Upload 8 files
Browse files- config.json +7 -2
- generation_config.json +1 -1
- model.safetensors +2 -2
- optimizer.pt +3 -0
- rng_state.pth +3 -0
- scheduler.pt +3 -0
- trainer_state.json +2630 -0
- training_args.bin +3 -0
config.json
CHANGED
@@ -1,16 +1,20 @@
|
|
1 |
{
|
|
|
2 |
"architectures": [
|
3 |
"T5ForConditionalGeneration"
|
4 |
],
|
|
|
5 |
"d_ff": 2816,
|
6 |
"d_kv": 64,
|
7 |
"d_model": 1024,
|
8 |
"decoder_start_token_id": 0,
|
|
|
9 |
"dropout_rate": 0.1,
|
10 |
"eos_token_id": 1,
|
11 |
"feed_forward_proj": "gated-gelu",
|
12 |
"initializer_factor": 1.0,
|
13 |
"is_encoder_decoder": true,
|
|
|
14 |
"layer_norm_epsilon": 1e-06,
|
15 |
"model_type": "t5",
|
16 |
"n_positions": 512,
|
@@ -22,7 +26,8 @@
|
|
22 |
"relative_attention_max_distance": 128,
|
23 |
"relative_attention_num_buckets": 32,
|
24 |
"tie_word_embeddings": false,
|
25 |
-
"
|
|
|
26 |
"use_cache": true,
|
27 |
-
"vocab_size":
|
28 |
}
|
|
|
1 |
{
|
2 |
+
"_name_or_path": "AndrewOficial/metrology",
|
3 |
"architectures": [
|
4 |
"T5ForConditionalGeneration"
|
5 |
],
|
6 |
+
"classifier_dropout": 0.0,
|
7 |
"d_ff": 2816,
|
8 |
"d_kv": 64,
|
9 |
"d_model": 1024,
|
10 |
"decoder_start_token_id": 0,
|
11 |
+
"dense_act_fn": "gelu_new",
|
12 |
"dropout_rate": 0.1,
|
13 |
"eos_token_id": 1,
|
14 |
"feed_forward_proj": "gated-gelu",
|
15 |
"initializer_factor": 1.0,
|
16 |
"is_encoder_decoder": true,
|
17 |
+
"is_gated_act": true,
|
18 |
"layer_norm_epsilon": 1e-06,
|
19 |
"model_type": "t5",
|
20 |
"n_positions": 512,
|
|
|
26 |
"relative_attention_max_distance": 128,
|
27 |
"relative_attention_num_buckets": 32,
|
28 |
"tie_word_embeddings": false,
|
29 |
+
"torch_dtype": "float32",
|
30 |
+
"transformers_version": "4.44.2",
|
31 |
"use_cache": true,
|
32 |
+
"vocab_size": 2788
|
33 |
}
|
generation_config.json
CHANGED
@@ -3,5 +3,5 @@
|
|
3 |
"decoder_start_token_id": 0,
|
4 |
"eos_token_id": 1,
|
5 |
"pad_token_id": 0,
|
6 |
-
"transformers_version": "4.
|
7 |
}
|
|
|
3 |
"decoder_start_token_id": 0,
|
4 |
"eos_token_id": 1,
|
5 |
"pad_token_id": 0,
|
6 |
+
"transformers_version": "4.44.2"
|
7 |
}
|
model.safetensors
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
-
size
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:6829bdc431b5624a096b9b65977634ea848472214f88e03b383dfb1f68c8cc00
|
3 |
+
size 2892315528
|
optimizer.pt
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:fdc6ca94e880c2911bf8e187221e120265f344e8963edb2bb9e06b81ccf1f202
|
3 |
+
size 5784969920
|
rng_state.pth
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:311ea95b7728e2cdf457db1738d98fce000acb7dfee84a4db68a81f1be9b5ea0
|
3 |
+
size 14244
|
scheduler.pt
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:f9ddaac7ec3041d112f4f48d29dde8e7471b07b9dda42d84634915c3381e11ec
|
3 |
+
size 1064
|
trainer_state.json
ADDED
@@ -0,0 +1,2630 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"best_metric": null,
|
3 |
+
"best_model_checkpoint": null,
|
4 |
+
"epoch": 10.0,
|
5 |
+
"eval_steps": 500,
|
6 |
+
"global_step": 3710,
|
7 |
+
"is_hyper_param_search": false,
|
8 |
+
"is_local_process_zero": true,
|
9 |
+
"is_world_process_zero": true,
|
10 |
+
"log_history": [
|
11 |
+
{
|
12 |
+
"epoch": 0.026954177897574125,
|
13 |
+
"grad_norm": 93.09680938720703,
|
14 |
+
"learning_rate": 7.692307692307694e-06,
|
15 |
+
"loss": 5.4416,
|
16 |
+
"step": 10
|
17 |
+
},
|
18 |
+
{
|
19 |
+
"epoch": 0.05390835579514825,
|
20 |
+
"grad_norm": 1869.5386962890625,
|
21 |
+
"learning_rate": 1.5384615384615387e-05,
|
22 |
+
"loss": 4.4275,
|
23 |
+
"step": 20
|
24 |
+
},
|
25 |
+
{
|
26 |
+
"epoch": 0.08086253369272237,
|
27 |
+
"grad_norm": 29.11223602294922,
|
28 |
+
"learning_rate": 2.307692307692308e-05,
|
29 |
+
"loss": 2.2122,
|
30 |
+
"step": 30
|
31 |
+
},
|
32 |
+
{
|
33 |
+
"epoch": 0.1078167115902965,
|
34 |
+
"grad_norm": 45.186805725097656,
|
35 |
+
"learning_rate": 3.0769230769230774e-05,
|
36 |
+
"loss": 1.9827,
|
37 |
+
"step": 40
|
38 |
+
},
|
39 |
+
{
|
40 |
+
"epoch": 0.1347708894878706,
|
41 |
+
"grad_norm": 88.7030029296875,
|
42 |
+
"learning_rate": 3.846153846153846e-05,
|
43 |
+
"loss": 1.9525,
|
44 |
+
"step": 50
|
45 |
+
},
|
46 |
+
{
|
47 |
+
"epoch": 0.16172506738544473,
|
48 |
+
"grad_norm": 11.193861961364746,
|
49 |
+
"learning_rate": 4.615384615384616e-05,
|
50 |
+
"loss": 1.401,
|
51 |
+
"step": 60
|
52 |
+
},
|
53 |
+
{
|
54 |
+
"epoch": 0.18867924528301888,
|
55 |
+
"grad_norm": 1.629717469215393,
|
56 |
+
"learning_rate": 4.993141289437586e-05,
|
57 |
+
"loss": 1.4541,
|
58 |
+
"step": 70
|
59 |
+
},
|
60 |
+
{
|
61 |
+
"epoch": 0.215633423180593,
|
62 |
+
"grad_norm": 6.375326156616211,
|
63 |
+
"learning_rate": 4.9794238683127575e-05,
|
64 |
+
"loss": 1.2295,
|
65 |
+
"step": 80
|
66 |
+
},
|
67 |
+
{
|
68 |
+
"epoch": 0.24258760107816713,
|
69 |
+
"grad_norm": 2.1516671180725098,
|
70 |
+
"learning_rate": 4.965706447187929e-05,
|
71 |
+
"loss": 1.8626,
|
72 |
+
"step": 90
|
73 |
+
},
|
74 |
+
{
|
75 |
+
"epoch": 0.2695417789757412,
|
76 |
+
"grad_norm": 2.126923084259033,
|
77 |
+
"learning_rate": 4.9519890260631e-05,
|
78 |
+
"loss": 1.5149,
|
79 |
+
"step": 100
|
80 |
+
},
|
81 |
+
{
|
82 |
+
"epoch": 0.29649595687331537,
|
83 |
+
"grad_norm": 3.0875656604766846,
|
84 |
+
"learning_rate": 4.938271604938271e-05,
|
85 |
+
"loss": 1.7612,
|
86 |
+
"step": 110
|
87 |
+
},
|
88 |
+
{
|
89 |
+
"epoch": 0.32345013477088946,
|
90 |
+
"grad_norm": 3.2731056213378906,
|
91 |
+
"learning_rate": 4.924554183813443e-05,
|
92 |
+
"loss": 1.1813,
|
93 |
+
"step": 120
|
94 |
+
},
|
95 |
+
{
|
96 |
+
"epoch": 0.3504043126684636,
|
97 |
+
"grad_norm": 1.5794568061828613,
|
98 |
+
"learning_rate": 4.9108367626886145e-05,
|
99 |
+
"loss": 1.2228,
|
100 |
+
"step": 130
|
101 |
+
},
|
102 |
+
{
|
103 |
+
"epoch": 0.37735849056603776,
|
104 |
+
"grad_norm": 1.0991185903549194,
|
105 |
+
"learning_rate": 4.8971193415637865e-05,
|
106 |
+
"loss": 2.1294,
|
107 |
+
"step": 140
|
108 |
+
},
|
109 |
+
{
|
110 |
+
"epoch": 0.40431266846361186,
|
111 |
+
"grad_norm": 0.7588101029396057,
|
112 |
+
"learning_rate": 4.883401920438958e-05,
|
113 |
+
"loss": 1.6584,
|
114 |
+
"step": 150
|
115 |
+
},
|
116 |
+
{
|
117 |
+
"epoch": 0.431266846361186,
|
118 |
+
"grad_norm": 0.6242086887359619,
|
119 |
+
"learning_rate": 4.86968449931413e-05,
|
120 |
+
"loss": 1.1689,
|
121 |
+
"step": 160
|
122 |
+
},
|
123 |
+
{
|
124 |
+
"epoch": 0.4582210242587601,
|
125 |
+
"grad_norm": 5.792896270751953,
|
126 |
+
"learning_rate": 4.855967078189301e-05,
|
127 |
+
"loss": 1.3696,
|
128 |
+
"step": 170
|
129 |
+
},
|
130 |
+
{
|
131 |
+
"epoch": 0.48517520215633425,
|
132 |
+
"grad_norm": 2.379605770111084,
|
133 |
+
"learning_rate": 4.842249657064472e-05,
|
134 |
+
"loss": 1.2535,
|
135 |
+
"step": 180
|
136 |
+
},
|
137 |
+
{
|
138 |
+
"epoch": 0.5121293800539084,
|
139 |
+
"grad_norm": 0.7756444811820984,
|
140 |
+
"learning_rate": 4.8285322359396435e-05,
|
141 |
+
"loss": 1.2733,
|
142 |
+
"step": 190
|
143 |
+
},
|
144 |
+
{
|
145 |
+
"epoch": 0.5390835579514824,
|
146 |
+
"grad_norm": 0.9577376246452332,
|
147 |
+
"learning_rate": 4.814814814814815e-05,
|
148 |
+
"loss": 1.366,
|
149 |
+
"step": 200
|
150 |
+
},
|
151 |
+
{
|
152 |
+
"epoch": 0.5660377358490566,
|
153 |
+
"grad_norm": 0.7913112640380859,
|
154 |
+
"learning_rate": 4.801097393689987e-05,
|
155 |
+
"loss": 0.8251,
|
156 |
+
"step": 210
|
157 |
+
},
|
158 |
+
{
|
159 |
+
"epoch": 0.5929919137466307,
|
160 |
+
"grad_norm": 1.1061924695968628,
|
161 |
+
"learning_rate": 4.787379972565158e-05,
|
162 |
+
"loss": 1.0164,
|
163 |
+
"step": 220
|
164 |
+
},
|
165 |
+
{
|
166 |
+
"epoch": 0.6199460916442049,
|
167 |
+
"grad_norm": 0.854999840259552,
|
168 |
+
"learning_rate": 4.773662551440329e-05,
|
169 |
+
"loss": 1.2698,
|
170 |
+
"step": 230
|
171 |
+
},
|
172 |
+
{
|
173 |
+
"epoch": 0.6469002695417789,
|
174 |
+
"grad_norm": 0.9977573156356812,
|
175 |
+
"learning_rate": 4.7599451303155006e-05,
|
176 |
+
"loss": 1.2353,
|
177 |
+
"step": 240
|
178 |
+
},
|
179 |
+
{
|
180 |
+
"epoch": 0.6738544474393531,
|
181 |
+
"grad_norm": 1.3502521514892578,
|
182 |
+
"learning_rate": 4.7462277091906725e-05,
|
183 |
+
"loss": 1.3206,
|
184 |
+
"step": 250
|
185 |
+
},
|
186 |
+
{
|
187 |
+
"epoch": 0.7008086253369272,
|
188 |
+
"grad_norm": 0.7636290192604065,
|
189 |
+
"learning_rate": 4.732510288065844e-05,
|
190 |
+
"loss": 1.2825,
|
191 |
+
"step": 260
|
192 |
+
},
|
193 |
+
{
|
194 |
+
"epoch": 0.7277628032345014,
|
195 |
+
"grad_norm": 1.700835108757019,
|
196 |
+
"learning_rate": 4.718792866941015e-05,
|
197 |
+
"loss": 1.0656,
|
198 |
+
"step": 270
|
199 |
+
},
|
200 |
+
{
|
201 |
+
"epoch": 0.7547169811320755,
|
202 |
+
"grad_norm": 0.7844202518463135,
|
203 |
+
"learning_rate": 4.7050754458161864e-05,
|
204 |
+
"loss": 1.4337,
|
205 |
+
"step": 280
|
206 |
+
},
|
207 |
+
{
|
208 |
+
"epoch": 0.7816711590296496,
|
209 |
+
"grad_norm": 2.4619789123535156,
|
210 |
+
"learning_rate": 4.691358024691358e-05,
|
211 |
+
"loss": 1.2912,
|
212 |
+
"step": 290
|
213 |
+
},
|
214 |
+
{
|
215 |
+
"epoch": 0.8086253369272237,
|
216 |
+
"grad_norm": 2.110649347305298,
|
217 |
+
"learning_rate": 4.6776406035665296e-05,
|
218 |
+
"loss": 1.1539,
|
219 |
+
"step": 300
|
220 |
+
},
|
221 |
+
{
|
222 |
+
"epoch": 0.8355795148247979,
|
223 |
+
"grad_norm": 1.690564751625061,
|
224 |
+
"learning_rate": 4.6639231824417016e-05,
|
225 |
+
"loss": 1.8889,
|
226 |
+
"step": 310
|
227 |
+
},
|
228 |
+
{
|
229 |
+
"epoch": 0.862533692722372,
|
230 |
+
"grad_norm": 2.4102346897125244,
|
231 |
+
"learning_rate": 4.650205761316873e-05,
|
232 |
+
"loss": 0.7576,
|
233 |
+
"step": 320
|
234 |
+
},
|
235 |
+
{
|
236 |
+
"epoch": 0.889487870619946,
|
237 |
+
"grad_norm": 1.0781276226043701,
|
238 |
+
"learning_rate": 4.636488340192044e-05,
|
239 |
+
"loss": 1.167,
|
240 |
+
"step": 330
|
241 |
+
},
|
242 |
+
{
|
243 |
+
"epoch": 0.9164420485175202,
|
244 |
+
"grad_norm": 1.4452636241912842,
|
245 |
+
"learning_rate": 4.622770919067216e-05,
|
246 |
+
"loss": 1.4974,
|
247 |
+
"step": 340
|
248 |
+
},
|
249 |
+
{
|
250 |
+
"epoch": 0.9433962264150944,
|
251 |
+
"grad_norm": 0.7777872681617737,
|
252 |
+
"learning_rate": 4.609053497942387e-05,
