jm9095 commited on
Commit ·
5c08a6d
1
Parent(s): f4178de
Pilot
Browse files- config.json +25 -0
- generation_config.json +5 -0
- model.safetensors +3 -0
- optimizer.pt +3 -0
- rng_state.pth +3 -0
- scheduler.pt +3 -0
- trainer_state.json +2823 -0
- training_args.bin +3 -0
config.json
ADDED
|
@@ -0,0 +1,25 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "/scratch/gpfs/BG11/logion-resources/cache/logion-bert-base-model",
|
| 3 |
+
"architectures": [
|
| 4 |
+
"BertForMaskedLM"
|
| 5 |
+
],
|
| 6 |
+
"attention_probs_dropout_prob": 0.1,
|
| 7 |
+
"classifier_dropout": null,
|
| 8 |
+
"hidden_act": "gelu",
|
| 9 |
+
"hidden_dropout_prob": 0.1,
|
| 10 |
+
"hidden_size": 768,
|
| 11 |
+
"initializer_range": 0.02,
|
| 12 |
+
"intermediate_size": 3072,
|
| 13 |
+
"layer_norm_eps": 1e-12,
|
| 14 |
+
"max_position_embeddings": 512,
|
| 15 |
+
"model_type": "bert",
|
| 16 |
+
"num_attention_heads": 12,
|
| 17 |
+
"num_hidden_layers": 12,
|
| 18 |
+
"pad_token_id": 0,
|
| 19 |
+
"position_embedding_type": "absolute",
|
| 20 |
+
"torch_dtype": "float32",
|
| 21 |
+
"transformers_version": "4.48.3",
|
| 22 |
+
"type_vocab_size": 2,
|
| 23 |
+
"use_cache": true,
|
| 24 |
+
"vocab_size": 50000
|
| 25 |
+
}
|
generation_config.json
ADDED
|
@@ -0,0 +1,5 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_from_model_config": true,
|
| 3 |
+
"pad_token_id": 0,
|
| 4 |
+
"transformers_version": "4.48.3"
|
| 5 |
+
}
|
model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:2ae2e8d99b4f73b431005486e6cc543e9cb3bc77c2a6bf9ab45ea6f440e38f91
|
| 3 |
+
size 497995232
|
optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:2d2a747ef1b9f357cae03a6945fad8ebbabc998da404c3062fc4cf5fb157b150
|
| 3 |
+
size 996111994
|
rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:45c0a83d33594387c957a7ffa4fbf7e539e07b4ddc50e3efbd985c826f540f39
|
| 3 |
+
size 14244
|
scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:347e7053a75d1510fc29ff59939af8dd5a1c48bcdb6ccd1a3fee5abb9605784a
|
| 3 |
+
size 1064
|
trainer_state.json
ADDED
|
@@ -0,0 +1,2823 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 1.3527668714523315,
|
| 3 |
+
"best_model_checkpoint": "/scratch/gpfs/BG11/suze-mdl-steps/model_combo_10.txt/checkpoint-22506",
|
| 4 |
+
"epoch": 46.118852459016395,
|
| 5 |
+
"eval_steps": 121,
|
| 6 |
+
"global_step": 22506,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.24795081967213115,
|
| 13 |
+
"grad_norm": 6.242262363433838,
|
| 14 |
+
"learning_rate": 1.990081967213115e-05,
|
| 15 |
+
"loss": 1.5517,
|
| 16 |
+
"step": 121
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.24795081967213115,
|
| 20 |
+
"eval_loss": 1.451869249343872,
|
| 21 |
+
"eval_runtime": 4.1207,
|
| 22 |
+
"eval_samples_per_second": 105.321,
|
| 23 |
+
"eval_steps_per_second": 13.347,
|
| 24 |
+
"step": 121
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.4959016393442623,
|
| 28 |
+
"grad_norm": 5.890889644622803,
|
| 29 |
+
"learning_rate": 1.9801639344262295e-05,
|
| 30 |
+
"loss": 1.5921,
|
| 31 |
+
"step": 242
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.4959016393442623,
|
| 35 |
+
"eval_loss": 1.410286784172058,
|
| 36 |
+
"eval_runtime": 4.1324,
|
| 37 |
+
"eval_samples_per_second": 105.023,
|
| 38 |
+
"eval_steps_per_second": 13.309,
|
| 39 |
+
"step": 242
|
| 40 |
+
},
|
| 41 |
+
{
|
| 42 |
+
"epoch": 0.7438524590163934,
|
| 43 |
+
"grad_norm": 5.299322128295898,
|
| 44 |
+
"learning_rate": 1.9702459016393446e-05,
|
| 45 |
+
"loss": 1.6134,
|
| 46 |
+
"step": 363
|
| 47 |
+
},
|
| 48 |
+
{
|
| 49 |
+
"epoch": 0.7438524590163934,
|
| 50 |
+
"eval_loss": 1.4782627820968628,
|
| 51 |
+
"eval_runtime": 4.1505,
|
| 52 |
+
"eval_samples_per_second": 104.567,
|
| 53 |
+
"eval_steps_per_second": 13.252,
|
| 54 |
+
"step": 363
|
| 55 |
+
},
|
| 56 |
+
{
|
| 57 |
+
"epoch": 0.9918032786885246,
|
| 58 |
+
"grad_norm": 5.992594242095947,
|
| 59 |
+
"learning_rate": 1.960327868852459e-05,
|
| 60 |
+
"loss": 1.554,
|
| 61 |
+
"step": 484
|
| 62 |
+
},
|
| 63 |
+
{
|
| 64 |
+
"epoch": 0.9918032786885246,
|
| 65 |
+
"eval_loss": 1.4408639669418335,
|
| 66 |
+
"eval_runtime": 4.1339,
|
| 67 |
+
"eval_samples_per_second": 104.987,
|
| 68 |
+
"eval_steps_per_second": 13.305,
|
| 69 |
+
"step": 484
|
| 70 |
+
},
|
| 71 |
+
{
|
| 72 |
+
"epoch": 1.2397540983606556,
|
| 73 |
+
"grad_norm": 5.52030611038208,
|
| 74 |
+
"learning_rate": 1.950409836065574e-05,
|
| 75 |
+
"loss": 1.5205,
|
| 76 |
+
"step": 605
|
| 77 |
+
},
|
| 78 |
+
{
|
| 79 |
+
"epoch": 1.2397540983606556,
|
| 80 |
+
"eval_loss": 1.4437590837478638,
|
| 81 |
+
"eval_runtime": 4.1338,
|
| 82 |
+
"eval_samples_per_second": 104.988,
|
| 83 |
+
"eval_steps_per_second": 13.305,
|
| 84 |
+
"step": 605
|
| 85 |
+
},
|
| 86 |
+
{
|
| 87 |
+
"epoch": 1.4877049180327868,
|
| 88 |
+
"grad_norm": 5.520483016967773,
|
| 89 |
+
"learning_rate": 1.9404918032786887e-05,
|
| 90 |
+
"loss": 1.5542,
|
| 91 |
+
"step": 726
|
| 92 |
+
},
|
| 93 |
+
{
|
| 94 |
+
"epoch": 1.4877049180327868,
|
| 95 |
+
"eval_loss": 1.4612840414047241,
|
| 96 |
+
"eval_runtime": 4.1326,
|
| 97 |
+
"eval_samples_per_second": 105.019,
|
| 98 |
+
"eval_steps_per_second": 13.309,
|
| 99 |
+
"step": 726
|
| 100 |
+
},
|
| 101 |
+
{
|
| 102 |
+
"epoch": 1.735655737704918,
|
| 103 |
+
"grad_norm": 5.218242168426514,
|
| 104 |
+
"learning_rate": 1.9305737704918036e-05,
|
| 105 |
+
"loss": 1.5079,
|
| 106 |
+
"step": 847
|
| 107 |
+
},
|
| 108 |
+
{
|
| 109 |
+
"epoch": 1.735655737704918,
|
| 110 |
+
"eval_loss": 1.4480838775634766,
|
| 111 |
+
"eval_runtime": 4.13,
|
| 112 |
+
"eval_samples_per_second": 105.086,
|
| 113 |
+
"eval_steps_per_second": 13.317,
|
| 114 |
+
"step": 847
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 1.9836065573770492,
|
| 118 |
+
"grad_norm": 5.828314304351807,
|
| 119 |
+
"learning_rate": 1.920655737704918e-05,
|
| 120 |
+
"loss": 1.525,
|
| 121 |
+
"step": 968
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 1.9836065573770492,
|
| 125 |
+
"eval_loss": 1.4295066595077515,
|
| 126 |
+
"eval_runtime": 4.129,
|
| 127 |
+
"eval_samples_per_second": 105.11,
|
| 128 |
+
"eval_steps_per_second": 13.32,
|
| 129 |
+
"step": 968
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 2.2315573770491803,
|
| 133 |
+
"grad_norm": 6.473424911499023,
|
| 134 |
+
"learning_rate": 1.910737704918033e-05,
|
| 135 |
+
"loss": 1.5396,
|
| 136 |
+
"step": 1089
|
| 137 |
+
},
|
| 138 |
+
{
|
| 139 |
+
"epoch": 2.2315573770491803,
|
| 140 |
+
"eval_loss": 1.4381992816925049,
|
| 141 |
+
"eval_runtime": 4.1264,
|
| 142 |
+
"eval_samples_per_second": 105.178,
|
| 143 |
+
"eval_steps_per_second": 13.329,
|
| 144 |
+
"step": 1089
|
| 145 |
+
},
|
| 146 |
+
{
|
| 147 |
+
"epoch": 2.4795081967213113,
|
| 148 |
+
"grad_norm": 5.705435276031494,
|
| 149 |
+
"learning_rate": 1.9008196721311477e-05,
|
| 150 |
+
"loss": 1.4863,
|
| 151 |
+
"step": 1210
|
| 152 |
+
},
|
| 153 |
+
{
|
| 154 |
+
"epoch": 2.4795081967213113,
|
| 155 |
+
"eval_loss": 1.4149571657180786,
|
| 156 |
+
"eval_runtime": 4.1261,
|
| 157 |
+
"eval_samples_per_second": 105.183,
|
| 158 |
+
"eval_steps_per_second": 13.33,
|
| 159 |
+
"step": 1210
|
| 160 |
+
},
|
| 161 |
+
{
|
| 162 |
+
"epoch": 2.7274590163934427,
|
| 163 |
+
"grad_norm": 6.025406837463379,
|
| 164 |
+
"learning_rate": 1.8909016393442625e-05,
|
| 165 |
+
"loss": 1.497,
|
| 166 |
+
"step": 1331
|
| 167 |
+
},
|
| 168 |
+
{
|
| 169 |
+
"epoch": 2.7274590163934427,
|
| 170 |
+
"eval_loss": 1.4381588697433472,
|
| 171 |
+
"eval_runtime": 4.1242,
|
| 172 |
+
"eval_samples_per_second": 105.232,
|
| 173 |
+
"eval_steps_per_second": 13.336,
|
| 174 |
+
"step": 1331
|
| 175 |
+
},
|
| 176 |
+
{
|
| 177 |
+
"epoch": 2.9754098360655736,
|
| 178 |
+
"grad_norm": 5.826422214508057,
|
| 179 |
+
"learning_rate": 1.8809836065573773e-05,
|
| 180 |
+
"loss": 1.5218,
|
| 181 |
+
"step": 1452
|
| 182 |
+
},
|
| 183 |
+
{
|
| 184 |
+
"epoch": 2.9754098360655736,
|
| 185 |
+
"eval_loss": 1.4456675052642822,
|
| 186 |
+
"eval_runtime": 4.1254,
|
| 187 |
+
"eval_samples_per_second": 105.203,
|
| 188 |
+
"eval_steps_per_second": 13.332,
|
| 189 |
+
"step": 1452
|
| 190 |
+
},
|
| 191 |
+
{
|
| 192 |
+
"epoch": 3.223360655737705,
|
| 193 |
+
"grad_norm": 5.673182010650635,
|
| 194 |
+
"learning_rate": 1.8710655737704918e-05,
|
| 195 |
+
"loss": 1.4819,
|
| 196 |
+
"step": 1573
|
| 197 |
+
},
|
| 198 |
+
{
|
| 199 |
+
"epoch": 3.223360655737705,
|
| 200 |
+
"eval_loss": 1.4289182424545288,
|
| 201 |
+
"eval_runtime": 4.1236,
|
| 202 |
+
"eval_samples_per_second": 105.249,
|
| 203 |
+
"eval_steps_per_second": 13.338,
|
| 204 |
+
"step": 1573
|
| 205 |
+
},
|
| 206 |
+
{
|
| 207 |
+
"epoch": 3.471311475409836,
|
| 208 |
+
"grad_norm": 5.801868438720703,
|
| 209 |
+
"learning_rate": 1.8611475409836066e-05,
|
| 210 |
+
"loss": 1.4607,
|
| 211 |
+
"step": 1694
|
| 212 |
+
},
|
| 213 |
+
{
|
| 214 |
+
"epoch": 3.471311475409836,
|
| 215 |
+
"eval_loss": 1.4265520572662354,
|
| 216 |
+
"eval_runtime": 4.1254,
|
| 217 |
+
"eval_samples_per_second": 105.202,
|
| 218 |
+
"eval_steps_per_second": 13.332,
|
| 219 |
+
"step": 1694
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 3.7192622950819674,
|
| 223 |
+
"grad_norm": 6.084210395812988,
|
| 224 |
+
"learning_rate": 1.8512295081967214e-05,
|
| 225 |
+
"loss": 1.5112,
|
| 226 |
+
"step": 1815
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 3.7192622950819674,
|
| 230 |
+
"eval_loss": 1.4315944910049438,
|
| 231 |
+
"eval_runtime": 4.1231,
|
| 232 |
+
"eval_samples_per_second": 105.26,
|
| 233 |
+
"eval_steps_per_second": 13.339,
|
| 234 |
+
"step": 1815
|
| 235 |
+
},
|
| 236 |
+
{
|
| 237 |
+
"epoch": 3.9672131147540983,
|
| 238 |
+
"grad_norm": 5.915734767913818,
|
| 239 |
+
"learning_rate": 1.8413114754098362e-05,
|
| 240 |
+
"loss": 1.4653,
|
| 241 |
+
"step": 1936
|
| 242 |
+
},
|
| 243 |
+
{
|
| 244 |
+
"epoch": 3.9672131147540983,
|
| 245 |
+
"eval_loss": 1.4230977296829224,
|
| 246 |
+
"eval_runtime": 4.1241,
|
| 247 |
+
"eval_samples_per_second": 105.235,
|
| 248 |
+
"eval_steps_per_second": 13.336,
|
| 249 |
+
"step": 1936
|
| 250 |
+
},
|
| 251 |
+
{
|
| 252 |
+
"epoch": 4.215163934426229,
|
| 253 |
+
"grad_norm": 5.40634822845459,
|
| 254 |
+
"learning_rate": 1.831393442622951e-05,
|
| 255 |
+
"loss": 1.4445,
|
| 256 |
+
"step": 2057
|
| 257 |
+
},
|
| 258 |
+
{
|
| 259 |
+
"epoch": 4.215163934426229,
|
| 260 |
+
"eval_loss": 1.4059760570526123,
|
| 261 |
+
"eval_runtime": 4.1255,
|
| 262 |
+
"eval_samples_per_second": 105.2,
|
| 263 |
+
"eval_steps_per_second": 13.332,
|
| 264 |
+
"step": 2057
|
| 265 |
+
},
|
| 266 |
+
{
|
| 267 |
+
"epoch": 4.463114754098361,
|
| 268 |
+
"grad_norm": 5.65101432800293,
|
| 269 |
+
"learning_rate": 1.8214754098360655e-05,
|
| 270 |
+
"loss": 1.4843,
|
| 271 |
+
"step": 2178
|
| 272 |
+
},
|
| 273 |
+
{
|
| 274 |
+
"epoch": 4.463114754098361,
|
| 275 |
+
"eval_loss": 1.4199328422546387,
|
| 276 |
+
"eval_runtime": 4.1264,
|
| 277 |
+
"eval_samples_per_second": 105.178,
|
| 278 |
+
"eval_steps_per_second": 13.329,
|
| 279 |
+
"step": 2178
|
| 280 |
+
},
|
| 281 |
+
{
|
| 282 |
+
"epoch": 4.711065573770492,
|
| 283 |
+
"grad_norm": 5.768662452697754,
|
| 284 |
+
"learning_rate": 1.8115573770491807e-05,
|
| 285 |
+
"loss": 1.4737,
|
| 286 |
+
"step": 2299
|
| 287 |
+
},
|
| 288 |
+
{
|
| 289 |
+
"epoch": 4.711065573770492,
|
| 290 |
+
"eval_loss": 1.4175463914871216,
|
| 291 |
+
"eval_runtime": 4.1255,
|
| 292 |
+
"eval_samples_per_second": 105.2,
|
| 293 |
+
"eval_steps_per_second": 13.332,
|
| 294 |
+
"step": 2299
|
| 295 |
+
},
|
| 296 |
+
{
|
| 297 |
+
"epoch": 4.959016393442623,
|
| 298 |
+
"grad_norm": 5.345828056335449,
|
| 299 |
+
"learning_rate": 1.8016393442622952e-05,
|
| 300 |
+
"loss": 1.4308,
|
| 301 |
+
"step": 2420
|
| 302 |
+
},
|
| 303 |
+
{
|
| 304 |
+
"epoch": 4.959016393442623,
|
| 305 |
+
"eval_loss": 1.3921349048614502,
|
| 306 |
+
"eval_runtime": 4.1245,
|
| 307 |
+
"eval_samples_per_second": 105.224,
|
| 308 |
