065 |
+
"loss": 0.0004437011666595936,
|
| 1066 |
+
"step": 1490
|
| 1067 |
+
},
|
| 1068 |
+
{
|
| 1069 |
+
"epoch": 1.6910935738444195,
|
| 1070 |
+
"grad_norm": 0.0049287667497992516,
|
| 1071 |
+
"learning_rate": 8.900804289544237e-06,
|
| 1072 |
+
"loss": 0.0002617941936478019,
|
| 1073 |
+
"step": 1500
|
| 1074 |
+
},
|
| 1075 |
+
{
|
| 1076 |
+
"epoch": 1.7023675310033823,
|
| 1077 |
+
"grad_norm": 0.0032671773806214333,
|
| 1078 |
+
"learning_rate": 8.824205285331291e-06,
|
| 1079 |
+
"loss": 0.0003260360099375248,
|
| 1080 |
+
"step": 1510
|
| 1081 |
+
},
|
| 1082 |
+
{
|
| 1083 |
+
"epoch": 1.713641488162345,
|
| 1084 |
+
"grad_norm": 0.0026306200306862593,
|
| 1085 |
+
"learning_rate": 8.747606281118347e-06,
|
| 1086 |
+
"loss": 0.040199264883995056,
|
| 1087 |
+
"step": 1520
|
| 1088 |
+
},
|
| 1089 |
+
{
|
| 1090 |
+
"epoch": 1.7249154453213078,
|
| 1091 |
+
"grad_norm": 0.002604865934699774,
|
| 1092 |
+
"learning_rate": 8.6710072769054e-06,
|
| 1093 |
+
"loss": 0.0021020427346229555,
|
| 1094 |
+
"step": 1530
|
| 1095 |
+
},
|
| 1096 |
+
{
|
| 1097 |
+
"epoch": 1.7361894024802704,
|
| 1098 |
+
"grad_norm": 0.004267075564712286,
|
| 1099 |
+
"learning_rate": 8.594408272692456e-06,
|
| 1100 |
+
"loss": 0.0005022045224905014,
|
| 1101 |
+
"step": 1540
|
| 1102 |
+
},
|
| 1103 |
+
{
|
| 1104 |
+
"epoch": 1.7474633596392333,
|
| 1105 |
+
"grad_norm": 0.003063610754907131,
|
| 1106 |
+
"learning_rate": 8.51780926847951e-06,
|
| 1107 |
+
"loss": 0.00021760372910648585,
|
| 1108 |
+
"step": 1550
|
| 1109 |
+
},
|
| 1110 |
+
{
|
| 1111 |
+
"epoch": 1.7587373167981961,
|
| 1112 |
+
"grad_norm": 0.002829270903021097,
|
| 1113 |
+
"learning_rate": 8.441210264266565e-06,
|
| 1114 |
+
"loss": 0.0003501762170344591,
|
| 1115 |
+
"step": 1560
|
| 1116 |
+
},
|
| 1117 |
+
{
|
| 1118 |
+
"epoch": 1.770011273957159,
|
| 1119 |
+
"grad_norm": 0.0024476037360727787,
|
| 1120 |
+
"learning_rate": 8.36461126005362e-06,
|
| 1121 |
+
"loss": 0.0002327218884602189,
|
| 1122 |
+
"step": 1570
|
| 1123 |
+
},
|
| 1124 |
+
{
|
| 1125 |
+
"epoch": 1.7812852311161218,
|
| 1126 |
+
"grad_norm": 0.003161813598126173,
|
| 1127 |
+
"learning_rate": 8.288012255840674e-06,
|
| 1128 |
+
"loss": 0.0002234887331724167,
|
| 1129 |
+
"step": 1580
|
| 1130 |
+
},
|
| 1131 |
+
{
|
| 1132 |
+
"epoch": 1.7925591882750846,
|
| 1133 |
+
"grad_norm": 0.002495658351108432,
|
| 1134 |
+
"learning_rate": 8.21141325162773e-06,
|
| 1135 |
+
"loss": 0.0002736057387664914,
|
| 1136 |
+
"step": 1590
|
| 1137 |
+
},
|
| 1138 |
+
{
|
| 1139 |
+
"epoch": 1.8038331454340475,
|
| 1140 |
+
"grad_norm": 0.0035063582472503185,
|
| 1141 |
+
"learning_rate": 8.134814247414784e-06,
|
| 1142 |
+
"loss": 0.0002091569360345602,
|
| 1143 |
+
"step": 1600
|
| 1144 |
+
},
|
| 1145 |
+
{
|
| 1146 |
+
"epoch": 1.8151071025930101,
|
| 1147 |
+
"grad_norm": 0.022207682952284813,
|
| 1148 |
+
"learning_rate": 8.05821524320184e-06,
|
| 1149 |
+
"loss": 0.04271211624145508,
|
| 1150 |
+
"step": 1610
|
| 1151 |
+
},
|
| 1152 |
+
{
|
| 1153 |
+
"epoch": 1.826381059751973,
|
| 1154 |
+
"grad_norm": 0.002612703014165163,
|
| 1155 |
+
"learning_rate": 7.981616238988895e-06,
|
| 1156 |
+
"loss": 0.00029878674540668727,
|
| 1157 |
+
"step": 1620
|
| 1158 |
+
},
|
| 1159 |
+
{
|
| 1160 |
+
"epoch": 1.8376550169109356,
|
| 1161 |
+
"grad_norm": 26.909603118896484,
|
| 1162 |
+
"learning_rate": 7.905017234775948e-06,
|
| 1163 |
+
"loss": 0.06723988056182861,
|
| 1164 |
+
"step": 1630
|
| 1165 |
+
},
|
| 1166 |
+
{
|
| 1167 |
+
"epoch": 1.8489289740698984,
|
| 1168 |
+
"grad_norm": 0.002532525220885873,
|
| 1169 |
+
"learning_rate": 7.828418230563002e-06,
|
| 1170 |
+
"loss": 0.00022608181461691856,
|
| 1171 |
+
"step": 1640
|
| 1172 |
+
},
|
| 1173 |
+
{
|
| 1174 |
+
"epoch": 1.8602029312288613,
|
| 1175 |
+
"grad_norm": 0.002735898597165942,
|
| 1176 |
+
"learning_rate": 7.751819226350058e-06,
|
| 1177 |
+
"loss": 0.08260895013809204,
|
| 1178 |
+
"step": 1650
|
| 1179 |
+
},
|
| 1180 |
+
{
|
| 1181 |
+
"epoch": 1.8714768883878241,
|
| 1182 |
+
"grad_norm": 0.01515868492424488,
|
| 1183 |
+
"learning_rate": 7.675220222137113e-06,
|
| 1184 |
+
"loss": 0.00023219306021928787,
|
| 1185 |
+
"step": 1660
|
| 1186 |
+
},
|
| 1187 |
+
{
|
| 1188 |
+
"epoch": 1.882750845546787,
|
| 1189 |
+
"grad_norm": 0.0035585863515734673,
|
| 1190 |
+
"learning_rate": 7.598621217924168e-06,
|
| 1191 |
+
"loss": 0.00048695728182792666,
|
| 1192 |
+
"step": 1670
|
| 1193 |
+
},
|
| 1194 |
+
{
|
| 1195 |
+
"epoch": 1.8940248027057498,
|
| 1196 |
+
"grad_norm": 0.00424063578248024,
|
| 1197 |
+
"learning_rate": 7.522022213711222e-06,
|
| 1198 |
+
"loss": 0.0003054874483495951,
|
| 1199 |
+
"step": 1680
|
| 1200 |
+
},
|
| 1201 |
+
{
|
| 1202 |
+
"epoch": 1.9052987598647126,
|
| 1203 |
+
"grad_norm": 0.004633221309632063,
|
| 1204 |
+
"learning_rate": 7.445423209498277e-06,
|
| 1205 |
+
"loss": 0.0031257856637239454,
|
| 1206 |
+
"step": 1690
|
| 1207 |
+
},
|
| 1208 |
+
{
|
| 1209 |
+
"epoch": 1.9165727170236753,
|
| 1210 |
+
"grad_norm": 0.0049049886874854565,
|
| 1211 |
+
"learning_rate": 7.368824205285332e-06,
|
| 1212 |
+
"loss": 0.00021440631244331598,
|
| 1213 |
+
"step": 1700
|
| 1214 |
+
},
|
| 1215 |
+
{
|
| 1216 |
+
"epoch": 1.927846674182638,
|
| 1217 |
+
"grad_norm": 0.01473376713693142,
|
| 1218 |
+
"learning_rate": 7.292225201072387e-06,
|
| 1219 |
+
"loss": 0.0002546449890360236,
|
| 1220 |
+
"step": 1710
|
| 1221 |
+
},
|
| 1222 |
+
{
|
| 1223 |
+
"epoch": 1.939120631341601,
|
| 1224 |
+
"grad_norm": 0.024918919429183006,
|
| 1225 |
+
"learning_rate": 7.215626196859441e-06,
|
| 1226 |
+
"loss": 0.0002916081575676799,
|
| 1227 |
+
"step": 1720
|
| 1228 |
+
},
|
| 1229 |
+
{
|
| 1230 |
+
"epoch": 1.9503945885005636,
|
| 1231 |
+
"grad_norm": 0.008383645676076412,
|
| 1232 |
+
"learning_rate": 7.139027192646496e-06,
|
| 1233 |
+
"loss": 0.0002473884029313922,
|
| 1234 |
+
"step": 1730
|
| 1235 |
+
},
|
| 1236 |
+
{
|
| 1237 |
+
"epoch": 1.9616685456595264,
|
| 1238 |
+
"grad_norm": 0.003205355489626527,
|
| 1239 |
+
"learning_rate": 7.062428188433551e-06,
|
| 1240 |
+
"loss": 0.00030408557504415514,
|
| 1241 |
+
"step": 1740
|
| 1242 |
+
},
|
| 1243 |
+
{
|
| 1244 |
+
"epoch": 1.9729425028184893,
|
| 1245 |
+
"grad_norm": 0.0037417688872665167,
|
| 1246 |
+
"learning_rate": 6.9858291842206064e-06,
|
| 1247 |
+
"loss": 0.00022464105859398842,
|
| 1248 |
+
"step": 1750
|
| 1249 |
+
},
|
| 1250 |
+
{
|
| 1251 |
+
"epoch": 1.984216459977452,
|
| 1252 |
+
"grad_norm": 0.0024626674130558968,
|
| 1253 |
+
"learning_rate": 6.90923018000766e-06,
|
| 1254 |
+
"loss": 0.0002444717800244689,
|
| 1255 |
+
"step": 1760
|
| 1256 |
+
},
|
| 1257 |
+
{
|
| 1258 |
+
"epoch": 1.995490417136415,
|
| 1259 |
+
"grad_norm": 0.0025584984105080366,
|
| 1260 |
+
"learning_rate": 6.832631175794715e-06,
|
| 1261 |
+
"loss": 0.00020699049346148968,
|
| 1262 |
+
"step": 1770
|
| 1263 |
+
},
|
| 1264 |
+
{
|
| 1265 |
+
"epoch": 2.0,
|
| 1266 |
+
"eval_avg_accuracy": 0.9983074753173484,
|
| 1267 |
+
"eval_explanation_accuracy": 0.9983074753173484,
|
| 1268 |
+
"eval_explanation_f1": 0.9966740576496674,
|
| 1269 |
+
"eval_loss": 0.017838388681411743,
|
| 1270 |
+
"eval_refusal_accuracy": 0.9983074753173484,
|
| 1271 |
+
"eval_refusal_f1": 0.9983079526226735,
|
| 1272 |
+
"eval_runtime": 8.1647,
|
| 1273 |
+
"eval_samples_per_second": 434.186,
|
| 1274 |
+
"eval_steps_per_second": 27.19,
|
| 1275 |
+
"step": 1774
|
| 1276 |
+
}
|
| 1277 |
+
],
|
| 1278 |
+
"logging_steps": 10,
|
| 1279 |
+
"max_steps": 2661,
|
| 1280 |
+
"num_input_tokens_seen": 0,
|
| 1281 |
+
"num_train_epochs": 3,
|
| 1282 |
+
"save_steps": 500,
|
| 1283 |
+
"stateful_callbacks": {
|
| 1284 |
+
"TrainerControl": {
|
| 1285 |
+
"args": {
|
| 1286 |
+
"should_epoch_stop": false,
|
| 1287 |
+
"should_evaluate": false,
|
| 1288 |
+
"should_log": false,
|
| 1289 |
+
"should_save": true,
|
| 1290 |
+
"should_training_stop": false
|
| 1291 |
+
},
|
| 1292 |
+
"attributes": {}
|
| 1293 |
+
}
|
| 1294 |
+
},
|
| 1295 |
+
"total_flos": 0.0,
|
| 1296 |
+
"train_batch_size": 16,
|
| 1297 |
+
"trial_name": null,
|
| 1298 |
+
"trial_params": null
|
| 1299 |
+
}
|
checkpoints/checkpoint-1774/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:78b3ded581335bef923598436ee3c2f3e4bb84d0d2d55e710db56437d81078ce
|
| 3 |
+
size 5201
|
checkpoints/checkpoint-2661/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ffdd6ed6fe76cd8763e094c26b800dc2850f205e8d5cc957e7fcc9b592e6eaa6
|
| 3 |
+
size 265475800
|
checkpoints/checkpoint-2661/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:b65497f3bef254d6680416630317167a1eb30ecb3b1fa6eae5f2beb134116737
|
| 3 |
+
size 531012363
|
checkpoints/checkpoint-2661/rng_state.pth
ADDED
|
Binary file (14.5 kB). View file
|
|
|
checkpoints/checkpoint-2661/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:56a637f58b237e2589670223ef1c7fcfb17748f2535d6bfe5d0a1b7495726d75
|
| 3 |
+
size 1465
|
checkpoints/checkpoint-2661/trainer_state.json
ADDED
|
@@ -0,0 +1,1935 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": 2661,
|
| 3 |
+
"best_metric": 0.9991537376586741,
|
| 4 |
+
"best_model_checkpoint": "models/classifier_model/checkpoints/checkpoint-2661",
|
| 5 |
+
"epoch": 3.0,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 2661,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.011273957158962795,
|
| 14 |
+
"grad_norm": 6.372345447540283,
|
| 15 |
+
"learning_rate": 3.6000000000000003e-06,
|
| 16 |
+
"loss": 1.4694642066955566,
|
| 17 |
+
"step": 10
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.02254791431792559,
|
| 21 |
+
"grad_norm": 3.5241456031799316,
|
| 22 |
+
"learning_rate": 7.600000000000001e-06,
|
| 23 |
+
"loss": 1.3229194641113282,
|
| 24 |
+
"step": 20