|
253 |
+
"loss": 0.7603,
|
254 |
+
"step": 350
|
255 |
+
},
|
256 |
+
{
|
257 |
+
"epoch": 0.9703504043126685,
|
258 |
+
"grad_norm": 1.9176746606826782,
|
259 |
+
"learning_rate": 4.5953360768175586e-05,
|
260 |
+
"loss": 1.2179,
|
261 |
+
"step": 360
|
262 |
+
},
|
263 |
+
{
|
264 |
+
"epoch": 0.9973045822102425,
|
265 |
+
"grad_norm": 0.7132334113121033,
|
266 |
+
"learning_rate": 4.58161865569273e-05,
|
267 |
+
"loss": 1.3606,
|
268 |
+
"step": 370
|
269 |
+
},
|
270 |
+
{
|
271 |
+
"epoch": 1.0242587601078168,
|
272 |
+
"grad_norm": 2.0499353408813477,
|
273 |
+
"learning_rate": 4.567901234567901e-05,
|
274 |
+
"loss": 1.5535,
|
275 |
+
"step": 380
|
276 |
+
},
|
277 |
+
{
|
278 |
+
"epoch": 1.0512129380053907,
|
279 |
+
"grad_norm": 1.237725019454956,
|
280 |
+
"learning_rate": 4.554183813443073e-05,
|
281 |
+
"loss": 1.2777,
|
282 |
+
"step": 390
|
283 |
+
},
|
284 |
+
{
|
285 |
+
"epoch": 1.0781671159029649,
|
286 |
+
"grad_norm": 0.8941486477851868,
|
287 |
+
"learning_rate": 4.5404663923182444e-05,
|
288 |
+
"loss": 1.098,
|
289 |
+
"step": 400
|
290 |
+
},
|
291 |
+
{
|
292 |
+
"epoch": 1.105121293800539,
|
293 |
+
"grad_norm": 3.052698850631714,
|
294 |
+
"learning_rate": 4.5267489711934157e-05,
|
295 |
+
"loss": 0.8203,
|
296 |
+
"step": 410
|
297 |
+
},
|
298 |
+
{
|
299 |
+
"epoch": 1.1320754716981132,
|
300 |
+
"grad_norm": 1.9386261701583862,
|
301 |
+
"learning_rate": 4.513031550068587e-05,
|
302 |
+
"loss": 1.3465,
|
303 |
+
"step": 420
|
304 |
+
},
|
305 |
+
{
|
306 |
+
"epoch": 1.1590296495956873,
|
307 |
+
"grad_norm": 1.5312304496765137,
|
308 |
+
"learning_rate": 4.499314128943759e-05,
|
309 |
+
"loss": 1.5988,
|
310 |
+
"step": 430
|
311 |
+
},
|
312 |
+
{
|
313 |
+
"epoch": 1.1859838274932615,
|
314 |
+
"grad_norm": 1.5553970336914062,
|
315 |
+
"learning_rate": 4.48559670781893e-05,
|
316 |
+
"loss": 1.3158,
|
317 |
+
"step": 440
|
318 |
+
},
|
319 |
+
{
|
320 |
+
"epoch": 1.2129380053908356,
|
321 |
+
"grad_norm": 0.9215840101242065,
|
322 |
+
"learning_rate": 4.4718792866941014e-05,
|
323 |
+
"loss": 1.1357,
|
324 |
+
"step": 450
|
325 |
+
},
|
326 |
+
{
|
327 |
+
"epoch": 1.2398921832884098,
|
328 |
+
"grad_norm": 0.8981826901435852,
|
329 |
+
"learning_rate": 4.4581618655692734e-05,
|
330 |
+
"loss": 1.4697,
|
331 |
+
"step": 460
|
332 |
+
},
|
333 |
+
{
|
334 |
+
"epoch": 1.266846361185984,
|
335 |
+
"grad_norm": 1.3505191802978516,
|
336 |
+
"learning_rate": 4.4444444444444447e-05,
|
337 |
+
"loss": 1.3321,
|
338 |
+
"step": 470
|
339 |
+
},
|
340 |
+
{
|
341 |
+
"epoch": 1.2938005390835579,
|
342 |
+
"grad_norm": 1.836127758026123,
|
343 |
+
"learning_rate": 4.4307270233196166e-05,
|
344 |
+
"loss": 0.7458,
|
345 |
+
"step": 480
|
346 |
+
},
|
347 |
+
{
|
348 |
+
"epoch": 1.320754716981132,
|
349 |
+
"grad_norm": 1.3519006967544556,
|
350 |
+
"learning_rate": 4.417009602194788e-05,
|
351 |
+
"loss": 0.9444,
|
352 |
+
"step": 490
|
353 |
+
},
|
354 |
+
{
|
355 |
+
"epoch": 1.3477088948787062,
|
356 |
+
"grad_norm": 0.8336719870567322,
|
357 |
+
"learning_rate": 4.403292181069959e-05,
|
358 |
+
"loss": 1.2264,
|
359 |
+
"step": 500
|
360 |
+
},
|
361 |
+
{
|
362 |
+
"epoch": 1.3746630727762803,
|
363 |
+
"grad_norm": 2.3419370651245117,
|
364 |
+
"learning_rate": 4.3895747599451304e-05,
|
365 |
+
"loss": 1.4784,
|
366 |
+
"step": 510
|
367 |
+
},
|
368 |
+
{
|
369 |
+
"epoch": 1.4016172506738545,
|
370 |
+
"grad_norm": 1.679734468460083,
|
371 |
+
"learning_rate": 4.3758573388203024e-05,
|
372 |
+
"loss": 0.9167,
|
373 |
+
"step": 520
|
374 |
+
},
|
375 |
+
{
|
376 |
+
"epoch": 1.4285714285714286,
|
377 |
+
"grad_norm": 0.7462900876998901,
|
378 |
+
"learning_rate": 4.3621399176954737e-05,
|
379 |
+
"loss": 0.8118,
|
380 |
+
"step": 530
|
381 |
+
},
|
382 |
+
{
|
383 |
+
"epoch": 1.4555256064690028,
|
384 |
+
"grad_norm": 1.11771559715271,
|
385 |
+
"learning_rate": 4.348422496570645e-05,
|
386 |
+
"loss": 1.2103,
|
387 |
+
"step": 540
|
388 |
+
},
|
389 |
+
{
|
390 |
+
"epoch": 1.482479784366577,
|
391 |
+
"grad_norm": 1.0246716737747192,
|
392 |
+
"learning_rate": 4.334705075445816e-05,
|
393 |
+
"loss": 0.9964,
|
394 |
+
"step": 550
|
395 |
+
},
|
396 |
+
{
|
397 |
+
"epoch": 1.509433962264151,
|
398 |
+
"grad_norm": 1.9072849750518799,
|
399 |
+
"learning_rate": 4.3209876543209875e-05,
|
400 |
+
"loss": 0.9468,
|
401 |
+
"step": 560
|
402 |
+
},
|
403 |
+
{
|
404 |
+
"epoch": 1.536388140161725,
|
405 |
+
"grad_norm": 2.465437889099121,
|
406 |
+
"learning_rate": 4.3072702331961594e-05,
|
407 |
+
"loss": 1.0798,
|
408 |
+
"step": 570
|
409 |
+
},
|
410 |
+
{
|
411 |
+
"epoch": 1.5633423180592994,
|
412 |
+
"grad_norm": 2.021115779876709,
|
413 |
+
"learning_rate": 4.293552812071331e-05,
|
414 |
+
"loss": 1.5825,
|
415 |
+
"step": 580
|
416 |
+
},
|
417 |
+
{
|
418 |
+
"epoch": 1.5902964959568733,
|
419 |
+
"grad_norm": 1.403136968612671,
|
420 |
+
"learning_rate": 4.279835390946502e-05,
|
421 |
+
"loss": 1.0247,
|
422 |
+
"step": 590
|
423 |
+
},
|
424 |
+
{
|
425 |
+
"epoch": 1.6172506738544474,
|
426 |
+
"grad_norm": 2.3880553245544434,
|
427 |
+
"learning_rate": 4.266117969821673e-05,
|
428 |
+
"loss": 0.7598,
|
429 |
+
"step": 600
|
430 |
+
},
|
431 |
+
{
|
432 |
+
"epoch": 1.6442048517520216,
|
433 |
+
"grad_norm": 1.278221607208252,
|
434 |
+
"learning_rate": 4.252400548696845e-05,
|
435 |
+
"loss": 1.2538,
|
436 |
+
"step": 610
|
437 |
+
},
|
438 |
+
{
|
439 |
+
"epoch": 1.6711590296495957,
|
440 |
+
"grad_norm": 1.4277830123901367,
|
441 |
+
"learning_rate": 4.2386831275720165e-05,
|
442 |
+
"loss": 1.0456,
|
443 |
+
"step": 620
|
444 |
+
},
|
445 |
+
{
|
446 |
+
"epoch": 1.6981132075471699,
|
447 |
+
"grad_norm": 1.370888113975525,
|
448 |
+
"learning_rate": 4.2249657064471884e-05,
|
449 |
+
"loss": 1.1437,
|
450 |
+
"step": 630
|
451 |
+
},
|
452 |
+
{
|
453 |
+
"epoch": 1.7250673854447438,
|
454 |
+
"grad_norm": 1.2015599012374878,
|
455 |
+
"learning_rate": 4.21124828532236e-05,
|
456 |
+
"loss": 1.309,
|
457 |
+
"step": 640
|
458 |
+
},
|
459 |
+
{
|
460 |
+
"epoch": 1.7520215633423182,
|
461 |
+
"grad_norm": 0.6258445978164673,
|
462 |
+
"learning_rate": 4.197530864197531e-05,
|
463 |
+
"loss": 0.7159,
|
464 |
+
"step": 650
|
465 |
+
},
|
466 |
+
{
|
467 |
+
"epoch": 1.778975741239892,
|
468 |
+
"grad_norm": 1.4056499004364014,
|
469 |
+
"learning_rate": 4.183813443072703e-05,
|
470 |
+
"loss": 0.9217,
|
471 |
+
"step": 660
|
472 |
+
},
|
473 |
+
{
|
474 |
+
"epoch": 1.8059299191374663,
|
475 |
+
"grad_norm": 2.2113542556762695,
|
476 |
+
"learning_rate": 4.170096021947874e-05,
|
477 |
+
"loss": 0.9003,
|
478 |
+
"step": 670
|
479 |
+
},
|
480 |
+
{
|
481 |
+
"epoch": 1.8328840970350404,
|
482 |
+
"grad_norm": 1.0116914510726929,
|
483 |
+
"learning_rate": 4.1563786008230455e-05,
|
484 |
+
"loss": 1.0246,
|
485 |
+
"step": 680
|
486 |
+
},
|
487 |
+
{
|
488 |
+
"epoch": 1.8598382749326146,
|
489 |
+
"grad_norm": 1.2756783962249756,
|
490 |
+
"learning_rate": 4.142661179698217e-05,
|
491 |
+
"loss": 1.2909,
|
492 |
+
"step": 690
|
493 |
+
},
|
494 |
+
{
|
495 |
+
"epoch": 1.8867924528301887,
|
496 |
+
"grad_norm": 1.2654672861099243,
|
497 |
+
"learning_rate": 4.128943758573389e-05,
|
498 |
+
"loss": 1.2136,
|
499 |
+
"step": 700
|
500 |
+
},
|
501 |
+
{
|
502 |
+
"epoch": 1.9137466307277629,
|
503 |
+
"grad_norm": 0.7074457406997681,
|
504 |
+
"learning_rate": 4.11522633744856e-05,
|
505 |
+
"loss": 1.4202,
|
506 |
+
"step": 710
|
507 |
+
},
|
508 |
+
{
|
509 |
+
"epoch": 1.940700808625337,
|
510 |
+
"grad_norm": 2.5393640995025635,
|
511 |
+
"learning_rate": 4.101508916323731e-05,
|
512 |
+
"loss": 0.959,
|
513 |
+
"step": 720
|
514 |
+
},
|
515 |
+
{
|
516 |
+
"epoch": 1.967654986522911,
|
517 |
+
"grad_norm": 1.194204330444336,
|
518 |
+
"learning_rate": 4.0877914951989025e-05,
|
519 |
+
"loss": 0.7979,
|
520 |
+
"step": 730
|
521 |
+
},
|
522 |
+
{
|
523 |
+
"epoch": 1.9946091644204853,
|
524 |
+
"grad_norm": 2.2830913066864014,
|
525 |
+
"learning_rate": 4.074074074074074e-05,
|
526 |
+
"loss": 0.9024,
|
527 |
+
"step": 740
|
528 |
+
},
|
529 |
+
{
|
530 |
+
"epoch": 2.0215633423180592,
|
531 |
+
"grad_norm": 1.4763576984405518,
|
532 |
+
"learning_rate": 4.060356652949246e-05,
|
533 |
+
"loss": 0.8169,
|
534 |
+
"step": 750
|
535 |
+
},
|
536 |
+
{
|
537 |
+
"epoch": 2.0485175202156336,
|
538 |
+
"grad_norm": 2.3150315284729004,
|
539 |
+
"learning_rate": 4.046639231824417e-05,
|
540 |
+
"loss": 1.1729,
|
541 |
+
"step": 760
|
542 |
+
},
|
543 |
+
{
|
544 |
+
"epoch": 2.0754716981132075,
|
545 |
+
"grad_norm": 1.5322928428649902,
|
546 |
+
"learning_rate": 4.032921810699588e-05,
|
547 |
+
"loss": 0.8163,
|
548 |
+
"step": 770
|
549 |
+
},
|
550 |
+
{
|
551 |
+
"epoch": 2.1024258760107815,
|
552 |
+
"grad_norm": 1.3889187574386597,
|
553 |
+
"learning_rate": 4.01920438957476e-05,
|
554 |
+
"loss": 1.1553,
|
555 |
+
"step": 780
|
556 |
+
},
|
557 |
+
{
|
558 |
+
"epoch": 2.129380053908356,
|
559 |
+
"grad_norm": 1.2974004745483398,
|
560 |
+
"learning_rate": 4.0054869684499315e-05,
|
561 |
+
"loss": 1.5599,
|
562 |
+
"step": 790
|
563 |
+
},
|
564 |
+
{
|
565 |
+
"epoch": 2.1563342318059298,
|
566 |
+
"grad_norm": 1.3070930242538452,
|
567 |
+
"learning_rate": 3.9917695473251035e-05,
|
568 |
+
"loss": 0.8604,
|
569 |
+
"step": 800
|
570 |
+
},
|
571 |
+
{
|
572 |
+
"epoch": 2.183288409703504,
|
573 |
+
"grad_norm": 2.8242597579956055,
|
574 |
+
"learning_rate": 3.978052126200275e-05,
|
575 |
+
"loss": 1.1088,
|
576 |
+
"step": 810
|
577 |
+
},
|
578 |
+
{
|
579 |
+
"epoch": 2.210242587601078,
|
580 |
+
"grad_norm": 1.1547167301177979,
|
581 |
+
"learning_rate": 3.964334705075446e-05,
|
582 |
+
"loss": 0.9212,
|
583 |
+
"step": 820
|
584 |
+
},
|
585 |
+
{
|
586 |
+
"epoch": 2.2371967654986524,
|
587 |
+
"grad_norm": 1.5743651390075684,
|
588 |
+
"learning_rate": 3.950617283950617e-05,
|
589 |
+
"loss": 1.0367,
|
590 |
+
"step": 830
|
591 |
+
},
|
592 |
+
{
|
593 |
+
"epoch": 2.2641509433962264,
|
594 |
+
"grad_norm": 1.0381134748458862,
|
595 |
+
"learning_rate": 3.936899862825789e-05,
|
596 |
+
"loss": 1.1694,
|
597 |
+
"step": 840
|
598 |
+
},
|
599 |
+
{
|
600 |
+
"epoch": 2.2911051212938007,
|
601 |
+
"grad_norm": 1.475040316581726,
|