+
"eval_steps_per_second": 13.335,
|
| 309 |
+
"step": 2420
|
| 310 |
+
},
|
| 311 |
+
{
|
| 312 |
+
"epoch": 5.206967213114754,
|
| 313 |
+
"grad_norm": 5.558488845825195,
|
| 314 |
+
"learning_rate": 1.79172131147541e-05,
|
| 315 |
+
"loss": 1.4288,
|
| 316 |
+
"step": 2541
|
| 317 |
+
},
|
| 318 |
+
{
|
| 319 |
+
"epoch": 5.206967213114754,
|
| 320 |
+
"eval_loss": 1.4271553754806519,
|
| 321 |
+
"eval_runtime": 4.123,
|
| 322 |
+
"eval_samples_per_second": 105.263,
|
| 323 |
+
"eval_steps_per_second": 13.34,
|
| 324 |
+
"step": 2541
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 5.454918032786885,
|
| 328 |
+
"grad_norm": 5.293428897857666,
|
| 329 |
+
"learning_rate": 1.7818032786885248e-05,
|
| 330 |
+
"loss": 1.4291,
|
| 331 |
+
"step": 2662
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 5.454918032786885,
|
| 335 |
+
"eval_loss": 1.430025577545166,
|
| 336 |
+
"eval_runtime": 4.1248,
|
| 337 |
+
"eval_samples_per_second": 105.218,
|
| 338 |
+
"eval_steps_per_second": 13.334,
|
| 339 |
+
"step": 2662
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 5.702868852459017,
|
| 343 |
+
"grad_norm": 5.373508453369141,
|
| 344 |
+
"learning_rate": 1.7718852459016396e-05,
|
| 345 |
+
"loss": 1.4616,
|
| 346 |
+
"step": 2783
|
| 347 |
+
},
|
| 348 |
+
{
|
| 349 |
+
"epoch": 5.702868852459017,
|
| 350 |
+
"eval_loss": 1.452791452407837,
|
| 351 |
+
"eval_runtime": 4.1221,
|
| 352 |
+
"eval_samples_per_second": 105.286,
|
| 353 |
+
"eval_steps_per_second": 13.343,
|
| 354 |
+
"step": 2783
|
| 355 |
+
},
|
| 356 |
+
{
|
| 357 |
+
"epoch": 5.950819672131147,
|
| 358 |
+
"grad_norm": 4.782499313354492,
|
| 359 |
+
"learning_rate": 1.761967213114754e-05,
|
| 360 |
+
"loss": 1.4241,
|
| 361 |
+
"step": 2904
|
| 362 |
+
},
|
| 363 |
+
{
|
| 364 |
+
"epoch": 5.950819672131147,
|
| 365 |
+
"eval_loss": 1.4025797843933105,
|
| 366 |
+
"eval_runtime": 4.1264,
|
| 367 |
+
"eval_samples_per_second": 105.176,
|
| 368 |
+
"eval_steps_per_second": 13.329,
|
| 369 |
+
"step": 2904
|
| 370 |
+
},
|
| 371 |
+
{
|
| 372 |
+
"epoch": 6.198770491803279,
|
| 373 |
+
"grad_norm": 5.344179630279541,
|
| 374 |
+
"learning_rate": 1.752049180327869e-05,
|
| 375 |
+
"loss": 1.4187,
|
| 376 |
+
"step": 3025
|
| 377 |
+
},
|
| 378 |
+
{
|
| 379 |
+
"epoch": 6.198770491803279,
|
| 380 |
+
"eval_loss": 1.4222002029418945,
|
| 381 |
+
"eval_runtime": 4.1238,
|
| 382 |
+
"eval_samples_per_second": 105.242,
|
| 383 |
+
"eval_steps_per_second": 13.337,
|
| 384 |
+
"step": 3025
|
| 385 |
+
},
|
| 386 |
+
{
|
| 387 |
+
"epoch": 6.44672131147541,
|
| 388 |
+
"grad_norm": 5.498497009277344,
|
| 389 |
+
"learning_rate": 1.7421311475409838e-05,
|
| 390 |
+
"loss": 1.4179,
|
| 391 |
+
"step": 3146
|
| 392 |
+
},
|
| 393 |
+
{
|
| 394 |
+
"epoch": 6.44672131147541,
|
| 395 |
+
"eval_loss": 1.425389051437378,
|
| 396 |
+
"eval_runtime": 4.1255,
|
| 397 |
+
"eval_samples_per_second": 105.201,
|
| 398 |
+
"eval_steps_per_second": 13.332,
|
| 399 |
+
"step": 3146
|
| 400 |
+
},
|
| 401 |
+
{
|
| 402 |
+
"epoch": 6.6946721311475414,
|
| 403 |
+
"grad_norm": 6.103499412536621,
|
| 404 |
+
"learning_rate": 1.7322131147540986e-05,
|
| 405 |
+
"loss": 1.4036,
|
| 406 |
+
"step": 3267
|
| 407 |
+
},
|
| 408 |
+
{
|
| 409 |
+
"epoch": 6.6946721311475414,
|
| 410 |
+
"eval_loss": 1.4255503416061401,
|
| 411 |
+
"eval_runtime": 4.1267,
|
| 412 |
+
"eval_samples_per_second": 105.168,
|
| 413 |
+
"eval_steps_per_second": 13.328,
|
| 414 |
+
"step": 3267
|
| 415 |
+
},
|
| 416 |
+
{
|
| 417 |
+
"epoch": 6.942622950819672,
|
| 418 |
+
"grad_norm": 5.6141462326049805,
|
| 419 |
+
"learning_rate": 1.7222950819672134e-05,
|
| 420 |
+
"loss": 1.4237,
|
| 421 |
+
"step": 3388
|
| 422 |
+
},
|
| 423 |
+
{
|
| 424 |
+
"epoch": 6.942622950819672,
|
| 425 |
+
"eval_loss": 1.4209522008895874,
|
| 426 |
+
"eval_runtime": 4.1246,
|
| 427 |
+
"eval_samples_per_second": 105.221,
|
| 428 |
+
"eval_steps_per_second": 13.334,
|
| 429 |
+
"step": 3388
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 7.190573770491803,
|
| 433 |
+
"grad_norm": 6.086358547210693,
|
| 434 |
+
"learning_rate": 1.712377049180328e-05,
|
| 435 |
+
"loss": 1.3858,
|
| 436 |
+
"step": 3509
|
| 437 |
+
},
|
| 438 |
+
{
|
| 439 |
+
"epoch": 7.190573770491803,
|
| 440 |
+
"eval_loss": 1.4121168851852417,
|
| 441 |
+
"eval_runtime": 4.1269,
|
| 442 |
+
"eval_samples_per_second": 105.163,
|
| 443 |
+
"eval_steps_per_second": 13.327,
|
| 444 |
+
"step": 3509
|
| 445 |
+
},
|
| 446 |
+
{
|
| 447 |
+
"epoch": 7.438524590163935,
|
| 448 |
+
"grad_norm": 5.1807637214660645,
|
| 449 |
+
"learning_rate": 1.7024590163934427e-05,
|
| 450 |
+
"loss": 1.4206,
|
| 451 |
+
"step": 3630
|
| 452 |
+
},
|
| 453 |
+
{
|
| 454 |
+
"epoch": 7.438524590163935,
|
| 455 |
+
"eval_loss": 1.421557903289795,
|
| 456 |
+
"eval_runtime": 4.1232,
|
| 457 |
+
"eval_samples_per_second": 105.259,
|
| 458 |
+
"eval_steps_per_second": 13.339,
|
| 459 |
+
"step": 3630
|
| 460 |
+
},
|
| 461 |
+
{
|
| 462 |
+
"epoch": 7.686475409836065,
|
| 463 |
+
"grad_norm": 5.515040397644043,
|
| 464 |
+
"learning_rate": 1.6925409836065575e-05,
|
| 465 |
+
"loss": 1.4213,
|
| 466 |
+
"step": 3751
|
| 467 |
+
},
|
| 468 |
+
{
|
| 469 |
+
"epoch": 7.686475409836065,
|
| 470 |
+
"eval_loss": 1.392737627029419,
|
| 471 |
+
"eval_runtime": 4.1243,
|
| 472 |
+
"eval_samples_per_second": 105.23,
|
| 473 |
+
"eval_steps_per_second": 13.336,
|
| 474 |
+
"step": 3751
|
| 475 |
+
},
|
| 476 |
+
{
|
| 477 |
+
"epoch": 7.934426229508197,
|
| 478 |
+
"grad_norm": 5.375150203704834,
|
| 479 |
+
"learning_rate": 1.6826229508196723e-05,
|
| 480 |
+
"loss": 1.3999,
|
| 481 |
+
"step": 3872
|
| 482 |
+
},
|
| 483 |
+
{
|
| 484 |
+
"epoch": 7.934426229508197,
|
| 485 |
+
"eval_loss": 1.425716519355774,
|
| 486 |
+
"eval_runtime": 4.1231,
|
| 487 |
+
"eval_samples_per_second": 105.259,
|
| 488 |
+
"eval_steps_per_second": 13.339,
|
| 489 |
+
"step": 3872
|
| 490 |
+
},
|
| 491 |
+
{
|
| 492 |
+
"epoch": 8.182377049180328,
|
| 493 |
+
"grad_norm": 5.733981609344482,
|
| 494 |
+
"learning_rate": 1.672704918032787e-05,
|
| 495 |
+
"loss": 1.4004,
|
| 496 |
+
"step": 3993
|
| 497 |
+
},
|
| 498 |
+
{
|
| 499 |
+
"epoch": 8.182377049180328,
|
| 500 |
+
"eval_loss": 1.4038772583007812,
|
| 501 |
+
"eval_runtime": 4.122,
|
| 502 |
+
"eval_samples_per_second": 105.29,
|
| 503 |
+
"eval_steps_per_second": 13.343,
|
| 504 |
+
"step": 3993
|
| 505 |
+
},
|
| 506 |
+
{
|
| 507 |
+
"epoch": 8.430327868852459,
|
| 508 |
+
"grad_norm": 5.9191083908081055,
|
| 509 |
+
"learning_rate": 1.6627868852459016e-05,
|
| 510 |
+
"loss": 1.3726,
|
| 511 |
+
"step": 4114
|
| 512 |
+
},
|
| 513 |
+
{
|
| 514 |
+
"epoch": 8.430327868852459,
|
| 515 |
+
"eval_loss": 1.3957583904266357,
|
| 516 |
+
"eval_runtime": 4.122,
|
| 517 |
+
"eval_samples_per_second": 105.288,
|
| 518 |
+
"eval_steps_per_second": 13.343,
|
| 519 |
+
"step": 4114
|
| 520 |
+
},
|
| 521 |
+
{
|
| 522 |
+
"epoch": 8.67827868852459,
|
| 523 |
+
"grad_norm": 5.487318992614746,
|
| 524 |
+
"learning_rate": 1.6528688524590168e-05,
|
| 525 |
+
"loss": 1.385,
|
| 526 |
+
"step": 4235
|
| 527 |
+
},
|
| 528 |
+
{
|
| 529 |
+
"epoch": 8.67827868852459,
|
| 530 |
+
"eval_loss": 1.457805871963501,
|
| 531 |
+
"eval_runtime": 4.177,
|
| 532 |
+
"eval_samples_per_second": 103.902,
|
| 533 |
+
"eval_steps_per_second": 13.167,
|
| 534 |
+
"step": 4235
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 8.926229508196721,
|
| 538 |
+
"grad_norm": 5.596704959869385,
|
| 539 |
+
"learning_rate": 1.6429508196721313e-05,
|
| 540 |
+
"loss": 1.401,
|
| 541 |
+
"step": 4356
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 8.926229508196721,
|
| 545 |
+
"eval_loss": 1.3986533880233765,
|
| 546 |
+
"eval_runtime": 4.1733,
|
| 547 |
+
"eval_samples_per_second": 103.994,
|
| 548 |
+
"eval_steps_per_second": 13.179,
|
| 549 |
+
"step": 4356
|
| 550 |
+
},
|
| 551 |
+
{
|
| 552 |
+
"epoch": 9.174180327868852,
|
| 553 |
+
"grad_norm": 5.978359222412109,
|
| 554 |
+
"learning_rate": 1.633032786885246e-05,
|
| 555 |
+
"loss": 1.3846,
|
| 556 |
+
"step": 4477
|
| 557 |
+
},
|
| 558 |
+
{
|
| 559 |
+
"epoch": 9.174180327868852,
|
| 560 |
+
"eval_loss": 1.4179773330688477,
|
| 561 |
+
"eval_runtime": 4.1466,
|
| 562 |
+
"eval_samples_per_second": 104.664,
|
| 563 |
+
"eval_steps_per_second": 13.264,
|
| 564 |
+
"step": 4477
|
| 565 |
+
},
|
| 566 |
+
{
|
| 567 |
+
"epoch": 9.422131147540984,
|
| 568 |
+
"grad_norm": 5.524637222290039,
|
| 569 |
+
"learning_rate": 1.6231147540983606e-05,
|
| 570 |
+
"loss": 1.3677,
|
| 571 |
+
"step": 4598
|
| 572 |
+
},
|
| 573 |
+
{
|
| 574 |
+
"epoch": 9.422131147540984,
|
| 575 |
+
"eval_loss": 1.41084623336792,
|
| 576 |
+
"eval_runtime": 4.1689,
|
| 577 |
+
"eval_samples_per_second": 104.103,
|
| 578 |
+
"eval_steps_per_second": 13.193,
|
| 579 |
+
"step": 4598
|
| 580 |
+
},
|
| 581 |
+
{
|
| 582 |
+
"epoch": 9.670081967213115,
|
| 583 |
+
"grad_norm": 5.55858039855957,
|
| 584 |
+
"learning_rate": 1.6131967213114757e-05,
|
| 585 |
+
"loss": 1.366,
|
| 586 |
+
"step": 4719
|
| 587 |
+
},
|
| 588 |
+
{
|
| 589 |
+
"epoch": 9.670081967213115,
|
| 590 |
+
"eval_loss": 1.404625415802002,
|
| 591 |
+
"eval_runtime": 4.1425,
|
| 592 |
+
"eval_samples_per_second": 104.768,
|
| 593 |
+
"eval_steps_per_second": 13.277,
|
| 594 |
+
"step": 4719
|
| 595 |
+
},
|
| 596 |
+
{
|
| 597 |
+
"epoch": 9.918032786885245,
|
| 598 |
+
"grad_norm": 5.1674275398254395,
|
| 599 |
+
"learning_rate": 1.6032786885245902e-05,
|
| 600 |
+
"loss": 1.3817,
|
| 601 |
+
"step": 4840
|
| 602 |
+
},
|
| 603 |
+
{
|
| 604 |
+
"epoch": 9.918032786885245,
|
| 605 |
+
"eval_loss": 1.4428249597549438,
|
| 606 |
+
"eval_runtime": 4.1564,
|
| 607 |
+
"eval_samples_per_second": 104.418,
|
| 608 |
+
"eval_steps_per_second": 13.233,
|
| 609 |
+
"step": 4840
|
| 610 |
+
},
|
| 611 |
+
{
|
| 612 |
+
"epoch": 10.165983606557377,
|
| 613 |
+
"grad_norm": 5.132227897644043,
|
| 614 |
+
"learning_rate": 1.593360655737705e-05,
|
| 615 |
+
"loss": 1.3653,
|
| 616 |
+
"step": 4961
|
| 617 |
+
},
|
| 618 |
+
{
|
| 619 |
+
"epoch": 10.165983606557377,
|
| 620 |
+
"eval_loss": 1.3900526762008667,
|
| 621 |
+
"eval_runtime": 4.153,
|
| 622 |
+
"eval_samples_per_second": 104.502,
|
| 623 |
+
"eval_steps_per_second": 13.243,
|
| 624 |
+
"step": 4961
|
| 625 |
+
},
|
| 626 |
+
{
|
| 627 |
+
"epoch": 10.413934426229508,
|
| 628 |
+
"grad_norm": 6.354508876800537,
|
| 629 |
+
"learning_rate": 1.5834426229508198e-05,
|
| 630 |
+
"loss": 1.3543,
|
| 631 |
+
"step": 5082
|
| 632 |
+
},
|
| 633 |
+
{
|
| 634 |
+
"epoch": 10.413934426229508,
|
| 635 |
+
"eval_loss": 1.415820598602295,
|
| 636 |
+
"eval_runtime": 4.155,
|
| 637 |
+
"eval_samples_per_second": 104.452,
|
| 638 |
+
"eval_steps_per_second": 13.237,
|
| 639 |
+
"step": 5082
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 10.66188524590164,
|
| 643 |
+
"grad_norm": 5.173457622528076,
|
| 644 |
+
"learning_rate": 1.5735245901639346e-05,
|
| 645 |
+
"loss": 1.3692,
|
| 646 |
+
"step": 5203
|
| 647 |
+
},
|
| 648 |
+
{
|
| 649 |
+
"epoch": 10.66188524590164,
|
| 650 |
+
"eval_loss": 1.3986164331436157,
|
| 651 |
+
"eval_runtime": 4.1614,
|
| 652 |
+
"eval_samples_per_second": 104.292,
|
| 653 |
+
"eval_steps_per_second": 13.217,
|
| 654 |
+
"step": 5203
|
| 655 |
+
},
|
| 656 |
+
{
|
| 657 |
+
"epoch": 10.90983606557377,
|
| 658 |
+
"grad_norm": 5.250792026519775,
|
| 659 |
+
"learning_rate": 1.5636065573770495e-05,
|
| 660 |
+
"loss": 1.364,
|
| 661 |
+
"step": 5324
|
| 662 |
+
},
|
| 663 |
+
{
|
| 664 |
+
"epoch": 10.90983606557377,
|
| 665 |
+
"eval_loss": 1.4086618423461914,
|
| 666 |
+
"eval_runtime": 4.1657,
|
| 667 |
+
"eval_samples_per_second": 104.185,
|
| 668 |
+
"eval_steps_per_second": 13.203,
|
| 669 |
+
"step": 5324
|
| 670 |
+
},
|
| 671 |
+
{
|
| 672 |
+
"epoch": 11.157786885245901,
|
| 673 |
+
"grad_norm": 4.989199161529541,
|
| 674 |
+
"learning_rate": 1.553688524590164e-05,
|
| 675 |
+
"loss": 1.3471,
|
| 676 |
