Model save
Browse files- README.md +4 -7
- all_results.json +7 -7
- config.json +1 -1
- model-00001-of-00004.safetensors +1 -1
- model-00002-of-00004.safetensors +1 -1
- model-00003-of-00004.safetensors +1 -1
- model-00004-of-00004.safetensors +1 -1
- train_results.json +7 -7
- trainer_state.json +2227 -14
- training_args.bin +1 -1
README.md
CHANGED
|
@@ -1,20 +1,17 @@
|
|
| 1 |
---
|
| 2 |
base_model: BSC-LT/salamandra-7b
|
| 3 |
-
datasets:
|
| 4 |
-
- latam-gpt/es-smoltalk
|
| 5 |
library_name: transformers
|
| 6 |
-
model_name:
|
| 7 |
tags:
|
| 8 |
- generated_from_trainer
|
| 9 |
-
- alignment-handbook
|
| 10 |
- trl
|
| 11 |
- sft
|
| 12 |
licence: license
|
| 13 |
---
|
| 14 |
|
| 15 |
-
# Model Card for
|
| 16 |
|
| 17 |
-
This model is a fine-tuned version of [BSC-LT/salamandra-7b](https://huggingface.co/BSC-LT/salamandra-7b)
|
| 18 |
It has been trained using [TRL](https://github.com/huggingface/trl).
|
| 19 |
|
| 20 |
## Quick start
|
|
@@ -30,7 +27,7 @@ print(output["generated_text"])
|
|
| 30 |
|
| 31 |
## Training procedure
|
| 32 |
|
| 33 |
-
[<img src="https://raw.githubusercontent.com/wandb/assets/main/wandb-github-badge-28.svg" alt="Visualize in Weights & Biases" width="150" height="24"/>](https://wandb.ai/sebcif-cenia/huggingface/runs/
|
| 34 |
|
| 35 |
|
| 36 |
This model was trained with SFT.
|
|
|
|
| 1 |
---
|
| 2 |
base_model: BSC-LT/salamandra-7b
|
|
|
|
|
|
|
| 3 |
library_name: transformers
|
| 4 |
+
model_name: salamandra-2B-instruct-smoltalk
|
| 5 |
tags:
|
| 6 |
- generated_from_trainer
|
|
|
|
| 7 |
- trl
|
| 8 |
- sft
|
| 9 |
licence: license
|
| 10 |
---
|
| 11 |
|
| 12 |
+
# Model Card for salamandra-2B-instruct-smoltalk
|
| 13 |
|
| 14 |
+
This model is a fine-tuned version of [BSC-LT/salamandra-7b](https://huggingface.co/BSC-LT/salamandra-7b).
|
| 15 |
It has been trained using [TRL](https://github.com/huggingface/trl).
|
| 16 |
|
| 17 |
## Quick start
|
|
|
|
| 27 |
|
| 28 |
## Training procedure
|
| 29 |
|
| 30 |
+
[<img src="https://raw.githubusercontent.com/wandb/assets/main/wandb-github-badge-28.svg" alt="Visualize in Weights & Biases" width="150" height="24"/>](https://wandb.ai/sebcif-cenia/huggingface/runs/2hr6al95)
|
| 31 |
|
| 32 |
|
| 33 |
This model was trained with SFT.
|
all_results.json
CHANGED
|
@@ -1,9 +1,9 @@
|
|
| 1 |
{
|
| 2 |
-
"epoch":
|
| 3 |
-
"total_flos":
|
| 4 |
-
"train_loss":
|
| 5 |
-
"train_runtime":
|
| 6 |
-
"train_samples":
|
| 7 |
-
"train_samples_per_second": 1.
|
| 8 |
-
"train_steps_per_second": 0.
