Training in progress, step 450, checkpoint
Browse files
last-checkpoint/README.md
CHANGED
|
@@ -1,14 +1,6 @@
|
|
| 1 |
---
|
| 2 |
base_model: unsloth/phi-4-reasoning-unsloth-bnb-4bit
|
| 3 |
library_name: peft
|
| 4 |
-
pipeline_tag: text-generation
|
| 5 |
-
tags:
|
| 6 |
-
- base_model:adapter:unsloth/phi-4-reasoning-unsloth-bnb-4bit
|
| 7 |
-
- lora
|
| 8 |
-
- sft
|
| 9 |
-
- transformers
|
| 10 |
-
- trl
|
| 11 |
-
- unsloth
|
| 12 |
---
|
| 13 |
|
| 14 |
# Model Card for Model ID
|
|
@@ -207,5 +199,4 @@ Carbon emissions can be estimated using the [Machine Learning Impact calculator]
|
|
| 207 |
[More Information Needed]
|
| 208 |
### Framework versions
|
| 209 |
|
| 210 |
-
- PEFT 0.14.0
|
| 211 |
-
- PEFT 0.18.1
|
|
|
|
| 1 |
---
|
| 2 |
base_model: unsloth/phi-4-reasoning-unsloth-bnb-4bit
|
| 3 |
library_name: peft
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 4 |
---
|
| 5 |
|
| 6 |
# Model Card for Model ID
|
|
|
|
| 199 |
[More Information Needed]
|
| 200 |
### Framework versions
|
| 201 |
|
| 202 |
+
- PEFT 0.14.0
|
|
|
last-checkpoint/adapter_model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 170415112
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:46446edb55375c8e30837b741ea6efdf301ab5e7c33a3e01a2228e7f2984fcfa
|
| 3 |
size 170415112
|
last-checkpoint/optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 86718091
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:a6a68463284b4ad2a06822b70257392d6961fee2250f7ab39d5edb82b4dcbafc
|
| 3 |
size 86718091
|
last-checkpoint/rng_state.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 14645
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:718a0f3db00824213036a2c0441849791319b7d9cf189065873bb26a7020738e
|
| 3 |
size 14645
|
last-checkpoint/scheduler.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1465
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:0c2fa43c6f5c9db389e161efa317fdc098f9dc594a3395c416087750a9a40f32
|
| 3 |
size 1465
|
last-checkpoint/trainer_state.json
CHANGED
|
@@ -2,9 +2,9 @@
|
|
| 2 |
"best_global_step": null,
|
| 3 |
"best_metric": null,
|
| 4 |
"best_model_checkpoint": null,
|
| 5 |
-
"epoch":
|
| 6 |
"eval_steps": 500,
|
| 7 |
-
"global_step":
|
| 8 |
"is_hyper_param_search": false,
|
| 9 |
"is_local_process_zero": true,
|
| 10 |
"is_world_process_zero": true,
|
|
@@ -2108,6 +2108,1056 @@
|
|
| 2108 |
"learning_rate": 6.786516853932583e-05,
|
| 2109 |
"loss": 0.8799,
|
| 2110 |
"step": 300
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 2111 |
}
|
| 2112 |
],
|
| 2113 |
"logging_steps": 1,
|
|
@@ -2122,12 +3172,12 @@
|
|
| 2122 |
"should_evaluate": false,
|
| 2123 |
"should_log": false,
|
| 2124 |
"should_save": true,
|
| 2125 |
-
"should_training_stop":
|
| 2126 |
},
|
| 2127 |
"attributes": {}
|
| 2128 |
}
|
| 2129 |
},
|
| 2130 |
-
"total_flos":
|
| 2131 |
"train_batch_size": 16,
|
| 2132 |
"trial_name": null,
|
| 2133 |
"trial_params": null
|
|
|
|
| 2 |
"best_global_step": null,
|
| 3 |
"best_metric": null,
|
| 4 |
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 5.0,
|
| 6 |
"eval_steps": 500,
|
| 7 |
+
"global_step": 450,
|
| 8 |
"is_hyper_param_search": false,
|
| 9 |
"is_local_process_zero": true,
|
| 10 |
"is_world_process_zero": true,
|
|
|
|
| 2108 |
"learning_rate": 6.786516853932583e-05,
|
| 2109 |
"loss": 0.8799,
|
| 2110 |
"step": 300
|
| 2111 |
+
},
|
| 2112 |
+
{
|
| 2113 |
+
"epoch": 3.3444444444444446,
|
| 2114 |
+
"grad_norm": 0.06960175186395645,
|
| 2115 |
+
"learning_rate": 6.741573033707866e-05,
|
| 2116 |
+
"loss": 0.8587,
|
| 2117 |
+
"step": 301
|
| 2118 |
+
},
|
| 2119 |
+
{
|
| 2120 |
+
"epoch": 3.3555555555555556,
|
| 2121 |
+
"grad_norm": 0.06346665322780609,
|
| 2122 |
+
"learning_rate": 6.696629213483147e-05,
|
| 2123 |
+
"loss": 0.8311,
|
| 2124 |
+
"step": 302
|
| 2125 |
+
},
|
| 2126 |
+
{
|
| 2127 |
+
"epoch": 3.3666666666666667,
|
| 2128 |
+
"grad_norm": 0.06519903242588043,
|
| 2129 |
+
"learning_rate": 6.651685393258428e-05,
|
| 2130 |
+
"loss": 0.8537,
|
| 2131 |
+
"step": 303
|
| 2132 |
+
},
|
| 2133 |
+
{
|
| 2134 |
+
"epoch": 3.3777777777777778,
|
| 2135 |
+
"grad_norm": 0.060187116265296936,
|
| 2136 |
+
"learning_rate": 6.606741573033708e-05,
|
| 2137 |
+
"loss": 0.8614,
|
| 2138 |
+
"step": 304
|
| 2139 |
+
},
|
| 2140 |
+
{
|
| 2141 |
+
"epoch": 3.388888888888889,
|
| 2142 |
+
"grad_norm": 0.06387775391340256,
|
| 2143 |
+
"learning_rate": 6.561797752808989e-05,
|
| 2144 |
+
"loss": 0.8795,
|
| 2145 |
+
"step": 305
|
| 2146 |
+
},
|
| 2147 |
+
{
|
| 2148 |
+
"epoch": 3.4,
|
| 2149 |
+
"grad_norm": 0.06667324900627136,
|
| 2150 |
+
"learning_rate": 6.51685393258427e-05,
|
| 2151 |
+
"loss": 0.8086,
|
| 2152 |
+
"step": 306