602 |
+
"learning_rate": 3.9231824417009605e-05,
|
603 |
+
"loss": 0.7031,
|
604 |
+
"step": 850
|
605 |
+
},
|
606 |
+
{
|
607 |
+
"epoch": 2.3180592991913747,
|
608 |
+
"grad_norm": 1.6957072019577026,
|
609 |
+
"learning_rate": 3.909465020576132e-05,
|
610 |
+
"loss": 0.6501,
|
611 |
+
"step": 860
|
612 |
+
},
|
613 |
+
{
|
614 |
+
"epoch": 2.3450134770889486,
|
615 |
+
"grad_norm": 1.2005444765090942,
|
616 |
+
"learning_rate": 3.895747599451303e-05,
|
617 |
+
"loss": 1.1483,
|
618 |
+
"step": 870
|
619 |
+
},
|
620 |
+
{
|
621 |
+
"epoch": 2.371967654986523,
|
622 |
+
"grad_norm": 0.9146741628646851,
|
623 |
+
"learning_rate": 3.8820301783264744e-05,
|
624 |
+
"loss": 0.6192,
|
625 |
+
"step": 880
|
626 |
+
},
|
627 |
+
{
|
628 |
+
"epoch": 2.398921832884097,
|
629 |
+
"grad_norm": 2.0229978561401367,
|
630 |
+
"learning_rate": 3.868312757201646e-05,
|
631 |
+
"loss": 1.0959,
|
632 |
+
"step": 890
|
633 |
+
},
|
634 |
+
{
|
635 |
+
"epoch": 2.4258760107816713,
|
636 |
+
"grad_norm": 1.864313006401062,
|
637 |
+
"learning_rate": 3.8545953360768176e-05,
|
638 |
+
"loss": 1.1319,
|
639 |
+
"step": 900
|
640 |
+
},
|
641 |
+
{
|
642 |
+
"epoch": 2.452830188679245,
|
643 |
+
"grad_norm": 2.050306558609009,
|
644 |
+
"learning_rate": 3.840877914951989e-05,
|
645 |
+
"loss": 0.8544,
|
646 |
+
"step": 910
|
647 |
+
},
|
648 |
+
{
|
649 |
+
"epoch": 2.4797843665768196,
|
650 |
+
"grad_norm": 0.8470428586006165,
|
651 |
+
"learning_rate": 3.82716049382716e-05,
|
652 |
+
"loss": 0.7248,
|
653 |
+
"step": 920
|
654 |
+
},
|
655 |
+
{
|
656 |
+
"epoch": 2.5067385444743935,
|
657 |
+
"grad_norm": 1.329987645149231,
|
658 |
+
"learning_rate": 3.813443072702332e-05,
|
659 |
+
"loss": 0.95,
|
660 |
+
"step": 930
|
661 |
+
},
|
662 |
+
{
|
663 |
+
"epoch": 2.533692722371968,
|
664 |
+
"grad_norm": 1.2895963191986084,
|
665 |
+
"learning_rate": 3.7997256515775034e-05,
|
666 |
+
"loss": 0.8183,
|
667 |
+
"step": 940
|
668 |
+
},
|
669 |
+
{
|
670 |
+
"epoch": 2.560646900269542,
|
671 |
+
"grad_norm": 1.2707000970840454,
|
672 |
+
"learning_rate": 3.786008230452675e-05,
|
673 |
+
"loss": 1.0259,
|
674 |
+
"step": 950
|
675 |
+
},
|
676 |
+
{
|
677 |
+
"epoch": 2.5876010781671157,
|
678 |
+
"grad_norm": 1.5575352907180786,
|
679 |
+
"learning_rate": 3.7722908093278466e-05,
|
680 |
+
"loss": 0.8121,
|
681 |
+
"step": 960
|
682 |
+
},
|
683 |
+
{
|
684 |
+
"epoch": 2.61455525606469,
|
685 |
+
"grad_norm": 1.4395464658737183,
|
686 |
+
"learning_rate": 3.758573388203018e-05,
|
687 |
+
"loss": 0.9093,
|
688 |
+
"step": 970
|
689 |
+
},
|
690 |
+
{
|
691 |
+
"epoch": 2.641509433962264,
|
692 |
+
"grad_norm": 1.8523131608963013,
|
693 |
+
"learning_rate": 3.74485596707819e-05,
|
694 |
+
"loss": 1.4859,
|
695 |
+
"step": 980
|
696 |
+
},
|
697 |
+
{
|
698 |
+
"epoch": 2.6684636118598384,
|
699 |
+
"grad_norm": 1.0986195802688599,
|
700 |
+
"learning_rate": 3.731138545953361e-05,
|
701 |
+
"loss": 0.9775,
|
702 |
+
"step": 990
|
703 |
+
},
|
704 |
+
{
|
705 |
+
"epoch": 2.6954177897574123,
|
706 |
+
"grad_norm": 1.1339648962020874,
|
707 |
+
"learning_rate": 3.7174211248285324e-05,
|
708 |
+
"loss": 1.1358,
|
709 |
+
"step": 1000
|
710 |
+
},
|
711 |
+
{
|
712 |
+
"epoch": 2.7223719676549867,
|
713 |
+
"grad_norm": 2.267329216003418,
|
714 |
+
"learning_rate": 3.7037037037037037e-05,
|
715 |
+
"loss": 0.8947,
|
716 |
+
"step": 1010
|
717 |
+
},
|
718 |
+
{
|
719 |
+
"epoch": 2.7493261455525606,
|
720 |
+
"grad_norm": 1.501046061515808,
|
721 |
+
"learning_rate": 3.6899862825788756e-05,
|
722 |
+
"loss": 1.3137,
|
723 |
+
"step": 1020
|
724 |
+
},
|
725 |
+
{
|
726 |
+
"epoch": 2.776280323450135,
|
727 |
+
"grad_norm": 3.182382345199585,
|
728 |
+
"learning_rate": 3.676268861454047e-05,
|
729 |
+
"loss": 1.4114,
|
730 |
+
"step": 1030
|
731 |
+
},
|
732 |
+
{
|
733 |
+
"epoch": 2.803234501347709,
|
734 |
+
"grad_norm": 1.9981257915496826,
|
735 |
+
"learning_rate": 3.662551440329218e-05,
|
736 |
+
"loss": 1.3174,
|
737 |
+
"step": 1040
|
738 |
+
},
|
739 |
+
{
|
740 |
+
"epoch": 2.830188679245283,
|
741 |
+
"grad_norm": 2.4771640300750732,
|
742 |
+
"learning_rate": 3.6488340192043894e-05,
|
743 |
+
"loss": 1.7184,
|
744 |
+
"step": 1050
|
745 |
+
},
|
746 |
+
{
|
747 |
+
"epoch": 2.857142857142857,
|
748 |
+
"grad_norm": 1.307492971420288,
|
749 |
+
"learning_rate": 3.635116598079561e-05,
|
750 |
+
"loss": 1.0643,
|
751 |
+
"step": 1060
|
752 |
+
},
|
753 |
+
{
|
754 |
+
"epoch": 2.884097035040431,
|
755 |
+
"grad_norm": 0.7365540862083435,
|
756 |
+
"learning_rate": 3.6213991769547327e-05,
|
757 |
+
"loss": 0.9895,
|
758 |
+
"step": 1070
|
759 |
+
},
|
760 |
+
{
|
761 |
+
"epoch": 2.9110512129380055,
|
762 |
+
"grad_norm": 1.7630441188812256,
|
763 |
+
"learning_rate": 3.607681755829904e-05,
|
764 |
+
"loss": 1.0862,
|
765 |
+
"step": 1080
|
766 |
+
},
|
767 |
+
{
|
768 |
+
"epoch": 2.9380053908355794,
|
769 |
+
"grad_norm": 1.731484055519104,
|
770 |
+
"learning_rate": 3.593964334705075e-05,
|
771 |
+
"loss": 0.7582,
|
772 |
+
"step": 1090
|
773 |
+
},
|
774 |
+
{
|
775 |
+
"epoch": 2.964959568733154,
|
776 |
+
"grad_norm": 1.7666038274765015,
|
777 |
+
"learning_rate": 3.580246913580247e-05,
|
778 |
+
"loss": 0.9193,
|
779 |
+
"step": 1100
|
780 |
+
},
|
781 |
+
{
|
782 |
+
"epoch": 2.9919137466307277,
|
783 |
+
"grad_norm": 1.3342130184173584,
|
784 |
+
"learning_rate": 3.566529492455419e-05,
|
785 |
+
"loss": 0.9391,
|
786 |
+
"step": 1110
|
787 |
+
},
|
788 |
+
{
|
789 |
+
"epoch": 3.018867924528302,
|
790 |
+
"grad_norm": 1.7891361713409424,
|
791 |
+
"learning_rate": 3.5528120713305904e-05,
|
792 |
+
"loss": 0.4177,
|
793 |
+
"step": 1120
|
794 |
+
},
|
795 |
+
{
|
796 |
+
"epoch": 3.045822102425876,
|
797 |
+
"grad_norm": 1.730653166770935,
|
798 |
+
"learning_rate": 3.539094650205762e-05,
|
799 |
+
"loss": 1.4771,
|
800 |
+
"step": 1130
|
801 |
+
},
|
802 |
+
{
|
803 |
+
"epoch": 3.07277628032345,
|
804 |
+
"grad_norm": 1.5244202613830566,
|
805 |
+
"learning_rate": 3.525377229080933e-05,
|
806 |
+
"loss": 0.6919,
|
807 |
+
"step": 1140
|
808 |
+
},
|
809 |
+
{
|
810 |
+
"epoch": 3.0997304582210243,
|
811 |
+
"grad_norm": 1.608145833015442,
|
812 |
+
"learning_rate": 3.511659807956104e-05,
|
813 |
+
"loss": 1.0934,
|
814 |
+
"step": 1150
|
815 |
+
},
|
816 |
+
{
|
817 |
+
"epoch": 3.1266846361185983,
|
818 |
+
"grad_norm": 2.0203192234039307,
|
819 |
+
"learning_rate": 3.497942386831276e-05,
|
820 |
+
"loss": 0.8308,
|
821 |
+
"step": 1160
|
822 |
+
},
|
823 |
+
{
|
824 |
+
"epoch": 3.1536388140161726,
|
825 |
+
"grad_norm": 1.7629802227020264,
|
826 |
+
"learning_rate": 3.4842249657064474e-05,
|
827 |
+
"loss": 0.8283,
|
828 |
+
"step": 1170
|
829 |
+
},
|
830 |
+
{
|
831 |
+
"epoch": 3.1805929919137466,
|
832 |
+
"grad_norm": 1.6114568710327148,
|
833 |
+
"learning_rate": 3.470507544581619e-05,
|
834 |
+
"loss": 1.1672,
|
835 |
+
"step": 1180
|
836 |
+
},
|
837 |
+
{
|
838 |
+
"epoch": 3.207547169811321,
|
839 |
+
"grad_norm": 1.5010507106781006,
|
840 |
+
"learning_rate": 3.45679012345679e-05,
|
841 |
+
"loss": 1.5866,
|
842 |
+
"step": 1190
|
843 |
+
},
|
844 |
+
{
|
845 |
+
"epoch": 3.234501347708895,
|
846 |
+
"grad_norm": 1.4426625967025757,
|
847 |
+
"learning_rate": 3.443072702331962e-05,
|
848 |
+
"loss": 0.8659,
|
849 |
+
"step": 1200
|
850 |
+
},
|
851 |
+
{
|
852 |
+
"epoch": 3.2614555256064692,
|
853 |
+
"grad_norm": 1.3204811811447144,
|
854 |
+
"learning_rate": 3.429355281207133e-05,
|
855 |
+
"loss": 0.643,
|
856 |
+
"step": 1210
|
857 |
+
},
|
858 |
+
{
|
859 |
+
"epoch": 3.288409703504043,
|
860 |
+
"grad_norm": 0.9203002452850342,
|
861 |
+
"learning_rate": 3.4156378600823045e-05,
|
862 |
+
"loss": 0.6825,
|
863 |
+
"step": 1220
|
864 |
+
},
|
865 |
+
{
|
866 |
+
"epoch": 3.315363881401617,
|
867 |
+
"grad_norm": 1.3211963176727295,
|
868 |
+
"learning_rate": 3.401920438957476e-05,
|
869 |
+
"loss": 1.1164,
|
870 |
+
"step": 1230
|
871 |
+
},
|
872 |
+
{
|
873 |
+
"epoch": 3.3423180592991915,
|
874 |
+
"grad_norm": 1.4665788412094116,
|
875 |
+
"learning_rate": 3.388203017832647e-05,
|
876 |
+
"loss": 0.7564,
|
877 |
+
"step": 1240
|
878 |
+
},
|
879 |
+
{
|
880 |
+
"epoch": 3.3692722371967654,
|
881 |
+
"grad_norm": 1.3217601776123047,
|
882 |
+
"learning_rate": 3.374485596707819e-05,
|
883 |
+
"loss": 0.8292,
|
884 |
+
"step": 1250
|
885 |
+
},
|
886 |
+
{
|
887 |
+
"epoch": 3.3962264150943398,
|
888 |
+
"grad_norm": 1.1328483819961548,
|
889 |
+
"learning_rate": 3.360768175582991e-05,
|
890 |
+
"loss": 0.9806,
|
891 |
+
"step": 1260
|
892 |
+
},
|
893 |
+
{
|
894 |
+
"epoch": 3.4231805929919137,
|
895 |
+
"grad_norm": 1.3456640243530273,
|
896 |
+
"learning_rate": 3.347050754458162e-05,
|
897 |
+
"loss": 0.8297,
|
898 |
+
"step": 1270
|
899 |
+
},
|
900 |
+
{
|
901 |
+
"epoch": 3.450134770889488,
|
902 |
+
"grad_norm": 0.9857800006866455,
|
903 |
+
"learning_rate": 3.3333333333333335e-05,
|
904 |
+
"loss": 0.6432,
|
905 |
+
"step": 1280
|
906 |
+
},
|
907 |
+
{
|
908 |
+
"epoch": 3.477088948787062,
|
909 |
+
"grad_norm": 1.5837385654449463,
|
910 |
+
"learning_rate": 3.3196159122085054e-05,
|
911 |
+
"loss": 0.6766,
|
912 |
+
"step": 1290
|
913 |
+
},
|
914 |
+
{
|
915 |
+
"epoch": 3.5040431266846364,
|
916 |
+
"grad_norm": 0.9640551805496216,
|
917 |
+
"learning_rate": 3.305898491083677e-05,
|
918 |
+
"loss": 0.9636,
|
919 |
+
"step": 1300
|
920 |
+
},
|
921 |
+
{
|
922 |
+
"epoch": 3.5309973045822103,
|
923 |
+
"grad_norm": 1.6151617765426636,
|
924 |
+
"learning_rate": 3.292181069958848e-05,
|
925 |
+
"loss": 0.5541,
|
926 |
+
"step": 1310
|
927 |
+
},
|
928 |
+
{
|
929 |
+
"epoch": 3.557951482479784,
|
930 |
+
"grad_norm": 1.0056880712509155,
|
931 |
+
"learning_rate": 3.278463648834019e-05,
|
932 |
+
"loss": 0.8266,
|
933 |
+
"step": 1320
|
934 |
+
},
|
935 |
+
{
|
936 |
+
"epoch": 3.5849056603773586,
|
937 |
+
"grad_norm": 2.169956684112549,
|
938 |
+
"learning_rate": 3.2647462277091905e-05,
|
939 |
+
"loss": 1.4843,
|
940 |
+
"step": 1330
|
941 |
+
},
|
942 |
+
{
|
943 |
+
"epoch": 3.6118598382749325,
|
944 |
+
"grad_norm": 1.395251989364624,
|
945 |
+
"learning_rate": 3.2510288065843625e-05,
|
946 |
+
"loss": 0.4211,
|
947 |
+
"step": 1340
|
948 |
+
},
|
949 |
+
{
|
950 |
+