+
"step": 5445
|
| 677 |
+
},
|
| 678 |
+
{
|
| 679 |
+
"epoch": 11.157786885245901,
|
| 680 |
+
"eval_loss": 1.4073328971862793,
|
| 681 |
+
"eval_runtime": 4.1729,
|
| 682 |
+
"eval_samples_per_second": 104.003,
|
| 683 |
+
"eval_steps_per_second": 13.18,
|
| 684 |
+
"step": 5445
|
| 685 |
+
},
|
| 686 |
+
{
|
| 687 |
+
"epoch": 11.405737704918034,
|
| 688 |
+
"grad_norm": 5.10518741607666,
|
| 689 |
+
"learning_rate": 1.5437704918032788e-05,
|
| 690 |
+
"loss": 1.3468,
|
| 691 |
+
"step": 5566
|
| 692 |
+
},
|
| 693 |
+
{
|
| 694 |
+
"epoch": 11.405737704918034,
|
| 695 |
+
"eval_loss": 1.37214994430542,
|
| 696 |
+
"eval_runtime": 4.1352,
|
| 697 |
+
"eval_samples_per_second": 104.953,
|
| 698 |
+
"eval_steps_per_second": 13.3,
|
| 699 |
+
"step": 5566
|
| 700 |
+
},
|
| 701 |
+
{
|
| 702 |
+
"epoch": 11.653688524590164,
|
| 703 |
+
"grad_norm": 5.374734878540039,
|
| 704 |
+
"learning_rate": 1.5338524590163936e-05,
|
| 705 |
+
"loss": 1.3465,
|
| 706 |
+
"step": 5687
|
| 707 |
+
},
|
| 708 |
+
{
|
| 709 |
+
"epoch": 11.653688524590164,
|
| 710 |
+
"eval_loss": 1.3922464847564697,
|
| 711 |
+
"eval_runtime": 4.1462,
|
| 712 |
+
"eval_samples_per_second": 104.675,
|
| 713 |
+
"eval_steps_per_second": 13.265,
|
| 714 |
+
"step": 5687
|
| 715 |
+
},
|
| 716 |
+
{
|
| 717 |
+
"epoch": 11.901639344262295,
|
| 718 |
+
"grad_norm": 5.95257568359375,
|
| 719 |
+
"learning_rate": 1.5239344262295084e-05,
|
| 720 |
+
"loss": 1.3421,
|
| 721 |
+
"step": 5808
|
| 722 |
+
},
|
| 723 |
+
{
|
| 724 |
+
"epoch": 11.901639344262295,
|
| 725 |
+
"eval_loss": 1.4019742012023926,
|
| 726 |
+
"eval_runtime": 4.1249,
|
| 727 |
+
"eval_samples_per_second": 105.214,
|
| 728 |
+
"eval_steps_per_second": 13.334,
|
| 729 |
+
"step": 5808
|
| 730 |
+
},
|
| 731 |
+
{
|
| 732 |
+
"epoch": 12.149590163934427,
|
| 733 |
+
"grad_norm": 5.823298931121826,
|
| 734 |
+
"learning_rate": 1.514016393442623e-05,
|
| 735 |
+
"loss": 1.3324,
|
| 736 |
+
"step": 5929
|
| 737 |
+
},
|
| 738 |
+
{
|
| 739 |
+
"epoch": 12.149590163934427,
|
| 740 |
+
"eval_loss": 1.4440032243728638,
|
| 741 |
+
"eval_runtime": 4.1244,
|
| 742 |
+
"eval_samples_per_second": 105.227,
|
| 743 |
+
"eval_steps_per_second": 13.335,
|
| 744 |
+
"step": 5929
|
| 745 |
+
},
|
| 746 |
+
{
|
| 747 |
+
"epoch": 12.397540983606557,
|
| 748 |
+
"grad_norm": 5.080052375793457,
|
| 749 |
+
"learning_rate": 1.5040983606557377e-05,
|
| 750 |
+
"loss": 1.2988,
|
| 751 |
+
"step": 6050
|
| 752 |
+
},
|
| 753 |
+
{
|
| 754 |
+
"epoch": 12.397540983606557,
|
| 755 |
+
"eval_loss": 1.3977147340774536,
|
| 756 |
+
"eval_runtime": 4.1245,
|
| 757 |
+
"eval_samples_per_second": 105.225,
|
| 758 |
+
"eval_steps_per_second": 13.335,
|
| 759 |
+
"step": 6050
|
| 760 |
+
},
|
| 761 |
+
{
|
| 762 |
+
"epoch": 12.645491803278688,
|
| 763 |
+
"grad_norm": 5.39788293838501,
|
| 764 |
+
"learning_rate": 1.4941803278688525e-05,
|
| 765 |
+
"loss": 1.3222,
|
| 766 |
+
"step": 6171
|
| 767 |
+
},
|
| 768 |
+
{
|
| 769 |
+
"epoch": 12.645491803278688,
|
| 770 |
+
"eval_loss": 1.4123514890670776,
|
| 771 |
+
"eval_runtime": 4.1278,
|
| 772 |
+
"eval_samples_per_second": 105.14,
|
| 773 |
+
"eval_steps_per_second": 13.324,
|
| 774 |
+
"step": 6171
|
| 775 |
+
},
|
| 776 |
+
{
|
| 777 |
+
"epoch": 12.89344262295082,
|
| 778 |
+
"grad_norm": 5.456956386566162,
|
| 779 |
+
"learning_rate": 1.4842622950819673e-05,
|
| 780 |
+
"loss": 1.3274,
|
| 781 |
+
"step": 6292
|
| 782 |
+
},
|
| 783 |
+
{
|
| 784 |
+
"epoch": 12.89344262295082,
|
| 785 |
+
"eval_loss": 1.3916068077087402,
|
| 786 |
+
"eval_runtime": 4.1237,
|
| 787 |
+
"eval_samples_per_second": 105.245,
|
| 788 |
+
"eval_steps_per_second": 13.338,
|
| 789 |
+
"step": 6292
|
| 790 |
+
},
|
| 791 |
+
{
|
| 792 |
+
"epoch": 13.14139344262295,
|
| 793 |
+
"grad_norm": 6.0245490074157715,
|
| 794 |
+
"learning_rate": 1.4743442622950822e-05,
|
| 795 |
+
"loss": 1.3313,
|
| 796 |
+
"step": 6413
|
| 797 |
+
},
|
| 798 |
+
{
|
| 799 |
+
"epoch": 13.14139344262295,
|
| 800 |
+
"eval_loss": 1.419415831565857,
|
| 801 |
+
"eval_runtime": 4.1248,
|
| 802 |
+
"eval_samples_per_second": 105.218,
|
| 803 |
+
"eval_steps_per_second": 13.334,
|
| 804 |
+
"step": 6413
|
| 805 |
+
},
|
| 806 |
+
{
|
| 807 |
+
"epoch": 13.389344262295081,
|
| 808 |
+
"grad_norm": 5.695260524749756,
|
| 809 |
+
"learning_rate": 1.4644262295081968e-05,
|
| 810 |
+
"loss": 1.3364,
|
| 811 |
+
"step": 6534
|
| 812 |
+
},
|
| 813 |
+
{
|
| 814 |
+
"epoch": 13.389344262295081,
|
| 815 |
+
"eval_loss": 1.4173094034194946,
|
| 816 |
+
"eval_runtime": 4.1242,
|
| 817 |
+
"eval_samples_per_second": 105.233,
|
| 818 |
+
"eval_steps_per_second": 13.336,
|
| 819 |
+
"step": 6534
|
| 820 |
+
},
|
| 821 |
+
{
|
| 822 |
+
"epoch": 13.637295081967213,
|
| 823 |
+
"grad_norm": 5.108427047729492,
|
| 824 |
+
"learning_rate": 1.4545081967213115e-05,
|
| 825 |
+
"loss": 1.3057,
|
| 826 |
+
"step": 6655
|
| 827 |
+
},
|
| 828 |
+
{
|
| 829 |
+
"epoch": 13.637295081967213,
|
| 830 |
+
"eval_loss": 1.4033712148666382,
|
| 831 |
+
"eval_runtime": 4.1221,
|
| 832 |
+
"eval_samples_per_second": 105.287,
|
| 833 |
+
"eval_steps_per_second": 13.343,
|
| 834 |
+
"step": 6655
|
| 835 |
+
},
|
| 836 |
+
{
|
| 837 |
+
"epoch": 13.885245901639344,
|
| 838 |
+
"grad_norm": 7.01995849609375,
|
| 839 |
+
"learning_rate": 1.4445901639344264e-05,
|
| 840 |
+
"loss": 1.3075,
|
| 841 |
+
"step": 6776
|
| 842 |
+
},
|
| 843 |
+
{
|
| 844 |
+
"epoch": 13.885245901639344,
|
| 845 |
+
"eval_loss": 1.4299310445785522,
|
| 846 |
+
"eval_runtime": 4.1219,
|
| 847 |
+
"eval_samples_per_second": 105.291,
|
| 848 |
+
"eval_steps_per_second": 13.343,
|
| 849 |
+
"step": 6776
|
| 850 |
+
},
|
| 851 |
+
{
|
| 852 |
+
"epoch": 14.133196721311476,
|
| 853 |
+
"grad_norm": 6.163614749908447,
|
| 854 |
+
"learning_rate": 1.4346721311475411e-05,
|
| 855 |
+
"loss": 1.319,
|
| 856 |
+
"step": 6897
|
| 857 |
+
},
|
| 858 |
+
{
|
| 859 |
+
"epoch": 14.133196721311476,
|
| 860 |
+
"eval_loss": 1.4211634397506714,
|
| 861 |
+
"eval_runtime": 4.1221,
|
| 862 |
+
"eval_samples_per_second": 105.286,
|
| 863 |
+
"eval_steps_per_second": 13.343,
|
| 864 |
+
"step": 6897
|
| 865 |
+
},
|
| 866 |
+
{
|
| 867 |
+
"epoch": 14.381147540983607,
|
| 868 |
+
"grad_norm": 4.972089767456055,
|
| 869 |
+
"learning_rate": 1.4247540983606557e-05,
|
| 870 |
+
"loss": 1.2971,
|
| 871 |
+
"step": 7018
|
| 872 |
+
},
|
| 873 |
+
{
|
| 874 |
+
"epoch": 14.381147540983607,
|
| 875 |
+
"eval_loss": 1.4057807922363281,
|
| 876 |
+
"eval_runtime": 4.125,
|
| 877 |
+
"eval_samples_per_second": 105.211,
|
| 878 |
+
"eval_steps_per_second": 13.333,
|
| 879 |
+
"step": 7018
|
| 880 |
+
},
|
| 881 |
+
{
|
| 882 |
+
"epoch": 14.629098360655737,
|
| 883 |
+
"grad_norm": 5.575927734375,
|
| 884 |
+
"learning_rate": 1.4148360655737706e-05,
|
| 885 |
+
"loss": 1.2984,
|
| 886 |
+
"step": 7139
|
| 887 |
+
},
|
| 888 |
+
{
|
| 889 |
+
"epoch": 14.629098360655737,
|
| 890 |
+
"eval_loss": 1.4220715761184692,
|
| 891 |
+
"eval_runtime": 4.1218,
|
| 892 |
+
"eval_samples_per_second": 105.294,
|
| 893 |
+
"eval_steps_per_second": 13.344,
|
| 894 |
+
"step": 7139
|
| 895 |
+
},
|
| 896 |
+
{
|
| 897 |
+
"epoch": 14.87704918032787,
|
| 898 |
+
"grad_norm": 5.530676364898682,
|
| 899 |
+
"learning_rate": 1.4049180327868854e-05,
|
| 900 |
+
"loss": 1.2919,
|
| 901 |
+
"step": 7260
|
| 902 |
+
},
|
| 903 |
+
{
|
| 904 |
+
"epoch": 14.87704918032787,
|
| 905 |
+
"eval_loss": 1.410522222518921,
|
| 906 |
+
"eval_runtime": 4.122,
|
| 907 |
+
"eval_samples_per_second": 105.288,
|
| 908 |
+
"eval_steps_per_second": 13.343,
|
| 909 |
+
"step": 7260
|
| 910 |
+
},
|
| 911 |
+
{
|
| 912 |
+
"epoch": 15.125,
|
| 913 |
+
"grad_norm": 5.002058029174805,
|
| 914 |
+
"learning_rate": 1.3950000000000002e-05,
|
| 915 |
+
"loss": 1.2886,
|
| 916 |
+
"step": 7381
|
| 917 |
+
},
|
| 918 |
+
{
|
| 919 |
+
"epoch": 15.125,
|
| 920 |
+
"eval_loss": 1.3917385339736938,
|
| 921 |
+
"eval_runtime": 4.1253,
|
| 922 |
+
"eval_samples_per_second": 105.203,
|
| 923 |
+
"eval_steps_per_second": 13.332,
|
| 924 |
+
"step": 7381
|
| 925 |
+
},
|
| 926 |
+
{
|
| 927 |
+
"epoch": 15.37295081967213,
|
| 928 |
+
"grad_norm": 4.9464874267578125,
|
| 929 |
+
"learning_rate": 1.3850819672131148e-05,
|
| 930 |
+
"loss": 1.3125,
|
| 931 |
+
"step": 7502
|
| 932 |
+
},
|
| 933 |
+
{
|
| 934 |
+
"epoch": 15.37295081967213,
|
| 935 |
+
"eval_loss": 1.3845915794372559,
|
| 936 |
+
"eval_runtime": 4.1253,
|
| 937 |
+
"eval_samples_per_second": 105.204,
|
| 938 |
+
"eval_steps_per_second": 13.332,
|
| 939 |
+
"step": 7502
|
| 940 |
+
},
|
| 941 |
+
{
|
| 942 |
+
"epoch": 15.620901639344263,
|
| 943 |
+
"grad_norm": 5.266591548919678,
|
| 944 |
+
"learning_rate": 1.3751639344262295e-05,
|
| 945 |
+
"loss": 1.2923,
|
| 946 |
+
"step": 7623
|
| 947 |
+
},
|
| 948 |
+
{
|
| 949 |
+
"epoch": 15.620901639344263,
|
| 950 |
+
"eval_loss": 1.4047952890396118,
|
| 951 |
+
"eval_runtime": 4.1253,
|
| 952 |
+
"eval_samples_per_second": 105.205,
|
| 953 |
+
"eval_steps_per_second": 13.332,
|
| 954 |
+
"step": 7623
|
| 955 |
+
},
|
| 956 |
+
{
|
| 957 |
+
"epoch": 15.868852459016393,
|
| 958 |
+
"grad_norm": 5.416749000549316,
|
| 959 |
+
"learning_rate": 1.3652459016393445e-05,
|
| 960 |
+
"loss": 1.3015,
|
| 961 |
+
"step": 7744
|
| 962 |
+
},
|
| 963 |
+
{
|
| 964 |
+
"epoch": 15.868852459016393,
|
| 965 |
+
"eval_loss": 1.428416132926941,
|
| 966 |
+
"eval_runtime": 4.1255,
|
| 967 |
+
"eval_samples_per_second": 105.2,
|
| 968 |
+
"eval_steps_per_second": 13.332,
|
| 969 |
+
"step": 7744
|
| 970 |
+
},
|
| 971 |
+
{
|
| 972 |
+
"epoch": 16.116803278688526,
|
| 973 |
+
"grad_norm": 4.929660797119141,
|
| 974 |
+
"learning_rate": 1.3553278688524591e-05,
|
| 975 |
+
"loss": 1.257,
|
| 976 |
+
"step": 7865
|
| 977 |
+
},
|
| 978 |
+
{
|
| 979 |
+
"epoch": 16.116803278688526,
|
| 980 |
+
"eval_loss": 1.4093120098114014,
|
| 981 |
+
"eval_runtime": 4.1273,
|
| 982 |
+
"eval_samples_per_second": 105.154,
|
| 983 |
+
"eval_steps_per_second": 13.326,
|
| 984 |
+
"step": 7865
|
| 985 |
+
},
|
| 986 |
+
{
|
| 987 |
+
"epoch": 16.364754098360656,
|
| 988 |
+
"grad_norm": 5.573082447052002,
|
| 989 |
+
"learning_rate": 1.3454098360655738e-05,
|
| 990 |
+
"loss": 1.2802,
|
| 991 |
+
"step": 7986
|
| 992 |
+
},
|
| 993 |
+
{
|
| 994 |
+
"epoch": 16.364754098360656,
|
| 995 |
+
"eval_loss": 1.4037296772003174,
|
| 996 |
+
"eval_runtime": 4.125,
|
| 997 |
+
"eval_samples_per_second": 105.212,
|
| 998 |
+
"eval_steps_per_second": 13.333,
|
| 999 |
+
"step": 7986
|
| 1000 |
+
},
|
| 1001 |
+
{
|
| 1002 |
+
"epoch": 16.612704918032787,
|
| 1003 |
+
"grad_norm": 5.440979480743408,
|
| 1004 |
+
"learning_rate": 1.3354918032786886e-05,
|
| 1005 |
+
"loss": 1.2671,
|
| 1006 |
+
"step": 8107
|
| 1007 |
+
},
|
| 1008 |
+
{
|
| 1009 |
+
"epoch": 16.612704918032787,
|
| 1010 |
+
"eval_loss": 1.392979383468628,
|
| 1011 |
+
"eval_runtime": 4.123,
|
| 1012 |
+
"eval_samples_per_second": 105.264,
|
| 1013 |
+
"eval_steps_per_second": 13.34,
|
| 1014 |
+
"step": 8107
|
| 1015 |
+
},
|
| 1016 |
+
{
|
| 1017 |
+
"epoch": 16.860655737704917,
|
| 1018 |
+
"grad_norm": 6.133276462554932,
|
| 1019 |
+
"learning_rate": 1.3255737704918034e-05,
|
| 1020 |
+
"loss": 1.2816,
|
| 1021 |
+
"step": 8228
|
| 1022 |
+
},
|
| 1023 |
+
{
|
| 1024 |
+
"epoch": 16.860655737704917,
|
| 1025 |
+
"eval_loss": 1.3886833190917969,
|
| 1026 |
+
"eval_runtime": 4.1222,
|
| 1027 |
+
"eval_samples_per_second": 105.284,
|
| 1028 |
+
"eval_steps_per_second": 13.342,
|
| 1029 |
+
"step": 8228
|
| 1030 |
+
},
|
| 1031 |
+
{
|
| 1032 |
+
"epoch": 17.108606557377048,
|
| 1033 |
+
"grad_norm": 5.235976219177246,
|
| 1034 |
+
"learning_rate": 1.3156557377049182e-05,
|
| 1035 |
+
"loss": 1.2669,
|
| 1036 |
+
"step": 8349
|
| 1037 |
+
},
|
| 1038 |
+
{
|
| 1039 |
+
"epoch": 17.108606557377048,
|
| 1040 |
+
"eval_loss": 1.4218270778656006,
|
| 1041 |
+
"eval_runtime": 4.1262,
|
| 1042 |
+
"eval_samples_per_second": 105.181,
|
| 1043 |
+
