Upload folder using huggingface_hub
Browse files- config.json +25 -0
- model.safetensors +3 -0
- optimizer.pt +3 -0
- rng_state_0.pth +3 -0
- rng_state_1.pth +3 -0
- scheduler.pt +3 -0
- trainer_state.json +2834 -0
- training_args.bin +3 -0
config.json
ADDED
|
@@ -0,0 +1,25 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"architectures": [
|
| 3 |
+
"BertModel"
|
| 4 |
+
],
|
| 5 |
+
"attention_probs_dropout_prob": 0.1,
|
| 6 |
+
"classifier_dropout": null,
|
| 7 |
+
"dtype": "float32",
|
| 8 |
+
"gradient_checkpointing": false,
|
| 9 |
+
"hidden_act": "gelu",
|
| 10 |
+
"hidden_dropout_prob": 0.1,
|
| 11 |
+
"hidden_size": 768,
|
| 12 |
+
"initializer_range": 0.02,
|
| 13 |
+
"intermediate_size": 3072,
|
| 14 |
+
"layer_norm_eps": 1e-12,
|
| 15 |
+
"max_position_embeddings": 512,
|
| 16 |
+
"model_type": "bert",
|
| 17 |
+
"num_attention_heads": 12,
|
| 18 |
+
"num_hidden_layers": 12,
|
| 19 |
+
"pad_token_id": 0,
|
| 20 |
+
"position_embedding_type": "absolute",
|
| 21 |
+
"transformers_version": "4.57.3",
|
| 22 |
+
"type_vocab_size": 2,
|
| 23 |
+
"use_cache": true,
|
| 24 |
+
"vocab_size": 30522
|
| 25 |
+
}
|
model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:57cb36b7ada7fcb3161b6714c1b0d5eaa3a26d9931505d21d7766aaea592cb41
|
| 3 |
+
size 437951328
|
optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:6872bdeb6ceb7c023c095cbc3538ebe06f6c3614b2cb9edf7ca98d5fff876dd4
|
| 3 |
+
size 871301195
|
rng_state_0.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:da2afed8ef5fedb8767b73d5e797d4e7acac10ea48ea1613c396dc26ee2b3bf9
|
| 3 |
+
size 14917
|
rng_state_1.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:cf42ccc321a3167e8cb22526c8c1ef07dc771be68e720aaf43466a8e7db83836
|
| 3 |
+
size 14917
|
scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:29b3fb56e014004eda566be571c0ce0e959f96903a2d160ed923e6293335a3b2
|
| 3 |
+
size 1465
|
trainer_state.json
ADDED
|
@@ -0,0 +1,2834 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": null,
|
| 3 |
+
"best_metric": null,
|
| 4 |
+
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 1.852366398073539,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 20000,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.004630915995183847,
|
| 14 |
+
"grad_norm": 51.771697998046875,
|
| 15 |
+
"learning_rate": 1.5123456790123457e-07,
|
| 16 |
+
"loss": 4.9315,
|
| 17 |
+
"step": 50
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.009261831990367694,
|
| 21 |
+
"grad_norm": 48.02553939819336,
|
| 22 |
+
"learning_rate": 3.055555555555556e-07,
|
| 23 |
+
"loss": 4.8131,
|
| 24 |
+
"step": 100
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.013892747985551542,
|
| 28 |
+
"grad_norm": 50.21088409423828,
|
| 29 |
+
"learning_rate": 4.5987654320987657e-07,
|
| 30 |
+
"loss": 4.148,
|
| 31 |
+
"step": 150
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.018523663980735388,
|
| 35 |
+
"grad_norm": 31.535659790039062,
|
| 36 |
+
"learning_rate": 6.141975308641976e-07,
|
| 37 |
+
"loss": 3.4211,
|
| 38 |
+
"step": 200
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.023154579975919236,
|
| 42 |
+
"grad_norm": 24.023160934448242,
|
| 43 |
+
"learning_rate": 7.685185185185186e-07,
|
| 44 |
+
"loss": 2.6902,
|
| 45 |
+
"step": 250
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.027785495971103084,
|
| 49 |
+
"grad_norm": 20.243444442749023,
|
| 50 |
+
"learning_rate": 9.228395061728395e-07,
|
| 51 |
+
"loss": 2.3382,
|
| 52 |
+
"step": 300
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 0.03241641196628693,
|
| 56 |
+
"grad_norm": 17.97671890258789,
|
| 57 |
+
"learning_rate": 1.0771604938271605e-06,
|
| 58 |
+
"loss": 2.1599,
|
| 59 |
+
"step": 350
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 0.037047327961470776,
|
| 63 |
+
"grad_norm": 18.1010684967041,
|
| 64 |
+
"learning_rate": 1.2314814814814815e-06,
|
| 65 |
+
"loss": 1.9409,
|
| 66 |
+
"step": 400
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 0.04167824395665463,
|
| 70 |
+
"grad_norm": 16.637313842773438,
|
| 71 |
+
"learning_rate": 1.3858024691358028e-06,
|
| 72 |
+
"loss": 1.9179,
|
| 73 |
+
"step": 450
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 0.04630915995183847,
|
| 77 |
+
"grad_norm": 22.70376205444336,
|
| 78 |
+
"learning_rate": 1.5401234567901236e-06,
|
| 79 |
+
"loss": 1.8012,
|
| 80 |
+
"step": 500
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.05094007594702232,
|
| 84 |
+
"grad_norm": 20.450454711914062,
|
| 85 |
+
"learning_rate": 1.6944444444444446e-06,
|
| 86 |
+
"loss": 1.6436,
|
| 87 |
+
"step": 550
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 0.05557099194220617,
|
| 91 |
+
"grad_norm": 16.855588912963867,
|
| 92 |
+
"learning_rate": 1.8487654320987655e-06,
|
| 93 |
+
"loss": 1.754,
|
| 94 |
+
"step": 600
|
| 95 |
+
},
|
| 96 |
+
{
|
| 97 |
+
"epoch": 0.06020190793739002,
|
| 98 |
+
"grad_norm": 15.478442192077637,
|
| 99 |
+
"learning_rate": 2.0030864197530867e-06,
|
| 100 |
+
"loss": 1.6117,
|
| 101 |
+
"step": 650
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"epoch": 0.06483282393257386,
|
| 105 |
+
"grad_norm": 14.401535034179688,
|
| 106 |
+
"learning_rate": 2.1574074074074073e-06,
|
| 107 |
+
"loss": 1.574,
|
| 108 |
+
"step": 700
|
| 109 |
+
},
|
| 110 |
+
{
|
| 111 |
+
"epoch": 0.06946373992775771,
|
| 112 |
+
"grad_norm": 19.8779296875,
|
| 113 |
+
"learning_rate": 2.311728395061729e-06,
|
| 114 |
+
"loss": 1.4455,
|
| 115 |
+
"step": 750
|
| 116 |
+
},
|
| 117 |
+
{
|
| 118 |
+
"epoch": 0.07409465592294155,
|
| 119 |
+
"grad_norm": 23.51628875732422,
|
| 120 |
+
"learning_rate": 2.4660493827160494e-06,
|
| 121 |
+
"loss": 1.5821,
|
| 122 |
+
"step": 800
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 0.0787255719181254,
|
| 126 |
+
"grad_norm": 13.794211387634277,
|
| 127 |
+
"learning_rate": 2.6203703703703705e-06,
|
| 128 |
+
"loss": 1.3384,
|
| 129 |
+
"step": 850
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 0.08335648791330925,
|
| 133 |
+
"grad_norm": 17.96605110168457,
|
| 134 |
+
"learning_rate": 2.7746913580246915e-06,
|
| 135 |
+
"loss": 1.3484,
|
| 136 |
+
"step": 900
|
| 137 |
+
},
|
| 138 |
+
{
|
| 139 |
+
"epoch": 0.0879874039084931,
|
| 140 |
+
"grad_norm": 18.876876831054688,
|
| 141 |
+
"learning_rate": 2.9290123456790126e-06,
|
| 142 |
+
"loss": 1.3963,
|
| 143 |
+
"step": 950
|
| 144 |
+
},
|
| 145 |
+
{
|
| 146 |
+
"epoch": 0.09261831990367694,
|
| 147 |
+
"grad_norm": 21.498750686645508,
|
| 148 |
+
"learning_rate": 3.0833333333333336e-06,
|
| 149 |
+
"loss": 1.3247,
|
| 150 |
+
"step": 1000
|
| 151 |
+
},
|
| 152 |
+
{
|
| 153 |
+
"epoch": 0.0972492358988608,
|
| 154 |
+
"grad_norm": 15.515962600708008,
|
| 155 |
+
"learning_rate": 3.2376543209876542e-06,
|
| 156 |
+
"loss": 1.2645,
|
| 157 |
+
"step": 1050
|
| 158 |
+
},
|
| 159 |
+
{
|
| 160 |
+
"epoch": 0.10188015189404465,
|
| 161 |
+
"grad_norm": 19.241085052490234,
|
| 162 |
+
"learning_rate": 3.3919753086419753e-06,
|
| 163 |
+
"loss": 1.2757,
|
| 164 |
+
"step": 1100
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 0.10651106788922848,
|
| 168 |
+
"grad_norm": 19.796600341796875,
|
| 169 |
+
"learning_rate": 3.5462962962962967e-06,
|
| 170 |
+
"loss": 1.2339,
|
| 171 |
+
"step": 1150
|
| 172 |
+
},
|
| 173 |
+
{
|
| 174 |
+
"epoch": 0.11114198388441233,
|
| 175 |
+
"grad_norm": 818.7846069335938,
|
| 176 |
+
"learning_rate": 3.7006172839506178e-06,
|
| 177 |
+
"loss": 1.2809,
|
| 178 |
+
"step": 1200
|
| 179 |
+
},
|
| 180 |
+
{
|
| 181 |
+
"epoch": 0.11577289987959619,
|
| 182 |
+
"grad_norm": 13.61767578125,
|
| 183 |
+
"learning_rate": 3.854938271604939e-06,
|
| 184 |
+
"loss": 1.185,
|
| 185 |
+
"step": 1250
|
| 186 |
+
},
|
| 187 |
+
{
|
| 188 |
+
"epoch": 0.12040381587478004,
|
| 189 |
+
"grad_norm": 17.602561950683594,
|
| 190 |
+
"learning_rate": 4.0092592592592594e-06,
|
| 191 |
+
"loss": 1.1874,
|
| 192 |
+
"step": 1300
|
| 193 |
+
},
|
| 194 |
+
{
|
| 195 |
+
"epoch": 0.1250347318699639,
|
| 196 |
+
"grad_norm": 17.445096969604492,
|
| 197 |
+
"learning_rate": 4.163580246913581e-06,
|
| 198 |
+
"loss": 1.273,
|
| 199 |
+
"step": 1350
|
| 200 |
+
},
|
| 201 |
+
{
|
| 202 |
+
"epoch": 0.12966564786514773,
|
| 203 |
+
"grad_norm": 18.69367790222168,
|
| 204 |
+
"learning_rate": 4.3179012345679015e-06,
|
| 205 |
+
"loss": 1.1443,
|
| 206 |
+
"step": 1400
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 0.13429656386033156,
|
| 210 |
+
"grad_norm": 16.09065055847168,
|
| 211 |
+
"learning_rate": 4.472222222222223e-06,
|
| 212 |
+
"loss": 1.2184,
|
| 213 |
+
"step": 1450
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"epoch": 0.13892747985551543,
|
| 217 |
+
"grad_norm": 16.429534912109375,
|
| 218 |
+
"learning_rate": 4.626543209876544e-06,
|
| 219 |
+
"loss": 1.1603,
|
| 220 |
+
"step": 1500
|
| 221 |
+
},
|
| 222 |
+
{
|
| 223 |
+
"epoch": 0.14355839585069927,
|
| 224 |
+
"grad_norm": 14.309248924255371,
|
| 225 |
+
"learning_rate": 4.780864197530864e-06,
|
| 226 |
+
"loss": 1.1405,
|
| 227 |
+
"step": 1550
|
| 228 |
+
},
|
| 229 |
+
{
|
| 230 |
+
"epoch": 0.1481893118458831,
|
| 231 |
+
"grad_norm": 15.651091575622559,
|
| 232 |
+
"learning_rate": 4.935185185185186e-06,
|
| 233 |
+
"loss": 1.1124,
|
| 234 |
+
"step": 1600
|
| 235 |
+
},
|
| 236 |
+
{
|
| 237 |
+
"epoch": 0.15282022784106697,
|
| 238 |
+
"grad_norm": 15.110349655151367,
|
| 239 |
+
"learning_rate": 5.089506172839507e-06,
|
| 240 |
+
"loss": 1.1964,
|
| 241 |
+
"step": 1650
|
| 242 |
+
},
|
| 243 |
+
{
|
| 244 |
+
"epoch": 0.1574511438362508,
|
| 245 |
+
"grad_norm": 14.926290512084961,
|
| 246 |
+
"learning_rate": 5.243827160493827e-06,
|
| 247 |
+
"loss": 1.0599,
|
| 248 |
+
"step": 1700
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 0.16208205983143464,
|
| 252 |
+
"grad_norm": 14.762713432312012,
|
| 253 |
+
"learning_rate": 5.398148148148149e-06,
|
| 254 |
+
"loss": 1.0663,
|
| 255 |
+
"step": 1750
|
| 256 |
+
},
|
| 257 |
+
{
|
| 258 |
+
"epoch": 0.1667129758266185,
|
| 259 |
+
"grad_norm": 13.759990692138672,
|
| 260 |
+
"learning_rate": 5.55246913580247e-06,
|
| 261 |
+
"loss": 1.0532,
|
| 262 |
+
"step": 1800
|
| 263 |
+
},
|
| 264 |
+
{
|
| 265 |
+
"epoch": 0.17134389182180235,
|
| 266 |
+
"grad_norm": 12.649763107299805,
|
| 267 |
+
"learning_rate": 5.7067901234567905e-06,
|
| 268 |
+
"loss": 1.1887,
|
| 269 |
+
"step": 1850
|
| 270 |
+
},
|
| 271 |
+
{
|
| 272 |
+
"epoch": 0.1759748078169862,
|
| 273 |
+
"grad_norm": 12.257176399230957,
|
| 274 |
+
"learning_rate": 5.861111111111112e-06,
|
| 275 |
+
"loss": 1.1441,
|
| 276 |
+
"step": 1900
|
| 277 |
+
},
|
| 278 |
+
{
|
| 279 |
+
"epoch": 0.18060572381217005,
|
| 280 |
+
"grad_norm": 13.522123336791992,
|
| 281 |
+
"learning_rate": 6.0154320987654326e-06,
|
| 282 |
+
"loss": 1.0646,
|
| 283 |
+
"step": 1950
|
| 284 |
+
},
|
| 285 |
+
{
|
| 286 |
+
"epoch": 0.1852366398073539,
|
| 287 |
+
"grad_norm": 13.428495407104492,
|
| 288 |
+
"learning_rate": 6.169753086419754e-06,
|
| 289 |
+
"loss": 1.0759,
|
| 290 |
+
"step": 2000
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 0.18986755580253775,
|
| 294 |
+
"grad_norm": 15.00213623046875,
|
| 295 |
+
"learning_rate": 6.324074074074075e-06,
|
| 296 |
+
"loss": 1.1082,
|
| 297 |
+
"step": 2050
|
| 298 |
+
},
|
| 299 |
+
{
|
| 300 |
+
"epoch": 0.1944984717977216,
|
| 301 |
+
"grad_norm": 14.586730003356934,
|
| 302 |
+
"learning_rate": 6.478395061728395e-06,
|
| 303 |
+
"loss": 1.1114,
|
| 304 |
+
"step": 2100
|
| 305 |
+
},
|
| 306 |
+
{
|
| 307 |
+
"epoch": 0.19912938779290543,
|
| 308 |
+
"grad_norm": 15.068018913269043,
|
| 309 |
+
"learning_rate": 6.632716049382717e-06,
