Upload rerank_model from local
Browse files- config.json +1 -1
- optimizer.pt +3 -0
- rng_state.pth +3 -0
- scheduler.pt +3 -0
- special_tokens_map.json +1 -1
- tokenizer_config.json +1 -1
- trainer_state.json +1252 -0
- training_args.bin +3 -0
config.json
CHANGED
|
@@ -48,4 +48,4 @@
|
|
| 48 |
"unpad_inputs": false,
|
| 49 |
"use_memory_efficient_attention": false,
|
| 50 |
"vocab_size": 250048
|
| 51 |
-
}
|
|
|
|
| 48 |
"unpad_inputs": false,
|
| 49 |
"use_memory_efficient_attention": false,
|
| 50 |
"vocab_size": 250048
|
| 51 |
+
}
|
optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:4a32ba179658a4500aef14adfde824f9fd1209b08f12c4879c82c1c1e424a677
|
| 3 |
+
size 2447795147
|
rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:153e458bd75fb3848db9b6f77fd3dbc202e8b9991637e5679559811445c889a7
|
| 3 |
+
size 14645
|
scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:d860214655dfba7a7d4a60019458f6430c807bb94c834bbbef71b2ce0819480e
|
| 3 |
+
size 1465
|
special_tokens_map.json
CHANGED
|
@@ -48,4 +48,4 @@
|
|
| 48 |
"rstrip": false,
|
| 49 |
"single_word": false
|
| 50 |
}
|
| 51 |
-
}
|
|
|
|
| 48 |
"rstrip": false,
|
| 49 |
"single_word": false
|
| 50 |
}
|
| 51 |
+
}
|
tokenizer_config.json
CHANGED
|
@@ -58,4 +58,4 @@
|
|
| 58 |
"truncation_side": "right",
|
| 59 |
"truncation_strategy": "longest_first",
|
| 60 |
"unk_token": "<unk>"
|
| 61 |
-
}
|
|
|
|
| 58 |
"truncation_side": "right",
|
| 59 |
"truncation_strategy": "longest_first",
|
| 60 |
"unk_token": "<unk>"
|
| 61 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,1252 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 0.0018289362778887153,
|
| 3 |
+
"best_model_checkpoint": "./results/checkpoint-1669",
|
| 4 |
+
"epoch": 3.998502994011976,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 1669,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.023952095808383235,
|
| 13 |
+
"grad_norm": 21869930.0,
|
| 14 |
+
"learning_rate": 2.3980815347721825e-07,
|
| 15 |
+
"loss": 0.9053,
|
| 16 |
+
"step": 10
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.04790419161676647,
|
| 20 |
+
"grad_norm": 19059822.0,
|
| 21 |
+
"learning_rate": 4.796163069544365e-07,
|
| 22 |
+
"loss": 0.7081,
|
| 23 |
+
"step": 20
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.0718562874251497,
|
| 27 |
+
"grad_norm": 11655885.0,
|
| 28 |
+
"learning_rate": 7.194244604316547e-07,
|
| 29 |
+
"loss": 0.5323,
|
| 30 |
+
"step": 30
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.09580838323353294,
|
| 34 |
+
"grad_norm": 3832326.5,
|
| 35 |
+
"learning_rate": 9.59232613908873e-07,
|
| 36 |
+
"loss": 0.3305,
|
| 37 |
+
"step": 40
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.11976047904191617,
|
| 41 |
+
"grad_norm": 2399766.25,
|
| 42 |
+
"learning_rate": 1.1990407673860912e-06,
|
| 43 |
+
"loss": 0.178,
|
| 44 |
+
"step": 50
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.1437125748502994,
|
| 48 |
+
"grad_norm": 707622.9375,
|
| 49 |
+
"learning_rate": 1.4388489208633094e-06,
|
| 50 |
+
"loss": 0.1208,
|
| 51 |
+
"step": 60
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.16766467065868262,
|
| 55 |
+
"grad_norm": 694203.75,
|
| 56 |
+
"learning_rate": 1.6786570743405278e-06,
|
| 57 |
+
"loss": 0.1082,
|
| 58 |
+
"step": 70
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.19161676646706588,
|
| 62 |
+
"grad_norm": 906404.875,
|
| 63 |
+
"learning_rate": 1.918465227817746e-06,
|
| 64 |
+
"loss": 0.085,
|
| 65 |
+
"step": 80
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.2155688622754491,
|
| 69 |
+
"grad_norm": 430721.75,
|
| 70 |
+
"learning_rate": 2.158273381294964e-06,
|
| 71 |
+
"loss": 0.0717,
|
| 72 |
+
"step": 90
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.23952095808383234,
|
| 76 |
+
"grad_norm": 347536.15625,
|
| 77 |
+
"learning_rate": 2.3980815347721824e-06,
|
| 78 |
+
"loss": 0.053,
|
| 79 |
+
"step": 100
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.2634730538922156,
|
| 83 |
+
"grad_norm": 364138.15625,
|
| 84 |
+
"learning_rate": 2.637889688249401e-06,
|
| 85 |
+
"loss": 0.045,
|
| 86 |
+
"step": 110
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.2874251497005988,
|
| 90 |
+
"grad_norm": 305517.90625,
|
| 91 |
+
"learning_rate": 2.877697841726619e-06,
|
| 92 |
+
"loss": 0.0366,
|
| 93 |
+
"step": 120
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.31137724550898205,
|
| 97 |
+
"grad_norm": 286330.90625,
|
| 98 |
+
"learning_rate": 3.1175059952038373e-06,
|
| 99 |
+
"loss": 0.0318,
|
| 100 |
+
"step": 130
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 0.33532934131736525,
|
| 104 |
+
"grad_norm": 221961.6875,
|
| 105 |
+
"learning_rate": 3.3573141486810557e-06,
|
| 106 |
+
"loss": 0.0262,
|
| 107 |
+
"step": 140
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.3592814371257485,
|
| 111 |
+
"grad_norm": 307385.9375,
|
| 112 |
+
"learning_rate": 3.5971223021582737e-06,
|
| 113 |
+
"loss": 0.0229,
|
| 114 |
+
"step": 150
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.38323353293413176,
|
| 118 |
+
"grad_norm": 362847.0,