"eval_steps_per_second": 13.329,
|
| 1044 |
+
"step": 8349
|
| 1045 |
+
},
|
| 1046 |
+
{
|
| 1047 |
+
"epoch": 17.35655737704918,
|
| 1048 |
+
"grad_norm": 5.8139519691467285,
|
| 1049 |
+
"learning_rate": 1.3057377049180329e-05,
|
| 1050 |
+
"loss": 1.2582,
|
| 1051 |
+
"step": 8470
|
| 1052 |
+
},
|
| 1053 |
+
{
|
| 1054 |
+
"epoch": 17.35655737704918,
|
| 1055 |
+
"eval_loss": 1.419556736946106,
|
| 1056 |
+
"eval_runtime": 4.1304,
|
| 1057 |
+
"eval_samples_per_second": 105.074,
|
| 1058 |
+
"eval_steps_per_second": 13.316,
|
| 1059 |
+
"step": 8470
|
| 1060 |
+
},
|
| 1061 |
+
{
|
| 1062 |
+
"epoch": 17.604508196721312,
|
| 1063 |
+
"grad_norm": 4.849159240722656,
|
| 1064 |
+
"learning_rate": 1.2958196721311475e-05,
|
| 1065 |
+
"loss": 1.2908,
|
| 1066 |
+
"step": 8591
|
| 1067 |
+
},
|
| 1068 |
+
{
|
| 1069 |
+
"epoch": 17.604508196721312,
|
| 1070 |
+
"eval_loss": 1.4165232181549072,
|
| 1071 |
+
"eval_runtime": 4.1257,
|
| 1072 |
+
"eval_samples_per_second": 105.193,
|
| 1073 |
+
"eval_steps_per_second": 13.331,
|
| 1074 |
+
"step": 8591
|
| 1075 |
+
},
|
| 1076 |
+
{
|
| 1077 |
+
"epoch": 17.852459016393443,
|
| 1078 |
+
"grad_norm": 5.0550360679626465,
|
| 1079 |
+
"learning_rate": 1.2859016393442625e-05,
|
| 1080 |
+
"loss": 1.2312,
|
| 1081 |
+
"step": 8712
|
| 1082 |
+
},
|
| 1083 |
+
{
|
| 1084 |
+
"epoch": 17.852459016393443,
|
| 1085 |
+
"eval_loss": 1.4022456407546997,
|
| 1086 |
+
"eval_runtime": 4.1255,
|
| 1087 |
+
"eval_samples_per_second": 105.198,
|
| 1088 |
+
"eval_steps_per_second": 13.332,
|
| 1089 |
+
"step": 8712
|
| 1090 |
+
},
|
| 1091 |
+
{
|
| 1092 |
+
"epoch": 18.100409836065573,
|
| 1093 |
+
"grad_norm": 5.549849987030029,
|
| 1094 |
+
"learning_rate": 1.2759836065573772e-05,
|
| 1095 |
+
"loss": 1.2664,
|
| 1096 |
+
"step": 8833
|
| 1097 |
+
},
|
| 1098 |
+
{
|
| 1099 |
+
"epoch": 18.100409836065573,
|
| 1100 |
+
"eval_loss": 1.3795406818389893,
|
| 1101 |
+
"eval_runtime": 4.1227,
|
| 1102 |
+
"eval_samples_per_second": 105.27,
|
| 1103 |
+
"eval_steps_per_second": 13.341,
|
| 1104 |
+
"step": 8833
|
| 1105 |
+
},
|
| 1106 |
+
{
|
| 1107 |
+
"epoch": 18.348360655737704,
|
| 1108 |
+
"grad_norm": 5.974192142486572,
|
| 1109 |
+
"learning_rate": 1.2660655737704918e-05,
|
| 1110 |
+
"loss": 1.25,
|
| 1111 |
+
"step": 8954
|
| 1112 |
+
},
|
| 1113 |
+
{
|
| 1114 |
+
"epoch": 18.348360655737704,
|
| 1115 |
+
"eval_loss": 1.4104958772659302,
|
| 1116 |
+
"eval_runtime": 4.1253,
|
| 1117 |
+
"eval_samples_per_second": 105.204,
|
| 1118 |
+
"eval_steps_per_second": 13.332,
|
| 1119 |
+
"step": 8954
|
| 1120 |
+
},
|
| 1121 |
+
{
|
| 1122 |
+
"epoch": 18.596311475409838,
|
| 1123 |
+
"grad_norm": 5.209727764129639,
|
| 1124 |
+
"learning_rate": 1.2561475409836066e-05,
|
| 1125 |
+
"loss": 1.259,
|
| 1126 |
+
"step": 9075
|
| 1127 |
+
},
|
| 1128 |
+
{
|
| 1129 |
+
"epoch": 18.596311475409838,
|
| 1130 |
+
"eval_loss": 1.3699523210525513,
|
| 1131 |
+
"eval_runtime": 4.1271,
|
| 1132 |
+
"eval_samples_per_second": 105.159,
|
| 1133 |
+
"eval_steps_per_second": 13.327,
|
| 1134 |
+
"step": 9075
|
| 1135 |
+
},
|
| 1136 |
+
{
|
| 1137 |
+
"epoch": 18.84426229508197,
|
| 1138 |
+
"grad_norm": 5.296604156494141,
|
| 1139 |
+
"learning_rate": 1.2462295081967215e-05,
|
| 1140 |
+
"loss": 1.2593,
|
| 1141 |
+
"step": 9196
|
| 1142 |
+
},
|
| 1143 |
+
{
|
| 1144 |
+
"epoch": 18.84426229508197,
|
| 1145 |
+
"eval_loss": 1.3999890089035034,
|
| 1146 |
+
"eval_runtime": 4.1243,
|
| 1147 |
+
"eval_samples_per_second": 105.231,
|
| 1148 |
+
"eval_steps_per_second": 13.336,
|
| 1149 |
+
"step": 9196
|
| 1150 |
+
},
|
| 1151 |
+
{
|
| 1152 |
+
"epoch": 19.0922131147541,
|
| 1153 |
+
"grad_norm": 5.599063396453857,
|
| 1154 |
+
"learning_rate": 1.2363114754098363e-05,
|
| 1155 |
+
"loss": 1.2479,
|
| 1156 |
+
"step": 9317
|
| 1157 |
+
},
|
| 1158 |
+
{
|
| 1159 |
+
"epoch": 19.0922131147541,
|
| 1160 |
+
"eval_loss": 1.4049623012542725,
|
| 1161 |
+
"eval_runtime": 4.1272,
|
| 1162 |
+
"eval_samples_per_second": 105.155,
|
| 1163 |
+
"eval_steps_per_second": 13.326,
|
| 1164 |
+
"step": 9317
|
| 1165 |
+
},
|
| 1166 |
+
{
|
| 1167 |
+
"epoch": 19.34016393442623,
|
| 1168 |
+
"grad_norm": 5.4184346199035645,
|
| 1169 |
+
"learning_rate": 1.226393442622951e-05,
|
| 1170 |
+
"loss": 1.2436,
|
| 1171 |
+
"step": 9438
|
| 1172 |
+
},
|
| 1173 |
+
{
|
| 1174 |
+
"epoch": 19.34016393442623,
|
| 1175 |
+
"eval_loss": 1.4119490385055542,
|
| 1176 |
+
"eval_runtime": 4.1249,
|
| 1177 |
+
"eval_samples_per_second": 105.213,
|
| 1178 |
+
"eval_steps_per_second": 13.334,
|
| 1179 |
+
"step": 9438
|
| 1180 |
+
},
|
| 1181 |
+
{
|
| 1182 |
+
"epoch": 19.58811475409836,
|
| 1183 |
+
"grad_norm": 4.9451189041137695,
|
| 1184 |
+
"learning_rate": 1.2164754098360656e-05,
|
| 1185 |
+
"loss": 1.2554,
|
| 1186 |
+
"step": 9559
|
| 1187 |
+
},
|
| 1188 |
+
{
|
| 1189 |
+
"epoch": 19.58811475409836,
|
| 1190 |
+
"eval_loss": 1.3934992551803589,
|
| 1191 |
+
"eval_runtime": 4.1257,
|
| 1192 |
+
"eval_samples_per_second": 105.195,
|
| 1193 |
+
"eval_steps_per_second": 13.331,
|
| 1194 |
+
"step": 9559
|
| 1195 |
+
},
|
| 1196 |
+
{
|
| 1197 |
+
"epoch": 19.83606557377049,
|
| 1198 |
+
"grad_norm": 5.6321845054626465,
|
| 1199 |
+
"learning_rate": 1.2065573770491806e-05,
|
| 1200 |
+
"loss": 1.2534,
|
| 1201 |
+
"step": 9680
|
| 1202 |
+
},
|
| 1203 |
+
{
|
| 1204 |
+
"epoch": 19.83606557377049,
|
| 1205 |
+
"eval_loss": 1.3982270956039429,
|
| 1206 |
+
"eval_runtime": 4.1253,
|
| 1207 |
+
"eval_samples_per_second": 105.206,
|
| 1208 |
+
"eval_steps_per_second": 13.333,
|
| 1209 |
+
"step": 9680
|
| 1210 |
+
},
|
| 1211 |
+
{
|
| 1212 |
+
"epoch": 20.084016393442624,
|
| 1213 |
+
"grad_norm": 5.988424301147461,
|
| 1214 |
+
"learning_rate": 1.1966393442622952e-05,
|
| 1215 |
+
"loss": 1.2402,
|
| 1216 |
+
"step": 9801
|
| 1217 |
+
},
|
| 1218 |
+
{
|
| 1219 |
+
"epoch": 20.084016393442624,
|
| 1220 |
+
"eval_loss": 1.4115673303604126,
|
| 1221 |
+
"eval_runtime": 4.1245,
|
| 1222 |
+
"eval_samples_per_second": 105.225,
|
| 1223 |
+
"eval_steps_per_second": 13.335,
|
| 1224 |
+
"step": 9801
|
| 1225 |
+
},
|
| 1226 |
+
{
|
| 1227 |
+
"epoch": 20.331967213114755,
|
| 1228 |
+
"grad_norm": 5.451227188110352,
|
| 1229 |
+
"learning_rate": 1.1867213114754099e-05,
|
| 1230 |
+
"loss": 1.2298,
|
| 1231 |
+
"step": 9922
|
| 1232 |
+
},
|
| 1233 |
+
{
|
| 1234 |
+
"epoch": 20.331967213114755,
|
| 1235 |
+
"eval_loss": 1.381482720375061,
|
| 1236 |
+
"eval_runtime": 4.1454,
|
| 1237 |
+
"eval_samples_per_second": 104.695,
|
| 1238 |
+
"eval_steps_per_second": 13.268,
|
| 1239 |
+
"step": 9922
|
| 1240 |
+
},
|
| 1241 |
+
{
|
| 1242 |
+
"epoch": 20.579918032786885,
|
| 1243 |
+
"grad_norm": 5.616585731506348,
|
| 1244 |
+
"learning_rate": 1.1768032786885247e-05,
|
| 1245 |
+
"loss": 1.2807,
|
| 1246 |
+
"step": 10043
|
| 1247 |
+
},
|
| 1248 |
+
{
|
| 1249 |
+
"epoch": 20.579918032786885,
|
| 1250 |
+
"eval_loss": 1.380122423171997,
|
| 1251 |
+
"eval_runtime": 4.1243,
|
| 1252 |
+
"eval_samples_per_second": 105.229,
|
| 1253 |
+
"eval_steps_per_second": 13.336,
|
| 1254 |
+
"step": 10043
|
| 1255 |
+
},
|
| 1256 |
+
{
|
| 1257 |
+
"epoch": 20.827868852459016,
|
| 1258 |
+
"grad_norm": 4.926662445068359,
|
| 1259 |
+
"learning_rate": 1.1668852459016395e-05,
|
| 1260 |
+
"loss": 1.2016,
|
| 1261 |
+
"step": 10164
|
| 1262 |
+
},
|
| 1263 |
+
{
|
| 1264 |
+
"epoch": 20.827868852459016,
|
| 1265 |
+
"eval_loss": 1.365037202835083,
|
| 1266 |
+
"eval_runtime": 4.1243,
|
| 1267 |
+
"eval_samples_per_second": 105.23,
|
| 1268 |
+
"eval_steps_per_second": 13.336,
|
| 1269 |
+
"step": 10164
|
| 1270 |
+
},
|
| 1271 |
+
{
|
| 1272 |
+
"epoch": 21.075819672131146,
|
| 1273 |
+
"grad_norm": 5.1953959465026855,
|
| 1274 |
+
"learning_rate": 1.1569672131147543e-05,
|
| 1275 |
+
"loss": 1.2534,
|
| 1276 |
+
"step": 10285
|
| 1277 |
+
},
|
| 1278 |
+
{
|
| 1279 |
+
"epoch": 21.075819672131146,
|
| 1280 |
+
"eval_loss": 1.419316291809082,
|
| 1281 |
+
"eval_runtime": 4.1263,
|
| 1282 |
+
"eval_samples_per_second": 105.18,
|
| 1283 |
+
"eval_steps_per_second": 13.329,
|
| 1284 |
+
"step": 10285
|
| 1285 |
+
},
|
| 1286 |
+
{
|
| 1287 |
+
"epoch": 21.32377049180328,
|
| 1288 |
+
"grad_norm": 5.846003532409668,
|
| 1289 |
+
"learning_rate": 1.147049180327869e-05,
|
| 1290 |
+
"loss": 1.2345,
|
| 1291 |
+
"step": 10406
|
| 1292 |
+
},
|
| 1293 |
+
{
|
| 1294 |
+
"epoch": 21.32377049180328,
|
| 1295 |
+
"eval_loss": 1.4061498641967773,
|
| 1296 |
+
"eval_runtime": 4.1277,
|
| 1297 |
+
"eval_samples_per_second": 105.143,
|
| 1298 |
+
"eval_steps_per_second": 13.325,
|
| 1299 |
+
"step": 10406
|
| 1300 |
+
},
|
| 1301 |
+
{
|
| 1302 |
+
"epoch": 21.57172131147541,
|
| 1303 |
+
"grad_norm": 5.565953731536865,
|
| 1304 |
+
"learning_rate": 1.1371311475409836e-05,
|
| 1305 |
+
"loss": 1.229,
|
| 1306 |
+
"step": 10527
|
| 1307 |
+
},
|
| 1308 |
+
{
|
| 1309 |
+
"epoch": 21.57172131147541,
|
| 1310 |
+
"eval_loss": 1.3947559595108032,
|
| 1311 |
+
"eval_runtime": 4.1245,
|
| 1312 |
+
"eval_samples_per_second": 105.226,
|
| 1313 |
+
"eval_steps_per_second": 13.335,
|
| 1314 |
+
"step": 10527
|
| 1315 |
+
},
|
| 1316 |
+
{
|
| 1317 |
+
"epoch": 21.81967213114754,
|
| 1318 |
+
"grad_norm": 4.984577655792236,
|
| 1319 |
+
"learning_rate": 1.1272131147540986e-05,
|
| 1320 |
+
"loss": 1.2148,
|
| 1321 |
+
"step": 10648
|
| 1322 |
+
},
|
| 1323 |
+
{
|
| 1324 |
+
"epoch": 21.81967213114754,
|
| 1325 |
+
"eval_loss": 1.398898720741272,
|
| 1326 |
+
"eval_runtime": 4.1264,
|
| 1327 |
+
"eval_samples_per_second": 105.176,
|
| 1328 |
+
"eval_steps_per_second": 13.329,
|
| 1329 |
+
"step": 10648
|
| 1330 |
+
},
|
| 1331 |
+
{
|
| 1332 |
+
"epoch": 22.067622950819672,
|
| 1333 |
+
"grad_norm": 6.103436470031738,
|
| 1334 |
+
"learning_rate": 1.1172950819672133e-05,
|
| 1335 |
+
"loss": 1.2339,
|
| 1336 |
+
"step": 10769
|
| 1337 |
+
},
|
| 1338 |
+
{
|
| 1339 |
+
"epoch": 22.067622950819672,
|
| 1340 |
+
"eval_loss": 1.4189534187316895,
|
| 1341 |
+
"eval_runtime": 4.1257,
|
| 1342 |
+
"eval_samples_per_second": 105.193,
|
| 1343 |
+
"eval_steps_per_second": 13.331,
|
| 1344 |
+
"step": 10769
|
| 1345 |
+
},
|
| 1346 |
+
{
|
| 1347 |
+
"epoch": 22.315573770491802,
|
| 1348 |
+
"grad_norm": 5.6105732917785645,
|
| 1349 |
+
"learning_rate": 1.1073770491803279e-05,
|
| 1350 |
+
"loss": 1.2476,
|
| 1351 |
+
"step": 10890
|
| 1352 |
+
},
|
| 1353 |
+
{
|
| 1354 |
+
"epoch": 22.315573770491802,
|
| 1355 |
+
"eval_loss": 1.4122945070266724,
|
| 1356 |
+
"eval_runtime": 4.1262,
|
| 1357 |
+
"eval_samples_per_second": 105.181,
|
| 1358 |
+
"eval_steps_per_second": 13.329,
|
| 1359 |
+
"step": 10890
|
| 1360 |
+
},
|
| 1361 |
+
{
|
| 1362 |
+
"epoch": 22.563524590163933,
|
| 1363 |
+
"grad_norm": 6.014953136444092,
|
| 1364 |
+
"learning_rate": 1.0974590163934427e-05,
|
| 1365 |
+
"loss": 1.2173,
|
| 1366 |
+
"step": 11011
|
| 1367 |
+
},
|
| 1368 |
+
{
|
| 1369 |
+
"epoch": 22.563524590163933,
|
| 1370 |
+
"eval_loss": 1.417721152305603,
|
| 1371 |
+
"eval_runtime": 4.1262,
|
| 1372 |
+
"eval_samples_per_second": 105.182,
|
| 1373 |
+
"eval_steps_per_second": 13.33,
|
| 1374 |
+
"step": 11011
|
| 1375 |
+
},
|
| 1376 |
+
{
|
| 1377 |
+
"epoch": 22.811475409836067,
|
| 1378 |
+
"grad_norm": 5.092090129852295,
|
| 1379 |
+
"learning_rate": 1.0875409836065575e-05,
|
| 1380 |
+
"loss": 1.2119,
|
| 1381 |
+
"step": 11132
|
| 1382 |
+
},
|
| 1383 |
+
{
|
| 1384 |
+
"epoch": 22.811475409836067,
|
| 1385 |
+
"eval_loss": 1.4074928760528564,
|
| 1386 |
+
"eval_runtime": 4.1242,
|
| 1387 |
+
"eval_samples_per_second": 105.233,
|
| 1388 |
+
"eval_steps_per_second": 13.336,
|
| 1389 |
+
"step": 11132
|
| 1390 |
+
},
|
| 1391 |
+
{
|
| 1392 |
+
"epoch": 23.059426229508198,
|
| 1393 |
+
"grad_norm": 5.808371067047119,
|
| 1394 |
+
"learning_rate": 1.0776229508196724e-05,
|
| 1395 |
+
"loss": 1.2049,
|
| 1396 |
+
"step": 11253
|
| 1397 |
+
},
|
| 1398 |
+
{
|
| 1399 |
+
"epoch": 23.059426229508198,
|
| 1400 |
+