|
| 9 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"epoch": 1.9981072555205048,
|
| 3 |
+
"total_flos": 7.945866149068014e+18,
|
| 4 |
+
"train_loss": 3.2101790476778542,
|
| 5 |
+
"train_runtime": 19192.2556,
|
| 6 |
+
"train_samples": 149962,
|
| 7 |
+
"train_samples_per_second": 1.982,
|
| 8 |
+
"train_steps_per_second": 0.083
|
| 9 |
}
|
config.json
CHANGED
|
@@ -26,6 +26,6 @@
|
|
| 26 |
"tie_word_embeddings": false,
|
| 27 |
"torch_dtype": "bfloat16",
|
| 28 |
"transformers_version": "4.48.1",
|
| 29 |
-
"use_cache":
|
| 30 |
"vocab_size": 256000
|
| 31 |
}
|
|
|
|
| 26 |
"tie_word_embeddings": false,
|
| 27 |
"torch_dtype": "bfloat16",
|
| 28 |
"transformers_version": "4.48.1",
|
| 29 |
+
"use_cache": false,
|
| 30 |
"vocab_size": 256000
|
| 31 |
}
|
model-00001-of-00004.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 4982973048
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:64a7414784126213b9f63308b449168b4e8cc8509b95f90ad12779b19f11a349
|
| 3 |
size 4982973048
|
model-00002-of-00004.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 4995660232
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:5b1d95823292ec0c8a38aab42207b8b801961d7f19bbe2b6af7c48468533f21d
|
| 3 |
size 4995660232
|
model-00003-of-00004.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 3460482936
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ca2a029d7d00d4cd3adef7fc12c107515d7b403c9b514186e872872e3e5bcf30
|
| 3 |
size 3460482936
|
model-00004-of-00004.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 2097152128
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:2dd8eb7425034444230fef957ab5e195030c02081a8cc496615ae0c73a816e4b
|
| 3 |
size 2097152128
|
train_results.json
CHANGED
|
@@ -1,9 +1,9 @@
|
|
| 1 |
{
|
| 2 |
-
"epoch":
|
| 3 |
-
"total_flos":
|
| 4 |
-
"train_loss":
|
| 5 |
-
"train_runtime":
|
| 6 |
-
"train_samples":
|
| 7 |
-
"train_samples_per_second": 1.
|
| 8 |
-
"train_steps_per_second": 0.
|
| 9 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"epoch": 1.9981072555205048,
|
| 3 |
+
"total_flos": 7.945866149068014e+18,
|
| 4 |
+
"train_loss": 3.2101790476778542,
|
| 5 |
+
"train_runtime": 19192.2556,
|
| 6 |
+
"train_samples": 149962,
|
| 7 |
+
"train_samples_per_second": 1.982,
|
| 8 |
+
"train_steps_per_second": 0.083
|
| 9 |
}
|
trainer_state.json
CHANGED
|
@@ -1,31 +1,2244 @@
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
-
"epoch":
|
| 5 |
"eval_steps": 500,
|
| 6 |
-
"global_step":
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
| 10 |
"log_history": [
|
| 11 |
{
|
| 12 |
-
"epoch":
|
| 13 |
-
"
|
| 14 |
-
"
|
|
|
|
| 15 |
"step": 1
|
| 16 |
},
|
| 17 |
{
|
| 18 |
-
"epoch":
|
| 19 |
-
"
|
| 20 |
-
"
|
| 21 |
-
"
|
| 22 |
-
"
|
| 23 |
-
|
| 24 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 25 |
}
|
| 26 |
],
|
| 27 |
"logging_steps": 5,
|
| 28 |
-
"max_steps":
|
| 29 |
"num_input_tokens_seen": 0,
|
| 30 |
"num_train_epochs": 2,
|
| 31 |
"save_steps": 500,
|
|
@@ -41,7 +2254,7 @@
|
|
| 41 |
"attributes": {}
|
| 42 |
}
|
| 43 |
},
|
| 44 |
-
"total_flos":
|
| 45 |
"train_batch_size": 1,
|
| 46 |
"trial_name": null,
|
| 47 |
"trial_params": null
|
|
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 1.9981072555205048,
|
| 5 |
"eval_steps": 500,
|
| 6 |
+
"global_step": 1584,
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
| 10 |
"log_history": [
|
| 11 |
{
|
| 12 |