"epoch": 3.638814016172507,
|
951 |
+
"grad_norm": 2.2781965732574463,
|
952 |
+
"learning_rate": 3.237311385459534e-05,
|
953 |
+
"loss": 0.799,
|
954 |
+
"step": 1350
|
955 |
+
},
|
956 |
+
{
|
957 |
+
"epoch": 3.665768194070081,
|
958 |
+
"grad_norm": 2.2836267948150635,
|
959 |
+
"learning_rate": 3.223593964334705e-05,
|
960 |
+
"loss": 0.7618,
|
961 |
+
"step": 1360
|
962 |
+
},
|
963 |
+
{
|
964 |
+
"epoch": 3.6927223719676547,
|
965 |
+
"grad_norm": 2.4063193798065186,
|
966 |
+
"learning_rate": 3.209876543209876e-05,
|
967 |
+
"loss": 1.3887,
|
968 |
+
"step": 1370
|
969 |
+
},
|
970 |
+
{
|
971 |
+
"epoch": 3.719676549865229,
|
972 |
+
"grad_norm": 1.7423830032348633,
|
973 |
+
"learning_rate": 3.196159122085048e-05,
|
974 |
+
"loss": 0.8476,
|
975 |
+
"step": 1380
|
976 |
+
},
|
977 |
+
{
|
978 |
+
"epoch": 3.7466307277628035,
|
979 |
+
"grad_norm": 2.047558307647705,
|
980 |
+
"learning_rate": 3.1824417009602195e-05,
|
981 |
+
"loss": 1.0701,
|
982 |
+
"step": 1390
|
983 |
+
},
|
984 |
+
{
|
985 |
+
"epoch": 3.7735849056603774,
|
986 |
+
"grad_norm": 1.421557068824768,
|
987 |
+
"learning_rate": 3.168724279835391e-05,
|
988 |
+
"loss": 1.6139,
|
989 |
+
"step": 1400
|
990 |
+
},
|
991 |
+
{
|
992 |
+
"epoch": 3.8005390835579513,
|
993 |
+
"grad_norm": 1.6829692125320435,
|
994 |
+
"learning_rate": 3.155006858710563e-05,
|
995 |
+
"loss": 0.8253,
|
996 |
+
"step": 1410
|
997 |
+
},
|
998 |
+
{
|
999 |
+
"epoch": 3.8274932614555257,
|
1000 |
+
"grad_norm": 0.9380494356155396,
|
1001 |
+
"learning_rate": 3.141289437585734e-05,
|
1002 |
+
"loss": 1.0122,
|
1003 |
+
"step": 1420
|
1004 |
+
},
|
1005 |
+
{
|
1006 |
+
"epoch": 3.8544474393530996,
|
1007 |
+
"grad_norm": 1.209517478942871,
|
1008 |
+
"learning_rate": 3.127572016460906e-05,
|
1009 |
+
"loss": 1.6921,
|
1010 |
+
"step": 1430
|
1011 |
+
},
|
1012 |
+
{
|
1013 |
+
"epoch": 3.881401617250674,
|
1014 |
+
"grad_norm": 1.2071969509124756,
|
1015 |
+
"learning_rate": 3.113854595336077e-05,
|
1016 |
+
"loss": 1.3358,
|
1017 |
+
"step": 1440
|
1018 |
+
},
|
1019 |
+
{
|
1020 |
+
"epoch": 3.908355795148248,
|
1021 |
+
"grad_norm": 1.6536906957626343,
|
1022 |
+
"learning_rate": 3.1001371742112486e-05,
|
1023 |
+
"loss": 0.7539,
|
1024 |
+
"step": 1450
|
1025 |
+
},
|
1026 |
+
{
|
1027 |
+
"epoch": 3.935309973045822,
|
1028 |
+
"grad_norm": 3.013587713241577,
|
1029 |
+
"learning_rate": 3.08641975308642e-05,
|
1030 |
+
"loss": 1.5292,
|
1031 |
+
"step": 1460
|
1032 |
+
},
|
1033 |
+
{
|
1034 |
+
"epoch": 3.9622641509433962,
|
1035 |
+
"grad_norm": 1.5310529470443726,
|
1036 |
+
"learning_rate": 3.072702331961592e-05,
|
1037 |
+
"loss": 0.8488,
|
1038 |
+
"step": 1470
|
1039 |
+
},
|
1040 |
+
{
|
1041 |
+
"epoch": 3.9892183288409706,
|
1042 |
+
"grad_norm": 1.6104665994644165,
|
1043 |
+
"learning_rate": 3.058984910836763e-05,
|
1044 |
+
"loss": 0.5511,
|
1045 |
+
"step": 1480
|
1046 |
+
},
|
1047 |
+
{
|
1048 |
+
"epoch": 4.0161725067385445,
|
1049 |
+
"grad_norm": 1.7617496252059937,
|
1050 |
+
"learning_rate": 3.0452674897119343e-05,
|
1051 |
+
"loss": 1.5566,
|
1052 |
+
"step": 1490
|
1053 |
+
},
|
1054 |
+
{
|
1055 |
+
"epoch": 4.0431266846361185,
|
1056 |
+
"grad_norm": 1.6060303449630737,
|
1057 |
+
"learning_rate": 3.0315500685871056e-05,
|
1058 |
+
"loss": 0.6705,
|
1059 |
+
"step": 1500
|
1060 |
+
},
|
1061 |
+
{
|
1062 |
+
"epoch": 4.070080862533692,
|
1063 |
+
"grad_norm": 1.2487995624542236,
|
1064 |
+
"learning_rate": 3.017832647462277e-05,
|
1065 |
+
"loss": 0.8386,
|
1066 |
+
"step": 1510
|
1067 |
+
},
|
1068 |
+
{
|
1069 |
+
"epoch": 4.097035040431267,
|
1070 |
+
"grad_norm": 1.5747429132461548,
|
1071 |
+
"learning_rate": 3.0041152263374488e-05,
|
1072 |
+
"loss": 1.5059,
|
1073 |
+
"step": 1520
|
1074 |
+
},
|
1075 |
+
{
|
1076 |
+
"epoch": 4.123989218328841,
|
1077 |
+
"grad_norm": 2.865582227706909,
|
1078 |
+
"learning_rate": 2.99039780521262e-05,
|
1079 |
+
"loss": 0.801,
|
1080 |
+
"step": 1530
|
1081 |
+
},
|
1082 |
+
{
|
1083 |
+
"epoch": 4.150943396226415,
|
1084 |
+
"grad_norm": 2.48677659034729,
|
1085 |
+
"learning_rate": 2.9766803840877917e-05,
|
1086 |
+
"loss": 1.1271,
|
1087 |
+
"step": 1540
|
1088 |
+
},
|
1089 |
+
{
|
1090 |
+
"epoch": 4.177897574123989,
|
1091 |
+
"grad_norm": 2.485933303833008,
|
1092 |
+
"learning_rate": 2.962962962962963e-05,
|
1093 |
+
"loss": 1.0503,
|
1094 |
+
"step": 1550
|
1095 |
+
},
|
1096 |
+
{
|
1097 |
+
"epoch": 4.204851752021563,
|
1098 |
+
"grad_norm": 2.0622434616088867,
|
1099 |
+
"learning_rate": 2.949245541838135e-05,
|
1100 |
+
"loss": 0.5068,
|
1101 |
+
"step": 1560
|
1102 |
+
},
|
1103 |
+
{
|
1104 |
+
"epoch": 4.231805929919138,
|
1105 |
+
"grad_norm": 0.7956051826477051,
|
1106 |
+
"learning_rate": 2.9355281207133062e-05,
|
1107 |
+
"loss": 1.0424,
|
1108 |
+
"step": 1570
|
1109 |
+
},
|
1110 |
+
{
|
1111 |
+
"epoch": 4.258760107816712,
|
1112 |
+
"grad_norm": 1.3042256832122803,
|
1113 |
+
"learning_rate": 2.9218106995884775e-05,
|
1114 |
+
"loss": 0.5118,
|
1115 |
+
"step": 1580
|
1116 |
+
},
|
1117 |
+
{
|
1118 |
+
"epoch": 4.285714285714286,
|
1119 |
+
"grad_norm": 2.2554943561553955,
|
1120 |
+
"learning_rate": 2.9080932784636488e-05,
|
1121 |
+
"loss": 0.9746,
|
1122 |
+
"step": 1590
|
1123 |
+
},
|
1124 |
+
{
|
1125 |
+
"epoch": 4.3126684636118595,
|
1126 |
+
"grad_norm": 2.549921751022339,
|
1127 |
+
"learning_rate": 2.8943758573388204e-05,
|
1128 |
+
"loss": 0.7228,
|
1129 |
+
"step": 1600
|
1130 |
+
},
|
1131 |
+
{
|
1132 |
+
"epoch": 4.339622641509434,
|
1133 |
+
"grad_norm": 1.333503246307373,
|
1134 |
+
"learning_rate": 2.880658436213992e-05,
|
1135 |
+
"loss": 0.9424,
|
1136 |
+
"step": 1610
|
1137 |
+
},
|
1138 |
+
{
|
1139 |
+
"epoch": 4.366576819407008,
|
1140 |
+
"grad_norm": 1.5245575904846191,
|
1141 |
+
"learning_rate": 2.8669410150891636e-05,
|
1142 |
+
"loss": 1.0232,
|
1143 |
+
"step": 1620
|
1144 |
+
},
|
1145 |
+
{
|
1146 |
+
"epoch": 4.393530997304582,
|
1147 |
+
"grad_norm": 5.889776229858398,
|
1148 |
+
"learning_rate": 2.853223593964335e-05,
|
1149 |
+
"loss": 0.9108,
|
1150 |
+
"step": 1630
|
1151 |
+
},
|
1152 |
+
{
|
1153 |
+
"epoch": 4.420485175202156,
|
1154 |
+
"grad_norm": 1.444460153579712,
|
1155 |
+
"learning_rate": 2.839506172839506e-05,
|
1156 |
+
"loss": 0.7664,
|
1157 |
+
"step": 1640
|
1158 |
+
},
|
1159 |
+
{
|
1160 |
+
"epoch": 4.44743935309973,
|
1161 |
+
"grad_norm": 2.633338689804077,
|
1162 |
+
"learning_rate": 2.825788751714678e-05,
|
1163 |
+
"loss": 0.9278,
|
1164 |
+
"step": 1650
|
1165 |
+
},
|
1166 |
+
{
|
1167 |
+
"epoch": 4.474393530997305,
|
1168 |
+
"grad_norm": 0.8140472769737244,
|
1169 |
+
"learning_rate": 2.8120713305898494e-05,
|
1170 |
+
"loss": 0.7608,
|
1171 |
+
"step": 1660
|
1172 |
+
},
|
1173 |
+
{
|
1174 |
+
"epoch": 4.501347708894879,
|
1175 |
+
"grad_norm": 1.0764524936676025,
|
1176 |
+
"learning_rate": 2.7983539094650207e-05,
|
1177 |
+
"loss": 0.8768,
|
1178 |
+
"step": 1670
|
1179 |
+
},
|
1180 |
+
{
|
1181 |
+
"epoch": 4.528301886792453,
|
1182 |
+
"grad_norm": 1.2656837701797485,
|
1183 |
+
"learning_rate": 2.784636488340192e-05,
|
1184 |
+
"loss": 1.2142,
|
1185 |
+
"step": 1680
|
1186 |
+
},
|
1187 |
+
{
|
1188 |
+
"epoch": 4.555256064690027,
|
1189 |
+
"grad_norm": 1.1997019052505493,
|
1190 |
+
"learning_rate": 2.7709190672153635e-05,
|
1191 |
+
"loss": 1.1219,
|
1192 |
+
"step": 1690
|
1193 |
+
},
|
1194 |
+
{
|
1195 |
+
"epoch": 4.5822102425876015,
|
1196 |
+
"grad_norm": 1.1328818798065186,
|
1197 |
+
"learning_rate": 2.757201646090535e-05,
|
1198 |
+
"loss": 0.7957,
|
1199 |
+
"step": 1700
|
1200 |
+
},
|
1201 |
+
{
|
1202 |
+
"epoch": 4.609164420485175,
|
1203 |
+
"grad_norm": 0.9673919677734375,
|
1204 |
+
"learning_rate": 2.7434842249657068e-05,
|
1205 |
+
"loss": 0.833,
|
1206 |
+
"step": 1710
|
1207 |
+
},
|
1208 |
+
{
|
1209 |
+
"epoch": 4.636118598382749,
|
1210 |
+
"grad_norm": 1.556420087814331,
|
1211 |
+
"learning_rate": 2.729766803840878e-05,
|
1212 |
+
"loss": 0.644,
|
1213 |
+
"step": 1720
|
1214 |
+
},
|
1215 |
+
{
|
1216 |
+
"epoch": 4.663072776280323,
|
1217 |
+
"grad_norm": 1.8422091007232666,
|
1218 |
+
"learning_rate": 2.7160493827160493e-05,
|
1219 |
+
"loss": 1.0444,
|
1220 |
+
"step": 1730
|
1221 |
+
},
|
1222 |
+
{
|
1223 |
+
"epoch": 4.690026954177897,
|
1224 |
+
"grad_norm": 1.5853588581085205,
|
1225 |
+
"learning_rate": 2.7023319615912206e-05,
|
1226 |
+
"loss": 0.8302,
|
1227 |
+
"step": 1740
|
1228 |
+
},
|
1229 |
+
{
|
1230 |
+
"epoch": 4.716981132075472,
|
1231 |
+
"grad_norm": 2.221689462661743,
|
1232 |
+
"learning_rate": 2.6886145404663926e-05,
|
1233 |
+
"loss": 0.8505,
|
1234 |
+
"step": 1750
|
1235 |
+
},
|
1236 |
+
{
|
1237 |
+
"epoch": 4.743935309973046,
|
1238 |
+
"grad_norm": 1.9763739109039307,
|
1239 |
+
"learning_rate": 2.6748971193415638e-05,
|
1240 |
+
"loss": 0.7795,
|
1241 |
+
"step": 1760
|
1242 |
+
},
|
1243 |
+
{
|
1244 |
+
"epoch": 4.77088948787062,
|
1245 |
+
"grad_norm": 2.094904661178589,
|
1246 |
+
"learning_rate": 2.6611796982167354e-05,
|
1247 |
+
"loss": 0.6406,
|
1248 |
+
"step": 1770
|
1249 |
+
},
|
1250 |
+
{
|
1251 |
+
"epoch": 4.797843665768194,
|
1252 |
+
"grad_norm": 2.045687437057495,
|
1253 |
+
"learning_rate": 2.6474622770919067e-05,
|
1254 |
+
"loss": 0.7323,
|
1255 |
+
"step": 1780
|
1256 |
+
},
|
1257 |
+
{
|
1258 |
+
"epoch": 4.824797843665769,
|
1259 |
+
"grad_norm": 2.2809979915618896,
|
1260 |
+
"learning_rate": 2.6337448559670787e-05,
|
1261 |
+
"loss": 0.783,
|
1262 |
+
"step": 1790
|
1263 |
+
},
|
1264 |
+
{
|
1265 |
+
"epoch": 4.8517520215633425,
|
1266 |
+
"grad_norm": 1.7300888299942017,
|
1267 |
+
"learning_rate": 2.62002743484225e-05,
|
1268 |
+
"loss": 0.6659,
|
1269 |
+
"step": 1800
|
1270 |
+
},
|
1271 |
+
{
|
1272 |
+
"epoch": 4.878706199460916,
|
1273 |
+
"grad_norm": 3.9680426120758057,
|
1274 |
+
"learning_rate": 2.6063100137174212e-05,
|
1275 |
+
"loss": 0.6866,
|
1276 |
+
"step": 1810
|
1277 |
+
},
|
1278 |
+
{
|
1279 |
+
"epoch": 4.90566037735849,
|
1280 |
+
"grad_norm": 1.8751779794692993,
|
1281 |
+
"learning_rate": 2.5925925925925925e-05,
|
1282 |
+
"loss": 0.9165,
|
1283 |
+
"step": 1820
|
1284 |
+
},
|
1285 |
+
{
|
1286 |
+
"epoch": 4.932614555256064,
|
1287 |
+
"grad_norm": 1.671337366104126,
