Commit
•
09d80d2
1
Parent(s):
c0a8bfa
End of training
Browse files- README.md +5 -3
- all_results.json +15 -0
- eval_results.json +9 -0
- train_results.json +10 -0
- trainer_state.json +3222 -0
README.md
CHANGED
|
@@ -1,5 +1,7 @@
|
|
| 1 |
---
|
| 2 |
library_name: transformers
|
|
|
|
|
|
|
| 3 |
license: apache-2.0
|
| 4 |
base_model: BEE-spoke-data/tFINE-680m-e32-d16-infinity_instruct-L1
|
| 5 |
tags:
|
|
@@ -14,10 +16,10 @@ should probably proofread and complete it, then remove this comment. -->
|
|
| 14 |
|
| 15 |
# tFINE-680m-e32-d16-infinity_instruct-L1-infinity-instruct-7m-T2T_en-1024
|
| 16 |
|
| 17 |
-
This model is a fine-tuned version of [BEE-spoke-data/tFINE-680m-e32-d16-infinity_instruct-L1](https://huggingface.co/BEE-spoke-data/tFINE-680m-e32-d16-infinity_instruct-L1) on
|
| 18 |
It achieves the following results on the evaluation set:
|
| 19 |
-
- Loss: 1.
|
| 20 |
-
- Num Input Tokens Seen:
|
| 21 |
|
| 22 |
## Model description
|
| 23 |
|
|
|
|
| 1 |
---
|
| 2 |
library_name: transformers
|
| 3 |
+
language:
|
| 4 |
+
- en
|
| 5 |
license: apache-2.0
|
| 6 |
base_model: BEE-spoke-data/tFINE-680m-e32-d16-infinity_instruct-L1
|
| 7 |
tags:
|
|
|
|
| 16 |
|
| 17 |
# tFINE-680m-e32-d16-infinity_instruct-L1-infinity-instruct-7m-T2T_en-1024
|
| 18 |
|
| 19 |
+
This model is a fine-tuned version of [BEE-spoke-data/tFINE-680m-e32-d16-infinity_instruct-L1](https://huggingface.co/BEE-spoke-data/tFINE-680m-e32-d16-infinity_instruct-L1) on the pszemraj/infinity-instruct-7m-T2T_en dataset.
|
| 20 |
It achieves the following results on the evaluation set:
|
| 21 |
+
- Loss: 1.3139
|
| 22 |
+
- Num Input Tokens Seen: 361724696
|
| 23 |
|
| 24 |
## Model description
|
| 25 |
|
all_results.json
ADDED
|
@@ -0,0 +1,15 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 0.9998574709003089,
|
| 3 |
+
"eval_loss": 1.3138757944107056,
|
| 4 |
+
"eval_runtime": 2.7934,
|
| 5 |
+
"eval_samples": 150,
|
| 6 |
+
"eval_samples_per_second": 53.699,
|
| 7 |
+
"eval_steps_per_second": 6.802,
|
| 8 |
+
"num_input_tokens_seen": 361724696,
|
| 9 |
+
"total_flos": 1.4115183327245763e+18,
|
| 10 |
+
"train_loss": 1.3691888915302182,
|
| 11 |
+
"train_runtime": 65409.2818,
|
| 12 |
+
"train_samples": 1010316,
|
| 13 |
+
"train_samples_per_second": 15.446,
|
| 14 |
+
"train_steps_per_second": 0.06
|
| 15 |
+
}
|
eval_results.json
ADDED
|
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 0.9998574709003089,
|
| 3 |
+
"eval_loss": 1.3138757944107056,
|
| 4 |
+
"eval_runtime": 2.7934,
|
| 5 |
+
"eval_samples": 150,
|
| 6 |
+
"eval_samples_per_second": 53.699,
|
| 7 |
+
"eval_steps_per_second": 6.802,
|
| 8 |
+
"num_input_tokens_seen": 361724696
|
| 9 |
+
}
|
train_results.json
ADDED
|
@@ -0,0 +1,10 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 0.9998574709003089,
|
| 3 |
+
"num_input_tokens_seen": 361724696,
|
| 4 |
+
"total_flos": 1.4115183327245763e+18,
|
| 5 |
+
"train_loss": 1.3691888915302182,
|
| 6 |
+
"train_runtime": 65409.2818,
|
| 7 |
+
"train_samples": 1010316,
|
| 8 |
+
"train_samples_per_second": 15.446,
|
| 9 |
+
"train_steps_per_second": 0.06
|
| 10 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,3222 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 0.9998574709003089,
|
| 5 |
+
"eval_steps": 1000,
|
| 6 |
+
"global_step": 3946,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.002533850661176657,
|
| 13 |
+
"grad_norm": 0.2532438337802887,
|
| 14 |
+
"learning_rate": 3.1645569620253167e-06,
|
| 15 |
+
"loss": 1.4823,
|
| 16 |
+
"num_input_tokens_seen": 930032,
|
| 17 |
+
"step": 10
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.005067701322353314,
|
| 21 |
+
"grad_norm": 0.2576071619987488,
|
| 22 |
+
"learning_rate": 6.329113924050633e-06,
|
| 23 |
+
"loss": 1.45,
|
| 24 |
+
"num_input_tokens_seen": 1834624,
|
| 25 |
+
"step": 20
|
| 26 |
+
},
|
| 27 |
+
{
|
| 28 |
+
"epoch": 0.007601551983529971,
|
| 29 |
+
"grad_norm": 0.27260521054267883,
|
| 30 |
+
"learning_rate": 9.49367088607595e-06,
|
| 31 |
+
"loss": 1.44,
|
| 32 |
+
"num_input_tokens_seen": 2731480,
|
| 33 |
+
"step": 30
|
| 34 |
+
},
|
| 35 |
+
{
|
| 36 |
+
"epoch": 0.010135402644706628,
|
| 37 |
+
"grad_norm": 0.2555162012577057,
|
| 38 |
+
"learning_rate": 1.2658227848101267e-05,
|
| 39 |
+
"loss": 1.4811,
|
| 40 |
+
"num_input_tokens_seen": 3620696,
|
| 41 |
+
"step": 40
|
| 42 |
+
},
|
| 43 |
+
{
|
| 44 |
+
"epoch": 0.012669253305883284,
|
| 45 |
+
"grad_norm": 0.24727804958820343,
|
| 46 |
+
"learning_rate": 1.5822784810126583e-05,
|
| 47 |
+
"loss": 1.4547,
|
| 48 |
+
"num_input_tokens_seen": 4537164,
|
| 49 |
+
"step": 50
|
| 50 |
+
},
|
| 51 |
+
{
|
| 52 |
+
"epoch": 0.015203103967059942,
|
| 53 |
+
"grad_norm": 0.26695573329925537,
|
| 54 |
+
"learning_rate": 1.89873417721519e-05,
|
| 55 |
+
"loss": 1.4288,
|
| 56 |
+
"num_input_tokens_seen": 5457344,
|
| 57 |
+
"step": 60
|
| 58 |
+
},
|
| 59 |
+
{
|
| 60 |
+
"epoch": 0.017736954628236597,
|
| 61 |
+
"grad_norm": 0.2801561653614044,
|
| 62 |
+
"learning_rate": 2.2151898734177217e-05,
|
| 63 |
+
"loss": 1.4569,
|
| 64 |
+
"num_input_tokens_seen": 6349292,
|
| 65 |
+
"step": 70
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.020270805289413257,
|
| 69 |
+
"grad_norm": 0.22158554196357727,
|
| 70 |
+
"learning_rate": 2.5e-05,
|
| 71 |
+
"loss": 1.4942,
|
| 72 |
+
"num_input_tokens_seen": 7245840,
|
| 73 |
+
"step": 80
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 0.022804655950589912,
|
| 77 |
+
"grad_norm": 0.26374679803848267,
|
| 78 |
+
"learning_rate": 2.5e-05,
|
| 79 |
+
"loss": 1.4492,
|
| 80 |
+
"num_input_tokens_seen": 8171092,
|
| 81 |
+
"step": 90
|
| 82 |
+
},
|
| 83 |
+
{
|
| 84 |
+
"epoch": 0.025338506611766568,
|
| 85 |
+
"grad_norm": 0.23668645322322845,
|
| 86 |
+
"learning_rate": 2.5e-05,
|
| 87 |
+
"loss": 1.4677,
|
| 88 |
+
"num_input_tokens_seen": 9093156,
|
| 89 |
+
"step": 100
|
| 90 |
+
},
|
| 91 |
+
{
|
| 92 |
+
"epoch": 0.027872357272943227,
|
| 93 |
+
"grad_norm": 0.25576356053352356,
|
| 94 |
+
"learning_rate": 2.5e-05,
|
| 95 |
+
"loss": 1.4109,
|
| 96 |
+
"num_input_tokens_seen": 9976384,
|
| 97 |
+
"step": 110
|
| 98 |
+
},
|
| 99 |
+
{
|
| 100 |
+
"epoch": 0.030406207934119883,
|
| 101 |
+
"grad_norm": 0.2770518660545349,
|
| 102 |
+
"learning_rate": 2.5e-05,
|
| 103 |
+
"loss": 1.4304,
|
| 104 |
+
"num_input_tokens_seen": 10906092,
|
| 105 |
+
"step": 120
|
| 106 |
+
},
|
| 107 |
+
{
|
| 108 |
+
"epoch": 0.03294005859529654,
|
| 109 |
+
"grad_norm": 0.2333258092403412,
|
| 110 |
+
"learning_rate": 2.5e-05,
|
| 111 |
+
"loss": 1.4142,
|
| 112 |
+
"num_input_tokens_seen": 11818744,
|
| 113 |
+
"step": 130
|
| 114 |
+
},
|
| 115 |
+
{
|
| 116 |
+
"epoch": 0.035473909256473195,
|
| 117 |
+
"grad_norm": 0.24696557223796844,
|
| 118 |
+
"learning_rate": 2.5e-05,
|
| 119 |
+
"loss": 1.4246,
|
| 120 |
+
"num_input_tokens_seen": 12743780,
|
| 121 |
+
"step": 140
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.038007759917649854,
|
| 125 |
+
"grad_norm": 0.2408542037010193,
|
| 126 |
+
"learning_rate": 2.5e-05,
|
| 127 |
+
"loss": 1.4475,
|
| 128 |
+
"num_input_tokens_seen": 13674048,
|
| 129 |
+
"step": 150
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 0.04054161057882651,
|
| 133 |
+
"grad_norm": 0.2496064305305481,
|
| 134 |
+
"learning_rate": 2.5e-05,
|
| 135 |
+
"loss": 1.4529,
|
| 136 |
+
"num_input_tokens_seen": 14538524,
|
| 137 |
+
"step": 160
|
| 138 |
+
},
|
| 139 |
+
{
|
| 140 |
+
"epoch": 0.043075461240003166,
|
| 141 |
+
"grad_norm": 0.2827187478542328,
|
| 142 |
+
"learning_rate": 2.5e-05,
|
| 143 |
+
"loss": 1.4542,
|
| 144 |
+
"num_input_tokens_seen": 15470540,
|
| 145 |
+
"step": 170
|
| 146 |
+
},
|
| 147 |
+
{
|
| 148 |
+
"epoch": 0.045609311901179825,
|
| 149 |
+
"grad_norm": 0.25148963928222656,
|
| 150 |
+
"learning_rate": 2.5e-05,
|
| 151 |
+
"loss": 1.4361,
|
| 152 |
+
"num_input_tokens_seen": 16422976,
|
| 153 |
+
"step": 180
|
| 154 |
+
},
|
| 155 |
+
{
|
| 156 |
+
"epoch": 0.048143162562356484,
|
| 157 |
+
"grad_norm": 0.24195212125778198,
|
| 158 |
+
"learning_rate": 2.5e-05,
|
| 159 |
+
"loss": 1.4775,
|
| 160 |
+
"num_input_tokens_seen": 17344648,
|
| 161 |
+
"step": 190
|
| 162 |
+
},
|
| 163 |
+
{
|
| 164 |
+
"epoch": 0.050677013223533136,
|
| 165 |
+
"grad_norm": 0.3068198263645172,
|
| 166 |
+
"learning_rate": 2.5e-05,
|
| 167 |
+
"loss": 1.4231,
|
| 168 |
+
"num_input_tokens_seen": 18245048,
|
| 169 |
+
"step": 200
|
| 170 |
+
},
|
| 171 |
+
{
|
| 172 |
+
"epoch": 0.053210863884709796,
|
| 173 |
+
"grad_norm": 0.24267973005771637,
|
| 174 |
+
"learning_rate": 2.5e-05,
|
| 175 |
+
"loss": 1.4354,
|
| 176 |
+
"num_input_tokens_seen": 19169384,
|
| 177 |
+
"step": 210
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.055744714545886455,
|
| 181 |
+
"grad_norm": 0.21026775240898132,
|
| 182 |
+
"learning_rate": 2.5e-05,
|
| 183 |
+
"loss": 1.4153,
|
| 184 |
+
"num_input_tokens_seen": 20096992,
|
| 185 |
+
"step": 220
|
| 186 |
+
},
|
| 187 |
+
{
|
| 188 |
+
"epoch": 0.05827856520706311,
|
| 189 |
+
"grad_norm": 0.21877512335777283,
|
| 190 |
+
"learning_rate": 2.5e-05,
|
| 191 |
+
"loss": 1.3941,
|
| 192 |
+
"num_input_tokens_seen": 21025604,
|
| 193 |
+
"step": 230
|
| 194 |
+
},
|
| 195 |
+
{
|
| 196 |
+
"epoch": 0.060812415868239766,
|
| 197 |
+
"grad_norm": 0.24055704474449158,
|
| 198 |
+
"learning_rate": 2.5e-05,
|
| 199 |
+
"loss": 1.4348,
|
| 200 |
+
"num_input_tokens_seen": 21935180,
|
| 201 |
+
"step": 240
|
| 202 |
+
},
|
| 203 |
+
{
|
| 204 |
+
"epoch": 0.06334626652941643,
|
| 205 |
+
"grad_norm": 0.24673806130886078,
|
| 206 |
+
"learning_rate": 2.5e-05,
|
| 207 |
+
"loss": 1.3719,
|
| 208 |
+
"num_input_tokens_seen": 22857776,
|
| 209 |
+
"step": 250
|
| 210 |
+
},
|
| 211 |
+
{
|
| 212 |
+
"epoch": 0.06588011719059308,
|
| 213 |
+
"grad_norm": 0.21661491692066193,
|
| 214 |
+
"learning_rate": 2.5e-05,
|
| 215 |
+
"loss": 1.4707,
|
| 216 |
+
"num_input_tokens_seen": 23805840,
|
| 217 |
+
"step": 260
|
| 218 |
+
},
|
| 219 |
+
{
|
| 220 |
+
"epoch": 0.06841396785176973,
|
| 221 |
+
"grad_norm": 0.2766810357570648,
|
| 222 |
+
"learning_rate": 2.5e-05,
|
| 223 |
+
"loss": 1.4558,
|
| 224 |
+
"num_input_tokens_seen": 24694772,
|
| 225 |
+
"step": 270
|
| 226 |
+
},
|
| 227 |
+
{
|
| 228 |
+
"epoch": 0.07094781851294639,
|
| 229 |
+
"grad_norm": 0.2665688097476959,
|
| 230 |
+
"learning_rate": 2.5e-05,
|
| 231 |
+
"loss": 1.4026,
|
| 232 |
+
"num_input_tokens_seen": 25637024,
|
| 233 |
+
"step": 280
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 0.07348166917412305,
|
| 237 |
+
"grad_norm": 0.2424854040145874,
|
| 238 |
+
"learning_rate": 2.5e-05,
|
| 239 |
+
"loss": 1.3998,
|
| 240 |
+
"num_input_tokens_seen": 26530332,
|
| 241 |
+
"step": 290
|
| 242 |
+
},
|
| 243 |
+
{
|
| 244 |
+
"epoch": 0.07601551983529971,
|
| 245 |
+
"grad_norm": 0.23512804508209229,
|
| 246 |
+
"learning_rate": 2.5e-05,
|
| 247 |
+
"loss": 1.4188,
|
| 248 |
+
"num_input_tokens_seen": 27449020,
|
| 249 |
+
"step": 300
|
| 250 |
+
},
|
| 251 |
+
{
|
| 252 |
+
"epoch": 0.07854937049647637,
|
| 253 |
+
"grad_norm": 0.23620112240314484,
|
| 254 |
+
"learning_rate": 2.5e-05,
|
| 255 |
+
"loss": 1.39,
|
| 256 |
+
"num_input_tokens_seen": 28367404,
|
| 257 |
+
"step": 310
|
| 258 |
+
},
|
| 259 |
+
{
|
| 260 |
+
"epoch": 0.08108322115765303,
|
| 261 |
+
"grad_norm": 0.2523897588253021,
|
| 262 |
+
"learning_rate": 2.5e-05,
|
| 263 |
+
"loss": 1.4487,
|
| 264 |
+
"num_input_tokens_seen": 29277896,
|
| 265 |
+
"step": 320
|
| 266 |
+
},
|
| 267 |
+
{
|
| 268 |
+
"epoch": 0.08361707181882967,
|
| 269 |
+
"grad_norm": 0.24064438045024872,
|
| 270 |
+
"learning_rate": 2.5e-05,
|
| 271 |
+
"loss": 1.4214,
|
| 272 |
+
"num_input_tokens_seen": 30200812,
|
| 273 |
+
"step": 330
|
| 274 |
+
},
|
| 275 |
+
{
|
| 276 |
+
"epoch": 0.08615092248000633,
|
| 277 |
+
"grad_norm": 0.2440669983625412,
|
| 278 |
+
"learning_rate": 2.5e-05,
|
| 279 |
+
"loss": 1.444,
|
| 280 |
+
"num_input_tokens_seen": 31158760,
|
| 281 |
+
"step": 340
|
| 282 |
+
},
|
| 283 |
+
{
|
| 284 |
+
"epoch": 0.08868477314118299,
|
| 285 |
+
"grad_norm": 0.22009992599487305,
|
| 286 |
+
"learning_rate": 2.5e-05,
|
| 287 |
+
"loss": 1.4127,
|
| 288 |
+
"num_input_tokens_seen": 32069424,
|
| 289 |
+
"step": 350
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 0.09121862380235965,
|
| 293 |
+
"grad_norm": 0.29601845145225525,
|
| 294 |
+
"learning_rate": 2.5e-05,
|
| 295 |
+
"loss": 1.4437,
|
| 296 |
+
"num_input_tokens_seen": 33009436,
|
| 297 |
+
"step": 360
|
| 298 |
+
},
|
| 299 |
+
{
|
| 300 |
+
"epoch": 0.09375247446353631,
|
| 301 |
+
"grad_norm": 0.2240906059741974,
|
| 302 |
+
"learning_rate": 2.5e-05,
|
| 303 |
+
"loss": 1.3871,
|
| 304 |
+
"num_input_tokens_seen": 33933612,
|
| 305 |
+
"step": 370
|
| 306 |
+
},
|
| 307 |
+
{
|
| 308 |
+
"epoch": 0.09628632512471297,
|
| 309 |
+
"grad_norm": 0.23164159059524536,
|
| 310 |
+
"learning_rate": 2.5e-05,
|
| 311 |
+
"loss": 1.4149,
|
| 312 |
+
"num_input_tokens_seen": 34839560,
|
| 313 |
+
"step": 380
|
| 314 |
+
},
|
| 315 |
+
{
|
| 316 |
+
"epoch": 0.09882017578588961,
|
| 317 |
+
"grad_norm": 0.335622638463974,
|
| 318 |
+
"learning_rate": 2.5e-05,
|
| 319 |
+
"loss": 1.427,
|
| 320 |
+
"num_input_tokens_seen": 35748032,
|
| 321 |
+
"step": 390
|
| 322 |
+
},
|
| 323 |
+
{
|
| 324 |
+
"epoch": 0.10135402644706627,
|
| 325 |
+
"grad_norm": 0.22885636985301971,
|
| 326 |
+
"learning_rate": 2.5e-05,
|
| 327 |
+
"loss": 1.4521,
|
| 328 |
+
"num_input_tokens_seen": 36672280,
|
| 329 |
+