|
| 2153 |
+
},
|
| 2154 |
+
{
|
| 2155 |
+
"epoch": 3.411111111111111,
|
| 2156 |
+
"grad_norm": 0.06455954164266586,
|
| 2157 |
+
"learning_rate": 6.47191011235955e-05,
|
| 2158 |
+
"loss": 0.7697,
|
| 2159 |
+
"step": 307
|
| 2160 |
+
},
|
| 2161 |
+
{
|
| 2162 |
+
"epoch": 3.422222222222222,
|
| 2163 |
+
"grad_norm": 0.06988281011581421,
|
| 2164 |
+
"learning_rate": 6.426966292134831e-05,
|
| 2165 |
+
"loss": 0.8821,
|
| 2166 |
+
"step": 308
|
| 2167 |
+
},
|
| 2168 |
+
{
|
| 2169 |
+
"epoch": 3.4333333333333336,
|
| 2170 |
+
"grad_norm": 0.07037835568189621,
|
| 2171 |
+
"learning_rate": 6.382022471910112e-05,
|
| 2172 |
+
"loss": 0.7936,
|
| 2173 |
+
"step": 309
|
| 2174 |
+
},
|
| 2175 |
+
{
|
| 2176 |
+
"epoch": 3.4444444444444446,
|
| 2177 |
+
"grad_norm": 0.060425762087106705,
|
| 2178 |
+
"learning_rate": 6.337078651685394e-05,
|
| 2179 |
+
"loss": 0.8403,
|
| 2180 |
+
"step": 310
|
| 2181 |
+
},
|
| 2182 |
+
{
|
| 2183 |
+
"epoch": 3.4555555555555557,
|
| 2184 |
+
"grad_norm": 0.06361618638038635,
|
| 2185 |
+
"learning_rate": 6.292134831460675e-05,
|
| 2186 |
+
"loss": 0.8605,
|
| 2187 |
+
"step": 311
|
| 2188 |
+
},
|
| 2189 |
+
{
|
| 2190 |
+
"epoch": 3.466666666666667,
|
| 2191 |
+
"grad_norm": 0.06325947493314743,
|
| 2192 |
+
"learning_rate": 6.247191011235956e-05,
|
| 2193 |
+
"loss": 0.9275,
|
| 2194 |
+
"step": 312
|
| 2195 |
+
},
|
| 2196 |
+
{
|
| 2197 |
+
"epoch": 3.477777777777778,
|
| 2198 |
+
"grad_norm": 0.054963476955890656,
|
| 2199 |
+
"learning_rate": 6.202247191011237e-05,
|
| 2200 |
+
"loss": 0.78,
|
| 2201 |
+
"step": 313
|
| 2202 |
+
},
|
| 2203 |
+
{
|
| 2204 |
+
"epoch": 3.488888888888889,
|
| 2205 |
+
"grad_norm": 0.09446276724338531,
|
| 2206 |
+
"learning_rate": 6.157303370786517e-05,
|
| 2207 |
+
"loss": 0.8164,
|
| 2208 |
+
"step": 314
|
| 2209 |
+
},
|
| 2210 |
+
{
|
| 2211 |
+
"epoch": 3.5,
|
| 2212 |
+
"grad_norm": 0.07087241113185883,
|
| 2213 |
+
"learning_rate": 6.112359550561798e-05,
|
| 2214 |
+
"loss": 0.9349,
|
| 2215 |
+
"step": 315
|
| 2216 |
+
},
|
| 2217 |
+
{
|
| 2218 |
+
"epoch": 3.511111111111111,
|
| 2219 |
+
"grad_norm": 0.057490210980176926,
|
| 2220 |
+
"learning_rate": 6.067415730337079e-05,
|
| 2221 |
+
"loss": 0.8346,
|
| 2222 |
+
"step": 316
|
| 2223 |
+
},
|
| 2224 |
+
{
|
| 2225 |
+
"epoch": 3.522222222222222,
|
| 2226 |
+
"grad_norm": 0.05834012106060982,
|
| 2227 |
+
"learning_rate": 6.0224719101123596e-05,
|
| 2228 |
+
"loss": 0.8413,
|
| 2229 |
+
"step": 317
|
| 2230 |
+
},
|
| 2231 |
+
{
|
| 2232 |
+
"epoch": 3.533333333333333,
|
| 2233 |
+
"grad_norm": 0.0664343386888504,
|
| 2234 |
+
"learning_rate": 5.977528089887641e-05,
|
| 2235 |
+
"loss": 0.8908,
|
| 2236 |
+
"step": 318
|
| 2237 |
+
},
|
| 2238 |
+
{
|
| 2239 |
+
"epoch": 3.5444444444444443,
|
| 2240 |
+
"grad_norm": 0.061850935220718384,
|
| 2241 |
+
"learning_rate": 5.932584269662922e-05,
|
| 2242 |
+
"loss": 0.8421,
|
| 2243 |
+
"step": 319
|
| 2244 |
+
},
|
| 2245 |
+
{
|
| 2246 |
+
"epoch": 3.5555555555555554,
|
| 2247 |
+
"grad_norm": 0.07486843317747116,
|
| 2248 |
+
"learning_rate": 5.8876404494382023e-05,
|
| 2249 |
+
"loss": 0.7768,
|
| 2250 |
+
"step": 320
|
| 2251 |
+
},
|
| 2252 |
+
{
|
| 2253 |
+
"epoch": 3.5666666666666664,
|
| 2254 |
+
"grad_norm": 0.06851966679096222,
|
| 2255 |
+
"learning_rate": 5.8426966292134835e-05,
|
| 2256 |
+
"loss": 0.7788,
|
| 2257 |
+
"step": 321
|
| 2258 |
+
},
|
| 2259 |
+
{
|
| 2260 |
+
"epoch": 3.5777777777777775,
|
| 2261 |
+
"grad_norm": 0.06413612514734268,
|
| 2262 |
+
"learning_rate": 5.7977528089887646e-05,
|
| 2263 |
+
"loss": 0.8243,
|
| 2264 |
+
"step": 322
|
| 2265 |
+
},
|
| 2266 |
+
{
|
| 2267 |
+
"epoch": 3.588888888888889,
|
| 2268 |
+
"grad_norm": 0.06908858567476273,
|
| 2269 |
+
"learning_rate": 5.752808988764046e-05,
|
| 2270 |
+
"loss": 0.8331,
|
| 2271 |
+
"step": 323
|
| 2272 |
+
},
|
| 2273 |
+
{
|
| 2274 |
+
"epoch": 3.6,
|
| 2275 |
+
"grad_norm": 0.056398071348667145,
|
| 2276 |
+
"learning_rate": 5.7078651685393256e-05,
|
| 2277 |
+
"loss": 0.8889,
|
| 2278 |
+
"step": 324
|
| 2279 |
+
},
|
| 2280 |
+
{
|
| 2281 |
+
"epoch": 3.611111111111111,
|
| 2282 |
+
"grad_norm": 0.06279837340116501,
|
| 2283 |
+
"learning_rate": 5.6629213483146074e-05,
|
| 2284 |
+
"loss": 0.7433,
|
| 2285 |
+
"step": 325
|
| 2286 |
+
},
|
| 2287 |
+
{
|
| 2288 |
+
"epoch": 3.6222222222222222,
|
| 2289 |
+
"grad_norm": 0.07156965136528015,
|
| 2290 |
+
"learning_rate": 5.6179775280898885e-05,
|
| 2291 |
+
"loss": 0.7733,
|
| 2292 |
+
"step": 326
|
| 2293 |
+
},
|
| 2294 |
+
{
|
| 2295 |
+
"epoch": 3.6333333333333333,
|
| 2296 |
+
"grad_norm": 0.058655962347984314,
|
| 2297 |
+
"learning_rate": 5.573033707865168e-05,