|
| 119 |
+
"learning_rate": 3.836930455635492e-06,
|
| 120 |
+
"loss": 0.0227,
|
| 121 |
+
"step": 160
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.40718562874251496,
|
| 125 |
+
"grad_norm": 225246.109375,
|
| 126 |
+
"learning_rate": 4.0767386091127105e-06,
|
| 127 |
+
"loss": 0.0211,
|
| 128 |
+
"step": 170
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.4311377245508982,
|
| 132 |
+
"grad_norm": 336017.8125,
|
| 133 |
+
"learning_rate": 4.316546762589928e-06,
|
| 134 |
+
"loss": 0.0188,
|
| 135 |
+
"step": 180
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.4550898203592814,
|
| 139 |
+
"grad_norm": 232213.890625,
|
| 140 |
+
"learning_rate": 4.5563549160671465e-06,
|
| 141 |
+
"loss": 0.0134,
|
| 142 |
+
"step": 190
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.47904191616766467,
|
| 146 |
+
"grad_norm": 189961.46875,
|
| 147 |
+
"learning_rate": 4.796163069544365e-06,
|
| 148 |
+
"loss": 0.016,
|
| 149 |
+
"step": 200
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.5029940119760479,
|
| 153 |
+
"grad_norm": 108848.3671875,
|
| 154 |
+
"learning_rate": 5.035971223021583e-06,
|
| 155 |
+
"loss": 0.0141,
|
| 156 |
+
"step": 210
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 0.5269461077844312,
|
| 160 |
+
"grad_norm": 158108.1875,
|
| 161 |
+
"learning_rate": 5.275779376498802e-06,
|
| 162 |
+
"loss": 0.0131,
|
| 163 |
+
"step": 220
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 0.5508982035928144,
|
| 167 |
+
"grad_norm": 257074.015625,
|
| 168 |
+
"learning_rate": 5.51558752997602e-06,
|
| 169 |
+
"loss": 0.0098,
|
| 170 |
+
"step": 230
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.5748502994011976,
|
| 174 |
+
"grad_norm": 181764.84375,
|
| 175 |
+
"learning_rate": 5.755395683453238e-06,
|
| 176 |
+
"loss": 0.011,
|
| 177 |
+
"step": 240
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.5988023952095808,
|
| 181 |
+
"grad_norm": 231444.921875,
|
| 182 |
+
"learning_rate": 5.995203836930457e-06,
|
| 183 |
+
"loss": 0.0135,
|
| 184 |
+
"step": 250
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 0.6227544910179641,
|
| 188 |
+
"grad_norm": 267698.96875,
|
| 189 |
+
"learning_rate": 6.2350119904076745e-06,
|
| 190 |
+
"loss": 0.0099,
|
| 191 |
+
"step": 260
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 0.6467065868263473,
|
| 195 |
+
"grad_norm": 168324.34375,
|
| 196 |
+
"learning_rate": 6.474820143884892e-06,
|
| 197 |
+
"loss": 0.0107,
|
| 198 |
+
"step": 270
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 0.6706586826347305,
|
| 202 |
+
"grad_norm": 174906.8125,
|
| 203 |
+
"learning_rate": 6.714628297362111e-06,
|
| 204 |
+
"loss": 0.013,
|
| 205 |
+
"step": 280
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 0.6946107784431138,
|
| 209 |
+
"grad_norm": 229340.8125,
|
| 210 |
+
"learning_rate": 6.954436450839329e-06,
|
| 211 |
+
"loss": 0.0125,
|
| 212 |
+
"step": 290
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.718562874251497,
|
| 216 |
+
"grad_norm": 141271.78125,
|
| 217 |
+
"learning_rate": 7.194244604316547e-06,
|
| 218 |
+
"loss": 0.0093,
|
| 219 |
+
"step": 300
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.7425149700598802,
|
| 223 |
+
"grad_norm": 65792.4296875,
|
| 224 |
+
"learning_rate": 7.434052757793766e-06,
|
| 225 |
+
"loss": 0.0107,
|
| 226 |
+
"step": 310
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 0.7664670658682635,
|
| 230 |
+
"grad_norm": 155497.109375,
|
| 231 |
+
"learning_rate": 7.673860911270984e-06,
|
| 232 |
+
"loss": 0.0078,
|
| 233 |
+
"step": 320
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 0.7904191616766467,
|
| 237 |
+
"grad_norm": 360102.5625,
|
| 238 |
+
"learning_rate": 7.913669064748202e-06,
|
| 239 |
+
"loss": 0.0077,
|
| 240 |
+
"step": 330
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 0.8143712574850299,
|
| 244 |
+
"grad_norm": 261093.21875,
|
| 245 |
+
"learning_rate": 8.153477218225421e-06,
|
| 246 |
+
"loss": 0.0061,
|
| 247 |
+
"step": 340
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 0.8383233532934131,
|
| 251 |
+
"grad_norm": 178080.28125,
|
| 252 |
+
"learning_rate": 8.393285371702639e-06,
|
| 253 |
+
"loss": 0.0082,
|
| 254 |
+
"step": 350
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 0.8622754491017964,
|
| 258 |
+
"grad_norm": 269176.5625,
|
| 259 |
+
"learning_rate": 8.633093525179856e-06,
|
| 260 |
+
"loss": 0.007,
|
| 261 |
+
"step": 360
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 0.8862275449101796,
|
| 265 |
+
"grad_norm": 409776.6875,
|
| 266 |
+
"learning_rate": 8.872901678657075e-06,
|
| 267 |
+
"loss": 0.009,
|
| 268 |
+
"step": 370
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 0.9101796407185628,
|
| 272 |
+
"grad_norm": 653130.125,
|
| 273 |
+
"learning_rate": 9.112709832134293e-06,
|
| 274 |
+
"loss": 0.0065,
|
| 275 |
+
"step": 380
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 0.9341317365269461,
|
| 279 |
+
"grad_norm": 85098.8671875,
|
| 280 |
+
"learning_rate": 9.35251798561151e-06,
|
| 281 |
+
"loss": 0.0036,
|
| 282 |