"step": 400
|
| 330 |
+
},
|
| 331 |
+
{
|
| 332 |
+
"epoch": 0.10388787710824293,
|
| 333 |
+
"grad_norm": 0.2555045783519745,
|
| 334 |
+
"learning_rate": 2.5e-05,
|
| 335 |
+
"loss": 1.4175,
|
| 336 |
+
"num_input_tokens_seen": 37599516,
|
| 337 |
+
"step": 410
|
| 338 |
+
},
|
| 339 |
+
{
|
| 340 |
+
"epoch": 0.10642172776941959,
|
| 341 |
+
"grad_norm": 0.24946229159832,
|
| 342 |
+
"learning_rate": 2.5e-05,
|
| 343 |
+
"loss": 1.4276,
|
| 344 |
+
"num_input_tokens_seen": 38529556,
|
| 345 |
+
"step": 420
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 0.10895557843059625,
|
| 349 |
+
"grad_norm": 0.24785666167736053,
|
| 350 |
+
"learning_rate": 2.5e-05,
|
| 351 |
+
"loss": 1.4007,
|
| 352 |
+
"num_input_tokens_seen": 39460044,
|
| 353 |
+
"step": 430
|
| 354 |
+
},
|
| 355 |
+
{
|
| 356 |
+
"epoch": 0.11148942909177291,
|
| 357 |
+
"grad_norm": 0.22006012499332428,
|
| 358 |
+
"learning_rate": 2.5e-05,
|
| 359 |
+
"loss": 1.4238,
|
| 360 |
+
"num_input_tokens_seen": 40369364,
|
| 361 |
+
"step": 440
|
| 362 |
+
},
|
| 363 |
+
{
|
| 364 |
+
"epoch": 0.11402327975294956,
|
| 365 |
+
"grad_norm": 0.26216018199920654,
|
| 366 |
+
"learning_rate": 2.5e-05,
|
| 367 |
+
"loss": 1.4318,
|
| 368 |
+
"num_input_tokens_seen": 41307640,
|
| 369 |
+
"step": 450
|
| 370 |
+
},
|
| 371 |
+
{
|
| 372 |
+
"epoch": 0.11655713041412621,
|
| 373 |
+
"grad_norm": 0.23494452238082886,
|
| 374 |
+
"learning_rate": 2.5e-05,
|
| 375 |
+
"loss": 1.417,
|
| 376 |
+
"num_input_tokens_seen": 42200280,
|
| 377 |
+
"step": 460
|
| 378 |
+
},
|
| 379 |
+
{
|
| 380 |
+
"epoch": 0.11909098107530287,
|
| 381 |
+
"grad_norm": 0.23429952561855316,
|
| 382 |
+
"learning_rate": 2.5e-05,
|
| 383 |
+
"loss": 1.4277,
|
| 384 |
+
"num_input_tokens_seen": 43112444,
|
| 385 |
+
"step": 470
|
| 386 |
+
},
|
| 387 |
+
{
|
| 388 |
+
"epoch": 0.12162483173647953,
|
| 389 |
+
"grad_norm": 0.2510409355163574,
|
| 390 |
+
"learning_rate": 2.5e-05,
|
| 391 |
+
"loss": 1.3853,
|
| 392 |
+
"num_input_tokens_seen": 44021860,
|
| 393 |
+
"step": 480
|
| 394 |
+
},
|
| 395 |
+
{
|
| 396 |
+
"epoch": 0.12415868239765619,
|
| 397 |
+
"grad_norm": 0.2570734918117523,
|
| 398 |
+
"learning_rate": 2.5e-05,
|
| 399 |
+
"loss": 1.4056,
|
| 400 |
+
"num_input_tokens_seen": 44938384,
|
| 401 |
+
"step": 490
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 0.12669253305883285,
|
| 405 |
+
"grad_norm": 0.23910905420780182,
|
| 406 |
+
"learning_rate": 2.5e-05,
|
| 407 |
+
"loss": 1.4303,
|
| 408 |
+
"num_input_tokens_seen": 45871544,
|
| 409 |
+
"step": 500
|
| 410 |
+
},
|
| 411 |
+
{
|
| 412 |
+
"epoch": 0.1292263837200095,
|
| 413 |
+
"grad_norm": 0.2258525788784027,
|
| 414 |
+
"learning_rate": 2.5e-05,
|
| 415 |
+
"loss": 1.4243,
|
| 416 |
+
"num_input_tokens_seen": 46798524,
|
| 417 |
+
"step": 510
|
| 418 |
+
},
|
| 419 |
+
{
|
| 420 |
+
"epoch": 0.13176023438118617,
|
| 421 |
+
"grad_norm": 0.21156556904315948,
|
| 422 |
+
"learning_rate": 2.5e-05,
|
| 423 |
+
"loss": 1.3965,
|
| 424 |
+
"num_input_tokens_seen": 47696192,
|
| 425 |
+
"step": 520
|
| 426 |
+
},
|
| 427 |
+
{
|
| 428 |
+
"epoch": 0.13429408504236282,
|
| 429 |
+
"grad_norm": 0.2665134370326996,
|
| 430 |
+
"learning_rate": 2.5e-05,
|
| 431 |
+
"loss": 1.393,
|
| 432 |
+
"num_input_tokens_seen": 48669228,
|
| 433 |
+
"step": 530
|
| 434 |
+
},
|
| 435 |
+
{
|
| 436 |
+
"epoch": 0.13682793570353946,
|
| 437 |
+
"grad_norm": 0.2551543414592743,
|
| 438 |
+
"learning_rate": 2.5e-05,
|
| 439 |
+
"loss": 1.4184,
|
| 440 |
+
"num_input_tokens_seen": 49616616,
|
| 441 |
+
"step": 540
|
| 442 |
+
},
|
| 443 |
+
{
|
| 444 |
+
"epoch": 0.13936178636471613,
|
| 445 |
+
"grad_norm": 0.2285103052854538,
|
| 446 |
+
"learning_rate": 2.5e-05,
|
| 447 |
+
"loss": 1.3964,
|
| 448 |
+
"num_input_tokens_seen": 50540636,
|
| 449 |
+
"step": 550
|
| 450 |
+
},
|
| 451 |
+
{
|
| 452 |
+
"epoch": 0.14189563702589278,
|
| 453 |
+
"grad_norm": 0.23576393723487854,
|
| 454 |
+
"learning_rate": 2.5e-05,
|
| 455 |
+
"loss": 1.4187,
|
| 456 |
+
"num_input_tokens_seen": 51440464,
|
| 457 |
+
"step": 560
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 0.14442948768706945,
|
| 461 |
+
"grad_norm": 0.22209148108959198,
|
| 462 |
+
"learning_rate": 2.5e-05,
|
| 463 |
+
"loss": 1.403,
|
| 464 |
+
"num_input_tokens_seen": 52315124,
|
| 465 |
+
"step": 570
|
| 466 |
+
},
|
| 467 |
+
{
|
| 468 |
+
"epoch": 0.1469633383482461,
|
| 469 |
+
"grad_norm": 0.23545274138450623,
|
| 470 |
+
"learning_rate": 2.5e-05,
|
| 471 |
+
"loss": 1.4313,
|
| 472 |
+
"num_input_tokens_seen": 53261804,
|
| 473 |
+
"step": 580
|
| 474 |
+
},
|
| 475 |
+
{
|
| 476 |
+
"epoch": 0.14949718900942277,
|
| 477 |
+
"grad_norm": 0.25153088569641113,
|
| 478 |
+
"learning_rate": 2.5e-05,
|
| 479 |
+
"loss": 1.3798,
|
| 480 |
+
"num_input_tokens_seen": 54106436,
|
| 481 |
+
"step": 590
|
| 482 |
+
},
|
| 483 |
+
{
|
| 484 |
+
"epoch": 0.15203103967059942,
|
| 485 |
+
"grad_norm": 0.23856191337108612,
|
| 486 |
+
"learning_rate": 2.5e-05,
|
| 487 |
+
"loss": 1.3679,
|
| 488 |
+
"num_input_tokens_seen": 55035052,
|
| 489 |
+
"step": 600
|
| 490 |
+
},
|
| 491 |
+
{
|
| 492 |
+
"epoch": 0.15456489033177606,
|
| 493 |
+
"grad_norm": 0.23667120933532715,
|
| 494 |
+
"learning_rate": 2.5e-05,
|
| 495 |
+
"loss": 1.4191,
|
| 496 |
+
"num_input_tokens_seen": 55935200,
|
| 497 |
+
"step": 610
|
| 498 |
+
},
|
| 499 |
+
{
|
| 500 |
+
"epoch": 0.15709874099295273,
|
| 501 |
+
"grad_norm": 0.26784512400627136,
|
| 502 |
+
"learning_rate": 2.5e-05,
|
| 503 |
+
"loss": 1.3684,
|
| 504 |
+
"num_input_tokens_seen": 56843340,
|
| 505 |
+
"step": 620
|
| 506 |
+
},
|
| 507 |
+
{
|
| 508 |
+
"epoch": 0.15963259165412938,
|
| 509 |
+
"grad_norm": 0.22612795233726501,
|
| 510 |
+
"learning_rate": 2.5e-05,
|
| 511 |
+
"loss": 1.436,
|
| 512 |
+
"num_input_tokens_seen": 57720808,
|
| 513 |
+
"step": 630
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 0.16216644231530605,
|
| 517 |
+
"grad_norm": 0.24946410953998566,
|
| 518 |
+
"learning_rate": 2.5e-05,
|
| 519 |
+
"loss": 1.426,
|
| 520 |
+
"num_input_tokens_seen": 58575924,
|
| 521 |
+
"step": 640
|
| 522 |
+
},
|
| 523 |
+
{
|
| 524 |
+
"epoch": 0.1647002929764827,
|
| 525 |
+
"grad_norm": 0.2528791129589081,
|
| 526 |
+
"learning_rate": 2.5e-05,
|
| 527 |
+
"loss": 1.4191,
|
| 528 |
+
"num_input_tokens_seen": 59484056,
|
| 529 |
+
"step": 650
|
| 530 |
+
},
|
| 531 |
+
{
|
| 532 |
+
"epoch": 0.16723414363765934,
|
| 533 |
+
"grad_norm": 0.21960842609405518,
|
| 534 |
+
"learning_rate": 2.5e-05,
|
| 535 |
+
"loss": 1.443,
|
| 536 |
+
"num_input_tokens_seen": 60382860,
|
| 537 |
+
"step": 660
|
| 538 |
+
},
|
| 539 |
+
{
|
| 540 |
+
"epoch": 0.16976799429883602,
|
| 541 |
+
"grad_norm": 0.2500540018081665,
|
| 542 |
+
"learning_rate": 2.5e-05,
|
| 543 |
+
"loss": 1.4284,
|
| 544 |
+
"num_input_tokens_seen": 61291764,
|
| 545 |
+
"step": 670
|
| 546 |
+
},
|
| 547 |
+
{
|
| 548 |
+
"epoch": 0.17230184496001266,
|
| 549 |
+
"grad_norm": 0.27140355110168457,
|
| 550 |
+
"learning_rate": 2.5e-05,
|
| 551 |
+
"loss": 1.3909,
|
| 552 |
+
"num_input_tokens_seen": 62183556,
|
| 553 |
+
"step": 680
|
| 554 |
+
},
|
| 555 |
+
{
|
| 556 |
+
"epoch": 0.17483569562118934,
|
| 557 |
+
"grad_norm": 0.22307205200195312,
|
| 558 |
+
"learning_rate": 2.5e-05,
|
| 559 |
+
"loss": 1.3682,
|
| 560 |
+
"num_input_tokens_seen": 63098340,
|
| 561 |
+
"step": 690
|
| 562 |
+
},
|
| 563 |
+
{
|
| 564 |
+
"epoch": 0.17736954628236598,
|
| 565 |
+
"grad_norm": 0.24494685232639313,
|
| 566 |
+
"learning_rate": 2.5e-05,
|
| 567 |
+
"loss": 1.3903,
|
| 568 |
+
"num_input_tokens_seen": 64000524,
|
| 569 |
+
"step": 700
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 0.17990339694354263,
|
| 573 |
+
"grad_norm": 0.2667907476425171,
|
| 574 |
+
"learning_rate": 2.5e-05,
|
| 575 |
+
"loss": 1.4569,
|
| 576 |
+
"num_input_tokens_seen": 64937424,
|
| 577 |
+
"step": 710
|
| 578 |
+
},
|
| 579 |
+
{
|
| 580 |
+
"epoch": 0.1824372476047193,
|
| 581 |
+
"grad_norm": 0.22164462506771088,
|
| 582 |
+
"learning_rate": 2.5e-05,
|
| 583 |
+
"loss": 1.3806,
|
| 584 |
+
"num_input_tokens_seen": 65822472,
|
| 585 |
+
"step": 720
|
| 586 |
+
},
|
| 587 |
+
{
|
| 588 |
+
"epoch": 0.18497109826589594,
|
| 589 |
+
"grad_norm": 0.23859019577503204,
|
| 590 |
+
"learning_rate": 2.5e-05,
|
| 591 |
+
"loss": 1.4003,
|
| 592 |
+
"num_input_tokens_seen": 66691752,
|
| 593 |
+
"step": 730
|
| 594 |
+
},
|
| 595 |
+
{
|
| 596 |
+
"epoch": 0.18750494892707262,
|
| 597 |
+
"grad_norm": 0.28847405314445496,
|
| 598 |
+
"learning_rate": 2.5e-05,
|
| 599 |
+
"loss": 1.4076,
|
| 600 |
+
"num_input_tokens_seen": 67658948,
|
| 601 |
+
"step": 740
|
| 602 |
+
},
|
| 603 |
+
{
|
| 604 |
+
"epoch": 0.19003879958824926,
|
| 605 |
+
"grad_norm": 0.2571374177932739,
|
| 606 |
+
"learning_rate": 2.5e-05,
|
| 607 |
+
"loss": 1.3924,
|
| 608 |
+
"num_input_tokens_seen": 68572048,
|
| 609 |
+
"step": 750
|
| 610 |
+
},
|
| 611 |
+
{
|
| 612 |
+
"epoch": 0.19257265024942594,
|
| 613 |
+
"grad_norm": 0.24991680681705475,
|
| 614 |
+
"learning_rate": 2.5e-05,
|
| 615 |
+
"loss": 1.4164,
|
| 616 |
+
"num_input_tokens_seen": 69502808,
|
| 617 |
+
"step": 760
|
| 618 |
+
},
|
| 619 |
+
{
|
| 620 |
+
"epoch": 0.19510650091060258,
|
| 621 |
+
"grad_norm": 0.23006725311279297,
|
| 622 |
+
"learning_rate": 2.5e-05,
|
| 623 |
+
"loss": 1.4019,
|
| 624 |
+
"num_input_tokens_seen": 70423124,
|
| 625 |
+
"step": 770
|
| 626 |
+
},
|
| 627 |
+
{
|
| 628 |
+
"epoch": 0.19764035157177923,
|
| 629 |
+
"grad_norm": 0.2484099566936493,
|
| 630 |
+
"learning_rate": 2.5e-05,
|
| 631 |
+
"loss": 1.3565,
|
| 632 |
+
"num_input_tokens_seen": 71271484,
|
| 633 |
+
"step": 780
|
| 634 |
+
},
|
| 635 |
+
{
|
| 636 |
+
"epoch": 0.2001742022329559,
|
| 637 |
+
"grad_norm": 0.2604601979255676,
|
| 638 |
+
"learning_rate": 2.5e-05,
|
| 639 |
+
"loss": 1.4098,
|
| 640 |
+
"num_input_tokens_seen": 72179604,
|
| 641 |
+
"step": 790
|
| 642 |
+
},
|
| 643 |
+
{
|
| 644 |
+
"epoch": 0.20270805289413255,
|
| 645 |
+
"grad_norm": 0.2681257724761963,
|
| 646 |
+
"learning_rate": 2.5e-05,
|
| 647 |
+
"loss": 1.4152,
|
| 648 |
+
"num_input_tokens_seen": 73085296,
|
| 649 |
+
"step": 800
|
| 650 |
+
},
|
| 651 |
+
{
|
| 652 |
+
"epoch": 0.20524190355530922,
|
| 653 |
+
"grad_norm": 0.20966367423534393,
|
| 654 |
+
"learning_rate": 2.5e-05,
|
| 655 |
+
"loss": 1.4006,
|
| 656 |
+
"num_input_tokens_seen": 74003640,
|
| 657 |
+
"step": 810
|
| 658 |
+
},
|
| 659 |
+
{
|
| 660 |
+
"epoch": 0.20777575421648586,
|
| 661 |
+
"grad_norm": 0.2371470183134079,
|
| 662 |
+
"learning_rate": 2.5e-05,
|
| 663 |
+
"loss": 1.3651,
|
| 664 |
+
"num_input_tokens_seen": 74957748,
|
| 665 |
+
"step": 820
|
| 666 |
+
},
|
| 667 |
+
{
|
| 668 |
+
"epoch": 0.2103096048776625,
|
| 669 |
+
"grad_norm": 0.24214884638786316,
|
| 670 |
+
"learning_rate": 2.5e-05,
|
| 671 |
+
"loss": 1.3969,
|
| 672 |
+
"num_input_tokens_seen": 75841664,
|
| 673 |
+
"step": 830
|
| 674 |
+
},
|
| 675 |
+
{
|
| 676 |
+
"epoch": 0.21284345553883918,
|
| 677 |
+
"grad_norm": 0.24258075654506683,
|
| 678 |
+
"learning_rate": 2.5e-05,
|
| 679 |
+
"loss": 1.4356,
|
| 680 |
+
"num_input_tokens_seen": 76765412,
|
| 681 |
+
"step": 840
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 0.21537730620001583,
|
| 685 |
+
"grad_norm": 0.25199827551841736,
|
| 686 |
+
"learning_rate": 2.5e-05,
|
| 687 |
+
"loss": 1.4171,
|
| 688 |
+
"num_input_tokens_seen": 77675892,
|
| 689 |
+
"step": 850
|
| 690 |
+
},
|
| 691 |
+
{
|
| 692 |
+
"epoch": 0.2179111568611925,
|
| 693 |
+
"grad_norm": 0.219390869140625,
|
| 694 |
+
"learning_rate": 2.5e-05,
|
| 695 |
+
"loss": 1.3713,
|
| 696 |
+
"num_input_tokens_seen": 78646236,
|
| 697 |
+
"step": 860
|
| 698 |
+
},
|
| 699 |
+
{
|
| 700 |
+
"epoch": 0.22044500752236915,
|
| 701 |
+
"grad_norm": 0.2546541690826416,
|
| 702 |
+
"learning_rate": 2.5e-05,
|
| 703 |
+
"loss": 1.4154,
|
| 704 |
+
"num_input_tokens_seen": 79594216,
|
| 705 |
+
"step": 870
|
| 706 |
+
},
|
| 707 |
+
{
|
| 708 |
+
"epoch": 0.22297885818354582,
|
| 709 |
+
"grad_norm": 0.28596746921539307,
|
| 710 |
+
"learning_rate": 2.5e-05,
|
| 711 |
+
"loss": 1.3981,
|
| 712 |
+
"num_input_tokens_seen": 80523804,
|
| 713 |
+
"step": 880
|
| 714 |
+
},
|
| 715 |
+
{
|
| 716 |
+
"epoch": 0.22551270884472246,
|
| 717 |
+
"grad_norm": 0.21436405181884766,
|
| 718 |
+
"learning_rate": 2.5e-05,
|
| 719 |
+
"loss": 1.3889,
|
| 720 |
+
"num_input_tokens_seen": 81405376,
|
| 721 |
+
"step": 890
|
| 722 |
+
},
|
| 723 |
+
{
|
| 724 |
+
"epoch": 0.2280465595058991,
|
| 725 |
+
"grad_norm": 0.2508715093135834,
|
| 726 |
+
"learning_rate": 2.5e-05,
|
| 727 |
+
"loss": 1.3682,
|
| 728 |
+
"num_input_tokens_seen": 82260336,
|
| 729 |
+
"step": 900
|
| 730 |
+
},
|
| 731 |
+
{
|
| 732 |
+
"epoch": 0.23058041016707578,
|
| 733 |
+
"grad_norm": 0.24959874153137207,
|
| 734 |
+
"learning_rate": 2.5e-05,
|
| 735 |
+
"loss": 1.3651,
|
| 736 |
+
"num_input_tokens_seen": 83190224,
|
| 737 |
+
"step": 910
|
| 738 |
+
},
|
| 739 |
+
{
|
| 740 |
+
"epoch": 0.23311426082825243,
|
| 741 |
+
"grad_norm": 0.27335524559020996,
|
| 742 |
+
"learning_rate": 2.5e-05,
|
| 743 |
+
"loss": 1.4221,
|
| 744 |
+
"num_input_tokens_seen": 84107372,
|
| 745 |
+
"step": 920
|
| 746 |
+
},
|
| 747 |
+
{
|
| 748 |
+
"epoch": 0.2356481114894291,
|
| 749 |
+
"grad_norm": 0.2550046443939209,
|
| 750 |
+
"learning_rate": 2.5e-05,
|
| 751 |
+
"loss": 1.4029,
|
| 752 |
+
"num_input_tokens_seen": 85024192,
|
| 753 |
+
"step": 930
|
| 754 |
+
},