"eval_loss": 1.4072319269180298,
|
| 1401 |
+
"eval_runtime": 4.131,
|
| 1402 |
+
"eval_samples_per_second": 105.06,
|
| 1403 |
+
"eval_steps_per_second": 13.314,
|
| 1404 |
+
"step": 11253
|
| 1405 |
+
},
|
| 1406 |
+
{
|
| 1407 |
+
"epoch": 23.307377049180328,
|
| 1408 |
+
"grad_norm": 5.78179407119751,
|
| 1409 |
+
"learning_rate": 1.067704918032787e-05,
|
| 1410 |
+
"loss": 1.1968,
|
| 1411 |
+
"step": 11374
|
| 1412 |
+
},
|
| 1413 |
+
{
|
| 1414 |
+
"epoch": 23.307377049180328,
|
| 1415 |
+
"eval_loss": 1.4036682844161987,
|
| 1416 |
+
"eval_runtime": 4.1268,
|
| 1417 |
+
"eval_samples_per_second": 105.165,
|
| 1418 |
+
"eval_steps_per_second": 13.327,
|
| 1419 |
+
"step": 11374
|
| 1420 |
+
},
|
| 1421 |
+
{
|
| 1422 |
+
"epoch": 23.55532786885246,
|
| 1423 |
+
"grad_norm": 5.584239482879639,
|
| 1424 |
+
"learning_rate": 1.0577868852459017e-05,
|
| 1425 |
+
"loss": 1.1862,
|
| 1426 |
+
"step": 11495
|
| 1427 |
+
},
|
| 1428 |
+
{
|
| 1429 |
+
"epoch": 23.55532786885246,
|
| 1430 |
+
"eval_loss": 1.4230456352233887,
|
| 1431 |
+
"eval_runtime": 4.1296,
|
| 1432 |
+
"eval_samples_per_second": 105.095,
|
| 1433 |
+
"eval_steps_per_second": 13.318,
|
| 1434 |
+
"step": 11495
|
| 1435 |
+
},
|
| 1436 |
+
{
|
| 1437 |
+
"epoch": 23.80327868852459,
|
| 1438 |
+
"grad_norm": 6.3946685791015625,
|
| 1439 |
+
"learning_rate": 1.0478688524590163e-05,
|
| 1440 |
+
"loss": 1.2407,
|
| 1441 |
+
"step": 11616
|
| 1442 |
+
},
|
| 1443 |
+
{
|
| 1444 |
+
"epoch": 23.80327868852459,
|
| 1445 |
+
"eval_loss": 1.3906227350234985,
|
| 1446 |
+
"eval_runtime": 4.1272,
|
| 1447 |
+
"eval_samples_per_second": 105.157,
|
| 1448 |
+
"eval_steps_per_second": 13.326,
|
| 1449 |
+
"step": 11616
|
| 1450 |
+
},
|
| 1451 |
+
{
|
| 1452 |
+
"epoch": 24.05122950819672,
|
| 1453 |
+
"grad_norm": 5.362001419067383,
|
| 1454 |
+
"learning_rate": 1.0379508196721313e-05,
|
| 1455 |
+
"loss": 1.2213,
|
| 1456 |
+
"step": 11737
|
| 1457 |
+
},
|
| 1458 |
+
{
|
| 1459 |
+
"epoch": 24.05122950819672,
|
| 1460 |
+
"eval_loss": 1.406900405883789,
|
| 1461 |
+
"eval_runtime": 4.1286,
|
| 1462 |
+
"eval_samples_per_second": 105.121,
|
| 1463 |
+
"eval_steps_per_second": 13.322,
|
| 1464 |
+
"step": 11737
|
| 1465 |
+
},
|
| 1466 |
+
{
|
| 1467 |
+
"epoch": 24.299180327868854,
|
| 1468 |
+
"grad_norm": 5.9641265869140625,
|
| 1469 |
+
"learning_rate": 1.028032786885246e-05,
|
| 1470 |
+
"loss": 1.2041,
|
| 1471 |
+
"step": 11858
|
| 1472 |
+
},
|
| 1473 |
+
{
|
| 1474 |
+
"epoch": 24.299180327868854,
|
| 1475 |
+
"eval_loss": 1.3982616662979126,
|
| 1476 |
+
"eval_runtime": 4.1239,
|
| 1477 |
+
"eval_samples_per_second": 105.24,
|
| 1478 |
+
"eval_steps_per_second": 13.337,
|
| 1479 |
+
"step": 11858
|
| 1480 |
+
},
|
| 1481 |
+
{
|
| 1482 |
+
"epoch": 24.547131147540984,
|
| 1483 |
+
"grad_norm": 5.0064568519592285,
|
| 1484 |
+
"learning_rate": 1.0181147540983608e-05,
|
| 1485 |
+
"loss": 1.1758,
|
| 1486 |
+
"step": 11979
|
| 1487 |
+
},
|
| 1488 |
+
{
|
| 1489 |
+
"epoch": 24.547131147540984,
|
| 1490 |
+
"eval_loss": 1.4083583354949951,
|
| 1491 |
+
"eval_runtime": 4.122,
|
| 1492 |
+
"eval_samples_per_second": 105.289,
|
| 1493 |
+
"eval_steps_per_second": 13.343,
|
| 1494 |
+
"step": 11979
|
| 1495 |
+
},
|
| 1496 |
+
{
|
| 1497 |
+
"epoch": 24.795081967213115,
|
| 1498 |
+
"grad_norm": 5.097125053405762,
|
| 1499 |
+
"learning_rate": 1.0081967213114754e-05,
|
| 1500 |
+
"loss": 1.2335,
|
| 1501 |
+
"step": 12100
|
| 1502 |
+
},
|
| 1503 |
+
{
|
| 1504 |
+
"epoch": 24.795081967213115,
|
| 1505 |
+
"eval_loss": 1.3814207315444946,
|
| 1506 |
+
"eval_runtime": 4.1238,
|
| 1507 |
+
"eval_samples_per_second": 105.243,
|
| 1508 |
+
"eval_steps_per_second": 13.337,
|
| 1509 |
+
"step": 12100
|
| 1510 |
+
},
|
| 1511 |
+
{
|
| 1512 |
+
"epoch": 25.043032786885245,
|
| 1513 |
+
"grad_norm": 5.703944206237793,
|
| 1514 |
+
"learning_rate": 9.982786885245902e-06,
|
| 1515 |
+
"loss": 1.2038,
|
| 1516 |
+
"step": 12221
|
| 1517 |
+
},
|
| 1518 |
+
{
|
| 1519 |
+
"epoch": 25.043032786885245,
|
| 1520 |
+
"eval_loss": 1.3744884729385376,
|
| 1521 |
+
"eval_runtime": 4.1288,
|
| 1522 |
+
"eval_samples_per_second": 105.115,
|
| 1523 |
+
"eval_steps_per_second": 13.321,
|
| 1524 |
+
"step": 12221
|
| 1525 |
+
},
|
| 1526 |
+
{
|
| 1527 |
+
"epoch": 25.290983606557376,
|
| 1528 |
+
"grad_norm": 6.049025535583496,
|
| 1529 |
+
"learning_rate": 9.88360655737705e-06,
|
| 1530 |
+
"loss": 1.1871,
|
| 1531 |
+
"step": 12342
|
| 1532 |
+
},
|
| 1533 |
+
{
|
| 1534 |
+
"epoch": 25.290983606557376,
|
| 1535 |
+
"eval_loss": 1.3687578439712524,
|
| 1536 |
+
"eval_runtime": 4.1247,
|
| 1537 |
+
"eval_samples_per_second": 105.22,
|
| 1538 |
+
"eval_steps_per_second": 13.334,
|
| 1539 |
+
"step": 12342
|
| 1540 |
+
},
|
| 1541 |
+
{
|
| 1542 |
+
"epoch": 25.53893442622951,
|
| 1543 |
+
"grad_norm": 5.135526180267334,
|
| 1544 |
+
"learning_rate": 9.784426229508197e-06,
|
| 1545 |
+
"loss": 1.2076,
|
| 1546 |
+
"step": 12463
|
| 1547 |
+
},
|
| 1548 |
+
{
|
| 1549 |
+
"epoch": 25.53893442622951,
|
| 1550 |
+
"eval_loss": 1.368115782737732,
|
| 1551 |
+
"eval_runtime": 4.1255,
|
| 1552 |
+
"eval_samples_per_second": 105.2,
|
| 1553 |
+
"eval_steps_per_second": 13.332,
|
| 1554 |
+
"step": 12463
|
| 1555 |
+
},
|
| 1556 |
+
{
|
| 1557 |
+
"epoch": 25.78688524590164,
|
| 1558 |
+
"grad_norm": 5.23483419418335,
|
| 1559 |
+
"learning_rate": 9.685245901639345e-06,
|
| 1560 |
+
"loss": 1.1994,
|
| 1561 |
+
"step": 12584
|
| 1562 |
+
},
|
| 1563 |
+
{
|
| 1564 |
+
"epoch": 25.78688524590164,
|
| 1565 |
+
"eval_loss": 1.397793173789978,
|
| 1566 |
+
"eval_runtime": 4.1272,
|
| 1567 |
+
"eval_samples_per_second": 105.155,
|
| 1568 |
+
"eval_steps_per_second": 13.326,
|
| 1569 |
+
"step": 12584
|
| 1570 |
+
},
|
| 1571 |
+
{
|
| 1572 |
+
"epoch": 26.03483606557377,
|
| 1573 |
+
"grad_norm": 5.80164098739624,
|
| 1574 |
+
"learning_rate": 9.586065573770492e-06,
|
| 1575 |
+
"loss": 1.2154,
|
| 1576 |
+
"step": 12705
|
| 1577 |
+
},
|
| 1578 |
+
{
|
| 1579 |
+
"epoch": 26.03483606557377,
|
| 1580 |
+
"eval_loss": 1.3746296167373657,
|
| 1581 |
+
"eval_runtime": 4.125,
|
| 1582 |
+
"eval_samples_per_second": 105.211,
|
| 1583 |
+
"eval_steps_per_second": 13.333,
|
| 1584 |
+
"step": 12705
|
| 1585 |
+
},
|
| 1586 |
+
{
|
| 1587 |
+
"epoch": 26.2827868852459,
|
| 1588 |
+
"grad_norm": 5.612881183624268,
|
| 1589 |
+
"learning_rate": 9.48688524590164e-06,
|
| 1590 |
+
"loss": 1.1805,
|
| 1591 |
+
"step": 12826
|
| 1592 |
+
},
|
| 1593 |
+
{
|
| 1594 |
+
"epoch": 26.2827868852459,
|
| 1595 |
+
"eval_loss": 1.4100968837738037,
|
| 1596 |
+
"eval_runtime": 4.1236,
|
| 1597 |
+
"eval_samples_per_second": 105.248,
|
| 1598 |
+
"eval_steps_per_second": 13.338,
|
| 1599 |
+
"step": 12826
|
| 1600 |
+
},
|
| 1601 |
+
{
|
| 1602 |
+
"epoch": 26.53073770491803,
|
| 1603 |
+
"grad_norm": 5.775158882141113,
|
| 1604 |
+
"learning_rate": 9.387704918032788e-06,
|
| 1605 |
+
"loss": 1.1533,
|
| 1606 |
+
"step": 12947
|
| 1607 |
+
},
|
| 1608 |
+
{
|
| 1609 |
+
"epoch": 26.53073770491803,
|
| 1610 |
+
"eval_loss": 1.387868046760559,
|
| 1611 |
+
"eval_runtime": 4.1263,
|
| 1612 |
+
"eval_samples_per_second": 105.178,
|
| 1613 |
+
"eval_steps_per_second": 13.329,
|
| 1614 |
+
"step": 12947
|
| 1615 |
+
},
|
| 1616 |
+
{
|
| 1617 |
+
"epoch": 26.778688524590162,
|
| 1618 |
+
"grad_norm": 5.682550430297852,
|
| 1619 |
+
"learning_rate": 9.288524590163936e-06,
|
| 1620 |
+
"loss": 1.1848,
|
| 1621 |
+
"step": 13068
|
| 1622 |
+
},
|
| 1623 |
+
{
|
| 1624 |
+
"epoch": 26.778688524590162,
|
| 1625 |
+
"eval_loss": 1.4003947973251343,
|
| 1626 |
+
"eval_runtime": 4.1236,
|
| 1627 |
+
"eval_samples_per_second": 105.247,
|
| 1628 |
+
"eval_steps_per_second": 13.338,
|
| 1629 |
+
"step": 13068
|
| 1630 |
+
},
|
| 1631 |
+
{
|
| 1632 |
+
"epoch": 27.026639344262296,
|
| 1633 |
+
"grad_norm": 5.132202625274658,
|
| 1634 |
+
"learning_rate": 9.189344262295083e-06,
|
| 1635 |
+
"loss": 1.2298,
|
| 1636 |
+
"step": 13189
|
| 1637 |
+
},
|
| 1638 |
+
{
|
| 1639 |
+
"epoch": 27.026639344262296,
|
| 1640 |
+
"eval_loss": 1.3953262567520142,
|
| 1641 |
+
"eval_runtime": 4.127,
|
| 1642 |
+
"eval_samples_per_second": 105.162,
|
| 1643 |
+
"eval_steps_per_second": 13.327,
|
| 1644 |
+
"step": 13189
|
| 1645 |
+
},
|
| 1646 |
+
{
|
| 1647 |
+
"epoch": 27.274590163934427,
|
| 1648 |
+
"grad_norm": 4.892816066741943,
|
| 1649 |
+
"learning_rate": 9.09016393442623e-06,
|
| 1650 |
+
"loss": 1.1734,
|
| 1651 |
+
"step": 13310
|
| 1652 |
+
},
|
| 1653 |
+
{
|
| 1654 |
+
"epoch": 27.274590163934427,
|
| 1655 |
+
"eval_loss": 1.3932636976242065,
|
| 1656 |
+
"eval_runtime": 4.1237,
|
| 1657 |
+
"eval_samples_per_second": 105.245,
|
| 1658 |
+
"eval_steps_per_second": 13.338,
|
| 1659 |
+
"step": 13310
|
| 1660 |
+
},
|
| 1661 |
+
{
|
| 1662 |
+
"epoch": 27.522540983606557,
|
| 1663 |
+
"grad_norm": 5.972357273101807,
|
| 1664 |
+
"learning_rate": 8.990983606557377e-06,
|
| 1665 |
+
"loss": 1.1857,
|
| 1666 |
+
"step": 13431
|
| 1667 |
+
},
|
| 1668 |
+
{
|
| 1669 |
+
"epoch": 27.522540983606557,
|
| 1670 |
+
"eval_loss": 1.373570203781128,
|
| 1671 |
+
"eval_runtime": 4.1234,
|
| 1672 |
+
"eval_samples_per_second": 105.254,
|
| 1673 |
+
"eval_steps_per_second": 13.339,
|
| 1674 |
+
"step": 13431
|
| 1675 |
+
},
|
| 1676 |
+
{
|
| 1677 |
+
"epoch": 27.770491803278688,
|
| 1678 |
+
"grad_norm": 5.362318992614746,
|
| 1679 |
+
"learning_rate": 8.891803278688526e-06,
|
| 1680 |
+
"loss": 1.1913,
|
| 1681 |
+
"step": 13552
|
| 1682 |
+
},
|
| 1683 |
+
{
|
| 1684 |
+
"epoch": 27.770491803278688,
|
| 1685 |
+
"eval_loss": 1.3928087949752808,
|
| 1686 |
+
"eval_runtime": 4.1266,
|
| 1687 |
+
"eval_samples_per_second": 105.171,
|
| 1688 |
+
"eval_steps_per_second": 13.328,
|
| 1689 |
+
"step": 13552
|
| 1690 |
+
},
|
| 1691 |
+
{
|
| 1692 |
+
"epoch": 28.01844262295082,
|
| 1693 |
+
"grad_norm": 6.388829708099365,
|
| 1694 |
+
"learning_rate": 8.792622950819672e-06,
|
| 1695 |
+
"loss": 1.1802,
|
| 1696 |
+
"step": 13673
|
| 1697 |
+
},
|
| 1698 |
+
{
|
| 1699 |
+
"epoch": 28.01844262295082,
|
| 1700 |
+
"eval_loss": 1.4076062440872192,
|
| 1701 |
+
"eval_runtime": 4.1256,
|
| 1702 |
+
"eval_samples_per_second": 105.196,
|
| 1703 |
+
"eval_steps_per_second": 13.331,
|
| 1704 |
+
"step": 13673
|
| 1705 |
+
},
|
| 1706 |
+
{
|
| 1707 |
+
"epoch": 28.266393442622952,
|
| 1708 |
+
"grad_norm": 5.451745510101318,
|
| 1709 |
+
"learning_rate": 8.69344262295082e-06,
|
| 1710 |
+
"loss": 1.1592,
|
| 1711 |
+
"step": 13794
|
| 1712 |
+
},
|
| 1713 |
+
{
|
| 1714 |
+
"epoch": 28.266393442622952,
|
| 1715 |
+
"eval_loss": 1.4036214351654053,
|
| 1716 |
+
"eval_runtime": 4.1264,
|
| 1717 |
+
"eval_samples_per_second": 105.177,
|
| 1718 |
+
"eval_steps_per_second": 13.329,
|
| 1719 |
+
"step": 13794
|
| 1720 |
+
},
|
| 1721 |
+
{
|
| 1722 |
+
"epoch": 28.514344262295083,
|
| 1723 |
+
"grad_norm": 5.723804473876953,
|
| 1724 |
+
"learning_rate": 8.594262295081968e-06,
|
| 1725 |
+
"loss": 1.159,
|
| 1726 |
+
"step": 13915
|
| 1727 |
+
},
|
| 1728 |
+
{
|
| 1729 |
+
"epoch": 28.514344262295083,
|
| 1730 |
+
"eval_loss": 1.4207592010498047,
|
| 1731 |
+
"eval_runtime": 4.1242,
|
| 1732 |
+
"eval_samples_per_second": 105.233,
|
| 1733 |
+
"eval_steps_per_second": 13.336,
|
| 1734 |
+
"step": 13915
|
| 1735 |
+
},
|
| 1736 |
+
{
|
| 1737 |
+
"epoch": 28.762295081967213,
|
| 1738 |
+
"grad_norm": 5.422088146209717,
|
| 1739 |
+
"learning_rate": 8.495081967213117e-06,
|
| 1740 |
+
"loss": 1.1692,
|
| 1741 |
+
"step": 14036
|
| 1742 |
+
},
|
| 1743 |
+
{
|
| 1744 |
+
"epoch": 28.762295081967213,
|
| 1745 |
+
"eval_loss": 1.4013181924819946,
|
| 1746 |
+
"eval_runtime": 4.1286,
|
| 1747 |
+
"eval_samples_per_second": 105.119,
|
| 1748 |
+
"eval_steps_per_second": 13.322,
|
| 1749 |
+
"step": 14036
|
| 1750 |
+
},
|
| 1751 |
+
{
|
| 1752 |
+
"epoch": 29.010245901639344,
|
| 1753 |
+
"grad_norm": 6.352570056915283,
|
| 1754 |
+
"learning_rate": 8.395901639344263e-06,
|
| 1755 |