+
"step": 390
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 0.9580838323353293,
|
| 286 |
+
"grad_norm": 427729.28125,
|
| 287 |
+
"learning_rate": 9.59232613908873e-06,
|
| 288 |
+
"loss": 0.0059,
|
| 289 |
+
"step": 400
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 0.9820359281437125,
|
| 293 |
+
"grad_norm": 151977.40625,
|
| 294 |
+
"learning_rate": 9.832134292565947e-06,
|
| 295 |
+
"loss": 0.0074,
|
| 296 |
+
"step": 410
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 0.9988023952095808,
|
| 300 |
+
"eval_accuracy": 0.8697995620683847,
|
| 301 |
+
"eval_f1": 0.8873013558827817,
|
| 302 |
+
"eval_loss": 0.009821193292737007,
|
| 303 |
+
"eval_precision": 0.7974318658280922,
|
| 304 |
+
"eval_recall": 1.0,
|
| 305 |
+
"eval_runtime": 156.9292,
|
| 306 |
+
"eval_samples_per_second": 37.832,
|
| 307 |
+
"eval_steps_per_second": 2.37,
|
| 308 |
+
"step": 417
|
| 309 |
+
},
|
| 310 |
+
{
|
| 311 |
+
"epoch": 1.0062874251497007,
|
| 312 |
+
"grad_norm": 229342.40625,
|
| 313 |
+
"learning_rate": 1.0071942446043167e-05,
|
| 314 |
+
"loss": 0.0093,
|
| 315 |
+
"step": 420
|
| 316 |
+
},
|
| 317 |
+
{
|
| 318 |
+
"epoch": 1.030239520958084,
|
| 319 |
+
"grad_norm": 105467.90625,
|
| 320 |
+
"learning_rate": 1.0311750599520384e-05,
|
| 321 |
+
"loss": 0.0042,
|
| 322 |
+
"step": 430
|
| 323 |
+
},
|
| 324 |
+
{
|
| 325 |
+
"epoch": 1.054191616766467,
|
| 326 |
+
"grad_norm": 71231.6171875,
|
| 327 |
+
"learning_rate": 1.0551558752997603e-05,
|
| 328 |
+
"loss": 0.0043,
|
| 329 |
+
"step": 440
|
| 330 |
+
},
|
| 331 |
+
{
|
| 332 |
+
"epoch": 1.0781437125748503,
|
| 333 |
+
"grad_norm": 147894.453125,
|
| 334 |
+
"learning_rate": 1.0791366906474821e-05,
|
| 335 |
+
"loss": 0.0037,
|
| 336 |
+
"step": 450
|
| 337 |
+
},
|
| 338 |
+
{
|
| 339 |
+
"epoch": 1.1020958083832335,
|
| 340 |
+
"grad_norm": 173828.546875,
|
| 341 |
+
"learning_rate": 1.103117505995204e-05,
|
| 342 |
+
"loss": 0.0044,
|
| 343 |
+
"step": 460
|
| 344 |
+
},
|
| 345 |
+
{
|
| 346 |
+
"epoch": 1.1260479041916167,
|
| 347 |
+
"grad_norm": 158015.078125,
|
| 348 |
+
"learning_rate": 1.1270983213429258e-05,
|
| 349 |
+
"loss": 0.0039,
|
| 350 |
+
"step": 470
|
| 351 |
+
},
|
| 352 |
+
{
|
| 353 |
+
"epoch": 1.15,
|
| 354 |
+
"grad_norm": 258871.4375,
|
| 355 |
+
"learning_rate": 1.1510791366906475e-05,
|
| 356 |
+
"loss": 0.0034,
|
| 357 |
+
"step": 480
|
| 358 |
+
},
|
| 359 |
+
{
|
| 360 |
+
"epoch": 1.1739520958083833,
|
| 361 |
+
"grad_norm": 182111.390625,
|
| 362 |
+
"learning_rate": 1.1750599520383695e-05,
|
| 363 |
+
"loss": 0.0066,
|
| 364 |
+
"step": 490
|
| 365 |
+
},
|
| 366 |
+
{
|
| 367 |
+
"epoch": 1.1979041916167665,
|
| 368 |
+
"grad_norm": 68393.5703125,
|
| 369 |
+
"learning_rate": 1.1990407673860914e-05,
|
| 370 |
+
"loss": 0.0031,
|
| 371 |
+
"step": 500
|
| 372 |
+
},
|
| 373 |
+
{
|
| 374 |
+
"epoch": 1.2218562874251497,
|
| 375 |
+
"grad_norm": 304475.46875,
|
| 376 |
+
"learning_rate": 1.223021582733813e-05,
|
| 377 |
+
"loss": 0.0026,
|
| 378 |
+
"step": 510
|
| 379 |
+
},
|
| 380 |
+
{
|
| 381 |
+
"epoch": 1.245808383233533,
|
| 382 |
+
"grad_norm": 149909.921875,
|
| 383 |
+
"learning_rate": 1.2470023980815349e-05,
|
| 384 |
+
"loss": 0.0051,
|
| 385 |
+
"step": 520
|
| 386 |
+
},
|
| 387 |
+
{
|
| 388 |
+
"epoch": 1.2697604790419161,
|
| 389 |
+
"grad_norm": 275816.8125,
|
| 390 |
+
"learning_rate": 1.2709832134292568e-05,
|
| 391 |
+
"loss": 0.0067,
|
| 392 |
+
"step": 530
|
| 393 |
+
},
|
| 394 |
+
{
|
| 395 |
+
"epoch": 1.2937125748502993,
|
| 396 |
+
"grad_norm": 343921.03125,
|
| 397 |
+
"learning_rate": 1.2949640287769784e-05,
|
| 398 |
+
"loss": 0.0052,
|
| 399 |
+
"step": 540
|
| 400 |
+
},
|
| 401 |
+
{
|
| 402 |
+
"epoch": 1.3176646706586825,
|
| 403 |
+
"grad_norm": 132028.453125,
|
| 404 |
+
"learning_rate": 1.3189448441247003e-05,
|
| 405 |
+
"loss": 0.0048,
|
| 406 |
+
"step": 550
|
| 407 |
+
},
|
| 408 |
+
{
|
| 409 |
+
"epoch": 1.341616766467066,
|
| 410 |
+
"grad_norm": 253700.5625,
|
| 411 |
+
"learning_rate": 1.3429256594724223e-05,
|
| 412 |
+
"loss": 0.0044,
|
| 413 |
+
"step": 560
|
| 414 |
+
},
|
| 415 |
+
{
|
| 416 |
+
"epoch": 1.3655688622754492,
|
| 417 |
+
"grad_norm": 66469.2265625,
|
| 418 |
+
"learning_rate": 1.3669064748201439e-05,
|
| 419 |
+
"loss": 0.0034,
|
| 420 |
+
"step": 570
|
| 421 |
+
},
|
| 422 |
+
{
|
| 423 |
+
"epoch": 1.3895209580838324,
|
| 424 |
+
"grad_norm": 266988.59375,
|
| 425 |
+
"learning_rate": 1.3908872901678658e-05,
|
| 426 |
+
"loss": 0.0066,
|
| 427 |
+
"step": 580
|
| 428 |
+
},
|
| 429 |
+
{
|
| 430 |
+
"epoch": 1.4134730538922156,
|
| 431 |
+
"grad_norm": 305832.90625,
|
| 432 |
+
"learning_rate": 1.4148681055155877e-05,
|
| 433 |
+
"loss": 0.0033,
|
| 434 |
+
"step": 590
|
| 435 |
+
},
|
| 436 |
+
{
|
| 437 |
+
"epoch": 1.4374251497005988,
|
| 438 |
+
"grad_norm": 356095.40625,
|
| 439 |
+
"learning_rate": 1.4388489208633095e-05,
|
| 440 |
+
"loss": 0.004,
|
| 441 |
+
"step": 600
|
| 442 |
+
},
|
| 443 |
+
{
|
| 444 |
+
"epoch": 1.461377245508982,
|
| 445 |
+
"grad_norm": 240207.0625,
|
| 446 |