|
| 755 |
+
{
|
| 756 |
+
"epoch": 0.23818196215060575,
|
| 757 |
+
"grad_norm": 0.23554718494415283,
|
| 758 |
+
"learning_rate": 2.5e-05,
|
| 759 |
+
"loss": 1.4045,
|
| 760 |
+
"num_input_tokens_seen": 85956220,
|
| 761 |
+
"step": 940
|
| 762 |
+
},
|
| 763 |
+
{
|
| 764 |
+
"epoch": 0.2407158128117824,
|
| 765 |
+
"grad_norm": 0.21662922203540802,
|
| 766 |
+
"learning_rate": 2.5e-05,
|
| 767 |
+
"loss": 1.3908,
|
| 768 |
+
"num_input_tokens_seen": 86858100,
|
| 769 |
+
"step": 950
|
| 770 |
+
},
|
| 771 |
+
{
|
| 772 |
+
"epoch": 0.24324966347295907,
|
| 773 |
+
"grad_norm": 0.22381572425365448,
|
| 774 |
+
"learning_rate": 2.5e-05,
|
| 775 |
+
"loss": 1.4337,
|
| 776 |
+
"num_input_tokens_seen": 87771400,
|
| 777 |
+
"step": 960
|
| 778 |
+
},
|
| 779 |
+
{
|
| 780 |
+
"epoch": 0.2457835141341357,
|
| 781 |
+
"grad_norm": 0.2680582106113434,
|
| 782 |
+
"learning_rate": 2.5e-05,
|
| 783 |
+
"loss": 1.4325,
|
| 784 |
+
"num_input_tokens_seen": 88675708,
|
| 785 |
+
"step": 970
|
| 786 |
+
},
|
| 787 |
+
{
|
| 788 |
+
"epoch": 0.24831736479531238,
|
| 789 |
+
"grad_norm": 0.22555038332939148,
|
| 790 |
+
"learning_rate": 2.5e-05,
|
| 791 |
+
"loss": 1.3741,
|
| 792 |
+
"num_input_tokens_seen": 89561964,
|
| 793 |
+
"step": 980
|
| 794 |
+
},
|
| 795 |
+
{
|
| 796 |
+
"epoch": 0.25085121545648903,
|
| 797 |
+
"grad_norm": 0.2812931537628174,
|
| 798 |
+
"learning_rate": 2.5e-05,
|
| 799 |
+
"loss": 1.4104,
|
| 800 |
+
"num_input_tokens_seen": 90488048,
|
| 801 |
+
"step": 990
|
| 802 |
+
},
|
| 803 |
+
{
|
| 804 |
+
"epoch": 0.2533850661176657,
|
| 805 |
+
"grad_norm": 0.23613446950912476,
|
| 806 |
+
"learning_rate": 2.5e-05,
|
| 807 |
+
"loss": 1.4008,
|
| 808 |
+
"num_input_tokens_seen": 91375832,
|
| 809 |
+
"step": 1000
|
| 810 |
+
},
|
| 811 |
+
{
|
| 812 |
+
"epoch": 0.2533850661176657,
|
| 813 |
+
"eval_loss": 1.4020060300827026,
|
| 814 |
+
"eval_runtime": 2.9465,
|
| 815 |
+
"eval_samples_per_second": 50.908,
|
| 816 |
+
"eval_steps_per_second": 6.448,
|
| 817 |
+
"num_input_tokens_seen": 91375832,
|
| 818 |
+
"step": 1000
|
| 819 |
+
},
|
| 820 |
+
{
|
| 821 |
+
"epoch": 0.2559189167788423,
|
| 822 |
+
"grad_norm": 0.2325298935174942,
|
| 823 |
+
"learning_rate": 2.5e-05,
|
| 824 |
+
"loss": 1.3544,
|
| 825 |
+
"num_input_tokens_seen": 92347240,
|
| 826 |
+
"step": 1010
|
| 827 |
+
},
|
| 828 |
+
{
|
| 829 |
+
"epoch": 0.258452767440019,
|
| 830 |
+
"grad_norm": 0.24142597615718842,
|
| 831 |
+
"learning_rate": 2.5e-05,
|
| 832 |
+
"loss": 1.3706,
|
| 833 |
+
"num_input_tokens_seen": 93237456,
|
| 834 |
+
"step": 1020
|
| 835 |
+
},
|
| 836 |
+
{
|
| 837 |
+
"epoch": 0.26098661810119567,
|
| 838 |
+
"grad_norm": 0.2356724739074707,
|
| 839 |
+
"learning_rate": 2.5e-05,
|
| 840 |
+
"loss": 1.3998,
|
| 841 |
+
"num_input_tokens_seen": 94145764,
|
| 842 |
+
"step": 1030
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"epoch": 0.26352046876237234,
|
| 846 |
+
"grad_norm": 0.243470698595047,
|
| 847 |
+
"learning_rate": 2.5e-05,
|
| 848 |
+
"loss": 1.4013,
|
| 849 |
+
"num_input_tokens_seen": 95055692,
|
| 850 |
+
"step": 1040
|
| 851 |
+
},
|
| 852 |
+
{
|
| 853 |
+
"epoch": 0.26605431942354896,
|
| 854 |
+
"grad_norm": 0.2412971556186676,
|
| 855 |
+
"learning_rate": 2.5e-05,
|
| 856 |
+
"loss": 1.373,
|
| 857 |
+
"num_input_tokens_seen": 95921656,
|
| 858 |
+
"step": 1050
|
| 859 |
+
},
|
| 860 |
+
{
|
| 861 |
+
"epoch": 0.26858817008472563,
|
| 862 |
+
"grad_norm": 0.2889567017555237,
|
| 863 |
+
"learning_rate": 2.5e-05,
|
| 864 |
+
"loss": 1.3898,
|
| 865 |
+
"num_input_tokens_seen": 96821452,
|
| 866 |
+
"step": 1060
|
| 867 |
+
},
|
| 868 |
+
{
|
| 869 |
+
"epoch": 0.2711220207459023,
|
| 870 |
+
"grad_norm": 0.23939931392669678,
|
| 871 |
+
"learning_rate": 2.5e-05,
|
| 872 |
+
"loss": 1.4088,
|
| 873 |
+
"num_input_tokens_seen": 97727612,
|
| 874 |
+
"step": 1070
|
| 875 |
+
},
|
| 876 |
+
{
|
| 877 |
+
"epoch": 0.2736558714070789,
|
| 878 |
+
"grad_norm": 0.25132742524147034,
|
| 879 |
+
"learning_rate": 2.5e-05,
|
| 880 |
+
"loss": 1.3853,
|
| 881 |
+
"num_input_tokens_seen": 98677952,
|
| 882 |
+
"step": 1080
|
| 883 |
+
},
|
| 884 |
+
{
|
| 885 |
+
"epoch": 0.2761897220682556,
|
| 886 |
+
"grad_norm": 0.2225540727376938,
|
| 887 |
+
"learning_rate": 2.5e-05,
|
| 888 |
+
"loss": 1.4041,
|
| 889 |
+
"num_input_tokens_seen": 99640748,
|
| 890 |
+
"step": 1090
|
| 891 |
+
},
|
| 892 |
+
{
|
| 893 |
+
"epoch": 0.27872357272943227,
|
| 894 |
+
"grad_norm": 0.24503560364246368,
|
| 895 |
+
"learning_rate": 2.5e-05,
|
| 896 |
+
"loss": 1.3719,
|
| 897 |
+
"num_input_tokens_seen": 100557008,
|
| 898 |
+
"step": 1100
|
| 899 |
+
},
|
| 900 |
+
{
|
| 901 |
+
"epoch": 0.28125742339060894,
|
| 902 |
+
"grad_norm": 0.2348717302083969,
|
| 903 |
+
"learning_rate": 2.5e-05,
|
| 904 |
+
"loss": 1.3937,
|
| 905 |
+
"num_input_tokens_seen": 101442164,
|
| 906 |
+
"step": 1110
|
| 907 |
+
},
|
| 908 |
+
{
|
| 909 |
+
"epoch": 0.28379127405178556,
|
| 910 |
+
"grad_norm": 0.24240590631961823,
|
| 911 |
+
"learning_rate": 2.5e-05,
|
| 912 |
+
"loss": 1.3641,
|
| 913 |
+
"num_input_tokens_seen": 102366056,
|
| 914 |
+
"step": 1120
|
| 915 |
+
},
|
| 916 |
+
{
|
| 917 |
+
"epoch": 0.28632512471296223,
|
| 918 |
+
"grad_norm": 0.2246118187904358,
|
| 919 |
+
"learning_rate": 2.5e-05,
|
| 920 |
+
"loss": 1.3631,
|
| 921 |
+
"num_input_tokens_seen": 103261480,
|
| 922 |
+
"step": 1130
|
| 923 |
+
},
|
| 924 |
+
{
|
| 925 |
+
"epoch": 0.2888589753741389,
|
| 926 |
+
"grad_norm": 0.2967662513256073,
|
| 927 |
+
"learning_rate": 2.5e-05,
|
| 928 |
+
"loss": 1.3883,
|
| 929 |
+
"num_input_tokens_seen": 104163484,
|
| 930 |
+
"step": 1140
|
| 931 |
+
},
|
| 932 |
+
{
|
| 933 |
+
"epoch": 0.2913928260353155,
|
| 934 |
+
"grad_norm": 0.24722802639007568,
|
| 935 |
+
"learning_rate": 2.5e-05,
|
| 936 |
+
"loss": 1.4444,
|
| 937 |
+
"num_input_tokens_seen": 105077064,
|
| 938 |
+
"step": 1150
|
| 939 |
+
},
|
| 940 |
+
{
|
| 941 |
+
"epoch": 0.2939266766964922,
|
| 942 |
+
"grad_norm": 0.2221587598323822,
|
| 943 |
+
"learning_rate": 2.5e-05,
|
| 944 |
+
"loss": 1.3809,
|
| 945 |
+
"num_input_tokens_seen": 105968728,
|
| 946 |
+
"step": 1160
|
| 947 |
+
},
|
| 948 |
+
{
|
| 949 |
+
"epoch": 0.29646052735766887,
|
| 950 |
+
"grad_norm": 0.23813994228839874,
|
| 951 |
+
"learning_rate": 2.5e-05,
|
| 952 |
+
"loss": 1.3941,
|
| 953 |
+
"num_input_tokens_seen": 106838388,
|
| 954 |
+
"step": 1170
|
| 955 |
+
},
|
| 956 |
+
{
|
| 957 |
+
"epoch": 0.29899437801884554,
|
| 958 |
+
"grad_norm": 0.24747894704341888,
|
| 959 |
+
"learning_rate": 2.5e-05,
|
| 960 |
+
"loss": 1.3812,
|
| 961 |
+
"num_input_tokens_seen": 107764200,
|
| 962 |
+
"step": 1180
|
| 963 |
+
},
|
| 964 |
+
{
|
| 965 |
+
"epoch": 0.30152822868002216,
|
| 966 |
+
"grad_norm": 0.26802727580070496,
|
| 967 |
+
"learning_rate": 2.5e-05,
|
| 968 |
+
"loss": 1.3717,
|
| 969 |
+
"num_input_tokens_seen": 108683176,
|
| 970 |
+
"step": 1190
|
| 971 |
+
},
|
| 972 |
+
{
|
| 973 |
+
"epoch": 0.30406207934119883,
|
| 974 |
+
"grad_norm": 0.27138280868530273,
|
| 975 |
+
"learning_rate": 2.5e-05,
|
| 976 |
+
"loss": 1.367,
|
| 977 |
+
"num_input_tokens_seen": 109606364,
|
| 978 |
+
"step": 1200
|
| 979 |
+
},
|
| 980 |
+
{
|
| 981 |
+
"epoch": 0.3065959300023755,
|
| 982 |
+
"grad_norm": 0.24378275871276855,
|
| 983 |
+
"learning_rate": 2.5e-05,
|
| 984 |
+
"loss": 1.3762,
|
| 985 |
+
"num_input_tokens_seen": 110518296,
|
| 986 |
+
"step": 1210
|
| 987 |
+
},
|
| 988 |
+
{
|
| 989 |
+
"epoch": 0.3091297806635521,
|
| 990 |
+
"grad_norm": 0.261106938123703,
|
| 991 |
+
"learning_rate": 2.5e-05,
|
| 992 |
+
"loss": 1.4227,
|
| 993 |
+
"num_input_tokens_seen": 111436828,
|
| 994 |
+
"step": 1220
|
| 995 |
+
},
|
| 996 |
+
{
|
| 997 |
+
"epoch": 0.3116636313247288,
|
| 998 |
+
"grad_norm": 0.2597008943557739,
|
| 999 |
+
"learning_rate": 2.5e-05,
|
| 1000 |
+
"loss": 1.3847,
|
| 1001 |
+
"num_input_tokens_seen": 112334112,
|
| 1002 |
+
"step": 1230
|
| 1003 |
+
},
|
| 1004 |
+
{
|
| 1005 |
+
"epoch": 0.31419748198590547,
|
| 1006 |
+
"grad_norm": 0.24535202980041504,
|
| 1007 |
+
"learning_rate": 2.5e-05,
|
| 1008 |
+
"loss": 1.3706,
|
| 1009 |
+
"num_input_tokens_seen": 113211652,
|
| 1010 |
+
"step": 1240
|
| 1011 |
+
},
|
| 1012 |
+
{
|
| 1013 |
+
"epoch": 0.3167313326470821,
|
| 1014 |
+
"grad_norm": 0.2770673632621765,
|
| 1015 |
+
"learning_rate": 2.5e-05,
|
| 1016 |
+
"loss": 1.3975,
|
| 1017 |
+
"num_input_tokens_seen": 114117744,
|
| 1018 |
+
"step": 1250
|
| 1019 |
+
},
|
| 1020 |
+
{
|
| 1021 |
+
"epoch": 0.31926518330825876,
|
| 1022 |
+
"grad_norm": 0.21976234018802643,
|
| 1023 |
+
"learning_rate": 2.5e-05,
|
| 1024 |
+
"loss": 1.4036,
|
| 1025 |
+
"num_input_tokens_seen": 115002568,
|
| 1026 |
+
"step": 1260
|
| 1027 |
+
},
|
| 1028 |
+
{
|
| 1029 |
+
"epoch": 0.32179903396943543,
|
| 1030 |
+
"grad_norm": 0.22749099135398865,
|
| 1031 |
+
"learning_rate": 2.5e-05,
|
| 1032 |
+
"loss": 1.3625,
|
| 1033 |
+
"num_input_tokens_seen": 115904964,
|
| 1034 |
+
"step": 1270
|
| 1035 |
+
},
|
| 1036 |
+
{
|
| 1037 |
+
"epoch": 0.3243328846306121,
|
| 1038 |
+
"grad_norm": 0.22470030188560486,
|
| 1039 |
+
"learning_rate": 2.5e-05,
|
| 1040 |
+
"loss": 1.3905,
|
| 1041 |
+
"num_input_tokens_seen": 116843732,
|
| 1042 |
+
"step": 1280
|
| 1043 |
+
},
|
| 1044 |
+
{
|
| 1045 |
+
"epoch": 0.3268667352917887,
|
| 1046 |
+
"grad_norm": 0.2671917974948883,
|
| 1047 |
+
"learning_rate": 2.5e-05,
|
| 1048 |
+
"loss": 1.3839,
|
| 1049 |
+
"num_input_tokens_seen": 117752200,
|
| 1050 |
+
"step": 1290
|
| 1051 |
+
},
|
| 1052 |
+
{
|
| 1053 |
+
"epoch": 0.3294005859529654,
|
| 1054 |
+
"grad_norm": 0.24347306787967682,
|
| 1055 |
+
"learning_rate": 2.5e-05,
|
| 1056 |
+
"loss": 1.36,
|
| 1057 |
+
"num_input_tokens_seen": 118656912,
|
| 1058 |
+
"step": 1300
|
| 1059 |
+
},
|
| 1060 |
+
{
|
| 1061 |
+
"epoch": 0.33193443661414207,
|
| 1062 |
+
"grad_norm": 0.22786876559257507,
|
| 1063 |
+
"learning_rate": 2.5e-05,
|
| 1064 |
+
"loss": 1.361,
|
| 1065 |
+
"num_input_tokens_seen": 119561700,
|
| 1066 |
+
"step": 1310
|
| 1067 |
+
},
|
| 1068 |
+
{
|
| 1069 |
+
"epoch": 0.3344682872753187,
|
| 1070 |
+
"grad_norm": 0.22891202569007874,
|
| 1071 |
+
"learning_rate": 2.5e-05,
|
| 1072 |
+
"loss": 1.3916,
|
| 1073 |
+
"num_input_tokens_seen": 120537120,
|
| 1074 |
+
"step": 1320
|
| 1075 |
+
},
|
| 1076 |
+
{
|
| 1077 |
+
"epoch": 0.33700213793649536,
|
| 1078 |
+
"grad_norm": 0.2579503357410431,
|
| 1079 |
+
"learning_rate": 2.5e-05,
|
| 1080 |
+
"loss": 1.4077,
|
| 1081 |
+
"num_input_tokens_seen": 121473416,
|
| 1082 |
+
"step": 1330
|
| 1083 |
+
},
|
| 1084 |
+
{
|
| 1085 |
+
"epoch": 0.33953598859767203,
|
| 1086 |
+
"grad_norm": 0.24670307338237762,
|
| 1087 |
+
"learning_rate": 2.5e-05,
|
| 1088 |
+
"loss": 1.4055,
|
| 1089 |
+
"num_input_tokens_seen": 122383356,
|
| 1090 |
+
"step": 1340
|
| 1091 |
+
},
|
| 1092 |
+
{
|
| 1093 |
+
"epoch": 0.3420698392588487,
|
| 1094 |
+
"grad_norm": 0.2923058569431305,
|
| 1095 |
+
"learning_rate": 2.5e-05,
|
| 1096 |
+
"loss": 1.3875,
|
| 1097 |
+
"num_input_tokens_seen": 123309020,
|
| 1098 |
+
"step": 1350
|
| 1099 |
+
},
|
| 1100 |
+
{
|
| 1101 |
+
"epoch": 0.3446036899200253,
|
| 1102 |
+
"grad_norm": 0.2256019562482834,
|
| 1103 |
+
"learning_rate": 2.5e-05,
|
| 1104 |
+
"loss": 1.3872,
|
| 1105 |
+
"num_input_tokens_seen": 124234924,
|
| 1106 |
+
"step": 1360
|
| 1107 |
+
},
|
| 1108 |
+
{
|
| 1109 |
+
"epoch": 0.347137540581202,
|
| 1110 |
+
"grad_norm": 0.2368822544813156,
|
| 1111 |
+
"learning_rate": 2.5e-05,
|
| 1112 |
+
"loss": 1.3969,
|
| 1113 |
+
"num_input_tokens_seen": 125162100,
|
| 1114 |
+
"step": 1370
|
| 1115 |
+
},
|
| 1116 |
+
{
|
| 1117 |
+
"epoch": 0.34967139124237867,
|
| 1118 |
+
"grad_norm": 0.2430727332830429,
|
| 1119 |
+
"learning_rate": 2.5e-05,
|
| 1120 |
+
"loss": 1.3638,
|
| 1121 |
+
"num_input_tokens_seen": 126113704,
|
| 1122 |
+
"step": 1380
|
| 1123 |
+
},
|
| 1124 |
+
{
|
| 1125 |
+
"epoch": 0.3522052419035553,
|
| 1126 |
+
"grad_norm": 0.23543952405452728,
|
| 1127 |
+
"learning_rate": 2.5e-05,
|
| 1128 |
+
"loss": 1.3642,
|
| 1129 |
+
"num_input_tokens_seen": 127052976,
|
| 1130 |
+
"step": 1390
|
| 1131 |
+
},
|
| 1132 |
+
{
|
| 1133 |
+
"epoch": 0.35473909256473196,
|
| 1134 |
+
"grad_norm": 0.24988651275634766,
|
| 1135 |
+
"learning_rate": 2.5e-05,
|
| 1136 |
+
"loss": 1.3784,
|
| 1137 |
+
"num_input_tokens_seen": 127996892,
|
| 1138 |
+
"step": 1400
|
| 1139 |
+
},
|
| 1140 |
+
{
|
| 1141 |
+
"epoch": 0.35727294322590863,
|
| 1142 |
+
"grad_norm": 0.2787221670150757,
|
| 1143 |
+
"learning_rate": 2.5e-05,
|
| 1144 |
+
"loss": 1.4052,
|
| 1145 |
+
"num_input_tokens_seen": 128935380,
|
| 1146 |
+
"step": 1410
|
| 1147 |
+
},
|
| 1148 |
+
{
|
| 1149 |
+
"epoch": 0.35980679388708525,
|
| 1150 |
+
"grad_norm": 0.24997858703136444,
|
| 1151 |
+
"learning_rate": 2.5e-05,
|
| 1152 |
+
"loss": 1.3791,
|
| 1153 |
+
"num_input_tokens_seen": 129871964,
|
| 1154 |
+
"step": 1420
|
| 1155 |
+
},
|
| 1156 |
+
{
|
| 1157 |
+
"epoch": 0.3623406445482619,
|
| 1158 |
+
"grad_norm": 0.24547652900218964,
|
| 1159 |
+
"learning_rate": 2.5e-05,
|
| 1160 |
+
"loss": 1.395,
|
| 1161 |
+
"num_input_tokens_seen": 130767084,
|
| 1162 |
+
"step": 1430
|
| 1163 |
+
},
|
| 1164 |
+
{
|
| 1165 |
+
"epoch": 0.3648744952094386,
|
| 1166 |
+
"grad_norm": 0.23068061470985413,
|
| 1167 |
+
"learning_rate": 2.5e-05,
|
| 1168 |
+
"loss": 1.3677,
|
| 1169 |
+
"num_input_tokens_seen": 131674508,
|
| 1170 |
+
"step": 1440
|
| 1171 |
+
},
|