+
"loss": 1.1904,
|
| 1756 |
+
"step": 14157
|
| 1757 |
+
},
|
| 1758 |
+
{
|
| 1759 |
+
"epoch": 29.010245901639344,
|
| 1760 |
+
"eval_loss": 1.404019832611084,
|
| 1761 |
+
"eval_runtime": 4.1262,
|
| 1762 |
+
"eval_samples_per_second": 105.181,
|
| 1763 |
+
"eval_steps_per_second": 13.329,
|
| 1764 |
+
"step": 14157
|
| 1765 |
+
},
|
| 1766 |
+
{
|
| 1767 |
+
"epoch": 29.258196721311474,
|
| 1768 |
+
"grad_norm": 5.4049177169799805,
|
| 1769 |
+
"learning_rate": 8.296721311475411e-06,
|
| 1770 |
+
"loss": 1.177,
|
| 1771 |
+
"step": 14278
|
| 1772 |
+
},
|
| 1773 |
+
{
|
| 1774 |
+
"epoch": 29.258196721311474,
|
| 1775 |
+
"eval_loss": 1.4221630096435547,
|
| 1776 |
+
"eval_runtime": 4.1266,
|
| 1777 |
+
"eval_samples_per_second": 105.171,
|
| 1778 |
+
"eval_steps_per_second": 13.328,
|
| 1779 |
+
"step": 14278
|
| 1780 |
+
},
|
| 1781 |
+
{
|
| 1782 |
+
"epoch": 29.506147540983605,
|
| 1783 |
+
"grad_norm": 5.744983196258545,
|
| 1784 |
+
"learning_rate": 8.197540983606558e-06,
|
| 1785 |
+
"loss": 1.1714,
|
| 1786 |
+
"step": 14399
|
| 1787 |
+
},
|
| 1788 |
+
{
|
| 1789 |
+
"epoch": 29.506147540983605,
|
| 1790 |
+
"eval_loss": 1.3760550022125244,
|
| 1791 |
+
"eval_runtime": 4.1269,
|
| 1792 |
+
"eval_samples_per_second": 105.164,
|
| 1793 |
+
"eval_steps_per_second": 13.327,
|
| 1794 |
+
"step": 14399
|
| 1795 |
+
},
|
| 1796 |
+
{
|
| 1797 |
+
"epoch": 29.75409836065574,
|
| 1798 |
+
"grad_norm": 5.267645359039307,
|
| 1799 |
+
"learning_rate": 8.098360655737706e-06,
|
| 1800 |
+
"loss": 1.1455,
|
| 1801 |
+
"step": 14520
|
| 1802 |
+
},
|
| 1803 |
+
{
|
| 1804 |
+
"epoch": 29.75409836065574,
|
| 1805 |
+
"eval_loss": 1.3890068531036377,
|
| 1806 |
+
"eval_runtime": 4.1267,
|
| 1807 |
+
"eval_samples_per_second": 105.168,
|
| 1808 |
+
"eval_steps_per_second": 13.328,
|
| 1809 |
+
"step": 14520
|
| 1810 |
+
},
|
| 1811 |
+
{
|
| 1812 |
+
"epoch": 30.00204918032787,
|
| 1813 |
+
"grad_norm": 5.940517425537109,
|
| 1814 |
+
"learning_rate": 7.999180327868852e-06,
|
| 1815 |
+
"loss": 1.1909,
|
| 1816 |
+
"step": 14641
|
| 1817 |
+
},
|
| 1818 |
+
{
|
| 1819 |
+
"epoch": 30.00204918032787,
|
| 1820 |
+
"eval_loss": 1.4213520288467407,
|
| 1821 |
+
"eval_runtime": 4.1245,
|
| 1822 |
+
"eval_samples_per_second": 105.226,
|
| 1823 |
+
"eval_steps_per_second": 13.335,
|
| 1824 |
+
"step": 14641
|
| 1825 |
+
},
|
| 1826 |
+
{
|
| 1827 |
+
"epoch": 30.25,
|
| 1828 |
+
"grad_norm": 4.8279643058776855,
|
| 1829 |
+
"learning_rate": 7.9e-06,
|
| 1830 |
+
"loss": 1.1729,
|
| 1831 |
+
"step": 14762
|
| 1832 |
+
},
|
| 1833 |
+
{
|
| 1834 |
+
"epoch": 30.25,
|
| 1835 |
+
"eval_loss": 1.387317419052124,
|
| 1836 |
+
"eval_runtime": 4.1264,
|
| 1837 |
+
"eval_samples_per_second": 105.177,
|
| 1838 |
+
"eval_steps_per_second": 13.329,
|
| 1839 |
+
"step": 14762
|
| 1840 |
+
},
|
| 1841 |
+
{
|
| 1842 |
+
"epoch": 30.49795081967213,
|
| 1843 |
+
"grad_norm": 5.653112411499023,
|
| 1844 |
+
"learning_rate": 7.800819672131149e-06,
|
| 1845 |
+
"loss": 1.1609,
|
| 1846 |
+
"step": 14883
|
| 1847 |
+
},
|
| 1848 |
+
{
|
| 1849 |
+
"epoch": 30.49795081967213,
|
| 1850 |
+
"eval_loss": 1.3974180221557617,
|
| 1851 |
+
"eval_runtime": 4.1211,
|
| 1852 |
+
"eval_samples_per_second": 105.311,
|
| 1853 |
+
"eval_steps_per_second": 13.346,
|
| 1854 |
+
"step": 14883
|
| 1855 |
+
},
|
| 1856 |
+
{
|
| 1857 |
+
"epoch": 30.74590163934426,
|
| 1858 |
+
"grad_norm": 5.124967098236084,
|
| 1859 |
+
"learning_rate": 7.701639344262295e-06,
|
| 1860 |
+
"loss": 1.1414,
|
| 1861 |
+
"step": 15004
|
| 1862 |
+
},
|
| 1863 |
+
{
|
| 1864 |
+
"epoch": 30.74590163934426,
|
| 1865 |
+
"eval_loss": 1.4140675067901611,
|
| 1866 |
+
"eval_runtime": 4.1217,
|
| 1867 |
+
"eval_samples_per_second": 105.297,
|
| 1868 |
+
"eval_steps_per_second": 13.344,
|
| 1869 |
+
"step": 15004
|
| 1870 |
+
},
|
| 1871 |
+
{
|
| 1872 |
+
"epoch": 30.993852459016395,
|
| 1873 |
+
"grad_norm": 5.296119689941406,
|
| 1874 |
+
"learning_rate": 7.6024590163934435e-06,
|
| 1875 |
+
"loss": 1.192,
|
| 1876 |
+
"step": 15125
|
| 1877 |
+
},
|
| 1878 |
+
{
|
| 1879 |
+
"epoch": 30.993852459016395,
|
| 1880 |
+
"eval_loss": 1.412348747253418,
|
| 1881 |
+
"eval_runtime": 4.1325,
|
| 1882 |
+
"eval_samples_per_second": 105.022,
|
| 1883 |
+
"eval_steps_per_second": 13.309,
|
| 1884 |
+
"step": 15125
|
| 1885 |
+
},
|
| 1886 |
+
{
|
| 1887 |
+
"epoch": 31.241803278688526,
|
| 1888 |
+
"grad_norm": 6.010870456695557,
|
| 1889 |
+
"learning_rate": 7.503278688524591e-06,
|
| 1890 |
+
"loss": 1.1565,
|
| 1891 |
+
"step": 15246
|
| 1892 |
+
},
|
| 1893 |
+
{
|
| 1894 |
+
"epoch": 31.241803278688526,
|
| 1895 |
+
"eval_loss": 1.3736042976379395,
|
| 1896 |
+
"eval_runtime": 4.1233,
|
| 1897 |
+
"eval_samples_per_second": 105.254,
|
| 1898 |
+
"eval_steps_per_second": 13.339,
|
| 1899 |
+
"step": 15246
|
| 1900 |
+
},
|
| 1901 |
+
{
|
| 1902 |
+
"epoch": 31.489754098360656,
|
| 1903 |
+
"grad_norm": 5.853750705718994,
|
| 1904 |
+
"learning_rate": 7.404098360655738e-06,
|
| 1905 |
+
"loss": 1.1371,
|
| 1906 |
+
"step": 15367
|
| 1907 |
+
},
|
| 1908 |
+
{
|
| 1909 |
+
"epoch": 31.489754098360656,
|
| 1910 |
+
"eval_loss": 1.3946257829666138,
|
| 1911 |
+
"eval_runtime": 4.1239,
|
| 1912 |
+
"eval_samples_per_second": 105.24,
|
| 1913 |
+
"eval_steps_per_second": 13.337,
|
| 1914 |
+
"step": 15367
|
| 1915 |
+
},
|
| 1916 |
+
{
|
| 1917 |
+
"epoch": 31.737704918032787,
|
| 1918 |
+
"grad_norm": 5.3367815017700195,
|
| 1919 |
+
"learning_rate": 7.304918032786886e-06,
|
| 1920 |
+
"loss": 1.1387,
|
| 1921 |
+
"step": 15488
|
| 1922 |
+
},
|
| 1923 |
+
{
|
| 1924 |
+
"epoch": 31.737704918032787,
|
| 1925 |
+
"eval_loss": 1.4051241874694824,
|
| 1926 |
+
"eval_runtime": 4.1214,
|
| 1927 |
+
"eval_samples_per_second": 105.303,
|
| 1928 |
+
"eval_steps_per_second": 13.345,
|
| 1929 |
+
"step": 15488
|
| 1930 |
+
},
|
| 1931 |
+
{
|
| 1932 |
+
"epoch": 31.985655737704917,
|
| 1933 |
+
"grad_norm": 4.613351821899414,
|
| 1934 |
+
"learning_rate": 7.205737704918034e-06,
|
| 1935 |
+
"loss": 1.1757,
|
| 1936 |
+
"step": 15609
|
| 1937 |
+
},
|
| 1938 |
+
{
|
| 1939 |
+
"epoch": 31.985655737704917,
|
| 1940 |
+
"eval_loss": 1.3825907707214355,
|
| 1941 |
+
"eval_runtime": 4.1236,
|
| 1942 |
+
"eval_samples_per_second": 105.248,
|
| 1943 |
+
"eval_steps_per_second": 13.338,
|
| 1944 |
+
"step": 15609
|
| 1945 |
+
},
|
| 1946 |
+
{
|
| 1947 |
+
"epoch": 32.23360655737705,
|
| 1948 |
+
"grad_norm": 5.111568450927734,
|
| 1949 |
+
"learning_rate": 7.106557377049181e-06,
|
| 1950 |
+
"loss": 1.1681,
|
| 1951 |
+
"step": 15730
|
| 1952 |
+
},
|
| 1953 |
+
{
|
| 1954 |
+
"epoch": 32.23360655737705,
|
| 1955 |
+
"eval_loss": 1.3923113346099854,
|
| 1956 |
+
"eval_runtime": 4.1224,
|
| 1957 |
+
"eval_samples_per_second": 105.278,
|
| 1958 |
+
"eval_steps_per_second": 13.342,
|
| 1959 |
+
"step": 15730
|
| 1960 |
+
},
|
| 1961 |
+
{
|
| 1962 |
+
"epoch": 32.48155737704918,
|
| 1963 |
+
"grad_norm": 5.637566089630127,
|
| 1964 |
+
"learning_rate": 7.007377049180328e-06,
|
| 1965 |
+
"loss": 1.1408,
|
| 1966 |
+
"step": 15851
|
| 1967 |
+
},
|
| 1968 |
+
{
|
| 1969 |
+
"epoch": 32.48155737704918,
|
| 1970 |
+
"eval_loss": 1.3988559246063232,
|
| 1971 |
+
"eval_runtime": 4.1222,
|
| 1972 |
+
"eval_samples_per_second": 105.283,
|
| 1973 |
+
"eval_steps_per_second": 13.342,
|
| 1974 |
+
"step": 15851
|
| 1975 |
+
},
|
| 1976 |
+
{
|
| 1977 |
+
"epoch": 32.72950819672131,
|
| 1978 |
+
"grad_norm": 5.146847724914551,
|
| 1979 |
+
"learning_rate": 6.9081967213114765e-06,
|
| 1980 |
+
"loss": 1.1424,
|
| 1981 |
+
"step": 15972
|
| 1982 |
+
},
|
| 1983 |
+
{
|
| 1984 |
+
"epoch": 32.72950819672131,
|
| 1985 |
+
"eval_loss": 1.3839340209960938,
|
| 1986 |
+
"eval_runtime": 4.124,
|
| 1987 |
+
"eval_samples_per_second": 105.237,
|
| 1988 |
+
"eval_steps_per_second": 13.336,
|
| 1989 |
+
"step": 15972
|
| 1990 |
+
},
|
| 1991 |
+
{
|
| 1992 |
+
"epoch": 32.97745901639344,
|
| 1993 |
+
"grad_norm": 4.275019645690918,
|
| 1994 |
+
"learning_rate": 6.809016393442624e-06,
|
| 1995 |
+
"loss": 1.1464,
|
| 1996 |
+
"step": 16093
|
| 1997 |
+
},
|
| 1998 |
+
{
|
| 1999 |
+
"epoch": 32.97745901639344,
|
| 2000 |
+
"eval_loss": 1.4004756212234497,
|
| 2001 |
+
"eval_runtime": 4.1251,
|
| 2002 |
+
"eval_samples_per_second": 105.211,
|
| 2003 |
+
"eval_steps_per_second": 13.333,
|
| 2004 |
+
"step": 16093
|
| 2005 |
+
},
|
| 2006 |
+
{
|
| 2007 |
+
"epoch": 33.22540983606557,
|
| 2008 |
+
"grad_norm": 4.860910892486572,
|
| 2009 |
+
"learning_rate": 6.709836065573771e-06,
|
| 2010 |
+
"loss": 1.145,
|
| 2011 |
+
"step": 16214
|
| 2012 |
+
},
|
| 2013 |
+
{
|
| 2014 |
+
"epoch": 33.22540983606557,
|
| 2015 |
+
"eval_loss": 1.380777359008789,
|
| 2016 |
+
"eval_runtime": 4.1264,
|
| 2017 |
+
"eval_samples_per_second": 105.177,
|
| 2018 |
+
"eval_steps_per_second": 13.329,
|
| 2019 |
+
"step": 16214
|
| 2020 |
+
},
|
| 2021 |
+
{
|
| 2022 |
+
"epoch": 33.47336065573771,
|
| 2023 |
+
"grad_norm": 5.620210647583008,
|
| 2024 |
+
"learning_rate": 6.6106557377049185e-06,
|
| 2025 |
+
"loss": 1.1134,
|
| 2026 |
+
"step": 16335
|
| 2027 |
+
},
|
| 2028 |
+
{
|
| 2029 |
+
"epoch": 33.47336065573771,
|
| 2030 |
+
"eval_loss": 1.414490818977356,
|
| 2031 |
+
"eval_runtime": 4.1242,
|
| 2032 |
+
"eval_samples_per_second": 105.231,
|
| 2033 |
+
"eval_steps_per_second": 13.336,
|
| 2034 |
+
"step": 16335
|
| 2035 |
+
},
|
| 2036 |
+
{
|
| 2037 |
+
"epoch": 33.721311475409834,
|
| 2038 |
+
"grad_norm": 5.423040866851807,
|
| 2039 |
+
"learning_rate": 6.511475409836067e-06,
|
| 2040 |
+
"loss": 1.1455,
|
| 2041 |
+
"step": 16456
|
| 2042 |
+
},
|
| 2043 |
+
{
|
| 2044 |
+
"epoch": 33.721311475409834,
|
| 2045 |
+
"eval_loss": 1.3776419162750244,
|
| 2046 |
+
"eval_runtime": 4.1244,
|
| 2047 |
+
"eval_samples_per_second": 105.227,
|
| 2048 |
+
"eval_steps_per_second": 13.335,
|
| 2049 |
+
"step": 16456
|
| 2050 |
+
},
|
| 2051 |
+
{
|
| 2052 |
+
"epoch": 33.96926229508197,
|
| 2053 |
+
"grad_norm": 4.687174320220947,
|
| 2054 |
+
"learning_rate": 6.412295081967213e-06,
|
| 2055 |
+
"loss": 1.1581,
|
| 2056 |
+
"step": 16577
|
| 2057 |
+
},
|
| 2058 |
+
{
|
| 2059 |
+
"epoch": 33.96926229508197,
|
| 2060 |
+
"eval_loss": 1.3872053623199463,
|
| 2061 |
+
"eval_runtime": 4.1225,
|
| 2062 |
+
"eval_samples_per_second": 105.276,
|
| 2063 |
+
"eval_steps_per_second": 13.341,
|
| 2064 |
+
"step": 16577
|
| 2065 |
+
},
|
| 2066 |
+
{
|
| 2067 |
+
"epoch": 34.217213114754095,
|
| 2068 |
+
"grad_norm": 4.981241703033447,
|
| 2069 |
+
"learning_rate": 6.313114754098361e-06,
|
| 2070 |
+
"loss": 1.1465,
|
| 2071 |
+
"step": 16698
|
| 2072 |
+
},
|
| 2073 |
+
{
|
| 2074 |
+
"epoch": 34.217213114754095,
|
| 2075 |
+
"eval_loss": 1.4010353088378906,
|
| 2076 |
+
"eval_runtime": 4.1257,
|
| 2077 |
+
"eval_samples_per_second": 105.195,
|
| 2078 |
+
"eval_steps_per_second": 13.331,
|
| 2079 |
+
"step": 16698
|
| 2080 |
+
},
|
| 2081 |
+
{
|
| 2082 |
+
"epoch": 34.46516393442623,
|
| 2083 |
+
"grad_norm": 5.649103164672852,
|
| 2084 |
+
"learning_rate": 6.213934426229509e-06,
|
| 2085 |
+
"loss": 1.1247,
|
| 2086 |
+
"step": 16819
|
| 2087 |
+
},
|
| 2088 |
+
{
|
| 2089 |
+
"epoch": 34.46516393442623,
|
| 2090 |
+
"eval_loss": 1.3929107189178467,
|
| 2091 |
+
"eval_runtime": 4.124,
|
| 2092 |
+
"eval_samples_per_second": 105.238,
|
| 2093 |
+
"eval_steps_per_second": 13.337,
|
| 2094 |
+
"step": 16819
|
| 2095 |
+
},
|
| 2096 |
+
{
|
| 2097 |
+
"epoch": 34.71311475409836,
|
| 2098 |
+
"grad_norm": 4.71021032333374,
|
| 2099 |
+
"learning_rate": 6.114754098360656e-06,
|
| 2100 |
+
"loss": 1.1658,
|
| 2101 |
+
"step": 16940
|
| 2102 |
+
},
|
| 2103 |
+
{
|
| 2104 |
+
"epoch": 34.71311475409836,
|
| 2105 |
+
"eval_loss": 1.4008798599243164,
|
| 2106 |
+
"eval_runtime": 4.1267,
|
| 2107 |
+
"eval_samples_per_second": 105.17,
|
| 2108 |
+
"eval_steps_per_second": 13.328,
|
| 2109 |
+
"step": 16940
|
| 2110 |
+
},
|
| 2111 |
+
{
|
| 2112 |
+
"epoch": 34.96106557377049,
|
| 2113 |
+