+
"epoch": 0.0012618296529968455,
|
| 13 |
+
"grad_norm": 1.1082602107573196,
|
| 14 |
+
"learning_rate": 1.8867924528301887e-06,
|
| 15 |
+
"loss": 1.3619,
|
| 16 |
"step": 1
|
| 17 |
},
|
| 18 |
{
|
| 19 |
+
"epoch": 0.006309148264984227,
|
| 20 |
+
"grad_norm": 0.5160716308680561,
|
| 21 |
+
"learning_rate": 9.433962264150942e-06,
|
| 22 |
+
"loss": 1.2587,
|
| 23 |
+
"step": 5
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.012618296529968454,
|
| 27 |
+
"grad_norm": 0.27778207375653746,
|
| 28 |
+
"learning_rate": 1.8867924528301885e-05,
|
| 29 |
+
"loss": 1.1806,
|
| 30 |
+
"step": 10
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.01892744479495268,
|
| 34 |
+
"grad_norm": 1.1679927805055026,
|
| 35 |
+
"learning_rate": 2.830188679245283e-05,
|
| 36 |
+
"loss": 1.1902,
|
| 37 |
+
"step": 15
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.025236593059936908,
|
| 41 |
+
"grad_norm": 0.21269694979373854,
|
| 42 |
+
"learning_rate": 3.773584905660377e-05,
|
| 43 |
+
"loss": 1.0417,
|
| 44 |
+
"step": 20
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.031545741324921134,
|
| 48 |
+
"grad_norm": 0.211706691454784,
|
| 49 |
+
"learning_rate": 4.716981132075471e-05,
|
| 50 |
+
"loss": 1.0067,
|
| 51 |
+
"step": 25
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.03785488958990536,
|
| 55 |
+
"grad_norm": 0.13721217987506173,
|
| 56 |
+
"learning_rate": 5.660377358490566e-05,
|
| 57 |
+
"loss": 0.9606,
|
| 58 |
+
"step": 30
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.04416403785488959,
|
| 62 |
+
"grad_norm": 0.11351584468319978,
|
| 63 |
+
"learning_rate": 6.60377358490566e-05,
|
| 64 |
+
"loss": 0.9455,
|
| 65 |
+
"step": 35
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.050473186119873815,
|
| 69 |
+
"grad_norm": 0.12496482045243724,
|
| 70 |
+
"learning_rate": 7.547169811320754e-05,
|
| 71 |
+
"loss": 0.9477,
|
| 72 |
+
"step": 40
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.056782334384858045,
|
| 76 |
+
"grad_norm": 0.1298636508268261,
|
| 77 |
+
"learning_rate": 8.490566037735848e-05,
|
| 78 |
+
"loss": 0.9582,
|
| 79 |
+
"step": 45
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.06309148264984227,
|
| 83 |
+
"grad_norm": 0.1467755476821997,
|
| 84 |
+
"learning_rate": 9.433962264150942e-05,
|
| 85 |
+
"loss": 0.9513,
|
| 86 |
+
"step": 50
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.0694006309148265,
|
| 90 |
+
"grad_norm": 0.13151314150620527,
|
| 91 |
+
"learning_rate": 0.00010377358490566036,
|
| 92 |
+
"loss": 0.9261,
|
| 93 |
+
"step": 55
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.07570977917981073,
|
| 97 |
+
"grad_norm": 0.13055155868282162,
|
| 98 |
+
"learning_rate": 0.00011320754716981132,
|
| 99 |
+
"loss": 0.9085,
|
| 100 |
+
"step": 60
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 0.08201892744479496,
|
| 104 |
+
"grad_norm": 0.1310447924936462,
|
| 105 |
+
"learning_rate": 0.00012264150943396227,
|
| 106 |
+
"loss": 0.9426,
|
| 107 |
+
"step": 65
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.08832807570977919,
|
| 111 |
+
"grad_norm": 0.128442056161394,
|
| 112 |
+
"learning_rate": 0.0001320754716981132,
|
| 113 |
+
"loss": 0.8889,
|
| 114 |
+
"step": 70
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.0946372239747634,
|
| 118 |
+
"grad_norm": 0.1271856500470796,
|
| 119 |
+
"learning_rate": 0.00014150943396226413,