+
"learning_rate": 1.4628297362110312e-05,
|
| 447 |
+
"loss": 0.006,
|
| 448 |
+
"step": 610
|
| 449 |
+
},
|
| 450 |
+
{
|
| 451 |
+
"epoch": 1.4853293413173652,
|
| 452 |
+
"grad_norm": 147457.96875,
|
| 453 |
+
"learning_rate": 1.4868105515587531e-05,
|
| 454 |
+
"loss": 0.0042,
|
| 455 |
+
"step": 620
|
| 456 |
+
},
|
| 457 |
+
{
|
| 458 |
+
"epoch": 1.5092814371257486,
|
| 459 |
+
"grad_norm": 148538.671875,
|
| 460 |
+
"learning_rate": 1.5107913669064749e-05,
|
| 461 |
+
"loss": 0.006,
|
| 462 |
+
"step": 630
|
| 463 |
+
},
|
| 464 |
+
{
|
| 465 |
+
"epoch": 1.5332335329341318,
|
| 466 |
+
"grad_norm": 85050.8359375,
|
| 467 |
+
"learning_rate": 1.534772182254197e-05,
|
| 468 |
+
"loss": 0.0024,
|
| 469 |
+
"step": 640
|
| 470 |
+
},
|
| 471 |
+
{
|
| 472 |
+
"epoch": 1.557185628742515,
|
| 473 |
+
"grad_norm": 211666.234375,
|
| 474 |
+
"learning_rate": 1.5587529976019188e-05,
|
| 475 |
+
"loss": 0.0034,
|
| 476 |
+
"step": 650
|
| 477 |
+
},
|
| 478 |
+
{
|
| 479 |
+
"epoch": 1.5811377245508982,
|
| 480 |
+
"grad_norm": 96871.3203125,
|
| 481 |
+
"learning_rate": 1.5827338129496403e-05,
|
| 482 |
+
"loss": 0.0058,
|
| 483 |
+
"step": 660
|
| 484 |
+
},
|
| 485 |
+
{
|
| 486 |
+
"epoch": 1.6050898203592814,
|
| 487 |
+
"grad_norm": 126729.3984375,
|
| 488 |
+
"learning_rate": 1.6067146282973623e-05,
|
| 489 |
+
"loss": 0.0022,
|
| 490 |
+
"step": 670
|
| 491 |
+
},
|
| 492 |
+
{
|
| 493 |
+
"epoch": 1.6290419161676648,
|
| 494 |
+
"grad_norm": 242174.453125,
|
| 495 |
+
"learning_rate": 1.6306954436450842e-05,
|
| 496 |
+
"loss": 0.0031,
|
| 497 |
+
"step": 680
|
| 498 |
+
},
|
| 499 |
+
{
|
| 500 |
+
"epoch": 1.6529940119760478,
|
| 501 |
+
"grad_norm": 86450.3984375,
|
| 502 |
+
"learning_rate": 1.6546762589928058e-05,
|
| 503 |
+
"loss": 0.0065,
|
| 504 |
+
"step": 690
|
| 505 |
+
},
|
| 506 |
+
{
|
| 507 |
+
"epoch": 1.6769461077844312,
|
| 508 |
+
"grad_norm": 218104.953125,
|
| 509 |
+
"learning_rate": 1.6786570743405277e-05,
|
| 510 |
+
"loss": 0.0018,
|
| 511 |
+
"step": 700
|
| 512 |
+
},
|
| 513 |
+
{
|
| 514 |
+
"epoch": 1.7008982035928144,
|
| 515 |
+
"grad_norm": 59439.62890625,
|
| 516 |
+
"learning_rate": 1.7026378896882496e-05,
|
| 517 |
+
"loss": 0.0028,
|
| 518 |
+
"step": 710
|
| 519 |
+
},
|
| 520 |
+
{
|
| 521 |
+
"epoch": 1.7248502994011976,
|
| 522 |
+
"grad_norm": 593990.1875,
|
| 523 |
+
"learning_rate": 1.7266187050359712e-05,
|
| 524 |
+
"loss": 0.0058,
|
| 525 |
+
"step": 720
|
| 526 |
+
},
|
| 527 |
+
{
|
| 528 |
+
"epoch": 1.7488023952095808,
|
| 529 |
+
"grad_norm": 75406.0546875,
|
| 530 |
+
"learning_rate": 1.750599520383693e-05,
|
| 531 |
+
"loss": 0.0024,
|
| 532 |
+
"step": 730
|
| 533 |
+
},
|
| 534 |
+
{
|
| 535 |
+
"epoch": 1.772754491017964,
|
| 536 |
+
"grad_norm": 339184.53125,
|
| 537 |
+
"learning_rate": 1.774580335731415e-05,
|
| 538 |
+
"loss": 0.0025,
|
| 539 |
+
"step": 740
|
| 540 |
+
},
|
| 541 |
+
{
|
| 542 |
+
"epoch": 1.7967065868263474,
|
| 543 |
+
"grad_norm": 50401.92578125,
|
| 544 |
+
"learning_rate": 1.7985611510791367e-05,
|
| 545 |
+
"loss": 0.0034,
|
| 546 |
+
"step": 750
|
| 547 |
+
},
|
| 548 |
+
{
|
| 549 |
+
"epoch": 1.8206586826347304,
|
| 550 |
+
"grad_norm": 149760.171875,
|
| 551 |
+
"learning_rate": 1.8225419664268586e-05,
|
| 552 |
+
"loss": 0.0007,
|
| 553 |
+
"step": 760
|
| 554 |
+
},
|
| 555 |
+
{
|
| 556 |
+
"epoch": 1.8446107784431138,
|
| 557 |
+
"grad_norm": 186960.296875,
|
| 558 |
+
"learning_rate": 1.8465227817745805e-05,
|
| 559 |
+
"loss": 0.0026,
|
| 560 |
+
"step": 770
|
| 561 |
+
},
|
| 562 |
+
{
|
| 563 |
+
"epoch": 1.868562874251497,
|
| 564 |
+
"grad_norm": 84101.9140625,
|
| 565 |
+
"learning_rate": 1.870503597122302e-05,
|
| 566 |
+
"loss": 0.0041,
|
| 567 |
+
"step": 780
|
| 568 |
+
},
|
| 569 |
+
{
|
| 570 |
+
"epoch": 1.8925149700598802,
|
| 571 |
+
"grad_norm": 158750.65625,
|
| 572 |
+
"learning_rate": 1.894484412470024e-05,
|
| 573 |
+
"loss": 0.0051,
|
| 574 |
+
"step": 790
|
| 575 |
+
},
|
| 576 |
+
{
|
| 577 |
+
"epoch": 1.9164670658682634,
|
| 578 |
+
"grad_norm": 148909.34375,
|
| 579 |
+
"learning_rate": 1.918465227817746e-05,
|
| 580 |
+
"loss": 0.002,
|
| 581 |
+
"step": 800
|
| 582 |
+
},
|
| 583 |
+
{
|
| 584 |
+
"epoch": 1.9404191616766466,
|
| 585 |
+
"grad_norm": 266416.875,
|
| 586 |
+
"learning_rate": 1.9424460431654675e-05,
|
| 587 |
+
"loss": 0.003,
|
| 588 |
+
"step": 810
|
| 589 |
+
},
|
| 590 |
+
{
|
| 591 |
+
"epoch": 1.96437125748503,
|
| 592 |
+
"grad_norm": 262396.09375,
|
| 593 |
+
"learning_rate": 1.9664268585131895e-05,
|
| 594 |
+
"loss": 0.0017,
|
| 595 |
+
"step": 820
|
| 596 |
+
},
|
| 597 |
+
{
|
| 598 |
+
"epoch": 1.988323353293413,
|
| 599 |
+
"grad_norm": 103789.453125,
|
| 600 |
+
"learning_rate": 1.9904076738609114e-05,
|
| 601 |
+
"loss": 0.0022,
|
| 602 |
+
"step": 830
|
| 603 |
+
},
|
| 604 |
+
{
|
| 605 |
+
"epoch": 1.9979041916167666,
|
| 606 |
+
"eval_accuracy": 0.9964628600303184,
|
| 607 |
+
"eval_f1": 0.9965613230718847,
|
| 608 |
+
"eval_loss": 0.00301022338680923,
|
| 609 |
+
"eval_precision": 0.9931462140992167,
|