| 1172 |
+
{
|
| 1173 |
+
"epoch": 0.36740834587061527,
|
| 1174 |
+
"grad_norm": 0.23524820804595947,
|
| 1175 |
+
"learning_rate": 2.5e-05,
|
| 1176 |
+
"loss": 1.4161,
|
| 1177 |
+
"num_input_tokens_seen": 132602416,
|
| 1178 |
+
"step": 1450
|
| 1179 |
+
},
|
| 1180 |
+
{
|
| 1181 |
+
"epoch": 0.3699421965317919,
|
| 1182 |
+
"grad_norm": 0.23469901084899902,
|
| 1183 |
+
"learning_rate": 2.5e-05,
|
| 1184 |
+
"loss": 1.3721,
|
| 1185 |
+
"num_input_tokens_seen": 133506196,
|
| 1186 |
+
"step": 1460
|
| 1187 |
+
},
|
| 1188 |
+
{
|
| 1189 |
+
"epoch": 0.37247604719296856,
|
| 1190 |
+
"grad_norm": 0.24987129867076874,
|
| 1191 |
+
"learning_rate": 2.5e-05,
|
| 1192 |
+
"loss": 1.4049,
|
| 1193 |
+
"num_input_tokens_seen": 134427152,
|
| 1194 |
+
"step": 1470
|
| 1195 |
+
},
|
| 1196 |
+
{
|
| 1197 |
+
"epoch": 0.37500989785414524,
|
| 1198 |
+
"grad_norm": 0.24462181329727173,
|
| 1199 |
+
"learning_rate": 2.5e-05,
|
| 1200 |
+
"loss": 1.3797,
|
| 1201 |
+
"num_input_tokens_seen": 135314244,
|
| 1202 |
+
"step": 1480
|
| 1203 |
+
},
|
| 1204 |
+
{
|
| 1205 |
+
"epoch": 0.37754374851532185,
|
| 1206 |
+
"grad_norm": 0.2653500437736511,
|
| 1207 |
+
"learning_rate": 2.5e-05,
|
| 1208 |
+
"loss": 1.3503,
|
| 1209 |
+
"num_input_tokens_seen": 136230948,
|
| 1210 |
+
"step": 1490
|
| 1211 |
+
},
|
| 1212 |
+
{
|
| 1213 |
+
"epoch": 0.3800775991764985,
|
| 1214 |
+
"grad_norm": 0.2400883287191391,
|
| 1215 |
+
"learning_rate": 2.5e-05,
|
| 1216 |
+
"loss": 1.3957,
|
| 1217 |
+
"num_input_tokens_seen": 137179452,
|
| 1218 |
+
"step": 1500
|
| 1219 |
+
},
|
| 1220 |
+
{
|
| 1221 |
+
"epoch": 0.3826114498376752,
|
| 1222 |
+
"grad_norm": 0.2289241999387741,
|
| 1223 |
+
"learning_rate": 2.5e-05,
|
| 1224 |
+
"loss": 1.3529,
|
| 1225 |
+
"num_input_tokens_seen": 138078404,
|
| 1226 |
+
"step": 1510
|
| 1227 |
+
},
|
| 1228 |
+
{
|
| 1229 |
+
"epoch": 0.3851453004988519,
|
| 1230 |
+
"grad_norm": 0.26289331912994385,
|
| 1231 |
+
"learning_rate": 2.5e-05,
|
| 1232 |
+
"loss": 1.4398,
|
| 1233 |
+
"num_input_tokens_seen": 138991724,
|
| 1234 |
+
"step": 1520
|
| 1235 |
+
},
|
| 1236 |
+
{
|
| 1237 |
+
"epoch": 0.3876791511600285,
|
| 1238 |
+
"grad_norm": 0.2165287286043167,
|
| 1239 |
+
"learning_rate": 2.5e-05,
|
| 1240 |
+
"loss": 1.41,
|
| 1241 |
+
"num_input_tokens_seen": 139933240,
|
| 1242 |
+
"step": 1530
|
| 1243 |
+
},
|
| 1244 |
+
{
|
| 1245 |
+
"epoch": 0.39021300182120516,
|
| 1246 |
+
"grad_norm": 0.29837462306022644,
|
| 1247 |
+
"learning_rate": 2.5e-05,
|
| 1248 |
+
"loss": 1.3662,
|
| 1249 |
+
"num_input_tokens_seen": 140836772,
|
| 1250 |
+
"step": 1540
|
| 1251 |
+
},
|
| 1252 |
+
{
|
| 1253 |
+
"epoch": 0.39274685248238184,
|
| 1254 |
+
"grad_norm": 0.24651922285556793,
|
| 1255 |
+
"learning_rate": 2.5e-05,
|
| 1256 |
+
"loss": 1.3412,
|
| 1257 |
+
"num_input_tokens_seen": 141744576,
|
| 1258 |
+
"step": 1550
|
| 1259 |
+
},
|
| 1260 |
+
{
|
| 1261 |
+
"epoch": 0.39528070314355845,
|
| 1262 |
+
"grad_norm": 0.29952993988990784,
|
| 1263 |
+
"learning_rate": 2.5e-05,
|
| 1264 |
+
"loss": 1.3907,
|
| 1265 |
+
"num_input_tokens_seen": 142624188,
|
| 1266 |
+
"step": 1560
|
| 1267 |
+
},
|
| 1268 |
+
{
|
| 1269 |
+
"epoch": 0.3978145538047351,
|
| 1270 |
+
"grad_norm": 0.2563650608062744,
|
| 1271 |
+
"learning_rate": 2.5e-05,
|
| 1272 |
+
"loss": 1.3858,
|
| 1273 |
+
"num_input_tokens_seen": 143554872,
|
| 1274 |
+
"step": 1570
|
| 1275 |
+
},
|
| 1276 |
+
{
|
| 1277 |
+
"epoch": 0.4003484044659118,
|
| 1278 |
+
"grad_norm": 0.2565977871417999,
|
| 1279 |
+
"learning_rate": 2.5e-05,
|
| 1280 |
+
"loss": 1.3732,
|
| 1281 |
+
"num_input_tokens_seen": 144477588,
|
| 1282 |
+
"step": 1580
|
| 1283 |
+
},
|
| 1284 |
+
{
|
| 1285 |
+
"epoch": 0.4028822551270885,
|
| 1286 |
+
"grad_norm": 0.2879079282283783,
|
| 1287 |
+
"learning_rate": 2.5e-05,
|
| 1288 |
+
"loss": 1.3692,
|
| 1289 |
+
"num_input_tokens_seen": 145354620,
|
| 1290 |
+
"step": 1590
|
| 1291 |
+
},
|
| 1292 |
+
{
|
| 1293 |
+
"epoch": 0.4054161057882651,
|
| 1294 |
+
"grad_norm": 0.2640700936317444,
|
| 1295 |
+
"learning_rate": 2.5e-05,
|
| 1296 |
+
"loss": 1.3909,
|
| 1297 |
+
"num_input_tokens_seen": 146266280,
|
| 1298 |
+
"step": 1600
|
| 1299 |
+
},
|
| 1300 |
+
{
|
| 1301 |
+
"epoch": 0.40794995644944176,
|
| 1302 |
+
"grad_norm": 0.26872700452804565,
|
| 1303 |
+
"learning_rate": 2.5e-05,
|
| 1304 |
+
"loss": 1.4051,
|
| 1305 |
+
"num_input_tokens_seen": 147165620,
|
| 1306 |
+
"step": 1610
|
| 1307 |
+
},
|
| 1308 |
+
{
|
| 1309 |
+
"epoch": 0.41048380711061844,
|
| 1310 |
+
"grad_norm": 0.2187357246875763,
|
| 1311 |
+
"learning_rate": 2.5e-05,
|
| 1312 |
+
"loss": 1.38,
|
| 1313 |
+
"num_input_tokens_seen": 148098344,
|
| 1314 |
+
"step": 1620
|
| 1315 |
+
},
|
| 1316 |
+
{
|
| 1317 |
+
"epoch": 0.41301765777179505,
|
| 1318 |
+
"grad_norm": 0.24293020367622375,
|
| 1319 |
+
"learning_rate": 2.5e-05,
|
| 1320 |
+
"loss": 1.3915,
|
| 1321 |
+
"num_input_tokens_seen": 149043924,
|
| 1322 |
+
"step": 1630
|
| 1323 |
+
},
|
| 1324 |
+
{
|
| 1325 |
+
"epoch": 0.41555150843297173,
|
| 1326 |
+
"grad_norm": 0.23092688620090485,
|
| 1327 |
+
"learning_rate": 2.5e-05,
|
| 1328 |
+
"loss": 1.4019,
|
| 1329 |
+
"num_input_tokens_seen": 149996036,
|
| 1330 |
+
"step": 1640
|
| 1331 |
+
},
|
| 1332 |
+
{
|
| 1333 |
+
"epoch": 0.4180853590941484,
|
| 1334 |
+
"grad_norm": 0.27063265442848206,
|
| 1335 |
+
"learning_rate": 2.5e-05,
|
| 1336 |
+
"loss": 1.3723,
|
| 1337 |
+
"num_input_tokens_seen": 150869152,
|
| 1338 |
+
"step": 1650
|
| 1339 |
+
},
|
| 1340 |
+
{
|
| 1341 |
+
"epoch": 0.420619209755325,
|
| 1342 |
+
"grad_norm": 0.25822359323501587,
|
| 1343 |
+
"learning_rate": 2.5e-05,
|
| 1344 |
+
"loss": 1.3682,
|
| 1345 |
+
"num_input_tokens_seen": 151783488,
|
| 1346 |
+
"step": 1660
|
| 1347 |
+
},
|
| 1348 |
+
{
|
| 1349 |
+
"epoch": 0.4231530604165017,
|
| 1350 |
+
"grad_norm": 0.269724041223526,
|
| 1351 |
+
"learning_rate": 2.5e-05,
|
| 1352 |
+
"loss": 1.3592,
|
| 1353 |
+
"num_input_tokens_seen": 152700960,
|
| 1354 |
+
"step": 1670
|
| 1355 |
+
},
|
| 1356 |
+
{
|
| 1357 |
+
"epoch": 0.42568691107767836,
|
| 1358 |
+
"grad_norm": 0.23563367128372192,
|
| 1359 |
+
"learning_rate": 2.5e-05,
|
| 1360 |
+
"loss": 1.3679,
|
| 1361 |
+
"num_input_tokens_seen": 153634040,
|
| 1362 |
+
"step": 1680
|
| 1363 |
+
},
|
| 1364 |
+
{
|
| 1365 |
+
"epoch": 0.42822076173885504,
|
| 1366 |
+
"grad_norm": 0.23306426405906677,
|
| 1367 |
+
"learning_rate": 2.5e-05,
|
| 1368 |
+
"loss": 1.3585,
|
| 1369 |
+
"num_input_tokens_seen": 154569656,
|
| 1370 |
+
"step": 1690
|
| 1371 |
+
},
|
| 1372 |
+
{
|
| 1373 |
+
"epoch": 0.43075461240003166,
|
| 1374 |
+
"grad_norm": 0.23761169612407684,
|
| 1375 |
+
"learning_rate": 2.5e-05,
|
| 1376 |
+
"loss": 1.3413,
|
| 1377 |
+
"num_input_tokens_seen": 155491724,
|
| 1378 |
+
"step": 1700
|
| 1379 |
+
},
|
| 1380 |
+
{
|
| 1381 |
+
"epoch": 0.43328846306120833,
|
| 1382 |
+
"grad_norm": 0.23138809204101562,
|
| 1383 |
+
"learning_rate": 2.5e-05,
|
| 1384 |
+
"loss": 1.3892,
|
| 1385 |
+
"num_input_tokens_seen": 156437340,
|
| 1386 |
+
"step": 1710
|
| 1387 |
+
},
|
| 1388 |
+
{
|
| 1389 |
+
"epoch": 0.435822313722385,
|
| 1390 |
+
"grad_norm": 0.24864792823791504,
|
| 1391 |
+
"learning_rate": 2.5e-05,
|
| 1392 |
+
"loss": 1.387,
|
| 1393 |
+
"num_input_tokens_seen": 157343056,
|
| 1394 |
+
"step": 1720
|
| 1395 |
+
},
|
| 1396 |
+
{
|
| 1397 |
+
"epoch": 0.4383561643835616,
|
| 1398 |
+
"grad_norm": 0.24503816664218903,
|
| 1399 |
+
"learning_rate": 2.5e-05,
|
| 1400 |
+
"loss": 1.3544,
|
| 1401 |
+
"num_input_tokens_seen": 158211084,
|
| 1402 |
+
"step": 1730
|
| 1403 |
+
},
|
| 1404 |
+
{
|
| 1405 |
+
"epoch": 0.4408900150447383,
|
| 1406 |
+
"grad_norm": 0.23860155045986176,
|
| 1407 |
+
"learning_rate": 2.5e-05,
|
| 1408 |
+
"loss": 1.3947,
|
| 1409 |
+
"num_input_tokens_seen": 159127644,
|
| 1410 |
+
"step": 1740
|
| 1411 |
+
},
|
| 1412 |
+
{
|
| 1413 |
+
"epoch": 0.44342386570591497,
|
| 1414 |
+
"grad_norm": 0.23359131813049316,
|
| 1415 |
+
"learning_rate": 2.5e-05,
|
| 1416 |
+
"loss": 1.3333,
|
| 1417 |
+
"num_input_tokens_seen": 160056144,
|
| 1418 |
+
"step": 1750
|
| 1419 |
+
},
|
| 1420 |
+
{
|
| 1421 |
+
"epoch": 0.44595771636709164,
|
| 1422 |
+
"grad_norm": 0.23289762437343597,
|
| 1423 |
+
"learning_rate": 2.5e-05,
|
| 1424 |
+
"loss": 1.4039,
|
| 1425 |
+
"num_input_tokens_seen": 161001352,
|
| 1426 |
+
"step": 1760
|
| 1427 |
+
},
|
| 1428 |
+
{
|
| 1429 |
+
"epoch": 0.44849156702826826,
|
| 1430 |
+
"grad_norm": 0.23038776218891144,
|
| 1431 |
+
"learning_rate": 2.5e-05,
|
| 1432 |
+
"loss": 1.3872,
|
| 1433 |
+
"num_input_tokens_seen": 161931048,
|
| 1434 |
+
"step": 1770
|
| 1435 |
+
},
|
| 1436 |
+
{
|
| 1437 |
+
"epoch": 0.45102541768944493,
|
| 1438 |
+
"grad_norm": 0.26440566778182983,
|
| 1439 |
+
"learning_rate": 2.5e-05,
|
| 1440 |
+
"loss": 1.372,
|
| 1441 |
+
"num_input_tokens_seen": 162861292,
|
| 1442 |
+
"step": 1780
|
| 1443 |
+
},
|
| 1444 |
+
{
|
| 1445 |
+
"epoch": 0.4535592683506216,
|
| 1446 |
+
"grad_norm": 0.2498098909854889,
|
| 1447 |
+
"learning_rate": 2.5e-05,
|
| 1448 |
+
"loss": 1.3287,
|
| 1449 |
+
"num_input_tokens_seen": 163797388,
|
| 1450 |
+
"step": 1790
|
| 1451 |
+
},
|
| 1452 |
+
{
|
| 1453 |
+
"epoch": 0.4560931190117982,
|
| 1454 |
+
"grad_norm": 0.2095261961221695,
|
| 1455 |
+
"learning_rate": 2.5e-05,
|
| 1456 |
+
"loss": 1.3778,
|
| 1457 |
+
"num_input_tokens_seen": 164671840,
|
| 1458 |
+
"step": 1800
|
| 1459 |
+
},
|
| 1460 |
+
{
|
| 1461 |
+
"epoch": 0.4586269696729749,
|
| 1462 |
+
"grad_norm": 0.2577464282512665,
|
| 1463 |
+
"learning_rate": 2.5e-05,
|
| 1464 |
+
"loss": 1.3821,
|
| 1465 |
+
"num_input_tokens_seen": 165619284,
|
| 1466 |
+
"step": 1810
|
| 1467 |
+
},
|
| 1468 |
+
{
|
| 1469 |
+
"epoch": 0.46116082033415157,
|
| 1470 |
+
"grad_norm": 0.23324383795261383,
|
| 1471 |
+
"learning_rate": 2.5e-05,
|
| 1472 |
+
"loss": 1.3921,
|
| 1473 |
+
"num_input_tokens_seen": 166521872,
|
| 1474 |
+
"step": 1820
|
| 1475 |
+
},
|
| 1476 |
+
{
|
| 1477 |
+
"epoch": 0.46369467099532824,
|
| 1478 |
+
"grad_norm": 0.23413369059562683,
|
| 1479 |
+
"learning_rate": 2.5e-05,
|
| 1480 |
+
"loss": 1.391,
|
| 1481 |
+
"num_input_tokens_seen": 167446436,
|
| 1482 |
+
"step": 1830
|
| 1483 |
+
},
|
| 1484 |
+
{
|
| 1485 |
+
"epoch": 0.46622852165650486,
|
| 1486 |
+
"grad_norm": 0.2720430791378021,
|
| 1487 |
+
"learning_rate": 2.5e-05,
|
| 1488 |
+
"loss": 1.359,
|
| 1489 |
+
"num_input_tokens_seen": 168356260,
|
| 1490 |
+
"step": 1840
|
| 1491 |
+
},
|
| 1492 |
+
{
|
| 1493 |
+
"epoch": 0.46876237231768153,
|
| 1494 |
+
"grad_norm": 0.2760706841945648,
|
| 1495 |
+
"learning_rate": 2.5e-05,
|
| 1496 |
+
"loss": 1.3498,
|
| 1497 |
+
"num_input_tokens_seen": 169262844,
|
| 1498 |
+
"step": 1850
|
| 1499 |
+
},
|
| 1500 |
+
{
|
| 1501 |
+
"epoch": 0.4712962229788582,
|
| 1502 |
+
"grad_norm": 0.27992355823516846,
|
| 1503 |
+
"learning_rate": 2.5e-05,
|
| 1504 |
+
"loss": 1.3984,
|
| 1505 |
+
"num_input_tokens_seen": 170164272,
|
| 1506 |
+
"step": 1860
|
| 1507 |
+
},
|
| 1508 |
+
{
|
| 1509 |
+
"epoch": 0.4738300736400348,
|
| 1510 |
+
"grad_norm": 0.23402582108974457,
|
| 1511 |
+
"learning_rate": 2.5e-05,
|
| 1512 |
+
"loss": 1.3667,
|
| 1513 |
+
"num_input_tokens_seen": 171067864,
|
| 1514 |
+
"step": 1870
|
| 1515 |
+
},
|
| 1516 |
+
{
|
| 1517 |
+
"epoch": 0.4763639243012115,
|
| 1518 |
+
"grad_norm": 0.29928284883499146,
|
| 1519 |
+
"learning_rate": 2.5e-05,
|
| 1520 |
+
"loss": 1.335,
|
| 1521 |
+
"num_input_tokens_seen": 172005232,
|
| 1522 |
+
"step": 1880
|
| 1523 |
+
},
|
| 1524 |
+
{
|
| 1525 |
+
"epoch": 0.47889777496238817,
|
| 1526 |
+
"grad_norm": 0.25357866287231445,
|
| 1527 |
+
"learning_rate": 2.5e-05,
|
| 1528 |
+
"loss": 1.3802,
|
| 1529 |
+
"num_input_tokens_seen": 172915708,
|
| 1530 |
+
"step": 1890
|
| 1531 |
+
},
|
| 1532 |
+
{
|
| 1533 |
+
"epoch": 0.4814316256235648,
|
| 1534 |
+
"grad_norm": 0.29246291518211365,
|
| 1535 |
+
"learning_rate": 2.5e-05,
|
| 1536 |
+
"loss": 1.3513,
|
| 1537 |
+
"num_input_tokens_seen": 173820476,
|
| 1538 |
+
"step": 1900
|
| 1539 |
+
},
|
| 1540 |
+
{
|
| 1541 |
+
"epoch": 0.48396547628474146,
|
| 1542 |
+
"grad_norm": 0.2792080342769623,
|
| 1543 |
+
"learning_rate": 2.5e-05,
|
| 1544 |
+
"loss": 1.3939,
|
| 1545 |
+
"num_input_tokens_seen": 174740920,
|
| 1546 |
+
"step": 1910
|
| 1547 |
+
},
|
| 1548 |
+
{
|
| 1549 |
+
"epoch": 0.48649932694591813,
|
| 1550 |
+
"grad_norm": 0.3099055588245392,
|
| 1551 |
+
"learning_rate": 2.5e-05,
|
| 1552 |
+
"loss": 1.3693,
|
| 1553 |
+
"num_input_tokens_seen": 175635720,
|
| 1554 |
+
"step": 1920
|
| 1555 |
+
},
|
| 1556 |
+
{
|
| 1557 |
+
"epoch": 0.4890331776070948,
|
| 1558 |
+
"grad_norm": 0.2375776320695877,
|
| 1559 |
+
"learning_rate": 2.5e-05,
|
| 1560 |
+
"loss": 1.3829,
|
| 1561 |
+
"num_input_tokens_seen": 176538688,
|
| 1562 |
+
"step": 1930
|
| 1563 |
+
},
|
| 1564 |
+
{
|
| 1565 |
+
"epoch": 0.4915670282682714,
|
| 1566 |
+
"grad_norm": 0.2295093983411789,
|
| 1567 |
+
"learning_rate": 2.5e-05,
|
| 1568 |
+
"loss": 1.3691,
|
| 1569 |
+
"num_input_tokens_seen": 177468420,
|
| 1570 |
+
"step": 1940
|
| 1571 |
+
},
|
| 1572 |
+
{
|
| 1573 |
+
"epoch": 0.4941008789294481,
|
| 1574 |
+
"grad_norm": 0.21639369428157806,
|
| 1575 |
+
"learning_rate": 2.5e-05,
|
| 1576 |
+
"loss": 1.3509,
|
| 1577 |
+
"num_input_tokens_seen": 178388296,
|
| 1578 |
+
"step": 1950
|
| 1579 |
+
},
|
| 1580 |
+
{
|
| 1581 |
+
"epoch": 0.49663472959062477,
|
| 1582 |