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.033821871476888386,
|
| 28 |
+
"grad_norm": 4.042177677154541,
|
| 29 |
+
"learning_rate": 1.16e-05,
|
| 30 |
+
"loss": 1.0544840812683105,
|
| 31 |
+
"step": 30
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.04509582863585118,
|
| 35 |
+
"grad_norm": 3.2139792442321777,
|
| 36 |
+
"learning_rate": 1.5600000000000003e-05,
|
| 37 |
+
"loss": 0.7191516399383545,
|
| 38 |
+
"step": 40
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.05636978579481398,
|
| 42 |
+
"grad_norm": 4.926011085510254,
|
| 43 |
+
"learning_rate": 1.9600000000000002e-05,
|
| 44 |
+
"loss": 0.40511221885681153,
|
| 45 |
+
"step": 50
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.06764374295377677,
|
| 49 |
+
"grad_norm": 4.614091873168945,
|
| 50 |
+
"learning_rate": 1.9931060896208353e-05,
|
| 51 |
+
"loss": 0.17188742160797119,
|
| 52 |
+
"step": 60
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 0.07891770011273957,
|
| 56 |
+
"grad_norm": 0.4526596665382385,
|
| 57 |
+
"learning_rate": 1.9854461891995405e-05,
|
| 58 |
+
"loss": 0.08630315661430359,
|
| 59 |
+
"step": 70
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 0.09019165727170236,
|
| 63 |
+
"grad_norm": 0.4451746940612793,
|
| 64 |
+
"learning_rate": 1.977786288778246e-05,
|
| 65 |
+
"loss": 0.03049740493297577,
|
| 66 |
+
"step": 80
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 0.10146561443066517,
|
| 70 |
+
"grad_norm": 0.13958168029785156,
|
| 71 |
+
"learning_rate": 1.9701263883569516e-05,
|
| 72 |
+
"loss": 0.012567885220050812,
|
| 73 |
+
"step": 90
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 0.11273957158962795,
|
| 77 |
+
"grad_norm": 0.06256264448165894,
|
| 78 |
+
"learning_rate": 1.962466487935657e-05,
|
| 79 |
+
"loss": 0.06832309961318969,
|
| 80 |
+
"step": 100
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.12401352874859076,
|
| 84 |
+
"grad_norm": 0.048310115933418274,
|
| 85 |
+
"learning_rate": 1.9548065875143624e-05,
|
| 86 |
+
"loss": 0.011482110619544983,
|
| 87 |
+
"step": 110
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 0.13528748590755355,
|
| 91 |
+
"grad_norm": 0.12978595495224,
|
| 92 |
+
"learning_rate": 1.947146687093068e-05,
|
| 93 |
+
"loss": 0.04678139984607697,
|
| 94 |
+
"step": 120
|
| 95 |
+
},
|
| 96 |
+
{
|
| 97 |
+
"epoch": 0.14656144306651633,
|
| 98 |
+
"grad_norm": 2.551255226135254,
|
| 99 |
+
"learning_rate": 1.9394867866717735e-05,
|
| 100 |
+
"loss": 0.11264426708221435,
|
| 101 |
+
"step": 130
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"epoch": 0.15783540022547915,
|
| 105 |
+
"grad_norm": 0.3405976891517639,
|
| 106 |
+
"learning_rate": 1.931826886250479e-05,
|
| 107 |
+
"loss": 0.031431844830513,
|
| 108 |
+
"step": 140
|
| 109 |
+
},
|
| 110 |
+
{
|
| 111 |
+
"epoch": 0.16910935738444194,
|
| 112 |
+
"grad_norm": 0.030345242470502853,
|
| 113 |
+
"learning_rate": 1.9241669858291846e-05,
|
| 114 |
+
"loss": 0.002859099581837654,
|
| 115 |
+
"step": 150
|
| 116 |
+
},
|
| 117 |
+
{
|
| 118 |
+
"epoch": 0.18038331454340473,
|
| 119 |
+
"grad_norm": 0.11610831320285797,
|
| 120 |
+
"learning_rate": 1.9165070854078898e-05,
|
| 121 |
+
"loss": 0.0036297302693128588,
|
| 122 |
+
"step": 160
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 0.19165727170236754,
|
| 126 |
+
"grad_norm": 0.0447811484336853,
|
| 127 |
+
"learning_rate": 1.9088471849865953e-05,
|
| 128 |
+
"loss": 0.003958044201135635,
|
| 129 |
+
"step": 170
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 0.20293122886133033,
|
| 133 |
+
"grad_norm": 0.02229328267276287,
|
| 134 |
+
"learning_rate": 1.901187284565301e-05,
|
| 135 |
+
"loss": 0.002498266473412514,
|
| 136 |
+
"step": 180
|
| 137 |
+
},
|
| 138 |
+
{
|
| 139 |
+
"epoch": 0.21420518602029312,
|
| 140 |
+
"grad_norm": 0.016443341970443726,
|
| 141 |
+
"learning_rate": 1.8935273841440064e-05,
|
| 142 |
+
"loss": 0.0019413232803344727,
|
| 143 |
+
"step": 190
|
| 144 |
+
},
|
| 145 |
+
{
|
| 146 |
+
"epoch": 0.2254791431792559,
|
| 147 |
+
"grad_norm": 0.025817960500717163,
|
| 148 |
+
"learning_rate": 1.8858674837227116e-05,
|
| 149 |
+
"loss": 0.10970286130905152,
|
| 150 |
+
"step": 200
|
| 151 |
+
},
|
| 152 |
+
{
|
| 153 |
+
"epoch": 0.23675310033821872,
|
| 154 |
+
"grad_norm": 0.030361849814653397,
|
| 155 |
+
"learning_rate": 1.8782075833014172e-05,
|
| 156 |
+
"loss": 0.034755557775497437,
|
| 157 |
+
"step": 210
|
| 158 |
+
},
|
| 159 |
+
{
|
| 160 |
+
"epoch": 0.2480270574971815,
|
| 161 |
+
"grad_norm": 0.6030261516571045,
|
| 162 |
+
"learning_rate": 1.8705476828801227e-05,
|
| 163 |
+
"loss": 0.0343275398015976,
|
| 164 |
+
"step": 220
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 0.2593010146561443,
|
| 168 |
+
"grad_norm": 0.016219334676861763,
|
| 169 |
+
"learning_rate": 1.8628877824588283e-05,
|
| 170 |
+
"loss": 0.020017309486865996,
|
| 171 |
+
"step": 230
|
| 172 |
+
},
|
| 173 |
+
{
|
| 174 |
+
"epoch": 0.2705749718151071,
|
| 175 |
+
"grad_norm": 0.013941165059804916,
|
| 176 |
+
"learning_rate": 1.8552278820375335e-05,
|
| 177 |
+
"loss": 0.0028075359761714934,
|
| 178 |
+
"step": 240
|
| 179 |
+
},
|
| 180 |
+
{
|
| 181 |
+
"epoch": 0.2818489289740699,
|
| 182 |
+
"grad_norm": 1.2171980142593384,
|
| 183 |
+
"learning_rate": 1.847567981616239e-05,
|
| 184 |
+
"loss": 0.0019562991335988046,
|
| 185 |
+
"step": 250
|
| 186 |
+
},
|
| 187 |
+
{
|
| 188 |
+
"epoch": 0.29312288613303267,
|
| 189 |
+
"grad_norm": 0.02682333253324032,
|
| 190 |
+
"learning_rate": 1.8399080811949446e-05,
|
| 191 |
+
"loss": 0.0012471863068640233,
|
| 192 |
+
"step": 260
|
| 193 |
+
},
|
| 194 |
+
{
|
| 195 |
+
"epoch": 0.3043968432919955,
|
| 196 |
+
"grad_norm": 0.013817372731864452,
|
| 197 |
+
"learning_rate": 1.83224818077365e-05,
|
| 198 |
+
"loss": 0.060174965858459474,
|
| 199 |
+
"step": 270
|
| 200 |
+
},
|
| 201 |
+
{
|
| 202 |
+
"epoch": 0.3156708004509583,
|
| 203 |
+
"grad_norm": 0.023286759853363037,
|
| 204 |
+
"learning_rate": 1.8245882803523557e-05,
|
| 205 |
+
"loss": 0.0016331907361745834,
|
| 206 |
+
"step": 280
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 0.3269447576099211,
|
| 210 |
+
"grad_norm": 13.508488655090332,
|
| 211 |
+
"learning_rate": 1.8169283799310612e-05,
|
| 212 |
+
"loss": 0.010956646502017975,
|
| 213 |
+
"step": 290
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"epoch": 0.3382187147688839,
|
| 217 |
+
"grad_norm": 0.012220031581819057,
|
| 218 |
+
"learning_rate": 1.8092684795097664e-05,
|
| 219 |
+
"loss": 0.04829314053058624,
|
| 220 |
+
"step": 300
|
| 221 |
+
},
|
| 222 |
+
{
|
| 223 |
+
"epoch": 0.34949267192784667,
|
| 224 |
+
"grad_norm": 21.441791534423828,
|
| 225 |
+
"learning_rate": 1.801608579088472e-05,
|
| 226 |
+
"loss": 0.007926396280527114,
|
| 227 |
+
"step": 310
|
| 228 |
+
},
|
| 229 |
+
{
|
| 230 |
+
"epoch": 0.36076662908680945,
|
| 231 |
+
"grad_norm": 0.8508244156837463,
|
| 232 |
+
"learning_rate": 1.7939486786671775e-05,
|
| 233 |
+
"loss": 0.0012166524305939675,
|
| 234 |
+
"step": 320
|
| 235 |
+
},
|
| 236 |
+
{
|
| 237 |
+
"epoch": 0.37204058624577224,
|
| 238 |
+
"grad_norm": 0.01957777701318264,
|
| 239 |
+
"learning_rate": 1.7862887782458827e-05,
|
| 240 |
+
"loss": 0.0012863848358392715,
|
| 241 |
+
"step": 330
|
| 242 |
+
},
|
| 243 |
+
{
|
| 244 |
+
"epoch": 0.3833145434047351,
|
| 245 |
+
"grad_norm": 0.013677588663995266,
|
| 246 |
+
"learning_rate": 1.7786288778245883e-05,
|
| 247 |
+
"loss": 0.0008751305751502514,
|
| 248 |
+
"step": 340
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 0.3945885005636979,
|
| 252 |
+
"grad_norm": 0.02119549550116062,
|
| 253 |
+
"learning_rate": 1.7709689774032938e-05,
|
| 254 |
+
"loss": 0.05260458588600159,
|
| 255 |
+
"step": 350
|
| 256 |
+
},
|
| 257 |
+
{
|
| 258 |
+
"epoch": 0.40586245772266066,
|
| 259 |
+
"grad_norm": 0.02179444581270218,
|
| 260 |
+
"learning_rate": 1.7633090769819994e-05,
|
| 261 |
+
"loss": 0.01002354547381401,
|
| 262 |
+
"step": 360
|
| 263 |
+
},
|
| 264 |
+
{
|
| 265 |
+
"epoch": 0.41713641488162345,
|
| 266 |
+
"grad_norm": 0.01586312986910343,
|
| 267 |
+
"learning_rate": 1.755649176560705e-05,
|
| 268 |
+
"loss": 0.0009520040825009346,
|
| 269 |
+
"step": 370
|
| 270 |
+
},
|
| 271 |
+
{
|
| 272 |
+
"epoch": 0.42841037204058624,
|
| 273 |
+
"grad_norm": 0.00843344908207655,
|
| 274 |
+
"learning_rate": 1.7479892761394105e-05,
|
| 275 |
+
"loss": 0.0007581338752061129,
|
| 276 |
+
"step": 380
|
| 277 |
+
},
|
| 278 |
+
{
|
| 279 |
+
"epoch": 0.43968432919954903,
|
| 280 |
+
"grad_norm": 0.0289614200592041,
|
| 281 |
+
"learning_rate": 1.740329375718116e-05,
|
| 282 |
+
"loss": 0.08814518451690674,
|
| 283 |
+
"step": 390
|
| 284 |
+
},
|
| 285 |
+
{
|
| 286 |
+
"epoch": 0.4509582863585118,
|
| 287 |
+
"grad_norm": 0.20195287466049194,
|
| 288 |
+
"learning_rate": 1.7326694752968215e-05,
|
| 289 |
+
"loss": 0.02897307574748993,
|
| 290 |
+
"step": 400
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 0.46223224351747466,
|
| 294 |
+
"grad_norm": 0.2169051319360733,
|
| 295 |
+
"learning_rate": 1.7250095748755268e-05,
|
| 296 |
+
"loss": 0.01624799370765686,
|
| 297 |
+
"step": 410
|
| 298 |
+
},
|
| 299 |
+
{
|
| 300 |
+
"epoch": 0.47350620067643745,
|
| 301 |
+
"grad_norm": 0.009392499923706055,
|
| 302 |
+
"learning_rate": 1.7173496744542323e-05,
|
| 303 |
+
"loss": 0.036326673626899716,
|
| 304 |
+
"step": 420
|
| 305 |
+
},
|
| 306 |
+
{
|
| 307 |
+
"epoch": 0.48478015783540024,
|
| 308 |
+
"grad_norm": 0.008052267134189606,
|
| 309 |
+
"learning_rate": 1.7096897740329375e-05,
|
| 310 |
+
"loss": 0.0008220113813877106,
|
| 311 |
+
"step": 430
|
| 312 |
+
},
|
| 313 |
+
{
|
| 314 |
+
"epoch": 0.496054114994363,
|
| 315 |
+