| 610 |
+
"eval_recall": 1.0,
|
| 611 |
+
"eval_runtime": 156.7957,
|
| 612 |
+
"eval_samples_per_second": 37.865,
|
| 613 |
+
"eval_steps_per_second": 2.373,
|
| 614 |
+
"step": 834
|
| 615 |
+
},
|
| 616 |
+
{
|
| 617 |
+
"epoch": 2.0125748502994014,
|
| 618 |
+
"grad_norm": 38590.04296875,
|
| 619 |
+
"learning_rate": 1.996402877697842e-05,
|
| 620 |
+
"loss": 0.0023,
|
| 621 |
+
"step": 840
|
| 622 |
+
},
|
| 623 |
+
{
|
| 624 |
+
"epoch": 2.0365269461077844,
|
| 625 |
+
"grad_norm": 76568.2109375,
|
| 626 |
+
"learning_rate": 1.9904076738609114e-05,
|
| 627 |
+
"loss": 0.0032,
|
| 628 |
+
"step": 850
|
| 629 |
+
},
|
| 630 |
+
{
|
| 631 |
+
"epoch": 2.060479041916168,
|
| 632 |
+
"grad_norm": 106512.125,
|
| 633 |
+
"learning_rate": 1.984412470023981e-05,
|
| 634 |
+
"loss": 0.0024,
|
| 635 |
+
"step": 860
|
| 636 |
+
},
|
| 637 |
+
{
|
| 638 |
+
"epoch": 2.084431137724551,
|
| 639 |
+
"grad_norm": 156610.90625,
|
| 640 |
+
"learning_rate": 1.9784172661870504e-05,
|
| 641 |
+
"loss": 0.0017,
|
| 642 |
+
"step": 870
|
| 643 |
+
},
|
| 644 |
+
{
|
| 645 |
+
"epoch": 2.108383233532934,
|
| 646 |
+
"grad_norm": 81875.859375,
|
| 647 |
+
"learning_rate": 1.97242206235012e-05,
|
| 648 |
+
"loss": 0.0028,
|
| 649 |
+
"step": 880
|
| 650 |
+
},
|
| 651 |
+
{
|
| 652 |
+
"epoch": 2.132335329341317,
|
| 653 |
+
"grad_norm": 138606.71875,
|
| 654 |
+
"learning_rate": 1.9664268585131895e-05,
|
| 655 |
+
"loss": 0.0014,
|
| 656 |
+
"step": 890
|
| 657 |
+
},
|
| 658 |
+
{
|
| 659 |
+
"epoch": 2.1562874251497006,
|
| 660 |
+
"grad_norm": 381988.15625,
|
| 661 |
+
"learning_rate": 1.960431654676259e-05,
|
| 662 |
+
"loss": 0.0033,
|
| 663 |
+
"step": 900
|
| 664 |
+
},
|
| 665 |
+
{
|
| 666 |
+
"epoch": 2.180239520958084,
|
| 667 |
+
"grad_norm": 83168.578125,
|
| 668 |
+
"learning_rate": 1.954436450839329e-05,
|
| 669 |
+
"loss": 0.001,
|
| 670 |
+
"step": 910
|
| 671 |
+
},
|
| 672 |
+
{
|
| 673 |
+
"epoch": 2.204191616766467,
|
| 674 |
+
"grad_norm": 20675.666015625,
|
| 675 |
+
"learning_rate": 1.9484412470023982e-05,
|
| 676 |
+
"loss": 0.0006,
|
| 677 |
+
"step": 920
|
| 678 |
+
},
|
| 679 |
+
{
|
| 680 |
+
"epoch": 2.2281437125748504,
|
| 681 |
+
"grad_norm": 278388.1875,
|
| 682 |
+
"learning_rate": 1.9424460431654675e-05,
|
| 683 |
+
"loss": 0.0008,
|
| 684 |
+
"step": 930
|
| 685 |
+
},
|
| 686 |
+
{
|
| 687 |
+
"epoch": 2.2520958083832334,
|
| 688 |
+
"grad_norm": 89630.8828125,
|
| 689 |
+
"learning_rate": 1.9364508393285372e-05,
|
| 690 |
+
"loss": 0.0008,
|
| 691 |
+
"step": 940
|
| 692 |
+
},
|
| 693 |
+
{
|
| 694 |
+
"epoch": 2.276047904191617,
|
| 695 |
+
"grad_norm": 172995.15625,
|
| 696 |
+
"learning_rate": 1.930455635491607e-05,
|
| 697 |
+
"loss": 0.0033,
|
| 698 |
+
"step": 950
|
| 699 |
+
},
|
| 700 |
+
{
|
| 701 |
+
"epoch": 2.3,
|
| 702 |
+
"grad_norm": 17161.27734375,
|
| 703 |
+
"learning_rate": 1.9244604316546766e-05,
|
| 704 |
+
"loss": 0.0013,
|
| 705 |
+
"step": 960
|
| 706 |
+
},
|
| 707 |
+
{
|
| 708 |
+
"epoch": 2.3239520958083832,
|
| 709 |
+
"grad_norm": 57101.74609375,
|
| 710 |
+
"learning_rate": 1.918465227817746e-05,
|
| 711 |
+
"loss": 0.0025,
|
| 712 |
+
"step": 970
|
| 713 |
+
},
|
| 714 |
+
{
|
| 715 |
+
"epoch": 2.3479041916167667,
|
| 716 |
+
"grad_norm": 368897.71875,
|
| 717 |
+
"learning_rate": 1.9124700239808156e-05,
|
| 718 |
+
"loss": 0.0042,
|
| 719 |
+
"step": 980
|
| 720 |
+
},
|
| 721 |
+
{
|
| 722 |
+
"epoch": 2.3718562874251496,
|
| 723 |
+
"grad_norm": 33615.3828125,
|
| 724 |
+
"learning_rate": 1.906474820143885e-05,
|
| 725 |
+
"loss": 0.0004,
|
| 726 |
+
"step": 990
|
| 727 |
+
},
|
| 728 |
+
{
|
| 729 |
+
"epoch": 2.395808383233533,
|
| 730 |
+
"grad_norm": 46844.67578125,
|
| 731 |
+
"learning_rate": 1.9004796163069547e-05,
|
| 732 |
+
"loss": 0.0011,
|
| 733 |
+
"step": 1000
|
| 734 |
+
},
|
| 735 |
+
{
|
| 736 |
+
"epoch": 2.419760479041916,
|
| 737 |
+
"grad_norm": 70535.828125,
|
| 738 |
+
"learning_rate": 1.894484412470024e-05,
|
| 739 |
+
"loss": 0.0005,
|
| 740 |
+
"step": 1010
|
| 741 |
+
},
|
| 742 |
+
{
|
| 743 |
+
"epoch": 2.4437125748502995,
|
| 744 |
+
"grad_norm": 41407.75,
|
| 745 |
+
"learning_rate": 1.8884892086330937e-05,
|
| 746 |
+
"loss": 0.0036,
|
| 747 |
+
"step": 1020
|
| 748 |
+
},
|
| 749 |
+
{
|
| 750 |
+
"epoch": 2.4676646706586824,
|
| 751 |
+
"grad_norm": 41804.3671875,
|
| 752 |
+
"learning_rate": 1.8824940047961634e-05,
|
| 753 |
+
"loss": 0.0005,
|
| 754 |
+
"step": 1030
|
| 755 |
+
},
|
| 756 |
+
{
|
| 757 |
+
"epoch": 2.491616766467066,
|
| 758 |
+
"grad_norm": 17560.765625,
|
| 759 |
+
"learning_rate": 1.8764988009592328e-05,
|
| 760 |
+
"loss": 0.0013,
|
| 761 |
+
"step": 1040
|
| 762 |
+
},
|
| 763 |
+
{
|
| 764 |
+
"epoch": 2.5155688622754493,
|
| 765 |
+
"grad_norm": 11264.94921875,
|
| 766 |
+
"learning_rate": 1.870503597122302e-05,
|
| 767 |
+
"loss": 0.0024,
|
| 768 |
+
"step": 1050
|
| 769 |
+
},
|
| 770 |
+
{
|
| 771 |
+
"epoch": 2.5395209580838323,
|
| 772 |
+
"grad_norm": 102580.203125,
|
| 773 |
+
"learning_rate": 1.8645083932853718e-05,
|
| 774 |
+
"loss": 0.0004,
|
| 775 |