|
| 310 |
+
"loss": 1.0588,
|
| 311 |
+
"step": 2150
|
| 312 |
+
},
|
| 313 |
+
{
|
| 314 |
+
"epoch": 0.2037603037880893,
|
| 315 |
+
"grad_norm": 11.068521499633789,
|
| 316 |
+
"learning_rate": 6.787037037037037e-06,
|
| 317 |
+
"loss": 0.9906,
|
| 318 |
+
"step": 2200
|
| 319 |
+
},
|
| 320 |
+
{
|
| 321 |
+
"epoch": 0.20839121978327313,
|
| 322 |
+
"grad_norm": 15.2577486038208,
|
| 323 |
+
"learning_rate": 6.941358024691358e-06,
|
| 324 |
+
"loss": 1.0344,
|
| 325 |
+
"step": 2250
|
| 326 |
+
},
|
| 327 |
+
{
|
| 328 |
+
"epoch": 0.21302213577845697,
|
| 329 |
+
"grad_norm": 15.963299751281738,
|
| 330 |
+
"learning_rate": 7.0956790123456794e-06,
|
| 331 |
+
"loss": 0.9656,
|
| 332 |
+
"step": 2300
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 0.21765305177364083,
|
| 336 |
+
"grad_norm": 12.144145011901855,
|
| 337 |
+
"learning_rate": 7.25e-06,
|
| 338 |
+
"loss": 0.9795,
|
| 339 |
+
"step": 2350
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 0.22228396776882467,
|
| 343 |
+
"grad_norm": 16.015628814697266,
|
| 344 |
+
"learning_rate": 7.404320987654322e-06,
|
| 345 |
+
"loss": 0.8968,
|
| 346 |
+
"step": 2400
|
| 347 |
+
},
|
| 348 |
+
{
|
| 349 |
+
"epoch": 0.2269148837640085,
|
| 350 |
+
"grad_norm": 13.311681747436523,
|
| 351 |
+
"learning_rate": 7.558641975308643e-06,
|
| 352 |
+
"loss": 0.9386,
|
| 353 |
+
"step": 2450
|
| 354 |
+
},
|
| 355 |
+
{
|
| 356 |
+
"epoch": 0.23154579975919237,
|
| 357 |
+
"grad_norm": 15.793661117553711,
|
| 358 |
+
"learning_rate": 7.712962962962964e-06,
|
| 359 |
+
"loss": 0.9716,
|
| 360 |
+
"step": 2500
|
| 361 |
+
},
|
| 362 |
+
{
|
| 363 |
+
"epoch": 0.2361767157543762,
|
| 364 |
+
"grad_norm": 9.333974838256836,
|
| 365 |
+
"learning_rate": 7.867283950617284e-06,
|
| 366 |
+
"loss": 0.9593,
|
| 367 |
+
"step": 2550
|
| 368 |
+
},
|
| 369 |
+
{
|
| 370 |
+
"epoch": 0.24080763174956007,
|
| 371 |
+
"grad_norm": 11.530397415161133,
|
| 372 |
+
"learning_rate": 8.021604938271605e-06,
|
| 373 |
+
"loss": 0.9795,
|
| 374 |
+
"step": 2600
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 0.2454385477447439,
|
| 378 |
+
"grad_norm": 12.683008193969727,
|
| 379 |
+
"learning_rate": 8.175925925925925e-06,
|
| 380 |
+
"loss": 1.0464,
|
| 381 |
+
"step": 2650
|
| 382 |
+
},
|
| 383 |
+
{
|
| 384 |
+
"epoch": 0.2500694637399278,
|
| 385 |
+
"grad_norm": 13.96157455444336,
|
| 386 |
+
"learning_rate": 8.330246913580248e-06,
|
| 387 |
+
"loss": 0.9384,
|
| 388 |
+
"step": 2700
|
| 389 |
+
},
|
| 390 |
+
{
|
| 391 |
+
"epoch": 0.2547003797351116,
|
| 392 |
+
"grad_norm": 15.856661796569824,
|
| 393 |
+
"learning_rate": 8.484567901234568e-06,
|
| 394 |
+
"loss": 0.9062,
|
| 395 |
+
"step": 2750
|
| 396 |
+
},
|
| 397 |
+
{
|
| 398 |
+
"epoch": 0.25933129573029545,
|
| 399 |
+
"grad_norm": 14.501742362976074,
|
| 400 |
+
"learning_rate": 8.63888888888889e-06,
|
| 401 |
+
"loss": 0.8873,
|
| 402 |
+
"step": 2800
|
| 403 |
+
},
|
| 404 |
+
{
|
| 405 |
+
"epoch": 0.2639622117254793,
|
| 406 |
+
"grad_norm": 11.515716552734375,
|
| 407 |
+
"learning_rate": 8.793209876543211e-06,
|
| 408 |
+
"loss": 1.0624,
|
| 409 |
+
"step": 2850
|
| 410 |
+
},
|
| 411 |
+
{
|
| 412 |
+
"epoch": 0.2685931277206631,
|
| 413 |
+
"grad_norm": 19.777027130126953,
|
| 414 |
+
"learning_rate": 8.947530864197532e-06,
|
| 415 |
+
"loss": 1.0515,
|
| 416 |
+
"step": 2900
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 0.273224043715847,
|
| 420 |
+
"grad_norm": 12.990806579589844,
|
| 421 |
+
"learning_rate": 9.101851851851853e-06,
|
| 422 |
+
"loss": 0.9574,
|
| 423 |
+
"step": 2950
|
| 424 |
+
},
|
| 425 |
+
{
|
| 426 |
+
"epoch": 0.27785495971103086,
|
| 427 |
+
"grad_norm": 135.59896850585938,
|
| 428 |
+
"learning_rate": 9.256172839506173e-06,
|
| 429 |
+
"loss": 0.9481,
|
| 430 |
+
"step": 3000
|
| 431 |
+
},
|
| 432 |
+
{
|
| 433 |
+
"epoch": 0.2824858757062147,
|
| 434 |
+
"grad_norm": 13.494345664978027,
|
| 435 |
+
"learning_rate": 9.410493827160495e-06,
|
| 436 |
+
"loss": 1.0147,
|
| 437 |
+
"step": 3050
|
| 438 |
+
},
|
| 439 |
+
{
|
| 440 |
+
"epoch": 0.28711679170139853,
|
| 441 |
+
"grad_norm": 15.373414993286133,
|
| 442 |
+
"learning_rate": 9.564814814814816e-06,
|
| 443 |
+
"loss": 0.9794,
|
| 444 |
+
"step": 3100
|
| 445 |
+
},
|
| 446 |
+
{
|
| 447 |
+
"epoch": 0.29174770769658237,
|
| 448 |
+
"grad_norm": 17.2181396484375,
|
| 449 |
+
"learning_rate": 9.719135802469137e-06,
|
| 450 |
+
"loss": 0.925,
|
| 451 |
+
"step": 3150
|
| 452 |
+
},
|
| 453 |
+
{
|
| 454 |
+
"epoch": 0.2963786236917662,
|
| 455 |
+
"grad_norm": 15.999231338500977,
|
| 456 |
+
"learning_rate": 9.873456790123457e-06,
|
| 457 |
+
"loss": 0.8622,
|
| 458 |
+
"step": 3200
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 0.3010095396869501,
|
| 462 |
+
"grad_norm": 12.339484214782715,
|
| 463 |
+
"learning_rate": 9.996912627354123e-06,
|
| 464 |
+
"loss": 0.8761,
|
| 465 |
+
"step": 3250
|
| 466 |
+
},
|
| 467 |
+
{
|
| 468 |
+
"epoch": 0.30564045568213394,
|
| 469 |
+
"grad_norm": 12.637123107910156,
|
| 470 |
+
"learning_rate": 9.979760557099242e-06,
|
| 471 |
+
"loss": 0.9473,
|
| 472 |
+
"step": 3300
|
| 473 |
+
},
|
| 474 |
+
{
|
| 475 |
+
"epoch": 0.3102713716773178,
|
| 476 |
+
"grad_norm": 11.328486442565918,
|
| 477 |
+
"learning_rate": 9.962608486844364e-06,
|
| 478 |
+
"loss": 0.9392,
|
| 479 |
+
"step": 3350
|
| 480 |
+
},
|
| 481 |
+
{
|
| 482 |
+
"epoch": 0.3149022876725016,
|
| 483 |
+
"grad_norm": 16.28173828125,
|
| 484 |
+
"learning_rate": 9.945456416589483e-06,
|
| 485 |
+
"loss": 0.871,
|
| 486 |
+
"step": 3400
|
| 487 |
+
},
|
| 488 |
+
{
|
| 489 |
+
"epoch": 0.31953320366768545,
|
| 490 |
+
"grad_norm": 12.468025207519531,
|
| 491 |
+
"learning_rate": 9.928304346334603e-06,
|
| 492 |
+
"loss": 0.8408,
|
| 493 |
+
"step": 3450
|
| 494 |
+
},
|
| 495 |
+
{
|
| 496 |
+
"epoch": 0.3241641196628693,
|
| 497 |
+
"grad_norm": 8.507662773132324,
|
| 498 |
+
"learning_rate": 9.911152276079725e-06,
|
| 499 |
+
"loss": 0.8601,
|
| 500 |
+
"step": 3500
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 0.3287950356580532,
|
| 504 |
+
"grad_norm": 13.56063461303711,
|
| 505 |
+
"learning_rate": 9.894000205824844e-06,
|
| 506 |
+
"loss": 0.8586,
|
| 507 |
+
"step": 3550
|
| 508 |
+
},
|
| 509 |
+
{
|
| 510 |
+
"epoch": 0.333425951653237,
|
| 511 |
+
"grad_norm": 13.81701374053955,
|
| 512 |
+
"learning_rate": 9.876848135569964e-06,
|
| 513 |
+
"loss": 0.9191,
|
| 514 |
+
"step": 3600
|
| 515 |
+
},
|
| 516 |
+
{
|
| 517 |
+
"epoch": 0.33805686764842086,
|
| 518 |
+
"grad_norm": 14.561142921447754,
|
| 519 |
+
"learning_rate": 9.859696065315084e-06,
|
| 520 |
+
"loss": 0.9144,
|
| 521 |
+
"step": 3650
|
| 522 |
+
},
|
| 523 |
+
{
|
| 524 |
+
"epoch": 0.3426877836436047,
|
| 525 |
+
"grad_norm": 12.986917495727539,
|
| 526 |
+
"learning_rate": 9.842543995060204e-06,
|
| 527 |
+
"loss": 0.8209,
|
| 528 |
+
"step": 3700
|
| 529 |
+
},
|
| 530 |
+
{
|
| 531 |
+
"epoch": 0.34731869963878853,
|
| 532 |
+
"grad_norm": 13.493863105773926,
|
| 533 |
+
"learning_rate": 9.825391924805325e-06,
|
| 534 |
+
"loss": 0.9236,
|
| 535 |
+
"step": 3750
|
| 536 |
+
},
|
| 537 |
+
{
|
| 538 |
+
"epoch": 0.3519496156339724,
|
| 539 |
+
"grad_norm": 14.126349449157715,
|
| 540 |
+
"learning_rate": 9.808239854550445e-06,
|
| 541 |
+
"loss": 0.8461,
|
| 542 |
+
"step": 3800
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 0.35658053162915626,
|
| 546 |
+
"grad_norm": 10.988364219665527,
|
| 547 |
+
"learning_rate": 9.791087784295566e-06,
|
| 548 |
+
"loss": 0.8421,
|
| 549 |
+
"step": 3850
|
| 550 |
+
},
|
| 551 |
+
{
|
| 552 |
+
"epoch": 0.3612114476243401,
|
| 553 |
+
"grad_norm": 11.538872718811035,
|
| 554 |
+
"learning_rate": 9.773935714040686e-06,
|
| 555 |
+
"loss": 0.7806,
|
| 556 |
+
"step": 3900
|
| 557 |
+
},
|
| 558 |
+
{
|
| 559 |
+
"epoch": 0.36584236361952394,
|
| 560 |
+
"grad_norm": 19.18027687072754,
|
| 561 |
+
"learning_rate": 9.756783643785806e-06,
|
| 562 |
+
"loss": 0.8667,
|
| 563 |
+
"step": 3950
|
| 564 |
+
},
|
| 565 |
+
{
|
| 566 |
+
"epoch": 0.3704732796147078,
|
| 567 |
+
"grad_norm": 15.992369651794434,
|
| 568 |
+
"learning_rate": 9.739631573530925e-06,
|
| 569 |
+
"loss": 0.8681,
|
| 570 |
+
"step": 4000
|
| 571 |
+
},
|
| 572 |
+
{
|
| 573 |
+
"epoch": 0.3751041956098916,
|
| 574 |
+
"grad_norm": 9.232007026672363,
|
| 575 |
+
"learning_rate": 9.722479503276047e-06,
|
| 576 |
+
"loss": 0.8877,
|
| 577 |
+
"step": 4050
|
| 578 |
+
},
|
| 579 |
+
{
|
| 580 |
+
"epoch": 0.3797351116050755,
|
| 581 |
+
"grad_norm": 13.309873580932617,
|
| 582 |
+
"learning_rate": 9.705327433021166e-06,
|
| 583 |
+
"loss": 0.8167,
|
| 584 |
+
"step": 4100
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 0.38436602760025934,
|
| 588 |
+
"grad_norm": 17.893497467041016,
|
| 589 |
+
"learning_rate": 9.688175362766286e-06,
|
| 590 |
+
"loss": 0.8315,
|
| 591 |
+
"step": 4150
|
| 592 |
+
},
|
| 593 |
+
{
|
| 594 |
+
"epoch": 0.3889969435954432,
|
| 595 |
+
"grad_norm": 16.278261184692383,
|
| 596 |
+
"learning_rate": 9.671023292511406e-06,
|
| 597 |
+
"loss": 0.8194,
|
| 598 |
+
"step": 4200
|
| 599 |
+
},
|
| 600 |
+
{
|
| 601 |
+
"epoch": 0.393627859590627,
|
| 602 |
+
"grad_norm": 18.51329803466797,
|
| 603 |
+
"learning_rate": 9.653871222256527e-06,
|
| 604 |
+
"loss": 0.8781,
|
| 605 |
+
"step": 4250
|
| 606 |
+
},
|
| 607 |
+
{
|
| 608 |
+
"epoch": 0.39825877558581085,
|
| 609 |
+
"grad_norm": 13.017523765563965,
|
| 610 |
+
"learning_rate": 9.636719152001647e-06,
|
| 611 |
+
"loss": 0.8474,
|
| 612 |
+
"step": 4300
|
| 613 |
+
},
|
| 614 |
+
{
|
| 615 |
+
"epoch": 0.40288969158099475,
|
| 616 |
+
"grad_norm": 13.606826782226562,
|
| 617 |
+
"learning_rate": 9.619567081746768e-06,
|
| 618 |
+
"loss": 0.886,
|
| 619 |
+
"step": 4350
|
| 620 |
+
},
|
| 621 |
+
{
|
| 622 |
+
"epoch": 0.4075206075761786,
|
| 623 |
+
"grad_norm": 15.338058471679688,
|
| 624 |
+
"learning_rate": 9.602415011491888e-06,
|
| 625 |
+
"loss": 0.8296,
|
| 626 |
+
"step": 4400
|
| 627 |
+
},
|
| 628 |
+
{
|
| 629 |
+
"epoch": 0.4121515235713624,
|
| 630 |
+
"grad_norm": 15.04211711883545,
|
| 631 |
+
"learning_rate": 9.585262941237008e-06,
|
| 632 |
+
"loss": 0.8103,
|
| 633 |
+
"step": 4450
|
| 634 |
+
},
|
| 635 |
+
{
|
| 636 |
+
"epoch": 0.41678243956654626,
|
| 637 |
+
"grad_norm": 22.55345344543457,
|
| 638 |
+
"learning_rate": 9.568110870982128e-06,
|
| 639 |
+
"loss": 0.8835,
|
| 640 |
+
"step": 4500
|
| 641 |
+
},
|
| 642 |
+
{
|
| 643 |
+
"epoch": 0.4214133555617301,
|
| 644 |
+
"grad_norm": 17.12868881225586,
|
| 645 |
+
"learning_rate": 9.550958800727247e-06,
|
| 646 |
+
"loss": 0.8298,
|
| 647 |
+
"step": 4550
|
| 648 |
+
},
|
| 649 |
+
{
|
| 650 |
+
"epoch": 0.42604427155691393,
|
| 651 |
+
"grad_norm": 12.845903396606445,
|
| 652 |
+
"learning_rate": 9.533806730472369e-06,
|
| 653 |
+
"loss": 0.8916,
|
| 654 |
+
"step": 4600
|
| 655 |
+
},
|
| 656 |
+
{
|
| 657 |
+
"epoch": 0.4306751875520978,
|
| 658 |
+
"grad_norm": 11.995304107666016,
|
| 659 |
+
"learning_rate": 9.516654660217489e-06,
|
| 660 |
+
"loss": 0.8882,
|
| 661 |
+
"step": 4650
|
| 662 |
+
},
|
| 663 |
+
{
|
| 664 |
+
"epoch": 0.43530610354728166,
|
| 665 |
+
"grad_norm": 15.846125602722168,
|
| 666 |
+
"learning_rate": 9.49950258996261e-06,
|
| 667 |
+
"loss": 0.8282,
|
| 668 |
+
"step": 4700
|
| 669 |
+
},
|
| 670 |
+
{
|
| 671 |
+
"epoch": 0.4399370195424655,
|
| 672 |
+
"grad_norm": 10.837459564208984,
|
| 673 |
+
"learning_rate": 9.48235051970773e-06,
|
| 674 |
+
"loss": 0.8809,
|
| 675 |
+
"step": 4750
|
| 676 |
+
},
|
| 677 |
+
{
|
| 678 |
+
"epoch": 0.44456793553764934,
|
| 679 |
+