"grad_norm": 0.09780994057655334,
|
| 316 |
+
"learning_rate": 1.702029873611643e-05,
|
| 317 |
+
"loss": 0.0006682470440864563,
|
| 318 |
+
"step": 440
|
| 319 |
+
},
|
| 320 |
+
{
|
| 321 |
+
"epoch": 0.5073280721533259,
|
| 322 |
+
"grad_norm": 0.09860614687204361,
|
| 323 |
+
"learning_rate": 1.6943699731903486e-05,
|
| 324 |
+
"loss": 0.0006374984513968229,
|
| 325 |
+
"step": 450
|
| 326 |
+
},
|
| 327 |
+
{
|
| 328 |
+
"epoch": 0.5186020293122886,
|
| 329 |
+
"grad_norm": 0.007464884780347347,
|
| 330 |
+
"learning_rate": 1.686710072769054e-05,
|
| 331 |
+
"loss": 0.0005574076902121305,
|
| 332 |
+
"step": 460
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 0.5298759864712514,
|
| 336 |
+
"grad_norm": 0.0052289823070168495,
|
| 337 |
+
"learning_rate": 1.6790501723477597e-05,
|
| 338 |
+
"loss": 0.0011871119029819965,
|
| 339 |
+
"step": 470
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 0.5411499436302142,
|
| 343 |
+
"grad_norm": 0.03991198539733887,
|
| 344 |
+
"learning_rate": 1.6713902719264652e-05,
|
| 345 |
+
"loss": 0.004352982714772224,
|
| 346 |
+
"step": 480
|
| 347 |
+
},
|
| 348 |
+
{
|
| 349 |
+
"epoch": 0.552423900789177,
|
| 350 |
+
"grad_norm": 0.005847420543432236,
|
| 351 |
+
"learning_rate": 1.6637303715051705e-05,
|
| 352 |
+
"loss": 0.023742355406284332,
|
| 353 |
+
"step": 490
|
| 354 |
+
},
|
| 355 |
+
{
|
| 356 |
+
"epoch": 0.5636978579481398,
|
| 357 |
+
"grad_norm": 0.007102746516466141,
|
| 358 |
+
"learning_rate": 1.656070471083876e-05,
|
| 359 |
+
"loss": 0.0033394105732440947,
|
| 360 |
+
"step": 500
|
| 361 |
+
},
|
| 362 |
+
{
|
| 363 |
+
"epoch": 0.5749718151071026,
|
| 364 |
+
"grad_norm": 0.00974489375948906,
|
| 365 |
+
"learning_rate": 1.6484105706625815e-05,
|
| 366 |
+
"loss": 0.0006546341348439455,
|
| 367 |
+
"step": 510
|
| 368 |
+
},
|
| 369 |
+
{
|
| 370 |
+
"epoch": 0.5862457722660653,
|
| 371 |
+
"grad_norm": 0.011177345179021358,
|
| 372 |
+
"learning_rate": 1.640750670241287e-05,
|
| 373 |
+
"loss": 0.0005123908165842294,
|
| 374 |
+
"step": 520
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 0.5975197294250282,
|
| 378 |
+
"grad_norm": 0.45112329721450806,
|
| 379 |
+
"learning_rate": 1.6330907698199926e-05,
|
| 380 |
+
"loss": 0.06648756265640259,
|
| 381 |
+
"step": 530
|
| 382 |
+
},
|
| 383 |
+
{
|
| 384 |
+
"epoch": 0.608793686583991,
|
| 385 |
+
"grad_norm": 0.060878559947013855,
|
| 386 |
+
"learning_rate": 1.625430869398698e-05,
|
| 387 |
+
"loss": 0.03919784724712372,
|
| 388 |
+
"step": 540
|
| 389 |
+
},
|
| 390 |
+
{
|
| 391 |
+
"epoch": 0.6200676437429538,
|
| 392 |
+
"grad_norm": 0.01804164983332157,
|
| 393 |
+
"learning_rate": 1.6177709689774034e-05,
|
| 394 |
+
"loss": 0.0012153918854892255,
|
| 395 |
+
"step": 550
|
| 396 |
+
},
|
| 397 |
+
{
|
| 398 |
+
"epoch": 0.6313416009019166,
|
| 399 |
+
"grad_norm": 0.0058420985005795956,
|
| 400 |
+
"learning_rate": 1.610111068556109e-05,
|
| 401 |
+
"loss": 0.0008219529874622822,
|
| 402 |
+
"step": 560
|
| 403 |
+
},
|
| 404 |
+
{
|
| 405 |
+
"epoch": 0.6426155580608793,
|
| 406 |
+
"grad_norm": 0.011106841266155243,
|
| 407 |
+
"learning_rate": 1.602451168134814e-05,
|
| 408 |
+
"loss": 0.0005111472215503454,
|
| 409 |
+
"step": 570
|
| 410 |
+
},
|
| 411 |
+
{
|
| 412 |
+
"epoch": 0.6538895152198422,
|
| 413 |
+
"grad_norm": 0.005702407564967871,
|
| 414 |
+
"learning_rate": 1.5947912677135197e-05,
|
| 415 |
+
"loss": 0.00047746491618454455,
|
| 416 |
+
"step": 580
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 0.6651634723788049,
|
| 420 |
+
"grad_norm": 0.006169776897877455,
|
| 421 |
+
"learning_rate": 1.5871313672922252e-05,
|
| 422 |
+
"loss": 0.0448866218328476,
|
| 423 |
+
"step": 590
|
| 424 |
+
},
|
| 425 |
+
{
|
| 426 |
+
"epoch": 0.6764374295377678,
|
| 427 |
+
"grad_norm": 15.104083061218262,
|
| 428 |
+
"learning_rate": 1.5794714668709308e-05,
|
| 429 |
+
"loss": 0.006607178598642349,
|
| 430 |
+
"step": 600
|
| 431 |
+
},
|
| 432 |
+
{
|
| 433 |
+
"epoch": 0.6877113866967306,
|
| 434 |
+
"grad_norm": 0.004718424286693335,
|
| 435 |
+
"learning_rate": 1.5718115664496363e-05,
|
| 436 |
+
"loss": 0.0009398681111633777,
|
| 437 |
+
"step": 610
|
| 438 |
+
},
|
| 439 |
+
{
|
| 440 |
+
"epoch": 0.6989853438556933,
|
| 441 |
+
"grad_norm": 0.033868566155433655,
|
| 442 |
+
"learning_rate": 1.564151666028342e-05,
|
| 443 |
+
"loss": 0.11185940504074096,
|
| 444 |
+
"step": 620
|
| 445 |
+
},
|
| 446 |
+
{
|
| 447 |
+
"epoch": 0.7102593010146562,
|
| 448 |
+
"grad_norm": 0.06708049029111862,
|
| 449 |
+
"learning_rate": 1.5564917656070474e-05,
|
| 450 |
+
"loss": 0.025441676378250122,
|
| 451 |
+
"step": 630
|
| 452 |
+
},
|
| 453 |
+
{
|
| 454 |
+
"epoch": 0.7215332581736189,
|
| 455 |
+
"grad_norm": 0.04639710858464241,
|
| 456 |
+
"learning_rate": 1.5488318651857526e-05,
|
| 457 |
+
"loss": 0.04628426432609558,
|
| 458 |
+
"step": 640
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 0.7328072153325818,
|
| 462 |
+
"grad_norm": 0.012839795090258121,
|
| 463 |
+
"learning_rate": 1.5411719647644582e-05,
|
| 464 |
+
"loss": 0.001783803105354309,
|
| 465 |
+
"step": 650
|
| 466 |
+
},
|
| 467 |
+
{
|
| 468 |
+
"epoch": 0.7440811724915445,
|
| 469 |
+
"grad_norm": 0.29861995577812195,
|
| 470 |
+
"learning_rate": 1.5335120643431637e-05,
|
| 471 |
+
"loss": 0.046310481429100034,
|
| 472 |
+
"step": 660
|
| 473 |
+
},
|
| 474 |
+
{
|
| 475 |
+
"epoch": 0.7553551296505073,
|
| 476 |
+
"grad_norm": 0.23833876848220825,
|
| 477 |
+
"learning_rate": 1.5258521639218691e-05,
|
| 478 |
+
"loss": 0.03593424260616303,
|
| 479 |
+
"step": 670
|
| 480 |
+
},
|
| 481 |
+
{
|
| 482 |
+
"epoch": 0.7666290868094702,
|
| 483 |
+
"grad_norm": 0.016054296866059303,
|
| 484 |
+
"learning_rate": 1.5181922635005747e-05,
|
| 485 |
+
"loss": 0.01490943431854248,
|
| 486 |
+
"step": 680
|
| 487 |
+
},
|
| 488 |
+
{
|
| 489 |
+
"epoch": 0.7779030439684329,
|
| 490 |
+
"grad_norm": 25.791549682617188,
|
| 491 |
+
"learning_rate": 1.51053236307928e-05,
|
| 492 |
+
"loss": 0.057176172733306885,
|
| 493 |
+
"step": 690
|
| 494 |
+
},
|
| 495 |
+
{
|
| 496 |
+
"epoch": 0.7891770011273957,
|
| 497 |
+
"grad_norm": 3.5461034774780273,
|
| 498 |
+
"learning_rate": 1.5028724626579856e-05,
|
| 499 |
+
"loss": 0.07008817791938782,
|
| 500 |
+
"step": 700
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 0.8004509582863585,
|
| 504 |
+
"grad_norm": 8.450429916381836,
|
| 505 |
+
"learning_rate": 1.4952125622366911e-05,
|
| 506 |
+
"loss": 0.081763756275177,
|
| 507 |
+
"step": 710
|
| 508 |
+
},
|
| 509 |
+
{
|
| 510 |
+
"epoch": 0.8117249154453213,
|
| 511 |
+
"grad_norm": 0.015128090977668762,
|
| 512 |
+
"learning_rate": 1.4875526618153965e-05,
|
| 513 |
+
"loss": 0.005710385367274284,
|
| 514 |
+
"step": 720
|
| 515 |
+
},
|
| 516 |
+
{
|
| 517 |
+
"epoch": 0.8229988726042841,
|
| 518 |
+
"grad_norm": 0.006377412937581539,
|
| 519 |
+
"learning_rate": 1.479892761394102e-05,
|
| 520 |
+
"loss": 0.0007186648901551962,
|
| 521 |
+
"step": 730
|
| 522 |
+
},
|
| 523 |
+
{
|
| 524 |
+
"epoch": 0.8342728297632469,
|
| 525 |
+
"grad_norm": 0.009410051628947258,
|
| 526 |
+
"learning_rate": 1.4722328609728074e-05,
|
| 527 |
+
"loss": 0.0129980206489563,
|
| 528 |
+
"step": 740
|
| 529 |
+
},
|
| 530 |
+
{
|
| 531 |
+
"epoch": 0.8455467869222097,
|
| 532 |
+
"grad_norm": 0.00972823053598404,
|
| 533 |
+
"learning_rate": 1.4645729605515128e-05,
|
| 534 |
+
"loss": 0.0006546798162162304,
|
| 535 |
+
"step": 750
|
| 536 |
+
},
|
| 537 |
+
{
|
| 538 |
+
"epoch": 0.8568207440811725,
|
| 539 |
+
"grad_norm": 0.013078595511615276,
|
| 540 |
+
"learning_rate": 1.4569130601302184e-05,
|
| 541 |
+
"loss": 0.058796089887619016,
|
| 542 |
+
"step": 760
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 0.8680947012401353,
|
| 546 |
+
"grad_norm": 0.005399172194302082,
|
| 547 |
+
"learning_rate": 1.4492531597089239e-05,
|
| 548 |
+
"loss": 0.004361886158585548,
|
| 549 |
+
"step": 770
|
| 550 |
+
},
|
| 551 |
+
{
|
| 552 |
+
"epoch": 0.8793686583990981,
|
| 553 |
+
"grad_norm": 0.005994090810418129,
|
| 554 |
+
"learning_rate": 1.4415932592876295e-05,
|
| 555 |
+
"loss": 0.0005545248743146658,
|
| 556 |
+
"step": 780
|
| 557 |
+
},
|
| 558 |
+
{
|
| 559 |
+
"epoch": 0.8906426155580609,
|
| 560 |
+
"grad_norm": 7.531558990478516,
|
| 561 |
+
"learning_rate": 1.4339333588663348e-05,
|
| 562 |
+
"loss": 0.03603121936321259,
|
| 563 |
+
"step": 790
|
| 564 |
+
},
|
| 565 |
+
{
|
| 566 |
+
"epoch": 0.9019165727170236,
|
| 567 |
+
"grad_norm": 0.026947202160954475,
|
| 568 |
+
"learning_rate": 1.4262734584450404e-05,
|
| 569 |
+
"loss": 0.0008070439100265503,
|
| 570 |
+
"step": 800
|
| 571 |
+
},
|
| 572 |
+
{
|
| 573 |
+
"epoch": 0.9131905298759865,
|
| 574 |
+
"grad_norm": 0.1804335117340088,
|
| 575 |
+
"learning_rate": 1.418613558023746e-05,
|
| 576 |
+
"loss": 0.0011755324900150298,
|
| 577 |
+
"step": 810
|
| 578 |
+
},
|
| 579 |
+
{
|
| 580 |
+
"epoch": 0.9244644870349493,
|
| 581 |
+
"grad_norm": 0.02349584922194481,
|
| 582 |
+
"learning_rate": 1.4109536576024511e-05,
|
| 583 |
+
"loss": 0.0017551105469465257,
|
| 584 |
+
"step": 820
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 0.9357384441939121,
|
| 588 |
+
"grad_norm": 0.006294341292232275,
|
| 589 |
+
"learning_rate": 1.4032937571811567e-05,
|
| 590 |
+
"loss": 0.0004298537038266659,
|
| 591 |
+
"step": 830
|
| 592 |
+
},
|
| 593 |
+
{
|
| 594 |
+
"epoch": 0.9470124013528749,
|
| 595 |
+
"grad_norm": 0.0048352451995015144,
|
| 596 |
+
"learning_rate": 1.3956338567598622e-05,
|
| 597 |
+
"loss": 0.0004683373495936394,
|
| 598 |
+
"step": 840
|
| 599 |
+
},
|
| 600 |
+
{
|
| 601 |
+
"epoch": 0.9582863585118376,
|
| 602 |
+
"grad_norm": 0.010727937333285809,
|
| 603 |
+
"learning_rate": 1.3879739563385676e-05,