"grad_norm": 5.667390823364258,
|
| 2114 |
+
"learning_rate": 6.015573770491803e-06,
|
| 2115 |
+
"loss": 1.1465,
|
| 2116 |
+
"step": 17061
|
| 2117 |
+
},
|
| 2118 |
+
{
|
| 2119 |
+
"epoch": 34.96106557377049,
|
| 2120 |
+
"eval_loss": 1.3762644529342651,
|
| 2121 |
+
"eval_runtime": 4.1235,
|
| 2122 |
+
"eval_samples_per_second": 105.25,
|
| 2123 |
+
"eval_steps_per_second": 13.338,
|
| 2124 |
+
"step": 17061
|
| 2125 |
+
},
|
| 2126 |
+
{
|
| 2127 |
+
"epoch": 35.209016393442624,
|
| 2128 |
+
"grad_norm": 5.304030418395996,
|
| 2129 |
+
"learning_rate": 5.916393442622951e-06,
|
| 2130 |
+
"loss": 1.1292,
|
| 2131 |
+
"step": 17182
|
| 2132 |
+
},
|
| 2133 |
+
{
|
| 2134 |
+
"epoch": 35.209016393442624,
|
| 2135 |
+
"eval_loss": 1.3737713098526,
|
| 2136 |
+
"eval_runtime": 4.1273,
|
| 2137 |
+
"eval_samples_per_second": 105.153,
|
| 2138 |
+
"eval_steps_per_second": 13.326,
|
| 2139 |
+
"step": 17182
|
| 2140 |
+
},
|
| 2141 |
+
{
|
| 2142 |
+
"epoch": 35.45696721311475,
|
| 2143 |
+
"grad_norm": 5.078277587890625,
|
| 2144 |
+
"learning_rate": 5.817213114754099e-06,
|
| 2145 |
+
"loss": 1.0997,
|
| 2146 |
+
"step": 17303
|
| 2147 |
+
},
|
| 2148 |
+
{
|
| 2149 |
+
"epoch": 35.45696721311475,
|
| 2150 |
+
"eval_loss": 1.3623582124710083,
|
| 2151 |
+
"eval_runtime": 4.1241,
|
| 2152 |
+
"eval_samples_per_second": 105.236,
|
| 2153 |
+
"eval_steps_per_second": 13.336,
|
| 2154 |
+
"step": 17303
|
| 2155 |
+
},
|
| 2156 |
+
{
|
| 2157 |
+
"epoch": 35.704918032786885,
|
| 2158 |
+
"grad_norm": 5.220365524291992,
|
| 2159 |
+
"learning_rate": 5.718032786885246e-06,
|
| 2160 |
+
"loss": 1.1463,
|
| 2161 |
+
"step": 17424
|
| 2162 |
+
},
|
| 2163 |
+
{
|
| 2164 |
+
"epoch": 35.704918032786885,
|
| 2165 |
+
"eval_loss": 1.3832011222839355,
|
| 2166 |
+
"eval_runtime": 4.1256,
|
| 2167 |
+
"eval_samples_per_second": 105.197,
|
| 2168 |
+
"eval_steps_per_second": 13.331,
|
| 2169 |
+
"step": 17424
|
| 2170 |
+
},
|
| 2171 |
+
{
|
| 2172 |
+
"epoch": 35.95286885245902,
|
| 2173 |
+
"grad_norm": 5.473215103149414,
|
| 2174 |
+
"learning_rate": 5.618852459016394e-06,
|
| 2175 |
+
"loss": 1.1426,
|
| 2176 |
+
"step": 17545
|
| 2177 |
+
},
|
| 2178 |
+
{
|
| 2179 |
+
"epoch": 35.95286885245902,
|
| 2180 |
+
"eval_loss": 1.3841850757598877,
|
| 2181 |
+
"eval_runtime": 4.124,
|
| 2182 |
+
"eval_samples_per_second": 105.237,
|
| 2183 |
+
"eval_steps_per_second": 13.336,
|
| 2184 |
+
"step": 17545
|
| 2185 |
+
},
|
| 2186 |
+
{
|
| 2187 |
+
"epoch": 36.200819672131146,
|
| 2188 |
+
"grad_norm": 5.864542484283447,
|
| 2189 |
+
"learning_rate": 5.519672131147541e-06,
|
| 2190 |
+
"loss": 1.1237,
|
| 2191 |
+
"step": 17666
|
| 2192 |
+
},
|
| 2193 |
+
{
|
| 2194 |
+
"epoch": 36.200819672131146,
|
| 2195 |
+
"eval_loss": 1.4124358892440796,
|
| 2196 |
+
"eval_runtime": 4.1246,
|
| 2197 |
+
"eval_samples_per_second": 105.223,
|
| 2198 |
+
"eval_steps_per_second": 13.335,
|
| 2199 |
+
"step": 17666
|
| 2200 |
+
},
|
| 2201 |
+
{
|
| 2202 |
+
"epoch": 36.44877049180328,
|
| 2203 |
+
"grad_norm": 5.416541576385498,
|
| 2204 |
+
"learning_rate": 5.420491803278689e-06,
|
| 2205 |
+
"loss": 1.0961,
|
| 2206 |
+
"step": 17787
|
| 2207 |
+
},
|
| 2208 |
+
{
|
| 2209 |
+
"epoch": 36.44877049180328,
|
| 2210 |
+
"eval_loss": 1.3820034265518188,
|
| 2211 |
+
"eval_runtime": 4.1254,
|
| 2212 |
+
"eval_samples_per_second": 105.203,
|
| 2213 |
+
"eval_steps_per_second": 13.332,
|
| 2214 |
+
"step": 17787
|
| 2215 |
+
},
|
| 2216 |
+
{
|
| 2217 |
+
"epoch": 36.69672131147541,
|
| 2218 |
+
"grad_norm": 5.523916721343994,
|
| 2219 |
+
"learning_rate": 5.3213114754098365e-06,
|
| 2220 |
+
"loss": 1.1677,
|
| 2221 |
+
"step": 17908
|
| 2222 |
+
},
|
| 2223 |
+
{
|
| 2224 |
+
"epoch": 36.69672131147541,
|
| 2225 |
+
"eval_loss": 1.391912817955017,
|
| 2226 |
+
"eval_runtime": 4.124,
|
| 2227 |
+
"eval_samples_per_second": 105.238,
|
| 2228 |
+
"eval_steps_per_second": 13.337,
|
| 2229 |
+
"step": 17908
|
| 2230 |
+
},
|
| 2231 |
+
{
|
| 2232 |
+
"epoch": 36.94467213114754,
|
| 2233 |
+
"grad_norm": 6.368148326873779,
|
| 2234 |
+
"learning_rate": 5.222131147540984e-06,
|
| 2235 |
+
"loss": 1.1142,
|
| 2236 |
+
"step": 18029
|
| 2237 |
+
},
|
| 2238 |
+
{
|
| 2239 |
+
"epoch": 36.94467213114754,
|
| 2240 |
+
"eval_loss": 1.427830457687378,
|
| 2241 |
+
"eval_runtime": 4.1227,
|
| 2242 |
+
"eval_samples_per_second": 105.27,
|
| 2243 |
+
"eval_steps_per_second": 13.341,
|
| 2244 |
+
"step": 18029
|
| 2245 |
+
},
|
| 2246 |
+
{
|
| 2247 |
+
"epoch": 37.192622950819676,
|
| 2248 |
+
"grad_norm": 4.822169303894043,
|
| 2249 |
+
"learning_rate": 5.122950819672131e-06,
|
| 2250 |
+
"loss": 1.146,
|
| 2251 |
+
"step": 18150
|
| 2252 |
+
},
|
| 2253 |
+
{
|
| 2254 |
+
"epoch": 37.192622950819676,
|
| 2255 |
+
"eval_loss": 1.3914594650268555,
|
| 2256 |
+
"eval_runtime": 4.1302,
|
| 2257 |
+
"eval_samples_per_second": 105.081,
|
| 2258 |
+
"eval_steps_per_second": 13.317,
|
| 2259 |
+
"step": 18150
|
| 2260 |
+
},
|
| 2261 |
+
{
|
| 2262 |
+
"epoch": 37.4405737704918,
|
| 2263 |
+
"grad_norm": 5.081784248352051,
|
| 2264 |
+
"learning_rate": 5.023770491803279e-06,
|
| 2265 |
+
"loss": 1.1352,
|
| 2266 |
+
"step": 18271
|
| 2267 |
+
},
|
| 2268 |
+
{
|
| 2269 |
+
"epoch": 37.4405737704918,
|
| 2270 |
+
"eval_loss": 1.38656485080719,
|
| 2271 |
+
"eval_runtime": 4.1258,
|
| 2272 |
+
"eval_samples_per_second": 105.192,
|
| 2273 |
+
"eval_steps_per_second": 13.331,
|
| 2274 |
+
"step": 18271
|
| 2275 |
+
},
|
| 2276 |
+
{
|
| 2277 |
+
"epoch": 37.68852459016394,
|
| 2278 |
+
"grad_norm": 5.341165065765381,
|
| 2279 |
+
"learning_rate": 4.924590163934427e-06,
|
| 2280 |
+
"loss": 1.1353,
|
| 2281 |
+
"step": 18392
|
| 2282 |
+
},
|
| 2283 |
+
{
|
| 2284 |
+
"epoch": 37.68852459016394,
|
| 2285 |
+
"eval_loss": 1.3952445983886719,
|
| 2286 |
+
"eval_runtime": 4.1257,
|
| 2287 |
+
"eval_samples_per_second": 105.195,
|
| 2288 |
+
"eval_steps_per_second": 13.331,
|
| 2289 |
+
"step": 18392
|
| 2290 |
+
},
|
| 2291 |
+
{
|
| 2292 |
+
"epoch": 37.93647540983606,
|
| 2293 |
+
"grad_norm": 5.281919002532959,
|
| 2294 |
+
"learning_rate": 4.825409836065574e-06,
|
| 2295 |
+
"loss": 1.1099,
|
| 2296 |
+
"step": 18513
|
| 2297 |
+
},
|
| 2298 |
+
{
|
| 2299 |
+
"epoch": 37.93647540983606,
|
| 2300 |
+
"eval_loss": 1.3772515058517456,
|
| 2301 |
+
"eval_runtime": 4.1261,
|
| 2302 |
+
"eval_samples_per_second": 105.184,
|
| 2303 |
+
"eval_steps_per_second": 13.33,
|
| 2304 |
+
"step": 18513
|
| 2305 |
+
},
|
| 2306 |
+
{
|
| 2307 |
+
"epoch": 38.1844262295082,
|
| 2308 |
+
"grad_norm": 5.485556125640869,
|
| 2309 |
+
"learning_rate": 4.726229508196722e-06,
|
| 2310 |
+
"loss": 1.1088,
|
| 2311 |
+
"step": 18634
|
| 2312 |
+
},
|
| 2313 |
+
{
|
| 2314 |
+
"epoch": 38.1844262295082,
|
| 2315 |
+
"eval_loss": 1.3951290845870972,
|
| 2316 |
+
"eval_runtime": 4.1248,
|
| 2317 |
+
"eval_samples_per_second": 105.218,
|
| 2318 |
+
"eval_steps_per_second": 13.334,
|
| 2319 |
+
"step": 18634
|
| 2320 |
+
},
|
| 2321 |
+
{
|
| 2322 |
+
"epoch": 38.432377049180324,
|
| 2323 |
+
"grad_norm": 4.903234004974365,
|
| 2324 |
+
"learning_rate": 4.6270491803278695e-06,
|
| 2325 |
+
"loss": 1.1141,
|
| 2326 |
+
"step": 18755
|
| 2327 |
+
},
|
| 2328 |
+
{
|
| 2329 |
+
"epoch": 38.432377049180324,
|
| 2330 |
+
"eval_loss": 1.4029154777526855,
|
| 2331 |
+
"eval_runtime": 4.1245,
|
| 2332 |
+
"eval_samples_per_second": 105.224,
|
| 2333 |
+
"eval_steps_per_second": 13.335,
|
| 2334 |
+
"step": 18755
|
| 2335 |
+
},
|
| 2336 |
+
{
|
| 2337 |
+
"epoch": 38.68032786885246,
|
| 2338 |
+
"grad_norm": 6.011325359344482,
|
| 2339 |
+
"learning_rate": 4.527868852459017e-06,
|
| 2340 |
+
"loss": 1.1115,
|
| 2341 |
+
"step": 18876
|
| 2342 |
+
},
|
| 2343 |
+
{
|
| 2344 |
+
"epoch": 38.68032786885246,
|
| 2345 |
+
"eval_loss": 1.3774739503860474,
|
| 2346 |
+
"eval_runtime": 4.1241,
|
| 2347 |
+
"eval_samples_per_second": 105.234,
|
| 2348 |
+
"eval_steps_per_second": 13.336,
|
| 2349 |
+
"step": 18876
|
| 2350 |
+
},
|
| 2351 |
+
{
|
| 2352 |
+
"epoch": 38.92827868852459,
|
| 2353 |
+
"grad_norm": 5.690525531768799,
|
| 2354 |
+
"learning_rate": 4.428688524590164e-06,
|
| 2355 |
+
"loss": 1.1552,
|
| 2356 |
+
"step": 18997
|
| 2357 |
+
},
|
| 2358 |
+
{
|
| 2359 |
+
"epoch": 38.92827868852459,
|
| 2360 |
+
"eval_loss": 1.3836060762405396,
|
| 2361 |
+
"eval_runtime": 4.1288,
|
| 2362 |
+
"eval_samples_per_second": 105.116,
|
| 2363 |
+
"eval_steps_per_second": 13.321,
|
| 2364 |
+
"step": 18997
|
| 2365 |
+
},
|
| 2366 |
+
{
|
| 2367 |
+
"epoch": 39.17622950819672,
|
| 2368 |
+
"grad_norm": 4.86214017868042,
|
| 2369 |
+
"learning_rate": 4.329508196721312e-06,
|
| 2370 |
+
"loss": 1.1291,
|
| 2371 |
+
"step": 19118
|
| 2372 |
+
},
|
| 2373 |
+
{
|
| 2374 |
+
"epoch": 39.17622950819672,
|
| 2375 |
+
"eval_loss": 1.3931368589401245,
|
| 2376 |
+
"eval_runtime": 4.1259,
|
| 2377 |
+
"eval_samples_per_second": 105.189,
|
| 2378 |
+
"eval_steps_per_second": 13.33,
|
| 2379 |
+
"step": 19118
|
| 2380 |
+
},
|
| 2381 |
+
{
|
| 2382 |
+
"epoch": 39.424180327868854,
|
| 2383 |
+
"grad_norm": 5.337844371795654,
|
| 2384 |
+
"learning_rate": 4.23032786885246e-06,
|
| 2385 |
+
"loss": 1.1261,
|
| 2386 |
+
"step": 19239
|
| 2387 |
+
},
|
| 2388 |
+
{
|
| 2389 |
+
"epoch": 39.424180327868854,
|
| 2390 |
+
"eval_loss": 1.393588662147522,
|
| 2391 |
+
"eval_runtime": 4.1276,
|
| 2392 |
+
"eval_samples_per_second": 105.146,
|
| 2393 |
+
"eval_steps_per_second": 13.325,
|
| 2394 |
+
"step": 19239
|
| 2395 |
+
},
|
| 2396 |
+
{
|
| 2397 |
+
"epoch": 39.67213114754098,
|
| 2398 |
+
"grad_norm": 5.509841442108154,
|
| 2399 |
+
"learning_rate": 4.131147540983607e-06,
|
| 2400 |
+
"loss": 1.1209,
|
| 2401 |
+
"step": 19360
|
| 2402 |
+
},
|
| 2403 |
+
{
|
| 2404 |
+
"epoch": 39.67213114754098,
|
| 2405 |
+
"eval_loss": 1.4069132804870605,
|
| 2406 |
+
"eval_runtime": 4.1261,
|
| 2407 |
+
"eval_samples_per_second": 105.184,
|
| 2408 |
+
"eval_steps_per_second": 13.33,
|
| 2409 |
+
"step": 19360
|
| 2410 |
+
},
|
| 2411 |
+
{
|
| 2412 |
+
"epoch": 39.920081967213115,
|
| 2413 |
+
"grad_norm": 4.939957141876221,
|
| 2414 |
+
"learning_rate": 4.031967213114754e-06,
|
| 2415 |
+
"loss": 1.124,
|
| 2416 |
+
"step": 19481
|
| 2417 |
+
},
|
| 2418 |
+
{
|
| 2419 |
+
"epoch": 39.920081967213115,
|
| 2420 |
+
"eval_loss": 1.391053318977356,
|
| 2421 |
+
"eval_runtime": 4.1267,
|
| 2422 |
+
"eval_samples_per_second": 105.169,
|
| 2423 |
+
"eval_steps_per_second": 13.328,
|
| 2424 |
+
"step": 19481
|
| 2425 |
+
},
|
| 2426 |
+
{
|
| 2427 |
+
"epoch": 40.16803278688525,
|
| 2428 |
+
"grad_norm": 5.964232444763184,
|
| 2429 |
+
"learning_rate": 3.932786885245902e-06,
|
| 2430 |
+
"loss": 1.1374,
|
| 2431 |
+
"step": 19602
|
| 2432 |
+
},
|
| 2433 |
+
{
|
| 2434 |
+
"epoch": 40.16803278688525,
|
| 2435 |
+
"eval_loss": 1.3899863958358765,
|
| 2436 |
+
"eval_runtime": 4.1261,
|
| 2437 |
+
"eval_samples_per_second": 105.183,
|
| 2438 |
+
"eval_steps_per_second": 13.33,
|
| 2439 |
+
"step": 19602
|
| 2440 |
+
},
|
| 2441 |
+
{
|
| 2442 |
+
"epoch": 40.415983606557376,
|
| 2443 |
+
"grad_norm": 5.4043049812316895,
|
| 2444 |
+
"learning_rate": 3.833606557377049e-06,
|
| 2445 |
+
"loss": 1.1079,
|
| 2446 |
+
"step": 19723
|
| 2447 |
+
},
|
| 2448 |
+
{
|
| 2449 |
+
"epoch": 40.415983606557376,
|
| 2450 |
+
"eval_loss": 1.386367678642273,
|
| 2451 |
+
"eval_runtime": 4.127,
|
| 2452 |
+
"eval_samples_per_second": 105.162,
|
| 2453 |
+
"eval_steps_per_second": 13.327,
|
| 2454 |
+
"step": 19723
|
| 2455 |
+
},
|
| 2456 |
+
{
|
| 2457 |
+
"epoch": 40.66393442622951,
|
| 2458 |
+
"grad_norm": 5.620067596435547,
|
| 2459 |
+
"learning_rate": 3.734426229508197e-06,
|
| 2460 |
+
"loss": 1.1286,
|
| 2461 |
+
"step": 19844
|
| 2462 |
+
},
|
| 2463 |
+
{
|
| 2464 |
+
"epoch": 40.66393442622951,
|
| 2465 |
+
"eval_loss": 1.3876351118087769,
|
| 2466 |
+
"eval_runtime": 4.1244,
|
| 2467 |
+
"eval_samples_per_second": 105.228,
|
| 2468 |
+
"eval_steps_per_second": 13.335,
|