"grad_norm": 12.318596839904785,
|
| 680 |
+
"learning_rate": 9.46519844945285e-06,
|
| 681 |
+
"loss": 0.8353,
|
| 682 |
+
"step": 4800
|
| 683 |
+
},
|
| 684 |
+
{
|
| 685 |
+
"epoch": 0.4491988515328332,
|
| 686 |
+
"grad_norm": 9.558602333068848,
|
| 687 |
+
"learning_rate": 9.448046379197971e-06,
|
| 688 |
+
"loss": 0.8337,
|
| 689 |
+
"step": 4850
|
| 690 |
+
},
|
| 691 |
+
{
|
| 692 |
+
"epoch": 0.453829767528017,
|
| 693 |
+
"grad_norm": 19.20343780517578,
|
| 694 |
+
"learning_rate": 9.43089430894309e-06,
|
| 695 |
+
"loss": 0.8147,
|
| 696 |
+
"step": 4900
|
| 697 |
+
},
|
| 698 |
+
{
|
| 699 |
+
"epoch": 0.4584606835232009,
|
| 700 |
+
"grad_norm": 17.662939071655273,
|
| 701 |
+
"learning_rate": 9.41374223868821e-06,
|
| 702 |
+
"loss": 0.827,
|
| 703 |
+
"step": 4950
|
| 704 |
+
},
|
| 705 |
+
{
|
| 706 |
+
"epoch": 0.46309159951838474,
|
| 707 |
+
"grad_norm": 13.364691734313965,
|
| 708 |
+
"learning_rate": 9.39659016843333e-06,
|
| 709 |
+
"loss": 0.863,
|
| 710 |
+
"step": 5000
|
| 711 |
+
},
|
| 712 |
+
{
|
| 713 |
+
"epoch": 0.4677225155135686,
|
| 714 |
+
"grad_norm": 13.197148323059082,
|
| 715 |
+
"learning_rate": 9.37943809817845e-06,
|
| 716 |
+
"loss": 0.7999,
|
| 717 |
+
"step": 5050
|
| 718 |
+
},
|
| 719 |
+
{
|
| 720 |
+
"epoch": 0.4723534315087524,
|
| 721 |
+
"grad_norm": 15.571122169494629,
|
| 722 |
+
"learning_rate": 9.362286027923571e-06,
|
| 723 |
+
"loss": 0.8459,
|
| 724 |
+
"step": 5100
|
| 725 |
+
},
|
| 726 |
+
{
|
| 727 |
+
"epoch": 0.47698434750393626,
|
| 728 |
+
"grad_norm": 17.007863998413086,
|
| 729 |
+
"learning_rate": 9.345133957668691e-06,
|
| 730 |
+
"loss": 0.8159,
|
| 731 |
+
"step": 5150
|
| 732 |
+
},
|
| 733 |
+
{
|
| 734 |
+
"epoch": 0.48161526349912015,
|
| 735 |
+
"grad_norm": 14.772674560546875,
|
| 736 |
+
"learning_rate": 9.327981887413812e-06,
|
| 737 |
+
"loss": 0.8754,
|
| 738 |
+
"step": 5200
|
| 739 |
+
},
|
| 740 |
+
{
|
| 741 |
+
"epoch": 0.486246179494304,
|
| 742 |
+
"grad_norm": 18.637845993041992,
|
| 743 |
+
"learning_rate": 9.310829817158932e-06,
|
| 744 |
+
"loss": 0.8453,
|
| 745 |
+
"step": 5250
|
| 746 |
+
},
|
| 747 |
+
{
|
| 748 |
+
"epoch": 0.4908770954894878,
|
| 749 |
+
"grad_norm": 13.34103775024414,
|
| 750 |
+
"learning_rate": 9.293677746904052e-06,
|
| 751 |
+
"loss": 0.8148,
|
| 752 |
+
"step": 5300
|
| 753 |
+
},
|
| 754 |
+
{
|
| 755 |
+
"epoch": 0.49550801148467166,
|
| 756 |
+
"grad_norm": 15.411863327026367,
|
| 757 |
+
"learning_rate": 9.276525676649172e-06,
|
| 758 |
+
"loss": 0.7435,
|
| 759 |
+
"step": 5350
|
| 760 |
+
},
|
| 761 |
+
{
|
| 762 |
+
"epoch": 0.5001389274798556,
|
| 763 |
+
"grad_norm": 11.534259796142578,
|
| 764 |
+
"learning_rate": 9.259373606394293e-06,
|
| 765 |
+
"loss": 0.7787,
|
| 766 |
+
"step": 5400
|
| 767 |
+
},
|
| 768 |
+
{
|
| 769 |
+
"epoch": 0.5047698434750394,
|
| 770 |
+
"grad_norm": 10.385927200317383,
|
| 771 |
+
"learning_rate": 9.242221536139413e-06,
|
| 772 |
+
"loss": 0.7536,
|
| 773 |
+
"step": 5450
|
| 774 |
+
},
|
| 775 |
+
{
|
| 776 |
+
"epoch": 0.5094007594702232,
|
| 777 |
+
"grad_norm": 11.001605987548828,
|
| 778 |
+
"learning_rate": 9.225069465884532e-06,
|
| 779 |
+
"loss": 0.7829,
|
| 780 |
+
"step": 5500
|
| 781 |
+
},
|
| 782 |
+
{
|
| 783 |
+
"epoch": 0.5140316754654071,
|
| 784 |
+
"grad_norm": 12.940703392028809,
|
| 785 |
+
"learning_rate": 9.207917395629654e-06,
|
| 786 |
+
"loss": 0.797,
|
| 787 |
+
"step": 5550
|
| 788 |
+
},
|
| 789 |
+
{
|
| 790 |
+
"epoch": 0.5186625914605909,
|
| 791 |
+
"grad_norm": 13.790589332580566,
|
| 792 |
+
"learning_rate": 9.190765325374774e-06,
|
| 793 |
+
"loss": 0.802,
|
| 794 |
+
"step": 5600
|
| 795 |
+
},
|
| 796 |
+
{
|
| 797 |
+
"epoch": 0.5232935074557747,
|
| 798 |
+
"grad_norm": 14.535255432128906,
|
| 799 |
+
"learning_rate": 9.173613255119895e-06,
|
| 800 |
+
"loss": 0.8199,
|
| 801 |
+
"step": 5650
|
| 802 |
+
},
|
| 803 |
+
{
|
| 804 |
+
"epoch": 0.5279244234509586,
|
| 805 |
+
"grad_norm": 8.654829025268555,
|
| 806 |
+
"learning_rate": 9.156461184865015e-06,
|
| 807 |
+
"loss": 0.8698,
|
| 808 |
+
"step": 5700
|
| 809 |
+
},
|
| 810 |
+
{
|
| 811 |
+
"epoch": 0.5325553394461424,
|
| 812 |
+
"grad_norm": 10.871634483337402,
|
| 813 |
+
"learning_rate": 9.139309114610134e-06,
|
| 814 |
+
"loss": 0.7877,
|
| 815 |
+
"step": 5750
|
| 816 |
+
},
|
| 817 |
+
{
|
| 818 |
+
"epoch": 0.5371862554413263,
|
| 819 |
+
"grad_norm": 16.741178512573242,
|
| 820 |
+
"learning_rate": 9.122157044355254e-06,
|
| 821 |
+
"loss": 0.7441,
|
| 822 |
+
"step": 5800
|
| 823 |
+
},
|
| 824 |
+
{
|
| 825 |
+
"epoch": 0.5418171714365101,
|
| 826 |
+
"grad_norm": 12.232110023498535,
|
| 827 |
+
"learning_rate": 9.105004974100374e-06,
|
| 828 |
+
"loss": 0.7152,
|
| 829 |
+
"step": 5850
|
| 830 |
+
},
|
| 831 |
+
{
|
| 832 |
+
"epoch": 0.546448087431694,
|
| 833 |
+
"grad_norm": 10.189510345458984,
|
| 834 |
+
"learning_rate": 9.087852903845494e-06,
|
| 835 |
+
"loss": 0.7917,
|
| 836 |
+
"step": 5900
|
| 837 |
+
},
|
| 838 |
+
{
|
| 839 |
+
"epoch": 0.5510790034268779,
|
| 840 |
+
"grad_norm": 13.428311347961426,
|
| 841 |
+
"learning_rate": 9.070700833590615e-06,
|
| 842 |
+
"loss": 0.7825,
|
| 843 |
+
"step": 5950
|
| 844 |
+
},
|
| 845 |
+
{
|
| 846 |
+
"epoch": 0.5557099194220617,
|
| 847 |
+
"grad_norm": 7.749415397644043,
|
| 848 |
+
"learning_rate": 9.053548763335735e-06,
|
| 849 |
+
"loss": 0.7085,
|
| 850 |
+
"step": 6000
|
| 851 |
+
},
|
| 852 |
+
{
|
| 853 |
+
"epoch": 0.5603408354172456,
|
| 854 |
+
"grad_norm": 12.291046142578125,
|
| 855 |
+
"learning_rate": 9.036396693080856e-06,
|
| 856 |
+
"loss": 0.7457,
|
| 857 |
+
"step": 6050
|
| 858 |
+
},
|
| 859 |
+
{
|
| 860 |
+
"epoch": 0.5649717514124294,
|
| 861 |
+
"grad_norm": 8.931249618530273,
|
| 862 |
+
"learning_rate": 9.019244622825976e-06,
|
| 863 |
+
"loss": 0.7621,
|
| 864 |
+
"step": 6100
|
| 865 |
+
},
|
| 866 |
+
{
|
| 867 |
+
"epoch": 0.5696026674076132,
|
| 868 |
+
"grad_norm": 14.065079689025879,
|
| 869 |
+
"learning_rate": 9.002092552571096e-06,
|
| 870 |
+
"loss": 0.7981,
|
| 871 |
+
"step": 6150
|
| 872 |
+
},
|
| 873 |
+
{
|
| 874 |
+
"epoch": 0.5742335834027971,
|
| 875 |
+
"grad_norm": 14.24923038482666,
|
| 876 |
+
"learning_rate": 8.984940482316217e-06,
|
| 877 |
+
"loss": 0.7599,
|
| 878 |
+
"step": 6200
|
| 879 |
+
},
|
| 880 |
+
{
|
| 881 |
+
"epoch": 0.5788644993979809,
|
| 882 |
+
"grad_norm": 11.286952018737793,
|
| 883 |
+
"learning_rate": 8.967788412061337e-06,
|
| 884 |
+
"loss": 0.764,
|
| 885 |
+
"step": 6250
|
| 886 |
+
},
|
| 887 |
+
{
|
| 888 |
+
"epoch": 0.5834954153931647,
|
| 889 |
+
"grad_norm": 8.349452018737793,
|
| 890 |
+
"learning_rate": 8.950636341806457e-06,
|
| 891 |
+
"loss": 0.7148,
|
| 892 |
+
"step": 6300
|
| 893 |
+
},
|
| 894 |
+
{
|
| 895 |
+
"epoch": 0.5881263313883486,
|
| 896 |
+
"grad_norm": 18.04071617126465,
|
| 897 |
+
"learning_rate": 8.933484271551576e-06,
|
| 898 |
+
"loss": 0.7638,
|
| 899 |
+
"step": 6350
|
| 900 |
+
},
|
| 901 |
+
{
|
| 902 |
+
"epoch": 0.5927572473835324,
|
| 903 |
+
"grad_norm": 12.750236511230469,
|
| 904 |
+
"learning_rate": 8.916332201296696e-06,
|
| 905 |
+
"loss": 0.7638,
|
| 906 |
+
"step": 6400
|
| 907 |
+
},
|
| 908 |
+
{
|
| 909 |
+
"epoch": 0.5973881633787164,
|
| 910 |
+
"grad_norm": 10.140752792358398,
|
| 911 |
+
"learning_rate": 8.899180131041817e-06,
|
| 912 |
+
"loss": 0.7904,
|
| 913 |
+
"step": 6450
|
| 914 |
+
},
|
| 915 |
+
{
|
| 916 |
+
"epoch": 0.6020190793739002,
|
| 917 |
+
"grad_norm": 15.348838806152344,
|
| 918 |
+
"learning_rate": 8.882028060786937e-06,
|
| 919 |
+
"loss": 0.8024,
|
| 920 |
+
"step": 6500
|
| 921 |
+
},
|
| 922 |
+
{
|
| 923 |
+
"epoch": 0.606649995369084,
|
| 924 |
+
"grad_norm": 12.342340469360352,
|
| 925 |
+
"learning_rate": 8.864875990532059e-06,
|
| 926 |
+
"loss": 0.8224,
|
| 927 |
+
"step": 6550
|
| 928 |
+
},
|
| 929 |
+
{
|
| 930 |
+
"epoch": 0.6112809113642679,
|
| 931 |
+
"grad_norm": 13.80968952178955,
|
| 932 |
+
"learning_rate": 8.847723920277178e-06,
|
| 933 |
+
"loss": 0.7567,
|
| 934 |
+
"step": 6600
|
| 935 |
+
},
|
| 936 |
+
{
|
| 937 |
+
"epoch": 0.6159118273594517,
|
| 938 |
+
"grad_norm": 13.79007625579834,
|
| 939 |
+
"learning_rate": 8.830571850022298e-06,
|
| 940 |
+
"loss": 0.8385,
|
| 941 |
+
"step": 6650
|
| 942 |
+
},
|
| 943 |
+
{
|
| 944 |
+
"epoch": 0.6205427433546356,
|
| 945 |
+
"grad_norm": 12.956084251403809,
|
| 946 |
+
"learning_rate": 8.813419779767418e-06,
|
| 947 |
+
"loss": 0.7652,
|
| 948 |
+
"step": 6700
|
| 949 |
+
},
|
| 950 |
+
{
|
| 951 |
+
"epoch": 0.6251736593498194,
|
| 952 |
+
"grad_norm": 17.817737579345703,
|
| 953 |
+
"learning_rate": 8.79626770951254e-06,
|
| 954 |
+
"loss": 0.7567,
|
| 955 |
+
"step": 6750
|
| 956 |
+
},
|
| 957 |
+
{
|
| 958 |
+
"epoch": 0.6298045753450032,
|
| 959 |
+
"grad_norm": 12.05063247680664,
|
| 960 |
+
"learning_rate": 8.779115639257659e-06,
|
| 961 |
+
"loss": 0.7012,
|
| 962 |
+
"step": 6800
|
| 963 |
+
},
|
| 964 |
+
{
|
| 965 |
+
"epoch": 0.6344354913401871,
|
| 966 |
+
"grad_norm": 7.249739646911621,
|
| 967 |
+
"learning_rate": 8.761963569002779e-06,
|
| 968 |
+
"loss": 0.7995,
|
| 969 |
+
"step": 6850
|
| 970 |
+
},
|
| 971 |
+
{
|
| 972 |
+
"epoch": 0.6390664073353709,
|
| 973 |
+
"grad_norm": 14.997950553894043,
|
| 974 |
+
"learning_rate": 8.7448114987479e-06,
|
| 975 |
+
"loss": 0.7259,
|
| 976 |
+
"step": 6900
|
| 977 |
+
},
|
| 978 |
+
{
|
| 979 |
+
"epoch": 0.6436973233305547,
|
| 980 |
+
"grad_norm": 4.541720390319824,
|
| 981 |
+
"learning_rate": 8.72765942849302e-06,
|
| 982 |
+
"loss": 0.7801,
|
| 983 |
+
"step": 6950
|
| 984 |
+
},
|
| 985 |
+
{
|
| 986 |
+
"epoch": 0.6483282393257386,
|
| 987 |
+
"grad_norm": 11.39427661895752,
|
| 988 |
+
"learning_rate": 8.710507358238141e-06,
|
| 989 |
+
"loss": 0.7132,
|
| 990 |
+
"step": 7000
|
| 991 |
+
},
|
| 992 |
+
{
|
| 993 |
+
"epoch": 0.6529591553209225,
|
| 994 |
+
"grad_norm": 12.994211196899414,
|
| 995 |
+
"learning_rate": 8.693355287983261e-06,
|
| 996 |
+
"loss": 0.7861,
|
| 997 |
+
"step": 7050
|
| 998 |
+
},
|
| 999 |
+
{
|
| 1000 |
+
"epoch": 0.6575900713161064,
|
| 1001 |
+
"grad_norm": 13.809148788452148,
|
| 1002 |
+
"learning_rate": 8.67620321772838e-06,
|
| 1003 |
+
"loss": 0.7418,
|
| 1004 |
+
"step": 7100
|
| 1005 |
+
},
|
| 1006 |
+
{
|
| 1007 |
+
"epoch": 0.6622209873112902,
|
| 1008 |
+
"grad_norm": 14.191105842590332,
|
| 1009 |
+
"learning_rate": 8.6590511474735e-06,
|
| 1010 |
+
"loss": 0.7191,
|
| 1011 |
+
"step": 7150
|
| 1012 |
+
},
|
| 1013 |
+
{
|
| 1014 |
+
"epoch": 0.666851903306474,
|
| 1015 |
+
"grad_norm": 13.280489921569824,
|
| 1016 |
+
"learning_rate": 8.64189907721862e-06,
|
| 1017 |
+
"loss": 0.7834,
|
| 1018 |
+
"step": 7200
|
| 1019 |
+
},
|
| 1020 |
+
{
|
| 1021 |
+
"epoch": 0.6714828193016579,
|
| 1022 |
+
"grad_norm": 12.677234649658203,
|
| 1023 |
+
"learning_rate": 8.62474700696374e-06,
|
| 1024 |
+
"loss": 0.752,
|
| 1025 |
+
"step": 7250
|
| 1026 |
+
},
|
| 1027 |
+
{
|
| 1028 |
+
"epoch": 0.6761137352968417,
|
| 1029 |
+
"grad_norm": 17.168846130371094,
|
| 1030 |
+
"learning_rate": 8.607594936708861e-06,
|
| 1031 |
+
"loss": 0.7715,
|
| 1032 |
+
"step": 7300
|
| 1033 |
+
},
|
| 1034 |
+
{
|
| 1035 |
+
"epoch": 0.6807446512920255,
|
| 1036 |
+
"grad_norm": 16.01129722595215,
|
| 1037 |
+
"learning_rate": 8.590442866453981e-06,
|
| 1038 |
+
"loss": 0.7613,
|
| 1039 |
+
"step": 7350
|
| 1040 |
+
},
|
| 1041 |
+
{
|
| 1042 |
+
"epoch": 0.6853755672872094,
|
| 1043 |
+
"grad_norm": 13.499317169189453,
|
| 1044 |
+
"learning_rate": 8.573290796199103e-06,
|
| 1045 |
+
"loss": 0.7688,
|
| 1046 |
+
"step": 7400
|
| 1047 |
+
},
|
| 1048 |
+
{
|