|
| 604 |
+
"loss": 0.0005296945571899414,
|
| 605 |
+
"step": 850
|
| 606 |
+
},
|
| 607 |
+
{
|
| 608 |
+
"epoch": 0.9695603156708005,
|
| 609 |
+
"grad_norm": 0.006586082745343447,
|
| 610 |
+
"learning_rate": 1.3803140559172731e-05,
|
| 611 |
+
"loss": 0.03151362538337708,
|
| 612 |
+
"step": 860
|
| 613 |
+
},
|
| 614 |
+
{
|
| 615 |
+
"epoch": 0.9808342728297632,
|
| 616 |
+
"grad_norm": 0.004769907798618078,
|
| 617 |
+
"learning_rate": 1.3726541554959787e-05,
|
| 618 |
+
"loss": 0.0008974189870059491,
|
| 619 |
+
"step": 870
|
| 620 |
+
},
|
| 621 |
+
{
|
| 622 |
+
"epoch": 0.992108229988726,
|
| 623 |
+
"grad_norm": 0.036083221435546875,
|
| 624 |
+
"learning_rate": 1.3649942550746842e-05,
|
| 625 |
+
"loss": 0.03273430764675141,
|
| 626 |
+
"step": 880
|
| 627 |
+
},
|
| 628 |
+
{
|
| 629 |
+
"epoch": 1.0,
|
| 630 |
+
"eval_avg_accuracy": 0.9984485190409027,
|
| 631 |
+
"eval_explanation_accuracy": 0.9983074753173484,
|
| 632 |
+
"eval_explanation_f1": 0.9966814159292036,
|
| 633 |
+
"eval_loss": 0.014823827892541885,
|
| 634 |
+
"eval_refusal_accuracy": 0.998589562764457,
|
| 635 |
+
"eval_refusal_f1": 0.9985919459307238,
|
| 636 |
+
"eval_runtime": 7.0154,
|
| 637 |
+
"eval_samples_per_second": 505.316,
|
| 638 |
+
"eval_steps_per_second": 31.645,
|
| 639 |
+
"step": 887
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 1.003382187147689,
|
| 643 |
+
"grad_norm": 0.09646238386631012,
|
| 644 |
+
"learning_rate": 1.3573343546533896e-05,
|
| 645 |
+
"loss": 0.000399038614705205,
|
| 646 |
+
"step": 890
|
| 647 |
+
},
|
| 648 |
+
{
|
| 649 |
+
"epoch": 1.0146561443066517,
|
| 650 |
+
"grad_norm": 0.013944474048912525,
|
| 651 |
+
"learning_rate": 1.3496744542320952e-05,
|
| 652 |
+
"loss": 0.0006968233734369278,
|
| 653 |
+
"step": 900
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 1.0259301014656144,
|
| 657 |
+
"grad_norm": 0.005086906719952822,
|
| 658 |
+
"learning_rate": 1.3420145538108005e-05,
|
| 659 |
+
"loss": 0.0006555594969540835,
|
| 660 |
+
"step": 910
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"epoch": 1.0372040586245772,
|
| 664 |
+
"grad_norm": 0.004092383664101362,
|
| 665 |
+
"learning_rate": 1.334354653389506e-05,
|
| 666 |
+
"loss": 0.0004657072015106678,
|
| 667 |
+
"step": 920
|
| 668 |
+
},
|
| 669 |
+
{
|
| 670 |
+
"epoch": 1.04847801578354,
|
| 671 |
+
"grad_norm": 1.3944780826568604,
|
| 672 |
+
"learning_rate": 1.3266947529682115e-05,
|
| 673 |
+
"loss": 0.0009271042421460151,
|
| 674 |
+
"step": 930
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 1.059751972942503,
|
| 678 |
+
"grad_norm": 0.004638408776372671,
|
| 679 |
+
"learning_rate": 1.319034852546917e-05,
|
| 680 |
+
"loss": 0.00035289865918457507,
|
| 681 |
+
"step": 940
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 1.0710259301014655,
|
| 685 |
+
"grad_norm": 0.004089709371328354,
|
| 686 |
+
"learning_rate": 1.3113749521256226e-05,
|
| 687 |
+
"loss": 0.0005369900725781917,
|
| 688 |
+
"step": 950
|
| 689 |
+
},
|
| 690 |
+
{
|
| 691 |
+
"epoch": 1.0822998872604284,
|
| 692 |
+
"grad_norm": 0.003287773346528411,
|
| 693 |
+
"learning_rate": 1.303715051704328e-05,
|
| 694 |
+
"loss": 0.04581061005592346,
|
| 695 |
+
"step": 960
|
| 696 |
+
},
|
| 697 |
+
{
|
| 698 |
+
"epoch": 1.0935738444193912,
|
| 699 |
+
"grad_norm": 0.015667209401726723,
|
| 700 |
+
"learning_rate": 1.2960551512830335e-05,
|
| 701 |
+
"loss": 0.0005310946144163609,
|
| 702 |
+
"step": 970
|
| 703 |
+
},
|
| 704 |
+
{
|
| 705 |
+
"epoch": 1.104847801578354,
|
| 706 |
+
"grad_norm": 0.03467559069395065,
|
| 707 |
+
"learning_rate": 1.288395250861739e-05,
|
| 708 |
+
"loss": 0.0005203377455472946,
|
| 709 |
+
"step": 980
|
| 710 |
+
},
|
| 711 |
+
{
|
| 712 |
+
"epoch": 1.116121758737317,
|
| 713 |
+
"grad_norm": 0.00872899405658245,
|
| 714 |
+
"learning_rate": 1.2807353504404442e-05,
|
| 715 |
+
"loss": 0.0024556104093790055,
|
| 716 |
+
"step": 990
|
| 717 |
+
},
|
| 718 |
+
{
|
| 719 |
+
"epoch": 1.1273957158962795,
|
| 720 |
+
"grad_norm": 0.003907124977558851,
|
| 721 |
+
"learning_rate": 1.2730754500191498e-05,
|
| 722 |
+
"loss": 0.0003500781022012234,
|
| 723 |
+
"step": 1000
|
| 724 |
+
},
|
| 725 |
+
{
|
| 726 |
+
"epoch": 1.1386696730552424,
|
| 727 |
+
"grad_norm": 4.096004009246826,
|
| 728 |
+
"learning_rate": 1.2654155495978553e-05,
|
| 729 |
+
"loss": 0.04666096866130829,
|
| 730 |
+
"step": 1010
|
| 731 |
+
},
|
| 732 |
+
{
|
| 733 |
+
"epoch": 1.1499436302142052,
|
| 734 |
+
"grad_norm": 0.03219756856560707,
|
| 735 |
+
"learning_rate": 1.2577556491765607e-05,
|
| 736 |
+
"loss": 0.00044811274856328966,
|
| 737 |
+
"step": 1020
|
| 738 |
+
},
|
| 739 |
+
{
|
| 740 |
+
"epoch": 1.161217587373168,
|
| 741 |
+
"grad_norm": 0.007884475402534008,
|
| 742 |
+
"learning_rate": 1.2500957487552663e-05,
|
| 743 |
+
"loss": 0.004064217209815979,
|
| 744 |
+
"step": 1030
|
| 745 |
+
},
|
| 746 |
+
{
|
| 747 |
+
"epoch": 1.172491544532131,
|
| 748 |
+
"grad_norm": 0.01001141220331192,
|
| 749 |
+
"learning_rate": 1.2424358483339718e-05,
|
| 750 |
+
"loss": 0.08857936263084412,
|
| 751 |
+
"step": 1040
|
| 752 |
+
},
|
| 753 |
+
{
|
| 754 |
+
"epoch": 1.1837655016910935,
|
| 755 |
+
"grad_norm": 0.005555972456932068,
|
| 756 |
+
"learning_rate": 1.2347759479126774e-05,
|
| 757 |
+
"loss": 0.0006975364871323108,
|
| 758 |
+
"step": 1050
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"epoch": 1.1950394588500564,
|
| 762 |
+
"grad_norm": 0.14273758232593536,
|
| 763 |
+
"learning_rate": 1.2271160474913827e-05,
|
| 764 |
+
"loss": 0.0048938382416963576,
|
| 765 |
+
"step": 1060
|
| 766 |
+
},
|
| 767 |
+
{
|
| 768 |
+
"epoch": 1.2063134160090192,
|
| 769 |
+
"grad_norm": 0.005249196197837591,
|
| 770 |
+
"learning_rate": 1.2194561470700881e-05,
|
| 771 |
+
"loss": 0.00038248966448009016,
|
| 772 |
+
"step": 1070
|
| 773 |
+
},
|
| 774 |
+
{
|
| 775 |
+
"epoch": 1.217587373167982,
|
| 776 |
+
"grad_norm": 0.004775014240294695,
|
| 777 |
+
"learning_rate": 1.2117962466487937e-05,
|
| 778 |
+
"loss": 0.00037580323405563833,
|
| 779 |
+
"step": 1080
|
| 780 |
+
},
|
| 781 |
+
{
|
| 782 |
+
"epoch": 1.2288613303269447,
|
| 783 |
+
"grad_norm": 0.005466618109494448,
|
| 784 |
+
"learning_rate": 1.204136346227499e-05,
|
| 785 |
+
"loss": 0.00038032194133847953,
|
| 786 |
+
"step": 1090
|
| 787 |
+
},
|
| 788 |
+
{
|
| 789 |
+
"epoch": 1.2401352874859075,
|
| 790 |
+
"grad_norm": 0.009438618086278439,
|
| 791 |
+
"learning_rate": 1.1964764458062046e-05,
|
| 792 |
+
"loss": 0.0026585064828395844,
|
| 793 |
+
"step": 1100
|
| 794 |
+
},
|
| 795 |
+
{
|
| 796 |
+
"epoch": 1.2514092446448704,
|
| 797 |
+
"grad_norm": 0.0035781103651970625,
|
| 798 |
+
"learning_rate": 1.1888165453849101e-05,
|
| 799 |
+
"loss": 0.000338529609143734,
|
| 800 |
+
"step": 1110
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 1.2626832018038332,
|
| 804 |
+
"grad_norm": 0.008734864182770252,
|
| 805 |
+
"learning_rate": 1.1811566449636157e-05,
|
| 806 |
+
"loss": 0.0003517736680805683,
|
| 807 |
+
"step": 1120
|
| 808 |
+
},
|
| 809 |
+
{
|
| 810 |
+
"epoch": 1.273957158962796,
|
| 811 |
+
"grad_norm": 0.0038098106160759926,
|
| 812 |
+
"learning_rate": 1.173496744542321e-05,
|
| 813 |
+
"loss": 0.0003093461273238063,
|
| 814 |
+
"step": 1130
|
| 815 |
+
},
|
| 816 |
+
{
|
| 817 |
+
"epoch": 1.2852311161217587,
|
| 818 |
+
"grad_norm": 0.004756222013384104,
|
| 819 |
+
"learning_rate": 1.1658368441210266e-05,
|
| 820 |
+
"loss": 0.00035437943879514935,
|
| 821 |
+
"step": 1140
|
| 822 |
+
},
|
| 823 |
+
{
|
| 824 |
+
"epoch": 1.2965050732807215,
|
| 825 |
+
"grad_norm": 0.0054916017688810825,
|
| 826 |
+
"learning_rate": 1.1581769436997321e-05,
|
| 827 |
+
"loss": 0.03466772437095642,
|
| 828 |
+
"step": 1150
|
| 829 |
+
},
|
| 830 |
+
{
|
| 831 |
+
"epoch": 1.3077790304396844,
|
| 832 |
+
"grad_norm": 0.004026887938380241,
|
| 833 |
+
"learning_rate": 1.1505170432784374e-05,
|
| 834 |
+
"loss": 0.018098314106464387,
|
| 835 |
+
"step": 1160
|
| 836 |
+
},
|
| 837 |
+
{
|
| 838 |
+
"epoch": 1.3190529875986472,
|
| 839 |
+
"grad_norm": 0.010358154773712158,
|
| 840 |
+
"learning_rate": 1.1428571428571429e-05,
|
| 841 |
+
"loss": 0.0005080571398139,
|
| 842 |
+
"step": 1170
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"epoch": 1.3303269447576098,
|
| 846 |
+
"grad_norm": 0.006685700733214617,
|
| 847 |
+
"learning_rate": 1.1351972424358484e-05,
|
| 848 |
+
"loss": 0.0006039659958332777,
|
| 849 |
+
"step": 1180
|
| 850 |
+
},
|
| 851 |
+
{
|
| 852 |
+
"epoch": 1.3416009019165727,
|
| 853 |
+
"grad_norm": 0.041763920336961746,
|
| 854 |
+
"learning_rate": 1.1275373420145538e-05,
|
| 855 |
+
"loss": 0.0003080849768593907,
|
| 856 |
+
"step": 1190
|
| 857 |
+
},
|
| 858 |
+
{
|
| 859 |
+
"epoch": 1.3528748590755355,
|
| 860 |
+
"grad_norm": 0.0038634659722447395,
|
| 861 |
+
"learning_rate": 1.1198774415932594e-05,
|
| 862 |
+
"loss": 0.00025509917177259924,
|
| 863 |
+
"step": 1200
|
| 864 |
+
},
|
| 865 |
+
{
|
| 866 |
+
"epoch": 1.3641488162344984,
|
| 867 |
+
"grad_norm": 0.007722954265773296,
|
| 868 |
+
"learning_rate": 1.112217541171965e-05,
|
| 869 |
+
"loss": 0.0002860145643353462,
|
| 870 |
+
"step": 1210
|
| 871 |
+
},
|
| 872 |
+
{
|
| 873 |
+
"epoch": 1.3754227733934612,
|
| 874 |
+
"grad_norm": 0.003805127227678895,
|
| 875 |
+
"learning_rate": 1.1045576407506705e-05,
|
| 876 |
+
"loss": 0.006610937416553497,
|
| 877 |
+
"step": 1220
|
| 878 |
+
},
|
| 879 |
+
{
|
| 880 |
+
"epoch": 1.386696730552424,
|
| 881 |
+
"grad_norm": 0.003794404910877347,
|
| 882 |
+
"learning_rate": 1.0968977403293758e-05,
|
| 883 |
+
"loss": 0.0002713572233915329,
|
| 884 |
+
"step": 1230
|
| 885 |
+
},
|
| 886 |
+
{
|
| 887 |
+
"epoch": 1.3979706877113867,
|
| 888 |
+