+
"grad_norm": 0.26756080985069275,
|
| 1583 |
+
"learning_rate": 2.5e-05,
|
| 1584 |
+
"loss": 1.3761,
|
| 1585 |
+
"num_input_tokens_seen": 179341380,
|
| 1586 |
+
"step": 1960
|
| 1587 |
+
},
|
| 1588 |
+
{
|
| 1589 |
+
"epoch": 0.4991685802518014,
|
| 1590 |
+
"grad_norm": 0.21319729089736938,
|
| 1591 |
+
"learning_rate": 2.5e-05,
|
| 1592 |
+
"loss": 1.3803,
|
| 1593 |
+
"num_input_tokens_seen": 180256564,
|
| 1594 |
+
"step": 1970
|
| 1595 |
+
},
|
| 1596 |
+
{
|
| 1597 |
+
"epoch": 0.5017024309129781,
|
| 1598 |
+
"grad_norm": 0.2565974295139313,
|
| 1599 |
+
"learning_rate": 2.5e-05,
|
| 1600 |
+
"loss": 1.3763,
|
| 1601 |
+
"num_input_tokens_seen": 181117020,
|
| 1602 |
+
"step": 1980
|
| 1603 |
+
},
|
| 1604 |
+
{
|
| 1605 |
+
"epoch": 0.5042362815741547,
|
| 1606 |
+
"grad_norm": 0.30257830023765564,
|
| 1607 |
+
"learning_rate": 2.5e-05,
|
| 1608 |
+
"loss": 1.3671,
|
| 1609 |
+
"num_input_tokens_seen": 182027528,
|
| 1610 |
+
"step": 1990
|
| 1611 |
+
},
|
| 1612 |
+
{
|
| 1613 |
+
"epoch": 0.5067701322353314,
|
| 1614 |
+
"grad_norm": 0.23474013805389404,
|
| 1615 |
+
"learning_rate": 2.5e-05,
|
| 1616 |
+
"loss": 1.3456,
|
| 1617 |
+
"num_input_tokens_seen": 182939052,
|
| 1618 |
+
"step": 2000
|
| 1619 |
+
},
|
| 1620 |
+
{
|
| 1621 |
+
"epoch": 0.5067701322353314,
|
| 1622 |
+
"eval_loss": 1.3669419288635254,
|
| 1623 |
+
"eval_runtime": 2.8409,
|
| 1624 |
+
"eval_samples_per_second": 52.801,
|
| 1625 |
+
"eval_steps_per_second": 6.688,
|
| 1626 |
+
"num_input_tokens_seen": 182939052,
|
| 1627 |
+
"step": 2000
|
| 1628 |
+
},
|
| 1629 |
+
{
|
| 1630 |
+
"epoch": 0.509303982896508,
|
| 1631 |
+
"grad_norm": 0.2144283950328827,
|
| 1632 |
+
"learning_rate": 2.5e-05,
|
| 1633 |
+
"loss": 1.37,
|
| 1634 |
+
"num_input_tokens_seen": 183841188,
|
| 1635 |
+
"step": 2010
|
| 1636 |
+
},
|
| 1637 |
+
{
|
| 1638 |
+
"epoch": 0.5118378335576846,
|
| 1639 |
+
"grad_norm": 0.2299591451883316,
|
| 1640 |
+
"learning_rate": 2.5e-05,
|
| 1641 |
+
"loss": 1.3436,
|
| 1642 |
+
"num_input_tokens_seen": 184804372,
|
| 1643 |
+
"step": 2020
|
| 1644 |
+
},
|
| 1645 |
+
{
|
| 1646 |
+
"epoch": 0.5143716842188614,
|
| 1647 |
+
"grad_norm": 0.2291470170021057,
|
| 1648 |
+
"learning_rate": 2.5e-05,
|
| 1649 |
+
"loss": 1.38,
|
| 1650 |
+
"num_input_tokens_seen": 185696628,
|
| 1651 |
+
"step": 2030
|
| 1652 |
+
},
|
| 1653 |
+
{
|
| 1654 |
+
"epoch": 0.516905534880038,
|
| 1655 |
+
"grad_norm": 0.25624164938926697,
|
| 1656 |
+
"learning_rate": 2.5e-05,
|
| 1657 |
+
"loss": 1.3741,
|
| 1658 |
+
"num_input_tokens_seen": 186584108,
|
| 1659 |
+
"step": 2040
|
| 1660 |
+
},
|
| 1661 |
+
{
|
| 1662 |
+
"epoch": 0.5194393855412147,
|
| 1663 |
+
"grad_norm": 0.2826102077960968,
|
| 1664 |
+
"learning_rate": 2.5e-05,
|
| 1665 |
+
"loss": 1.3786,
|
| 1666 |
+
"num_input_tokens_seen": 187491532,
|
| 1667 |
+
"step": 2050
|
| 1668 |
+
},
|
| 1669 |
+
{
|
| 1670 |
+
"epoch": 0.5219732362023913,
|
| 1671 |
+
"grad_norm": 0.23644354939460754,
|
| 1672 |
+
"learning_rate": 2.5e-05,
|
| 1673 |
+
"loss": 1.3119,
|
| 1674 |
+
"num_input_tokens_seen": 188398308,
|
| 1675 |
+
"step": 2060
|
| 1676 |
+
},
|
| 1677 |
+
{
|
| 1678 |
+
"epoch": 0.524507086863568,
|
| 1679 |
+
"grad_norm": 0.2631579041481018,
|
| 1680 |
+
"learning_rate": 2.5e-05,
|
| 1681 |
+
"loss": 1.3596,
|
| 1682 |
+
"num_input_tokens_seen": 189270772,
|
| 1683 |
+
"step": 2070
|
| 1684 |
+
},
|
| 1685 |
+
{
|
| 1686 |
+
"epoch": 0.5270409375247447,
|
| 1687 |
+
"grad_norm": 0.24663548171520233,
|
| 1688 |
+
"learning_rate": 2.5e-05,
|
| 1689 |
+
"loss": 1.3833,
|
| 1690 |
+
"num_input_tokens_seen": 190188192,
|
| 1691 |
+
"step": 2080
|
| 1692 |
+
},
|
| 1693 |
+
{
|
| 1694 |
+
"epoch": 0.5295747881859213,
|
| 1695 |
+
"grad_norm": 0.21753673255443573,
|
| 1696 |
+
"learning_rate": 2.5e-05,
|
| 1697 |
+
"loss": 1.3746,
|
| 1698 |
+
"num_input_tokens_seen": 191125784,
|
| 1699 |
+
"step": 2090
|
| 1700 |
+
},
|
| 1701 |
+
{
|
| 1702 |
+
"epoch": 0.5321086388470979,
|
| 1703 |
+
"grad_norm": 0.2312672883272171,
|
| 1704 |
+
"learning_rate": 2.5e-05,
|
| 1705 |
+
"loss": 1.3558,
|
| 1706 |
+
"num_input_tokens_seen": 192010984,
|
| 1707 |
+
"step": 2100
|
| 1708 |
+
},
|
| 1709 |
+
{
|
| 1710 |
+
"epoch": 0.5346424895082746,
|
| 1711 |
+
"grad_norm": 0.2641030251979828,
|
| 1712 |
+
"learning_rate": 2.5e-05,
|
| 1713 |
+
"loss": 1.3436,
|
| 1714 |
+
"num_input_tokens_seen": 192947832,
|
| 1715 |
+
"step": 2110
|
| 1716 |
+
},
|
| 1717 |
+
{
|
| 1718 |
+
"epoch": 0.5371763401694513,
|
| 1719 |
+
"grad_norm": 0.2314285784959793,
|
| 1720 |
+
"learning_rate": 2.5e-05,
|
| 1721 |
+
"loss": 1.3889,
|
| 1722 |
+
"num_input_tokens_seen": 193836096,
|
| 1723 |
+
"step": 2120
|
| 1724 |
+
},
|
| 1725 |
+
{
|
| 1726 |
+
"epoch": 0.5397101908306279,
|
| 1727 |
+
"grad_norm": 0.2117050439119339,
|
| 1728 |
+
"learning_rate": 2.5e-05,
|
| 1729 |
+
"loss": 1.3636,
|
| 1730 |
+
"num_input_tokens_seen": 194752188,
|
| 1731 |
+
"step": 2130
|
| 1732 |
+
},
|
| 1733 |
+
{
|
| 1734 |
+
"epoch": 0.5422440414918046,
|
| 1735 |
+
"grad_norm": 0.24790892004966736,
|
| 1736 |
+
"learning_rate": 2.5e-05,
|
| 1737 |
+
"loss": 1.3577,
|
| 1738 |
+
"num_input_tokens_seen": 195659416,
|
| 1739 |
+
"step": 2140
|
| 1740 |
+
},
|
| 1741 |
+
{
|
| 1742 |
+
"epoch": 0.5447778921529812,
|
| 1743 |
+
"grad_norm": 0.253757119178772,
|
| 1744 |
+
"learning_rate": 2.5e-05,
|
| 1745 |
+
"loss": 1.3767,
|
| 1746 |
+
"num_input_tokens_seen": 196584176,
|
| 1747 |
+
"step": 2150
|
| 1748 |
+
},
|
| 1749 |
+
{
|
| 1750 |
+
"epoch": 0.5473117428141578,
|
| 1751 |
+
"grad_norm": 0.2629224359989166,
|
| 1752 |
+
"learning_rate": 2.5e-05,
|
| 1753 |
+
"loss": 1.3771,
|
| 1754 |
+
"num_input_tokens_seen": 197456816,
|
| 1755 |
+
"step": 2160
|
| 1756 |
+
},
|
| 1757 |
+
{
|
| 1758 |
+
"epoch": 0.5498455934753346,
|
| 1759 |
+
"grad_norm": 0.2274072915315628,
|
| 1760 |
+
"learning_rate": 2.5e-05,
|
| 1761 |
+
"loss": 1.3633,
|
| 1762 |
+
"num_input_tokens_seen": 198358444,
|
| 1763 |
+
"step": 2170
|
| 1764 |
+
},
|
| 1765 |
+
{
|
| 1766 |
+
"epoch": 0.5523794441365112,
|
| 1767 |
+
"grad_norm": 0.2630630135536194,
|
| 1768 |
+
"learning_rate": 2.5e-05,
|
| 1769 |
+
"loss": 1.3702,
|
| 1770 |
+
"num_input_tokens_seen": 199246040,
|
| 1771 |
+
"step": 2180
|
| 1772 |
+
},
|
| 1773 |
+
{
|
| 1774 |
+
"epoch": 0.5549132947976878,
|
| 1775 |
+
"grad_norm": 0.24167053401470184,
|
| 1776 |
+
"learning_rate": 2.5e-05,
|
| 1777 |
+
"loss": 1.3785,
|
| 1778 |
+
"num_input_tokens_seen": 200167412,
|
| 1779 |
+
"step": 2190
|
| 1780 |
+
},
|
| 1781 |
+
{
|
| 1782 |
+
"epoch": 0.5574471454588645,
|
| 1783 |
+
"grad_norm": 0.2560918927192688,
|
| 1784 |
+
"learning_rate": 2.5e-05,
|
| 1785 |
+
"loss": 1.3757,
|
| 1786 |
+
"num_input_tokens_seen": 201090512,
|
| 1787 |
+
"step": 2200
|
| 1788 |
+
},
|
| 1789 |
+
{
|
| 1790 |
+
"epoch": 0.5599809961200412,
|
| 1791 |
+
"grad_norm": 0.23884332180023193,
|
| 1792 |
+
"learning_rate": 2.5e-05,
|
| 1793 |
+
"loss": 1.3642,
|
| 1794 |
+
"num_input_tokens_seen": 202070196,
|
| 1795 |
+
"step": 2210
|
| 1796 |
+
},
|
| 1797 |
+
{
|
| 1798 |
+
"epoch": 0.5625148467812179,
|
| 1799 |
+
"grad_norm": 0.25141972303390503,
|
| 1800 |
+
"learning_rate": 2.5e-05,
|
| 1801 |
+
"loss": 1.3669,
|
| 1802 |
+
"num_input_tokens_seen": 203015232,
|
| 1803 |
+
"step": 2220
|
| 1804 |
+
},
|
| 1805 |
+
{
|
| 1806 |
+
"epoch": 0.5650486974423945,
|
| 1807 |
+
"grad_norm": 0.20563028752803802,
|
| 1808 |
+
"learning_rate": 2.5e-05,
|
| 1809 |
+
"loss": 1.3622,
|
| 1810 |
+
"num_input_tokens_seen": 203955992,
|
| 1811 |
+
"step": 2230
|
| 1812 |
+
},
|
| 1813 |
+
{
|
| 1814 |
+
"epoch": 0.5675825481035711,
|
| 1815 |
+
"grad_norm": 0.26771050691604614,
|
| 1816 |
+
"learning_rate": 2.5e-05,
|
| 1817 |
+
"loss": 1.3551,
|
| 1818 |
+
"num_input_tokens_seen": 204867084,
|
| 1819 |
+
"step": 2240
|
| 1820 |
+
},
|
| 1821 |
+
{
|
| 1822 |
+
"epoch": 0.5701163987647478,
|
| 1823 |
+
"grad_norm": 0.2185191512107849,
|
| 1824 |
+
"learning_rate": 2.5e-05,
|
| 1825 |
+
"loss": 1.3574,
|
| 1826 |
+
"num_input_tokens_seen": 205818444,
|
| 1827 |
+
"step": 2250
|
| 1828 |
+
},
|
| 1829 |
+
{
|
| 1830 |
+
"epoch": 0.5726502494259245,
|
| 1831 |
+
"grad_norm": 0.23736274242401123,
|
| 1832 |
+
"learning_rate": 2.5e-05,
|
| 1833 |
+
"loss": 1.3478,
|
| 1834 |
+
"num_input_tokens_seen": 206727340,
|
| 1835 |
+
"step": 2260
|
| 1836 |
+
},
|
| 1837 |
+
{
|
| 1838 |
+
"epoch": 0.5751841000871011,
|
| 1839 |
+
"grad_norm": 0.2208438366651535,
|
| 1840 |
+
"learning_rate": 2.5e-05,
|
| 1841 |
+
"loss": 1.3576,
|
| 1842 |
+
"num_input_tokens_seen": 207682956,
|
| 1843 |
+
"step": 2270
|
| 1844 |
+
},
|
| 1845 |
+
{
|
| 1846 |
+
"epoch": 0.5777179507482778,
|
| 1847 |
+
"grad_norm": 0.215751051902771,
|
| 1848 |
+
"learning_rate": 2.5e-05,
|
| 1849 |
+
"loss": 1.3105,
|
| 1850 |
+
"num_input_tokens_seen": 208613224,
|
| 1851 |
+
"step": 2280
|
| 1852 |
+
},
|
| 1853 |
+
{
|
| 1854 |
+
"epoch": 0.5802518014094544,
|
| 1855 |
+
"grad_norm": 0.24414047598838806,
|
| 1856 |
+
"learning_rate": 2.5e-05,
|
| 1857 |
+
"loss": 1.3637,
|
| 1858 |
+
"num_input_tokens_seen": 209480700,
|
| 1859 |
+
"step": 2290
|
| 1860 |
+
},
|
| 1861 |
+
{
|
| 1862 |
+
"epoch": 0.582785652070631,
|
| 1863 |
+
"grad_norm": 0.27234476804733276,
|
| 1864 |
+
"learning_rate": 2.5e-05,
|
| 1865 |
+
"loss": 1.3648,
|
| 1866 |
+
"num_input_tokens_seen": 210380616,
|
| 1867 |
+
"step": 2300
|
| 1868 |
+
},
|
| 1869 |
+
{
|
| 1870 |
+
"epoch": 0.5853195027318078,
|
| 1871 |
+
"grad_norm": 0.23880694806575775,
|
| 1872 |
+
"learning_rate": 2.5e-05,
|
| 1873 |
+
"loss": 1.3452,
|
| 1874 |
+
"num_input_tokens_seen": 211323472,
|
| 1875 |
+
"step": 2310
|
| 1876 |
+
},
|
| 1877 |
+
{
|
| 1878 |
+
"epoch": 0.5878533533929844,
|
| 1879 |
+
"grad_norm": 0.24618738889694214,
|
| 1880 |
+
"learning_rate": 2.5e-05,
|
| 1881 |
+
"loss": 1.3357,
|
| 1882 |
+
"num_input_tokens_seen": 212269424,
|
| 1883 |
+
"step": 2320
|
| 1884 |
+
},
|
| 1885 |
+
{
|
| 1886 |
+
"epoch": 0.590387204054161,
|
| 1887 |
+
"grad_norm": 0.2280731499195099,
|
| 1888 |
+
"learning_rate": 2.5e-05,
|
| 1889 |
+
"loss": 1.3808,
|
| 1890 |
+
"num_input_tokens_seen": 213236052,
|
| 1891 |
+
"step": 2330
|
| 1892 |
+
},
|
| 1893 |
+
{
|
| 1894 |
+
"epoch": 0.5929210547153377,
|
| 1895 |
+
"grad_norm": 0.2641889452934265,
|
| 1896 |
+
"learning_rate": 2.5e-05,
|
| 1897 |
+
"loss": 1.3635,
|
| 1898 |
+
"num_input_tokens_seen": 214193180,
|
| 1899 |
+
"step": 2340
|
| 1900 |
+
},
|
| 1901 |
+
{
|
| 1902 |
+
"epoch": 0.5954549053765144,
|
| 1903 |
+
"grad_norm": 0.24398839473724365,
|
| 1904 |
+
"learning_rate": 2.5e-05,
|
| 1905 |
+
"loss": 1.3157,
|
| 1906 |
+
"num_input_tokens_seen": 215145888,
|
| 1907 |
+
"step": 2350
|
| 1908 |
+
},
|
| 1909 |
+
{
|
| 1910 |
+
"epoch": 0.5979887560376911,
|
| 1911 |
+
"grad_norm": 0.29194214940071106,
|
| 1912 |
+
"learning_rate": 2.5e-05,
|
| 1913 |
+
"loss": 1.3809,
|
| 1914 |
+
"num_input_tokens_seen": 216076328,
|
| 1915 |
+
"step": 2360
|
| 1916 |
+
},
|
| 1917 |
+
{
|
| 1918 |
+
"epoch": 0.6005226066988677,
|
| 1919 |
+
"grad_norm": 0.23668240010738373,
|
| 1920 |
+
"learning_rate": 2.5e-05,
|
| 1921 |
+
"loss": 1.3723,
|
| 1922 |
+
"num_input_tokens_seen": 216957792,
|
| 1923 |
+
"step": 2370
|
| 1924 |
+
},
|
| 1925 |
+
{
|
| 1926 |
+
"epoch": 0.6030564573600443,
|
| 1927 |
+
"grad_norm": 0.2053728848695755,
|
| 1928 |
+
"learning_rate": 2.5e-05,
|
| 1929 |
+
"loss": 1.3106,
|
| 1930 |
+
"num_input_tokens_seen": 217923088,
|
| 1931 |
+
"step": 2380
|
| 1932 |
+
},
|
| 1933 |
+
{
|
| 1934 |
+
"epoch": 0.605590308021221,
|
| 1935 |
+
"grad_norm": 0.2571648061275482,
|
| 1936 |
+
"learning_rate": 2.5e-05,
|
| 1937 |
+
"loss": 1.3636,
|
| 1938 |
+
"num_input_tokens_seen": 218831976,
|
| 1939 |
+
"step": 2390
|
| 1940 |
+
},
|
| 1941 |
+
{
|
| 1942 |
+
"epoch": 0.6081241586823977,
|
| 1943 |
+
"grad_norm": 0.25352680683135986,
|
| 1944 |
+
"learning_rate": 2.5e-05,
|
| 1945 |
+
"loss": 1.3448,
|
| 1946 |
+
"num_input_tokens_seen": 219756636,
|
| 1947 |
+
"step": 2400
|
| 1948 |
+
},
|
| 1949 |
+
{
|
| 1950 |
+
"epoch": 0.6106580093435743,
|
| 1951 |
+
"grad_norm": 0.23342467844486237,
|
| 1952 |
+
"learning_rate": 2.5e-05,
|
| 1953 |
+
"loss": 1.3908,
|
| 1954 |
+
"num_input_tokens_seen": 220660172,
|
| 1955 |
+
"step": 2410
|
| 1956 |
+
},
|
| 1957 |
+
{
|
| 1958 |
+
"epoch": 0.613191860004751,
|
| 1959 |
+
"grad_norm": 0.24378784000873566,
|
| 1960 |
+
"learning_rate": 2.5e-05,
|
| 1961 |
+
"loss": 1.3631,
|
| 1962 |
+
"num_input_tokens_seen": 221559444,
|
| 1963 |
+
"step": 2420
|
| 1964 |
+
},
|
| 1965 |
+
{
|
| 1966 |
+
"epoch": 0.6157257106659276,
|
| 1967 |
+
"grad_norm": 0.23902441561222076,
|
| 1968 |
+
"learning_rate": 2.5e-05,
|
| 1969 |
+
"loss": 1.3389,
|
| 1970 |
+
"num_input_tokens_seen": 222484304,
|
| 1971 |
+
"step": 2430
|
| 1972 |
+
},
|
| 1973 |
+
{
|
| 1974 |
+
"epoch": 0.6182595613271042,
|
| 1975 |
+
"grad_norm": 0.24430356919765472,
|
| 1976 |
+
"learning_rate": 2.5e-05,
|
| 1977 |
+
"loss": 1.3741,
|
| 1978 |
+
"num_input_tokens_seen": 223424636,
|
| 1979 |
+
"step": 2440
|
| 1980 |
+
},
|
| 1981 |
+
{
|
| 1982 |
+
"epoch": 0.620793411988281,
|
| 1983 |
+
"grad_norm": 0.22024385631084442,
|
| 1984 |
+
"learning_rate": 2.5e-05,
|
| 1985 |
+
"loss": 1.3173,
|
| 1986 |
+
"num_input_tokens_seen": 224336328,
|
| 1987 |
+
"step": 2450
|
| 1988 |
+
},
|
| 1989 |
+
{
|
| 1990 |
+
"epoch": 0.6233272626494576,
|
| 1991 |
+
"grad_norm": 0.2540358304977417,
|