|
1288 |
+
"learning_rate": 2.5788751714677638e-05,
|
1289 |
+
"loss": 0.846,
|
1290 |
+
"step": 1830
|
1291 |
+
},
|
1292 |
+
{
|
1293 |
+
"epoch": 4.959568733153639,
|
1294 |
+
"grad_norm": 2.150132179260254,
|
1295 |
+
"learning_rate": 2.5651577503429357e-05,
|
1296 |
+
"loss": 1.2381,
|
1297 |
+
"step": 1840
|
1298 |
+
},
|
1299 |
+
{
|
1300 |
+
"epoch": 4.986522911051213,
|
1301 |
+
"grad_norm": 1.2727786302566528,
|
1302 |
+
"learning_rate": 2.551440329218107e-05,
|
1303 |
+
"loss": 0.7399,
|
1304 |
+
"step": 1850
|
1305 |
+
},
|
1306 |
+
{
|
1307 |
+
"epoch": 5.013477088948787,
|
1308 |
+
"grad_norm": 1.6058365106582642,
|
1309 |
+
"learning_rate": 2.5377229080932786e-05,
|
1310 |
+
"loss": 1.3821,
|
1311 |
+
"step": 1860
|
1312 |
+
},
|
1313 |
+
{
|
1314 |
+
"epoch": 5.040431266846361,
|
1315 |
+
"grad_norm": 0.873196542263031,
|
1316 |
+
"learning_rate": 2.52400548696845e-05,
|
1317 |
+
"loss": 1.0094,
|
1318 |
+
"step": 1870
|
1319 |
+
},
|
1320 |
+
{
|
1321 |
+
"epoch": 5.067385444743936,
|
1322 |
+
"grad_norm": 2.6777074337005615,
|
1323 |
+
"learning_rate": 2.510288065843622e-05,
|
1324 |
+
"loss": 0.956,
|
1325 |
+
"step": 1880
|
1326 |
+
},
|
1327 |
+
{
|
1328 |
+
"epoch": 5.09433962264151,
|
1329 |
+
"grad_norm": 2.2035608291625977,
|
1330 |
+
"learning_rate": 2.496570644718793e-05,
|
1331 |
+
"loss": 1.0327,
|
1332 |
+
"step": 1890
|
1333 |
+
},
|
1334 |
+
{
|
1335 |
+
"epoch": 5.121293800539084,
|
1336 |
+
"grad_norm": 1.9999809265136719,
|
1337 |
+
"learning_rate": 2.4828532235939644e-05,
|
1338 |
+
"loss": 0.9534,
|
1339 |
+
"step": 1900
|
1340 |
+
},
|
1341 |
+
{
|
1342 |
+
"epoch": 5.1482479784366575,
|
1343 |
+
"grad_norm": 1.1194771528244019,
|
1344 |
+
"learning_rate": 2.4691358024691357e-05,
|
1345 |
+
"loss": 0.6969,
|
1346 |
+
"step": 1910
|
1347 |
+
},
|
1348 |
+
{
|
1349 |
+
"epoch": 5.175202156334231,
|
1350 |
+
"grad_norm": 2.1058928966522217,
|
1351 |
+
"learning_rate": 2.4554183813443073e-05,
|
1352 |
+
"loss": 1.077,
|
1353 |
+
"step": 1920
|
1354 |
+
},
|
1355 |
+
{
|
1356 |
+
"epoch": 5.202156334231806,
|
1357 |
+
"grad_norm": 1.584088921546936,
|
1358 |
+
"learning_rate": 2.441700960219479e-05,
|
1359 |
+
"loss": 1.0808,
|
1360 |
+
"step": 1930
|
1361 |
+
},
|
1362 |
+
{
|
1363 |
+
"epoch": 5.22911051212938,
|
1364 |
+
"grad_norm": 2.4530153274536133,
|
1365 |
+
"learning_rate": 2.4279835390946505e-05,
|
1366 |
+
"loss": 1.1974,
|
1367 |
+
"step": 1940
|
1368 |
+
},
|
1369 |
+
{
|
1370 |
+
"epoch": 5.256064690026954,
|
1371 |
+
"grad_norm": 1.3323501348495483,
|
1372 |
+
"learning_rate": 2.4142661179698218e-05,
|
1373 |
+
"loss": 1.0003,
|
1374 |
+
"step": 1950
|
1375 |
+
},
|
1376 |
+
{
|
1377 |
+
"epoch": 5.283018867924528,
|
1378 |
+
"grad_norm": 2.1293039321899414,
|
1379 |
+
"learning_rate": 2.4005486968449934e-05,
|
1380 |
+
"loss": 1.2501,
|
1381 |
+
"step": 1960
|
1382 |
+
},
|
1383 |
+
{
|
1384 |
+
"epoch": 5.309973045822103,
|
1385 |
+
"grad_norm": 2.1004316806793213,
|
1386 |
+
"learning_rate": 2.3868312757201647e-05,
|
1387 |
+
"loss": 1.0852,
|
1388 |
+
"step": 1970
|
1389 |
+
},
|
1390 |
+
{
|
1391 |
+
"epoch": 5.336927223719677,
|
1392 |
+
"grad_norm": 2.6623051166534424,
|
1393 |
+
"learning_rate": 2.3731138545953363e-05,
|
1394 |
+
"loss": 0.6813,
|
1395 |
+
"step": 1980
|
1396 |
+
},
|
1397 |
+
{
|
1398 |
+
"epoch": 5.363881401617251,
|
1399 |
+
"grad_norm": 1.0680763721466064,
|
1400 |
+
"learning_rate": 2.3593964334705075e-05,
|
1401 |
+
"loss": 0.6413,
|
1402 |
+
"step": 1990
|
1403 |
+
},
|
1404 |
+
{
|
1405 |
+
"epoch": 5.390835579514825,
|
1406 |
+
"grad_norm": 1.4404453039169312,
|
1407 |
+
"learning_rate": 2.345679012345679e-05,
|
1408 |
+
"loss": 0.8624,
|
1409 |
+
"step": 2000
|
1410 |
+
},
|
1411 |
+
{
|
1412 |
+
"epoch": 5.4177897574123985,
|
1413 |
+
"grad_norm": 2.0535359382629395,
|
1414 |
+
"learning_rate": 2.3319615912208508e-05,
|
1415 |
+
"loss": 0.8049,
|
1416 |
+
"step": 2010
|
1417 |
+
},
|
1418 |
+
{
|
1419 |
+
"epoch": 5.444743935309973,
|
1420 |
+
"grad_norm": 1.0573314428329468,
|
1421 |
+
"learning_rate": 2.318244170096022e-05,
|
1422 |
+
"loss": 0.8271,
|
1423 |
+
"step": 2020
|
1424 |
+
},
|
1425 |
+
{
|
1426 |
+
"epoch": 5.471698113207547,
|
1427 |
+
"grad_norm": 3.0816051959991455,
|
1428 |
+
"learning_rate": 2.3045267489711937e-05,
|
1429 |
+
"loss": 1.2149,
|
1430 |
+
"step": 2030
|
1431 |
+
},
|
1432 |
+
{
|
1433 |
+
"epoch": 5.498652291105121,
|
1434 |
+
"grad_norm": 0.8514117002487183,
|
1435 |
+
"learning_rate": 2.290809327846365e-05,
|
1436 |
+
"loss": 0.6105,
|
1437 |
+
"step": 2040
|
1438 |
+
},
|
1439 |
+
{
|
1440 |
+
"epoch": 5.525606469002695,
|
1441 |
+
"grad_norm": 2.2031140327453613,
|
1442 |
+
"learning_rate": 2.2770919067215366e-05,
|
1443 |
+
"loss": 0.7386,
|
1444 |
+
"step": 2050
|
1445 |
+
},
|
1446 |
+
{
|
1447 |
+
"epoch": 5.55256064690027,
|
1448 |
+
"grad_norm": 0.6656814217567444,
|
1449 |
+
"learning_rate": 2.2633744855967078e-05,
|
1450 |
+
"loss": 0.64,
|
1451 |
+
"step": 2060
|
1452 |
+
},
|
1453 |
+
{
|
1454 |
+
"epoch": 5.579514824797844,
|
1455 |
+
"grad_norm": 1.3257490396499634,
|
1456 |
+
"learning_rate": 2.2496570644718794e-05,
|
1457 |
+
"loss": 0.5246,
|
1458 |
+
"step": 2070
|
1459 |
+
},
|
1460 |
+
{
|
1461 |
+
"epoch": 5.606469002695418,
|
1462 |
+
"grad_norm": 1.9381606578826904,
|
1463 |
+
"learning_rate": 2.2359396433470507e-05,
|
1464 |
+
"loss": 0.8002,
|
1465 |
+
"step": 2080
|
1466 |
+
},
|
1467 |
+
{
|
1468 |
+
"epoch": 5.633423180592992,
|
1469 |
+
"grad_norm": 1.5016483068466187,
|
1470 |
+
"learning_rate": 2.2222222222222223e-05,
|
1471 |
+
"loss": 0.7496,
|
1472 |
+
"step": 2090
|
1473 |
+
},
|
1474 |
+
{
|
1475 |
+
"epoch": 5.660377358490566,
|
1476 |
+
"grad_norm": 1.4559197425842285,
|
1477 |
+
"learning_rate": 2.208504801097394e-05,
|
1478 |
+
"loss": 0.4904,
|
1479 |
+
"step": 2100
|
1480 |
+
},
|
1481 |
+
{
|
1482 |
+
"epoch": 5.6873315363881405,
|
1483 |
+
"grad_norm": 1.7636557817459106,
|
1484 |
+
"learning_rate": 2.1947873799725652e-05,
|
1485 |
+
"loss": 0.6664,
|
1486 |
+
"step": 2110
|
1487 |
+
},
|
1488 |
+
{
|
1489 |
+
"epoch": 5.714285714285714,
|
1490 |
+
"grad_norm": 1.5143455266952515,
|
1491 |
+
"learning_rate": 2.1810699588477368e-05,
|
1492 |
+
"loss": 0.9208,
|
1493 |
+
"step": 2120
|
1494 |
+
},
|
1495 |
+
{
|
1496 |
+
"epoch": 5.741239892183288,
|
1497 |
+
"grad_norm": 2.0838751792907715,
|
1498 |
+
"learning_rate": 2.167352537722908e-05,
|
1499 |
+
"loss": 0.3976,
|
1500 |
+
"step": 2130
|
1501 |
+
},
|
1502 |
+
{
|
1503 |
+
"epoch": 5.768194070080862,
|
1504 |
+
"grad_norm": 1.7315948009490967,
|
1505 |
+
"learning_rate": 2.1536351165980797e-05,
|
1506 |
+
"loss": 0.6964,
|
1507 |
+
"step": 2140
|
1508 |
+
},
|
1509 |
+
{
|
1510 |
+
"epoch": 5.795148247978437,
|
1511 |
+
"grad_norm": 2.732347249984741,
|
1512 |
+
"learning_rate": 2.139917695473251e-05,
|
1513 |
+
"loss": 1.4496,
|
1514 |
+
"step": 2150
|
1515 |
+
},
|
1516 |
+
{
|
1517 |
+
"epoch": 5.822102425876011,
|
1518 |
+
"grad_norm": 1.661921501159668,
|
1519 |
+
"learning_rate": 2.1262002743484226e-05,
|
1520 |
+
"loss": 0.8408,
|
1521 |
+
"step": 2160
|
1522 |
+
},
|
1523 |
+
{
|
1524 |
+
"epoch": 5.849056603773585,
|
1525 |
+
"grad_norm": 1.123995065689087,
|
1526 |
+
"learning_rate": 2.1124828532235942e-05,
|
1527 |
+
"loss": 0.5594,
|
1528 |
+
"step": 2170
|
1529 |
+
},
|
1530 |
+
{
|
1531 |
+
"epoch": 5.876010781671159,
|
1532 |
+
"grad_norm": 1.3693888187408447,
|
1533 |
+
"learning_rate": 2.0987654320987655e-05,
|
1534 |
+
"loss": 1.0366,
|
1535 |
+
"step": 2180
|
1536 |
+
},
|
1537 |
+
{
|
1538 |
+
"epoch": 5.902964959568733,
|
1539 |
+
"grad_norm": 1.589400053024292,
|
1540 |
+
"learning_rate": 2.085048010973937e-05,
|
1541 |
+
"loss": 1.1677,
|
1542 |
+
"step": 2190
|
1543 |
+
},
|
1544 |
+
{
|
1545 |
+
"epoch": 5.929919137466308,
|
1546 |
+
"grad_norm": 1.3596928119659424,
|
1547 |
+
"learning_rate": 2.0713305898491084e-05,
|
1548 |
+
"loss": 0.6217,
|
1549 |
+
"step": 2200
|
1550 |
+
},
|
1551 |
+
{
|
1552 |
+
"epoch": 5.9568733153638815,
|
1553 |
+
"grad_norm": 1.5818856954574585,
|
1554 |
+
"learning_rate": 2.05761316872428e-05,
|
1555 |
+
"loss": 1.0682,
|
1556 |
+
"step": 2210
|
1557 |
+
},
|
1558 |
+
{
|
1559 |
+
"epoch": 5.9838274932614555,
|
1560 |
+
"grad_norm": 2.694117546081543,
|
1561 |
+
"learning_rate": 2.0438957475994513e-05,
|
1562 |
+
"loss": 0.9281,
|
1563 |
+
"step": 2220
|
1564 |
+
},
|
1565 |
+
{
|
1566 |
+
"epoch": 6.010781671159029,
|
1567 |
+
"grad_norm": 1.463112473487854,
|
1568 |
+
"learning_rate": 2.030178326474623e-05,
|
1569 |
+
"loss": 0.7868,
|
1570 |
+
"step": 2230
|
1571 |
+
},
|
1572 |
+
{
|
1573 |
+
"epoch": 6.037735849056604,
|
1574 |
+
"grad_norm": 1.8323993682861328,
|
1575 |
+
"learning_rate": 2.016460905349794e-05,
|
1576 |
+
"loss": 0.6472,
|
1577 |
+
"step": 2240
|
1578 |
+
},
|
1579 |
+
{
|
1580 |
+
"epoch": 6.064690026954178,
|
1581 |
+
"grad_norm": 2.3480756282806396,
|
1582 |
+
"learning_rate": 2.0027434842249658e-05,
|
1583 |
+
"loss": 0.6044,
|
1584 |
+
"step": 2250
|
1585 |
+
},
|
1586 |
+
{
|
1587 |
+
"epoch": 6.091644204851752,
|
1588 |
+
"grad_norm": 3.0445172786712646,
|
1589 |
+
"learning_rate": 1.9890260631001374e-05,
|
1590 |
+
"loss": 0.5516,
|
1591 |
+
"step": 2260
|
1592 |
+
},
|
1593 |
+
{
|
1594 |
+
"epoch": 6.118598382749326,
|
1595 |
+
"grad_norm": 2.489713430404663,
|
1596 |
+
"learning_rate": 1.9753086419753087e-05,
|
1597 |
+
"loss": 0.5861,
|
1598 |
+
"step": 2270
|
1599 |
+
},
|
1600 |
+
{
|
1601 |
+
"epoch": 6.1455525606469,
|
1602 |
+
"grad_norm": 2.1993355751037598,
|
1603 |
+
"learning_rate": 1.9615912208504803e-05,
|
1604 |
+
"loss": 0.6938,
|
1605 |
+
"step": 2280
|
1606 |
+
},
|
1607 |
+
{
|
1608 |
+
"epoch": 6.172506738544475,
|
1609 |
+
"grad_norm": 1.914015769958496,
|
1610 |
+
"learning_rate": 1.9478737997256515e-05,
|
1611 |
+
"loss": 0.8575,
|
1612 |
+
"step": 2290
|
1613 |
+
},
|
1614 |
+
{
|
1615 |
+
"epoch": 6.199460916442049,
|
1616 |
+
"grad_norm": 2.359715223312378,
|
1617 |
+
"learning_rate": 1.934156378600823e-05,
|
1618 |
+
"loss": 0.6444,
|
1619 |
+
"step": 2300
|
1620 |
+
},
|
1621 |
+
{
|
1622 |
+
"epoch": 6.226415094339623,
|
1623 |
+
"grad_norm": 0.865803062915802,