"grad_norm": 0.0029455258045345545,
|
| 889 |
+
"learning_rate": 1.0892378399080812e-05,
|
| 890 |
+
"loss": 0.0002766825258731842,
|
| 891 |
+
"step": 1240
|
| 892 |
+
},
|
| 893 |
+
{
|
| 894 |
+
"epoch": 1.4092446448703495,
|
| 895 |
+
"grad_norm": 0.0034676918294280767,
|
| 896 |
+
"learning_rate": 1.0815779394867868e-05,
|
| 897 |
+
"loss": 0.0002525749383494258,
|
| 898 |
+
"step": 1250
|
| 899 |
+
},
|
| 900 |
+
{
|
| 901 |
+
"epoch": 1.4205186020293123,
|
| 902 |
+
"grad_norm": 0.0027564624324440956,
|
| 903 |
+
"learning_rate": 1.0739180390654921e-05,
|
| 904 |
+
"loss": 0.024026018381118775,
|
| 905 |
+
"step": 1260
|
| 906 |
+
},
|
| 907 |
+
{
|
| 908 |
+
"epoch": 1.431792559188275,
|
| 909 |
+
"grad_norm": 0.004176211077719927,
|
| 910 |
+
"learning_rate": 1.0662581386441977e-05,
|
| 911 |
+
"loss": 0.04461250603199005,
|
| 912 |
+
"step": 1270
|
| 913 |
+
},
|
| 914 |
+
{
|
| 915 |
+
"epoch": 1.4430665163472378,
|
| 916 |
+
"grad_norm": 0.7301909327507019,
|
| 917 |
+
"learning_rate": 1.0585982382229032e-05,
|
| 918 |
+
"loss": 0.0013874154537916183,
|
| 919 |
+
"step": 1280
|
| 920 |
+
},
|
| 921 |
+
{
|
| 922 |
+
"epoch": 1.4543404735062007,
|
| 923 |
+
"grad_norm": 0.006072559859603643,
|
| 924 |
+
"learning_rate": 1.0509383378016086e-05,
|
| 925 |
+
"loss": 0.0008793852292001247,
|
| 926 |
+
"step": 1290
|
| 927 |
+
},
|
| 928 |
+
{
|
| 929 |
+
"epoch": 1.4656144306651635,
|
| 930 |
+
"grad_norm": 0.003349730744957924,
|
| 931 |
+
"learning_rate": 1.0432784373803142e-05,
|
| 932 |
+
"loss": 0.11177332401275634,
|
| 933 |
+
"step": 1300
|
| 934 |
+
},
|
| 935 |
+
{
|
| 936 |
+
"epoch": 1.4768883878241263,
|
| 937 |
+
"grad_norm": 0.0037300498224794865,
|
| 938 |
+
"learning_rate": 1.0356185369590197e-05,
|
| 939 |
+
"loss": 0.02746536731719971,
|
| 940 |
+
"step": 1310
|
| 941 |
+
},
|
| 942 |
+
{
|
| 943 |
+
"epoch": 1.4881623449830892,
|
| 944 |
+
"grad_norm": 0.0067192791029810905,
|
| 945 |
+
"learning_rate": 1.027958636537725e-05,
|
| 946 |
+
"loss": 0.0006490409374237061,
|
| 947 |
+
"step": 1320
|
| 948 |
+
},
|
| 949 |
+
{
|
| 950 |
+
"epoch": 1.4994363021420518,
|
| 951 |
+
"grad_norm": 0.007033471018075943,
|
| 952 |
+
"learning_rate": 1.0202987361164305e-05,
|
| 953 |
+
"loss": 0.00473451130092144,
|
| 954 |
+
"step": 1330
|
| 955 |
+
},
|
| 956 |
+
{
|
| 957 |
+
"epoch": 1.5107102593010147,
|
| 958 |
+
"grad_norm": 0.005484273657202721,
|
| 959 |
+
"learning_rate": 1.012638835695136e-05,
|
| 960 |
+
"loss": 0.0010361320339143277,
|
| 961 |
+
"step": 1340
|
| 962 |
+
},
|
| 963 |
+
{
|
| 964 |
+
"epoch": 1.5219842164599775,
|
| 965 |
+
"grad_norm": 0.003371752332895994,
|
| 966 |
+
"learning_rate": 1.0049789352738416e-05,
|
| 967 |
+
"loss": 0.00029721735045313836,
|
| 968 |
+
"step": 1350
|
| 969 |
+
},
|
| 970 |
+
{
|
| 971 |
+
"epoch": 1.5332581736189401,
|
| 972 |
+
"grad_norm": 0.003974182065576315,
|
| 973 |
+
"learning_rate": 9.97319034852547e-06,
|
| 974 |
+
"loss": 0.00026908733416348697,
|
| 975 |
+
"step": 1360
|
| 976 |
+
},
|
| 977 |
+
{
|
| 978 |
+
"epoch": 1.544532130777903,
|
| 979 |
+
"grad_norm": 0.003871662775054574,
|
| 980 |
+
"learning_rate": 9.896591344312525e-06,
|
| 981 |
+
"loss": 0.0018835226073861122,
|
| 982 |
+
"step": 1370
|
| 983 |
+
},
|
| 984 |
+
{
|
| 985 |
+
"epoch": 1.5558060879368658,
|
| 986 |
+
"grad_norm": 0.0033102212473750114,
|
| 987 |
+
"learning_rate": 9.81999234009958e-06,
|
| 988 |
+
"loss": 0.0002722503151744604,
|
| 989 |
+
"step": 1380
|
| 990 |
+
},
|
| 991 |
+
{
|
| 992 |
+
"epoch": 1.5670800450958287,
|
| 993 |
+
"grad_norm": 0.0036641808692365885,
|
| 994 |
+
"learning_rate": 9.743393335886634e-06,
|
| 995 |
+
"loss": 0.0002578678075224161,
|
| 996 |
+
"step": 1390
|
| 997 |
+
},
|
| 998 |
+
{
|
| 999 |
+
"epoch": 1.5783540022547915,
|
| 1000 |
+
"grad_norm": 0.0043280464597046375,
|
| 1001 |
+
"learning_rate": 9.66679433167369e-06,
|
| 1002 |
+
"loss": 0.000738756125792861,
|
| 1003 |
+
"step": 1400
|
| 1004 |
+
},
|
| 1005 |
+
{
|
| 1006 |
+
"epoch": 1.5896279594137543,
|
| 1007 |
+
"grad_norm": 0.004314345773309469,
|
| 1008 |
+
"learning_rate": 9.590195327460743e-06,
|
| 1009 |
+
"loss": 0.009816586226224899,
|
| 1010 |
+
"step": 1410
|
| 1011 |
+
},
|
| 1012 |
+
{
|
| 1013 |
+
"epoch": 1.6009019165727172,
|
| 1014 |
+
"grad_norm": 0.0035737520083785057,
|
| 1015 |
+
"learning_rate": 9.513596323247799e-06,
|
| 1016 |
+
"loss": 0.0002666026121005416,
|
| 1017 |
+
"step": 1420
|
| 1018 |
+
},
|
| 1019 |
+
{
|
| 1020 |
+
"epoch": 1.6121758737316798,
|
| 1021 |
+
"grad_norm": 0.006729898974299431,
|
| 1022 |
+
"learning_rate": 9.436997319034853e-06,
|
| 1023 |
+
"loss": 0.04153406620025635,
|
| 1024 |
+
"step": 1430
|
| 1025 |
+
},
|
| 1026 |
+
{
|
| 1027 |
+
"epoch": 1.6234498308906427,
|
| 1028 |
+
"grad_norm": 0.003635931294411421,
|
| 1029 |
+
"learning_rate": 9.360398314821908e-06,
|
| 1030 |
+
"loss": 0.0002528228098526597,
|
| 1031 |
+
"step": 1440
|
| 1032 |
+
},
|
| 1033 |
+
{
|
| 1034 |
+
"epoch": 1.6347237880496053,
|
| 1035 |
+
"grad_norm": 0.003705125069245696,
|
| 1036 |
+
"learning_rate": 9.283799310608964e-06,
|
| 1037 |
+
"loss": 0.0006835301406681538,
|
| 1038 |
+
"step": 1450
|
| 1039 |
+
},
|
| 1040 |
+
{
|
| 1041 |
+
"epoch": 1.6459977452085681,
|
| 1042 |
+
"grad_norm": 0.03155064582824707,
|
| 1043 |
+
"learning_rate": 9.207200306396017e-06,
|
| 1044 |
+
"loss": 0.0004594951402395964,
|
| 1045 |
+
"step": 1460
|
| 1046 |
+
},
|
| 1047 |
+
{
|
| 1048 |
+
"epoch": 1.657271702367531,
|
| 1049 |
+
"grad_norm": 0.005488164257258177,
|
| 1050 |
+
"learning_rate": 9.130601302183073e-06,
|
| 1051 |
+
"loss": 0.0002626030007377267,
|
| 1052 |
+
"step": 1470
|
| 1053 |
+
},
|
| 1054 |
+
{
|
| 1055 |
+
"epoch": 1.6685456595264938,
|
| 1056 |
+
"grad_norm": 0.004056842997670174,
|
| 1057 |
+
"learning_rate": 9.054002297970127e-06,
|
| 1058 |
+
"loss": 0.00027129012160003184,
|
| 1059 |
+
"step": 1480
|
| 1060 |
+
},
|
| 1061 |
+
{
|
| 1062 |
+
"epoch": 1.6798196166854567,
|
| 1063 |
+
"grad_norm": 0.004610530100762844,
|
| 1064 |
+
"learning_rate": 8.977403293757182e-06,
|
| 1065 |
+
"loss": 0.0004437011666595936,
|
| 1066 |
+
"step": 1490
|
| 1067 |
+
},
|
| 1068 |
+
{
|
| 1069 |
+
"epoch": 1.6910935738444195,
|
| 1070 |
+
"grad_norm": 0.0049287667497992516,
|
| 1071 |
+
"learning_rate": 8.900804289544237e-06,
|
| 1072 |
+
"loss": 0.0002617941936478019,
|
| 1073 |
+
"step": 1500
|
| 1074 |
+
},
|
| 1075 |
+
{
|
| 1076 |
+
"epoch": 1.7023675310033823,
|
| 1077 |
+
"grad_norm": 0.0032671773806214333,
|
| 1078 |
+
"learning_rate": 8.824205285331291e-06,
|
| 1079 |
+
"loss": 0.0003260360099375248,
|
| 1080 |
+
"step": 1510
|
| 1081 |
+
},
|
| 1082 |
+
{
|
| 1083 |
+
"epoch": 1.713641488162345,
|
| 1084 |
+
"grad_norm": 0.0026306200306862593,
|
| 1085 |
+
"learning_rate": 8.747606281118347e-06,
|
| 1086 |
+
"loss": 0.040199264883995056,
|
| 1087 |
+
"step": 1520
|
| 1088 |
+
},
|
| 1089 |
+
{
|
| 1090 |
+
"epoch": 1.7249154453213078,
|
| 1091 |
+
"grad_norm": 0.002604865934699774,
|
| 1092 |
+
"learning_rate": 8.6710072769054e-06,
|
| 1093 |
+
"loss": 0.0021020427346229555,
|
| 1094 |
+
"step": 1530
|
| 1095 |
+
},
|
| 1096 |
+
{
|
| 1097 |
+
"epoch": 1.7361894024802704,
|
| 1098 |
+
"grad_norm": 0.004267075564712286,
|
| 1099 |
+
"learning_rate": 8.594408272692456e-06,
|
| 1100 |
+
"loss": 0.0005022045224905014,
|
| 1101 |
+
"step": 1540
|
| 1102 |
+
},
|
| 1103 |
+
{
|
| 1104 |
+
"epoch": 1.7474633596392333,
|
| 1105 |
+
"grad_norm": 0.003063610754907131,
|
| 1106 |
+
"learning_rate": 8.51780926847951e-06,
|
| 1107 |
+
"loss": 0.00021760372910648585,
|
| 1108 |
+
"step": 1550
|
| 1109 |
+
},
|
| 1110 |
+
{
|
| 1111 |
+
"epoch": 1.7587373167981961,
|
| 1112 |
+
"grad_norm": 0.002829270903021097,
|
| 1113 |
+
"learning_rate": 8.441210264266565e-06,
|
| 1114 |
+
"loss": 0.0003501762170344591,
|
| 1115 |
+
"step": 1560
|
| 1116 |
+
},
|
| 1117 |
+
{
|
| 1118 |
+
"epoch": 1.770011273957159,
|
| 1119 |
+
"grad_norm": 0.0024476037360727787,
|
| 1120 |
+
"learning_rate": 8.36461126005362e-06,
|
| 1121 |
+
"loss": 0.0002327218884602189,
|
| 1122 |
+
"step": 1570
|
| 1123 |
+
},
|
| 1124 |
+
{
|
| 1125 |
+
"epoch": 1.7812852311161218,
|
| 1126 |
+
"grad_norm": 0.003161813598126173,
|
| 1127 |
+
"learning_rate": 8.288012255840674e-06,
|
| 1128 |
+
"loss": 0.0002234887331724167,
|
| 1129 |
+
"step": 1580
|
| 1130 |
+
},
|
| 1131 |
+
{
|
| 1132 |
+
"epoch": 1.7925591882750846,
|
| 1133 |
+
"grad_norm": 0.002495658351108432,
|
| 1134 |
+
"learning_rate": 8.21141325162773e-06,
|
| 1135 |
+
"loss": 0.0002736057387664914,
|
| 1136 |
+
"step": 1590
|
| 1137 |
+
},
|
| 1138 |
+
{
|
| 1139 |
+
"epoch": 1.8038331454340475,
|
| 1140 |
+
"grad_norm": 0.0035063582472503185,
|
| 1141 |
+
"learning_rate": 8.134814247414784e-06,
|
| 1142 |
+
"loss": 0.0002091569360345602,
|
| 1143 |
+
"step": 1600
|
| 1144 |
+
},
|
| 1145 |
+
{
|
| 1146 |
+
"epoch": 1.8151071025930101,
|
| 1147 |
+
"grad_norm": 0.022207682952284813,
|
| 1148 |
+
"learning_rate": 8.05821524320184e-06,
|
| 1149 |
+
"loss": 0.04271211624145508,
|
| 1150 |
+
"step": 1610
|
| 1151 |
+
},
|
| 1152 |
+
{
|
| 1153 |
+
"epoch": 1.826381059751973,
|
| 1154 |
+
"grad_norm": 0.002612703014165163,
|
| 1155 |
+
"learning_rate": 7.981616238988895e-06,
|
| 1156 |
+
"loss": 0.00029878674540668727,
|
| 1157 |
+
"step": 1620
|
| 1158 |
+
},
|
| 1159 |
+
{
|
| 1160 |
+
"epoch": 1.8376550169109356,
|
| 1161 |
+
"grad_norm": 26.909603118896484,
|
| 1162 |
+
"learning_rate": 7.905017234775948e-06,
|
| 1163 |
+
"loss": 0.06723988056182861,
|
| 1164 |
+
"step": 1630
|
| 1165 |
+
},
|
| 1166 |
+
{
|
| 1167 |
+
"epoch": 1.8489289740698984,
|
| 1168 |
+
"grad_norm": 0.002532525220885873,
|
| 1169 |
+
"learning_rate": 7.828418230563002e-06,
|
| 1170 |
+