| 1992 |
+
"learning_rate": 2.5e-05,
|
| 1993 |
+
"loss": 1.3551,
|
| 1994 |
+
"num_input_tokens_seen": 225268812,
|
| 1995 |
+
"step": 2460
|
| 1996 |
+
},
|
| 1997 |
+
{
|
| 1998 |
+
"epoch": 0.6258611133106342,
|
| 1999 |
+
"grad_norm": 0.30823466181755066,
|
| 2000 |
+
"learning_rate": 2.5e-05,
|
| 2001 |
+
"loss": 1.3315,
|
| 2002 |
+
"num_input_tokens_seen": 226203392,
|
| 2003 |
+
"step": 2470
|
| 2004 |
+
},
|
| 2005 |
+
{
|
| 2006 |
+
"epoch": 0.6283949639718109,
|
| 2007 |
+
"grad_norm": 0.22996842861175537,
|
| 2008 |
+
"learning_rate": 2.5e-05,
|
| 2009 |
+
"loss": 1.3647,
|
| 2010 |
+
"num_input_tokens_seen": 227073928,
|
| 2011 |
+
"step": 2480
|
| 2012 |
+
},
|
| 2013 |
+
{
|
| 2014 |
+
"epoch": 0.6309288146329876,
|
| 2015 |
+
"grad_norm": 0.22297543287277222,
|
| 2016 |
+
"learning_rate": 2.5e-05,
|
| 2017 |
+
"loss": 1.3673,
|
| 2018 |
+
"num_input_tokens_seen": 227988144,
|
| 2019 |
+
"step": 2490
|
| 2020 |
+
},
|
| 2021 |
+
{
|
| 2022 |
+
"epoch": 0.6334626652941642,
|
| 2023 |
+
"grad_norm": 0.2600548267364502,
|
| 2024 |
+
"learning_rate": 2.5e-05,
|
| 2025 |
+
"loss": 1.3417,
|
| 2026 |
+
"num_input_tokens_seen": 228908304,
|
| 2027 |
+
"step": 2500
|
| 2028 |
+
},
|
| 2029 |
+
{
|
| 2030 |
+
"epoch": 0.6359965159553409,
|
| 2031 |
+
"grad_norm": 0.27056604623794556,
|
| 2032 |
+
"learning_rate": 2.5e-05,
|
| 2033 |
+
"loss": 1.2998,
|
| 2034 |
+
"num_input_tokens_seen": 229859596,
|
| 2035 |
+
"step": 2510
|
| 2036 |
+
},
|
| 2037 |
+
{
|
| 2038 |
+
"epoch": 0.6385303666165175,
|
| 2039 |
+
"grad_norm": 0.22515636682510376,
|
| 2040 |
+
"learning_rate": 2.5e-05,
|
| 2041 |
+
"loss": 1.3605,
|
| 2042 |
+
"num_input_tokens_seen": 230760960,
|
| 2043 |
+
"step": 2520
|
| 2044 |
+
},
|
| 2045 |
+
{
|
| 2046 |
+
"epoch": 0.6410642172776942,
|
| 2047 |
+
"grad_norm": 0.33911067247390747,
|
| 2048 |
+
"learning_rate": 2.5e-05,
|
| 2049 |
+
"loss": 1.3648,
|
| 2050 |
+
"num_input_tokens_seen": 231683832,
|
| 2051 |
+
"step": 2530
|
| 2052 |
+
},
|
| 2053 |
+
{
|
| 2054 |
+
"epoch": 0.6435980679388709,
|
| 2055 |
+
"grad_norm": 0.2713491916656494,
|
| 2056 |
+
"learning_rate": 2.5e-05,
|
| 2057 |
+
"loss": 1.3581,
|
| 2058 |
+
"num_input_tokens_seen": 232586192,
|
| 2059 |
+
"step": 2540
|
| 2060 |
+
},
|
| 2061 |
+
{
|
| 2062 |
+
"epoch": 0.6461319186000475,
|
| 2063 |
+
"grad_norm": 0.22554545104503632,
|
| 2064 |
+
"learning_rate": 2.5e-05,
|
| 2065 |
+
"loss": 1.3217,
|
| 2066 |
+
"num_input_tokens_seen": 233513620,
|
| 2067 |
+
"step": 2550
|
| 2068 |
+
},
|
| 2069 |
+
{
|
| 2070 |
+
"epoch": 0.6486657692612242,
|
| 2071 |
+
"grad_norm": 0.23459571599960327,
|
| 2072 |
+
"learning_rate": 2.5e-05,
|
| 2073 |
+
"loss": 1.3185,
|
| 2074 |
+
"num_input_tokens_seen": 234405628,
|
| 2075 |
+
"step": 2560
|
| 2076 |
+
},
|
| 2077 |
+
{
|
| 2078 |
+
"epoch": 0.6511996199224008,
|
| 2079 |
+
"grad_norm": 0.22022689878940582,
|
| 2080 |
+
"learning_rate": 2.5e-05,
|
| 2081 |
+
"loss": 1.3724,
|
| 2082 |
+
"num_input_tokens_seen": 235287208,
|
| 2083 |
+
"step": 2570
|
| 2084 |
+
},
|
| 2085 |
+
{
|
| 2086 |
+
"epoch": 0.6537334705835774,
|
| 2087 |
+
"grad_norm": 0.2207019031047821,
|
| 2088 |
+
"learning_rate": 2.5e-05,
|
| 2089 |
+
"loss": 1.3871,
|
| 2090 |
+
"num_input_tokens_seen": 236206532,
|
| 2091 |
+
"step": 2580
|
| 2092 |
+
},
|
| 2093 |
+
{
|
| 2094 |
+
"epoch": 0.6562673212447542,
|
| 2095 |
+
"grad_norm": 0.286006897687912,
|
| 2096 |
+
"learning_rate": 2.5e-05,
|
| 2097 |
+
"loss": 1.338,
|
| 2098 |
+
"num_input_tokens_seen": 237132236,
|
| 2099 |
+
"step": 2590
|
| 2100 |
+
},
|
| 2101 |
+
{
|
| 2102 |
+
"epoch": 0.6588011719059308,
|
| 2103 |
+
"grad_norm": 0.24479633569717407,
|
| 2104 |
+
"learning_rate": 2.5e-05,
|
| 2105 |
+
"loss": 1.3636,
|
| 2106 |
+
"num_input_tokens_seen": 238036544,
|
| 2107 |
+
"step": 2600
|
| 2108 |
+
},
|
| 2109 |
+
{
|
| 2110 |
+
"epoch": 0.6613350225671074,
|
| 2111 |
+
"grad_norm": 0.21694402396678925,
|
| 2112 |
+
"learning_rate": 2.5e-05,
|
| 2113 |
+
"loss": 1.3711,
|
| 2114 |
+
"num_input_tokens_seen": 238978380,
|
| 2115 |
+
"step": 2610
|
| 2116 |
+
},
|
| 2117 |
+
{
|
| 2118 |
+
"epoch": 0.6638688732282841,
|
| 2119 |
+
"grad_norm": 0.22491593658924103,
|
| 2120 |
+
"learning_rate": 2.5e-05,
|
| 2121 |
+
"loss": 1.3516,
|
| 2122 |
+
"num_input_tokens_seen": 239893524,
|
| 2123 |
+
"step": 2620
|
| 2124 |
+
},
|
| 2125 |
+
{
|
| 2126 |
+
"epoch": 0.6664027238894608,
|
| 2127 |
+
"grad_norm": 0.24287302792072296,
|
| 2128 |
+
"learning_rate": 2.5e-05,
|
| 2129 |
+
"loss": 1.3333,
|
| 2130 |
+
"num_input_tokens_seen": 240753560,
|
| 2131 |
+
"step": 2630
|
| 2132 |
+
},
|
| 2133 |
+
{
|
| 2134 |
+
"epoch": 0.6689365745506374,
|
| 2135 |
+
"grad_norm": 0.24059581756591797,
|
| 2136 |
+
"learning_rate": 2.5e-05,
|
| 2137 |
+
"loss": 1.3172,
|
| 2138 |
+
"num_input_tokens_seen": 241689616,
|
| 2139 |
+
"step": 2640
|
| 2140 |
+
},
|
| 2141 |
+
{
|
| 2142 |
+
"epoch": 0.6714704252118141,
|
| 2143 |
+
"grad_norm": 0.24688631296157837,
|
| 2144 |
+
"learning_rate": 2.5e-05,
|
| 2145 |
+
"loss": 1.3377,
|
| 2146 |
+
"num_input_tokens_seen": 242618896,
|
| 2147 |
+
"step": 2650
|
| 2148 |
+
},
|
| 2149 |
+
{
|
| 2150 |
+
"epoch": 0.6740042758729907,
|
| 2151 |
+
"grad_norm": 0.2412404716014862,
|
| 2152 |
+
"learning_rate": 2.5e-05,
|
| 2153 |
+
"loss": 1.3512,
|
| 2154 |
+
"num_input_tokens_seen": 243555264,
|
| 2155 |
+
"step": 2660
|
| 2156 |
+
},
|
| 2157 |
+
{
|
| 2158 |
+
"epoch": 0.6765381265341673,
|
| 2159 |
+
"grad_norm": 0.23944397270679474,
|
| 2160 |
+
"learning_rate": 2.5e-05,
|
| 2161 |
+
"loss": 1.3806,
|
| 2162 |
+
"num_input_tokens_seen": 244450244,
|
| 2163 |
+
"step": 2670
|
| 2164 |
+
},
|
| 2165 |
+
{
|
| 2166 |
+
"epoch": 0.6790719771953441,
|
| 2167 |
+
"grad_norm": 0.24713559448719025,
|
| 2168 |
+
"learning_rate": 2.5e-05,
|
| 2169 |
+
"loss": 1.3251,
|
| 2170 |
+
"num_input_tokens_seen": 245398672,
|
| 2171 |
+
"step": 2680
|
| 2172 |
+
},
|
| 2173 |
+
{
|
| 2174 |
+
"epoch": 0.6816058278565207,
|
| 2175 |
+
"grad_norm": 0.31667396426200867,
|
| 2176 |
+
"learning_rate": 2.5e-05,
|
| 2177 |
+
"loss": 1.3642,
|
| 2178 |
+
"num_input_tokens_seen": 246320464,
|
| 2179 |
+
"step": 2690
|
| 2180 |
+
},
|
| 2181 |
+
{
|
| 2182 |
+
"epoch": 0.6841396785176974,
|
| 2183 |
+
"grad_norm": 0.250383585691452,
|
| 2184 |
+
"learning_rate": 2.5e-05,
|
| 2185 |
+
"loss": 1.3329,
|
| 2186 |
+
"num_input_tokens_seen": 247248308,
|
| 2187 |
+
"step": 2700
|
| 2188 |
+
},
|
| 2189 |
+
{
|
| 2190 |
+
"epoch": 0.686673529178874,
|
| 2191 |
+
"grad_norm": 0.2263907939195633,
|
| 2192 |
+
"learning_rate": 2.5e-05,
|
| 2193 |
+
"loss": 1.3281,
|
| 2194 |
+
"num_input_tokens_seen": 248202884,
|
| 2195 |
+
"step": 2710
|
| 2196 |
+
},
|
| 2197 |
+
{
|
| 2198 |
+
"epoch": 0.6892073798400506,
|
| 2199 |
+
"grad_norm": 0.24522219598293304,
|
| 2200 |
+
"learning_rate": 2.5e-05,
|
| 2201 |
+
"loss": 1.3477,
|
| 2202 |
+
"num_input_tokens_seen": 249166112,
|
| 2203 |
+
"step": 2720
|
| 2204 |
+
},
|
| 2205 |
+
{
|
| 2206 |
+
"epoch": 0.6917412305012274,
|
| 2207 |
+
"grad_norm": 0.22159820795059204,
|
| 2208 |
+
"learning_rate": 2.5e-05,
|
| 2209 |
+
"loss": 1.3192,
|
| 2210 |
+
"num_input_tokens_seen": 250077904,
|
| 2211 |
+
"step": 2730
|
| 2212 |
+
},
|
| 2213 |
+
{
|
| 2214 |
+
"epoch": 0.694275081162404,
|
| 2215 |
+
"grad_norm": 0.2300739735364914,
|
| 2216 |
+
"learning_rate": 2.5e-05,
|
| 2217 |
+
"loss": 1.3336,
|
| 2218 |
+
"num_input_tokens_seen": 251012120,
|
| 2219 |
+
"step": 2740
|
| 2220 |
+
},
|
| 2221 |
+
{
|
| 2222 |
+
"epoch": 0.6968089318235806,
|
| 2223 |
+
"grad_norm": 0.22758354246616364,
|
| 2224 |
+
"learning_rate": 2.5e-05,
|
| 2225 |
+
"loss": 1.3964,
|
| 2226 |
+
"num_input_tokens_seen": 251934920,
|
| 2227 |
+
"step": 2750
|
| 2228 |
+
},
|
| 2229 |
+
{
|
| 2230 |
+
"epoch": 0.6993427824847573,
|
| 2231 |
+
"grad_norm": 0.2598190903663635,
|
| 2232 |
+
"learning_rate": 2.5e-05,
|
| 2233 |
+
"loss": 1.3311,
|
| 2234 |
+
"num_input_tokens_seen": 252877580,
|
| 2235 |
+
"step": 2760
|
| 2236 |
+
},
|
| 2237 |
+
{
|
| 2238 |
+
"epoch": 0.701876633145934,
|
| 2239 |
+
"grad_norm": 0.23178431391716003,
|
| 2240 |
+
"learning_rate": 2.5e-05,
|
| 2241 |
+
"loss": 1.3453,
|
| 2242 |
+
"num_input_tokens_seen": 253792028,
|
| 2243 |
+
"step": 2770
|
| 2244 |
+
},
|
| 2245 |
+
{
|
| 2246 |
+
"epoch": 0.7044104838071106,
|
| 2247 |
+
"grad_norm": 0.26508447527885437,
|
| 2248 |
+
"learning_rate": 2.5e-05,
|
| 2249 |
+
"loss": 1.3635,
|
| 2250 |
+
"num_input_tokens_seen": 254742856,
|
| 2251 |
+
"step": 2780
|
| 2252 |
+
},
|
| 2253 |
+
{
|
| 2254 |
+
"epoch": 0.7069443344682873,
|
| 2255 |
+
"grad_norm": 0.263509601354599,
|
| 2256 |
+
"learning_rate": 2.5e-05,
|
| 2257 |
+
"loss": 1.3584,
|
| 2258 |
+
"num_input_tokens_seen": 255676980,
|
| 2259 |
+
"step": 2790
|
| 2260 |
+
},
|
| 2261 |
+
{
|
| 2262 |
+
"epoch": 0.7094781851294639,
|
| 2263 |
+
"grad_norm": 0.25076207518577576,
|
| 2264 |
+
"learning_rate": 2.5e-05,
|
| 2265 |
+
"loss": 1.3654,
|
| 2266 |
+
"num_input_tokens_seen": 256607480,
|
| 2267 |
+
"step": 2800
|
| 2268 |
+
},
|
| 2269 |
+
{
|
| 2270 |
+
"epoch": 0.7120120357906405,
|
| 2271 |
+
"grad_norm": 0.3114246726036072,
|
| 2272 |
+
"learning_rate": 2.5e-05,
|
| 2273 |
+
"loss": 1.3626,
|
| 2274 |
+
"num_input_tokens_seen": 257486156,
|
| 2275 |
+
"step": 2810
|
| 2276 |
+
},
|
| 2277 |
+
{
|
| 2278 |
+
"epoch": 0.7145458864518173,
|
| 2279 |
+
"grad_norm": 0.2184561789035797,
|
| 2280 |
+
"learning_rate": 2.5e-05,
|
| 2281 |
+
"loss": 1.3481,
|
| 2282 |
+
"num_input_tokens_seen": 258406168,
|
| 2283 |
+
"step": 2820
|
| 2284 |
+
},
|
| 2285 |
+
{
|
| 2286 |
+
"epoch": 0.7170797371129939,
|
| 2287 |
+
"grad_norm": 0.27279725670814514,
|
| 2288 |
+
"learning_rate": 2.5e-05,
|
| 2289 |
+
"loss": 1.3358,
|
| 2290 |
+
"num_input_tokens_seen": 259298936,
|
| 2291 |
+
"step": 2830
|
| 2292 |
+
},
|
| 2293 |
+
{
|
| 2294 |
+
"epoch": 0.7196135877741705,
|
| 2295 |
+
"grad_norm": 0.23473051190376282,
|
| 2296 |
+
"learning_rate": 2.5e-05,
|
| 2297 |
+
"loss": 1.3157,
|
| 2298 |
+
"num_input_tokens_seen": 260214884,
|
| 2299 |
+
"step": 2840
|
| 2300 |
+
},
|
| 2301 |
+
{
|
| 2302 |
+
"epoch": 0.7221474384353472,
|
| 2303 |
+
"grad_norm": 0.2273094654083252,
|
| 2304 |
+
"learning_rate": 2.5e-05,
|
| 2305 |
+
"loss": 1.3695,
|
| 2306 |
+
"num_input_tokens_seen": 261150656,
|
| 2307 |
+
"step": 2850
|
| 2308 |
+
},
|
| 2309 |
+
{
|
| 2310 |
+
"epoch": 0.7246812890965239,
|
| 2311 |
+
"grad_norm": 0.23328402638435364,
|
| 2312 |
+
"learning_rate": 2.5e-05,
|
| 2313 |
+
"loss": 1.3491,
|
| 2314 |
+
"num_input_tokens_seen": 262090748,
|
| 2315 |
+
"step": 2860
|
| 2316 |
+
},
|
| 2317 |
+
{
|
| 2318 |
+
"epoch": 0.7272151397577006,
|
| 2319 |
+
"grad_norm": 0.27058523893356323,
|
| 2320 |
+
"learning_rate": 2.5e-05,
|
| 2321 |
+
"loss": 1.3164,
|
| 2322 |
+
"num_input_tokens_seen": 263047956,
|
| 2323 |
+
"step": 2870
|
| 2324 |
+
},
|
| 2325 |
+
{
|
| 2326 |
+
"epoch": 0.7297489904188772,
|
| 2327 |
+
"grad_norm": 0.26919999718666077,
|
| 2328 |
+
"learning_rate": 2.5e-05,
|
| 2329 |
+
"loss": 1.3429,
|
| 2330 |
+
"num_input_tokens_seen": 263952708,
|
| 2331 |
+
"step": 2880
|
| 2332 |
+
},
|
| 2333 |
+
{
|
| 2334 |
+
"epoch": 0.7322828410800538,
|
| 2335 |
+
"grad_norm": 0.2629719078540802,
|
| 2336 |
+
"learning_rate": 2.5e-05,
|
| 2337 |
+
"loss": 1.3736,
|
| 2338 |
+
"num_input_tokens_seen": 264850904,
|
| 2339 |
+
"step": 2890
|
| 2340 |
+
},
|
| 2341 |
+
{
|
| 2342 |
+
"epoch": 0.7348166917412305,
|
| 2343 |
+
"grad_norm": 0.2600915729999542,
|
| 2344 |
+
"learning_rate": 2.5e-05,
|
| 2345 |
+
"loss": 1.3179,
|
| 2346 |
+
"num_input_tokens_seen": 265795528,
|
| 2347 |
+
"step": 2900
|
| 2348 |
+
},
|
| 2349 |
+
{
|
| 2350 |
+
"epoch": 0.7373505424024072,
|
| 2351 |
+
"grad_norm": 0.29251357913017273,
|
| 2352 |
+
"learning_rate": 2.5e-05,
|
| 2353 |
+
"loss": 1.3671,
|
| 2354 |
+
"num_input_tokens_seen": 266703240,
|
| 2355 |
+
"step": 2910
|
| 2356 |
+
},
|
| 2357 |
+
{
|
| 2358 |
+
"epoch": 0.7398843930635838,
|
| 2359 |
+
"grad_norm": 0.23803594708442688,
|
| 2360 |
+
"learning_rate": 2.5e-05,
|
| 2361 |
+
"loss": 1.3632,
|
| 2362 |
+
"num_input_tokens_seen": 267637720,
|
| 2363 |
+
"step": 2920
|
| 2364 |
+
},
|
| 2365 |
+
{
|
| 2366 |
+
"epoch": 0.7424182437247605,
|
| 2367 |
+
"grad_norm": 0.24492381513118744,
|
| 2368 |
+
"learning_rate": 2.5e-05,
|
| 2369 |
+
"loss": 1.3275,
|
| 2370 |
+
"num_input_tokens_seen": 268547588,
|
| 2371 |
+
"step": 2930
|
| 2372 |
+
},
|
| 2373 |
+
{
|
| 2374 |
+
"epoch": 0.7449520943859371,
|
| 2375 |
+
"grad_norm": 0.2277376800775528,
|
| 2376 |
+
"learning_rate": 2.5e-05,
|
| 2377 |
+
"loss": 1.3058,
|
| 2378 |
+
"num_input_tokens_seen": 269503056,
|
| 2379 |
+
"step": 2940
|
| 2380 |
+
},
|
| 2381 |
+
{
|
| 2382 |
+
"epoch": 0.7474859450471137,
|
| 2383 |
+
"grad_norm": 0.22645527124404907,
|
| 2384 |
+
"learning_rate": 2.5e-05,
|
| 2385 |
+
"loss": 1.3462,
|
| 2386 |
+
"num_input_tokens_seen": 270372524,
|
| 2387 |
+
"step": 2950
|
| 2388 |
+
},
|
| 2389 |
+
{
|
| 2390 |
+
"epoch": 0.7500197957082905,
|
| 2391 |
+
"grad_norm": 0.27738144993782043,
|
| 2392 |
+
"learning_rate": 2.5e-05,
|
| 2393 |
+
"loss": 1.2953,
|
| 2394 |
+
"num_input_tokens_seen": 271255520,
|
| 2395 |
+
"step": 2960
|
| 2396 |
+
},
|
| 2397 |
+
{
|
| 2398 |
+
"epoch": 0.7525536463694671,
|
| 2399 |
+
"grad_norm": 0.2460719496011734,
|
| 2400 |
+
"learning_rate": 2.5e-05,
|
| 2401 |
+
"loss": 1.3291,
|
| 2402 |
+