|
1624 |
+
"learning_rate": 1.9204389574759944e-05,
|
1625 |
+
"loss": 0.6543,
|
1626 |
+
"step": 2310
|
1627 |
+
},
|
1628 |
+
{
|
1629 |
+
"epoch": 6.2533692722371965,
|
1630 |
+
"grad_norm": 1.363655686378479,
|
1631 |
+
"learning_rate": 1.906721536351166e-05,
|
1632 |
+
"loss": 0.9914,
|
1633 |
+
"step": 2320
|
1634 |
+
},
|
1635 |
+
{
|
1636 |
+
"epoch": 6.280323450134771,
|
1637 |
+
"grad_norm": 2.174215316772461,
|
1638 |
+
"learning_rate": 1.8930041152263377e-05,
|
1639 |
+
"loss": 0.6777,
|
1640 |
+
"step": 2330
|
1641 |
+
},
|
1642 |
+
{
|
1643 |
+
"epoch": 6.307277628032345,
|
1644 |
+
"grad_norm": 1.1655148267745972,
|
1645 |
+
"learning_rate": 1.879286694101509e-05,
|
1646 |
+
"loss": 0.6198,
|
1647 |
+
"step": 2340
|
1648 |
+
},
|
1649 |
+
{
|
1650 |
+
"epoch": 6.334231805929919,
|
1651 |
+
"grad_norm": 2.751349449157715,
|
1652 |
+
"learning_rate": 1.8655692729766806e-05,
|
1653 |
+
"loss": 1.3854,
|
1654 |
+
"step": 2350
|
1655 |
+
},
|
1656 |
+
{
|
1657 |
+
"epoch": 6.361185983827493,
|
1658 |
+
"grad_norm": 1.7437028884887695,
|
1659 |
+
"learning_rate": 1.8518518518518518e-05,
|
1660 |
+
"loss": 0.6498,
|
1661 |
+
"step": 2360
|
1662 |
+
},
|
1663 |
+
{
|
1664 |
+
"epoch": 6.388140161725067,
|
1665 |
+
"grad_norm": 2.6876518726348877,
|
1666 |
+
"learning_rate": 1.8381344307270234e-05,
|
1667 |
+
"loss": 0.8044,
|
1668 |
+
"step": 2370
|
1669 |
+
},
|
1670 |
+
{
|
1671 |
+
"epoch": 6.415094339622642,
|
1672 |
+
"grad_norm": 2.60074520111084,
|
1673 |
+
"learning_rate": 1.8244170096021947e-05,
|
1674 |
+
"loss": 1.155,
|
1675 |
+
"step": 2380
|
1676 |
+
},
|
1677 |
+
{
|
1678 |
+
"epoch": 6.442048517520216,
|
1679 |
+
"grad_norm": 2.644638776779175,
|
1680 |
+
"learning_rate": 1.8106995884773663e-05,
|
1681 |
+
"loss": 0.8515,
|
1682 |
+
"step": 2390
|
1683 |
+
},
|
1684 |
+
{
|
1685 |
+
"epoch": 6.46900269541779,
|
1686 |
+
"grad_norm": 3.353593587875366,
|
1687 |
+
"learning_rate": 1.7969821673525376e-05,
|
1688 |
+
"loss": 0.5867,
|
1689 |
+
"step": 2400
|
1690 |
+
},
|
1691 |
+
{
|
1692 |
+
"epoch": 6.495956873315364,
|
1693 |
+
"grad_norm": 1.9784125089645386,
|
1694 |
+
"learning_rate": 1.7832647462277096e-05,
|
1695 |
+
"loss": 0.6024,
|
1696 |
+
"step": 2410
|
1697 |
+
},
|
1698 |
+
{
|
1699 |
+
"epoch": 6.5229110512129385,
|
1700 |
+
"grad_norm": 1.7953509092330933,
|
1701 |
+
"learning_rate": 1.769547325102881e-05,
|
1702 |
+
"loss": 1.2158,
|
1703 |
+
"step": 2420
|
1704 |
+
},
|
1705 |
+
{
|
1706 |
+
"epoch": 6.549865229110512,
|
1707 |
+
"grad_norm": 2.232206344604492,
|
1708 |
+
"learning_rate": 1.755829903978052e-05,
|
1709 |
+
"loss": 0.8265,
|
1710 |
+
"step": 2430
|
1711 |
+
},
|
1712 |
+
{
|
1713 |
+
"epoch": 6.576819407008086,
|
1714 |
+
"grad_norm": 1.054999828338623,
|
1715 |
+
"learning_rate": 1.7421124828532237e-05,
|
1716 |
+
"loss": 0.7372,
|
1717 |
+
"step": 2440
|
1718 |
+
},
|
1719 |
+
{
|
1720 |
+
"epoch": 6.60377358490566,
|
1721 |
+
"grad_norm": 1.8623944520950317,
|
1722 |
+
"learning_rate": 1.728395061728395e-05,
|
1723 |
+
"loss": 1.4106,
|
1724 |
+
"step": 2450
|
1725 |
+
},
|
1726 |
+
{
|
1727 |
+
"epoch": 6.630727762803234,
|
1728 |
+
"grad_norm": 2.689140796661377,
|
1729 |
+
"learning_rate": 1.7146776406035666e-05,
|
1730 |
+
"loss": 0.9934,
|
1731 |
+
"step": 2460
|
1732 |
+
},
|
1733 |
+
{
|
1734 |
+
"epoch": 6.657681940700809,
|
1735 |
+
"grad_norm": 1.7552543878555298,
|
1736 |
+
"learning_rate": 1.700960219478738e-05,
|
1737 |
+
"loss": 0.8231,
|
1738 |
+
"step": 2470
|
1739 |
+
},
|
1740 |
+
{
|
1741 |
+
"epoch": 6.684636118598383,
|
1742 |
+
"grad_norm": 3.4416518211364746,
|
1743 |
+
"learning_rate": 1.6872427983539095e-05,
|
1744 |
+
"loss": 0.7788,
|
1745 |
+
"step": 2480
|
1746 |
+
},
|
1747 |
+
{
|
1748 |
+
"epoch": 6.711590296495957,
|
1749 |
+
"grad_norm": 3.726334571838379,
|
1750 |
+
"learning_rate": 1.673525377229081e-05,
|
1751 |
+
"loss": 0.7407,
|
1752 |
+
"step": 2490
|
1753 |
+
},
|
1754 |
+
{
|
1755 |
+
"epoch": 6.738544474393531,
|
1756 |
+
"grad_norm": 2.932966470718384,
|
1757 |
+
"learning_rate": 1.6598079561042527e-05,
|
1758 |
+
"loss": 1.0143,
|
1759 |
+
"step": 2500
|
1760 |
+
},
|
1761 |
+
{
|
1762 |
+
"epoch": 6.765498652291106,
|
1763 |
+
"grad_norm": 1.8736401796340942,
|
1764 |
+
"learning_rate": 1.646090534979424e-05,
|
1765 |
+
"loss": 0.8365,
|
1766 |
+
"step": 2510
|
1767 |
+
},
|
1768 |
+
{
|
1769 |
+
"epoch": 6.7924528301886795,
|
1770 |
+
"grad_norm": 2.0926668643951416,
|
1771 |
+
"learning_rate": 1.6323731138545953e-05,
|
1772 |
+
"loss": 0.7736,
|
1773 |
+
"step": 2520
|
1774 |
+
},
|
1775 |
+
{
|
1776 |
+
"epoch": 6.819407008086253,
|
1777 |
+
"grad_norm": 3.3576009273529053,
|
1778 |
+
"learning_rate": 1.618655692729767e-05,
|
1779 |
+
"loss": 1.1465,
|
1780 |
+
"step": 2530
|
1781 |
+
},
|
1782 |
+
{
|
1783 |
+
"epoch": 6.846361185983827,
|
1784 |
+
"grad_norm": 2.4661567211151123,
|
1785 |
+
"learning_rate": 1.604938271604938e-05,
|
1786 |
+
"loss": 0.7863,
|
1787 |
+
"step": 2540
|
1788 |
+
},
|
1789 |
+
{
|
1790 |
+
"epoch": 6.873315363881401,
|
1791 |
+
"grad_norm": 1.88754141330719,
|
1792 |
+
"learning_rate": 1.5912208504801098e-05,
|
1793 |
+
"loss": 0.9081,
|
1794 |
+
"step": 2550
|
1795 |
+
},
|
1796 |
+
{
|
1797 |
+
"epoch": 6.900269541778976,
|
1798 |
+
"grad_norm": 47.142337799072266,
|
1799 |
+
"learning_rate": 1.5775034293552814e-05,
|
1800 |
+
"loss": 0.8261,
|
1801 |
+
"step": 2560
|
1802 |
+
},
|
1803 |
+
{
|
1804 |
+
"epoch": 6.92722371967655,
|
1805 |
+
"grad_norm": 2.473158359527588,
|
1806 |
+
"learning_rate": 1.563786008230453e-05,
|
1807 |
+
"loss": 0.4822,
|
1808 |
+
"step": 2570
|
1809 |
+
},
|
1810 |
+
{
|
1811 |
+
"epoch": 6.954177897574124,
|
1812 |
+
"grad_norm": 1.6194536685943604,
|
1813 |
+
"learning_rate": 1.5500685871056243e-05,
|
1814 |
+
"loss": 0.856,
|
1815 |
+
"step": 2580
|
1816 |
+
},
|
1817 |
+
{
|
1818 |
+
"epoch": 6.981132075471698,
|
1819 |
+
"grad_norm": 5.963684558868408,
|
1820 |
+
"learning_rate": 1.536351165980796e-05,
|
1821 |
+
"loss": 0.7661,
|
1822 |
+
"step": 2590
|
1823 |
+
},
|
1824 |
+
{
|
1825 |
+
"epoch": 7.008086253369272,
|
1826 |
+
"grad_norm": 1.8639130592346191,
|
1827 |
+
"learning_rate": 1.5226337448559672e-05,
|
1828 |
+
"loss": 1.388,
|
1829 |
+
"step": 2600
|
1830 |
+
},
|
1831 |
+
{
|
1832 |
+
"epoch": 7.035040431266847,
|
1833 |
+
"grad_norm": 3.447125196456909,
|
1834 |
+
"learning_rate": 1.5089163237311384e-05,
|
1835 |
+
"loss": 0.7474,
|
1836 |
+
"step": 2610
|
1837 |
+
},
|
1838 |
+
{
|
1839 |
+
"epoch": 7.061994609164421,
|
1840 |
+
"grad_norm": 2.3289992809295654,
|
1841 |
+
"learning_rate": 1.49519890260631e-05,
|
1842 |
+
"loss": 0.5907,
|
1843 |
+
"step": 2620
|
1844 |
+
},
|
1845 |
+
{
|
1846 |
+
"epoch": 7.0889487870619945,
|
1847 |
+
"grad_norm": 1.342872977256775,
|
1848 |
+
"learning_rate": 1.4814814814814815e-05,
|
1849 |
+
"loss": 0.8798,
|
1850 |
+
"step": 2630
|
1851 |
+
},
|
1852 |
+
{
|
1853 |
+
"epoch": 7.115902964959568,
|
1854 |
+
"grad_norm": 1.3832533359527588,
|
1855 |
+
"learning_rate": 1.4677640603566531e-05,
|
1856 |
+
"loss": 1.0841,
|
1857 |
+
"step": 2640
|
1858 |
+
},
|
1859 |
+
{
|
1860 |
+
"epoch": 7.142857142857143,
|
1861 |
+
"grad_norm": 1.0350087881088257,
|
1862 |
+
"learning_rate": 1.4540466392318244e-05,
|
1863 |
+
"loss": 1.2334,
|
1864 |
+
"step": 2650
|
1865 |
+
},
|
1866 |
+
{
|
1867 |
+
"epoch": 7.169811320754717,
|
1868 |
+
"grad_norm": 4.020228862762451,
|
1869 |
+
"learning_rate": 1.440329218106996e-05,
|
1870 |
+
"loss": 0.8947,
|
1871 |
+
"step": 2660
|
1872 |
+
},
|
1873 |
+
{
|
1874 |
+
"epoch": 7.196765498652291,
|
1875 |
+
"grad_norm": 2.982022523880005,
|
1876 |
+
"learning_rate": 1.4266117969821674e-05,
|
1877 |
+
"loss": 0.9667,
|
1878 |
+
"step": 2670
|
1879 |
+
},
|
1880 |
+
{
|
1881 |
+
"epoch": 7.223719676549865,
|
1882 |
+
"grad_norm": 2.171691417694092,
|
1883 |
+
"learning_rate": 1.412894375857339e-05,
|
1884 |
+
"loss": 0.7494,
|
1885 |
+
"step": 2680
|
1886 |
+
},
|
1887 |
+
{
|
1888 |
+
"epoch": 7.250673854447439,
|
1889 |
+
"grad_norm": 2.717907667160034,
|
1890 |
+
"learning_rate": 1.3991769547325103e-05,
|
1891 |
+
"loss": 0.6898,
|
1892 |
+
"step": 2690
|
1893 |
+
},
|
1894 |
+
{
|
1895 |
+
"epoch": 7.277628032345014,
|
1896 |
+
"grad_norm": 2.5439579486846924,
|
1897 |
+
"learning_rate": 1.3854595336076818e-05,
|
1898 |
+
"loss": 0.4311,
|
1899 |
+
"step": 2700
|
1900 |
+
},
|
1901 |
+
{
|
1902 |
+
"epoch": 7.304582210242588,
|
1903 |
+
"grad_norm": 0.7530654072761536,
|
1904 |
+
"learning_rate": 1.3717421124828534e-05,
|
1905 |
+
"loss": 0.9008,
|
1906 |
+
"step": 2710
|
1907 |
+
},
|
1908 |
+
{
|
1909 |
+
"epoch": 7.331536388140162,
|
1910 |
+
"grad_norm": 0.637667179107666,
|
1911 |
+
"learning_rate": 1.3580246913580247e-05,
|
1912 |
+
"loss": 1.087,
|
1913 |
+
"step": 2720
|
1914 |
+
},
|
1915 |
+
{
|
1916 |
+
"epoch": 7.3584905660377355,
|
1917 |
+
"grad_norm": 1.1592116355895996,
|
1918 |
+
"learning_rate": 1.3443072702331963e-05,
|
1919 |
+
"loss": 0.6196,
|
1920 |
+
"step": 2730
|
1921 |
+
},
|
1922 |
+
{
|
1923 |
+
"epoch": 7.38544474393531,
|
1924 |
+
"grad_norm": 2.054795503616333,
|
1925 |
+
"learning_rate": 1.3305898491083677e-05,
|
1926 |
+
"loss": 0.7915,
|
1927 |
+
"step": 2740
|
1928 |
+
},
|
1929 |
+
{
|
1930 |
+
"epoch": 7.412398921832884,
|
1931 |
+
"grad_norm": 1.4349403381347656,
|
1932 |
+
"learning_rate": 1.3168724279835393e-05,
|
1933 |
+
"loss": 0.6572,
|
1934 |
+
"step": 2750
|
1935 |
+
},
|
1936 |
+
{
|
1937 |
+
"epoch": 7.439353099730458,
|
1938 |
+
"grad_norm": 1.0293610095977783,
|
1939 |
+
"learning_rate": 1.3031550068587106e-05,
|
1940 |
+
"loss": 1.0812,
|
1941 |
+
"step": 2760
|
1942 |
+
},
|
1943 |
+
{
|
1944 |
+
"epoch": 7.466307277628032,
|
1945 |
+
"grad_norm": 2.4686598777770996,
|
1946 |
+
"learning_rate": 1.2894375857338819e-05,
|
1947 |
+
"loss": 0.9702,
|
1948 |
+
"step": 2770
|
1949 |
+
},
|
1950 |
+
{
|
1951 |
+
"epoch": 7.493261455525606,
|
1952 |
+
"grad_norm": 1.4416760206222534,
|
1953 |
+
"learning_rate": 1.2757201646090535e-05,
|
1954 |
+
"loss": 1.0412,
|
1955 |
+
"step": 2780
|
1956 |
+
},
|
1957 |
+
{
|
1958 |
+
"epoch": 7.520215633423181,
|
1959 |
+
"grad_norm": 3.614410400390625,
|
1960 |