| 2469 |
+
"step": 19844
|
| 2470 |
+
},
|
| 2471 |
+
{
|
| 2472 |
+
"epoch": 40.91188524590164,
|
| 2473 |
+
"grad_norm": 5.134771347045898,
|
| 2474 |
+
"learning_rate": 3.635245901639344e-06,
|
| 2475 |
+
"loss": 1.103,
|
| 2476 |
+
"step": 19965
|
| 2477 |
+
},
|
| 2478 |
+
{
|
| 2479 |
+
"epoch": 40.91188524590164,
|
| 2480 |
+
"eval_loss": 1.401169776916504,
|
| 2481 |
+
"eval_runtime": 4.125,
|
| 2482 |
+
"eval_samples_per_second": 105.212,
|
| 2483 |
+
"eval_steps_per_second": 13.333,
|
| 2484 |
+
"step": 19965
|
| 2485 |
+
},
|
| 2486 |
+
{
|
| 2487 |
+
"epoch": 41.15983606557377,
|
| 2488 |
+
"grad_norm": 5.033725261688232,
|
| 2489 |
+
"learning_rate": 3.536065573770492e-06,
|
| 2490 |
+
"loss": 1.1017,
|
| 2491 |
+
"step": 20086
|
| 2492 |
+
},
|
| 2493 |
+
{
|
| 2494 |
+
"epoch": 41.15983606557377,
|
| 2495 |
+
"eval_loss": 1.413968801498413,
|
| 2496 |
+
"eval_runtime": 4.1265,
|
| 2497 |
+
"eval_samples_per_second": 105.175,
|
| 2498 |
+
"eval_steps_per_second": 13.329,
|
| 2499 |
+
"step": 20086
|
| 2500 |
+
},
|
| 2501 |
+
{
|
| 2502 |
+
"epoch": 41.407786885245905,
|
| 2503 |
+
"grad_norm": 5.851089000701904,
|
| 2504 |
+
"learning_rate": 3.4368852459016393e-06,
|
| 2505 |
+
"loss": 1.1036,
|
| 2506 |
+
"step": 20207
|
| 2507 |
+
},
|
| 2508 |
+
{
|
| 2509 |
+
"epoch": 41.407786885245905,
|
| 2510 |
+
"eval_loss": 1.369694709777832,
|
| 2511 |
+
"eval_runtime": 4.1274,
|
| 2512 |
+
"eval_samples_per_second": 105.151,
|
| 2513 |
+
"eval_steps_per_second": 13.326,
|
| 2514 |
+
"step": 20207
|
| 2515 |
+
},
|
| 2516 |
+
{
|
| 2517 |
+
"epoch": 41.65573770491803,
|
| 2518 |
+
"grad_norm": 5.828238010406494,
|
| 2519 |
+
"learning_rate": 3.337704918032787e-06,
|
| 2520 |
+
"loss": 1.1052,
|
| 2521 |
+
"step": 20328
|
| 2522 |
+
},
|
| 2523 |
+
{
|
| 2524 |
+
"epoch": 41.65573770491803,
|
| 2525 |
+
"eval_loss": 1.4037874937057495,
|
| 2526 |
+
"eval_runtime": 4.1236,
|
| 2527 |
+
"eval_samples_per_second": 105.247,
|
| 2528 |
+
"eval_steps_per_second": 13.338,
|
| 2529 |
+
"step": 20328
|
| 2530 |
+
},
|
| 2531 |
+
{
|
| 2532 |
+
"epoch": 41.903688524590166,
|
| 2533 |
+
"grad_norm": 5.322448253631592,
|
| 2534 |
+
"learning_rate": 3.2385245901639344e-06,
|
| 2535 |
+
"loss": 1.0977,
|
| 2536 |
+
"step": 20449
|
| 2537 |
+
},
|
| 2538 |
+
{
|
| 2539 |
+
"epoch": 41.903688524590166,
|
| 2540 |
+
"eval_loss": 1.3905967473983765,
|
| 2541 |
+
"eval_runtime": 4.1253,
|
| 2542 |
+
"eval_samples_per_second": 105.204,
|
| 2543 |
+
"eval_steps_per_second": 13.332,
|
| 2544 |
+
"step": 20449
|
| 2545 |
+
},
|
| 2546 |
+
{
|
| 2547 |
+
"epoch": 42.15163934426229,
|
| 2548 |
+
"grad_norm": 5.0981526374816895,
|
| 2549 |
+
"learning_rate": 3.139344262295082e-06,
|
| 2550 |
+
"loss": 1.1343,
|
| 2551 |
+
"step": 20570
|
| 2552 |
+
},
|
| 2553 |
+
{
|
| 2554 |
+
"epoch": 42.15163934426229,
|
| 2555 |
+
"eval_loss": 1.3920131921768188,
|
| 2556 |
+
"eval_runtime": 4.1259,
|
| 2557 |
+
"eval_samples_per_second": 105.19,
|
| 2558 |
+
"eval_steps_per_second": 13.331,
|
| 2559 |
+
"step": 20570
|
| 2560 |
+
},
|
| 2561 |
+
{
|
| 2562 |
+
"epoch": 42.39959016393443,
|
| 2563 |
+
"grad_norm": 5.493677616119385,
|
| 2564 |
+
"learning_rate": 3.0401639344262295e-06,
|
| 2565 |
+
"loss": 1.103,
|
| 2566 |
+
"step": 20691
|
| 2567 |
+
},
|
| 2568 |
+
{
|
| 2569 |
+
"epoch": 42.39959016393443,
|
| 2570 |
+
"eval_loss": 1.387071132659912,
|
| 2571 |
+
"eval_runtime": 4.1293,
|
| 2572 |
+
"eval_samples_per_second": 105.102,
|
| 2573 |
+
"eval_steps_per_second": 13.319,
|
| 2574 |
+
"step": 20691
|
| 2575 |
+
},
|
| 2576 |
+
{
|
| 2577 |
+
"epoch": 42.64754098360656,
|
| 2578 |
+
"grad_norm": 5.632537364959717,
|
| 2579 |
+
"learning_rate": 2.9409836065573772e-06,
|
| 2580 |
+
"loss": 1.1092,
|
| 2581 |
+
"step": 20812
|
| 2582 |
+
},
|
| 2583 |
+
{
|
| 2584 |
+
"epoch": 42.64754098360656,
|
| 2585 |
+
"eval_loss": 1.387178659439087,
|
| 2586 |
+
"eval_runtime": 4.1265,
|
| 2587 |
+
"eval_samples_per_second": 105.175,
|
| 2588 |
+
"eval_steps_per_second": 13.329,
|
| 2589 |
+
"step": 20812
|
| 2590 |
+
},
|
| 2591 |
+
{
|
| 2592 |
+
"epoch": 42.89549180327869,
|
| 2593 |
+
"grad_norm": 5.351899147033691,
|
| 2594 |
+
"learning_rate": 2.8418032786885246e-06,
|
| 2595 |
+
"loss": 1.0858,
|
| 2596 |
+
"step": 20933
|
| 2597 |
+
},
|
| 2598 |
+
{
|
| 2599 |
+
"epoch": 42.89549180327869,
|
| 2600 |
+
"eval_loss": 1.3848626613616943,
|
| 2601 |
+
"eval_runtime": 4.1255,
|
| 2602 |
+
"eval_samples_per_second": 105.199,
|
| 2603 |
+
"eval_steps_per_second": 13.332,
|
| 2604 |
+
"step": 20933
|
| 2605 |
+
},
|
| 2606 |
+
{
|
| 2607 |
+
"epoch": 43.14344262295082,
|
| 2608 |
+
"grad_norm": 7.1183247566223145,
|
| 2609 |
+
"learning_rate": 2.7426229508196723e-06,
|
| 2610 |
+
"loss": 1.0998,
|
| 2611 |
+
"step": 21054
|
| 2612 |
+
},
|
| 2613 |
+
{
|
| 2614 |
+
"epoch": 43.14344262295082,
|
| 2615 |
+
"eval_loss": 1.3888462781906128,
|
| 2616 |
+
"eval_runtime": 4.1286,
|
| 2617 |
+
"eval_samples_per_second": 105.12,
|
| 2618 |
+
"eval_steps_per_second": 13.322,
|
| 2619 |
+
"step": 21054
|
| 2620 |
+
},
|
| 2621 |
+
{
|
| 2622 |
+
"epoch": 43.39139344262295,
|
| 2623 |
+
"grad_norm": 5.993564128875732,
|
| 2624 |
+
"learning_rate": 2.6434426229508197e-06,
|
| 2625 |
+
"loss": 1.1082,
|
| 2626 |
+
"step": 21175
|
| 2627 |
+
},
|
| 2628 |
+
{
|
| 2629 |
+
"epoch": 43.39139344262295,
|
| 2630 |
+
"eval_loss": 1.3925265073776245,
|
| 2631 |
+
"eval_runtime": 4.1287,
|
| 2632 |
+
"eval_samples_per_second": 105.118,
|
| 2633 |
+
"eval_steps_per_second": 13.321,
|
| 2634 |
+
"step": 21175
|
| 2635 |
+
},
|
| 2636 |
+
{
|
| 2637 |
+
"epoch": 43.63934426229508,
|
| 2638 |
+
"grad_norm": 4.941683292388916,
|
| 2639 |
+
"learning_rate": 2.5442622950819674e-06,
|
| 2640 |
+
"loss": 1.1112,
|
| 2641 |
+
"step": 21296
|
| 2642 |
+
},
|
| 2643 |
+
{
|
| 2644 |
+
"epoch": 43.63934426229508,
|
| 2645 |
+
"eval_loss": 1.4261186122894287,
|
| 2646 |
+
"eval_runtime": 4.1253,
|
| 2647 |
+
"eval_samples_per_second": 105.205,
|
| 2648 |
+
"eval_steps_per_second": 13.332,
|
| 2649 |
+
"step": 21296
|
| 2650 |
+
},
|
| 2651 |
+
{
|
| 2652 |
+
"epoch": 43.88729508196721,
|
| 2653 |
+
"grad_norm": 5.67835807800293,
|
| 2654 |
+
"learning_rate": 2.4450819672131148e-06,
|
| 2655 |
+
"loss": 1.0969,
|
| 2656 |
+
"step": 21417
|
| 2657 |
+
},
|
| 2658 |
+
{
|
| 2659 |
+
"epoch": 43.88729508196721,
|
| 2660 |
+
"eval_loss": 1.3613173961639404,
|
| 2661 |
+
"eval_runtime": 4.1247,
|
| 2662 |
+
"eval_samples_per_second": 105.219,
|
| 2663 |
+
"eval_steps_per_second": 13.334,
|
| 2664 |
+
"step": 21417
|
| 2665 |
+
},
|
| 2666 |
+
{
|
| 2667 |
+
"epoch": 44.135245901639344,
|
| 2668 |
+
"grad_norm": 5.428313255310059,
|
| 2669 |
+
"learning_rate": 2.3459016393442625e-06,
|
| 2670 |
+
"loss": 1.112,
|
| 2671 |
+
"step": 21538
|
| 2672 |
+
},
|
| 2673 |
+
{
|
| 2674 |
+
"epoch": 44.135245901639344,
|
| 2675 |
+
"eval_loss": 1.387001395225525,
|
| 2676 |
+
"eval_runtime": 4.1267,
|
| 2677 |
+
"eval_samples_per_second": 105.17,
|
| 2678 |
+
"eval_steps_per_second": 13.328,
|
| 2679 |
+
"step": 21538
|
| 2680 |
+
},
|
| 2681 |
+
{
|
| 2682 |
+
"epoch": 44.38319672131148,
|
| 2683 |
+
"grad_norm": 5.432636260986328,
|
| 2684 |
+
"learning_rate": 2.24672131147541e-06,
|
| 2685 |
+
"loss": 1.0919,
|
| 2686 |
+
"step": 21659
|
| 2687 |
+
},
|
| 2688 |
+
{
|
| 2689 |
+
"epoch": 44.38319672131148,
|
| 2690 |
+
"eval_loss": 1.4082472324371338,
|
| 2691 |
+
"eval_runtime": 4.1279,
|
| 2692 |
+
"eval_samples_per_second": 105.139,
|
| 2693 |
+
"eval_steps_per_second": 13.324,
|
| 2694 |
+
"step": 21659
|
| 2695 |
+
},
|
| 2696 |
+
{
|
| 2697 |
+
"epoch": 44.631147540983605,
|
| 2698 |
+
"grad_norm": 5.585594177246094,
|
| 2699 |
+
"learning_rate": 2.1475409836065576e-06,
|
| 2700 |
+
"loss": 1.0921,
|
| 2701 |
+
"step": 21780
|
| 2702 |
+
},
|
| 2703 |
+
{
|
| 2704 |
+
"epoch": 44.631147540983605,
|
| 2705 |
+
"eval_loss": 1.3896039724349976,
|
| 2706 |
+
"eval_runtime": 4.1241,
|
| 2707 |
+
"eval_samples_per_second": 105.234,
|
| 2708 |
+
"eval_steps_per_second": 13.336,
|
| 2709 |
+
"step": 21780
|
| 2710 |
+
},
|
| 2711 |
+
{
|
| 2712 |
+
"epoch": 44.87909836065574,
|
| 2713 |
+
"grad_norm": 5.751293659210205,
|
| 2714 |
+
"learning_rate": 2.048360655737705e-06,
|
| 2715 |
+
"loss": 1.1125,
|
| 2716 |
+
"step": 21901
|
| 2717 |
+
},
|
| 2718 |
+
{
|
| 2719 |
+
"epoch": 44.87909836065574,
|
| 2720 |
+
"eval_loss": 1.3886334896087646,
|
| 2721 |
+
"eval_runtime": 4.1257,
|
| 2722 |
+
"eval_samples_per_second": 105.194,
|
| 2723 |
+
"eval_steps_per_second": 13.331,
|
| 2724 |
+
"step": 21901
|
| 2725 |
+
},
|
| 2726 |
+
{
|
| 2727 |
+
"epoch": 45.127049180327866,
|
| 2728 |
+
"grad_norm": 5.387383937835693,
|
| 2729 |
+
"learning_rate": 1.9491803278688527e-06,
|
| 2730 |
+
"loss": 1.0884,
|
| 2731 |
+
"step": 22022
|
| 2732 |
+
},
|
| 2733 |
+
{
|
| 2734 |
+
"epoch": 45.127049180327866,
|
| 2735 |
+
"eval_loss": 1.3778773546218872,
|
| 2736 |
+
"eval_runtime": 4.1273,
|
| 2737 |
+
"eval_samples_per_second": 105.154,
|
| 2738 |
+
"eval_steps_per_second": 13.326,
|
| 2739 |
+
"step": 22022
|
| 2740 |
+
},
|
| 2741 |
+
{
|
| 2742 |
+
"epoch": 45.375,
|
| 2743 |
+
"grad_norm": 4.957601070404053,
|
| 2744 |
+
"learning_rate": 1.85e-06,
|
| 2745 |
+
"loss": 1.0855,
|
| 2746 |
+
"step": 22143
|
| 2747 |
+
},
|
| 2748 |
+
{
|
| 2749 |
+
"epoch": 45.375,
|
| 2750 |
+
"eval_loss": 1.3757483959197998,
|
| 2751 |
+
"eval_runtime": 4.1246,
|
| 2752 |
+
"eval_samples_per_second": 105.223,
|
| 2753 |
+
"eval_steps_per_second": 13.335,
|
| 2754 |
+
"step": 22143
|
| 2755 |
+
},
|
| 2756 |
+
{
|
| 2757 |
+
"epoch": 45.622950819672134,
|
| 2758 |
+
"grad_norm": 5.841992378234863,
|
| 2759 |
+
"learning_rate": 1.7508196721311476e-06,
|
| 2760 |
+
"loss": 1.1281,
|
| 2761 |
+
"step": 22264
|
| 2762 |
+
},
|
| 2763 |
+
{
|
| 2764 |
+
"epoch": 45.622950819672134,
|
| 2765 |
+
"eval_loss": 1.385141372680664,
|
| 2766 |
+
"eval_runtime": 4.132,
|
| 2767 |
+
"eval_samples_per_second": 105.034,
|
| 2768 |
+
"eval_steps_per_second": 13.311,
|
| 2769 |
+
"step": 22264
|
| 2770 |
+
},
|
| 2771 |
+
{
|
| 2772 |
+
"epoch": 45.87090163934426,
|
| 2773 |
+
"grad_norm": 5.011885643005371,
|
| 2774 |
+
"learning_rate": 1.6516393442622952e-06,
|
| 2775 |
+
"loss": 1.0856,
|
| 2776 |
+
"step": 22385
|
| 2777 |
+
},
|
| 2778 |
+
{
|
| 2779 |
+
"epoch": 45.87090163934426,
|
| 2780 |
+
"eval_loss": 1.364372968673706,
|
| 2781 |
+
"eval_runtime": 4.1268,
|
| 2782 |
+
"eval_samples_per_second": 105.167,
|
| 2783 |
+
"eval_steps_per_second": 13.328,
|
| 2784 |
+
"step": 22385
|
| 2785 |
+
},
|
| 2786 |
+
{
|
| 2787 |
+
"epoch": 46.118852459016395,
|
| 2788 |
+
"grad_norm": 5.6464996337890625,
|
| 2789 |
+
"learning_rate": 1.5524590163934427e-06,
|
| 2790 |
+
"loss": 1.0963,
|
| 2791 |
+
"step": 22506
|
| 2792 |
+
},
|
| 2793 |
+
{
|
| 2794 |
+
"epoch": 46.118852459016395,
|
| 2795 |
+
"eval_loss": 1.3527668714523315,
|
| 2796 |
+
"eval_runtime": 4.122,
|
| 2797 |
+
"eval_samples_per_second": 105.289,
|
| 2798 |
+
"eval_steps_per_second": 13.343,
|
| 2799 |
+
"step": 22506
|
| 2800 |
+
}
|
| 2801 |
+
],
|
| 2802 |
+
"logging_steps": 121,
|
| 2803 |
+
"max_steps": 24400,
|
| 2804 |
+
"num_input_tokens_seen": 0,
|
| 2805 |
+
"num_train_epochs": 50,
|
| 2806 |
+
"save_steps": 121,
|
| 2807 |
+
"stateful_callbacks": {
|
| 2808 |
+
"TrainerControl": {
|
| 2809 |
+
"args": {
|
| 2810 |
+
"should_epoch_stop": false,
|
| 2811 |
+
"should_evaluate": false,
|
| 2812 |
+
"should_log": false,
|
| 2813 |
+
"should_save": true,
|
| 2814 |
+
"should_training_stop": false
|
| 2815 |
+
},
|
| 2816 |
+
"attributes": {}
|
| 2817 |
+
}
|
| 2818 |
+
},
|
| 2819 |
+
"total_flos": 4.732761177640141e+16,
|
| 2820 |
+
"train_batch_size": 8,
|
| 2821 |
+
"trial_name": null,
|
| 2822 |
+
"trial_params": null
|
| 2823 |
+
}
|
training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:4f927187e98d4800d729512fbebdabfe4add1bf40c51222bc94b304579b654e5
|
| 3 |
+
size 5368
|