"num_input_tokens_seen": 272173512,
|
| 2403 |
+
"step": 2970
|
| 2404 |
+
},
|
| 2405 |
+
{
|
| 2406 |
+
"epoch": 0.7550874970306437,
|
| 2407 |
+
"grad_norm": 0.23774035274982452,
|
| 2408 |
+
"learning_rate": 2.5e-05,
|
| 2409 |
+
"loss": 1.3105,
|
| 2410 |
+
"num_input_tokens_seen": 273082396,
|
| 2411 |
+
"step": 2980
|
| 2412 |
+
},
|
| 2413 |
+
{
|
| 2414 |
+
"epoch": 0.7576213476918204,
|
| 2415 |
+
"grad_norm": 0.2344847470521927,
|
| 2416 |
+
"learning_rate": 2.5e-05,
|
| 2417 |
+
"loss": 1.3379,
|
| 2418 |
+
"num_input_tokens_seen": 273951600,
|
| 2419 |
+
"step": 2990
|
| 2420 |
+
},
|
| 2421 |
+
{
|
| 2422 |
+
"epoch": 0.760155198352997,
|
| 2423 |
+
"grad_norm": 0.2422836273908615,
|
| 2424 |
+
"learning_rate": 2.5e-05,
|
| 2425 |
+
"loss": 1.3437,
|
| 2426 |
+
"num_input_tokens_seen": 274855796,
|
| 2427 |
+
"step": 3000
|
| 2428 |
+
},
|
| 2429 |
+
{
|
| 2430 |
+
"epoch": 0.760155198352997,
|
| 2431 |
+
"eval_loss": 1.3378311395645142,
|
| 2432 |
+
"eval_runtime": 2.7862,
|
| 2433 |
+
"eval_samples_per_second": 53.837,
|
| 2434 |
+
"eval_steps_per_second": 6.819,
|
| 2435 |
+
"num_input_tokens_seen": 274855796,
|
| 2436 |
+
"step": 3000
|
| 2437 |
+
},
|
| 2438 |
+
{
|
| 2439 |
+
"epoch": 0.7626890490141738,
|
| 2440 |
+
"grad_norm": 0.2418714016675949,
|
| 2441 |
+
"learning_rate": 2.5e-05,
|
| 2442 |
+
"loss": 1.3683,
|
| 2443 |
+
"num_input_tokens_seen": 275793364,
|
| 2444 |
+
"step": 3010
|
| 2445 |
+
},
|
| 2446 |
+
{
|
| 2447 |
+
"epoch": 0.7652228996753504,
|
| 2448 |
+
"grad_norm": 0.2433195561170578,
|
| 2449 |
+
"learning_rate": 2.5e-05,
|
| 2450 |
+
"loss": 1.3397,
|
| 2451 |
+
"num_input_tokens_seen": 276766688,
|
| 2452 |
+
"step": 3020
|
| 2453 |
+
},
|
| 2454 |
+
{
|
| 2455 |
+
"epoch": 0.767756750336527,
|
| 2456 |
+
"grad_norm": 0.2531881034374237,
|
| 2457 |
+
"learning_rate": 2.5e-05,
|
| 2458 |
+
"loss": 1.3069,
|
| 2459 |
+
"num_input_tokens_seen": 277692944,
|
| 2460 |
+
"step": 3030
|
| 2461 |
+
},
|
| 2462 |
+
{
|
| 2463 |
+
"epoch": 0.7702906009977037,
|
| 2464 |
+
"grad_norm": 0.228854700922966,
|
| 2465 |
+
"learning_rate": 2.5e-05,
|
| 2466 |
+
"loss": 1.3467,
|
| 2467 |
+
"num_input_tokens_seen": 278633648,
|
| 2468 |
+
"step": 3040
|
| 2469 |
+
},
|
| 2470 |
+
{
|
| 2471 |
+
"epoch": 0.7728244516588804,
|
| 2472 |
+
"grad_norm": 0.21645446121692657,
|
| 2473 |
+
"learning_rate": 2.5e-05,
|
| 2474 |
+
"loss": 1.2949,
|
| 2475 |
+
"num_input_tokens_seen": 279542668,
|
| 2476 |
+
"step": 3050
|
| 2477 |
+
},
|
| 2478 |
+
{
|
| 2479 |
+
"epoch": 0.775358302320057,
|
| 2480 |
+
"grad_norm": 0.2668648362159729,
|
| 2481 |
+
"learning_rate": 2.5e-05,
|
| 2482 |
+
"loss": 1.3272,
|
| 2483 |
+
"num_input_tokens_seen": 280474528,
|
| 2484 |
+
"step": 3060
|
| 2485 |
+
},
|
| 2486 |
+
{
|
| 2487 |
+
"epoch": 0.7778921529812337,
|
| 2488 |
+
"grad_norm": 0.26199036836624146,
|
| 2489 |
+
"learning_rate": 2.5e-05,
|
| 2490 |
+
"loss": 1.3395,
|
| 2491 |
+
"num_input_tokens_seen": 281383776,
|
| 2492 |
+
"step": 3070
|
| 2493 |
+
},
|
| 2494 |
+
{
|
| 2495 |
+
"epoch": 0.7804260036424103,
|
| 2496 |
+
"grad_norm": 0.23948872089385986,
|
| 2497 |
+
"learning_rate": 2.5e-05,
|
| 2498 |
+
"loss": 1.3534,
|
| 2499 |
+
"num_input_tokens_seen": 282297260,
|
| 2500 |
+
"step": 3080
|
| 2501 |
+
},
|
| 2502 |
+
{
|
| 2503 |
+
"epoch": 0.7829598543035869,
|
| 2504 |
+
"grad_norm": 0.2561713755130768,
|
| 2505 |
+
"learning_rate": 2.5e-05,
|
| 2506 |
+
"loss": 1.3251,
|
| 2507 |
+
"num_input_tokens_seen": 283169516,
|
| 2508 |
+
"step": 3090
|
| 2509 |
+
},
|
| 2510 |
+
{
|
| 2511 |
+
"epoch": 0.7854937049647637,
|
| 2512 |
+
"grad_norm": 0.26099705696105957,
|
| 2513 |
+
"learning_rate": 2.5e-05,
|
| 2514 |
+
"loss": 1.3394,
|
| 2515 |
+
"num_input_tokens_seen": 284109700,
|
| 2516 |
+
"step": 3100
|
| 2517 |
+
},
|
| 2518 |
+
{
|
| 2519 |
+
"epoch": 0.7880275556259403,
|
| 2520 |
+
"grad_norm": 0.23930218815803528,
|
| 2521 |
+
"learning_rate": 2.5e-05,
|
| 2522 |
+
"loss": 1.3242,
|
| 2523 |
+
"num_input_tokens_seen": 285031264,
|
| 2524 |
+
"step": 3110
|
| 2525 |
+
},
|
| 2526 |
+
{
|
| 2527 |
+
"epoch": 0.7905614062871169,
|
| 2528 |
+
"grad_norm": 0.23478297889232635,
|
| 2529 |
+
"learning_rate": 2.5e-05,
|
| 2530 |
+
"loss": 1.3647,
|
| 2531 |
+
"num_input_tokens_seen": 285943620,
|
| 2532 |
+
"step": 3120
|
| 2533 |
+
},
|
| 2534 |
+
{
|
| 2535 |
+
"epoch": 0.7930952569482936,
|
| 2536 |
+
"grad_norm": 0.24018226563930511,
|
| 2537 |
+
"learning_rate": 2.5e-05,
|
| 2538 |
+
"loss": 1.3166,
|
| 2539 |
+
"num_input_tokens_seen": 286819840,
|
| 2540 |
+
"step": 3130
|
| 2541 |
+
},
|
| 2542 |
+
{
|
| 2543 |
+
"epoch": 0.7956291076094703,
|
| 2544 |
+
"grad_norm": 0.22437995672225952,
|
| 2545 |
+
"learning_rate": 2.5e-05,
|
| 2546 |
+
"loss": 1.3418,
|
| 2547 |
+
"num_input_tokens_seen": 287731640,
|
| 2548 |
+
"step": 3140
|
| 2549 |
+
},
|
| 2550 |
+
{
|
| 2551 |
+
"epoch": 0.7981629582706469,
|
| 2552 |
+
"grad_norm": 0.2912137806415558,
|
| 2553 |
+
"learning_rate": 2.5e-05,
|
| 2554 |
+
"loss": 1.3336,
|
| 2555 |
+
"num_input_tokens_seen": 288650768,
|
| 2556 |
+
"step": 3150
|
| 2557 |
+
},
|
| 2558 |
+
{
|
| 2559 |
+
"epoch": 0.8006968089318236,
|
| 2560 |
+
"grad_norm": 0.27003979682922363,
|
| 2561 |
+
"learning_rate": 2.5e-05,
|
| 2562 |
+
"loss": 1.3094,
|
| 2563 |
+
"num_input_tokens_seen": 289579424,
|
| 2564 |
+
"step": 3160
|
| 2565 |
+
},
|
| 2566 |
+
{
|
| 2567 |
+
"epoch": 0.8032306595930002,
|
| 2568 |
+
"grad_norm": 0.24906513094902039,
|
| 2569 |
+
"learning_rate": 2.5e-05,
|
| 2570 |
+
"loss": 1.3089,
|
| 2571 |
+
"num_input_tokens_seen": 290506080,
|
| 2572 |
+
"step": 3170
|
| 2573 |
+
},
|
| 2574 |
+
{
|
| 2575 |
+
"epoch": 0.805764510254177,
|
| 2576 |
+
"grad_norm": 0.2620064616203308,
|
| 2577 |
+
"learning_rate": 2.5e-05,
|
| 2578 |
+
"loss": 1.3741,
|
| 2579 |
+
"num_input_tokens_seen": 291447632,
|
| 2580 |
+
"step": 3180
|
| 2581 |
+
},
|
| 2582 |
+
{
|
| 2583 |
+
"epoch": 0.8082983609153536,
|
| 2584 |
+
"grad_norm": 0.22881096601486206,
|
| 2585 |
+
"learning_rate": 2.5e-05,
|
| 2586 |
+
"loss": 1.3601,
|
| 2587 |
+
"num_input_tokens_seen": 292382736,
|
| 2588 |
+
"step": 3190
|
| 2589 |
+
},
|
| 2590 |
+
{
|
| 2591 |
+
"epoch": 0.8108322115765302,
|
| 2592 |
+
"grad_norm": 0.23649707436561584,
|
| 2593 |
+
"learning_rate": 2.5e-05,
|
| 2594 |
+
"loss": 1.3212,
|
| 2595 |
+
"num_input_tokens_seen": 293339376,
|
| 2596 |
+
"step": 3200
|
| 2597 |
+
},
|
| 2598 |
+
{
|
| 2599 |
+
"epoch": 0.8133660622377069,
|
| 2600 |
+
"grad_norm": 0.22773633897304535,
|
| 2601 |
+
"learning_rate": 2.5e-05,
|
| 2602 |
+
"loss": 1.3124,
|
| 2603 |
+
"num_input_tokens_seen": 294273900,
|
| 2604 |
+
"step": 3210
|
| 2605 |
+
},
|
| 2606 |
+
{
|
| 2607 |
+
"epoch": 0.8158999128988835,
|
| 2608 |
+
"grad_norm": 0.23439520597457886,
|
| 2609 |
+
"learning_rate": 2.5e-05,
|
| 2610 |
+
"loss": 1.3104,
|
| 2611 |
+
"num_input_tokens_seen": 295167620,
|
| 2612 |
+
"step": 3220
|
| 2613 |
+
},
|
| 2614 |
+
{
|
| 2615 |
+
"epoch": 0.8184337635600601,
|
| 2616 |
+
"grad_norm": 0.2587607800960541,
|
| 2617 |
+
"learning_rate": 2.5e-05,
|
| 2618 |
+
"loss": 1.3378,
|
| 2619 |
+
"num_input_tokens_seen": 296070252,
|
| 2620 |
+
"step": 3230
|
| 2621 |
+
},
|
| 2622 |
+
{
|
| 2623 |
+
"epoch": 0.8209676142212369,
|
| 2624 |
+
"grad_norm": 0.2375950813293457,
|
| 2625 |
+
"learning_rate": 2.5e-05,
|
| 2626 |
+
"loss": 1.3608,
|
| 2627 |
+
"num_input_tokens_seen": 296964880,
|
| 2628 |
+
"step": 3240
|
| 2629 |
+
},
|
| 2630 |
+
{
|
| 2631 |
+
"epoch": 0.8235014648824135,
|
| 2632 |
+
"grad_norm": 0.217642143368721,
|
| 2633 |
+
"learning_rate": 2.5e-05,
|
| 2634 |
+
"loss": 1.3711,
|
| 2635 |
+
"num_input_tokens_seen": 297861584,
|
| 2636 |
+
"step": 3250
|
| 2637 |
+
},
|
| 2638 |
+
{
|
| 2639 |
+
"epoch": 0.8260353155435901,
|
| 2640 |
+
"grad_norm": 0.24903365969657898,
|
| 2641 |
+
"learning_rate": 2.5e-05,
|
| 2642 |
+
"loss": 1.3759,
|
| 2643 |
+
"num_input_tokens_seen": 298763600,
|
| 2644 |
+
"step": 3260
|
| 2645 |
+
},
|
| 2646 |
+
{
|
| 2647 |
+
"epoch": 0.8285691662047668,
|
| 2648 |
+
"grad_norm": 0.25492629408836365,
|
| 2649 |
+
"learning_rate": 2.5e-05,
|
| 2650 |
+
"loss": 1.336,
|
| 2651 |
+
"num_input_tokens_seen": 299655852,
|
| 2652 |
+
"step": 3270
|
| 2653 |
+
},
|
| 2654 |
+
{
|
| 2655 |
+
"epoch": 0.8311030168659435,
|
| 2656 |
+
"grad_norm": 0.26514139771461487,
|
| 2657 |
+
"learning_rate": 2.5e-05,
|
| 2658 |
+
"loss": 1.3294,
|
| 2659 |
+
"num_input_tokens_seen": 300539872,
|
| 2660 |
+
"step": 3280
|
| 2661 |
+
},
|
| 2662 |
+
{
|
| 2663 |
+
"epoch": 0.8336368675271201,
|
| 2664 |
+
"grad_norm": 0.23889844119548798,
|
| 2665 |
+
"learning_rate": 2.5e-05,
|
| 2666 |
+
"loss": 1.3845,
|
| 2667 |
+
"num_input_tokens_seen": 301433356,
|
| 2668 |
+
"step": 3290
|
| 2669 |
+
},
|
| 2670 |
+
{
|
| 2671 |
+
"epoch": 0.8361707181882968,
|
| 2672 |
+
"grad_norm": 0.23075729608535767,
|
| 2673 |
+
"learning_rate": 2.5e-05,
|
| 2674 |
+
"loss": 1.3359,
|
| 2675 |
+
"num_input_tokens_seen": 302358284,
|
| 2676 |
+
"step": 3300
|
| 2677 |
+
},
|
| 2678 |
+
{
|
| 2679 |
+
"epoch": 0.8387045688494734,
|
| 2680 |
+
"grad_norm": 0.28124797344207764,
|
| 2681 |
+
"learning_rate": 2.5e-05,
|
| 2682 |
+
"loss": 1.3663,
|
| 2683 |
+
"num_input_tokens_seen": 303293764,
|
| 2684 |
+
"step": 3310
|
| 2685 |
+
},
|
| 2686 |
+
{
|
| 2687 |
+
"epoch": 0.84123841951065,
|
| 2688 |
+
"grad_norm": 0.30670827627182007,
|
| 2689 |
+
"learning_rate": 2.5e-05,
|
| 2690 |
+
"loss": 1.335,
|
| 2691 |
+
"num_input_tokens_seen": 304171336,
|
| 2692 |
+
"step": 3320
|
| 2693 |
+
},
|
| 2694 |
+
{
|
| 2695 |
+
"epoch": 0.8437722701718268,
|
| 2696 |
+
"grad_norm": 0.22578497231006622,
|
| 2697 |
+
"learning_rate": 2.5e-05,
|
| 2698 |
+
"loss": 1.327,
|
| 2699 |
+
"num_input_tokens_seen": 305091264,
|
| 2700 |
+
"step": 3330
|
| 2701 |
+
},
|
| 2702 |
+
{
|
| 2703 |
+
"epoch": 0.8463061208330034,
|
| 2704 |
+
"grad_norm": 0.22120265662670135,
|
| 2705 |
+
"learning_rate": 2.5e-05,
|
| 2706 |
+
"loss": 1.3509,
|
| 2707 |
+
"num_input_tokens_seen": 306010588,
|
| 2708 |
+
"step": 3340
|
| 2709 |
+
},
|
| 2710 |
+
{
|
| 2711 |
+
"epoch": 0.8488399714941801,
|
| 2712 |
+
"grad_norm": 0.2477473020553589,
|
| 2713 |
+
"learning_rate": 2.5e-05,
|
| 2714 |
+
"loss": 1.3565,
|
| 2715 |
+
"num_input_tokens_seen": 306940328,
|
| 2716 |
+
"step": 3350
|
| 2717 |
+
},
|
| 2718 |
+
{
|
| 2719 |
+
"epoch": 0.8513738221553567,
|
| 2720 |
+
"grad_norm": 0.2530181109905243,
|
| 2721 |
+
"learning_rate": 2.5e-05,
|
| 2722 |
+
"loss": 1.2936,
|
| 2723 |
+
"num_input_tokens_seen": 307838056,
|
| 2724 |
+
"step": 3360
|
| 2725 |
+
},
|
| 2726 |
+
{
|
| 2727 |
+
"epoch": 0.8539076728165333,
|
| 2728 |
+
"grad_norm": 0.2556324303150177,
|
| 2729 |
+
"learning_rate": 2.5e-05,
|
| 2730 |
+
"loss": 1.3002,
|
| 2731 |
+
"num_input_tokens_seen": 308773220,
|
| 2732 |
+
"step": 3370
|
| 2733 |
+
},
|
| 2734 |
+
{
|
| 2735 |
+
"epoch": 0.8564415234777101,
|
| 2736 |
+
"grad_norm": 0.24870575964450836,
|
| 2737 |
+
"learning_rate": 2.5e-05,
|
| 2738 |
+
"loss": 1.3086,
|
| 2739 |
+
"num_input_tokens_seen": 309713036,
|
| 2740 |
+
"step": 3380
|
| 2741 |
+
},
|
| 2742 |
+
{
|
| 2743 |
+
"epoch": 0.8589753741388867,
|
| 2744 |
+
"grad_norm": 0.22579419612884521,
|
| 2745 |
+
"learning_rate": 2.5e-05,
|
| 2746 |
+
"loss": 1.3238,
|
| 2747 |
+
"num_input_tokens_seen": 310676544,
|
| 2748 |
+
"step": 3390
|
| 2749 |
+
},
|
| 2750 |
+
{
|
| 2751 |
+
"epoch": 0.8615092248000633,
|
| 2752 |
+
"grad_norm": 0.26896366477012634,
|
| 2753 |
+
"learning_rate": 2.5e-05,
|
| 2754 |
+
"loss": 1.3518,
|
| 2755 |
+
"num_input_tokens_seen": 311609100,
|
| 2756 |
+
"step": 3400
|
| 2757 |
+
},
|
| 2758 |
+
{
|
| 2759 |
+
"epoch": 0.86404307546124,
|
| 2760 |
+
"grad_norm": 0.23491699993610382,
|
| 2761 |
+
"learning_rate": 2.5e-05,
|
| 2762 |
+
"loss": 1.3478,
|
| 2763 |
+
"num_input_tokens_seen": 312541060,
|
| 2764 |
+
"step": 3410
|
| 2765 |
+
},
|
| 2766 |
+
{
|
| 2767 |
+
"epoch": 0.8665769261224167,
|
| 2768 |
+
"grad_norm": 0.21398873627185822,
|
| 2769 |
+
"learning_rate": 2.5e-05,
|
| 2770 |
+
"loss": 1.317,
|
| 2771 |
+
"num_input_tokens_seen": 313464680,
|
| 2772 |
+
"step": 3420
|
| 2773 |
+
},
|
| 2774 |
+
{
|
| 2775 |
+
"epoch": 0.8691107767835933,
|
| 2776 |
+
"grad_norm": 0.2201145589351654,
|
| 2777 |
+
"learning_rate": 2.5e-05,
|
| 2778 |
+
"loss": 1.3203,
|
| 2779 |
+
"num_input_tokens_seen": 314362092,
|
| 2780 |
+
"step": 3430
|
| 2781 |
+
},
|
| 2782 |
+
{
|
| 2783 |
+
"epoch": 0.87164462744477,
|
| 2784 |
+
"grad_norm": 0.23937499523162842,
|
| 2785 |
+
"learning_rate": 2.5e-05,
|
| 2786 |
+
"loss": 1.3594,
|
| 2787 |
+
"num_input_tokens_seen": 315286788,
|
| 2788 |
+
"step": 3440
|
| 2789 |
+
},
|
| 2790 |
+
{
|
| 2791 |
+
"epoch": 0.8741784781059466,
|
| 2792 |
+
"grad_norm": 0.2299693375825882,
|
| 2793 |
+
"learning_rate": 2.5e-05,
|
| 2794 |
+
"loss": 1.359,
|
| 2795 |
+
"num_input_tokens_seen": 316199708,
|
| 2796 |
+
"step": 3450
|
| 2797 |
+
},
|
| 2798 |
+
{
|
| 2799 |
+
"epoch": 0.8767123287671232,
|
| 2800 |
+
"grad_norm": 0.21679440140724182,
|
| 2801 |
+
"learning_rate": 2.5e-05,
|
| 2802 |
+
"loss": 1.3372,
|
| 2803 |
+
"num_input_tokens_seen": 317082032,
|
| 2804 |
+
"step": 3460
|
| 2805 |
+
},
|
| 2806 |
+
{
|
| 2807 |
+
"epoch": 0.8792461794283,
|
| 2808 |
+
"grad_norm": 0.23869968950748444,
|
| 2809 |
+
"learning_rate": 2.5e-05,
|
| 2810 |
+
"loss": 1.2815,
|
| 2811 |
+
"num_input_tokens_seen": 317999160,
|
| 2812 |
+