+
"learning_rate": 1.262002743484225e-05,
|
1961 |
+
"loss": 1.2559,
|
1962 |
+
"step": 2790
|
1963 |
+
},
|
1964 |
+
{
|
1965 |
+
"epoch": 7.547169811320755,
|
1966 |
+
"grad_norm": 0.8973541855812073,
|
1967 |
+
"learning_rate": 1.2482853223593966e-05,
|
1968 |
+
"loss": 0.4049,
|
1969 |
+
"step": 2800
|
1970 |
+
},
|
1971 |
+
{
|
1972 |
+
"epoch": 7.574123989218329,
|
1973 |
+
"grad_norm": 1.4669396877288818,
|
1974 |
+
"learning_rate": 1.2345679012345678e-05,
|
1975 |
+
"loss": 0.764,
|
1976 |
+
"step": 2810
|
1977 |
+
},
|
1978 |
+
{
|
1979 |
+
"epoch": 7.601078167115903,
|
1980 |
+
"grad_norm": 2.6035633087158203,
|
1981 |
+
"learning_rate": 1.2208504801097394e-05,
|
1982 |
+
"loss": 0.7421,
|
1983 |
+
"step": 2820
|
1984 |
+
},
|
1985 |
+
{
|
1986 |
+
"epoch": 7.628032345013477,
|
1987 |
+
"grad_norm": 3.8187856674194336,
|
1988 |
+
"learning_rate": 1.2071330589849109e-05,
|
1989 |
+
"loss": 0.8026,
|
1990 |
+
"step": 2830
|
1991 |
+
},
|
1992 |
+
{
|
1993 |
+
"epoch": 7.654986522911051,
|
1994 |
+
"grad_norm": 1.5178192853927612,
|
1995 |
+
"learning_rate": 1.1934156378600823e-05,
|
1996 |
+
"loss": 0.8348,
|
1997 |
+
"step": 2840
|
1998 |
+
},
|
1999 |
+
{
|
2000 |
+
"epoch": 7.681940700808625,
|
2001 |
+
"grad_norm": 1.0527846813201904,
|
2002 |
+
"learning_rate": 1.1796982167352538e-05,
|
2003 |
+
"loss": 0.7379,
|
2004 |
+
"step": 2850
|
2005 |
+
},
|
2006 |
+
{
|
2007 |
+
"epoch": 7.708894878706199,
|
2008 |
+
"grad_norm": 2.7532355785369873,
|
2009 |
+
"learning_rate": 1.1659807956104254e-05,
|
2010 |
+
"loss": 1.4819,
|
2011 |
+
"step": 2860
|
2012 |
+
},
|
2013 |
+
{
|
2014 |
+
"epoch": 7.735849056603773,
|
2015 |
+
"grad_norm": 1.9591217041015625,
|
2016 |
+
"learning_rate": 1.1522633744855968e-05,
|
2017 |
+
"loss": 0.4296,
|
2018 |
+
"step": 2870
|
2019 |
+
},
|
2020 |
+
{
|
2021 |
+
"epoch": 7.762803234501348,
|
2022 |
+
"grad_norm": 2.7292425632476807,
|
2023 |
+
"learning_rate": 1.1385459533607683e-05,
|
2024 |
+
"loss": 0.6324,
|
2025 |
+
"step": 2880
|
2026 |
+
},
|
2027 |
+
{
|
2028 |
+
"epoch": 7.789757412398922,
|
2029 |
+
"grad_norm": 2.3577399253845215,
|
2030 |
+
"learning_rate": 1.1248285322359397e-05,
|
2031 |
+
"loss": 0.5892,
|
2032 |
+
"step": 2890
|
2033 |
+
},
|
2034 |
+
{
|
2035 |
+
"epoch": 7.816711590296496,
|
2036 |
+
"grad_norm": 1.4059489965438843,
|
2037 |
+
"learning_rate": 1.1111111111111112e-05,
|
2038 |
+
"loss": 0.5519,
|
2039 |
+
"step": 2900
|
2040 |
+
},
|
2041 |
+
{
|
2042 |
+
"epoch": 7.84366576819407,
|
2043 |
+
"grad_norm": 1.9480534791946411,
|
2044 |
+
"learning_rate": 1.0973936899862826e-05,
|
2045 |
+
"loss": 0.9436,
|
2046 |
+
"step": 2910
|
2047 |
+
},
|
2048 |
+
{
|
2049 |
+
"epoch": 7.870619946091644,
|
2050 |
+
"grad_norm": 2.3746042251586914,
|
2051 |
+
"learning_rate": 1.083676268861454e-05,
|
2052 |
+
"loss": 0.3955,
|
2053 |
+
"step": 2920
|
2054 |
+
},
|
2055 |
+
{
|
2056 |
+
"epoch": 7.8975741239892185,
|
2057 |
+
"grad_norm": 1.2576045989990234,
|
2058 |
+
"learning_rate": 1.0699588477366255e-05,
|
2059 |
+
"loss": 0.604,
|
2060 |
+
"step": 2930
|
2061 |
+
},
|
2062 |
+
{
|
2063 |
+
"epoch": 7.9245283018867925,
|
2064 |
+
"grad_norm": 2.509427309036255,
|
2065 |
+
"learning_rate": 1.0562414266117971e-05,
|
2066 |
+
"loss": 0.7358,
|
2067 |
+
"step": 2940
|
2068 |
+
},
|
2069 |
+
{
|
2070 |
+
"epoch": 7.951482479784366,
|
2071 |
+
"grad_norm": 1.8080178499221802,
|
2072 |
+
"learning_rate": 1.0425240054869686e-05,
|
2073 |
+
"loss": 0.6726,
|
2074 |
+
"step": 2950
|
2075 |
+
},
|
2076 |
+
{
|
2077 |
+
"epoch": 7.97843665768194,
|
2078 |
+
"grad_norm": 3.254493474960327,
|
2079 |
+
"learning_rate": 1.02880658436214e-05,
|
2080 |
+
"loss": 0.6716,
|
2081 |
+
"step": 2960
|
2082 |
+
},
|
2083 |
+
{
|
2084 |
+
"epoch": 8.005390835579515,
|
2085 |
+
"grad_norm": 1.1409560441970825,
|
2086 |
+
"learning_rate": 1.0150891632373114e-05,
|
2087 |
+
"loss": 0.8359,
|
2088 |
+
"step": 2970
|
2089 |
+
},
|
2090 |
+
{
|
2091 |
+
"epoch": 8.032345013477089,
|
2092 |
+
"grad_norm": 2.8074634075164795,
|
2093 |
+
"learning_rate": 1.0013717421124829e-05,
|
2094 |
+
"loss": 0.8532,
|
2095 |
+
"step": 2980
|
2096 |
+
},
|
2097 |
+
{
|
2098 |
+
"epoch": 8.059299191374663,
|
2099 |
+
"grad_norm": 1.6796783208847046,
|
2100 |
+
"learning_rate": 9.876543209876543e-06,
|
2101 |
+
"loss": 1.1366,
|
2102 |
+
"step": 2990
|
2103 |
+
},
|
2104 |
+
{
|
2105 |
+
"epoch": 8.086253369272237,
|
2106 |
+
"grad_norm": 0.6373213529586792,
|
2107 |
+
"learning_rate": 9.739368998628258e-06,
|
2108 |
+
"loss": 0.5121,
|
2109 |
+
"step": 3000
|
2110 |
+
},
|
2111 |
+
{
|
2112 |
+
"epoch": 8.11320754716981,
|
2113 |
+
"grad_norm": 1.404941201210022,
|
2114 |
+
"learning_rate": 9.602194787379972e-06,
|
2115 |
+
"loss": 0.7168,
|
2116 |
+
"step": 3010
|
2117 |
+
},
|
2118 |
+
{
|
2119 |
+
"epoch": 8.140161725067385,
|
2120 |
+
"grad_norm": 3.0173532962799072,
|
2121 |
+
"learning_rate": 9.465020576131688e-06,
|
2122 |
+
"loss": 0.6495,
|
2123 |
+
"step": 3020
|
2124 |
+
},
|
2125 |
+
{
|
2126 |
+
"epoch": 8.167115902964959,
|
2127 |
+
"grad_norm": 0.9373369812965393,
|
2128 |
+
"learning_rate": 9.327846364883403e-06,
|
2129 |
+
"loss": 1.1245,
|
2130 |
+
"step": 3030
|
2131 |
+
},
|
2132 |
+
{
|
2133 |
+
"epoch": 8.194070080862534,
|
2134 |
+
"grad_norm": 1.8687936067581177,
|
2135 |
+
"learning_rate": 9.190672153635117e-06,
|
2136 |
+
"loss": 0.7253,
|
2137 |
+
"step": 3040
|
2138 |
+
},
|
2139 |
+
{
|
2140 |
+
"epoch": 8.221024258760108,
|
2141 |
+
"grad_norm": 0.9592246413230896,
|
2142 |
+
"learning_rate": 9.053497942386832e-06,
|
2143 |
+
"loss": 1.2584,
|
2144 |
+
"step": 3050
|
2145 |
+
},
|
2146 |
+
{
|
2147 |
+
"epoch": 8.247978436657682,
|
2148 |
+
"grad_norm": 2.668527126312256,
|
2149 |
+
"learning_rate": 8.916323731138548e-06,
|
2150 |
+
"loss": 0.6645,
|
2151 |
+
"step": 3060
|
2152 |
+
},
|
2153 |
+
{
|
2154 |
+
"epoch": 8.274932614555256,
|
2155 |
+
"grad_norm": 0.8130900859832764,
|
2156 |
+
"learning_rate": 8.77914951989026e-06,
|
2157 |
+
"loss": 0.5244,
|
2158 |
+
"step": 3070
|
2159 |
+
},
|
2160 |
+
{
|
2161 |
+
"epoch": 8.30188679245283,
|
2162 |
+
"grad_norm": 1.980900764465332,
|
2163 |
+
"learning_rate": 8.641975308641975e-06,
|
2164 |
+
"loss": 0.5843,
|
2165 |
+
"step": 3080
|
2166 |
+
},
|
2167 |
+
{
|
2168 |
+
"epoch": 8.328840970350404,
|
2169 |
+
"grad_norm": 1.7426378726959229,
|
2170 |
+
"learning_rate": 8.50480109739369e-06,
|
2171 |
+
"loss": 0.9521,
|
2172 |
+
"step": 3090
|
2173 |
+
},
|
2174 |
+
{
|
2175 |
+
"epoch": 8.355795148247978,
|
2176 |
+
"grad_norm": 2.6671996116638184,
|
2177 |
+
"learning_rate": 8.367626886145406e-06,
|
2178 |
+
"loss": 0.8797,
|
2179 |
+
"step": 3100
|
2180 |
+
},
|
2181 |
+
{
|
2182 |
+
"epoch": 8.382749326145552,
|
2183 |
+
"grad_norm": 2.629798650741577,
|
2184 |
+
"learning_rate": 8.23045267489712e-06,
|
2185 |
+
"loss": 0.5629,
|
2186 |
+
"step": 3110
|
2187 |
+
},
|
2188 |
+
{
|
2189 |
+
"epoch": 8.409703504043126,
|
2190 |
+
"grad_norm": 1.723059058189392,
|
2191 |
+
"learning_rate": 8.093278463648834e-06,
|
2192 |
+
"loss": 0.5738,
|
2193 |
+
"step": 3120
|
2194 |
+
},
|
2195 |
+
{
|
2196 |
+
"epoch": 8.436657681940702,
|
2197 |
+
"grad_norm": 0.5404053926467896,
|
2198 |
+
"learning_rate": 7.956104252400549e-06,
|
2199 |
+
"loss": 0.72,
|
2200 |
+
"step": 3130
|
2201 |
+
},
|
2202 |
+
{
|
2203 |
+
"epoch": 8.463611859838275,
|
2204 |
+
"grad_norm": 2.8987836837768555,
|
2205 |
+
"learning_rate": 7.818930041152265e-06,
|
2206 |
+
"loss": 0.7432,
|
2207 |
+
"step": 3140
|
2208 |
+
},
|
2209 |
+
{
|
2210 |
+
"epoch": 8.49056603773585,
|
2211 |
+
"grad_norm": 1.2646089792251587,
|
2212 |
+
"learning_rate": 7.68175582990398e-06,
|
2213 |
+
"loss": 0.5356,
|
2214 |
+
"step": 3150
|
2215 |
+
},
|
2216 |
+
{
|
2217 |
+
"epoch": 8.517520215633423,
|
2218 |
+
"grad_norm": 3.3997156620025635,
|
2219 |
+
"learning_rate": 7.544581618655692e-06,
|
2220 |
+
"loss": 0.8056,
|
2221 |
+
"step": 3160
|
2222 |
+
},
|
2223 |
+
{
|
2224 |
+
"epoch": 8.544474393530997,
|
2225 |
+
"grad_norm": 2.264604330062866,
|
2226 |
+
"learning_rate": 7.4074074074074075e-06,
|
2227 |
+
"loss": 1.0019,
|
2228 |
+
"step": 3170
|
2229 |
+
},
|
2230 |
+
{
|
2231 |
+
"epoch": 8.571428571428571,
|
2232 |
+
"grad_norm": 1.962279200553894,
|
2233 |
+
"learning_rate": 7.270233196159122e-06,
|
2234 |
+
"loss": 0.6352,
|
2235 |
+
"step": 3180
|
2236 |
+
},
|
2237 |
+
{
|
2238 |
+
"epoch": 8.598382749326145,
|
2239 |
+
"grad_norm": 2.4050302505493164,
|
2240 |
+
"learning_rate": 7.133058984910837e-06,
|
2241 |
+
"loss": 0.6712,
|
2242 |
+
"step": 3190
|
2243 |
+
},
|
2244 |
+
{
|
2245 |
+
"epoch": 8.625336927223719,
|
2246 |
+
"grad_norm": 1.2828004360198975,
|
2247 |
+
"learning_rate": 6.995884773662552e-06,
|
2248 |
+
"loss": 0.625,
|
2249 |
+
"step": 3200
|
2250 |
+
},
|
2251 |
+
{
|
2252 |
+
"epoch": 8.652291105121293,
|
2253 |
+
"grad_norm": 3.306525230407715,
|
2254 |
+
"learning_rate": 6.858710562414267e-06,
|
2255 |
+
"loss": 0.7727,
|
2256 |
+
"step": 3210
|
2257 |
+
},
|
2258 |
+
{
|
2259 |
+
"epoch": 8.679245283018869,
|
2260 |
+
"grad_norm": 1.3902239799499512,
|
2261 |
+
"learning_rate": 6.721536351165981e-06,
|
2262 |
+
"loss": 0.4857,
|
2263 |
+
"step": 3220
|
2264 |
+
},
|
2265 |
+
{
|
2266 |
+
"epoch": 8.706199460916443,
|
2267 |
+
"grad_norm": 3.665642261505127,
|
2268 |
+
"learning_rate": 6.584362139917697e-06,
|
2269 |
+
"loss": 0.9572,
|
2270 |
+
"step": 3230
|
2271 |
+
},
|
2272 |
+
{
|
2273 |
+
"epoch": 8.733153638814017,
|
2274 |
+
"grad_norm": 2.6966567039489746,
|
2275 |
+
"learning_rate": 6.447187928669409e-06,
|
2276 |
+
"loss": 0.739,
|
2277 |
+
"step": 3240
|
2278 |
+
},
|
2279 |
+
{
|
2280 |
+
"epoch": 8.76010781671159,
|
2281 |
+
"grad_norm": 2.127206325531006,
|
2282 |
+
"learning_rate": 6.310013717421125e-06,
|
2283 |
+
"loss": 0.8983,
|
2284 |
+
"step": 3250
|
2285 |
+
},
|
2286 |
+
{
|
2287 |
+
"epoch": 8.787061994609164,
|
2288 |
+
"grad_norm": 2.1118884086608887,
|
2289 |
+
"learning_rate": 6.172839506172839e-06,
|
2290 |
+
"loss": 1.0385,
|
2291 |
+
"step": 3260
|
2292 |
+
},
|
2293 |
+
{
|
2294 |
+
"epoch": 8.814016172506738,
|
2295 |
+
"grad_norm": 1.6121476888656616,
|
2296 |
+
"learning_rate": 6.0356652949245544e-06,