"loss": 0.00022608181461691856,
|
| 1171 |
+
"step": 1640
|
| 1172 |
+
},
|
| 1173 |
+
{
|
| 1174 |
+
"epoch": 1.8602029312288613,
|
| 1175 |
+
"grad_norm": 0.002735898597165942,
|
| 1176 |
+
"learning_rate": 7.751819226350058e-06,
|
| 1177 |
+
"loss": 0.08260895013809204,
|
| 1178 |
+
"step": 1650
|
| 1179 |
+
},
|
| 1180 |
+
{
|
| 1181 |
+
"epoch": 1.8714768883878241,
|
| 1182 |
+
"grad_norm": 0.01515868492424488,
|
| 1183 |
+
"learning_rate": 7.675220222137113e-06,
|
| 1184 |
+
"loss": 0.00023219306021928787,
|
| 1185 |
+
"step": 1660
|
| 1186 |
+
},
|
| 1187 |
+
{
|
| 1188 |
+
"epoch": 1.882750845546787,
|
| 1189 |
+
"grad_norm": 0.0035585863515734673,
|
| 1190 |
+
"learning_rate": 7.598621217924168e-06,
|
| 1191 |
+
"loss": 0.00048695728182792666,
|
| 1192 |
+
"step": 1670
|
| 1193 |
+
},
|
| 1194 |
+
{
|
| 1195 |
+
"epoch": 1.8940248027057498,
|
| 1196 |
+
"grad_norm": 0.00424063578248024,
|
| 1197 |
+
"learning_rate": 7.522022213711222e-06,
|
| 1198 |
+
"loss": 0.0003054874483495951,
|
| 1199 |
+
"step": 1680
|
| 1200 |
+
},
|
| 1201 |
+
{
|
| 1202 |
+
"epoch": 1.9052987598647126,
|
| 1203 |
+
"grad_norm": 0.004633221309632063,
|
| 1204 |
+
"learning_rate": 7.445423209498277e-06,
|
| 1205 |
+
"loss": 0.0031257856637239454,
|
| 1206 |
+
"step": 1690
|
| 1207 |
+
},
|
| 1208 |
+
{
|
| 1209 |
+
"epoch": 1.9165727170236753,
|
| 1210 |
+
"grad_norm": 0.0049049886874854565,
|
| 1211 |
+
"learning_rate": 7.368824205285332e-06,
|
| 1212 |
+
"loss": 0.00021440631244331598,
|
| 1213 |
+
"step": 1700
|
| 1214 |
+
},
|
| 1215 |
+
{
|
| 1216 |
+
"epoch": 1.927846674182638,
|
| 1217 |
+
"grad_norm": 0.01473376713693142,
|
| 1218 |
+
"learning_rate": 7.292225201072387e-06,
|
| 1219 |
+
"loss": 0.0002546449890360236,
|
| 1220 |
+
"step": 1710
|
| 1221 |
+
},
|
| 1222 |
+
{
|
| 1223 |
+
"epoch": 1.939120631341601,
|
| 1224 |
+
"grad_norm": 0.024918919429183006,
|
| 1225 |
+
"learning_rate": 7.215626196859441e-06,
|
| 1226 |
+
"loss": 0.0002916081575676799,
|
| 1227 |
+
"step": 1720
|
| 1228 |
+
},
|
| 1229 |
+
{
|
| 1230 |
+
"epoch": 1.9503945885005636,
|
| 1231 |
+
"grad_norm": 0.008383645676076412,
|
| 1232 |
+
"learning_rate": 7.139027192646496e-06,
|
| 1233 |
+
"loss": 0.0002473884029313922,
|
| 1234 |
+
"step": 1730
|
| 1235 |
+
},
|
| 1236 |
+
{
|
| 1237 |
+
"epoch": 1.9616685456595264,
|
| 1238 |
+
"grad_norm": 0.003205355489626527,
|
| 1239 |
+
"learning_rate": 7.062428188433551e-06,
|
| 1240 |
+
"loss": 0.00030408557504415514,
|
| 1241 |
+
"step": 1740
|
| 1242 |
+
},
|
| 1243 |
+
{
|
| 1244 |
+
"epoch": 1.9729425028184893,
|
| 1245 |
+
"grad_norm": 0.0037417688872665167,
|
| 1246 |
+
"learning_rate": 6.9858291842206064e-06,
|
| 1247 |
+
"loss": 0.00022464105859398842,
|
| 1248 |
+
"step": 1750
|
| 1249 |
+
},
|
| 1250 |
+
{
|
| 1251 |
+
"epoch": 1.984216459977452,
|
| 1252 |
+
"grad_norm": 0.0024626674130558968,
|
| 1253 |
+
"learning_rate": 6.90923018000766e-06,
|
| 1254 |
+
"loss": 0.0002444717800244689,
|
| 1255 |
+
"step": 1760
|
| 1256 |
+
},
|
| 1257 |
+
{
|
| 1258 |
+
"epoch": 1.995490417136415,
|
| 1259 |
+
"grad_norm": 0.0025584984105080366,
|
| 1260 |
+
"learning_rate": 6.832631175794715e-06,
|
| 1261 |
+
"loss": 0.00020699049346148968,
|
| 1262 |
+
"step": 1770
|
| 1263 |
+
},
|
| 1264 |
+
{
|
| 1265 |
+
"epoch": 2.0,
|
| 1266 |
+
"eval_avg_accuracy": 0.9983074753173484,
|
| 1267 |
+
"eval_explanation_accuracy": 0.9983074753173484,
|
| 1268 |
+
"eval_explanation_f1": 0.9966740576496674,
|
| 1269 |
+
"eval_loss": 0.017838388681411743,
|
| 1270 |
+
"eval_refusal_accuracy": 0.9983074753173484,
|
| 1271 |
+
"eval_refusal_f1": 0.9983079526226735,
|
| 1272 |
+
"eval_runtime": 8.1647,
|
| 1273 |
+
"eval_samples_per_second": 434.186,
|
| 1274 |
+
"eval_steps_per_second": 27.19,
|
| 1275 |
+
"step": 1774
|
| 1276 |
+
},
|
| 1277 |
+
{
|
| 1278 |
+
"epoch": 2.006764374295378,
|
| 1279 |
+
"grad_norm": 0.0024250212591141462,
|
| 1280 |
+
"learning_rate": 6.75603217158177e-06,
|
| 1281 |
+
"loss": 0.00018842160934582353,
|
| 1282 |
+
"step": 1780
|
| 1283 |
+
},
|
| 1284 |
+
{
|
| 1285 |
+
"epoch": 2.0180383314543406,
|
| 1286 |
+
"grad_norm": 5.130032539367676,
|
| 1287 |
+
"learning_rate": 6.679433167368825e-06,
|
| 1288 |
+
"loss": 0.0034407805651426315,
|
| 1289 |
+
"step": 1790
|
| 1290 |
+
},
|
| 1291 |
+
{
|
| 1292 |
+
"epoch": 2.0293122886133035,
|
| 1293 |
+
"grad_norm": 0.0062168375588953495,
|
| 1294 |
+
"learning_rate": 6.602834163155879e-06,
|
| 1295 |
+
"loss": 0.00020795222371816635,
|
| 1296 |
+
"step": 1800
|
| 1297 |
+
},
|
| 1298 |
+
{
|
| 1299 |
+
"epoch": 2.040586245772266,
|
| 1300 |
+
"grad_norm": 0.002221637172624469,
|
| 1301 |
+
"learning_rate": 6.526235158942934e-06,
|
| 1302 |
+
"loss": 0.0002968248911201954,
|
| 1303 |
+
"step": 1810
|
| 1304 |
+
},
|
| 1305 |
+
{
|
| 1306 |
+
"epoch": 2.0518602029312287,
|
| 1307 |
+
"grad_norm": 0.002862810157239437,
|
| 1308 |
+
"learning_rate": 6.449636154729989e-06,
|
| 1309 |
+
"loss": 0.05430693626403808,
|
| 1310 |
+
"step": 1820
|
| 1311 |
+
},
|
| 1312 |
+
{
|
| 1313 |
+
"epoch": 2.0631341600901916,
|
| 1314 |
+
"grad_norm": 0.0035138484090566635,
|
| 1315 |
+
"learning_rate": 6.373037150517044e-06,
|
| 1316 |
+
"loss": 0.0002286258852109313,
|
| 1317 |
+
"step": 1830
|
| 1318 |
+
},
|
| 1319 |
+
{
|
| 1320 |
+
"epoch": 2.0744081172491544,
|
| 1321 |
+
"grad_norm": 0.0028608141001313925,
|
| 1322 |
+
"learning_rate": 6.296438146304099e-06,
|
| 1323 |
+
"loss": 0.00018272822489961982,
|
| 1324 |
+
"step": 1840
|
| 1325 |
+
},
|
| 1326 |
+
{
|
| 1327 |
+
"epoch": 2.0856820744081173,
|
| 1328 |
+
"grad_norm": 0.002431184286251664,
|
| 1329 |
+
"learning_rate": 6.219839142091153e-06,
|
| 1330 |
+
"loss": 0.00018496520351618528,
|
| 1331 |
+
"step": 1850
|
| 1332 |
+
},
|
| 1333 |
+
{
|
| 1334 |
+
"epoch": 2.09695603156708,
|
| 1335 |
+
"grad_norm": 0.0025748296175152063,
|
| 1336 |
+
"learning_rate": 6.143240137878208e-06,
|
| 1337 |
+
"loss": 0.00017570253694429994,
|
| 1338 |
+
"step": 1860
|
| 1339 |
+
},
|
| 1340 |
+
{
|
| 1341 |
+
"epoch": 2.108229988726043,
|
| 1342 |
+
"grad_norm": 0.0030030717607587576,
|
| 1343 |
+
"learning_rate": 6.066641133665263e-06,
|
| 1344 |
+
"loss": 0.0017355974763631822,
|
| 1345 |
+
"step": 1870
|
| 1346 |
+
},
|
| 1347 |
+
{
|
| 1348 |
+
"epoch": 2.119503945885006,
|
| 1349 |
+
"grad_norm": 0.004449139349162579,
|
| 1350 |
+
"learning_rate": 5.990042129452318e-06,
|
| 1351 |
+
"loss": 0.00039322059601545336,
|
| 1352 |
+
"step": 1880
|
| 1353 |
+
},
|
| 1354 |
+
{
|
| 1355 |
+
"epoch": 2.1307779030439686,
|
| 1356 |
+
"grad_norm": 0.0028530703857541084,
|
| 1357 |
+
"learning_rate": 5.913443125239372e-06,
|
| 1358 |
+
"loss": 0.00020313062705099583,
|
| 1359 |
+
"step": 1890
|
| 1360 |
+
},
|
| 1361 |
+
{
|
| 1362 |
+
"epoch": 2.142051860202931,
|
| 1363 |
+
"grad_norm": 0.002087223343551159,
|
| 1364 |
+
"learning_rate": 5.8368441210264275e-06,
|
| 1365 |
+
"loss": 0.00018089411314576865,
|
| 1366 |
+
"step": 1900
|
| 1367 |
+
},
|
| 1368 |
+
{
|
| 1369 |
+
"epoch": 2.153325817361894,
|
| 1370 |
+
"grad_norm": 0.002159111900255084,
|
| 1371 |
+
"learning_rate": 5.760245116813482e-06,
|
| 1372 |
+
"loss": 0.00019215191714465617,
|
| 1373 |
+
"step": 1910
|
| 1374 |
+
},
|
| 1375 |
+
{
|
| 1376 |
+
"epoch": 2.1645997745208567,
|
| 1377 |
+
"grad_norm": 0.00597824202850461,
|
| 1378 |
+
"learning_rate": 5.683646112600537e-06,
|
| 1379 |
+
"loss": 0.0001949567347764969,
|
| 1380 |
+
"step": 1920
|
| 1381 |
+
},
|
| 1382 |
+
{
|
| 1383 |
+
"epoch": 2.1758737316798196,
|
| 1384 |
+
"grad_norm": 0.0023554086219519377,
|
| 1385 |
+
"learning_rate": 5.607047108387591e-06,
|
| 1386 |
+
"loss": 0.00044615250080823896,
|
| 1387 |
+
"step": 1930
|
| 1388 |
+
},
|
| 1389 |
+
{
|
| 1390 |
+
"epoch": 2.1871476888387824,
|
| 1391 |
+
"grad_norm": 0.00262109050527215,
|
| 1392 |
+
"learning_rate": 5.530448104174646e-06,
|
| 1393 |
+
"loss": 0.00020969314500689505,
|
| 1394 |
+
"step": 1940
|
| 1395 |
+
},
|
| 1396 |
+
{
|
| 1397 |
+
"epoch": 2.1984216459977453,
|
| 1398 |
+
"grad_norm": 0.00705662602558732,
|
| 1399 |
+
"learning_rate": 5.453849099961701e-06,
|
| 1400 |
+
"loss": 0.00018844108562916518,
|
| 1401 |
+
"step": 1950
|
| 1402 |
+
},
|
| 1403 |
+
{
|
| 1404 |
+
"epoch": 2.209695603156708,
|
| 1405 |
+
"grad_norm": 0.0033652179408818483,
|
| 1406 |
+
"learning_rate": 5.377250095748756e-06,
|
| 1407 |
+
"loss": 0.00018388747703284024,
|
| 1408 |
+
"step": 1960
|
| 1409 |
+
},
|
| 1410 |
+
{
|
| 1411 |
+
"epoch": 2.220969560315671,
|
| 1412 |
+
"grad_norm": 0.0029013140592724085,
|
| 1413 |
+
"learning_rate": 5.30065109153581e-06,
|
| 1414 |
+
"loss": 0.00023608410265296698,
|
| 1415 |
+
"step": 1970
|
| 1416 |
+
},
|
| 1417 |
+
{
|
| 1418 |
+
"epoch": 2.232243517474634,
|
| 1419 |
+
"grad_norm": 0.0023085789289325476,
|
| 1420 |
+
"learning_rate": 5.224052087322865e-06,
|
| 1421 |
+
"loss": 0.04250820279121399,
|
| 1422 |
+
"step": 1980
|
| 1423 |
+
},
|
| 1424 |
+
{
|
| 1425 |
+
"epoch": 2.243517474633596,
|
| 1426 |
+
"grad_norm": 0.002931625349447131,
|
| 1427 |
+
"learning_rate": 5.14745308310992e-06,
|
| 1428 |
+
"loss": 0.0001782746287062764,
|
| 1429 |
+
"step": 1990
|
| 1430 |
+
},
|
| 1431 |
+
{
|
| 1432 |
+
"epoch": 2.254791431792559,
|
| 1433 |
+
"grad_norm": 0.014203422702848911,
|
| 1434 |
+
"learning_rate": 5.070854078896975e-06,
|
| 1435 |
+
"loss": 0.00047393282875418663,
|
| 1436 |
+
"step": 2000
|
| 1437 |
+
},
|
| 1438 |
+
{
|
| 1439 |
+
"epoch": 2.266065388951522,
|
| 1440 |
+
"grad_norm": 0.002238233108073473,
|
| 1441 |
+
"learning_rate": 4.99425507468403e-06,
|
| 1442 |
+
"loss": 0.00019599059596657752,
|
| 1443 |
+
"step": 2010
|
| 1444 |
+
},
|
| 1445 |
+
{
|
| 1446 |
+
"epoch": 2.2773393461104847,
|
| 1447 |
+
"grad_norm": 0.0791219025850296,