| 1049 |
+
"epoch": 0.6900064832823932,
|
| 1050 |
+
"grad_norm": 12.428091049194336,
|
| 1051 |
+
"learning_rate": 8.556138725944222e-06,
|
| 1052 |
+
"loss": 0.6496,
|
| 1053 |
+
"step": 7450
|
| 1054 |
+
},
|
| 1055 |
+
{
|
| 1056 |
+
"epoch": 0.6946373992775771,
|
| 1057 |
+
"grad_norm": 13.319454193115234,
|
| 1058 |
+
"learning_rate": 8.538986655689342e-06,
|
| 1059 |
+
"loss": 0.7926,
|
| 1060 |
+
"step": 7500
|
| 1061 |
+
},
|
| 1062 |
+
{
|
| 1063 |
+
"epoch": 0.6992683152727609,
|
| 1064 |
+
"grad_norm": 16.478046417236328,
|
| 1065 |
+
"learning_rate": 8.521834585434463e-06,
|
| 1066 |
+
"loss": 0.775,
|
| 1067 |
+
"step": 7550
|
| 1068 |
+
},
|
| 1069 |
+
{
|
| 1070 |
+
"epoch": 0.7038992312679448,
|
| 1071 |
+
"grad_norm": 17.74137306213379,
|
| 1072 |
+
"learning_rate": 8.504682515179583e-06,
|
| 1073 |
+
"loss": 0.6816,
|
| 1074 |
+
"step": 7600
|
| 1075 |
+
},
|
| 1076 |
+
{
|
| 1077 |
+
"epoch": 0.7085301472631287,
|
| 1078 |
+
"grad_norm": 11.713212966918945,
|
| 1079 |
+
"learning_rate": 8.487530444924703e-06,
|
| 1080 |
+
"loss": 0.8058,
|
| 1081 |
+
"step": 7650
|
| 1082 |
+
},
|
| 1083 |
+
{
|
| 1084 |
+
"epoch": 0.7131610632583125,
|
| 1085 |
+
"grad_norm": 7.7651872634887695,
|
| 1086 |
+
"learning_rate": 8.470378374669823e-06,
|
| 1087 |
+
"loss": 0.7829,
|
| 1088 |
+
"step": 7700
|
| 1089 |
+
},
|
| 1090 |
+
{
|
| 1091 |
+
"epoch": 0.7177919792534964,
|
| 1092 |
+
"grad_norm": 16.14422035217285,
|
| 1093 |
+
"learning_rate": 8.453226304414944e-06,
|
| 1094 |
+
"loss": 0.7515,
|
| 1095 |
+
"step": 7750
|
| 1096 |
+
},
|
| 1097 |
+
{
|
| 1098 |
+
"epoch": 0.7224228952486802,
|
| 1099 |
+
"grad_norm": 11.298089981079102,
|
| 1100 |
+
"learning_rate": 8.436074234160064e-06,
|
| 1101 |
+
"loss": 0.7801,
|
| 1102 |
+
"step": 7800
|
| 1103 |
+
},
|
| 1104 |
+
{
|
| 1105 |
+
"epoch": 0.727053811243864,
|
| 1106 |
+
"grad_norm": 13.805248260498047,
|
| 1107 |
+
"learning_rate": 8.418922163905185e-06,
|
| 1108 |
+
"loss": 0.7604,
|
| 1109 |
+
"step": 7850
|
| 1110 |
+
},
|
| 1111 |
+
{
|
| 1112 |
+
"epoch": 0.7316847272390479,
|
| 1113 |
+
"grad_norm": 14.030163764953613,
|
| 1114 |
+
"learning_rate": 8.401770093650305e-06,
|
| 1115 |
+
"loss": 0.8276,
|
| 1116 |
+
"step": 7900
|
| 1117 |
+
},
|
| 1118 |
+
{
|
| 1119 |
+
"epoch": 0.7363156432342317,
|
| 1120 |
+
"grad_norm": 15.09167766571045,
|
| 1121 |
+
"learning_rate": 8.384618023395425e-06,
|
| 1122 |
+
"loss": 0.7044,
|
| 1123 |
+
"step": 7950
|
| 1124 |
+
},
|
| 1125 |
+
{
|
| 1126 |
+
"epoch": 0.7409465592294155,
|
| 1127 |
+
"grad_norm": 11.88632583618164,
|
| 1128 |
+
"learning_rate": 8.367465953140544e-06,
|
| 1129 |
+
"loss": 0.6987,
|
| 1130 |
+
"step": 8000
|
| 1131 |
+
},
|
| 1132 |
+
{
|
| 1133 |
+
"epoch": 0.7455774752245994,
|
| 1134 |
+
"grad_norm": 8.257684707641602,
|
| 1135 |
+
"learning_rate": 8.350313882885664e-06,
|
| 1136 |
+
"loss": 0.7309,
|
| 1137 |
+
"step": 8050
|
| 1138 |
+
},
|
| 1139 |
+
{
|
| 1140 |
+
"epoch": 0.7502083912197832,
|
| 1141 |
+
"grad_norm": 16.112545013427734,
|
| 1142 |
+
"learning_rate": 8.333161812630786e-06,
|
| 1143 |
+
"loss": 0.7171,
|
| 1144 |
+
"step": 8100
|
| 1145 |
+
},
|
| 1146 |
+
{
|
| 1147 |
+
"epoch": 0.7548393072149672,
|
| 1148 |
+
"grad_norm": 16.633085250854492,
|
| 1149 |
+
"learning_rate": 8.316009742375905e-06,
|
| 1150 |
+
"loss": 0.7689,
|
| 1151 |
+
"step": 8150
|
| 1152 |
+
},
|
| 1153 |
+
{
|
| 1154 |
+
"epoch": 0.759470223210151,
|
| 1155 |
+
"grad_norm": 15.874983787536621,
|
| 1156 |
+
"learning_rate": 8.298857672121025e-06,
|
| 1157 |
+
"loss": 0.7185,
|
| 1158 |
+
"step": 8200
|
| 1159 |
+
},
|
| 1160 |
+
{
|
| 1161 |
+
"epoch": 0.7641011392053348,
|
| 1162 |
+
"grad_norm": 14.1043119430542,
|
| 1163 |
+
"learning_rate": 8.281705601866146e-06,
|
| 1164 |
+
"loss": 0.7266,
|
| 1165 |
+
"step": 8250
|
| 1166 |
+
},
|
| 1167 |
+
{
|
| 1168 |
+
"epoch": 0.7687320552005187,
|
| 1169 |
+
"grad_norm": 9.457536697387695,
|
| 1170 |
+
"learning_rate": 8.264553531611266e-06,
|
| 1171 |
+
"loss": 0.8332,
|
| 1172 |
+
"step": 8300
|
| 1173 |
+
},
|
| 1174 |
+
{
|
| 1175 |
+
"epoch": 0.7733629711957025,
|
| 1176 |
+
"grad_norm": 13.245748519897461,
|
| 1177 |
+
"learning_rate": 8.247401461356388e-06,
|
| 1178 |
+
"loss": 0.7343,
|
| 1179 |
+
"step": 8350
|
| 1180 |
+
},
|
| 1181 |
+
{
|
| 1182 |
+
"epoch": 0.7779938871908864,
|
| 1183 |
+
"grad_norm": 15.276327133178711,
|
| 1184 |
+
"learning_rate": 8.230249391101507e-06,
|
| 1185 |
+
"loss": 0.7036,
|
| 1186 |
+
"step": 8400
|
| 1187 |
+
},
|
| 1188 |
+
{
|
| 1189 |
+
"epoch": 0.7826248031860702,
|
| 1190 |
+
"grad_norm": 7.606396198272705,
|
| 1191 |
+
"learning_rate": 8.213097320846627e-06,
|
| 1192 |
+
"loss": 0.6989,
|
| 1193 |
+
"step": 8450
|
| 1194 |
+
},
|
| 1195 |
+
{
|
| 1196 |
+
"epoch": 0.787255719181254,
|
| 1197 |
+
"grad_norm": 11.166284561157227,
|
| 1198 |
+
"learning_rate": 8.195945250591747e-06,
|
| 1199 |
+
"loss": 0.6699,
|
| 1200 |
+
"step": 8500
|
| 1201 |
+
},
|
| 1202 |
+
{
|
| 1203 |
+
"epoch": 0.7918866351764379,
|
| 1204 |
+
"grad_norm": 13.374491691589355,
|
| 1205 |
+
"learning_rate": 8.178793180336866e-06,
|
| 1206 |
+
"loss": 0.7964,
|
| 1207 |
+
"step": 8550
|
| 1208 |
+
},
|
| 1209 |
+
{
|
| 1210 |
+
"epoch": 0.7965175511716217,
|
| 1211 |
+
"grad_norm": 9.178342819213867,
|
| 1212 |
+
"learning_rate": 8.161641110081988e-06,
|
| 1213 |
+
"loss": 0.7728,
|
| 1214 |
+
"step": 8600
|
| 1215 |
+
},
|
| 1216 |
+
{
|
| 1217 |
+
"epoch": 0.8011484671668055,
|
| 1218 |
+
"grad_norm": 13.058328628540039,
|
| 1219 |
+
"learning_rate": 8.144489039827108e-06,
|
| 1220 |
+
"loss": 0.778,
|
| 1221 |
+
"step": 8650
|
| 1222 |
+
},
|
| 1223 |
+
{
|
| 1224 |
+
"epoch": 0.8057793831619895,
|
| 1225 |
+
"grad_norm": 13.078104019165039,
|
| 1226 |
+
"learning_rate": 8.127336969572227e-06,
|
| 1227 |
+
"loss": 0.8046,
|
| 1228 |
+
"step": 8700
|
| 1229 |
+
},
|
| 1230 |
+
{
|
| 1231 |
+
"epoch": 0.8104102991571733,
|
| 1232 |
+
"grad_norm": 8.841740608215332,
|
| 1233 |
+
"learning_rate": 8.110184899317349e-06,
|
| 1234 |
+
"loss": 0.7621,
|
| 1235 |
+
"step": 8750
|
| 1236 |
+
},
|
| 1237 |
+
{
|
| 1238 |
+
"epoch": 0.8150412151523572,
|
| 1239 |
+
"grad_norm": 10.117661476135254,
|
| 1240 |
+
"learning_rate": 8.093032829062469e-06,
|
| 1241 |
+
"loss": 0.6871,
|
| 1242 |
+
"step": 8800
|
| 1243 |
+
},
|
| 1244 |
+
{
|
| 1245 |
+
"epoch": 0.819672131147541,
|
| 1246 |
+
"grad_norm": 13.581981658935547,
|
| 1247 |
+
"learning_rate": 8.075880758807588e-06,
|
| 1248 |
+
"loss": 0.7598,
|
| 1249 |
+
"step": 8850
|
| 1250 |
+
},
|
| 1251 |
+
{
|
| 1252 |
+
"epoch": 0.8243030471427248,
|
| 1253 |
+
"grad_norm": 18.15890884399414,
|
| 1254 |
+
"learning_rate": 8.05872868855271e-06,
|
| 1255 |
+
"loss": 0.7241,
|
| 1256 |
+
"step": 8900
|
| 1257 |
+
},
|
| 1258 |
+
{
|
| 1259 |
+
"epoch": 0.8289339631379087,
|
| 1260 |
+
"grad_norm": 10.792105674743652,
|
| 1261 |
+
"learning_rate": 8.04157661829783e-06,
|
| 1262 |
+
"loss": 0.7827,
|
| 1263 |
+
"step": 8950
|
| 1264 |
+
},
|
| 1265 |
+
{
|
| 1266 |
+
"epoch": 0.8335648791330925,
|
| 1267 |
+
"grad_norm": 11.7811861038208,
|
| 1268 |
+
"learning_rate": 8.024424548042949e-06,
|
| 1269 |
+
"loss": 0.6763,
|
| 1270 |
+
"step": 9000
|
| 1271 |
+
},
|
| 1272 |
+
{
|
| 1273 |
+
"epoch": 0.8381957951282764,
|
| 1274 |
+
"grad_norm": 14.736871719360352,
|
| 1275 |
+
"learning_rate": 8.007272477788069e-06,
|
| 1276 |
+
"loss": 0.6663,
|
| 1277 |
+
"step": 9050
|
| 1278 |
+
},
|
| 1279 |
+
{
|
| 1280 |
+
"epoch": 0.8428267111234602,
|
| 1281 |
+
"grad_norm": 8.909809112548828,
|
| 1282 |
+
"learning_rate": 7.99012040753319e-06,
|
| 1283 |
+
"loss": 0.7726,
|
| 1284 |
+
"step": 9100
|
| 1285 |
+
},
|
| 1286 |
+
{
|
| 1287 |
+
"epoch": 0.847457627118644,
|
| 1288 |
+
"grad_norm": 11.672173500061035,
|
| 1289 |
+
"learning_rate": 7.97296833727831e-06,
|
| 1290 |
+
"loss": 0.7877,
|
| 1291 |
+
"step": 9150
|
| 1292 |
+
},
|
| 1293 |
+
{
|
| 1294 |
+
"epoch": 0.8520885431138279,
|
| 1295 |
+
"grad_norm": 17.882083892822266,
|
| 1296 |
+
"learning_rate": 7.955816267023431e-06,
|
| 1297 |
+
"loss": 0.7163,
|
| 1298 |
+
"step": 9200
|
| 1299 |
+
},
|
| 1300 |
+
{
|
| 1301 |
+
"epoch": 0.8567194591090118,
|
| 1302 |
+
"grad_norm": 10.789363861083984,
|
| 1303 |
+
"learning_rate": 7.938664196768551e-06,
|
| 1304 |
+
"loss": 0.6821,
|
| 1305 |
+
"step": 9250
|
| 1306 |
+
},
|
| 1307 |
+
{
|
| 1308 |
+
"epoch": 0.8613503751041957,
|
| 1309 |
+
"grad_norm": 9.205415725708008,
|
| 1310 |
+
"learning_rate": 7.921512126513671e-06,
|
| 1311 |
+
"loss": 0.7975,
|
| 1312 |
+
"step": 9300
|
| 1313 |
+
},
|
| 1314 |
+
{
|
| 1315 |
+
"epoch": 0.8659812910993795,
|
| 1316 |
+
"grad_norm": 10.62431812286377,
|
| 1317 |
+
"learning_rate": 7.90436005625879e-06,
|
| 1318 |
+
"loss": 0.6893,
|
| 1319 |
+
"step": 9350
|
| 1320 |
+
},
|
| 1321 |
+
{
|
| 1322 |
+
"epoch": 0.8706122070945633,
|
| 1323 |
+
"grad_norm": 7.258960247039795,
|
| 1324 |
+
"learning_rate": 7.88720798600391e-06,
|
| 1325 |
+
"loss": 0.8111,
|
| 1326 |
+
"step": 9400
|
| 1327 |
+
},
|
| 1328 |
+
{
|
| 1329 |
+
"epoch": 0.8752431230897472,
|
| 1330 |
+
"grad_norm": 16.368818283081055,
|
| 1331 |
+
"learning_rate": 7.870055915749032e-06,
|
| 1332 |
+
"loss": 0.7149,
|
| 1333 |
+
"step": 9450
|
| 1334 |
+
},
|
| 1335 |
+
{
|
| 1336 |
+
"epoch": 0.879874039084931,
|
| 1337 |
+
"grad_norm": 7.597476482391357,
|
| 1338 |
+
"learning_rate": 7.852903845494151e-06,
|
| 1339 |
+
"loss": 0.6776,
|
| 1340 |
+
"step": 9500
|
| 1341 |
+
},
|
| 1342 |
+
{
|
| 1343 |
+
"epoch": 0.8845049550801148,
|
| 1344 |
+
"grad_norm": 14.393339157104492,
|
| 1345 |
+
"learning_rate": 7.835751775239271e-06,
|
| 1346 |
+
"loss": 0.6311,
|
| 1347 |
+
"step": 9550
|
| 1348 |
+
},
|
| 1349 |
+
{
|
| 1350 |
+
"epoch": 0.8891358710752987,
|
| 1351 |
+
"grad_norm": 11.397239685058594,
|
| 1352 |
+
"learning_rate": 7.818599704984393e-06,
|
| 1353 |
+
"loss": 0.7019,
|
| 1354 |
+
"step": 9600
|
| 1355 |
+
},
|
| 1356 |
+
{
|
| 1357 |
+
"epoch": 0.8937667870704825,
|
| 1358 |
+
"grad_norm": 8.419168472290039,
|
| 1359 |
+
"learning_rate": 7.801447634729512e-06,
|
| 1360 |
+
"loss": 0.7242,
|
| 1361 |
+
"step": 9650
|
| 1362 |
+
},
|
| 1363 |
+
{
|
| 1364 |
+
"epoch": 0.8983977030656664,
|
| 1365 |
+
"grad_norm": 11.137995719909668,
|
| 1366 |
+
"learning_rate": 7.784295564474634e-06,
|
| 1367 |
+
"loss": 0.6679,
|
| 1368 |
+
"step": 9700
|
| 1369 |
+
},
|
| 1370 |
+
{
|
| 1371 |
+
"epoch": 0.9030286190608502,
|
| 1372 |
+
"grad_norm": 12.533097267150879,
|
| 1373 |
+
"learning_rate": 7.767143494219754e-06,
|
| 1374 |
+
"loss": 0.7113,
|
| 1375 |
+
"step": 9750
|
| 1376 |
+
},
|
| 1377 |
+
{
|
| 1378 |
+
"epoch": 0.907659535056034,
|
| 1379 |
+
"grad_norm": 8.744980812072754,
|
| 1380 |
+
"learning_rate": 7.749991423964873e-06,
|
| 1381 |
+
"loss": 0.6634,
|
| 1382 |
+
"step": 9800
|
| 1383 |
+
},
|
| 1384 |
+
{
|
| 1385 |
+
"epoch": 0.912290451051218,
|
| 1386 |
+
"grad_norm": 12.743056297302246,
|
| 1387 |
+
"learning_rate": 7.732839353709993e-06,
|
| 1388 |
+
"loss": 0.7264,
|
| 1389 |
+
"step": 9850
|
| 1390 |
+
},
|
| 1391 |
+
{
|
| 1392 |
+
"epoch": 0.9169213670464018,
|
| 1393 |
+
"grad_norm": 14.499612808227539,
|
| 1394 |
+
"learning_rate": 7.715687283455113e-06,
|
| 1395 |
+
"loss": 0.6815,
|
| 1396 |
+
"step": 9900
|
| 1397 |
+
},
|
| 1398 |
+
{
|
| 1399 |
+
"epoch": 0.9215522830415857,
|
| 1400 |
+
"grad_norm": 12.846882820129395,
|
| 1401 |
+
"learning_rate": 7.698535213200234e-06,
|
| 1402 |
+
"loss": 0.7602,
|
| 1403 |
+
"step": 9950
|
| 1404 |
+
},
|
| 1405 |
+
{
|
| 1406 |
+
"epoch": 0.9261831990367695,
|
| 1407 |
+
"grad_norm": 10.059849739074707,