|
| 2298 |
+
"loss": 0.7912,
|
| 2299 |
+
"step": 327
|
| 2300 |
+
},
|
| 2301 |
+
{
|
| 2302 |
+
"epoch": 3.6444444444444444,
|
| 2303 |
+
"grad_norm": 0.06260320544242859,
|
| 2304 |
+
"learning_rate": 5.5280898876404495e-05,
|
| 2305 |
+
"loss": 0.8112,
|
| 2306 |
+
"step": 328
|
| 2307 |
+
},
|
| 2308 |
+
{
|
| 2309 |
+
"epoch": 3.6555555555555554,
|
| 2310 |
+
"grad_norm": 0.08468577265739441,
|
| 2311 |
+
"learning_rate": 5.4831460674157306e-05,
|
| 2312 |
+
"loss": 0.8475,
|
| 2313 |
+
"step": 329
|
| 2314 |
+
},
|
| 2315 |
+
{
|
| 2316 |
+
"epoch": 3.6666666666666665,
|
| 2317 |
+
"grad_norm": 0.06323560327291489,
|
| 2318 |
+
"learning_rate": 5.438202247191011e-05,
|
| 2319 |
+
"loss": 0.9234,
|
| 2320 |
+
"step": 330
|
| 2321 |
+
},
|
| 2322 |
+
{
|
| 2323 |
+
"epoch": 3.677777777777778,
|
| 2324 |
+
"grad_norm": 0.0636834055185318,
|
| 2325 |
+
"learning_rate": 5.393258426966292e-05,
|
| 2326 |
+
"loss": 0.8859,
|
| 2327 |
+
"step": 331
|
| 2328 |
+
},
|
| 2329 |
+
{
|
| 2330 |
+
"epoch": 3.688888888888889,
|
| 2331 |
+
"grad_norm": 0.06598466634750366,
|
| 2332 |
+
"learning_rate": 5.3483146067415734e-05,
|
| 2333 |
+
"loss": 0.8385,
|
| 2334 |
+
"step": 332
|
| 2335 |
+
},
|
| 2336 |
+
{
|
| 2337 |
+
"epoch": 3.7,
|
| 2338 |
+
"grad_norm": 0.0580470897257328,
|
| 2339 |
+
"learning_rate": 5.3033707865168545e-05,
|
| 2340 |
+
"loss": 0.8485,
|
| 2341 |
+
"step": 333
|
| 2342 |
+
},
|
| 2343 |
+
{
|
| 2344 |
+
"epoch": 3.7111111111111112,
|
| 2345 |
+
"grad_norm": 0.06465502083301544,
|
| 2346 |
+
"learning_rate": 5.258426966292135e-05,
|
| 2347 |
+
"loss": 0.8595,
|
| 2348 |
+
"step": 334
|
| 2349 |
+
},
|
| 2350 |
+
{
|
| 2351 |
+
"epoch": 3.7222222222222223,
|
| 2352 |
+
"grad_norm": 0.06161164864897728,
|
| 2353 |
+
"learning_rate": 5.213483146067416e-05,
|
| 2354 |
+
"loss": 0.8341,
|
| 2355 |
+
"step": 335
|
| 2356 |
+
},
|
| 2357 |
+
{
|
| 2358 |
+
"epoch": 3.7333333333333334,
|
| 2359 |
+
"grad_norm": 0.07357273995876312,
|
| 2360 |
+
"learning_rate": 5.168539325842697e-05,
|
| 2361 |
+
"loss": 0.7917,
|
| 2362 |
+
"step": 336
|
| 2363 |
+
},
|
| 2364 |
+
{
|
| 2365 |
+
"epoch": 3.7444444444444445,
|
| 2366 |
+
"grad_norm": 0.06672197580337524,
|
| 2367 |
+
"learning_rate": 5.123595505617977e-05,
|
| 2368 |
+
"loss": 0.7472,
|
| 2369 |
+
"step": 337
|
| 2370 |
+
},
|
| 2371 |
+
{
|
| 2372 |
+
"epoch": 3.7555555555555555,
|
| 2373 |
+
"grad_norm": 0.06582935154438019,
|
| 2374 |
+
"learning_rate": 5.078651685393259e-05,
|
| 2375 |
+
"loss": 0.9154,
|
| 2376 |
+
"step": 338
|
| 2377 |
+
},
|
| 2378 |
+
{
|
| 2379 |
+
"epoch": 3.7666666666666666,
|
| 2380 |
+
"grad_norm": 0.06434327363967896,
|
| 2381 |
+
"learning_rate": 5.03370786516854e-05,
|
| 2382 |
+
"loss": 0.7815,
|
| 2383 |
+
"step": 339
|
| 2384 |
+
},
|
| 2385 |
+
{
|
| 2386 |
+
"epoch": 3.7777777777777777,
|
| 2387 |
+
"grad_norm": 0.06618902832269669,
|
| 2388 |
+
"learning_rate": 4.9887640449438205e-05,
|
| 2389 |
+
"loss": 0.8484,
|
| 2390 |
+
"step": 340
|
| 2391 |
+
},
|
| 2392 |
+
{
|
| 2393 |
+
"epoch": 3.7888888888888888,
|
| 2394 |
+
"grad_norm": 0.05907148867845535,
|
| 2395 |
+
"learning_rate": 4.943820224719101e-05,
|
| 2396 |
+
"loss": 0.8342,
|
| 2397 |
+
"step": 341
|
| 2398 |
+
},
|
| 2399 |
+
{
|
| 2400 |
+
"epoch": 3.8,
|
| 2401 |
+
"grad_norm": 0.0687561109662056,
|
| 2402 |
+
"learning_rate": 4.898876404494382e-05,
|
| 2403 |
+
"loss": 0.7931,
|
| 2404 |
+
"step": 342
|
| 2405 |
+
},
|
| 2406 |
+
{
|
| 2407 |
+
"epoch": 3.811111111111111,
|
| 2408 |
+
"grad_norm": 0.06155551224946976,
|
| 2409 |
+
"learning_rate": 4.853932584269663e-05,
|
| 2410 |
+
"loss": 0.8265,
|
| 2411 |
+
"step": 343
|
| 2412 |
+
},
|
| 2413 |
+
{
|
| 2414 |
+
"epoch": 3.822222222222222,
|
| 2415 |
+
"grad_norm": 0.06934966892004013,
|
| 2416 |
+
"learning_rate": 4.808988764044944e-05,
|
| 2417 |
+
"loss": 0.8432,
|
| 2418 |
+
"step": 344
|
| 2419 |
+
},
|
| 2420 |
+
{
|
| 2421 |
+
"epoch": 3.8333333333333335,
|
| 2422 |
+
"grad_norm": 0.06328413635492325,
|
| 2423 |
+
"learning_rate": 4.764044943820225e-05,
|
| 2424 |
+
"loss": 0.9138,
|
| 2425 |
+
"step": 345
|
| 2426 |
+
},
|
| 2427 |
+
{
|
| 2428 |
+
"epoch": 3.8444444444444446,
|
| 2429 |
+
"grad_norm": 0.07244782894849777,
|
| 2430 |
+
"learning_rate": 4.719101123595506e-05,
|
| 2431 |
+
"loss": 0.8205,
|
| 2432 |
+
"step": 346
|
| 2433 |
+
},
|
| 2434 |
+
{
|
| 2435 |
+
"epoch": 3.8555555555555556,
|
| 2436 |
+
"grad_norm": 0.06606698781251907,
|
| 2437 |
+
"learning_rate": 4.674157303370787e-05,
|
| 2438 |
+
"loss": 0.7833,
|
| 2439 |
+
"step": 347
|
| 2440 |
+
},
|
| 2441 |
+
{
|
| 2442 |
+
"epoch": 3.8666666666666667,
|
| 2443 |
+
"grad_norm": 0.06491662561893463,
|