+
"step": 1060
|
| 776 |
+
},
|
| 777 |
+
{
|
| 778 |
+
"epoch": 2.5634730538922157,
|
| 779 |
+
"grad_norm": 19371.921875,
|
| 780 |
+
"learning_rate": 1.8585131894484415e-05,
|
| 781 |
+
"loss": 0.0021,
|
| 782 |
+
"step": 1070
|
| 783 |
+
},
|
| 784 |
+
{
|
| 785 |
+
"epoch": 2.5874251497005987,
|
| 786 |
+
"grad_norm": 59045.21484375,
|
| 787 |
+
"learning_rate": 1.8525179856115108e-05,
|
| 788 |
+
"loss": 0.0039,
|
| 789 |
+
"step": 1080
|
| 790 |
+
},
|
| 791 |
+
{
|
| 792 |
+
"epoch": 2.611377245508982,
|
| 793 |
+
"grad_norm": 37528.83984375,
|
| 794 |
+
"learning_rate": 1.8465227817745805e-05,
|
| 795 |
+
"loss": 0.0015,
|
| 796 |
+
"step": 1090
|
| 797 |
+
},
|
| 798 |
+
{
|
| 799 |
+
"epoch": 2.635329341317365,
|
| 800 |
+
"grad_norm": 484952.59375,
|
| 801 |
+
"learning_rate": 1.8405275779376502e-05,
|
| 802 |
+
"loss": 0.0023,
|
| 803 |
+
"step": 1100
|
| 804 |
+
},
|
| 805 |
+
{
|
| 806 |
+
"epoch": 2.6592814371257485,
|
| 807 |
+
"grad_norm": 14117.0771484375,
|
| 808 |
+
"learning_rate": 1.8345323741007196e-05,
|
| 809 |
+
"loss": 0.0017,
|
| 810 |
+
"step": 1110
|
| 811 |
+
},
|
| 812 |
+
{
|
| 813 |
+
"epoch": 2.683233532934132,
|
| 814 |
+
"grad_norm": 276284.59375,
|
| 815 |
+
"learning_rate": 1.8285371702637892e-05,
|
| 816 |
+
"loss": 0.0016,
|
| 817 |
+
"step": 1120
|
| 818 |
+
},
|
| 819 |
+
{
|
| 820 |
+
"epoch": 2.707185628742515,
|
| 821 |
+
"grad_norm": 40982.03515625,
|
| 822 |
+
"learning_rate": 1.8225419664268586e-05,
|
| 823 |
+
"loss": 0.001,
|
| 824 |
+
"step": 1130
|
| 825 |
+
},
|
| 826 |
+
{
|
| 827 |
+
"epoch": 2.7311377245508983,
|
| 828 |
+
"grad_norm": 101166.234375,
|
| 829 |
+
"learning_rate": 1.8165467625899283e-05,
|
| 830 |
+
"loss": 0.0009,
|
| 831 |
+
"step": 1140
|
| 832 |
+
},
|
| 833 |
+
{
|
| 834 |
+
"epoch": 2.7550898203592813,
|
| 835 |
+
"grad_norm": 51041.46875,
|
| 836 |
+
"learning_rate": 1.810551558752998e-05,
|
| 837 |
+
"loss": 0.0013,
|
| 838 |
+
"step": 1150
|
| 839 |
+
},
|
| 840 |
+
{
|
| 841 |
+
"epoch": 2.7790419161676647,
|
| 842 |
+
"grad_norm": 204233.671875,
|
| 843 |
+
"learning_rate": 1.8045563549160673e-05,
|
| 844 |
+
"loss": 0.0016,
|
| 845 |
+
"step": 1160
|
| 846 |
+
},
|
| 847 |
+
{
|
| 848 |
+
"epoch": 2.8029940119760477,
|
| 849 |
+
"grad_norm": 143324.921875,
|
| 850 |
+
"learning_rate": 1.7985611510791367e-05,
|
| 851 |
+
"loss": 0.0026,
|
| 852 |
+
"step": 1170
|
| 853 |
+
},
|
| 854 |
+
{
|
| 855 |
+
"epoch": 2.826946107784431,
|
| 856 |
+
"grad_norm": 44342.2734375,
|
| 857 |
+
"learning_rate": 1.7925659472422064e-05,
|
| 858 |
+
"loss": 0.0023,
|
| 859 |
+
"step": 1180
|
| 860 |
+
},
|
| 861 |
+
{
|
| 862 |
+
"epoch": 2.8508982035928145,
|
| 863 |
+
"grad_norm": 75831.65625,
|
| 864 |
+
"learning_rate": 1.786570743405276e-05,
|
| 865 |
+
"loss": 0.0016,
|
| 866 |
+
"step": 1190
|
| 867 |
+
},
|
| 868 |
+
{
|
| 869 |
+
"epoch": 2.8748502994011975,
|
| 870 |
+
"grad_norm": 96995.328125,
|
| 871 |
+
"learning_rate": 1.7805755395683454e-05,
|
| 872 |
+
"loss": 0.0003,
|
| 873 |
+
"step": 1200
|
| 874 |
+
},
|
| 875 |
+
{
|
| 876 |
+
"epoch": 2.898802395209581,
|
| 877 |
+
"grad_norm": 110547.0546875,
|
| 878 |
+
"learning_rate": 1.774580335731415e-05,
|
| 879 |
+
"loss": 0.0003,
|
| 880 |
+
"step": 1210
|
| 881 |
+
},
|
| 882 |
+
{
|
| 883 |
+
"epoch": 2.922754491017964,
|
| 884 |
+
"grad_norm": 30139.875,
|
| 885 |
+
"learning_rate": 1.7685851318944848e-05,
|
| 886 |
+
"loss": 0.0018,
|
| 887 |
+
"step": 1220
|
| 888 |
+
},
|
| 889 |
+
{
|
| 890 |
+
"epoch": 2.9467065868263473,
|
| 891 |
+
"grad_norm": 42453.53515625,
|
| 892 |
+
"learning_rate": 1.762589928057554e-05,
|
| 893 |
+
"loss": 0.0002,
|
| 894 |
+
"step": 1230
|
| 895 |
+
},
|
| 896 |
+
{
|
| 897 |
+
"epoch": 2.9706586826347303,
|
| 898 |
+
"grad_norm": 49538.61328125,
|
| 899 |
+
"learning_rate": 1.7565947242206235e-05,
|
| 900 |
+
"loss": 0.001,
|
| 901 |
+
"step": 1240
|
| 902 |
+
},
|
| 903 |
+
{
|
| 904 |
+
"epoch": 2.9946107784431137,
|
| 905 |
+
"grad_norm": 121568.9453125,
|
| 906 |
+
"learning_rate": 1.750599520383693e-05,
|
| 907 |
+
"loss": 0.0007,
|
| 908 |
+
"step": 1250
|
| 909 |
+
},
|
| 910 |
+
{
|
| 911 |
+
"epoch": 2.9994011976047905,
|
| 912 |
+
"eval_accuracy": 0.9823143001515917,
|
| 913 |
+
"eval_f1": 0.9830398966241318,
|
| 914 |
+
"eval_loss": 0.0027729119174182415,
|
| 915 |
+
"eval_precision": 0.9666454891994918,
|
| 916 |
+
"eval_recall": 1.0,
|
| 917 |
+
"eval_runtime": 156.768,
|
| 918 |
+
"eval_samples_per_second": 37.871,
|
| 919 |
+
"eval_steps_per_second": 2.373,
|
| 920 |
+
"step": 1252
|
| 921 |
+
},
|
| 922 |
+
{
|
| 923 |
+
"epoch": 3.0188622754491017,
|
| 924 |
+
"grad_norm": 54773.9140625,
|
| 925 |
+
"learning_rate": 1.744604316546763e-05,
|
| 926 |
+
"loss": 0.0008,
|
| 927 |
+
"step": 1260
|
| 928 |
+
},
|
| 929 |
+
{
|
| 930 |
+
"epoch": 3.042814371257485,
|
| 931 |
+
"grad_norm": 130051.09375,
|
| 932 |
+
"learning_rate": 1.7386091127098322e-05,
|
| 933 |
+
"loss": 0.0004,
|
| 934 |
+
"step": 1270
|
| 935 |
+
},
|
| 936 |
+
{
|
| 937 |
+
"epoch": 3.066766467065868,
|
| 938 |
+