|
| 1408 |
+
"learning_rate": 7.681383142945354e-06,
|
| 1409 |
+
"loss": 0.6744,
|
| 1410 |
+
"step": 10000
|
| 1411 |
+
},
|
| 1412 |
+
{
|
| 1413 |
+
"epoch": 0.9308141150319533,
|
| 1414 |
+
"grad_norm": 13.844738006591797,
|
| 1415 |
+
"learning_rate": 7.664231072690475e-06,
|
| 1416 |
+
"loss": 0.7137,
|
| 1417 |
+
"step": 10050
|
| 1418 |
+
},
|
| 1419 |
+
{
|
| 1420 |
+
"epoch": 0.9354450310271372,
|
| 1421 |
+
"grad_norm": 8.811226844787598,
|
| 1422 |
+
"learning_rate": 7.647079002435595e-06,
|
| 1423 |
+
"loss": 0.7499,
|
| 1424 |
+
"step": 10100
|
| 1425 |
+
},
|
| 1426 |
+
{
|
| 1427 |
+
"epoch": 0.940075947022321,
|
| 1428 |
+
"grad_norm": 14.390593528747559,
|
| 1429 |
+
"learning_rate": 7.629926932180715e-06,
|
| 1430 |
+
"loss": 0.7108,
|
| 1431 |
+
"step": 10150
|
| 1432 |
+
},
|
| 1433 |
+
{
|
| 1434 |
+
"epoch": 0.9447068630175048,
|
| 1435 |
+
"grad_norm": 13.213685035705566,
|
| 1436 |
+
"learning_rate": 7.6127748619258345e-06,
|
| 1437 |
+
"loss": 0.7291,
|
| 1438 |
+
"step": 10200
|
| 1439 |
+
},
|
| 1440 |
+
{
|
| 1441 |
+
"epoch": 0.9493377790126887,
|
| 1442 |
+
"grad_norm": 12.737407684326172,
|
| 1443 |
+
"learning_rate": 7.595622791670956e-06,
|
| 1444 |
+
"loss": 0.6997,
|
| 1445 |
+
"step": 10250
|
| 1446 |
+
},
|
| 1447 |
+
{
|
| 1448 |
+
"epoch": 0.9539686950078725,
|
| 1449 |
+
"grad_norm": 10.98259449005127,
|
| 1450 |
+
"learning_rate": 7.578470721416076e-06,
|
| 1451 |
+
"loss": 0.7107,
|
| 1452 |
+
"step": 10300
|
| 1453 |
+
},
|
| 1454 |
+
{
|
| 1455 |
+
"epoch": 0.9585996110030564,
|
| 1456 |
+
"grad_norm": 12.873851776123047,
|
| 1457 |
+
"learning_rate": 7.561318651161195e-06,
|
| 1458 |
+
"loss": 0.6858,
|
| 1459 |
+
"step": 10350
|
| 1460 |
+
},
|
| 1461 |
+
{
|
| 1462 |
+
"epoch": 0.9632305269982403,
|
| 1463 |
+
"grad_norm": 9.490121841430664,
|
| 1464 |
+
"learning_rate": 7.544166580906316e-06,
|
| 1465 |
+
"loss": 0.7393,
|
| 1466 |
+
"step": 10400
|
| 1467 |
+
},
|
| 1468 |
+
{
|
| 1469 |
+
"epoch": 0.9678614429934241,
|
| 1470 |
+
"grad_norm": 12.453715324401855,
|
| 1471 |
+
"learning_rate": 7.527014510651436e-06,
|
| 1472 |
+
"loss": 0.7221,
|
| 1473 |
+
"step": 10450
|
| 1474 |
+
},
|
| 1475 |
+
{
|
| 1476 |
+
"epoch": 0.972492358988608,
|
| 1477 |
+
"grad_norm": 21.79591941833496,
|
| 1478 |
+
"learning_rate": 7.509862440396557e-06,
|
| 1479 |
+
"loss": 0.7092,
|
| 1480 |
+
"step": 10500
|
| 1481 |
+
},
|
| 1482 |
+
{
|
| 1483 |
+
"epoch": 0.9771232749837918,
|
| 1484 |
+
"grad_norm": 14.497958183288574,
|
| 1485 |
+
"learning_rate": 7.492710370141677e-06,
|
| 1486 |
+
"loss": 0.7017,
|
| 1487 |
+
"step": 10550
|
| 1488 |
+
},
|
| 1489 |
+
{
|
| 1490 |
+
"epoch": 0.9817541909789756,
|
| 1491 |
+
"grad_norm": 16.157657623291016,
|
| 1492 |
+
"learning_rate": 7.475558299886797e-06,
|
| 1493 |
+
"loss": 0.761,
|
| 1494 |
+
"step": 10600
|
| 1495 |
+
},
|
| 1496 |
+
{
|
| 1497 |
+
"epoch": 0.9863851069741595,
|
| 1498 |
+
"grad_norm": 9.551992416381836,
|
| 1499 |
+
"learning_rate": 7.458406229631917e-06,
|
| 1500 |
+
"loss": 0.7494,
|
| 1501 |
+
"step": 10650
|
| 1502 |
+
},
|
| 1503 |
+
{
|
| 1504 |
+
"epoch": 0.9910160229693433,
|
| 1505 |
+
"grad_norm": 12.097346305847168,
|
| 1506 |
+
"learning_rate": 7.441254159377037e-06,
|
| 1507 |
+
"loss": 0.6412,
|
| 1508 |
+
"step": 10700
|
| 1509 |
+
},
|
| 1510 |
+
{
|
| 1511 |
+
"epoch": 0.9956469389645272,
|
| 1512 |
+
"grad_norm": 6.892613887786865,
|
| 1513 |
+
"learning_rate": 7.4241020891221575e-06,
|
| 1514 |
+
"loss": 0.7003,
|
| 1515 |
+
"step": 10750
|
| 1516 |
+
},
|
| 1517 |
+
{
|
| 1518 |
+
"epoch": 1.000277854959711,
|
| 1519 |
+
"grad_norm": 10.388261795043945,
|
| 1520 |
+
"learning_rate": 7.406950018867278e-06,
|
| 1521 |
+
"loss": 0.6807,
|
| 1522 |
+
"step": 10800
|
| 1523 |
+
},
|
| 1524 |
+
{
|
| 1525 |
+
"epoch": 1.004908770954895,
|
| 1526 |
+
"grad_norm": 7.812435150146484,
|
| 1527 |
+
"learning_rate": 7.389797948612399e-06,
|
| 1528 |
+
"loss": 0.5981,
|
| 1529 |
+
"step": 10850
|
| 1530 |
+
},
|
| 1531 |
+
{
|
| 1532 |
+
"epoch": 1.0095396869500788,
|
| 1533 |
+
"grad_norm": 7.479443550109863,
|
| 1534 |
+
"learning_rate": 7.372645878357518e-06,
|
| 1535 |
+
"loss": 0.5875,
|
| 1536 |
+
"step": 10900
|
| 1537 |
+
},
|
| 1538 |
+
{
|
| 1539 |
+
"epoch": 1.0141706029452626,
|
| 1540 |
+
"grad_norm": 14.079961776733398,
|
| 1541 |
+
"learning_rate": 7.355493808102638e-06,
|
| 1542 |
+
"loss": 0.6712,
|
| 1543 |
+
"step": 10950
|
| 1544 |
+
},
|
| 1545 |
+
{
|
| 1546 |
+
"epoch": 1.0188015189404465,
|
| 1547 |
+
"grad_norm": 13.788300514221191,
|
| 1548 |
+
"learning_rate": 7.338341737847759e-06,
|
| 1549 |
+
"loss": 0.6216,
|
| 1550 |
+
"step": 11000
|
| 1551 |
+
},
|
| 1552 |
+
{
|
| 1553 |
+
"epoch": 1.0234324349356303,
|
| 1554 |
+
"grad_norm": 10.6414213180542,
|
| 1555 |
+
"learning_rate": 7.321189667592879e-06,
|
| 1556 |
+
"loss": 0.6274,
|
| 1557 |
+
"step": 11050
|
| 1558 |
+
},
|
| 1559 |
+
{
|
| 1560 |
+
"epoch": 1.0280633509308141,
|
| 1561 |
+
"grad_norm": 8.648374557495117,
|
| 1562 |
+
"learning_rate": 7.304037597338e-06,
|
| 1563 |
+
"loss": 0.602,
|
| 1564 |
+
"step": 11100
|
| 1565 |
+
},
|
| 1566 |
+
{
|
| 1567 |
+
"epoch": 1.032694266925998,
|
| 1568 |
+
"grad_norm": 12.511055946350098,
|
| 1569 |
+
"learning_rate": 7.2868855270831195e-06,
|
| 1570 |
+
"loss": 0.5893,
|
| 1571 |
+
"step": 11150
|
| 1572 |
+
},
|
| 1573 |
+
{
|
| 1574 |
+
"epoch": 1.0373251829211818,
|
| 1575 |
+
"grad_norm": 14.385446548461914,
|
| 1576 |
+
"learning_rate": 7.269733456828239e-06,
|
| 1577 |
+
"loss": 0.6545,
|
| 1578 |
+
"step": 11200
|
| 1579 |
+
},
|
| 1580 |
+
{
|
| 1581 |
+
"epoch": 1.0419560989163656,
|
| 1582 |
+
"grad_norm": 15.328493118286133,
|
| 1583 |
+
"learning_rate": 7.25258138657336e-06,
|
| 1584 |
+
"loss": 0.6003,
|
| 1585 |
+
"step": 11250
|
| 1586 |
+
},
|
| 1587 |
+
{
|
| 1588 |
+
"epoch": 1.0465870149115495,
|
| 1589 |
+
"grad_norm": 13.367624282836914,
|
| 1590 |
+
"learning_rate": 7.23542931631848e-06,
|
| 1591 |
+
"loss": 0.6151,
|
| 1592 |
+
"step": 11300
|
| 1593 |
+
},
|
| 1594 |
+
{
|
| 1595 |
+
"epoch": 1.0512179309067333,
|
| 1596 |
+
"grad_norm": 7.708530426025391,
|
| 1597 |
+
"learning_rate": 7.218277246063601e-06,
|
| 1598 |
+
"loss": 0.626,
|
| 1599 |
+
"step": 11350
|
| 1600 |
+
},
|
| 1601 |
+
{
|
| 1602 |
+
"epoch": 1.0558488469019172,
|
| 1603 |
+
"grad_norm": 12.538946151733398,
|
| 1604 |
+
"learning_rate": 7.201125175808721e-06,
|
| 1605 |
+
"loss": 0.652,
|
| 1606 |
+
"step": 11400
|
| 1607 |
+
},
|
| 1608 |
+
{
|
| 1609 |
+
"epoch": 1.060479762897101,
|
| 1610 |
+
"grad_norm": 12.275097846984863,
|
| 1611 |
+
"learning_rate": 7.1839731055538404e-06,
|
| 1612 |
+
"loss": 0.6398,
|
| 1613 |
+
"step": 11450
|
| 1614 |
+
},
|
| 1615 |
+
{
|
| 1616 |
+
"epoch": 1.0651106788922848,
|
| 1617 |
+
"grad_norm": 13.180695533752441,
|
| 1618 |
+
"learning_rate": 7.166821035298961e-06,
|
| 1619 |
+
"loss": 0.6825,
|
| 1620 |
+
"step": 11500
|
| 1621 |
+
},
|
| 1622 |
+
{
|
| 1623 |
+
"epoch": 1.0697415948874687,
|
| 1624 |
+
"grad_norm": 15.09621810913086,
|
| 1625 |
+
"learning_rate": 7.149668965044081e-06,
|
| 1626 |
+
"loss": 0.6298,
|
| 1627 |
+
"step": 11550
|
| 1628 |
+
},
|
| 1629 |
+
{
|
| 1630 |
+
"epoch": 1.0743725108826525,
|
| 1631 |
+
"grad_norm": 13.7881441116333,
|
| 1632 |
+
"learning_rate": 7.132516894789202e-06,
|
| 1633 |
+
"loss": 0.5414,
|
| 1634 |
+
"step": 11600
|
| 1635 |
+
},
|
| 1636 |
+
{
|
| 1637 |
+
"epoch": 1.0790034268778363,
|
| 1638 |
+
"grad_norm": 10.31930160522461,
|
| 1639 |
+
"learning_rate": 7.115364824534322e-06,
|
| 1640 |
+
"loss": 0.6619,
|
| 1641 |
+
"step": 11650
|
| 1642 |
+
},
|
| 1643 |
+
{
|
| 1644 |
+
"epoch": 1.0836343428730202,
|
| 1645 |
+
"grad_norm": 10.23030948638916,
|
| 1646 |
+
"learning_rate": 7.0982127542794425e-06,
|
| 1647 |
+
"loss": 0.6333,
|
| 1648 |
+
"step": 11700
|
| 1649 |
+
},
|
| 1650 |
+
{
|
| 1651 |
+
"epoch": 1.088265258868204,
|
| 1652 |
+
"grad_norm": 13.737223625183105,
|
| 1653 |
+
"learning_rate": 7.081060684024562e-06,
|
| 1654 |
+
"loss": 0.6896,
|
| 1655 |
+
"step": 11750
|
| 1656 |
+
},
|
| 1657 |
+
{
|
| 1658 |
+
"epoch": 1.092896174863388,
|
| 1659 |
+
"grad_norm": 10.776179313659668,
|
| 1660 |
+
"learning_rate": 7.063908613769682e-06,
|
| 1661 |
+
"loss": 0.6393,
|
| 1662 |
+
"step": 11800
|
| 1663 |
+
},
|
| 1664 |
+
{
|
| 1665 |
+
"epoch": 1.097527090858572,
|
| 1666 |
+
"grad_norm": 14.287514686584473,
|
| 1667 |
+
"learning_rate": 7.046756543514803e-06,
|
| 1668 |
+
"loss": 0.6185,
|
| 1669 |
+
"step": 11850
|
| 1670 |
+
},
|
| 1671 |
+
{
|
| 1672 |
+
"epoch": 1.1021580068537558,
|
| 1673 |
+
"grad_norm": 7.316586017608643,
|
| 1674 |
+
"learning_rate": 7.029604473259923e-06,
|
| 1675 |
+
"loss": 0.5542,
|
| 1676 |
+
"step": 11900
|
| 1677 |
+
},
|
| 1678 |
+
{
|
| 1679 |
+
"epoch": 1.1067889228489396,
|
| 1680 |
+
"grad_norm": 10.488934516906738,
|
| 1681 |
+
"learning_rate": 7.012452403005044e-06,
|
| 1682 |
+
"loss": 0.6723,
|
| 1683 |
+
"step": 11950
|
| 1684 |
+
},
|
| 1685 |
+
{
|
| 1686 |
+
"epoch": 1.1114198388441234,
|
| 1687 |
+
"grad_norm": 15.451316833496094,
|
| 1688 |
+
"learning_rate": 6.995300332750163e-06,
|
| 1689 |
+
"loss": 0.5714,
|
| 1690 |
+
"step": 12000
|
| 1691 |
+
},
|
| 1692 |
+
{
|
| 1693 |
+
"epoch": 1.1160507548393073,
|
| 1694 |
+
"grad_norm": 14.99089241027832,
|
| 1695 |
+
"learning_rate": 6.978148262495283e-06,
|
| 1696 |
+
"loss": 0.6006,
|
| 1697 |
+
"step": 12050
|
| 1698 |
+
},
|
| 1699 |
+
{
|
| 1700 |
+
"epoch": 1.120681670834491,
|
| 1701 |
+
"grad_norm": 12.084638595581055,
|
| 1702 |
+
"learning_rate": 6.960996192240404e-06,
|
| 1703 |
+
"loss": 0.5881,
|
| 1704 |
+
"step": 12100
|
| 1705 |
+
},
|
| 1706 |
+
{
|
| 1707 |
+
"epoch": 1.125312586829675,
|
| 1708 |
+
"grad_norm": 13.337446212768555,
|
| 1709 |
+
"learning_rate": 6.943844121985524e-06,
|
| 1710 |
+
"loss": 0.6301,
|
| 1711 |
+
"step": 12150
|
| 1712 |
+
},
|
| 1713 |
+
{
|
| 1714 |
+
"epoch": 1.1299435028248588,
|
| 1715 |
+
"grad_norm": 15.448657989501953,
|
| 1716 |
+
"learning_rate": 6.926692051730645e-06,
|
| 1717 |
+
"loss": 0.6238,
|
| 1718 |
+
"step": 12200
|
| 1719 |
+
},
|
| 1720 |
+
{
|
| 1721 |
+
"epoch": 1.1345744188200426,
|
| 1722 |
+
"grad_norm": 8.770259857177734,
|
| 1723 |
+
"learning_rate": 6.909539981475765e-06,
|
| 1724 |
+
"loss": 0.5843,
|
| 1725 |
+
"step": 12250
|
| 1726 |
+
},
|
| 1727 |
+
{
|
| 1728 |
+
"epoch": 1.1392053348152265,
|
| 1729 |
+
"grad_norm": 10.406338691711426,
|
| 1730 |
+
"learning_rate": 6.892387911220884e-06,
|
| 1731 |
+
"loss": 0.5978,
|
| 1732 |
+
"step": 12300
|
| 1733 |
+
},
|
| 1734 |
+
{
|
| 1735 |
+
"epoch": 1.1438362508104103,
|
| 1736 |
+
"grad_norm": 13.003678321838379,
|
| 1737 |
+
"learning_rate": 6.875235840966005e-06,
|
| 1738 |
+
"loss": 0.6475,
|
| 1739 |
+
"step": 12350
|
| 1740 |
+
},
|
| 1741 |
+
{
|
| 1742 |
+
"epoch": 1.1484671668055941,
|
| 1743 |
+
"grad_norm": 13.714174270629883,
|
| 1744 |
+
"learning_rate": 6.8580837707111255e-06,
|
| 1745 |
+
"loss": 0.5861,
|
| 1746 |
+
"step": 12400
|
| 1747 |
+
},
|
| 1748 |
+
{
|
| 1749 |
+
"epoch": 1.153098082800778,
|
| 1750 |
+
"grad_norm": 9.3059720993042,
|
| 1751 |
+
"learning_rate": 6.840931700456246e-06,
|
| 1752 |
+
"loss": 0.6743,
|
| 1753 |
+
"step": 12450
|
| 1754 |
+
},
|
| 1755 |
+
{
|
| 1756 |
+
"epoch": 1.1577289987959618,
|
| 1757 |
+
"grad_norm": 13.990130424499512,
|
| 1758 |
+
"learning_rate": 6.823779630201366e-06,
|
| 1759 |
+
"loss": 0.6724,
|
| 1760 |
+
"step": 12500
|
| 1761 |
+
},
|
| 1762 |
+
{
|
| 1763 |
+
"epoch": 1.1623599147911456,
|
| 1764 |
+
"grad_norm": 13.609382629394531,
|