| 2444 |
+
"learning_rate": 4.629213483146068e-05,
|
| 2445 |
+
"loss": 0.834,
|
| 2446 |
+
"step": 348
|
| 2447 |
+
},
|
| 2448 |
+
{
|
| 2449 |
+
"epoch": 3.8777777777777778,
|
| 2450 |
+
"grad_norm": 0.059926439076662064,
|
| 2451 |
+
"learning_rate": 4.584269662921348e-05,
|
| 2452 |
+
"loss": 0.7969,
|
| 2453 |
+
"step": 349
|
| 2454 |
+
},
|
| 2455 |
+
{
|
| 2456 |
+
"epoch": 3.888888888888889,
|
| 2457 |
+
"grad_norm": 0.060968831181526184,
|
| 2458 |
+
"learning_rate": 4.539325842696629e-05,
|
| 2459 |
+
"loss": 0.846,
|
| 2460 |
+
"step": 350
|
| 2461 |
+
},
|
| 2462 |
+
{
|
| 2463 |
+
"epoch": 3.9,
|
| 2464 |
+
"grad_norm": 0.07196994125843048,
|
| 2465 |
+
"learning_rate": 4.4943820224719104e-05,
|
| 2466 |
+
"loss": 0.8191,
|
| 2467 |
+
"step": 351
|
| 2468 |
+
},
|
| 2469 |
+
{
|
| 2470 |
+
"epoch": 3.911111111111111,
|
| 2471 |
+
"grad_norm": 0.061406128108501434,
|
| 2472 |
+
"learning_rate": 4.4494382022471916e-05,
|
| 2473 |
+
"loss": 0.8128,
|
| 2474 |
+
"step": 352
|
| 2475 |
+
},
|
| 2476 |
+
{
|
| 2477 |
+
"epoch": 3.9222222222222225,
|
| 2478 |
+
"grad_norm": 0.05814112350344658,
|
| 2479 |
+
"learning_rate": 4.404494382022472e-05,
|
| 2480 |
+
"loss": 0.8453,
|
| 2481 |
+
"step": 353
|
| 2482 |
+
},
|
| 2483 |
+
{
|
| 2484 |
+
"epoch": 3.9333333333333336,
|
| 2485 |
+
"grad_norm": 0.06283029168844223,
|
| 2486 |
+
"learning_rate": 4.3595505617977525e-05,
|
| 2487 |
+
"loss": 0.899,
|
| 2488 |
+
"step": 354
|
| 2489 |
+
},
|
| 2490 |
+
{
|
| 2491 |
+
"epoch": 3.9444444444444446,
|
| 2492 |
+
"grad_norm": 0.06355167180299759,
|
| 2493 |
+
"learning_rate": 4.314606741573034e-05,
|
| 2494 |
+
"loss": 0.7727,
|
| 2495 |
+
"step": 355
|
| 2496 |
+
},
|
| 2497 |
+
{
|
| 2498 |
+
"epoch": 3.9555555555555557,
|
| 2499 |
+
"grad_norm": 0.06236105412244797,
|
| 2500 |
+
"learning_rate": 4.269662921348315e-05,
|
| 2501 |
+
"loss": 0.8153,
|
| 2502 |
+
"step": 356
|
| 2503 |
+
},
|
| 2504 |
+
{
|
| 2505 |
+
"epoch": 3.966666666666667,
|
| 2506 |
+
"grad_norm": 0.07224887609481812,
|
| 2507 |
+
"learning_rate": 4.224719101123596e-05,
|
| 2508 |
+
"loss": 0.7292,
|
| 2509 |
+
"step": 357
|
| 2510 |
+
},
|
| 2511 |
+
{
|
| 2512 |
+
"epoch": 3.977777777777778,
|
| 2513 |
+
"grad_norm": 0.06908698379993439,
|
| 2514 |
+
"learning_rate": 4.1797752808988764e-05,
|
| 2515 |
+
"loss": 0.7171,
|
| 2516 |
+
"step": 358
|
| 2517 |
+
},
|
| 2518 |
+
{
|
| 2519 |
+
"epoch": 3.988888888888889,
|
| 2520 |
+
"grad_norm": 0.06701194494962692,
|
| 2521 |
+
"learning_rate": 4.1348314606741576e-05,
|
| 2522 |
+
"loss": 0.8333,
|
| 2523 |
+
"step": 359
|
| 2524 |
+
},
|
| 2525 |
+
{
|
| 2526 |
+
"epoch": 4.0,
|
| 2527 |
+
"grad_norm": 0.06868044286966324,
|
| 2528 |
+
"learning_rate": 4.089887640449439e-05,
|
| 2529 |
+
"loss": 0.7593,
|
| 2530 |
+
"step": 360
|
| 2531 |
+
},
|
| 2532 |
+
{
|
| 2533 |
+
"epoch": 4.011111111111111,
|
| 2534 |
+
"grad_norm": 0.059760935604572296,
|
| 2535 |
+
"learning_rate": 4.044943820224719e-05,
|
| 2536 |
+
"loss": 0.8347,
|
| 2537 |
+
"step": 361
|
| 2538 |
+
},
|
| 2539 |
+
{
|
| 2540 |
+
"epoch": 4.022222222222222,
|
| 2541 |
+
"grad_norm": 0.06333184987306595,
|
| 2542 |
+
"learning_rate": 4e-05,
|
| 2543 |
+
"loss": 0.8283,
|
| 2544 |
+
"step": 362
|
| 2545 |
+
},
|
| 2546 |
+
{
|
| 2547 |
+
"epoch": 4.033333333333333,
|
| 2548 |
+
"grad_norm": 0.05701572820544243,
|
| 2549 |
+
"learning_rate": 3.955056179775281e-05,
|
| 2550 |
+
"loss": 0.8445,
|
| 2551 |
+
"step": 363
|
| 2552 |
+
},
|
| 2553 |
+
{
|
| 2554 |
+
"epoch": 4.044444444444444,
|
| 2555 |
+
"grad_norm": 0.057465020567178726,
|
| 2556 |
+
"learning_rate": 3.910112359550562e-05,
|
| 2557 |
+
"loss": 0.8068,
|
| 2558 |
+
"step": 364
|
| 2559 |
+
},
|
| 2560 |
+
{
|
| 2561 |
+
"epoch": 4.055555555555555,
|
| 2562 |
+
"grad_norm": 0.07226772606372833,
|
| 2563 |
+
"learning_rate": 3.865168539325843e-05,
|
| 2564 |
+
"loss": 0.8497,
|
| 2565 |
+
"step": 365
|
| 2566 |
+
},
|
| 2567 |
+
{
|
| 2568 |
+
"epoch": 4.066666666666666,
|
| 2569 |
+
"grad_norm": 0.07886355370283127,
|
| 2570 |
+
"learning_rate": 3.8202247191011236e-05,
|
| 2571 |
+
"loss": 0.8044,
|
| 2572 |
+
"step": 366
|
| 2573 |
+
},
|
| 2574 |
+
{
|
| 2575 |
+
"epoch": 4.0777777777777775,
|
| 2576 |
+
"grad_norm": 0.071531742811203,
|
| 2577 |
+
"learning_rate": 3.775280898876405e-05,
|
| 2578 |
+
"loss": 0.7713,
|
| 2579 |
+
"step": 367
|
| 2580 |
+
},
|
| 2581 |
+
{
|
| 2582 |
+
"epoch": 4.088888888888889,
|
| 2583 |
+
"grad_norm": 0.07198038697242737,
|
| 2584 |
+
"learning_rate": 3.730337078651686e-05,
|
| 2585 |
+
"loss": 0.8685,
|
| 2586 |
+
"step": 368
|
| 2587 |
+
},
|
| 2588 |
+
{
|
| 2589 |
+
"epoch": 4.1,
|
| 2590 |
+
"grad_norm": 0.07044125348329544,