|
| 120 |
+
"loss": 0.9279,
|
| 121 |
+
"step": 75
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.10094637223974763,
|
| 125 |
+
"grad_norm": 0.15157546026263177,
|
| 126 |
+
"learning_rate": 0.00015094339622641508,
|
| 127 |
+
"loss": 0.8658,
|
| 128 |
+
"step": 80
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.10725552050473186,
|
| 132 |
+
"grad_norm": 0.12660512962301268,
|
| 133 |
+
"learning_rate": 0.00016037735849056602,
|
| 134 |
+
"loss": 0.9468,
|
| 135 |
+
"step": 85
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.11356466876971609,
|
| 139 |
+
"grad_norm": 0.15115851743166572,
|
| 140 |
+
"learning_rate": 0.00016981132075471697,
|
| 141 |
+
"loss": 0.9333,
|
| 142 |
+
"step": 90
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.11987381703470032,
|
| 146 |
+
"grad_norm": 0.12185915010225536,
|
| 147 |
+
"learning_rate": 0.0001792452830188679,
|
| 148 |
+
"loss": 0.9563,
|
| 149 |
+
"step": 95
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.12618296529968454,
|
| 153 |
+
"grad_norm": 0.15870917924232272,
|
| 154 |
+
"learning_rate": 0.00018867924528301883,
|
| 155 |
+
"loss": 0.96,
|
| 156 |
+
"step": 100
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 0.13249211356466878,
|
| 160 |
+
"grad_norm": 0.26045717159940285,
|
| 161 |
+
"learning_rate": 0.00019811320754716978,
|
| 162 |
+
"loss": 0.938,
|
| 163 |
+
"step": 105
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 0.138801261829653,
|
| 167 |
+
"grad_norm": 0.14033587532787,
|
| 168 |
+
"learning_rate": 0.00020754716981132072,
|
| 169 |
+
"loss": 0.9387,
|
| 170 |
+
"step": 110
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.14511041009463724,
|
| 174 |
+
"grad_norm": 0.15061172203573397,
|
| 175 |
+
"learning_rate": 0.00021698113207547167,
|
| 176 |
+
"loss": 0.9608,
|
| 177 |
+
"step": 115
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.15141955835962145,
|
| 181 |
+
"grad_norm": 0.167702609373909,
|
| 182 |
+
"learning_rate": 0.00022641509433962264,
|
| 183 |
+
"loss": 1.0191,
|
| 184 |
+
"step": 120
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 0.15772870662460567,
|
| 188 |
+
"grad_norm": 0.1508042853813776,
|
| 189 |
+
"learning_rate": 0.0002358490566037736,
|
| 190 |
+
"loss": 0.9555,
|
| 191 |
+
"step": 125
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 0.1640378548895899,
|
| 195 |
+
"grad_norm": 0.2119642871692552,
|
| 196 |
+
"learning_rate": 0.00024528301886792453,
|
| 197 |
+
"loss": 1.0223,
|
| 198 |
+
"step": 130
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 0.17034700315457413,
|
| 202 |
+
"grad_norm": 0.18918503147120555,
|
| 203 |
+
"learning_rate": 0.0002547169811320755,
|
| 204 |
+
"loss": 0.9611,
|
| 205 |
+
"step": 135
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 0.17665615141955837,
|
| 209 |
+
"grad_norm": 0.12816601187099438,
|
| 210 |
+
"learning_rate": 0.0002641509433962264,
|
| 211 |
+
"loss": 0.9742,
|
| 212 |
+
"step": 140
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.1829652996845426,
|
| 216 |
+
"grad_norm": 0.2401662977028052,
|
| 217 |
+
"learning_rate": 0.0002735849056603773,
|
| 218 |
+
"loss": 0.9817,
|
| 219 |
+
"step": 145
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.1892744479495268,
|
| 223 |
+
"grad_norm": 0.4355491352923922,
|
| 224 |
+
"learning_rate": 0.00028301886792452826,
|
| 225 |
+
"loss": 1.0259,
|
| 226 |