| 1765 |
+
"learning_rate": 6.8066275599464855e-06,
|
| 1766 |
+
"loss": 0.5501,
|
| 1767 |
+
"step": 12550
|
| 1768 |
+
},
|
| 1769 |
+
{
|
| 1770 |
+
"epoch": 1.1669908307863295,
|
| 1771 |
+
"grad_norm": 7.106515884399414,
|
| 1772 |
+
"learning_rate": 6.789475489691606e-06,
|
| 1773 |
+
"loss": 0.5339,
|
| 1774 |
+
"step": 12600
|
| 1775 |
+
},
|
| 1776 |
+
{
|
| 1777 |
+
"epoch": 1.1716217467815133,
|
| 1778 |
+
"grad_norm": 17.91675567626953,
|
| 1779 |
+
"learning_rate": 6.7723234194367275e-06,
|
| 1780 |
+
"loss": 0.6094,
|
| 1781 |
+
"step": 12650
|
| 1782 |
+
},
|
| 1783 |
+
{
|
| 1784 |
+
"epoch": 1.1762526627766972,
|
| 1785 |
+
"grad_norm": 9.104899406433105,
|
| 1786 |
+
"learning_rate": 6.755171349181847e-06,
|
| 1787 |
+
"loss": 0.6913,
|
| 1788 |
+
"step": 12700
|
| 1789 |
+
},
|
| 1790 |
+
{
|
| 1791 |
+
"epoch": 1.180883578771881,
|
| 1792 |
+
"grad_norm": 13.601178169250488,
|
| 1793 |
+
"learning_rate": 6.738019278926967e-06,
|
| 1794 |
+
"loss": 0.6087,
|
| 1795 |
+
"step": 12750
|
| 1796 |
+
},
|
| 1797 |
+
{
|
| 1798 |
+
"epoch": 1.1855144947670648,
|
| 1799 |
+
"grad_norm": 12.020017623901367,
|
| 1800 |
+
"learning_rate": 6.7208672086720876e-06,
|
| 1801 |
+
"loss": 0.6051,
|
| 1802 |
+
"step": 12800
|
| 1803 |
+
},
|
| 1804 |
+
{
|
| 1805 |
+
"epoch": 1.190145410762249,
|
| 1806 |
+
"grad_norm": 11.219183921813965,
|
| 1807 |
+
"learning_rate": 6.703715138417207e-06,
|
| 1808 |
+
"loss": 0.6096,
|
| 1809 |
+
"step": 12850
|
| 1810 |
+
},
|
| 1811 |
+
{
|
| 1812 |
+
"epoch": 1.1947763267574327,
|
| 1813 |
+
"grad_norm": 10.890588760375977,
|
| 1814 |
+
"learning_rate": 6.686563068162327e-06,
|
| 1815 |
+
"loss": 0.5875,
|
| 1816 |
+
"step": 12900
|
| 1817 |
+
},
|
| 1818 |
+
{
|
| 1819 |
+
"epoch": 1.1994072427526166,
|
| 1820 |
+
"grad_norm": 9.199064254760742,
|
| 1821 |
+
"learning_rate": 6.6694109979074484e-06,
|
| 1822 |
+
"loss": 0.6276,
|
| 1823 |
+
"step": 12950
|
| 1824 |
+
},
|
| 1825 |
+
{
|
| 1826 |
+
"epoch": 1.2040381587478004,
|
| 1827 |
+
"grad_norm": 13.708924293518066,
|
| 1828 |
+
"learning_rate": 6.652258927652568e-06,
|
| 1829 |
+
"loss": 0.608,
|
| 1830 |
+
"step": 13000
|
| 1831 |
+
},
|
| 1832 |
+
{
|
| 1833 |
+
"epoch": 1.2086690747429842,
|
| 1834 |
+
"grad_norm": 7.073580741882324,
|
| 1835 |
+
"learning_rate": 6.635106857397689e-06,
|
| 1836 |
+
"loss": 0.5901,
|
| 1837 |
+
"step": 13050
|
| 1838 |
+
},
|
| 1839 |
+
{
|
| 1840 |
+
"epoch": 1.213299990738168,
|
| 1841 |
+
"grad_norm": 10.898780822753906,
|
| 1842 |
+
"learning_rate": 6.6179547871428085e-06,
|
| 1843 |
+
"loss": 0.554,
|
| 1844 |
+
"step": 13100
|
| 1845 |
+
},
|
| 1846 |
+
{
|
| 1847 |
+
"epoch": 1.217930906733352,
|
| 1848 |
+
"grad_norm": 13.31747055053711,
|
| 1849 |
+
"learning_rate": 6.600802716887928e-06,
|
| 1850 |
+
"loss": 0.5498,
|
| 1851 |
+
"step": 13150
|
| 1852 |
+
},
|
| 1853 |
+
{
|
| 1854 |
+
"epoch": 1.2225618227285358,
|
| 1855 |
+
"grad_norm": 10.6091890335083,
|
| 1856 |
+
"learning_rate": 6.58365064663305e-06,
|
| 1857 |
+
"loss": 0.6398,
|
| 1858 |
+
"step": 13200
|
| 1859 |
+
},
|
| 1860 |
+
{
|
| 1861 |
+
"epoch": 1.2271927387237196,
|
| 1862 |
+
"grad_norm": 11.425141334533691,
|
| 1863 |
+
"learning_rate": 6.566498576378169e-06,
|
| 1864 |
+
"loss": 0.6532,
|
| 1865 |
+
"step": 13250
|
| 1866 |
+
},
|
| 1867 |
+
{
|
| 1868 |
+
"epoch": 1.2318236547189034,
|
| 1869 |
+
"grad_norm": 16.974721908569336,
|
| 1870 |
+
"learning_rate": 6.54934650612329e-06,
|
| 1871 |
+
"loss": 0.6373,
|
| 1872 |
+
"step": 13300
|
| 1873 |
+
},
|
| 1874 |
+
{
|
| 1875 |
+
"epoch": 1.2364545707140873,
|
| 1876 |
+
"grad_norm": 15.951576232910156,
|
| 1877 |
+
"learning_rate": 6.53219443586841e-06,
|
| 1878 |
+
"loss": 0.5556,
|
| 1879 |
+
"step": 13350
|
| 1880 |
+
},
|
| 1881 |
+
{
|
| 1882 |
+
"epoch": 1.241085486709271,
|
| 1883 |
+
"grad_norm": 13.966357231140137,
|
| 1884 |
+
"learning_rate": 6.515042365613529e-06,
|
| 1885 |
+
"loss": 0.6114,
|
| 1886 |
+
"step": 13400
|
| 1887 |
+
},
|
| 1888 |
+
{
|
| 1889 |
+
"epoch": 1.245716402704455,
|
| 1890 |
+
"grad_norm": 10.22027587890625,
|
| 1891 |
+
"learning_rate": 6.49789029535865e-06,
|
| 1892 |
+
"loss": 0.5842,
|
| 1893 |
+
"step": 13450
|
| 1894 |
+
},
|
| 1895 |
+
{
|
| 1896 |
+
"epoch": 1.2503473186996388,
|
| 1897 |
+
"grad_norm": 14.317681312561035,
|
| 1898 |
+
"learning_rate": 6.4807382251037706e-06,
|
| 1899 |
+
"loss": 0.6064,
|
| 1900 |
+
"step": 13500
|
| 1901 |
+
},
|
| 1902 |
+
{
|
| 1903 |
+
"epoch": 1.2549782346948226,
|
| 1904 |
+
"grad_norm": 9.962833404541016,
|
| 1905 |
+
"learning_rate": 6.463586154848891e-06,
|
| 1906 |
+
"loss": 0.5688,
|
| 1907 |
+
"step": 13550
|
| 1908 |
+
},
|
| 1909 |
+
{
|
| 1910 |
+
"epoch": 1.2596091506900065,
|
| 1911 |
+
"grad_norm": 5.116006374359131,
|
| 1912 |
+
"learning_rate": 6.446434084594011e-06,
|
| 1913 |
+
"loss": 0.5714,
|
| 1914 |
+
"step": 13600
|
| 1915 |
+
},
|
| 1916 |
+
{
|
| 1917 |
+
"epoch": 1.2642400666851903,
|
| 1918 |
+
"grad_norm": 9.99418830871582,
|
| 1919 |
+
"learning_rate": 6.4292820143391314e-06,
|
| 1920 |
+
"loss": 0.6329,
|
| 1921 |
+
"step": 13650
|
| 1922 |
+
},
|
| 1923 |
+
{
|
| 1924 |
+
"epoch": 1.2688709826803741,
|
| 1925 |
+
"grad_norm": 12.337767601013184,
|
| 1926 |
+
"learning_rate": 6.412129944084251e-06,
|
| 1927 |
+
"loss": 0.5907,
|
| 1928 |
+
"step": 13700
|
| 1929 |
+
},
|
| 1930 |
+
{
|
| 1931 |
+
"epoch": 1.273501898675558,
|
| 1932 |
+
"grad_norm": 7.793351173400879,
|
| 1933 |
+
"learning_rate": 6.394977873829373e-06,
|
| 1934 |
+
"loss": 0.5418,
|
| 1935 |
+
"step": 13750
|
| 1936 |
+
},
|
| 1937 |
+
{
|
| 1938 |
+
"epoch": 1.2781328146707418,
|
| 1939 |
+
"grad_norm": 6.3952436447143555,
|
| 1940 |
+
"learning_rate": 6.377825803574492e-06,
|
| 1941 |
+
"loss": 0.6304,
|
| 1942 |
+
"step": 13800
|
| 1943 |
+
},
|
| 1944 |
+
{
|
| 1945 |
+
"epoch": 1.2827637306659256,
|
| 1946 |
+
"grad_norm": 10.685275077819824,
|
| 1947 |
+
"learning_rate": 6.360673733319612e-06,
|
| 1948 |
+
"loss": 0.5713,
|
| 1949 |
+
"step": 13850
|
| 1950 |
+
},
|
| 1951 |
+
{
|
| 1952 |
+
"epoch": 1.2873946466611095,
|
| 1953 |
+
"grad_norm": 12.462202072143555,
|
| 1954 |
+
"learning_rate": 6.343521663064733e-06,
|
| 1955 |
+
"loss": 0.64,
|
| 1956 |
+
"step": 13900
|
| 1957 |
+
},
|
| 1958 |
+
{
|
| 1959 |
+
"epoch": 1.2920255626562933,
|
| 1960 |
+
"grad_norm": 11.661152839660645,
|
| 1961 |
+
"learning_rate": 6.326369592809852e-06,
|
| 1962 |
+
"loss": 0.7054,
|
| 1963 |
+
"step": 13950
|
| 1964 |
+
},
|
| 1965 |
+
{
|
| 1966 |
+
"epoch": 1.2966564786514772,
|
| 1967 |
+
"grad_norm": 11.020445823669434,
|
| 1968 |
+
"learning_rate": 6.309217522554974e-06,
|
| 1969 |
+
"loss": 0.6831,
|
| 1970 |
+
"step": 14000
|
| 1971 |
+
},
|
| 1972 |
+
{
|
| 1973 |
+
"epoch": 1.301287394646661,
|
| 1974 |
+
"grad_norm": 12.878907203674316,
|
| 1975 |
+
"learning_rate": 6.2920654523000935e-06,
|
| 1976 |
+
"loss": 0.5805,
|
| 1977 |
+
"step": 14050
|
| 1978 |
+
},
|
| 1979 |
+
{
|
| 1980 |
+
"epoch": 1.3059183106418448,
|
| 1981 |
+
"grad_norm": 7.879404544830322,
|
| 1982 |
+
"learning_rate": 6.274913382045213e-06,
|
| 1983 |
+
"loss": 0.5727,
|
| 1984 |
+
"step": 14100
|
| 1985 |
+
},
|
| 1986 |
+
{
|
| 1987 |
+
"epoch": 1.3105492266370289,
|
| 1988 |
+
"grad_norm": 13.666654586791992,
|
| 1989 |
+
"learning_rate": 6.257761311790334e-06,
|
| 1990 |
+
"loss": 0.5393,
|
| 1991 |
+
"step": 14150
|
| 1992 |
+
},
|
| 1993 |
+
{
|
| 1994 |
+
"epoch": 1.3151801426322127,
|
| 1995 |
+
"grad_norm": 13.355966567993164,
|
| 1996 |
+
"learning_rate": 6.2406092415354535e-06,
|
| 1997 |
+
"loss": 0.5949,
|
| 1998 |
+
"step": 14200
|
| 1999 |
+
},
|
| 2000 |
+
{
|
| 2001 |
+
"epoch": 1.3198110586273966,
|
| 2002 |
+
"grad_norm": 9.979362487792969,
|
| 2003 |
+
"learning_rate": 6.223457171280573e-06,
|
| 2004 |
+
"loss": 0.5794,
|
| 2005 |
+
"step": 14250
|
| 2006 |
+
},
|
| 2007 |
+
{
|
| 2008 |
+
"epoch": 1.3244419746225804,
|
| 2009 |
+
"grad_norm": 14.988487243652344,
|
| 2010 |
+
"learning_rate": 6.206305101025695e-06,
|
| 2011 |
+
"loss": 0.61,
|
| 2012 |
+
"step": 14300
|
| 2013 |
+
},
|
| 2014 |
+
{
|
| 2015 |
+
"epoch": 1.3290728906177642,
|
| 2016 |
+
"grad_norm": 13.34450626373291,
|
| 2017 |
+
"learning_rate": 6.1891530307708144e-06,
|
| 2018 |
+
"loss": 0.6704,
|
| 2019 |
+
"step": 14350
|
| 2020 |
+
},
|
| 2021 |
+
{
|
| 2022 |
+
"epoch": 1.333703806612948,
|
| 2023 |
+
"grad_norm": 12.021376609802246,
|
| 2024 |
+
"learning_rate": 6.172000960515935e-06,
|
| 2025 |
+
"loss": 0.5829,
|
| 2026 |
+
"step": 14400
|
| 2027 |
+
},
|
| 2028 |
+
{
|
| 2029 |
+
"epoch": 1.338334722608132,
|
| 2030 |
+
"grad_norm": 11.892980575561523,
|
| 2031 |
+
"learning_rate": 6.154848890261055e-06,
|
| 2032 |
+
"loss": 0.5603,
|
| 2033 |
+
"step": 14450
|
| 2034 |
+
},
|
| 2035 |
+
{
|
| 2036 |
+
"epoch": 1.3429656386033157,
|
| 2037 |
+
"grad_norm": 15.183904647827148,
|
| 2038 |
+
"learning_rate": 6.1376968200061745e-06,
|
| 2039 |
+
"loss": 0.6163,
|
| 2040 |
+
"step": 14500
|
| 2041 |
+
},
|
| 2042 |
+
{
|
| 2043 |
+
"epoch": 1.3475965545984996,
|
| 2044 |
+
"grad_norm": 14.723098754882812,
|
| 2045 |
+
"learning_rate": 6.120544749751296e-06,
|
| 2046 |
+
"loss": 0.7141,
|
| 2047 |
+
"step": 14550
|
| 2048 |
+
},
|
| 2049 |
+
{
|
| 2050 |
+
"epoch": 1.3522274705936834,
|
| 2051 |
+
"grad_norm": 7.206448554992676,
|
| 2052 |
+
"learning_rate": 6.103392679496416e-06,
|
| 2053 |
+
"loss": 0.5413,
|
| 2054 |
+
"step": 14600
|
| 2055 |
+
},
|
| 2056 |
+
{
|
| 2057 |
+
"epoch": 1.3568583865888673,
|
| 2058 |
+
"grad_norm": 14.650779724121094,
|
| 2059 |
+
"learning_rate": 6.086240609241536e-06,
|
| 2060 |
+
"loss": 0.5972,
|
| 2061 |
+
"step": 14650
|
| 2062 |
+
},
|
| 2063 |
+
{
|
| 2064 |
+
"epoch": 1.361489302584051,
|
| 2065 |
+
"grad_norm": 11.455196380615234,
|
| 2066 |
+
"learning_rate": 6.069088538986656e-06,
|
| 2067 |
+
"loss": 0.6104,
|
| 2068 |
+
"step": 14700
|
| 2069 |
+
},
|
| 2070 |
+
{
|
| 2071 |
+
"epoch": 1.366120218579235,
|
| 2072 |
+
"grad_norm": 10.497644424438477,
|
| 2073 |
+
"learning_rate": 6.0519364687317765e-06,
|
| 2074 |
+
"loss": 0.6147,
|
| 2075 |
+
"step": 14750
|
| 2076 |
+
},
|
| 2077 |
+
{
|
| 2078 |
+
"epoch": 1.3707511345744188,
|
| 2079 |
+
"grad_norm": 15.974631309509277,
|
| 2080 |
+
"learning_rate": 6.034784398476896e-06,
|
| 2081 |
+
"loss": 0.6319,
|
| 2082 |
+
"step": 14800
|
| 2083 |
+
},
|
| 2084 |
+
{
|
| 2085 |
+
"epoch": 1.3753820505696026,
|
| 2086 |
+
"grad_norm": 13.41238021850586,
|
| 2087 |
+
"learning_rate": 6.017632328222018e-06,
|
| 2088 |
+
"loss": 0.6277,
|
| 2089 |
+
"step": 14850
|
| 2090 |
+
},
|
| 2091 |
+
{
|
| 2092 |
+
"epoch": 1.3800129665647864,
|
| 2093 |
+
"grad_norm": 11.693626403808594,
|
| 2094 |
+
"learning_rate": 6.000480257967137e-06,
|
| 2095 |
+
"loss": 0.5986,
|
| 2096 |
+
"step": 14900
|
| 2097 |
+
},
|
| 2098 |
+
{
|
| 2099 |
+
"epoch": 1.3846438825599703,
|
| 2100 |
+
"grad_norm": 14.90977668762207,
|
| 2101 |
+
"learning_rate": 5.983328187712257e-06,
|
| 2102 |
+
"loss": 0.5409,
|
| 2103 |
+
"step": 14950
|
| 2104 |
+
},
|
| 2105 |
+
{
|
| 2106 |
+
"epoch": 1.3892747985551543,
|
| 2107 |
+
"grad_norm": 15.576924324035645,
|
| 2108 |
+
"learning_rate": 5.966176117457378e-06,
|
| 2109 |
+
"loss": 0.6396,
|
| 2110 |
+
"step": 15000
|
| 2111 |
+
},
|
| 2112 |
+
{
|
| 2113 |
+
"epoch": 1.3939057145503382,
|
| 2114 |
+
"grad_norm": 12.302413940429688,
|
| 2115 |
+
"learning_rate": 5.9490240472024974e-06,
|
| 2116 |
+
"loss": 0.599,
|
| 2117 |
+
"step": 15050
|
| 2118 |
+
},
|
| 2119 |
+
{
|
| 2120 |
+
"epoch": 1.398536630545522,
|
| 2121 |
+
"grad_norm": 11.971318244934082,
|
| 2122 |
+