"step": 3470
|
| 2813 |
+
},
|
| 2814 |
+
{
|
| 2815 |
+
"epoch": 0.8817800300894766,
|
| 2816 |
+
"grad_norm": 0.24342550337314606,
|
| 2817 |
+
"learning_rate": 2.5e-05,
|
| 2818 |
+
"loss": 1.3222,
|
| 2819 |
+
"num_input_tokens_seen": 318945628,
|
| 2820 |
+
"step": 3480
|
| 2821 |
+
},
|
| 2822 |
+
{
|
| 2823 |
+
"epoch": 0.8843138807506532,
|
| 2824 |
+
"grad_norm": 0.23146317899227142,
|
| 2825 |
+
"learning_rate": 2.5e-05,
|
| 2826 |
+
"loss": 1.3188,
|
| 2827 |
+
"num_input_tokens_seen": 319892264,
|
| 2828 |
+
"step": 3490
|
| 2829 |
+
},
|
| 2830 |
+
{
|
| 2831 |
+
"epoch": 0.8868477314118299,
|
| 2832 |
+
"grad_norm": 0.27557140588760376,
|
| 2833 |
+
"learning_rate": 2.5e-05,
|
| 2834 |
+
"loss": 1.3065,
|
| 2835 |
+
"num_input_tokens_seen": 320815992,
|
| 2836 |
+
"step": 3500
|
| 2837 |
+
},
|
| 2838 |
+
{
|
| 2839 |
+
"epoch": 0.8893815820730065,
|
| 2840 |
+
"grad_norm": 0.24911952018737793,
|
| 2841 |
+
"learning_rate": 2.5e-05,
|
| 2842 |
+
"loss": 1.3275,
|
| 2843 |
+
"num_input_tokens_seen": 321703172,
|
| 2844 |
+
"step": 3510
|
| 2845 |
+
},
|
| 2846 |
+
{
|
| 2847 |
+
"epoch": 0.8919154327341833,
|
| 2848 |
+
"grad_norm": 0.2727194130420685,
|
| 2849 |
+
"learning_rate": 2.5e-05,
|
| 2850 |
+
"loss": 1.3297,
|
| 2851 |
+
"num_input_tokens_seen": 322642588,
|
| 2852 |
+
"step": 3520
|
| 2853 |
+
},
|
| 2854 |
+
{
|
| 2855 |
+
"epoch": 0.8944492833953599,
|
| 2856 |
+
"grad_norm": 0.242356538772583,
|
| 2857 |
+
"learning_rate": 2.5e-05,
|
| 2858 |
+
"loss": 1.2881,
|
| 2859 |
+
"num_input_tokens_seen": 323529188,
|
| 2860 |
+
"step": 3530
|
| 2861 |
+
},
|
| 2862 |
+
{
|
| 2863 |
+
"epoch": 0.8969831340565365,
|
| 2864 |
+
"grad_norm": 0.21331574022769928,
|
| 2865 |
+
"learning_rate": 2.5e-05,
|
| 2866 |
+
"loss": 1.2861,
|
| 2867 |
+
"num_input_tokens_seen": 324438984,
|
| 2868 |
+
"step": 3540
|
| 2869 |
+
},
|
| 2870 |
+
{
|
| 2871 |
+
"epoch": 0.8995169847177132,
|
| 2872 |
+
"grad_norm": 0.28540030121803284,
|
| 2873 |
+
"learning_rate": 2.5e-05,
|
| 2874 |
+
"loss": 1.3361,
|
| 2875 |
+
"num_input_tokens_seen": 325302632,
|
| 2876 |
+
"step": 3550
|
| 2877 |
+
},
|
| 2878 |
+
{
|
| 2879 |
+
"epoch": 0.9020508353788899,
|
| 2880 |
+
"grad_norm": 0.2721042037010193,
|
| 2881 |
+
"learning_rate": 2.5e-05,
|
| 2882 |
+
"loss": 1.3377,
|
| 2883 |
+
"num_input_tokens_seen": 326223312,
|
| 2884 |
+
"step": 3560
|
| 2885 |
+
},
|
| 2886 |
+
{
|
| 2887 |
+
"epoch": 0.9045846860400665,
|
| 2888 |
+
"grad_norm": 0.235883429646492,
|
| 2889 |
+
"learning_rate": 2.5e-05,
|
| 2890 |
+
"loss": 1.3603,
|
| 2891 |
+
"num_input_tokens_seen": 327174992,
|
| 2892 |
+
"step": 3570
|
| 2893 |
+
},
|
| 2894 |
+
{
|
| 2895 |
+
"epoch": 0.9071185367012432,
|
| 2896 |
+
"grad_norm": 0.2746555507183075,
|
| 2897 |
+
"learning_rate": 2.5e-05,
|
| 2898 |
+
"loss": 1.3497,
|
| 2899 |
+
"num_input_tokens_seen": 328087740,
|
| 2900 |
+
"step": 3580
|
| 2901 |
+
},
|
| 2902 |
+
{
|
| 2903 |
+
"epoch": 0.9096523873624198,
|
| 2904 |
+
"grad_norm": 0.21206247806549072,
|
| 2905 |
+
"learning_rate": 2.5e-05,
|
| 2906 |
+
"loss": 1.3192,
|
| 2907 |
+
"num_input_tokens_seen": 329015496,
|
| 2908 |
+
"step": 3590
|
| 2909 |
+
},
|
| 2910 |
+
{
|
| 2911 |
+
"epoch": 0.9121862380235964,
|
| 2912 |
+
"grad_norm": 0.24580571055412292,
|
| 2913 |
+
"learning_rate": 2.5e-05,
|
| 2914 |
+
"loss": 1.2958,
|
| 2915 |
+
"num_input_tokens_seen": 329914504,
|
| 2916 |
+
"step": 3600
|
| 2917 |
+
},
|
| 2918 |
+
{
|
| 2919 |
+
"epoch": 0.9147200886847732,
|
| 2920 |
+
"grad_norm": 0.2298029512166977,
|
| 2921 |
+
"learning_rate": 2.5e-05,
|
| 2922 |
+
"loss": 1.2955,
|
| 2923 |
+
"num_input_tokens_seen": 330861412,
|
| 2924 |
+
"step": 3610
|
| 2925 |
+
},
|
| 2926 |
+
{
|
| 2927 |
+
"epoch": 0.9172539393459498,
|
| 2928 |
+
"grad_norm": 0.20944957435131073,
|
| 2929 |
+
"learning_rate": 2.5e-05,
|
| 2930 |
+
"loss": 1.3413,
|
| 2931 |
+
"num_input_tokens_seen": 331705132,
|
| 2932 |
+
"step": 3620
|
| 2933 |
+
},
|
| 2934 |
+
{
|
| 2935 |
+
"epoch": 0.9197877900071264,
|
| 2936 |
+
"grad_norm": 0.26745468378067017,
|
| 2937 |
+
"learning_rate": 2.5e-05,
|
| 2938 |
+
"loss": 1.3528,
|
| 2939 |
+
"num_input_tokens_seen": 332613612,
|
| 2940 |
+
"step": 3630
|
| 2941 |
+
},
|
| 2942 |
+
{
|
| 2943 |
+
"epoch": 0.9223216406683031,
|
| 2944 |
+
"grad_norm": 0.23441898822784424,
|
| 2945 |
+
"learning_rate": 2.5e-05,
|
| 2946 |
+
"loss": 1.3125,
|
| 2947 |
+
"num_input_tokens_seen": 333546464,
|
| 2948 |
+
"step": 3640
|
| 2949 |
+
},
|
| 2950 |
+
{
|
| 2951 |
+
"epoch": 0.9248554913294798,
|
| 2952 |
+
"grad_norm": 0.25231051445007324,
|
| 2953 |
+
"learning_rate": 2.5e-05,
|
| 2954 |
+
"loss": 1.3264,
|
| 2955 |
+
"num_input_tokens_seen": 334449860,
|
| 2956 |
+
"step": 3650
|
| 2957 |
+
},
|
| 2958 |
+
{
|
| 2959 |
+
"epoch": 0.9273893419906565,
|
| 2960 |
+
"grad_norm": 0.22412322461605072,
|
| 2961 |
+
"learning_rate": 2.5e-05,
|
| 2962 |
+
"loss": 1.3159,
|
| 2963 |
+
"num_input_tokens_seen": 335390600,
|
| 2964 |
+
"step": 3660
|
| 2965 |
+
},
|
| 2966 |
+
{
|
| 2967 |
+
"epoch": 0.9299231926518331,
|
| 2968 |
+
"grad_norm": 0.23513691127300262,
|
| 2969 |
+
"learning_rate": 2.5e-05,
|
| 2970 |
+
"loss": 1.3115,
|
| 2971 |
+
"num_input_tokens_seen": 336327464,
|
| 2972 |
+
"step": 3670
|
| 2973 |
+
},
|
| 2974 |
+
{
|
| 2975 |
+
"epoch": 0.9324570433130097,
|
| 2976 |
+
"grad_norm": 0.22470693290233612,
|
| 2977 |
+
"learning_rate": 2.5e-05,
|
| 2978 |
+
"loss": 1.3214,
|
| 2979 |
+
"num_input_tokens_seen": 337241700,
|
| 2980 |
+
"step": 3680
|
| 2981 |
+
},
|
| 2982 |
+
{
|
| 2983 |
+
"epoch": 0.9349908939741864,
|
| 2984 |
+
"grad_norm": 0.24091310799121857,
|
| 2985 |
+
"learning_rate": 2.5e-05,
|
| 2986 |
+
"loss": 1.3306,
|
| 2987 |
+
"num_input_tokens_seen": 338184552,
|
| 2988 |
+
"step": 3690
|
| 2989 |
+
},
|
| 2990 |
+
{
|
| 2991 |
+
"epoch": 0.9375247446353631,
|
| 2992 |
+
"grad_norm": 0.23601089417934418,
|
| 2993 |
+
"learning_rate": 2.5e-05,
|
| 2994 |
+
"loss": 1.2856,
|
| 2995 |
+
"num_input_tokens_seen": 339109296,
|
| 2996 |
+
"step": 3700
|
| 2997 |
+
},
|
| 2998 |
+
{
|
| 2999 |
+
"epoch": 0.9400585952965397,
|
| 3000 |
+
"grad_norm": 0.23559744656085968,
|
| 3001 |
+
"learning_rate": 2.5e-05,
|
| 3002 |
+
"loss": 1.293,
|
| 3003 |
+
"num_input_tokens_seen": 340010148,
|
| 3004 |
+
"step": 3710
|
| 3005 |
+
},
|
| 3006 |
+
{
|
| 3007 |
+
"epoch": 0.9425924459577164,
|
| 3008 |
+
"grad_norm": 0.2477143257856369,
|
| 3009 |
+
"learning_rate": 2.5e-05,
|
| 3010 |
+
"loss": 1.3226,
|
| 3011 |
+
"num_input_tokens_seen": 340905016,
|
| 3012 |
+
"step": 3720
|
| 3013 |
+
},
|
| 3014 |
+
{
|
| 3015 |
+
"epoch": 0.945126296618893,
|
| 3016 |
+
"grad_norm": 0.2724590599536896,
|
| 3017 |
+
"learning_rate": 2.5e-05,
|
| 3018 |
+
"loss": 1.3063,
|
| 3019 |
+
"num_input_tokens_seen": 341861552,
|
| 3020 |
+
"step": 3730
|
| 3021 |
+
},
|
| 3022 |
+
{
|
| 3023 |
+
"epoch": 0.9476601472800696,
|
| 3024 |
+
"grad_norm": 0.23112662136554718,
|
| 3025 |
+
"learning_rate": 2.5e-05,
|
| 3026 |
+
"loss": 1.3099,
|
| 3027 |
+
"num_input_tokens_seen": 342806136,
|
| 3028 |
+
"step": 3740
|
| 3029 |
+
},
|
| 3030 |
+
{
|
| 3031 |
+
"epoch": 0.9501939979412464,
|
| 3032 |
+
"grad_norm": 0.2522134780883789,
|
| 3033 |
+
"learning_rate": 2.5e-05,
|
| 3034 |
+
"loss": 1.2874,
|
| 3035 |
+
"num_input_tokens_seen": 343741672,
|
| 3036 |
+
"step": 3750
|
| 3037 |
+
},
|
| 3038 |
+
{
|
| 3039 |
+
"epoch": 0.952727848602423,
|
| 3040 |
+
"grad_norm": 0.23056572675704956,
|
| 3041 |
+
"learning_rate": 2.5e-05,
|
| 3042 |
+
"loss": 1.3069,
|
| 3043 |
+
"num_input_tokens_seen": 344641984,
|
| 3044 |
+
"step": 3760
|
| 3045 |
+
},
|
| 3046 |
+
{
|
| 3047 |
+
"epoch": 0.9552616992635996,
|
| 3048 |
+
"grad_norm": 0.2758452892303467,
|
| 3049 |
+
"learning_rate": 2.5e-05,
|
| 3050 |
+
"loss": 1.2951,
|
| 3051 |
+
"num_input_tokens_seen": 345553040,
|
| 3052 |
+
"step": 3770
|
| 3053 |
+
},
|
| 3054 |
+
{
|
| 3055 |
+
"epoch": 0.9577955499247763,
|
| 3056 |
+
"grad_norm": 0.2210364043712616,
|
| 3057 |
+
"learning_rate": 2.5e-05,
|
| 3058 |
+
"loss": 1.288,
|
| 3059 |
+
"num_input_tokens_seen": 346455716,
|
| 3060 |
+
"step": 3780
|
| 3061 |
+
},
|
| 3062 |
+
{
|
| 3063 |
+
"epoch": 0.960329400585953,
|
| 3064 |
+
"grad_norm": 0.24254508316516876,
|
| 3065 |
+
"learning_rate": 2.5e-05,
|
| 3066 |
+
"loss": 1.3527,
|
| 3067 |
+
"num_input_tokens_seen": 347362188,
|
| 3068 |
+
"step": 3790
|
| 3069 |
+
},
|
| 3070 |
+
{
|
| 3071 |
+
"epoch": 0.9628632512471296,
|
| 3072 |
+
"grad_norm": 0.2317672073841095,
|
| 3073 |
+
"learning_rate": 2.5e-05,
|
| 3074 |
+
"loss": 1.2872,
|
| 3075 |
+
"num_input_tokens_seen": 348323636,
|
| 3076 |
+
"step": 3800
|
| 3077 |
+
},
|
| 3078 |
+
{
|
| 3079 |
+
"epoch": 0.9653971019083063,
|
| 3080 |
+
"grad_norm": 0.25921356678009033,
|
| 3081 |
+
"learning_rate": 2.5e-05,
|
| 3082 |
+
"loss": 1.326,
|
| 3083 |
+
"num_input_tokens_seen": 349187636,
|
| 3084 |
+
"step": 3810
|
| 3085 |
+
},
|
| 3086 |
+
{
|
| 3087 |
+
"epoch": 0.9679309525694829,
|
| 3088 |
+
"grad_norm": 0.24803981184959412,
|
| 3089 |
+
"learning_rate": 2.5e-05,
|
| 3090 |
+
"loss": 1.2919,
|
| 3091 |
+
"num_input_tokens_seen": 350146896,
|
| 3092 |
+
"step": 3820
|
| 3093 |
+
},
|
| 3094 |
+
{
|
| 3095 |
+
"epoch": 0.9704648032306596,
|
| 3096 |
+
"grad_norm": 0.27010080218315125,
|
| 3097 |
+
"learning_rate": 2.5e-05,
|
| 3098 |
+
"loss": 1.3511,
|
| 3099 |
+
"num_input_tokens_seen": 351082648,
|
| 3100 |
+
"step": 3830
|
| 3101 |
+
},
|
| 3102 |
+
{
|
| 3103 |
+
"epoch": 0.9729986538918363,
|
| 3104 |
+
"grad_norm": 0.3154395520687103,
|
| 3105 |
+
"learning_rate": 2.5e-05,
|
| 3106 |
+
"loss": 1.328,
|
| 3107 |
+
"num_input_tokens_seen": 351973288,
|
| 3108 |
+
"step": 3840
|
| 3109 |
+
},
|
| 3110 |
+
{
|
| 3111 |
+
"epoch": 0.9755325045530129,
|
| 3112 |
+
"grad_norm": 0.27058759331703186,
|
| 3113 |
+
"learning_rate": 2.5e-05,
|
| 3114 |
+
"loss": 1.2797,
|
| 3115 |
+
"num_input_tokens_seen": 352899120,
|
| 3116 |
+
"step": 3850
|
| 3117 |
+
},
|
| 3118 |
+
{
|
| 3119 |
+
"epoch": 0.9780663552141896,
|
| 3120 |
+
"grad_norm": 0.22412972152233124,
|
| 3121 |
+
"learning_rate": 2.5e-05,
|
| 3122 |
+
"loss": 1.3193,
|
| 3123 |
+
"num_input_tokens_seen": 353825356,
|
| 3124 |
+
"step": 3860
|
| 3125 |
+
},
|
| 3126 |
+
{
|
| 3127 |
+
"epoch": 0.9806002058753662,
|
| 3128 |
+
"grad_norm": 0.3295518755912781,
|
| 3129 |
+
"learning_rate": 2.5e-05,
|
| 3130 |
+
"loss": 1.324,
|
| 3131 |
+
"num_input_tokens_seen": 354778268,
|
| 3132 |
+
"step": 3870
|
| 3133 |
+
},
|
| 3134 |
+
{
|
| 3135 |
+
"epoch": 0.9831340565365428,
|
| 3136 |
+
"grad_norm": 0.20455938577651978,
|
| 3137 |
+
"learning_rate": 2.5e-05,
|
| 3138 |
+
"loss": 1.3359,
|
| 3139 |
+
"num_input_tokens_seen": 355687292,
|
| 3140 |
+
"step": 3880
|
| 3141 |
+
},
|
| 3142 |
+
{
|
| 3143 |
+
"epoch": 0.9856679071977196,
|
| 3144 |
+
"grad_norm": 0.22574731707572937,
|
| 3145 |
+
"learning_rate": 2.5e-05,
|
| 3146 |
+
"loss": 1.3081,
|
| 3147 |
+
"num_input_tokens_seen": 356581252,
|
| 3148 |
+
"step": 3890
|
| 3149 |
+
},
|
| 3150 |
+
{
|
| 3151 |
+
"epoch": 0.9882017578588962,
|
| 3152 |
+
"grad_norm": 0.25318706035614014,
|
| 3153 |
+
"learning_rate": 2.5e-05,
|
| 3154 |
+
"loss": 1.3327,
|
| 3155 |
+
"num_input_tokens_seen": 357531400,
|
| 3156 |
+
"step": 3900
|
| 3157 |
+
},
|
| 3158 |
+
{
|
| 3159 |
+
"epoch": 0.9907356085200728,
|
| 3160 |
+
"grad_norm": 0.25423163175582886,
|
| 3161 |
+
"learning_rate": 2.5e-05,
|
| 3162 |
+
"loss": 1.3269,
|
| 3163 |
+
"num_input_tokens_seen": 358429676,
|
| 3164 |
+
"step": 3910
|
| 3165 |
+
},
|
| 3166 |
+
{
|
| 3167 |
+
"epoch": 0.9932694591812495,
|
| 3168 |
+
"grad_norm": 0.23770791292190552,
|
| 3169 |
+
"learning_rate": 2.5e-05,
|
| 3170 |
+
"loss": 1.2942,
|
| 3171 |
+
"num_input_tokens_seen": 359328932,
|
| 3172 |
+
"step": 3920
|
| 3173 |
+
},
|
| 3174 |
+
{
|
| 3175 |
+
"epoch": 0.9958033098424262,
|
| 3176 |
+
"grad_norm": 0.23878265917301178,
|
| 3177 |
+
"learning_rate": 2.5e-05,
|
| 3178 |
+
"loss": 1.3295,
|
| 3179 |
+
"num_input_tokens_seen": 360264552,
|
| 3180 |
+
"step": 3930
|
| 3181 |
+
},
|
| 3182 |
+
{
|
| 3183 |
+
"epoch": 0.9983371605036028,
|
| 3184 |
+
"grad_norm": 0.2264624685049057,
|
| 3185 |
+
"learning_rate": 2.5e-05,
|
| 3186 |
+
"loss": 1.3224,
|
| 3187 |
+
"num_input_tokens_seen": 361179728,
|
| 3188 |
+
"step": 3940
|
| 3189 |
+
},
|
| 3190 |
+
{
|
| 3191 |
+
"epoch": 0.9998574709003089,
|
| 3192 |
+
"num_input_tokens_seen": 361724696,
|
| 3193 |
+
"step": 3946,
|
| 3194 |
+
"total_flos": 1.4115183327245763e+18,
|
| 3195 |
+
"train_loss": 1.3691888915302182,
|
| 3196 |
+
"train_runtime": 65409.2818,
|
| 3197 |
+
"train_samples_per_second": 15.446,
|
| 3198 |
+
"train_steps_per_second": 0.06
|
| 3199 |
+
}
|
| 3200 |
+
],
|
| 3201 |
+
"logging_steps": 10,
|
| 3202 |
+
"max_steps": 3946,
|
| 3203 |
+
"num_input_tokens_seen": 361724696,
|
| 3204 |
+
"num_train_epochs": 1,
|
| 3205 |
+
"save_steps": 500,
|
| 3206 |
+
"stateful_callbacks": {
|
| 3207 |
+
"TrainerControl": {
|
| 3208 |
+
"args": {
|
| 3209 |
+
"should_epoch_stop": false,
|
| 3210 |
+
"should_evaluate": false,
|
| 3211 |
+
"should_log": false,
|
| 3212 |
+
"should_save": true,
|
| 3213 |
+
"should_training_stop": true
|
| 3214 |
+
},
|
| 3215 |
+
"attributes": {}
|
| 3216 |
+
}
|
| 3217 |
+
},
|
| 3218 |
+
"total_flos": 1.4115183327245763e+18,
|
| 3219 |
+
"train_batch_size": 4,
|
| 3220 |
+
"trial_name": null,
|
| 3221 |
+
"trial_params": null
|
| 3222 |
+
}
|