+
"step": 150
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 0.19558359621451105,
|
| 230 |
+
"grad_norm": 33.426776604032355,
|
| 231 |
+
"learning_rate": 0.0002924528301886792,
|
| 232 |
+
"loss": 1.7801,
|
| 233 |
+
"step": 155
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 0.20189274447949526,
|
| 237 |
+
"grad_norm": 23.589648572658916,
|
| 238 |
+
"learning_rate": 0.0002999996354718226,
|
| 239 |
+
"loss": 8.488,
|
| 240 |
+
"step": 160
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 0.2082018927444795,
|
| 244 |
+
"grad_norm": 49.43734950307383,
|
| 245 |
+
"learning_rate": 0.0002999868771716451,
|
| 246 |
+
"loss": 14.2628,
|
| 247 |
+
"step": 165
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 0.21451104100946372,
|
| 251 |
+
"grad_norm": 9.659465955823281,
|
| 252 |
+
"learning_rate": 0.00029995589423430187,
|
| 253 |
+
"loss": 10.1064,
|
| 254 |
+
"step": 170
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 0.22082018927444794,
|
| 258 |
+
"grad_norm": 3.232393240271477,
|
| 259 |
+
"learning_rate": 0.0002999066904244743,
|
| 260 |
+
"loss": 8.747,
|
| 261 |
+
"step": 175
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 0.22712933753943218,
|
| 265 |
+
"grad_norm": 2.0596614328111316,
|
| 266 |
+
"learning_rate": 0.0002998392717208292,
|
| 267 |
+
"loss": 7.8676,
|
| 268 |
+
"step": 180
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 0.2334384858044164,
|
| 272 |
+
"grad_norm": 1.4670633636737682,
|
| 273 |
+
"learning_rate": 0.0002997536463152927,
|
| 274 |
+
"loss": 7.7991,
|
| 275 |
+
"step": 185
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 0.23974763406940064,
|
| 279 |
+
"grad_norm": 1.206234973599562,
|
| 280 |
+
"learning_rate": 0.0002996498246120547,
|
| 281 |
+
"loss": 7.6306,
|
| 282 |
+
"step": 190
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 0.24605678233438485,
|
| 286 |
+
"grad_norm": 1.860628033188109,
|
| 287 |
+
"learning_rate": 0.0002995278192263046,
|
| 288 |
+
"loss": 7.5877,
|
| 289 |
+
"step": 195
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 0.25236593059936907,
|
| 293 |
+
"grad_norm": 2.4361387350882953,
|
| 294 |
+
"learning_rate": 0.0002993876449826987,
|
| 295 |
+
"loss": 7.695,
|
| 296 |
+
"step": 200
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 0.2586750788643533,
|
| 300 |
+
"grad_norm": 1.3576011870434976,
|
| 301 |
+
"learning_rate": 0.0002992293189135584,
|
| 302 |
+
"loss": 7.6365,
|
| 303 |
+
"step": 205
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 0.26498422712933756,
|
| 307 |
+
"grad_norm": 1.7464049616707837,
|
| 308 |
+
"learning_rate": 0.0002990528602568014,
|
| 309 |
+
"loss": 7.644,
|
| 310 |
+
"step": 210
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 0.27129337539432175,
|
| 314 |
+
"grad_norm": 1.3351952885823426,
|
| 315 |
+
"learning_rate": 0.0002988582904536033,
|
| 316 |
+
"loss": 7.5704,
|
| 317 |
+
"step": 215
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 0.277602523659306,
|
| 321 |
+
"grad_norm": 7.73259516055185,
|
| 322 |
+
"learning_rate": 0.00029864563314579295,
|
| 323 |
+
"loss": 7.5867,
|
| 324 |
+
"step": 220
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 0.28391167192429023,
|
| 328 |
+
"grad_norm": 2.116522648519502,
|
| 329 |
+
"learning_rate": 0.00029841491417297933,
|
| 330 |
+
"loss": 7.5697,
|
| 331 |
+
"step": 225