"learning_rate": 5.931871976947619e-06,
|
| 2123 |
+
"loss": 0.6512,
|
| 2124 |
+
"step": 15100
|
| 2125 |
+
},
|
| 2126 |
+
{
|
| 2127 |
+
"epoch": 1.4031675465407059,
|
| 2128 |
+
"grad_norm": 9.926959037780762,
|
| 2129 |
+
"learning_rate": 5.914719906692739e-06,
|
| 2130 |
+
"loss": 0.5849,
|
| 2131 |
+
"step": 15150
|
| 2132 |
+
},
|
| 2133 |
+
{
|
| 2134 |
+
"epoch": 1.4077984625358897,
|
| 2135 |
+
"grad_norm": 4.981182098388672,
|
| 2136 |
+
"learning_rate": 5.897567836437858e-06,
|
| 2137 |
+
"loss": 0.582,
|
| 2138 |
+
"step": 15200
|
| 2139 |
+
},
|
| 2140 |
+
{
|
| 2141 |
+
"epoch": 1.4124293785310735,
|
| 2142 |
+
"grad_norm": 13.111713409423828,
|
| 2143 |
+
"learning_rate": 5.880415766182979e-06,
|
| 2144 |
+
"loss": 0.6674,
|
| 2145 |
+
"step": 15250
|
| 2146 |
+
},
|
| 2147 |
+
{
|
| 2148 |
+
"epoch": 1.4170602945262574,
|
| 2149 |
+
"grad_norm": 14.456388473510742,
|
| 2150 |
+
"learning_rate": 5.863263695928099e-06,
|
| 2151 |
+
"loss": 0.6413,
|
| 2152 |
+
"step": 15300
|
| 2153 |
+
},
|
| 2154 |
+
{
|
| 2155 |
+
"epoch": 1.4216912105214412,
|
| 2156 |
+
"grad_norm": 12.475411415100098,
|
| 2157 |
+
"learning_rate": 5.84611162567322e-06,
|
| 2158 |
+
"loss": 0.6454,
|
| 2159 |
+
"step": 15350
|
| 2160 |
+
},
|
| 2161 |
+
{
|
| 2162 |
+
"epoch": 1.426322126516625,
|
| 2163 |
+
"grad_norm": 14.428534507751465,
|
| 2164 |
+
"learning_rate": 5.82895955541834e-06,
|
| 2165 |
+
"loss": 0.5802,
|
| 2166 |
+
"step": 15400
|
| 2167 |
+
},
|
| 2168 |
+
{
|
| 2169 |
+
"epoch": 1.4309530425118089,
|
| 2170 |
+
"grad_norm": 10.774445533752441,
|
| 2171 |
+
"learning_rate": 5.8118074851634595e-06,
|
| 2172 |
+
"loss": 0.6026,
|
| 2173 |
+
"step": 15450
|
| 2174 |
+
},
|
| 2175 |
+
{
|
| 2176 |
+
"epoch": 1.4355839585069927,
|
| 2177 |
+
"grad_norm": 7.127171039581299,
|
| 2178 |
+
"learning_rate": 5.79465541490858e-06,
|
| 2179 |
+
"loss": 0.5861,
|
| 2180 |
+
"step": 15500
|
| 2181 |
+
},
|
| 2182 |
+
{
|
| 2183 |
+
"epoch": 1.4402148745021766,
|
| 2184 |
+
"grad_norm": 20.572057723999023,
|
| 2185 |
+
"learning_rate": 5.7775033446537e-06,
|
| 2186 |
+
"loss": 0.6995,
|
| 2187 |
+
"step": 15550
|
| 2188 |
+
},
|
| 2189 |
+
{
|
| 2190 |
+
"epoch": 1.4448457904973604,
|
| 2191 |
+
"grad_norm": 9.042718887329102,
|
| 2192 |
+
"learning_rate": 5.7603512743988195e-06,
|
| 2193 |
+
"loss": 0.602,
|
| 2194 |
+
"step": 15600
|
| 2195 |
+
},
|
| 2196 |
+
{
|
| 2197 |
+
"epoch": 1.4494767064925442,
|
| 2198 |
+
"grad_norm": 10.351805686950684,
|
| 2199 |
+
"learning_rate": 5.743199204143941e-06,
|
| 2200 |
+
"loss": 0.5973,
|
| 2201 |
+
"step": 15650
|
| 2202 |
+
},
|
| 2203 |
+
{
|
| 2204 |
+
"epoch": 1.454107622487728,
|
| 2205 |
+
"grad_norm": 20.35566520690918,
|
| 2206 |
+
"learning_rate": 5.726047133889061e-06,
|
| 2207 |
+
"loss": 0.5642,
|
| 2208 |
+
"step": 15700
|
| 2209 |
+
},
|
| 2210 |
+
{
|
| 2211 |
+
"epoch": 1.458738538482912,
|
| 2212 |
+
"grad_norm": 12.212347984313965,
|
| 2213 |
+
"learning_rate": 5.708895063634181e-06,
|
| 2214 |
+
"loss": 0.5624,
|
| 2215 |
+
"step": 15750
|
| 2216 |
+
},
|
| 2217 |
+
{
|
| 2218 |
+
"epoch": 1.4633694544780957,
|
| 2219 |
+
"grad_norm": 7.124228477478027,
|
| 2220 |
+
"learning_rate": 5.691742993379301e-06,
|
| 2221 |
+
"loss": 0.5613,
|
| 2222 |
+
"step": 15800
|
| 2223 |
+
},
|
| 2224 |
+
{
|
| 2225 |
+
"epoch": 1.4680003704732796,
|
| 2226 |
+
"grad_norm": 18.672344207763672,
|
| 2227 |
+
"learning_rate": 5.6745909231244216e-06,
|
| 2228 |
+
"loss": 0.5946,
|
| 2229 |
+
"step": 15850
|
| 2230 |
+
},
|
| 2231 |
+
{
|
| 2232 |
+
"epoch": 1.4726312864684634,
|
| 2233 |
+
"grad_norm": 15.380640029907227,
|
| 2234 |
+
"learning_rate": 5.657438852869542e-06,
|
| 2235 |
+
"loss": 0.6391,
|
| 2236 |
+
"step": 15900
|
| 2237 |
+
},
|
| 2238 |
+
{
|
| 2239 |
+
"epoch": 1.4772622024636473,
|
| 2240 |
+
"grad_norm": 13.782407760620117,
|
| 2241 |
+
"learning_rate": 5.640286782614663e-06,
|
| 2242 |
+
"loss": 0.5824,
|
| 2243 |
+
"step": 15950
|
| 2244 |
+
},
|
| 2245 |
+
{
|
| 2246 |
+
"epoch": 1.481893118458831,
|
| 2247 |
+
"grad_norm": 10.316747665405273,
|
| 2248 |
+
"learning_rate": 5.6231347123597825e-06,
|
| 2249 |
+
"loss": 0.5761,
|
| 2250 |
+
"step": 16000
|
| 2251 |
+
},
|
| 2252 |
+
{
|
| 2253 |
+
"epoch": 1.486524034454015,
|
| 2254 |
+
"grad_norm": 14.089554786682129,
|
| 2255 |
+
"learning_rate": 5.605982642104902e-06,
|
| 2256 |
+
"loss": 0.6171,
|
| 2257 |
+
"step": 16050
|
| 2258 |
+
},
|
| 2259 |
+
{
|
| 2260 |
+
"epoch": 1.4911549504491988,
|
| 2261 |
+
"grad_norm": 11.800119400024414,
|
| 2262 |
+
"learning_rate": 5.588830571850023e-06,
|
| 2263 |
+
"loss": 0.6445,
|
| 2264 |
+
"step": 16100
|
| 2265 |
+
},
|
| 2266 |
+
{
|
| 2267 |
+
"epoch": 1.4957858664443826,
|
| 2268 |
+
"grad_norm": 13.505830764770508,
|
| 2269 |
+
"learning_rate": 5.5716785015951425e-06,
|
| 2270 |
+
"loss": 0.5833,
|
| 2271 |
+
"step": 16150
|
| 2272 |
+
},
|
| 2273 |
+
{
|
| 2274 |
+
"epoch": 1.5004167824395664,
|
| 2275 |
+
"grad_norm": 11.643794059753418,
|
| 2276 |
+
"learning_rate": 5.554526431340264e-06,
|
| 2277 |
+
"loss": 0.5479,
|
| 2278 |
+
"step": 16200
|
| 2279 |
+
},
|
| 2280 |
+
{
|
| 2281 |
+
"epoch": 1.5050476984347503,
|
| 2282 |
+
"grad_norm": 16.229856491088867,
|
| 2283 |
+
"learning_rate": 5.537374361085384e-06,
|
| 2284 |
+
"loss": 0.603,
|
| 2285 |
+
"step": 16250
|
| 2286 |
+
},
|
| 2287 |
+
{
|
| 2288 |
+
"epoch": 1.5096786144299341,
|
| 2289 |
+
"grad_norm": 23.821744918823242,
|
| 2290 |
+
"learning_rate": 5.520222290830503e-06,
|
| 2291 |
+
"loss": 0.5961,
|
| 2292 |
+
"step": 16300
|
| 2293 |
+
},
|
| 2294 |
+
{
|
| 2295 |
+
"epoch": 1.514309530425118,
|
| 2296 |
+
"grad_norm": 16.5296630859375,
|
| 2297 |
+
"learning_rate": 5.503070220575624e-06,
|
| 2298 |
+
"loss": 0.5738,
|
| 2299 |
+
"step": 16350
|
| 2300 |
+
},
|
| 2301 |
+
{
|
| 2302 |
+
"epoch": 1.5189404464203018,
|
| 2303 |
+
"grad_norm": 14.164803504943848,
|
| 2304 |
+
"learning_rate": 5.485918150320744e-06,
|
| 2305 |
+
"loss": 0.5495,
|
| 2306 |
+
"step": 16400
|
| 2307 |
+
},
|
| 2308 |
+
{
|
| 2309 |
+
"epoch": 1.5235713624154856,
|
| 2310 |
+
"grad_norm": 10.112049102783203,
|
| 2311 |
+
"learning_rate": 5.468766080065865e-06,
|
| 2312 |
+
"loss": 0.5656,
|
| 2313 |
+
"step": 16450
|
| 2314 |
+
},
|
| 2315 |
+
{
|
| 2316 |
+
"epoch": 1.5282022784106695,
|
| 2317 |
+
"grad_norm": 11.996503829956055,
|
| 2318 |
+
"learning_rate": 5.451614009810985e-06,
|
| 2319 |
+
"loss": 0.6021,
|
| 2320 |
+
"step": 16500
|
| 2321 |
+
},
|
| 2322 |
+
{
|
| 2323 |
+
"epoch": 1.5328331944058535,
|
| 2324 |
+
"grad_norm": 7.8670525550842285,
|
| 2325 |
+
"learning_rate": 5.4344619395561046e-06,
|
| 2326 |
+
"loss": 0.6592,
|
| 2327 |
+
"step": 16550
|
| 2328 |
+
},
|
| 2329 |
+
{
|
| 2330 |
+
"epoch": 1.5374641104010374,
|
| 2331 |
+
"grad_norm": 14.625728607177734,
|
| 2332 |
+
"learning_rate": 5.417309869301225e-06,
|
| 2333 |
+
"loss": 0.6173,
|
| 2334 |
+
"step": 16600
|
| 2335 |
+
},
|
| 2336 |
+
{
|
| 2337 |
+
"epoch": 1.5420950263962212,
|
| 2338 |
+
"grad_norm": 14.137253761291504,
|
| 2339 |
+
"learning_rate": 5.400157799046345e-06,
|
| 2340 |
+
"loss": 0.5709,
|
| 2341 |
+
"step": 16650
|
| 2342 |
+
},
|
| 2343 |
+
{
|
| 2344 |
+
"epoch": 1.546725942391405,
|
| 2345 |
+
"grad_norm": 9.287371635437012,
|
| 2346 |
+
"learning_rate": 5.383005728791466e-06,
|
| 2347 |
+
"loss": 0.6174,
|
| 2348 |
+
"step": 16700
|
| 2349 |
+
},
|
| 2350 |
+
{
|
| 2351 |
+
"epoch": 1.5513568583865889,
|
| 2352 |
+
"grad_norm": 12.439033508300781,
|
| 2353 |
+
"learning_rate": 5.365853658536586e-06,
|
| 2354 |
+
"loss": 0.5798,
|
| 2355 |
+
"step": 16750
|
| 2356 |
+
},
|
| 2357 |
+
{
|
| 2358 |
+
"epoch": 1.5559877743817727,
|
| 2359 |
+
"grad_norm": 7.709993362426758,
|
| 2360 |
+
"learning_rate": 5.348701588281706e-06,
|
| 2361 |
+
"loss": 0.571,
|
| 2362 |
+
"step": 16800
|
| 2363 |
+
},
|
| 2364 |
+
{
|
| 2365 |
+
"epoch": 1.5606186903769566,
|
| 2366 |
+
"grad_norm": 8.796138763427734,
|
| 2367 |
+
"learning_rate": 5.331549518026826e-06,
|
| 2368 |
+
"loss": 0.5744,
|
| 2369 |
+
"step": 16850
|
| 2370 |
+
},
|
| 2371 |
+
{
|
| 2372 |
+
"epoch": 1.5652496063721404,
|
| 2373 |
+
"grad_norm": 7.53299617767334,
|
| 2374 |
+
"learning_rate": 5.314397447771946e-06,
|
| 2375 |
+
"loss": 0.6213,
|
| 2376 |
+
"step": 16900
|
| 2377 |
+
},
|
| 2378 |
+
{
|
| 2379 |
+
"epoch": 1.5698805223673242,
|
| 2380 |
+
"grad_norm": 13.07565689086914,
|
| 2381 |
+
"learning_rate": 5.297245377517067e-06,
|
| 2382 |
+
"loss": 0.586,
|
| 2383 |
+
"step": 16950
|
| 2384 |
+
},
|
| 2385 |
+
{
|
| 2386 |
+
"epoch": 1.574511438362508,
|
| 2387 |
+
"grad_norm": 8.85544490814209,
|
| 2388 |
+
"learning_rate": 5.280093307262187e-06,
|
| 2389 |
+
"loss": 0.5868,
|
| 2390 |
+
"step": 17000
|
| 2391 |
+
},
|
| 2392 |
+
{
|
| 2393 |
+
"epoch": 1.579142354357692,
|
| 2394 |
+
"grad_norm": 11.888114929199219,
|
| 2395 |
+
"learning_rate": 5.262941237007308e-06,
|
| 2396 |
+
"loss": 0.6038,
|
| 2397 |
+
"step": 17050
|
| 2398 |
+
},
|
| 2399 |
+
{
|
| 2400 |
+
"epoch": 1.583773270352876,
|
| 2401 |
+
"grad_norm": 15.81039047241211,
|
| 2402 |
+
"learning_rate": 5.2457891667524275e-06,
|
| 2403 |
+
"loss": 0.5684,
|
| 2404 |
+
"step": 17100
|
| 2405 |
+
},
|
| 2406 |
+
{
|
| 2407 |
+
"epoch": 1.5884041863480598,
|
| 2408 |
+
"grad_norm": 11.658292770385742,
|
| 2409 |
+
"learning_rate": 5.228637096497547e-06,
|
| 2410 |
+
"loss": 0.63,
|
| 2411 |
+
"step": 17150
|
| 2412 |
+
},
|
| 2413 |
+
{
|
| 2414 |
+
"epoch": 1.5930351023432436,
|
| 2415 |
+
"grad_norm": 12.549324989318848,
|
| 2416 |
+
"learning_rate": 5.211485026242668e-06,
|
| 2417 |
+
"loss": 0.5832,
|
| 2418 |
+
"step": 17200
|
| 2419 |
+
},
|
| 2420 |
+
{
|
| 2421 |
+
"epoch": 1.5976660183384275,
|
| 2422 |
+
"grad_norm": 13.756429672241211,
|
| 2423 |
+
"learning_rate": 5.194332955987788e-06,
|
| 2424 |
+
"loss": 0.566,
|
| 2425 |
+
"step": 17250
|
| 2426 |
+
},
|
| 2427 |
+
{
|
| 2428 |
+
"epoch": 1.6022969343336113,
|
| 2429 |
+
"grad_norm": 13.937918663024902,
|
| 2430 |
+
"learning_rate": 5.177180885732909e-06,
|
| 2431 |
+
"loss": 0.5913,
|
| 2432 |
+
"step": 17300
|
| 2433 |
+
},
|
| 2434 |
+
{
|
| 2435 |
+
"epoch": 1.6069278503287951,
|
| 2436 |
+
"grad_norm": 15.999268531799316,
|
| 2437 |
+
"learning_rate": 5.160028815478029e-06,
|
| 2438 |
+
"loss": 0.6455,
|
| 2439 |
+
"step": 17350
|
| 2440 |
+
},
|
| 2441 |
+
{
|
| 2442 |
+
"epoch": 1.611558766323979,
|
| 2443 |
+
"grad_norm": 11.241209983825684,
|
| 2444 |
+
"learning_rate": 5.1428767452231485e-06,
|
| 2445 |
+
"loss": 0.5808,
|
| 2446 |
+
"step": 17400
|
| 2447 |
+
},
|
| 2448 |
+
{
|
| 2449 |
+
"epoch": 1.6161896823191628,
|
| 2450 |
+
"grad_norm": 10.14393424987793,
|
| 2451 |
+
"learning_rate": 5.125724674968269e-06,
|
| 2452 |
+
"loss": 0.54,
|
| 2453 |
+
"step": 17450
|
| 2454 |
+
},
|
| 2455 |
+
{
|
| 2456 |
+
"epoch": 1.6208205983143467,
|
| 2457 |
+
"grad_norm": 10.87530517578125,
|
| 2458 |
+
"learning_rate": 5.108572604713389e-06,
|
| 2459 |
+
"loss": 0.6425,
|
| 2460 |
+
"step": 17500
|
| 2461 |
+
},
|
| 2462 |
+
{
|
| 2463 |
+
"epoch": 1.6254515143095305,
|
| 2464 |
+
"grad_norm": 14.635629653930664,
|
| 2465 |
+
"learning_rate": 5.09142053445851e-06,
|
| 2466 |
+
"loss": 0.5994,
|
| 2467 |
+
"step": 17550
|
| 2468 |
+
},
|
| 2469 |
+
{
|
| 2470 |
+
"epoch": 1.6300824303047143,
|
| 2471 |
+
"grad_norm": 17.629318237304688,
|
| 2472 |
+
"learning_rate": 5.07426846420363e-06,
|
| 2473 |
+
"loss": 0.5827,
|
| 2474 |
+
"step": 17600
|
| 2475 |
+
},
|
| 2476 |
+
{
|
| 2477 |
+
"epoch": 1.6347133462998982,
|
| 2478 |
+
"grad_norm": 14.779356002807617,
|
| 2479 |
+
"learning_rate": 5.05711639394875e-06,