|
2297 |
+
"loss": 0.4564,
|
2298 |
+
"step": 3270
|
2299 |
+
},
|
2300 |
+
{
|
2301 |
+
"epoch": 8.840970350404312,
|
2302 |
+
"grad_norm": 3.346813440322876,
|
2303 |
+
"learning_rate": 5.898491083676269e-06,
|
2304 |
+
"loss": 0.9239,
|
2305 |
+
"step": 3280
|
2306 |
+
},
|
2307 |
+
{
|
2308 |
+
"epoch": 8.867924528301886,
|
2309 |
+
"grad_norm": 2.808685541152954,
|
2310 |
+
"learning_rate": 5.761316872427984e-06,
|
2311 |
+
"loss": 1.0246,
|
2312 |
+
"step": 3290
|
2313 |
+
},
|
2314 |
+
{
|
2315 |
+
"epoch": 8.89487870619946,
|
2316 |
+
"grad_norm": 0.6503840684890747,
|
2317 |
+
"learning_rate": 5.624142661179699e-06,
|
2318 |
+
"loss": 0.4066,
|
2319 |
+
"step": 3300
|
2320 |
+
},
|
2321 |
+
{
|
2322 |
+
"epoch": 8.921832884097036,
|
2323 |
+
"grad_norm": 2.2090868949890137,
|
2324 |
+
"learning_rate": 5.486968449931413e-06,
|
2325 |
+
"loss": 0.8856,
|
2326 |
+
"step": 3310
|
2327 |
+
},
|
2328 |
+
{
|
2329 |
+
"epoch": 8.94878706199461,
|
2330 |
+
"grad_norm": 2.4998106956481934,
|
2331 |
+
"learning_rate": 5.3497942386831275e-06,
|
2332 |
+
"loss": 1.0539,
|
2333 |
+
"step": 3320
|
2334 |
+
},
|
2335 |
+
{
|
2336 |
+
"epoch": 8.975741239892184,
|
2337 |
+
"grad_norm": 1.3785419464111328,
|
2338 |
+
"learning_rate": 5.212620027434843e-06,
|
2339 |
+
"loss": 1.0206,
|
2340 |
+
"step": 3330
|
2341 |
+
},
|
2342 |
+
{
|
2343 |
+
"epoch": 9.002695417789758,
|
2344 |
+
"grad_norm": 2.4248452186584473,
|
2345 |
+
"learning_rate": 5.075445816186557e-06,
|
2346 |
+
"loss": 0.6985,
|
2347 |
+
"step": 3340
|
2348 |
+
},
|
2349 |
+
{
|
2350 |
+
"epoch": 9.029649595687331,
|
2351 |
+
"grad_norm": 2.76381254196167,
|
2352 |
+
"learning_rate": 4.938271604938272e-06,
|
2353 |
+
"loss": 0.8552,
|
2354 |
+
"step": 3350
|
2355 |
+
},
|
2356 |
+
{
|
2357 |
+
"epoch": 9.056603773584905,
|
2358 |
+
"grad_norm": 0.44387710094451904,
|
2359 |
+
"learning_rate": 4.801097393689986e-06,
|
2360 |
+
"loss": 0.7421,
|
2361 |
+
"step": 3360
|
2362 |
+
},
|
2363 |
+
{
|
2364 |
+
"epoch": 9.08355795148248,
|
2365 |
+
"grad_norm": 1.8279584646224976,
|
2366 |
+
"learning_rate": 4.663923182441701e-06,
|
2367 |
+
"loss": 0.6471,
|
2368 |
+
"step": 3370
|
2369 |
+
},
|
2370 |
+
{
|
2371 |
+
"epoch": 9.110512129380053,
|
2372 |
+
"grad_norm": 4.515883922576904,
|
2373 |
+
"learning_rate": 4.526748971193416e-06,
|
2374 |
+
"loss": 0.8336,
|
2375 |
+
"step": 3380
|
2376 |
+
},
|
2377 |
+
{
|
2378 |
+
"epoch": 9.137466307277627,
|
2379 |
+
"grad_norm": 2.7579915523529053,
|
2380 |
+
"learning_rate": 4.38957475994513e-06,
|
2381 |
+
"loss": 1.0069,
|
2382 |
+
"step": 3390
|
2383 |
+
},
|
2384 |
+
{
|
2385 |
+
"epoch": 9.164420485175203,
|
2386 |
+
"grad_norm": 2.561863660812378,
|
2387 |
+
"learning_rate": 4.252400548696845e-06,
|
2388 |
+
"loss": 0.8613,
|
2389 |
+
"step": 3400
|
2390 |
+
},
|
2391 |
+
{
|
2392 |
+
"epoch": 9.191374663072777,
|
2393 |
+
"grad_norm": 3.5827927589416504,
|
2394 |
+
"learning_rate": 4.11522633744856e-06,
|
2395 |
+
"loss": 0.7177,
|
2396 |
+
"step": 3410
|
2397 |
+
},
|
2398 |
+
{
|
2399 |
+
"epoch": 9.21832884097035,
|
2400 |
+
"grad_norm": 1.109462022781372,
|
2401 |
+
"learning_rate": 3.9780521262002744e-06,
|
2402 |
+
"loss": 0.6787,
|
2403 |
+
"step": 3420
|
2404 |
+
},
|
2405 |
+
{
|
2406 |
+
"epoch": 9.245283018867925,
|
2407 |
+
"grad_norm": 2.3508827686309814,
|
2408 |
+
"learning_rate": 3.84087791495199e-06,
|
2409 |
+
"loss": 0.7854,
|
2410 |
+
"step": 3430
|
2411 |
+
},
|
2412 |
+
{
|
2413 |
+
"epoch": 9.272237196765499,
|
2414 |
+
"grad_norm": 3.206239938735962,
|
2415 |
+
"learning_rate": 3.7037037037037037e-06,
|
2416 |
+
"loss": 0.6083,
|
2417 |
+
"step": 3440
|
2418 |
+
},
|
2419 |
+
{
|
2420 |
+
"epoch": 9.299191374663073,
|
2421 |
+
"grad_norm": 5.175552845001221,
|
2422 |
+
"learning_rate": 3.5665294924554186e-06,
|
2423 |
+
"loss": 0.6398,
|
2424 |
+
"step": 3450
|
2425 |
+
},
|
2426 |
+
{
|
2427 |
+
"epoch": 9.326145552560646,
|
2428 |
+
"grad_norm": 1.54671311378479,
|
2429 |
+
"learning_rate": 3.4293552812071335e-06,
|
2430 |
+
"loss": 1.3045,
|
2431 |
+
"step": 3460
|
2432 |
+
},
|
2433 |
+
{
|
2434 |
+
"epoch": 9.35309973045822,
|
2435 |
+
"grad_norm": 1.7365717887878418,
|
2436 |
+
"learning_rate": 3.2921810699588483e-06,
|
2437 |
+
"loss": 0.6332,
|
2438 |
+
"step": 3470
|
2439 |
+
},
|
2440 |
+
{
|
2441 |
+
"epoch": 9.380053908355794,
|
2442 |
+
"grad_norm": 1.5034387111663818,
|
2443 |
+
"learning_rate": 3.1550068587105624e-06,
|
2444 |
+
"loss": 0.6763,
|
2445 |
+
"step": 3480
|
2446 |
+
},
|
2447 |
+
{
|
2448 |
+
"epoch": 9.40700808625337,
|
2449 |
+
"grad_norm": 3.266697645187378,
|
2450 |
+
"learning_rate": 3.0178326474622772e-06,
|
2451 |
+
"loss": 0.8033,
|
2452 |
+
"step": 3490
|
2453 |
+
},
|
2454 |
+
{
|
2455 |
+
"epoch": 9.433962264150944,
|
2456 |
+
"grad_norm": 1.4214359521865845,
|
2457 |
+
"learning_rate": 2.880658436213992e-06,
|
2458 |
+
"loss": 0.748,
|
2459 |
+
"step": 3500
|
2460 |
+
},
|
2461 |
+
{
|
2462 |
+
"epoch": 9.460916442048518,
|
2463 |
+
"grad_norm": 2.0403311252593994,
|
2464 |
+
"learning_rate": 2.7434842249657065e-06,
|
2465 |
+
"loss": 0.862,
|
2466 |
+
"step": 3510
|
2467 |
+
},
|
2468 |
+
{
|
2469 |
+
"epoch": 9.487870619946092,
|
2470 |
+
"grad_norm": 1.678673505783081,
|
2471 |
+
"learning_rate": 2.6063100137174214e-06,
|
2472 |
+
"loss": 0.6221,
|
2473 |
+
"step": 3520
|
2474 |
+
},
|
2475 |
+
{
|
2476 |
+
"epoch": 9.514824797843666,
|
2477 |
+
"grad_norm": 2.317265033721924,
|
2478 |
+
"learning_rate": 2.469135802469136e-06,
|
2479 |
+
"loss": 0.697,
|
2480 |
+
"step": 3530
|
2481 |
+
},
|
2482 |
+
{
|
2483 |
+
"epoch": 9.54177897574124,
|
2484 |
+
"grad_norm": 2.5591745376586914,
|
2485 |
+
"learning_rate": 2.3319615912208507e-06,
|
2486 |
+
"loss": 0.6209,
|
2487 |
+
"step": 3540
|
2488 |
+
},
|
2489 |
+
{
|
2490 |
+
"epoch": 9.568733153638814,
|
2491 |
+
"grad_norm": 4.573774337768555,
|
2492 |
+
"learning_rate": 2.194787379972565e-06,
|
2493 |
+
"loss": 0.6689,
|
2494 |
+
"step": 3550
|
2495 |
+
},
|
2496 |
+
{
|
2497 |
+
"epoch": 9.595687331536388,
|
2498 |
+
"grad_norm": 1.756940484046936,
|
2499 |
+
"learning_rate": 2.05761316872428e-06,
|
2500 |
+
"loss": 0.9016,
|
2501 |
+
"step": 3560
|
2502 |
+
},
|
2503 |
+
{
|
2504 |
+
"epoch": 9.622641509433961,
|
2505 |
+
"grad_norm": 1.913479208946228,
|
2506 |
+
"learning_rate": 1.920438957475995e-06,
|
2507 |
+
"loss": 0.7732,
|
2508 |
+
"step": 3570
|
2509 |
+
},
|
2510 |
+
{
|
2511 |
+
"epoch": 9.649595687331537,
|
2512 |
+
"grad_norm": 3.303154230117798,
|
2513 |
+
"learning_rate": 1.7832647462277093e-06,
|
2514 |
+
"loss": 1.0092,
|
2515 |
+
"step": 3580
|
2516 |
+
},
|
2517 |
+
{
|
2518 |
+
"epoch": 9.676549865229111,
|
2519 |
+
"grad_norm": 2.583827018737793,
|
2520 |
+
"learning_rate": 1.6460905349794242e-06,
|
2521 |
+
"loss": 1.1913,
|
2522 |
+
"step": 3590
|
2523 |
+
},
|
2524 |
+
{
|
2525 |
+
"epoch": 9.703504043126685,
|
2526 |
+
"grad_norm": 0.6483349800109863,
|
2527 |
+
"learning_rate": 1.5089163237311386e-06,
|
2528 |
+
"loss": 0.8234,
|
2529 |
+
"step": 3600
|
2530 |
+
},
|
2531 |
+
{
|
2532 |
+
"epoch": 9.730458221024259,
|
2533 |
+
"grad_norm": 3.4046685695648193,
|
2534 |
+
"learning_rate": 1.3717421124828533e-06,
|
2535 |
+
"loss": 1.1062,
|
2536 |
+
"step": 3610
|
2537 |
+
},
|
2538 |
+
{
|
2539 |
+
"epoch": 9.757412398921833,
|
2540 |
+
"grad_norm": 1.3275524377822876,
|
2541 |
+
"learning_rate": 1.234567901234568e-06,
|
2542 |
+
"loss": 0.5139,
|
2543 |
+
"step": 3620
|
2544 |
+
},
|
2545 |
+
{
|
2546 |
+
"epoch": 9.784366576819407,
|
2547 |
+
"grad_norm": 2.8111588954925537,
|
2548 |
+
"learning_rate": 1.0973936899862826e-06,
|
2549 |
+
"loss": 0.7716,
|
2550 |
+
"step": 3630
|
2551 |
+
},
|
2552 |
+
{
|
2553 |
+
"epoch": 9.81132075471698,
|
2554 |
+
"grad_norm": 1.3000844717025757,
|
2555 |
+
"learning_rate": 9.602194787379974e-07,
|
2556 |
+
"loss": 0.4894,
|
2557 |
+
"step": 3640
|
2558 |
+
},
|
2559 |
+
{
|
2560 |
+
"epoch": 9.838274932614555,
|
2561 |
+
"grad_norm": 2.7694694995880127,
|
2562 |
+
"learning_rate": 8.230452674897121e-07,
|
2563 |
+
"loss": 0.5647,
|
2564 |
+
"step": 3650
|
2565 |
+
},
|
2566 |
+
{
|
2567 |
+
"epoch": 9.865229110512129,
|
2568 |
+
"grad_norm": 1.6062043905258179,
|
2569 |
+
"learning_rate": 6.858710562414266e-07,
|
2570 |
+
"loss": 0.9145,
|
2571 |
+
"step": 3660
|
2572 |
+
},
|
2573 |
+
{
|
2574 |
+
"epoch": 9.892183288409704,
|
2575 |
+
"grad_norm": 1.323443055152893,
|
2576 |
+
"learning_rate": 5.486968449931413e-07,
|
2577 |
+
"loss": 0.7909,
|
2578 |
+
"step": 3670
|
2579 |
+
},
|
2580 |
+
{
|
2581 |
+
"epoch": 9.919137466307278,
|
2582 |
+
"grad_norm": 3.0170159339904785,
|
2583 |
+
"learning_rate": 4.1152263374485604e-07,
|
2584 |
+
"loss": 0.7927,
|
2585 |
+
"step": 3680
|
2586 |
+
},
|
2587 |
+
{
|
2588 |
+
"epoch": 9.946091644204852,
|
2589 |
+
"grad_norm": 3.5468873977661133,
|
2590 |
+
"learning_rate": 2.7434842249657064e-07,
|
2591 |
+
"loss": 0.7568,
|
2592 |
+
"step": 3690
|
2593 |
+
},
|
2594 |
+
{
|
2595 |
+
"epoch": 9.973045822102426,
|
2596 |
+
"grad_norm": 1.8293559551239014,
|
2597 |
+
"learning_rate": 1.3717421124828532e-07,
|
2598 |
+
"loss": 0.6616,
|
2599 |
+
"step": 3700
|
2600 |
+
},
|
2601 |
+
{
|
2602 |
+
"epoch": 10.0,
|
2603 |
+
"grad_norm": 2.379469633102417,
|
2604 |
+
"learning_rate": 0.0,
|
2605 |
+
"loss": 0.3192,
|
2606 |
+
"step": 3710
|
2607 |
+
}
|
2608 |
+
],
|
2609 |
+
"logging_steps": 10,
|
2610 |
+
"max_steps": 3710,
|
2611 |
+
"num_input_tokens_seen": 0,
|
2612 |
+
"num_train_epochs": 10,
|
2613 |
+
"save_steps": 10000,
|
2614 |
+
"stateful_callbacks": {
|
2615 |
+
"TrainerControl": {
|
2616 |
+
"args": {
|
2617 |
+
"should_epoch_stop": false,
|
2618 |
+
"should_evaluate": false,
|
2619 |
+
"should_log": false,
|
2620 |
+
"should_save": true,
|
2621 |
+
"should_training_stop": true
|
2622 |
+
},
|
2623 |
+
"attributes": {}
|
2624 |
+
}
|
2625 |
+
},
|
2626 |
+
"total_flos": 3062070293667840.0,
|
2627 |
+
"train_batch_size": 1,
|
2628 |
+
"trial_name": null,
|
2629 |
+
"trial_params": null
|
2630 |
+
}
|
training_args.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:503c99a98dc4b9930816df475c72f5df1f7dc0cf3562e030e4261835eda54780
|
3 |
+
size 5112
|