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 0.2902208201892745,
|
| 335 |
+
"grad_norm": 0.9672426637939436,
|
| 336 |
+
"learning_rate": 0.00029816616156941235,
|
| 337 |
+
"loss": 7.3697,
|
| 338 |
+
"step": 230
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 0.29652996845425866,
|
| 342 |
+
"grad_norm": 0.9133420299395039,
|
| 343 |
+
"learning_rate": 0.0002978994055605757,
|
| 344 |
+
"loss": 7.2885,
|
| 345 |
+
"step": 235
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 0.3028391167192429,
|
| 349 |
+
"grad_norm": 0.89577573837346,
|
| 350 |
+
"learning_rate": 0.00029761467855951503,
|
| 351 |
+
"loss": 7.197,
|
| 352 |
+
"step": 240
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 0.30914826498422715,
|
| 356 |
+
"grad_norm": 0.6210703156303241,
|
| 357 |
+
"learning_rate": 0.00029731201516289887,
|
| 358 |
+
"loss": 7.1349,
|
| 359 |
+
"step": 245
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 0.31545741324921134,
|
| 363 |
+
"grad_norm": 0.808850185153722,
|
| 364 |
+
"learning_rate": 0.0002969914521468152,
|
| 365 |
+
"loss": 7.0638,
|
| 366 |
+
"step": 250
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 0.3217665615141956,
|
| 370 |
+
"grad_norm": 1.85410405853387,
|
| 371 |
+
"learning_rate": 0.0002966530284623027,
|
| 372 |
+
"loss": 7.0124,
|
| 373 |
+
"step": 255
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 0.3280757097791798,
|
| 377 |
+
"grad_norm": 3.7075003580164614,
|
| 378 |
+
"learning_rate": 0.000296296785230618,
|
| 379 |
+
"loss": 7.075,
|
| 380 |
+
"step": 260
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 0.334384858044164,
|
| 384 |
+
"grad_norm": 2.008079065287935,
|
| 385 |
+
"learning_rate": 0.0002959227657382389,
|
| 386 |
+
"loss": 7.0068,
|
| 387 |
+
"step": 265
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 0.34069400630914826,
|
| 391 |
+
"grad_norm": 1.6477223709054638,
|
| 392 |
+
"learning_rate": 0.0002955310154316049,
|
| 393 |
+
"loss": 6.9867,
|
| 394 |
+
"step": 270
|
| 395 |
+
},
|
| 396 |
+
{
|
| 397 |
+
"epoch": 0.3470031545741325,
|
| 398 |
+
"grad_norm": 1.0664175079492362,
|
| 399 |
+
"learning_rate": 0.00029512158191159516,
|
| 400 |
+
"loss": 6.8709,
|
| 401 |
+
"step": 275
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 0.35331230283911674,
|
| 405 |
+
"grad_norm": 2.5543016292614396,
|
| 406 |
+
"learning_rate": 0.00029469451492774424,
|
| 407 |
+
"loss": 6.8063,
|
| 408 |
+
"step": 280
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"epoch": 0.35962145110410093,
|
| 412 |
+
"grad_norm": 1.338219244815538,
|
| 413 |
+
"learning_rate": 0.0002942498663721975,
|
| 414 |
+
"loss": 6.7921,
|
| 415 |
+
"step": 285
|
| 416 |
+
},
|
| 417 |
+
{
|
| 418 |
+
"epoch": 0.3659305993690852,
|
| 419 |
+
"grad_norm": 0.9958330815807348,
|
| 420 |
+
"learning_rate": 0.0002937876902734056,
|
| 421 |
+
"loss": 6.6918,
|
| 422 |
+
"step": 290
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 0.3722397476340694,
|
| 426 |
+
"grad_norm": 1.027445657732641,
|
| 427 |
+
"learning_rate": 0.0002933080427895596,
|
| 428 |
+
"loss": 6.6539,
|
| 429 |
+
"step": 295
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 0.3785488958990536,
|
| 433 |
+
"grad_norm": 0.9231046972681346,
|
| 434 |
+
"learning_rate": 0.0002928109822017673,
|
| 435 |
+
"loss": 6.5347,
|
| 436 |
+
"step": 300
|
| 437 |
+
},
|
| 438 |
+
{
|
| 439 |
+
"epoch": 0.38485804416403785,
|
| 440 |