|
| 2591 |
+
"learning_rate": 3.685393258426966e-05,
|
| 2592 |
+
"loss": 0.767,
|
| 2593 |
+
"step": 369
|
| 2594 |
+
},
|
| 2595 |
+
{
|
| 2596 |
+
"epoch": 4.111111111111111,
|
| 2597 |
+
"grad_norm": 0.07338492572307587,
|
| 2598 |
+
"learning_rate": 3.6404494382022475e-05,
|
| 2599 |
+
"loss": 0.7563,
|
| 2600 |
+
"step": 370
|
| 2601 |
+
},
|
| 2602 |
+
{
|
| 2603 |
+
"epoch": 4.122222222222222,
|
| 2604 |
+
"grad_norm": 0.06580597162246704,
|
| 2605 |
+
"learning_rate": 3.595505617977528e-05,
|
| 2606 |
+
"loss": 0.8228,
|
| 2607 |
+
"step": 371
|
| 2608 |
+
},
|
| 2609 |
+
{
|
| 2610 |
+
"epoch": 4.133333333333334,
|
| 2611 |
+
"grad_norm": 0.06712008267641068,
|
| 2612 |
+
"learning_rate": 3.550561797752809e-05,
|
| 2613 |
+
"loss": 0.8644,
|
| 2614 |
+
"step": 372
|
| 2615 |
+
},
|
| 2616 |
+
{
|
| 2617 |
+
"epoch": 4.144444444444445,
|
| 2618 |
+
"grad_norm": 0.0610797144472599,
|
| 2619 |
+
"learning_rate": 3.50561797752809e-05,
|
| 2620 |
+
"loss": 0.7718,
|
| 2621 |
+
"step": 373
|
| 2622 |
+
},
|
| 2623 |
+
{
|
| 2624 |
+
"epoch": 4.155555555555556,
|
| 2625 |
+
"grad_norm": 0.07456668466329575,
|
| 2626 |
+
"learning_rate": 3.460674157303371e-05,
|
| 2627 |
+
"loss": 0.7259,
|
| 2628 |
+
"step": 374
|
| 2629 |
+
},
|
| 2630 |
+
{
|
| 2631 |
+
"epoch": 4.166666666666667,
|
| 2632 |
+
"grad_norm": 0.05761990696191788,
|
| 2633 |
+
"learning_rate": 3.415730337078652e-05,
|
| 2634 |
+
"loss": 0.8152,
|
| 2635 |
+
"step": 375
|
| 2636 |
+
},
|
| 2637 |
+
{
|
| 2638 |
+
"epoch": 4.177777777777778,
|
| 2639 |
+
"grad_norm": 0.06759507954120636,
|
| 2640 |
+
"learning_rate": 3.370786516853933e-05,
|
| 2641 |
+
"loss": 0.8665,
|
| 2642 |
+
"step": 376
|
| 2643 |
+
},
|
| 2644 |
+
{
|
| 2645 |
+
"epoch": 4.188888888888889,
|
| 2646 |
+
"grad_norm": 0.06531881541013718,
|
| 2647 |
+
"learning_rate": 3.325842696629214e-05,
|
| 2648 |
+
"loss": 0.8211,
|
| 2649 |
+
"step": 377
|
| 2650 |
+
},
|
| 2651 |
+
{
|
| 2652 |
+
"epoch": 4.2,
|
| 2653 |
+
"grad_norm": 0.06870734691619873,
|
| 2654 |
+
"learning_rate": 3.2808988764044946e-05,
|
| 2655 |
+
"loss": 0.7547,
|
| 2656 |
+
"step": 378
|
| 2657 |
+
},
|
| 2658 |
+
{
|
| 2659 |
+
"epoch": 4.211111111111111,
|
| 2660 |
+
"grad_norm": 0.06056152656674385,
|
| 2661 |
+
"learning_rate": 3.235955056179775e-05,
|
| 2662 |
+
"loss": 0.7779,
|
| 2663 |
+
"step": 379
|
| 2664 |
+
},
|
| 2665 |
+
{
|
| 2666 |
+
"epoch": 4.222222222222222,
|
| 2667 |
+
"grad_norm": 0.06231197342276573,
|
| 2668 |
+
"learning_rate": 3.191011235955056e-05,
|
| 2669 |
+
"loss": 0.767,
|
| 2670 |
+
"step": 380
|
| 2671 |
+
},
|
| 2672 |
+
{
|
| 2673 |
+
"epoch": 4.233333333333333,
|
| 2674 |
+
"grad_norm": 0.06242848560214043,
|
| 2675 |
+
"learning_rate": 3.1460674157303374e-05,
|
| 2676 |
+
"loss": 0.8473,
|
| 2677 |
+
"step": 381
|
| 2678 |
+
},
|
| 2679 |
+
{
|
| 2680 |
+
"epoch": 4.2444444444444445,
|
| 2681 |
+
"grad_norm": 0.06435809284448624,
|
| 2682 |
+
"learning_rate": 3.1011235955056185e-05,
|
| 2683 |
+
"loss": 0.898,
|
| 2684 |
+
"step": 382
|
| 2685 |
+
},
|
| 2686 |
+
{
|
| 2687 |
+
"epoch": 4.2555555555555555,
|
| 2688 |
+
"grad_norm": 0.06603720039129257,
|
| 2689 |
+
"learning_rate": 3.056179775280899e-05,
|
| 2690 |
+
"loss": 0.8338,
|
| 2691 |
+
"step": 383
|
| 2692 |
+
},
|
| 2693 |
+
{
|
| 2694 |
+
"epoch": 4.266666666666667,
|
| 2695 |
+
"grad_norm": 0.06830265372991562,
|
| 2696 |
+
"learning_rate": 3.0112359550561798e-05,
|
| 2697 |
+
"loss": 0.8323,
|
| 2698 |
+
"step": 384
|
| 2699 |
+
},
|
| 2700 |
+
{
|
| 2701 |
+
"epoch": 4.277777777777778,
|
| 2702 |
+
"grad_norm": 0.06457255780696869,
|
| 2703 |
+
"learning_rate": 2.966292134831461e-05,
|
| 2704 |
+
"loss": 0.8353,
|
| 2705 |
+
"step": 385
|
| 2706 |
+
},
|
| 2707 |
+
{
|
| 2708 |
+
"epoch": 4.288888888888889,
|
| 2709 |
+
"grad_norm": 0.06458742916584015,
|
| 2710 |
+
"learning_rate": 2.9213483146067417e-05,
|
| 2711 |
+
"loss": 0.7949,
|
| 2712 |
+
"step": 386
|
| 2713 |
+
},
|
| 2714 |
+
{
|
| 2715 |
+
"epoch": 4.3,
|
| 2716 |
+
"grad_norm": 0.07112720608711243,
|
| 2717 |
+
"learning_rate": 2.876404494382023e-05,
|
| 2718 |
+
"loss": 0.7822,
|
| 2719 |
+
"step": 387
|
| 2720 |
+
},
|
| 2721 |
+
{
|
| 2722 |
+
"epoch": 4.311111111111111,
|
| 2723 |
+
"grad_norm": 0.06759954988956451,
|
| 2724 |
+
"learning_rate": 2.8314606741573037e-05,
|
| 2725 |
+
"loss": 0.7434,
|
| 2726 |
+
"step": 388
|
| 2727 |
+
},
|
| 2728 |
+
{
|
| 2729 |
+
"epoch": 4.322222222222222,
|
| 2730 |
+
"grad_norm": 0.061707496643066406,
|
| 2731 |
+
"learning_rate": 2.786516853932584e-05,
|
| 2732 |
+
"loss": 0.7928,
|
| 2733 |
+
"step": 389
|
| 2734 |
+
},
|
| 2735 |
+
{
|
| 2736 |
+
"epoch": 4.333333333333333,
|