"grad_norm": 38819.87109375,
|
| 939 |
+
"learning_rate": 1.732613908872902e-05,
|
| 940 |
+
"loss": 0.0012,
|
| 941 |
+
"step": 1280
|
| 942 |
+
},
|
| 943 |
+
{
|
| 944 |
+
"epoch": 3.0907185628742515,
|
| 945 |
+
"grad_norm": 36351.796875,
|
| 946 |
+
"learning_rate": 1.7266187050359712e-05,
|
| 947 |
+
"loss": 0.0009,
|
| 948 |
+
"step": 1290
|
| 949 |
+
},
|
| 950 |
+
{
|
| 951 |
+
"epoch": 3.114670658682635,
|
| 952 |
+
"grad_norm": 71120.03125,
|
| 953 |
+
"learning_rate": 1.720623501199041e-05,
|
| 954 |
+
"loss": 0.0027,
|
| 955 |
+
"step": 1300
|
| 956 |
+
},
|
| 957 |
+
{
|
| 958 |
+
"epoch": 3.138622754491018,
|
| 959 |
+
"grad_norm": 50392.36328125,
|
| 960 |
+
"learning_rate": 1.7146282973621106e-05,
|
| 961 |
+
"loss": 0.0045,
|
| 962 |
+
"step": 1310
|
| 963 |
+
},
|
| 964 |
+
{
|
| 965 |
+
"epoch": 3.1625748502994013,
|
| 966 |
+
"grad_norm": 25353.765625,
|
| 967 |
+
"learning_rate": 1.70863309352518e-05,
|
| 968 |
+
"loss": 0.0016,
|
| 969 |
+
"step": 1320
|
| 970 |
+
},
|
| 971 |
+
{
|
| 972 |
+
"epoch": 3.1865269461077843,
|
| 973 |
+
"grad_norm": 55808.15625,
|
| 974 |
+
"learning_rate": 1.7026378896882496e-05,
|
| 975 |
+
"loss": 0.002,
|
| 976 |
+
"step": 1330
|
| 977 |
+
},
|
| 978 |
+
{
|
| 979 |
+
"epoch": 3.2104790419161677,
|
| 980 |
+
"grad_norm": 35924.5078125,
|
| 981 |
+
"learning_rate": 1.6966426858513193e-05,
|
| 982 |
+
"loss": 0.0002,
|
| 983 |
+
"step": 1340
|
| 984 |
+
},
|
| 985 |
+
{
|
| 986 |
+
"epoch": 3.2344311377245507,
|
| 987 |
+
"grad_norm": 8763.8037109375,
|
| 988 |
+
"learning_rate": 1.6906474820143887e-05,
|
| 989 |
+
"loss": 0.0005,
|
| 990 |
+
"step": 1350
|
| 991 |
+
},
|
| 992 |
+
{
|
| 993 |
+
"epoch": 3.258383233532934,
|
| 994 |
+
"grad_norm": 3576.546875,
|
| 995 |
+
"learning_rate": 1.684652278177458e-05,
|
| 996 |
+
"loss": 0.0004,
|
| 997 |
+
"step": 1360
|
| 998 |
+
},
|
| 999 |
+
{
|
| 1000 |
+
"epoch": 3.2823353293413176,
|
| 1001 |
+
"grad_norm": 53395.13671875,
|
| 1002 |
+
"learning_rate": 1.6786570743405277e-05,
|
| 1003 |
+
"loss": 0.0001,
|
| 1004 |
+
"step": 1370
|
| 1005 |
+
},
|
| 1006 |
+
{
|
| 1007 |
+
"epoch": 3.3062874251497005,
|
| 1008 |
+
"grad_norm": 14970.0224609375,
|
| 1009 |
+
"learning_rate": 1.6726618705035974e-05,
|
| 1010 |
+
"loss": 0.0001,
|
| 1011 |
+
"step": 1380
|
| 1012 |
+
},
|
| 1013 |
+
{
|
| 1014 |
+
"epoch": 3.330239520958084,
|
| 1015 |
+
"grad_norm": 62940.63671875,
|
| 1016 |
+
"learning_rate": 1.6666666666666667e-05,
|
| 1017 |
+
"loss": 0.0008,
|
| 1018 |
+
"step": 1390
|
| 1019 |
+
},
|
| 1020 |
+
{
|
| 1021 |
+
"epoch": 3.354191616766467,
|
| 1022 |
+
"grad_norm": 66767.4453125,
|
| 1023 |
+
"learning_rate": 1.660671462829736e-05,
|
| 1024 |
+
"loss": 0.001,
|
| 1025 |
+
"step": 1400
|
| 1026 |
+
},
|
| 1027 |
+
{
|
| 1028 |
+
"epoch": 3.3781437125748504,
|
| 1029 |
+
"grad_norm": 62226.42578125,
|
| 1030 |
+
"learning_rate": 1.6546762589928058e-05,
|
| 1031 |
+
"loss": 0.0015,
|
| 1032 |
+
"step": 1410
|
| 1033 |
+
},
|
| 1034 |
+
{
|
| 1035 |
+
"epoch": 3.4020958083832333,
|
| 1036 |
+
"grad_norm": 43982.64453125,
|
| 1037 |
+
"learning_rate": 1.6486810551558755e-05,
|
| 1038 |
+
"loss": 0.0018,
|
| 1039 |
+
"step": 1420
|
| 1040 |
+
},
|
| 1041 |
+
{
|
| 1042 |
+
"epoch": 3.4260479041916168,
|
| 1043 |
+
"grad_norm": 114928.5390625,
|
| 1044 |
+
"learning_rate": 1.6426858513189448e-05,
|
| 1045 |
+
"loss": 0.0036,
|
| 1046 |
+
"step": 1430
|
| 1047 |
+
},
|
| 1048 |
+
{
|
| 1049 |
+
"epoch": 3.45,
|
| 1050 |
+
"grad_norm": 101840.3515625,
|
| 1051 |
+
"learning_rate": 1.6366906474820145e-05,
|
| 1052 |
+
"loss": 0.0011,
|
| 1053 |
+
"step": 1440
|
| 1054 |
+
},
|
| 1055 |
+
{
|
| 1056 |
+
"epoch": 3.473952095808383,
|
| 1057 |
+
"grad_norm": 59594.5078125,
|
| 1058 |
+
"learning_rate": 1.6306954436450842e-05,
|
| 1059 |
+
"loss": 0.0033,
|
| 1060 |
+
"step": 1450
|
| 1061 |
+
},
|
| 1062 |
+
{
|
| 1063 |
+
"epoch": 3.4979041916167666,
|
| 1064 |
+
"grad_norm": 18318.77734375,
|
| 1065 |
+
"learning_rate": 1.6247002398081535e-05,
|
| 1066 |
+
"loss": 0.0004,
|
| 1067 |
+
"step": 1460
|
| 1068 |
+
},
|
| 1069 |
+
{
|
| 1070 |
+
"epoch": 3.5218562874251496,
|
| 1071 |
+
"grad_norm": 18281.22265625,
|
| 1072 |
+
"learning_rate": 1.6187050359712232e-05,
|
| 1073 |
+
"loss": 0.0003,
|
| 1074 |
+
"step": 1470
|
| 1075 |
+
},
|
| 1076 |
+
{
|
| 1077 |
+
"epoch": 3.545808383233533,
|
| 1078 |
+
"grad_norm": 13549.404296875,
|
| 1079 |
+
"learning_rate": 1.6127098321342926e-05,
|
| 1080 |
+
"loss": 0.0009,
|
| 1081 |
+
"step": 1480
|
| 1082 |
+
},
|
| 1083 |
+
{
|
| 1084 |
+
"epoch": 3.569760479041916,
|
| 1085 |
+
"grad_norm": 13404.5205078125,
|
| 1086 |
+
"learning_rate": 1.6067146282973623e-05,
|
| 1087 |
+
"loss": 0.0007,
|
| 1088 |
+
"step": 1490
|
| 1089 |
+
},
|
| 1090 |
+
{
|
| 1091 |
+
"epoch": 3.5937125748502994,
|
| 1092 |
+
"grad_norm": 36192.89453125,
|
| 1093 |
+
"learning_rate": 1.600719424460432e-05,
|
| 1094 |
+
"loss": 0.0002,
|
| 1095 |
+
"step": 1500
|
| 1096 |
+
},
|
| 1097 |
+
{
|
| 1098 |
+
"epoch": 3.617664670658683,
|
| 1099 |
+
"grad_norm": 38474.55859375,