|
| 1448 |
+
"learning_rate": 4.917656070471084e-06,
|
| 1449 |
+
"loss": 0.00023696895223110915,
|
| 1450 |
+
"step": 2020
|
| 1451 |
+
},
|
| 1452 |
+
{
|
| 1453 |
+
"epoch": 2.2886133032694476,
|
| 1454 |
+
"grad_norm": 0.003860430559143424,
|
| 1455 |
+
"learning_rate": 4.841057066258139e-06,
|
| 1456 |
+
"loss": 0.03795796036720276,
|
| 1457 |
+
"step": 2030
|
| 1458 |
+
},
|
| 1459 |
+
{
|
| 1460 |
+
"epoch": 2.2998872604284104,
|
| 1461 |
+
"grad_norm": 0.003089021658524871,
|
| 1462 |
+
"learning_rate": 4.764458062045194e-06,
|
| 1463 |
+
"loss": 0.0005109061021357775,
|
| 1464 |
+
"step": 2040
|
| 1465 |
+
},
|
| 1466 |
+
{
|
| 1467 |
+
"epoch": 2.3111612175873733,
|
| 1468 |
+
"grad_norm": 0.007778828032314777,
|
| 1469 |
+
"learning_rate": 4.6878590578322485e-06,
|
| 1470 |
+
"loss": 0.019122488796710968,
|
| 1471 |
+
"step": 2050
|
| 1472 |
+
},
|
| 1473 |
+
{
|
| 1474 |
+
"epoch": 2.322435174746336,
|
| 1475 |
+
"grad_norm": 0.010298606008291245,
|
| 1476 |
+
"learning_rate": 4.611260053619303e-06,
|
| 1477 |
+
"loss": 0.0003088724333792925,
|
| 1478 |
+
"step": 2060
|
| 1479 |
+
},
|
| 1480 |
+
{
|
| 1481 |
+
"epoch": 2.333709131905299,
|
| 1482 |
+
"grad_norm": 0.005622932221740484,
|
| 1483 |
+
"learning_rate": 4.534661049406358e-06,
|
| 1484 |
+
"loss": 0.0003277263371273875,
|
| 1485 |
+
"step": 2070
|
| 1486 |
+
},
|
| 1487 |
+
{
|
| 1488 |
+
"epoch": 2.344983089064262,
|
| 1489 |
+
"grad_norm": 0.0020355971064418554,
|
| 1490 |
+
"learning_rate": 4.458062045193413e-06,
|
| 1491 |
+
"loss": 0.0009053186513483524,
|
| 1492 |
+
"step": 2080
|
| 1493 |
+
},
|
| 1494 |
+
{
|
| 1495 |
+
"epoch": 2.356257046223224,
|
| 1496 |
+
"grad_norm": 0.002295851707458496,
|
| 1497 |
+
"learning_rate": 4.381463040980468e-06,
|
| 1498 |
+
"loss": 0.0001918965601362288,
|
| 1499 |
+
"step": 2090
|
| 1500 |
+
},
|
| 1501 |
+
{
|
| 1502 |
+
"epoch": 2.367531003382187,
|
| 1503 |
+
"grad_norm": 0.026092899963259697,
|
| 1504 |
+
"learning_rate": 4.3048640367675224e-06,
|
| 1505 |
+
"loss": 0.00019979043863713742,
|
| 1506 |
+
"step": 2100
|
| 1507 |
+
},
|
| 1508 |
+
{
|
| 1509 |
+
"epoch": 2.37880496054115,
|
| 1510 |
+
"grad_norm": 0.0031999810598790646,
|
| 1511 |
+
"learning_rate": 4.228265032554577e-06,
|
| 1512 |
+
"loss": 0.00046187713742256163,
|
| 1513 |
+
"step": 2110
|
| 1514 |
+
},
|
| 1515 |
+
{
|
| 1516 |
+
"epoch": 2.3900789177001127,
|
| 1517 |
+
"grad_norm": 7.394179344177246,
|
| 1518 |
+
"learning_rate": 4.1516660283416325e-06,
|
| 1519 |
+
"loss": 0.09607717394828796,
|
| 1520 |
+
"step": 2120
|
| 1521 |
+
},
|
| 1522 |
+
{
|
| 1523 |
+
"epoch": 2.4013528748590756,
|
| 1524 |
+
"grad_norm": 0.003000697586685419,
|
| 1525 |
+
"learning_rate": 4.075067024128686e-06,
|
| 1526 |
+
"loss": 0.0002733314177021384,
|
| 1527 |
+
"step": 2130
|
| 1528 |
+
},
|
| 1529 |
+
{
|
| 1530 |
+
"epoch": 2.4126268320180384,
|
| 1531 |
+
"grad_norm": 0.0024585698265582323,
|
| 1532 |
+
"learning_rate": 3.998468019915742e-06,
|
| 1533 |
+
"loss": 0.00020082059781998396,
|
| 1534 |
+
"step": 2140
|
| 1535 |
+
},
|
| 1536 |
+
{
|
| 1537 |
+
"epoch": 2.4239007891770012,
|
| 1538 |
+
"grad_norm": 0.009559140540659428,
|
| 1539 |
+
"learning_rate": 3.921869015702796e-06,
|
| 1540 |
+
"loss": 0.00020159895066171885,
|
| 1541 |
+
"step": 2150
|
| 1542 |
+
},
|
| 1543 |
+
{
|
| 1544 |
+
"epoch": 2.435174746335964,
|
| 1545 |
+
"grad_norm": 0.002507139928638935,
|
| 1546 |
+
"learning_rate": 3.845270011489851e-06,
|
| 1547 |
+
"loss": 0.00019097424810752273,
|
| 1548 |
+
"step": 2160
|
| 1549 |
+
},
|
| 1550 |
+
{
|
| 1551 |
+
"epoch": 2.4464487034949265,
|
| 1552 |
+
"grad_norm": 0.003593851812183857,
|
| 1553 |
+
"learning_rate": 3.7686710072769056e-06,
|
| 1554 |
+
"loss": 0.00019985807593911887,
|
| 1555 |
+
"step": 2170
|
| 1556 |
+
},
|
| 1557 |
+
{
|
| 1558 |
+
"epoch": 2.4577226606538893,
|
| 1559 |
+
"grad_norm": 0.0018217683536931872,
|
| 1560 |
+
"learning_rate": 3.6920720030639607e-06,
|
| 1561 |
+
"loss": 0.00018714518519118427,
|
| 1562 |
+
"step": 2180
|
| 1563 |
+
},
|
| 1564 |
+
{
|
| 1565 |
+
"epoch": 2.468996617812852,
|
| 1566 |
+
"grad_norm": 0.0026852686423808336,
|
| 1567 |
+
"learning_rate": 3.6154729988510153e-06,
|
| 1568 |
+
"loss": 0.00018492992967367172,
|
| 1569 |
+
"step": 2190
|
| 1570 |
+
},
|
| 1571 |
+
{
|
| 1572 |
+
"epoch": 2.480270574971815,
|
| 1573 |
+
"grad_norm": 0.004490630701184273,
|
| 1574 |
+
"learning_rate": 3.53887399463807e-06,
|
| 1575 |
+
"loss": 0.007024706155061722,
|
| 1576 |
+
"step": 2200
|
| 1577 |
+
},
|
| 1578 |
+
{
|
| 1579 |
+
"epoch": 2.491544532130778,
|
| 1580 |
+
"grad_norm": 0.0024169229436665773,
|
| 1581 |
+
"learning_rate": 3.4622749904251246e-06,
|
| 1582 |
+
"loss": 0.00021023082081228495,
|
| 1583 |
+
"step": 2210
|
| 1584 |
+
},
|
| 1585 |
+
{
|
| 1586 |
+
"epoch": 2.5028184892897407,
|
| 1587 |
+
"grad_norm": 0.0027928894851356745,
|
| 1588 |
+
"learning_rate": 3.3856759862121796e-06,
|
| 1589 |
+
"loss": 0.00023648575879633426,
|
| 1590 |
+
"step": 2220
|
| 1591 |
+
},
|
| 1592 |
+
{
|
| 1593 |
+
"epoch": 2.5140924464487036,
|
| 1594 |
+
"grad_norm": 0.0053055062890052795,
|
| 1595 |
+
"learning_rate": 3.3090769819992342e-06,
|
| 1596 |
+
"loss": 0.00020812514703720807,
|
| 1597 |
+
"step": 2230
|
| 1598 |
+
},
|
| 1599 |
+
{
|
| 1600 |
+
"epoch": 2.5253664036076664,
|
| 1601 |
+
"grad_norm": 0.0040803938172757626,
|
| 1602 |
+
"learning_rate": 3.2324779777862893e-06,
|
| 1603 |
+
"loss": 0.03706019520759583,
|
| 1604 |
+
"step": 2240
|
| 1605 |
+
},
|
| 1606 |
+
{
|
| 1607 |
+
"epoch": 2.5366403607666292,
|
| 1608 |
+
"grad_norm": 0.004515635780990124,
|
| 1609 |
+
"learning_rate": 3.1558789735733435e-06,
|
| 1610 |
+
"loss": 0.00022339699789881707,
|
| 1611 |
+
"step": 2250
|
| 1612 |
+
},
|
| 1613 |
+
{
|
| 1614 |
+
"epoch": 2.547914317925592,
|
| 1615 |
+
"grad_norm": 0.0072349319234490395,
|
| 1616 |
+
"learning_rate": 3.0792799693603985e-06,
|
| 1617 |
+
"loss": 0.0001945420168340206,
|
| 1618 |
+
"step": 2260
|
| 1619 |
+
},
|
| 1620 |
+
{
|
| 1621 |
+
"epoch": 2.559188275084555,
|
| 1622 |
+
"grad_norm": 0.0061413892544806,
|
| 1623 |
+
"learning_rate": 3.002680965147453e-06,
|
| 1624 |
+
"loss": 0.00021348206792026758,
|
| 1625 |
+
"step": 2270
|
| 1626 |
+
},
|
| 1627 |
+
{
|
| 1628 |
+
"epoch": 2.5704622322435173,
|
| 1629 |
+
"grad_norm": 0.0026650819927453995,
|
| 1630 |
+
"learning_rate": 2.926081960934508e-06,
|
| 1631 |
+
"loss": 0.00020323649514466524,
|
| 1632 |
+
"step": 2280
|
| 1633 |
+
},
|
| 1634 |
+
{
|
| 1635 |
+
"epoch": 2.58173618940248,
|
| 1636 |
+
"grad_norm": 0.006988232955336571,
|
| 1637 |
+
"learning_rate": 2.849482956721563e-06,
|
| 1638 |
+
"loss": 0.00023265560157597064,
|
| 1639 |
+
"step": 2290
|
| 1640 |
+
},
|
| 1641 |
+
{
|
| 1642 |
+
"epoch": 2.593010146561443,
|
| 1643 |
+
"grad_norm": 0.001989186042919755,
|
| 1644 |
+
"learning_rate": 2.772883952508618e-06,
|
| 1645 |
+
"loss": 0.00020218645222485064,
|
| 1646 |
+
"step": 2300
|
| 1647 |
+
},
|
| 1648 |
+
{
|
| 1649 |
+
"epoch": 2.604284103720406,
|
| 1650 |
+
"grad_norm": 0.7532305121421814,
|
| 1651 |
+
"learning_rate": 2.696284948295672e-06,
|
| 1652 |
+
"loss": 0.0005159636959433556,
|
| 1653 |
+
"step": 2310
|
| 1654 |
+
},
|
| 1655 |
+
{
|
| 1656 |
+
"epoch": 2.6155580608793687,
|
| 1657 |
+
"grad_norm": 0.004034025128930807,
|
| 1658 |
+
"learning_rate": 2.619685944082727e-06,
|
| 1659 |
+
"loss": 0.0002368612913414836,
|
| 1660 |
+
"step": 2320
|
| 1661 |
+
},
|
| 1662 |
+
{
|
| 1663 |
+
"epoch": 2.6268320180383316,
|
| 1664 |
+
"grad_norm": 0.022439252585172653,
|
| 1665 |
+
"learning_rate": 2.543086939869782e-06,
|
| 1666 |
+
"loss": 0.0002526436699554324,
|
| 1667 |
+
"step": 2330
|
| 1668 |
+
},
|
| 1669 |
+
{
|
| 1670 |
+
"epoch": 2.6381059751972944,
|
| 1671 |
+
"grad_norm": 0.0024465364404022694,
|
| 1672 |
+
"learning_rate": 2.4664879356568368e-06,
|
| 1673 |
+
"loss": 0.10790177583694457,
|
| 1674 |
+
"step": 2340
|
| 1675 |
+
},
|
| 1676 |
+
{
|
| 1677 |
+
"epoch": 2.649379932356257,
|
| 1678 |
+
"grad_norm": 0.0027746243868023157,
|
| 1679 |
+
"learning_rate": 2.3898889314438914e-06,
|
| 1680 |
+
"loss": 0.00025054800789803267,
|
| 1681 |
+
"step": 2350
|
| 1682 |
+
},
|
| 1683 |
+
{
|
| 1684 |
+
"epoch": 2.6606538895152196,
|
| 1685 |
+
"grad_norm": 0.0029607873875647783,
|
| 1686 |
+
"learning_rate": 2.313289927230946e-06,
|
| 1687 |
+
"loss": 0.00021450400818139314,
|
| 1688 |
+
"step": 2360
|
| 1689 |
+
},
|
| 1690 |
+
{
|
| 1691 |
+
"epoch": 2.6719278466741825,
|
| 1692 |
+
"grad_norm": 0.014295640401542187,
|
| 1693 |
+
"learning_rate": 2.236690923018001e-06,
|
| 1694 |
+
"loss": 0.0002021346241235733,
|
| 1695 |
+
"step": 2370
|
| 1696 |
+
},
|
| 1697 |
+
{
|
| 1698 |
+
"epoch": 2.6832018038331453,
|
| 1699 |
+
"grad_norm": 0.00496269715949893,
|
| 1700 |
+
"learning_rate": 2.1600919188050557e-06,
|
| 1701 |
+
"loss": 0.00023605772294104098,
|
| 1702 |
+
"step": 2380
|
| 1703 |
+
},
|
| 1704 |
+
{
|
| 1705 |
+
"epoch": 2.694475760992108,
|
| 1706 |
+
"grad_norm": 0.002372585004195571,
|
| 1707 |
+
"learning_rate": 2.0834929145921103e-06,
|
| 1708 |
+
"loss": 0.0013206375762820244,
|
| 1709 |
+
"step": 2390
|
| 1710 |
+
},
|
| 1711 |
+
{
|
| 1712 |
+
"epoch": 2.705749718151071,
|
| 1713 |
+
"grad_norm": 0.0025791102088987827,
|
| 1714 |
+
"learning_rate": 2.0068939103791653e-06,
|
| 1715 |
+
"loss": 0.00020619079004973172,
|
| 1716 |
+
"step": 2400
|
| 1717 |
+
},
|
| 1718 |
+
{
|
| 1719 |
+
"epoch": 2.717023675310034,
|
| 1720 |
+
"grad_norm": 0.0025324171874672174,
|
| 1721 |
+
"learning_rate": 1.93029490616622e-06,
|
| 1722 |
+
"loss": 0.00019635353237390518,
|
| 1723 |
+
"step": 2410
|
| 1724 |
+
},
|
| 1725 |
+
{
|
| 1726 |
+
"epoch": 2.7282976324689967,
|