| 2737 |
+
"grad_norm": 0.07415860146284103,
|
| 2738 |
+
"learning_rate": 2.7415730337078653e-05,
|
| 2739 |
+
"loss": 0.7564,
|
| 2740 |
+
"step": 390
|
| 2741 |
+
},
|
| 2742 |
+
{
|
| 2743 |
+
"epoch": 4.344444444444444,
|
| 2744 |
+
"grad_norm": 0.06894145160913467,
|
| 2745 |
+
"learning_rate": 2.696629213483146e-05,
|
| 2746 |
+
"loss": 0.7985,
|
| 2747 |
+
"step": 391
|
| 2748 |
+
},
|
| 2749 |
+
{
|
| 2750 |
+
"epoch": 4.355555555555555,
|
| 2751 |
+
"grad_norm": 0.06181568279862404,
|
| 2752 |
+
"learning_rate": 2.6516853932584273e-05,
|
| 2753 |
+
"loss": 0.8546,
|
| 2754 |
+
"step": 392
|
| 2755 |
+
},
|
| 2756 |
+
{
|
| 2757 |
+
"epoch": 4.366666666666666,
|
| 2758 |
+
"grad_norm": 0.0904124304652214,
|
| 2759 |
+
"learning_rate": 2.606741573033708e-05,
|
| 2760 |
+
"loss": 0.8677,
|
| 2761 |
+
"step": 393
|
| 2762 |
+
},
|
| 2763 |
+
{
|
| 2764 |
+
"epoch": 4.377777777777778,
|
| 2765 |
+
"grad_norm": 0.06363783776760101,
|
| 2766 |
+
"learning_rate": 2.5617977528089885e-05,
|
| 2767 |
+
"loss": 0.7788,
|
| 2768 |
+
"step": 394
|
| 2769 |
+
},
|
| 2770 |
+
{
|
| 2771 |
+
"epoch": 4.388888888888889,
|
| 2772 |
+
"grad_norm": 0.06672205030918121,
|
| 2773 |
+
"learning_rate": 2.51685393258427e-05,
|
| 2774 |
+
"loss": 0.8017,
|
| 2775 |
+
"step": 395
|
| 2776 |
+
},
|
| 2777 |
+
{
|
| 2778 |
+
"epoch": 4.4,
|
| 2779 |
+
"grad_norm": 0.07275419682264328,
|
| 2780 |
+
"learning_rate": 2.4719101123595505e-05,
|
| 2781 |
+
"loss": 0.7785,
|
| 2782 |
+
"step": 396
|
| 2783 |
+
},
|
| 2784 |
+
{
|
| 2785 |
+
"epoch": 4.411111111111111,
|
| 2786 |
+
"grad_norm": 0.05966269597411156,
|
| 2787 |
+
"learning_rate": 2.4269662921348316e-05,
|
| 2788 |
+
"loss": 0.8946,
|
| 2789 |
+
"step": 397
|
| 2790 |
+
},
|
| 2791 |
+
{
|
| 2792 |
+
"epoch": 4.4222222222222225,
|
| 2793 |
+
"grad_norm": 0.07346921414136887,
|
| 2794 |
+
"learning_rate": 2.3820224719101125e-05,
|
| 2795 |
+
"loss": 0.8088,
|
| 2796 |
+
"step": 398
|
| 2797 |
+
},
|
| 2798 |
+
{
|
| 2799 |
+
"epoch": 4.433333333333334,
|
| 2800 |
+
"grad_norm": 0.06261128932237625,
|
| 2801 |
+
"learning_rate": 2.3370786516853936e-05,
|
| 2802 |
+
"loss": 0.7987,
|
| 2803 |
+
"step": 399
|
| 2804 |
+
},
|
| 2805 |
+
{
|
| 2806 |
+
"epoch": 4.444444444444445,
|
| 2807 |
+
"grad_norm": 0.06262333691120148,
|
| 2808 |
+
"learning_rate": 2.292134831460674e-05,
|
| 2809 |
+
"loss": 0.8036,
|
| 2810 |
+
"step": 400
|
| 2811 |
+
},
|
| 2812 |
+
{
|
| 2813 |
+
"epoch": 4.455555555555556,
|
| 2814 |
+
"grad_norm": 0.061543941497802734,
|
| 2815 |
+
"learning_rate": 2.2471910112359552e-05,
|
| 2816 |
+
"loss": 0.7783,
|
| 2817 |
+
"step": 401
|
| 2818 |
+
},
|
| 2819 |
+
{
|
| 2820 |
+
"epoch": 4.466666666666667,
|
| 2821 |
+
"grad_norm": 0.061959441751241684,
|
| 2822 |
+
"learning_rate": 2.202247191011236e-05,
|
| 2823 |
+
"loss": 0.8373,
|
| 2824 |
+
"step": 402
|
| 2825 |
+
},
|
| 2826 |
+
{
|
| 2827 |
+
"epoch": 4.477777777777778,
|
| 2828 |
+
"grad_norm": 0.06770654022693634,
|
| 2829 |
+
"learning_rate": 2.157303370786517e-05,
|
| 2830 |
+
"loss": 0.843,
|
| 2831 |
+
"step": 403
|
| 2832 |
+
},
|
| 2833 |
+
{
|
| 2834 |
+
"epoch": 4.488888888888889,
|
| 2835 |
+
"grad_norm": 0.06563393771648407,
|
| 2836 |
+
"learning_rate": 2.112359550561798e-05,
|
| 2837 |
+
"loss": 0.7825,
|
| 2838 |
+
"step": 404
|
| 2839 |
+
},
|
| 2840 |
+
{
|
| 2841 |
+
"epoch": 4.5,
|
| 2842 |
+
"grad_norm": 0.07568306475877762,
|
| 2843 |
+
"learning_rate": 2.0674157303370788e-05,
|
| 2844 |
+
"loss": 0.8061,
|
| 2845 |
+
"step": 405
|
| 2846 |
+
},
|
| 2847 |
+
{
|
| 2848 |
+
"epoch": 4.511111111111111,
|
| 2849 |
+
"grad_norm": 0.06064269691705704,
|
| 2850 |
+
"learning_rate": 2.0224719101123596e-05,
|
| 2851 |
+
"loss": 0.7677,
|
| 2852 |
+
"step": 406
|
| 2853 |
+
},
|
| 2854 |
+
{
|
| 2855 |
+
"epoch": 4.522222222222222,
|
| 2856 |
+
"grad_norm": 0.06697507947683334,
|
| 2857 |
+
"learning_rate": 1.9775280898876404e-05,
|
| 2858 |
+
"loss": 0.838,
|
| 2859 |
+
"step": 407
|
| 2860 |
+
},
|
| 2861 |
+
{
|
| 2862 |
+
"epoch": 4.533333333333333,
|
| 2863 |
+
"grad_norm": 0.07317288964986801,
|
| 2864 |
+
"learning_rate": 1.9325842696629215e-05,
|
| 2865 |
+
"loss": 0.8156,
|
| 2866 |
+
"step": 408
|
| 2867 |
+
},
|
| 2868 |
+
{
|
| 2869 |
+
"epoch": 4.544444444444444,
|
| 2870 |
+
"grad_norm": 0.06405317783355713,
|
| 2871 |
+
"learning_rate": 1.8876404494382024e-05,
|
| 2872 |
+
"loss": 0.8283,
|
| 2873 |
+
"step": 409
|
| 2874 |
+
},
|
| 2875 |
+
{
|
| 2876 |
+
"epoch": 4.555555555555555,
|
| 2877 |
+
"grad_norm": 0.0716167539358139,
|
| 2878 |
+
"learning_rate": 1.842696629213483e-05,
|
| 2879 |
+
"loss": 0.8894,
|
| 2880 |
+
"step": 410
|
| 2881 |
+
},
|
| 2882 |
+
{