|
| 2480 |
+
"loss": 0.57,
|
| 2481 |
+
"step": 17650
|
| 2482 |
+
},
|
| 2483 |
+
{
|
| 2484 |
+
"epoch": 1.639344262295082,
|
| 2485 |
+
"grad_norm": 13.21997356414795,
|
| 2486 |
+
"learning_rate": 5.03996432369387e-06,
|
| 2487 |
+
"loss": 0.555,
|
| 2488 |
+
"step": 17700
|
| 2489 |
+
},
|
| 2490 |
+
{
|
| 2491 |
+
"epoch": 1.6439751782902658,
|
| 2492 |
+
"grad_norm": 10.608312606811523,
|
| 2493 |
+
"learning_rate": 5.02281225343899e-06,
|
| 2494 |
+
"loss": 0.6238,
|
| 2495 |
+
"step": 17750
|
| 2496 |
+
},
|
| 2497 |
+
{
|
| 2498 |
+
"epoch": 1.6486060942854497,
|
| 2499 |
+
"grad_norm": 10.145637512207031,
|
| 2500 |
+
"learning_rate": 5.005660183184111e-06,
|
| 2501 |
+
"loss": 0.6148,
|
| 2502 |
+
"step": 17800
|
| 2503 |
+
},
|
| 2504 |
+
{
|
| 2505 |
+
"epoch": 1.6532370102806335,
|
| 2506 |
+
"grad_norm": 13.766275405883789,
|
| 2507 |
+
"learning_rate": 4.988508112929231e-06,
|
| 2508 |
+
"loss": 0.6019,
|
| 2509 |
+
"step": 17850
|
| 2510 |
+
},
|
| 2511 |
+
{
|
| 2512 |
+
"epoch": 1.6578679262758174,
|
| 2513 |
+
"grad_norm": 12.229156494140625,
|
| 2514 |
+
"learning_rate": 4.971356042674351e-06,
|
| 2515 |
+
"loss": 0.6072,
|
| 2516 |
+
"step": 17900
|
| 2517 |
+
},
|
| 2518 |
+
{
|
| 2519 |
+
"epoch": 1.6624988422710012,
|
| 2520 |
+
"grad_norm": 13.928374290466309,
|
| 2521 |
+
"learning_rate": 4.954203972419471e-06,
|
| 2522 |
+
"loss": 0.6248,
|
| 2523 |
+
"step": 17950
|
| 2524 |
+
},
|
| 2525 |
+
{
|
| 2526 |
+
"epoch": 1.667129758266185,
|
| 2527 |
+
"grad_norm": 11.471216201782227,
|
| 2528 |
+
"learning_rate": 4.937051902164592e-06,
|
| 2529 |
+
"loss": 0.6054,
|
| 2530 |
+
"step": 18000
|
| 2531 |
+
},
|
| 2532 |
+
{
|
| 2533 |
+
"epoch": 1.6717606742613689,
|
| 2534 |
+
"grad_norm": 15.625840187072754,
|
| 2535 |
+
"learning_rate": 4.919899831909712e-06,
|
| 2536 |
+
"loss": 0.5306,
|
| 2537 |
+
"step": 18050
|
| 2538 |
+
},
|
| 2539 |
+
{
|
| 2540 |
+
"epoch": 1.6763915902565527,
|
| 2541 |
+
"grad_norm": 13.443790435791016,
|
| 2542 |
+
"learning_rate": 4.902747761654832e-06,
|
| 2543 |
+
"loss": 0.574,
|
| 2544 |
+
"step": 18100
|
| 2545 |
+
},
|
| 2546 |
+
{
|
| 2547 |
+
"epoch": 1.6810225062517365,
|
| 2548 |
+
"grad_norm": 9.181079864501953,
|
| 2549 |
+
"learning_rate": 4.885595691399953e-06,
|
| 2550 |
+
"loss": 0.5879,
|
| 2551 |
+
"step": 18150
|
| 2552 |
+
},
|
| 2553 |
+
{
|
| 2554 |
+
"epoch": 1.6856534222469204,
|
| 2555 |
+
"grad_norm": 11.0618314743042,
|
| 2556 |
+
"learning_rate": 4.868443621145073e-06,
|
| 2557 |
+
"loss": 0.5564,
|
| 2558 |
+
"step": 18200
|
| 2559 |
+
},
|
| 2560 |
+
{
|
| 2561 |
+
"epoch": 1.6902843382421042,
|
| 2562 |
+
"grad_norm": 11.787243843078613,
|
| 2563 |
+
"learning_rate": 4.851291550890193e-06,
|
| 2564 |
+
"loss": 0.5756,
|
| 2565 |
+
"step": 18250
|
| 2566 |
+
},
|
| 2567 |
+
{
|
| 2568 |
+
"epoch": 1.694915254237288,
|
| 2569 |
+
"grad_norm": 14.000703811645508,
|
| 2570 |
+
"learning_rate": 4.834139480635313e-06,
|
| 2571 |
+
"loss": 0.5347,
|
| 2572 |
+
"step": 18300
|
| 2573 |
+
},
|
| 2574 |
+
{
|
| 2575 |
+
"epoch": 1.699546170232472,
|
| 2576 |
+
"grad_norm": 17.239091873168945,
|
| 2577 |
+
"learning_rate": 4.8169874103804335e-06,
|
| 2578 |
+
"loss": 0.5733,
|
| 2579 |
+
"step": 18350
|
| 2580 |
+
},
|
| 2581 |
+
{
|
| 2582 |
+
"epoch": 1.7041770862276557,
|
| 2583 |
+
"grad_norm": 9.486557006835938,
|
| 2584 |
+
"learning_rate": 4.799835340125554e-06,
|
| 2585 |
+
"loss": 0.6062,
|
| 2586 |
+
"step": 18400
|
| 2587 |
+
},
|
| 2588 |
+
{
|
| 2589 |
+
"epoch": 1.7088080022228396,
|
| 2590 |
+
"grad_norm": 8.935564994812012,
|
| 2591 |
+
"learning_rate": 4.782683269870674e-06,
|
| 2592 |
+
"loss": 0.5582,
|
| 2593 |
+
"step": 18450
|
| 2594 |
+
},
|
| 2595 |
+
{
|
| 2596 |
+
"epoch": 1.7134389182180234,
|
| 2597 |
+
"grad_norm": 12.081872940063477,
|
| 2598 |
+
"learning_rate": 4.7655311996157935e-06,
|
| 2599 |
+
"loss": 0.5848,
|
| 2600 |
+
"step": 18500
|
| 2601 |
+
},
|
| 2602 |
+
{
|
| 2603 |
+
"epoch": 1.7180698342132072,
|
| 2604 |
+
"grad_norm": 11.821228981018066,
|
| 2605 |
+
"learning_rate": 4.748379129360914e-06,
|
| 2606 |
+
"loss": 0.537,
|
| 2607 |
+
"step": 18550
|
| 2608 |
+
},
|
| 2609 |
+
{
|
| 2610 |
+
"epoch": 1.722700750208391,
|
| 2611 |
+
"grad_norm": 17.22193145751953,
|
| 2612 |
+
"learning_rate": 4.731227059106035e-06,
|
| 2613 |
+
"loss": 0.5867,
|
| 2614 |
+
"step": 18600
|
| 2615 |
+
},
|
| 2616 |
+
{
|
| 2617 |
+
"epoch": 1.727331666203575,
|
| 2618 |
+
"grad_norm": 14.369434356689453,
|
| 2619 |
+
"learning_rate": 4.714074988851154e-06,
|
| 2620 |
+
"loss": 0.5729,
|
| 2621 |
+
"step": 18650
|
| 2622 |
+
},
|
| 2623 |
+
{
|
| 2624 |
+
"epoch": 1.7319625821987588,
|
| 2625 |
+
"grad_norm": 8.231170654296875,
|
| 2626 |
+
"learning_rate": 4.696922918596275e-06,
|
| 2627 |
+
"loss": 0.6438,
|
| 2628 |
+
"step": 18700
|
| 2629 |
+
},
|
| 2630 |
+
{
|
| 2631 |
+
"epoch": 1.7365934981939426,
|
| 2632 |
+
"grad_norm": 9.67603874206543,
|
| 2633 |
+
"learning_rate": 4.679770848341395e-06,
|
| 2634 |
+
"loss": 0.5475,
|
| 2635 |
+
"step": 18750
|
| 2636 |
+
},
|
| 2637 |
+
{
|
| 2638 |
+
"epoch": 1.7412244141891267,
|
| 2639 |
+
"grad_norm": 13.235024452209473,
|
| 2640 |
+
"learning_rate": 4.662618778086515e-06,
|
| 2641 |
+
"loss": 0.6442,
|
| 2642 |
+
"step": 18800
|
| 2643 |
+
},
|
| 2644 |
+
{
|
| 2645 |
+
"epoch": 1.7458553301843105,
|
| 2646 |
+
"grad_norm": 9.636026382446289,
|
| 2647 |
+
"learning_rate": 4.645466707831636e-06,
|
| 2648 |
+
"loss": 0.5909,
|
| 2649 |
+
"step": 18850
|
| 2650 |
+
},
|
| 2651 |
+
{
|
| 2652 |
+
"epoch": 1.7504862461794943,
|
| 2653 |
+
"grad_norm": 14.982551574707031,
|
| 2654 |
+
"learning_rate": 4.628314637576756e-06,
|
| 2655 |
+
"loss": 0.5283,
|
| 2656 |
+
"step": 18900
|
| 2657 |
+
},
|
| 2658 |
+
{
|
| 2659 |
+
"epoch": 1.7551171621746782,
|
| 2660 |
+
"grad_norm": 13.695237159729004,
|
| 2661 |
+
"learning_rate": 4.611162567321876e-06,
|
| 2662 |
+
"loss": 0.5789,
|
| 2663 |
+
"step": 18950
|
| 2664 |
+
},
|
| 2665 |
+
{
|
| 2666 |
+
"epoch": 1.759748078169862,
|
| 2667 |
+
"grad_norm": 16.833280563354492,
|
| 2668 |
+
"learning_rate": 4.594010497066996e-06,
|
| 2669 |
+
"loss": 0.604,
|
| 2670 |
+
"step": 19000
|
| 2671 |
+
},
|
| 2672 |
+
{
|
| 2673 |
+
"epoch": 1.7643789941650458,
|
| 2674 |
+
"grad_norm": 8.901232719421387,
|
| 2675 |
+
"learning_rate": 4.5768584268121165e-06,
|
| 2676 |
+
"loss": 0.5661,
|
| 2677 |
+
"step": 19050
|
| 2678 |
+
},
|
| 2679 |
+
{
|
| 2680 |
+
"epoch": 1.7690099101602297,
|
| 2681 |
+
"grad_norm": 8.41875171661377,
|
| 2682 |
+
"learning_rate": 4.559706356557237e-06,
|
| 2683 |
+
"loss": 0.5905,
|
| 2684 |
+
"step": 19100
|
| 2685 |
+
},
|
| 2686 |
+
{
|
| 2687 |
+
"epoch": 1.7736408261554135,
|
| 2688 |
+
"grad_norm": 12.987443923950195,
|
| 2689 |
+
"learning_rate": 4.542554286302357e-06,
|
| 2690 |
+
"loss": 0.5465,
|
| 2691 |
+
"step": 19150
|
| 2692 |
+
},
|
| 2693 |
+
{
|
| 2694 |
+
"epoch": 1.7782717421505974,
|
| 2695 |
+
"grad_norm": 7.876839637756348,
|
| 2696 |
+
"learning_rate": 4.525402216047477e-06,
|
| 2697 |
+
"loss": 0.5171,
|
| 2698 |
+
"step": 19200
|
| 2699 |
+
},
|
| 2700 |
+
{
|
| 2701 |
+
"epoch": 1.7829026581457812,
|
| 2702 |
+
"grad_norm": 7.976441860198975,
|
| 2703 |
+
"learning_rate": 4.508250145792598e-06,
|
| 2704 |
+
"loss": 0.5637,
|
| 2705 |
+
"step": 19250
|
| 2706 |
+
},
|
| 2707 |
+
{
|
| 2708 |
+
"epoch": 1.787533574140965,
|
| 2709 |
+
"grad_norm": 11.731218338012695,
|
| 2710 |
+
"learning_rate": 4.491098075537718e-06,
|
| 2711 |
+
"loss": 0.6353,
|
| 2712 |
+
"step": 19300
|
| 2713 |
+
},
|
| 2714 |
+
{
|
| 2715 |
+
"epoch": 1.792164490136149,
|
| 2716 |
+
"grad_norm": 12.95614242553711,
|
| 2717 |
+
"learning_rate": 4.473946005282838e-06,
|
| 2718 |
+
"loss": 0.5705,
|
| 2719 |
+
"step": 19350
|
| 2720 |
+
},
|
| 2721 |
+
{
|
| 2722 |
+
"epoch": 1.796795406131333,
|
| 2723 |
+
"grad_norm": 12.79031753540039,
|
| 2724 |
+
"learning_rate": 4.456793935027958e-06,
|
| 2725 |
+
"loss": 0.6166,
|
| 2726 |
+
"step": 19400
|
| 2727 |
+
},
|
| 2728 |
+
{
|
| 2729 |
+
"epoch": 1.8014263221265168,
|
| 2730 |
+
"grad_norm": 5.699042797088623,
|
| 2731 |
+
"learning_rate": 4.4396418647730786e-06,
|
| 2732 |
+
"loss": 0.5323,
|
| 2733 |
+
"step": 19450
|
| 2734 |
+
},
|
| 2735 |
+
{
|
| 2736 |
+
"epoch": 1.8060572381217006,
|
| 2737 |
+
"grad_norm": 12.937334060668945,
|
| 2738 |
+
"learning_rate": 4.422489794518199e-06,
|
| 2739 |
+
"loss": 0.5709,
|
| 2740 |
+
"step": 19500
|
| 2741 |
+
},
|
| 2742 |
+
{
|
| 2743 |
+
"epoch": 1.8106881541168844,
|
| 2744 |
+
"grad_norm": 14.37783145904541,
|
| 2745 |
+
"learning_rate": 4.405337724263319e-06,
|
| 2746 |
+
"loss": 0.5967,
|
| 2747 |
+
"step": 19550
|
| 2748 |
+
},
|
| 2749 |
+
{
|
| 2750 |
+
"epoch": 1.8153190701120683,
|
| 2751 |
+
"grad_norm": 11.934647560119629,
|
| 2752 |
+
"learning_rate": 4.3881856540084394e-06,
|
| 2753 |
+
"loss": 0.5723,
|
| 2754 |
+
"step": 19600
|
| 2755 |
+
},
|
| 2756 |
+
{
|
| 2757 |
+
"epoch": 1.8199499861072521,
|
| 2758 |
+
"grad_norm": 10.981630325317383,
|
| 2759 |
+
"learning_rate": 4.371033583753559e-06,
|
| 2760 |
+
"loss": 0.5652,
|
| 2761 |
+
"step": 19650
|
| 2762 |
+
},
|
| 2763 |
+
{
|
| 2764 |
+
"epoch": 1.824580902102436,
|
| 2765 |
+
"grad_norm": 17.21076202392578,
|
| 2766 |
+
"learning_rate": 4.35388151349868e-06,
|
| 2767 |
+
"loss": 0.5898,
|
| 2768 |
+
"step": 19700
|
| 2769 |
+
},
|
| 2770 |
+
{
|
| 2771 |
+
"epoch": 1.8292118180976198,
|
| 2772 |
+
"grad_norm": 13.629597663879395,
|
| 2773 |
+
"learning_rate": 4.3367294432438e-06,
|
| 2774 |
+
"loss": 0.6095,
|
| 2775 |
+
"step": 19750
|
| 2776 |
+
},
|
| 2777 |
+
{
|
| 2778 |
+
"epoch": 1.8338427340928036,
|
| 2779 |
+
"grad_norm": 8.095635414123535,
|
| 2780 |
+
"learning_rate": 4.31957737298892e-06,
|
| 2781 |
+
"loss": 0.5825,
|
| 2782 |
+
"step": 19800
|
| 2783 |
+
},
|
| 2784 |
+
{
|
| 2785 |
+
"epoch": 1.8384736500879875,
|
| 2786 |
+
"grad_norm": 8.601015090942383,
|
| 2787 |
+
"learning_rate": 4.30242530273404e-06,
|
| 2788 |
+
"loss": 0.587,
|
| 2789 |
+
"step": 19850
|
| 2790 |
+
},
|
| 2791 |
+
{
|
| 2792 |
+
"epoch": 1.8431045660831713,
|
| 2793 |
+
"grad_norm": 13.996216773986816,
|
| 2794 |
+
"learning_rate": 4.28527323247916e-06,
|
| 2795 |
+
"loss": 0.5412,
|
| 2796 |
+
"step": 19900
|
| 2797 |
+
},
|
| 2798 |
+
{
|
| 2799 |
+
"epoch": 1.8477354820783551,
|
| 2800 |
+
"grad_norm": 7.266920566558838,
|
| 2801 |
+
"learning_rate": 4.268121162224281e-06,
|
| 2802 |
+
"loss": 0.5759,
|
| 2803 |
+
"step": 19950
|
| 2804 |
+
},
|
| 2805 |
+
{
|
| 2806 |
+
"epoch": 1.852366398073539,
|
| 2807 |
+
"grad_norm": 12.852421760559082,
|
| 2808 |
+
"learning_rate": 4.2509690919694015e-06,
|
| 2809 |
+
"loss": 0.559,
|
| 2810 |
+
"step": 20000
|
| 2811 |
+
}
|
| 2812 |
+
],
|
| 2813 |
+
"logging_steps": 50,
|
| 2814 |
+
"max_steps": 32391,
|
| 2815 |
+
"num_input_tokens_seen": 0,
|
| 2816 |
+
"num_train_epochs": 3,
|
| 2817 |
+
"save_steps": 5000,
|
| 2818 |
+
"stateful_callbacks": {
|
| 2819 |
+
"TrainerControl": {
|
| 2820 |
+
"args": {
|
| 2821 |
+
"should_epoch_stop": false,
|
| 2822 |
+
"should_evaluate": false,
|
| 2823 |
+
"should_log": false,
|
| 2824 |
+
"should_save": true,
|
| 2825 |
+
"should_training_stop": false
|
| 2826 |
+
},
|
| 2827 |
+
"attributes": {}
|
| 2828 |
+
}
|
| 2829 |
+
},
|
| 2830 |
+
"total_flos": 0.0,
|
| 2831 |
+
"train_batch_size": 16,
|
| 2832 |
+
"trial_name": null,
|
| 2833 |
+
"trial_params": null
|
| 2834 |
+
}
|
training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:e26fac31905493d62258c683ffafa714e71b8e9dbd3df9ce891301cc5656c7ee
|
| 3 |
+
size 5969
|