|
| 1100 |
+
"learning_rate": 1.5947242206235013e-05,
|
| 1101 |
+
"loss": 0.0017,
|
| 1102 |
+
"step": 1510
|
| 1103 |
+
},
|
| 1104 |
+
{
|
| 1105 |
+
"epoch": 3.641616766467066,
|
| 1106 |
+
"grad_norm": 36329.5,
|
| 1107 |
+
"learning_rate": 1.5887290167865707e-05,
|
| 1108 |
+
"loss": 0.0006,
|
| 1109 |
+
"step": 1520
|
| 1110 |
+
},
|
| 1111 |
+
{
|
| 1112 |
+
"epoch": 3.665568862275449,
|
| 1113 |
+
"grad_norm": 839098.25,
|
| 1114 |
+
"learning_rate": 1.5827338129496403e-05,
|
| 1115 |
+
"loss": 0.0004,
|
| 1116 |
+
"step": 1530
|
| 1117 |
+
},
|
| 1118 |
+
{
|
| 1119 |
+
"epoch": 3.689520958083832,
|
| 1120 |
+
"grad_norm": 29361.619140625,
|
| 1121 |
+
"learning_rate": 1.57673860911271e-05,
|
| 1122 |
+
"loss": 0.001,
|
| 1123 |
+
"step": 1540
|
| 1124 |
+
},
|
| 1125 |
+
{
|
| 1126 |
+
"epoch": 3.7134730538922156,
|
| 1127 |
+
"grad_norm": 14026.3037109375,
|
| 1128 |
+
"learning_rate": 1.5707434052757794e-05,
|
| 1129 |
+
"loss": 0.0009,
|
| 1130 |
+
"step": 1550
|
| 1131 |
+
},
|
| 1132 |
+
{
|
| 1133 |
+
"epoch": 3.7374251497005986,
|
| 1134 |
+
"grad_norm": 17467.173828125,
|
| 1135 |
+
"learning_rate": 1.564748201438849e-05,
|
| 1136 |
+
"loss": 0.0016,
|
| 1137 |
+
"step": 1560
|
| 1138 |
+
},
|
| 1139 |
+
{
|
| 1140 |
+
"epoch": 3.761377245508982,
|
| 1141 |
+
"grad_norm": 27251.703125,
|
| 1142 |
+
"learning_rate": 1.5587529976019188e-05,
|
| 1143 |
+
"loss": 0.001,
|
| 1144 |
+
"step": 1570
|
| 1145 |
+
},
|
| 1146 |
+
{
|
| 1147 |
+
"epoch": 3.7853293413173654,
|
| 1148 |
+
"grad_norm": 54221.84765625,
|
| 1149 |
+
"learning_rate": 1.552757793764988e-05,
|
| 1150 |
+
"loss": 0.0002,
|
| 1151 |
+
"step": 1580
|
| 1152 |
+
},
|
| 1153 |
+
{
|
| 1154 |
+
"epoch": 3.8092814371257484,
|
| 1155 |
+
"grad_norm": 22032.01953125,
|
| 1156 |
+
"learning_rate": 1.5467625899280575e-05,
|
| 1157 |
+
"loss": 0.001,
|
| 1158 |
+
"step": 1590
|
| 1159 |
+
},
|
| 1160 |
+
{
|
| 1161 |
+
"epoch": 3.833233532934132,
|
| 1162 |
+
"grad_norm": 32913.08203125,
|
| 1163 |
+
"learning_rate": 1.540767386091127e-05,
|
| 1164 |
+
"loss": 0.0019,
|
| 1165 |
+
"step": 1600
|
| 1166 |
+
},
|
| 1167 |
+
{
|
| 1168 |
+
"epoch": 3.857185628742515,
|
| 1169 |
+
"grad_norm": 16616.380859375,
|
| 1170 |
+
"learning_rate": 1.534772182254197e-05,
|
| 1171 |
+
"loss": 0.0009,
|
| 1172 |
+
"step": 1610
|
| 1173 |
+
},
|
| 1174 |
+
{
|
| 1175 |
+
"epoch": 3.8811377245508982,
|
| 1176 |
+
"grad_norm": 80629.234375,
|
| 1177 |
+
"learning_rate": 1.5287769784172665e-05,
|
| 1178 |
+
"loss": 0.0001,
|
| 1179 |
+
"step": 1620
|
| 1180 |
+
},
|
| 1181 |
+
{
|
| 1182 |
+
"epoch": 3.905089820359281,
|
| 1183 |
+
"grad_norm": 16487.8671875,
|
| 1184 |
+
"learning_rate": 1.5227817745803359e-05,
|
| 1185 |
+
"loss": 0.0007,
|
| 1186 |
+
"step": 1630
|
| 1187 |
+
},
|
| 1188 |
+
{
|
| 1189 |
+
"epoch": 3.9290419161676646,
|
| 1190 |
+
"grad_norm": 11642.541015625,
|
| 1191 |
+
"learning_rate": 1.5167865707434052e-05,
|
| 1192 |
+
"loss": 0.0011,
|
| 1193 |
+
"step": 1640
|
| 1194 |
+
},
|
| 1195 |
+
{
|
| 1196 |
+
"epoch": 3.952994011976048,
|
| 1197 |
+
"grad_norm": 37791.38671875,
|
| 1198 |
+
"learning_rate": 1.5107913669064749e-05,
|
| 1199 |
+
"loss": 0.0004,
|
| 1200 |
+
"step": 1650
|
| 1201 |
+
},
|
| 1202 |
+
{
|
| 1203 |
+
"epoch": 3.976946107784431,
|
| 1204 |
+
"grad_norm": 16768.58203125,
|
| 1205 |
+
"learning_rate": 1.5047961630695444e-05,
|
| 1206 |
+
"loss": 0.0015,
|
| 1207 |
+
"step": 1660
|
| 1208 |
+
},
|
| 1209 |
+
{
|
| 1210 |
+
"epoch": 3.998502994011976,
|
| 1211 |
+
"eval_accuracy": 0.9984840828701365,
|
| 1212 |
+
"eval_f1": 0.9985233798195242,
|
| 1213 |
+
"eval_loss": 0.0018289362778887153,
|
| 1214 |
+
"eval_precision": 0.997051114023591,
|
| 1215 |
+
"eval_recall": 1.0,
|
| 1216 |
+
"eval_runtime": 157.3988,
|
| 1217 |
+
"eval_samples_per_second": 37.719,
|
| 1218 |
+
"eval_steps_per_second": 2.363,
|
| 1219 |
+
"step": 1669
|
| 1220 |
+
}
|
| 1221 |
+
],
|
| 1222 |
+
"logging_steps": 10,
|
| 1223 |
+
"max_steps": 4170,
|
| 1224 |
+
"num_input_tokens_seen": 0,
|
| 1225 |
+
"num_train_epochs": 10,
|
| 1226 |
+
"save_steps": 500,
|
| 1227 |
+
"stateful_callbacks": {
|
| 1228 |
+
"EarlyStoppingCallback": {
|
| 1229 |
+
"args": {
|
| 1230 |
+
"early_stopping_patience": 3,
|
| 1231 |
+
"early_stopping_threshold": 0.01
|
| 1232 |
+
},
|
| 1233 |
+
"attributes": {
|
| 1234 |
+
"early_stopping_patience_counter": 3
|
| 1235 |
+
}
|
| 1236 |
+
},
|
| 1237 |
+
"TrainerControl": {
|
| 1238 |
+
"args": {
|
| 1239 |
+
"should_epoch_stop": false,
|
| 1240 |
+
"should_evaluate": false,
|
| 1241 |
+
"should_log": false,
|
| 1242 |
+
"should_save": true,
|
| 1243 |
+
"should_training_stop": true
|
| 1244 |
+
},
|
| 1245 |
+
"attributes": {}
|
| 1246 |
+
}
|
| 1247 |
+
},
|
| 1248 |
+
"total_flos": 7.478687595449549e+16,
|
| 1249 |
+
"train_batch_size": 16,
|
| 1250 |
+
"trial_name": null,
|
| 1251 |
+
"trial_params": null
|
| 1252 |
+
}
|
training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:9615c5ea3e960e79d4b8cdc71a576ac4a577dbd92d77caf40421c0eddaea1df1
|
| 3 |
+
size 5713
|