|
| 2883 |
+
"epoch": 4.566666666666666,
|
| 2884 |
+
"grad_norm": 0.07410852611064911,
|
| 2885 |
+
"learning_rate": 1.797752808988764e-05,
|
| 2886 |
+
"loss": 0.7339,
|
| 2887 |
+
"step": 411
|
| 2888 |
+
},
|
| 2889 |
+
{
|
| 2890 |
+
"epoch": 4.5777777777777775,
|
| 2891 |
+
"grad_norm": 0.06780914962291718,
|
| 2892 |
+
"learning_rate": 1.752808988764045e-05,
|
| 2893 |
+
"loss": 0.7925,
|
| 2894 |
+
"step": 412
|
| 2895 |
+
},
|
| 2896 |
+
{
|
| 2897 |
+
"epoch": 4.588888888888889,
|
| 2898 |
+
"grad_norm": 0.065445177257061,
|
| 2899 |
+
"learning_rate": 1.707865168539326e-05,
|
| 2900 |
+
"loss": 0.8451,
|
| 2901 |
+
"step": 413
|
| 2902 |
+
},
|
| 2903 |
+
{
|
| 2904 |
+
"epoch": 4.6,
|
| 2905 |
+
"grad_norm": 0.06177813559770584,
|
| 2906 |
+
"learning_rate": 1.662921348314607e-05,
|
| 2907 |
+
"loss": 0.9403,
|
| 2908 |
+
"step": 414
|
| 2909 |
+
},
|
| 2910 |
+
{
|
| 2911 |
+
"epoch": 4.611111111111111,
|
| 2912 |
+
"grad_norm": 0.0611334890127182,
|
| 2913 |
+
"learning_rate": 1.6179775280898875e-05,
|
| 2914 |
+
"loss": 0.8064,
|
| 2915 |
+
"step": 415
|
| 2916 |
+
},
|
| 2917 |
+
{
|
| 2918 |
+
"epoch": 4.622222222222222,
|
| 2919 |
+
"grad_norm": 0.06584127992391586,
|
| 2920 |
+
"learning_rate": 1.5730337078651687e-05,
|
| 2921 |
+
"loss": 0.8344,
|
| 2922 |
+
"step": 416
|
| 2923 |
+
},
|
| 2924 |
+
{
|
| 2925 |
+
"epoch": 4.633333333333333,
|
| 2926 |
+
"grad_norm": 0.06387150287628174,
|
| 2927 |
+
"learning_rate": 1.5280898876404495e-05,
|
| 2928 |
+
"loss": 0.8348,
|
| 2929 |
+
"step": 417
|
| 2930 |
+
},
|
| 2931 |
+
{
|
| 2932 |
+
"epoch": 4.644444444444445,
|
| 2933 |
+
"grad_norm": 0.07291285693645477,
|
| 2934 |
+
"learning_rate": 1.4831460674157305e-05,
|
| 2935 |
+
"loss": 0.8291,
|
| 2936 |
+
"step": 418
|
| 2937 |
+
},
|
| 2938 |
+
{
|
| 2939 |
+
"epoch": 4.655555555555556,
|
| 2940 |
+
"grad_norm": 0.06709293276071548,
|
| 2941 |
+
"learning_rate": 1.4382022471910114e-05,
|
| 2942 |
+
"loss": 0.7971,
|
| 2943 |
+
"step": 419
|
| 2944 |
+
},
|
| 2945 |
+
{
|
| 2946 |
+
"epoch": 4.666666666666667,
|
| 2947 |
+
"grad_norm": 0.06627030670642853,
|
| 2948 |
+
"learning_rate": 1.393258426966292e-05,
|
| 2949 |
+
"loss": 0.7975,
|
| 2950 |
+
"step": 420
|
| 2951 |
+
},
|
| 2952 |
+
{
|
| 2953 |
+
"epoch": 4.677777777777778,
|
| 2954 |
+
"grad_norm": 0.06093061715364456,
|
| 2955 |
+
"learning_rate": 1.348314606741573e-05,
|
| 2956 |
+
"loss": 0.7644,
|
| 2957 |
+
"step": 421
|
| 2958 |
+
},
|
| 2959 |
+
{
|
| 2960 |
+
"epoch": 4.688888888888889,
|
| 2961 |
+
"grad_norm": 0.06513749808073044,
|
| 2962 |
+
"learning_rate": 1.303370786516854e-05,
|
| 2963 |
+
"loss": 0.8221,
|
| 2964 |
+
"step": 422
|
| 2965 |
+
},
|
| 2966 |
+
{
|
| 2967 |
+
"epoch": 4.7,
|
| 2968 |
+
"grad_norm": 0.06582172214984894,
|
| 2969 |
+
"learning_rate": 1.258426966292135e-05,
|
| 2970 |
+
"loss": 0.84,
|
| 2971 |
+
"step": 423
|
| 2972 |
+
},
|
| 2973 |
+
{
|
| 2974 |
+
"epoch": 4.711111111111111,
|
| 2975 |
+
"grad_norm": 0.05726146325469017,
|
| 2976 |
+
"learning_rate": 1.2134831460674158e-05,
|
| 2977 |
+
"loss": 0.8009,
|
| 2978 |
+
"step": 424
|
| 2979 |
+
},
|
| 2980 |
+
{
|
| 2981 |
+
"epoch": 4.722222222222222,
|
| 2982 |
+
"grad_norm": 0.060694370418787,
|
| 2983 |
+
"learning_rate": 1.1685393258426968e-05,
|
| 2984 |
+
"loss": 0.8697,
|
| 2985 |
+
"step": 425
|
| 2986 |
+
},
|
| 2987 |
+
{
|
| 2988 |
+
"epoch": 4.733333333333333,
|
| 2989 |
+
"grad_norm": 0.0613742358982563,
|
| 2990 |
+
"learning_rate": 1.1235955056179776e-05,
|
| 2991 |
+
"loss": 0.8642,
|
| 2992 |
+
"step": 426
|
| 2993 |
+
},
|
| 2994 |
+
{
|
| 2995 |
+
"epoch": 4.7444444444444445,
|
| 2996 |
+
"grad_norm": 0.07079113274812698,
|
| 2997 |
+
"learning_rate": 1.0786516853932586e-05,
|
| 2998 |
+
"loss": 0.7765,
|
| 2999 |
+
"step": 427
|
| 3000 |
+
},
|
| 3001 |
+
{
|
| 3002 |
+
"epoch": 4.7555555555555555,
|
| 3003 |
+
"grad_norm": 0.058669183403253555,
|
| 3004 |
+
"learning_rate": 1.0337078651685394e-05,
|
| 3005 |
+
"loss": 0.8284,
|
| 3006 |
+
"step": 428
|
| 3007 |
+
},
|
| 3008 |
+
{
|
| 3009 |
+
"epoch": 4.766666666666667,
|
| 3010 |
+
"grad_norm": 0.058568935841321945,
|
| 3011 |
+
"learning_rate": 9.887640449438202e-06,
|
| 3012 |
+
"loss": 0.8041,
|
| 3013 |
+
"step": 429
|
| 3014 |
+
},
|
| 3015 |
+
{
|
| 3016 |
+
"epoch": 4.777777777777778,
|
| 3017 |
+
"grad_norm": 0.3323858678340912,
|
| 3018 |
+
"learning_rate": 9.438202247191012e-06,
|
| 3019 |
+
"loss": 0.7972,
|
| 3020 |
+
"step": 430
|
| 3021 |
+
},
|
| 3022 |
+
{
|
| 3023 |
+
"epoch": 4.788888888888889,
|
| 3024 |
+
"grad_norm": 0.060114577412605286,
|
| 3025 |
+
"learning_rate": 8.98876404494382e-06,
|
| 3026 |
+
"loss": 0.918,
|
| 3027 |
+