| 1727 |
+
"grad_norm": 0.00920526497066021,
|
| 1728 |
+
"learning_rate": 1.8536959019532746e-06,
|
| 1729 |
+
"loss": 0.00018937309505417943,
|
| 1730 |
+
"step": 2420
|
| 1731 |
+
},
|
| 1732 |
+
{
|
| 1733 |
+
"epoch": 2.7395715896279595,
|
| 1734 |
+
"grad_norm": 0.002414870075881481,
|
| 1735 |
+
"learning_rate": 1.7770968977403294e-06,
|
| 1736 |
+
"loss": 0.0001852948684245348,
|
| 1737 |
+
"step": 2430
|
| 1738 |
+
},
|
| 1739 |
+
{
|
| 1740 |
+
"epoch": 2.7508455467869224,
|
| 1741 |
+
"grad_norm": 0.0018624997464939952,
|
| 1742 |
+
"learning_rate": 1.700497893527384e-06,
|
| 1743 |
+
"loss": 0.0006128895562142134,
|
| 1744 |
+
"step": 2440
|
| 1745 |
+
},
|
| 1746 |
+
{
|
| 1747 |
+
"epoch": 2.7621195039458852,
|
| 1748 |
+
"grad_norm": 0.0024695040192455053,
|
| 1749 |
+
"learning_rate": 1.623898889314439e-06,
|
| 1750 |
+
"loss": 0.00018132688710466028,
|
| 1751 |
+
"step": 2450
|
| 1752 |
+
},
|
| 1753 |
+
{
|
| 1754 |
+
"epoch": 2.773393461104848,
|
| 1755 |
+
"grad_norm": 0.0031408024951815605,
|
| 1756 |
+
"learning_rate": 1.547299885101494e-06,
|
| 1757 |
+
"loss": 0.0001800063531845808,
|
| 1758 |
+
"step": 2460
|
| 1759 |
+
},
|
| 1760 |
+
{
|
| 1761 |
+
"epoch": 2.7846674182638105,
|
| 1762 |
+
"grad_norm": 0.0027796358335763216,
|
| 1763 |
+
"learning_rate": 1.4707008808885488e-06,
|
| 1764 |
+
"loss": 0.00019140702206641436,
|
| 1765 |
+
"step": 2470
|
| 1766 |
+
},
|
| 1767 |
+
{
|
| 1768 |
+
"epoch": 2.7959413754227733,
|
| 1769 |
+
"grad_norm": 0.0023478721268475056,
|
| 1770 |
+
"learning_rate": 1.3941018766756034e-06,
|
| 1771 |
+
"loss": 0.00021520003210753202,
|
| 1772 |
+
"step": 2480
|
| 1773 |
+
},
|
| 1774 |
+
{
|
| 1775 |
+
"epoch": 2.807215332581736,
|
| 1776 |
+
"grad_norm": 0.011640374548733234,
|
| 1777 |
+
"learning_rate": 1.3175028724626582e-06,
|
| 1778 |
+
"loss": 0.0001937905792146921,
|
| 1779 |
+
"step": 2490
|
| 1780 |
+
},
|
| 1781 |
+
{
|
| 1782 |
+
"epoch": 2.818489289740699,
|
| 1783 |
+
"grad_norm": 0.0029908877331763506,
|
| 1784 |
+
"learning_rate": 1.2409038682497128e-06,
|
| 1785 |
+
"loss": 0.00016609376762062312,
|
| 1786 |
+
"step": 2500
|
| 1787 |
+
},
|
| 1788 |
+
{
|
| 1789 |
+
"epoch": 2.829763246899662,
|
| 1790 |
+
"grad_norm": 0.03514697775244713,
|
| 1791 |
+
"learning_rate": 1.1643048640367677e-06,
|
| 1792 |
+
"loss": 0.0536762535572052,
|
| 1793 |
+
"step": 2510
|
| 1794 |
+
},
|
| 1795 |
+
{
|
| 1796 |
+
"epoch": 2.8410372040586247,
|
| 1797 |
+
"grad_norm": 0.0031698185484856367,
|
| 1798 |
+
"learning_rate": 1.0877058598238225e-06,
|
| 1799 |
+
"loss": 0.00016408661613240839,
|
| 1800 |
+
"step": 2520
|
| 1801 |
+
},
|
| 1802 |
+
{
|
| 1803 |
+
"epoch": 2.852311161217587,
|
| 1804 |
+
"grad_norm": 0.0022064542863518,
|
| 1805 |
+
"learning_rate": 1.0111068556108771e-06,
|
| 1806 |
+
"loss": 0.0001742431544698775,
|
| 1807 |
+
"step": 2530
|
| 1808 |
+
},
|
| 1809 |
+
{
|
| 1810 |
+
"epoch": 2.86358511837655,
|
| 1811 |
+
"grad_norm": 0.0024360711686313152,
|
| 1812 |
+
"learning_rate": 9.34507851397932e-07,
|
| 1813 |
+
"loss": 0.00018918949645012616,
|
| 1814 |
+
"step": 2540
|
| 1815 |
+
},
|
| 1816 |
+
{
|
| 1817 |
+
"epoch": 2.874859075535513,
|
| 1818 |
+
"grad_norm": 0.0021841819398105145,
|
| 1819 |
+
"learning_rate": 8.579088471849867e-07,
|
| 1820 |
+
"loss": 0.00493232011795044,
|
| 1821 |
+
"step": 2550
|
| 1822 |
+
},
|
| 1823 |
+
{
|
| 1824 |
+
"epoch": 2.8861330326944756,
|
| 1825 |
+
"grad_norm": 0.0036376486532390118,
|
| 1826 |
+
"learning_rate": 7.813098429720414e-07,
|
| 1827 |
+
"loss": 0.00020034858025610446,
|
| 1828 |
+
"step": 2560
|
| 1829 |
+
},
|
| 1830 |
+
{
|
| 1831 |
+
"epoch": 2.8974069898534385,
|
| 1832 |
+
"grad_norm": 0.0025231263134628534,
|
| 1833 |
+
"learning_rate": 7.047108387590961e-07,
|
| 1834 |
+
"loss": 0.00018351208418607712,
|
| 1835 |
+
"step": 2570
|
| 1836 |
+
},
|
| 1837 |
+
{
|
| 1838 |
+
"epoch": 2.9086809470124013,
|
| 1839 |
+
"grad_norm": 0.0025046563241630793,
|
| 1840 |
+
"learning_rate": 6.28111834546151e-07,
|
| 1841 |
+
"loss": 0.00019472729181870818,
|
| 1842 |
+
"step": 2580
|
| 1843 |
+
},
|
| 1844 |
+
{
|
| 1845 |
+
"epoch": 2.919954904171364,
|
| 1846 |
+
"grad_norm": 0.002354441909119487,
|
| 1847 |
+
"learning_rate": 5.515128303332057e-07,
|
| 1848 |
+
"loss": 0.00017740146722644567,
|
| 1849 |
+
"step": 2590
|
| 1850 |
+
},
|
| 1851 |
+
{
|
| 1852 |
+
"epoch": 2.931228861330327,
|
| 1853 |
+
"grad_norm": 0.009609458968043327,
|
| 1854 |
+
"learning_rate": 4.749138261202605e-07,
|
| 1855 |
+
"loss": 0.00019160339143127204,
|
| 1856 |
+
"step": 2600
|
| 1857 |
+
},
|
| 1858 |
+
{
|
| 1859 |
+
"epoch": 2.94250281848929,
|
| 1860 |
+
"grad_norm": 0.0036235046572983265,
|
| 1861 |
+
"learning_rate": 3.9831482190731527e-07,
|
| 1862 |
+
"loss": 0.00022586036939173937,
|
| 1863 |
+
"step": 2610
|
| 1864 |
+
},
|
| 1865 |
+
{
|
| 1866 |
+
"epoch": 2.9537767756482527,
|
| 1867 |
+
"grad_norm": 0.0024371265899389982,
|
| 1868 |
+
"learning_rate": 3.2171581769437e-07,
|
| 1869 |
+
"loss": 0.0001714262878522277,
|
| 1870 |
+
"step": 2620
|
| 1871 |
+
},
|
| 1872 |
+
{
|
| 1873 |
+
"epoch": 2.9650507328072155,
|
| 1874 |
+
"grad_norm": 0.0023558461107313633,
|
| 1875 |
+
"learning_rate": 2.451168134814248e-07,
|
| 1876 |
+
"loss": 0.00018437257967889308,
|
| 1877 |
+
"step": 2630
|
| 1878 |
+
},
|
| 1879 |
+
{
|
| 1880 |
+
"epoch": 2.9763246899661784,
|
| 1881 |
+
"grad_norm": 0.02326689288020134,
|
| 1882 |
+
"learning_rate": 1.685178092684795e-07,
|
| 1883 |
+
"loss": 0.00022208874579519034,
|
| 1884 |
+
"step": 2640
|
| 1885 |
+
},
|
| 1886 |
+
{
|
| 1887 |
+
"epoch": 2.987598647125141,
|
| 1888 |
+
"grad_norm": 0.013761946000158787,
|
| 1889 |
+
"learning_rate": 9.19188050555343e-08,
|
| 1890 |
+
"loss": 0.0002170191379263997,
|
| 1891 |
+
"step": 2650
|
| 1892 |
+
},
|
| 1893 |
+
{
|
| 1894 |
+
"epoch": 2.9988726042841036,
|
| 1895 |
+
"grad_norm": 0.002142044249922037,
|
| 1896 |
+
"learning_rate": 1.531980084258905e-08,
|
| 1897 |
+
"loss": 0.0002309123519808054,
|
| 1898 |
+
"step": 2660
|
| 1899 |
+
},
|
| 1900 |
+
{
|
| 1901 |
+
"epoch": 3.0,
|
| 1902 |
+
"eval_avg_accuracy": 0.9991537376586741,
|
| 1903 |
+
"eval_explanation_accuracy": 0.9988716502115655,
|
| 1904 |
+
"eval_explanation_f1": 0.9977851605758582,
|
| 1905 |
+
"eval_loss": 0.010961832478642464,
|
| 1906 |
+
"eval_refusal_accuracy": 0.9994358251057828,
|
| 1907 |
+
"eval_refusal_f1": 0.9994363021420518,
|
| 1908 |
+
"eval_runtime": 8.2108,
|
| 1909 |
+
"eval_samples_per_second": 431.749,
|
| 1910 |
+
"eval_steps_per_second": 27.038,
|
| 1911 |
+
"step": 2661
|
| 1912 |
+
}
|
| 1913 |
+
],
|
| 1914 |
+
"logging_steps": 10,
|
| 1915 |
+
"max_steps": 2661,
|
| 1916 |
+
"num_input_tokens_seen": 0,
|
| 1917 |
+
"num_train_epochs": 3,
|
| 1918 |
+
"save_steps": 500,
|
| 1919 |
+
"stateful_callbacks": {
|
| 1920 |
+
"TrainerControl": {
|
| 1921 |
+
"args": {
|
| 1922 |
+
"should_epoch_stop": false,
|
| 1923 |
+
"should_evaluate": false,
|
| 1924 |
+
"should_log": false,
|
| 1925 |
+
"should_save": true,
|
| 1926 |
+
"should_training_stop": true
|
| 1927 |
+
},
|
| 1928 |
+
"attributes": {}
|
| 1929 |
+
}
|
| 1930 |
+
},
|
| 1931 |
+
"total_flos": 0.0,
|
| 1932 |
+
"train_batch_size": 16,
|
| 1933 |
+
"trial_name": null,
|
| 1934 |
+
"trial_params": null
|
| 1935 |
+
}
|
checkpoints/checkpoint-2661/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:78b3ded581335bef923598436ee3c2f3e4bb84d0d2d55e710db56437d81078ce
|
| 3 |
+
size 5201
|
classifier_heads.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 14839
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:c5d20caf7a45b5855a0567e334517a0fa09e00b733d4e499b78586b6c1c7f467
|
| 3 |
size 14839
|
tokenizer/tokenizer_config.json
CHANGED
|
@@ -1,56 +1,14 @@
|
|
| 1 |
{
|
| 2 |
-
"
|
| 3 |
-
"0": {
|
| 4 |
-
"content": "[PAD]",
|
| 5 |
-
"lstrip": false,
|
| 6 |
-
"normalized": false,
|
| 7 |
-
"rstrip": false,
|
| 8 |
-
"single_word": false,
|
| 9 |
-
"special": true
|
| 10 |
-
},
|
| 11 |
-
"100": {
|
| 12 |
-
"content": "[UNK]",
|
| 13 |
-
"lstrip": false,
|
| 14 |
-
"normalized": false,
|
| 15 |
-
"rstrip": false,
|
| 16 |
-
"single_word": false,
|
| 17 |
-
"special": true
|
| 18 |
-
},
|
| 19 |
-
"101": {
|
| 20 |
-
"content": "[CLS]",
|
| 21 |
-
"lstrip": false,
|
| 22 |
-
"normalized": false,
|
| 23 |
-
"rstrip": false,
|
| 24 |
-
"single_word": false,
|
| 25 |
-
"special": true
|
| 26 |
-
},
|
| 27 |
-
"102": {
|
| 28 |
-
"content": "[SEP]",
|
| 29 |
-
"lstrip": false,
|
| 30 |
-
"normalized": false,
|
| 31 |
-
"rstrip": false,
|
| 32 |
-
"single_word": false,
|
| 33 |
-
"special": true
|
| 34 |
-
},
|
| 35 |
-
"103": {
|
| 36 |
-
"content": "[MASK]",
|
| 37 |
-
"lstrip": false,
|
| 38 |
-
"normalized": false,
|
| 39 |
-
"rstrip": false,
|
| 40 |
-
"single_word": false,
|
| 41 |
-
"special": true
|
| 42 |
-
}
|
| 43 |
-
},
|
| 44 |
-
"clean_up_tokenization_spaces": false,
|
| 45 |
"cls_token": "[CLS]",
|
| 46 |
"do_lower_case": true,
|
| 47 |
-
"
|
| 48 |
"mask_token": "[MASK]",
|
| 49 |
"model_max_length": 512,
|
| 50 |
"pad_token": "[PAD]",
|
| 51 |
"sep_token": "[SEP]",
|
| 52 |
"strip_accents": null,
|
| 53 |
"tokenize_chinese_chars": true,
|
| 54 |
-
"tokenizer_class": "
|
| 55 |
"unk_token": "[UNK]"
|
| 56 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"backend": "tokenizers",
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 3 |
"cls_token": "[CLS]",
|
| 4 |
"do_lower_case": true,
|
| 5 |
+
"is_local": false,
|
| 6 |
"mask_token": "[MASK]",
|
| 7 |
"model_max_length": 512,
|
| 8 |
"pad_token": "[PAD]",
|
| 9 |
"sep_token": "[SEP]",
|
| 10 |
"strip_accents": null,
|
| 11 |
"tokenize_chinese_chars": true,
|
| 12 |
+
"tokenizer_class": "BertTokenizer",
|
| 13 |
"unk_token": "[UNK]"
|
| 14 |
}
|