"step": 431
|
| 3028 |
+
},
|
| 3029 |
+
{
|
| 3030 |
+
"epoch": 4.8,
|
| 3031 |
+
"grad_norm": 0.07039128243923187,
|
| 3032 |
+
"learning_rate": 8.53932584269663e-06,
|
| 3033 |
+
"loss": 0.7934,
|
| 3034 |
+
"step": 432
|
| 3035 |
+
},
|
| 3036 |
+
{
|
| 3037 |
+
"epoch": 4.811111111111111,
|
| 3038 |
+
"grad_norm": 0.06390310078859329,
|
| 3039 |
+
"learning_rate": 8.089887640449438e-06,
|
| 3040 |
+
"loss": 0.8117,
|
| 3041 |
+
"step": 433
|
| 3042 |
+
},
|
| 3043 |
+
{
|
| 3044 |
+
"epoch": 4.822222222222222,
|
| 3045 |
+
"grad_norm": 0.06396052241325378,
|
| 3046 |
+
"learning_rate": 7.640449438202247e-06,
|
| 3047 |
+
"loss": 0.7735,
|
| 3048 |
+
"step": 434
|
| 3049 |
+
},
|
| 3050 |
+
{
|
| 3051 |
+
"epoch": 4.833333333333333,
|
| 3052 |
+
"grad_norm": 0.12294171750545502,
|
| 3053 |
+
"learning_rate": 7.191011235955057e-06,
|
| 3054 |
+
"loss": 0.8392,
|
| 3055 |
+
"step": 435
|
| 3056 |
+
},
|
| 3057 |
+
{
|
| 3058 |
+
"epoch": 4.844444444444444,
|
| 3059 |
+
"grad_norm": 0.056867167353630066,
|
| 3060 |
+
"learning_rate": 6.741573033707865e-06,
|
| 3061 |
+
"loss": 0.7539,
|
| 3062 |
+
"step": 436
|
| 3063 |
+
},
|
| 3064 |
+
{
|
| 3065 |
+
"epoch": 4.855555555555555,
|
| 3066 |
+
"grad_norm": 0.06927565485239029,
|
| 3067 |
+
"learning_rate": 6.292134831460675e-06,
|
| 3068 |
+
"loss": 0.7451,
|
| 3069 |
+
"step": 437
|
| 3070 |
+
},
|
| 3071 |
+
{
|
| 3072 |
+
"epoch": 4.866666666666667,
|
| 3073 |
+
"grad_norm": 0.06880547106266022,
|
| 3074 |
+
"learning_rate": 5.842696629213484e-06,
|
| 3075 |
+
"loss": 0.799,
|
| 3076 |
+
"step": 438
|
| 3077 |
+
},
|
| 3078 |
+
{
|
| 3079 |
+
"epoch": 4.877777777777778,
|
| 3080 |
+
"grad_norm": 0.07145966589450836,
|
| 3081 |
+
"learning_rate": 5.393258426966293e-06,
|
| 3082 |
+
"loss": 0.8621,
|
| 3083 |
+
"step": 439
|
| 3084 |
+
},
|
| 3085 |
+
{
|
| 3086 |
+
"epoch": 4.888888888888889,
|
| 3087 |
+
"grad_norm": 0.06943900138139725,
|
| 3088 |
+
"learning_rate": 4.943820224719101e-06,
|
| 3089 |
+
"loss": 0.871,
|
| 3090 |
+
"step": 440
|
| 3091 |
+
},
|
| 3092 |
+
{
|
| 3093 |
+
"epoch": 4.9,
|
| 3094 |
+
"grad_norm": 0.14493992924690247,
|
| 3095 |
+
"learning_rate": 4.49438202247191e-06,
|
| 3096 |
+
"loss": 0.8927,
|
| 3097 |
+
"step": 441
|
| 3098 |
+
},
|
| 3099 |
+
{
|
| 3100 |
+
"epoch": 4.911111111111111,
|
| 3101 |
+
"grad_norm": 0.060017552226781845,
|
| 3102 |
+
"learning_rate": 4.044943820224719e-06,
|
| 3103 |
+
"loss": 0.8933,
|
| 3104 |
+
"step": 442
|
| 3105 |
+
},
|
| 3106 |
+
{
|
| 3107 |
+
"epoch": 4.9222222222222225,
|
| 3108 |
+
"grad_norm": 0.0633043721318245,
|
| 3109 |
+
"learning_rate": 3.5955056179775286e-06,
|
| 3110 |
+
"loss": 0.8325,
|
| 3111 |
+
"step": 443
|
| 3112 |
+
},
|
| 3113 |
+
{
|
| 3114 |
+
"epoch": 4.933333333333334,
|
| 3115 |
+
"grad_norm": 0.06329839676618576,
|
| 3116 |
+
"learning_rate": 3.1460674157303375e-06,
|
| 3117 |
+
"loss": 0.7598,
|
| 3118 |
+
"step": 444
|
| 3119 |
+
},
|
| 3120 |
+
{
|
| 3121 |
+
"epoch": 4.944444444444445,
|
| 3122 |
+
"grad_norm": 0.07061803340911865,
|
| 3123 |
+
"learning_rate": 2.6966292134831465e-06,
|
| 3124 |
+
"loss": 0.7973,
|
| 3125 |
+
"step": 445
|
| 3126 |
+
},
|
| 3127 |
+
{
|
| 3128 |
+
"epoch": 4.955555555555556,
|
| 3129 |
+
"grad_norm": 0.06513971835374832,
|
| 3130 |
+
"learning_rate": 2.247191011235955e-06,
|
| 3131 |
+
"loss": 0.8717,
|
| 3132 |
+
"step": 446
|
| 3133 |
+
},
|
| 3134 |
+
{
|
| 3135 |
+
"epoch": 4.966666666666667,
|
| 3136 |
+
"grad_norm": 0.07594099640846252,
|
| 3137 |
+
"learning_rate": 1.7977528089887643e-06,
|
| 3138 |
+
"loss": 0.7733,
|
| 3139 |
+
"step": 447
|
| 3140 |
+
},
|
| 3141 |
+
{
|
| 3142 |
+
"epoch": 4.977777777777778,
|
| 3143 |
+
"grad_norm": 0.06038981303572655,
|
| 3144 |
+
"learning_rate": 1.3483146067415732e-06,
|
| 3145 |
+
"loss": 0.7951,
|
| 3146 |
+
"step": 448
|
| 3147 |
+
},
|
| 3148 |
+
{
|
| 3149 |
+
"epoch": 4.988888888888889,
|
| 3150 |
+
"grad_norm": 0.061661411076784134,
|
| 3151 |
+
"learning_rate": 8.988764044943822e-07,
|
| 3152 |
+
"loss": 0.7623,
|
| 3153 |
+
"step": 449
|
| 3154 |
+
},
|
| 3155 |
+
{
|
| 3156 |
+
"epoch": 5.0,
|
| 3157 |
+
"grad_norm": 0.05862731486558914,
|
| 3158 |
+
"learning_rate": 4.494382022471911e-07,
|
| 3159 |
+
"loss": 0.819,
|
| 3160 |
+
"step": 450
|
| 3161 |
}
|
| 3162 |
],
|
| 3163 |
"logging_steps": 1,
|
|
|
|
| 3172 |
"should_evaluate": false,
|
| 3173 |
"should_log": false,
|
| 3174 |
"should_save": true,
|
| 3175 |
+
"should_training_stop": true
|
| 3176 |
},
|
| 3177 |
"attributes": {}
|
| 3178 |
}
|
| 3179 |
},
|
| 3180 |
+
"total_flos": 5.021161104605184e+18,
|
| 3181 |
"train_batch_size": 16,
|
| 3182 |
"trial_name": null,
|
| 3183 |
"trial_params": null
|