+
"grad_norm": 1.0746356041622689,
|
| 441 |
+
"learning_rate": 0.0002922965689069717,
|
| 442 |
+
"loss": 6.5037,
|
| 443 |
+
"step": 305
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 0.3911671924290221,
|
| 447 |
+
"grad_norm": 1.6425297364390479,
|
| 448 |
+
"learning_rate": 0.00029176486541061206,
|
| 449 |
+
"loss": 6.4267,
|
| 450 |
+
"step": 310
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"epoch": 0.39747634069400634,
|
| 454 |
+
"grad_norm": 1.006291003645608,
|
| 455 |
+
"learning_rate": 0.0002912159363190292,
|
| 456 |
+
"loss": 6.2811,
|
| 457 |
+
"step": 315
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 0.4037854889589905,
|
| 461 |
+
"grad_norm": 0.8858373034477988,
|
| 462 |
+
"learning_rate": 0.00029064984833161516,
|
| 463 |
+
"loss": 6.2622,
|
| 464 |
+
"step": 320
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 0.41009463722397477,
|
| 468 |
+
"grad_norm": 1.6818497207777416,
|
| 469 |
+
"learning_rate": 0.00029006667023270875,
|
| 470 |
+
"loss": 6.1731,
|
| 471 |
+
"step": 325
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 0.416403785488959,
|
| 475 |
+
"grad_norm": 1.0530445540331839,
|
| 476 |
+
"learning_rate": 0.00028946647288323766,
|
| 477 |
+
"loss": 6.1119,
|
| 478 |
+
"step": 330
|
| 479 |
+
},
|
| 480 |
+
{
|
| 481 |
+
"epoch": 0.4227129337539432,
|
| 482 |
+
"grad_norm": 0.774910061771403,
|
| 483 |
+
"learning_rate": 0.0002888493292121083,
|
| 484 |
+
"loss": 5.978,
|
| 485 |
+
"step": 335
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 0.42902208201892744,
|
| 489 |
+
"grad_norm": 0.9934460872479443,
|
| 490 |
+
"learning_rate": 0.00028821531420734415,
|
| 491 |
+
"loss": 5.9174,
|
| 492 |
+
"step": 340
|
| 493 |
+
},
|
| 494 |
+
{
|
| 495 |
+
"epoch": 0.4353312302839117,
|
| 496 |
+
"grad_norm": 1.7195239014940356,
|
| 497 |
+
"learning_rate": 0.0002875645049069745,
|
| 498 |
+
"loss": 5.8155,
|
| 499 |
+
"step": 345
|
| 500 |
+
},
|
| 501 |
+
{
|
| 502 |
+
"epoch": 0.4416403785488959,
|
| 503 |
+
"grad_norm": 0.9170318354561573,
|
| 504 |
+
"learning_rate": 0.0002868969803896732,
|
| 505 |
+
"loss": 5.8763,
|
| 506 |
+
"step": 350
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 0.4479495268138801,
|
| 510 |
+
"grad_norm": 0.9805790741458281,
|
| 511 |
+
"learning_rate": 0.00028621282176515055,
|
| 512 |
+
"loss": 5.6656,
|
| 513 |
+
"step": 355
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 0.45425867507886436,
|
| 517 |
+
"grad_norm": 0.6109164727980874,
|
| 518 |
+
"learning_rate": 0.0002855121121642972,
|
| 519 |
+
"loss": 5.6587,
|
| 520 |
+
"step": 360
|
| 521 |
+
},
|
| 522 |
+
{
|
| 523 |
+
"epoch": 0.4605678233438486,
|
| 524 |
+
"grad_norm": 0.9057557867003939,
|
| 525 |
+
"learning_rate": 0.0002847949367290833,
|
| 526 |
+
"loss": 5.5942,
|
| 527 |
+
"step": 365
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 0.4668769716088328,
|
| 531 |
+
"grad_norm": 0.9907647813457101,
|
| 532 |
+
"learning_rate": 0.0002840613826022133,
|
| 533 |
+
"loss": 5.5445,
|
| 534 |
+
"step": 370
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 0.47318611987381703,
|
| 538 |
+
"grad_norm": 0.5830919154959711,
|
| 539 |
+
"learning_rate": 0.000283311538916537,
|
| 540 |
+
"loss": 5.4069,
|
| 541 |
+
"step": 375
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 0.4794952681388013,
|
| 545 |
+
"grad_norm": 0.691363852310782,
|
| 546 |
+
"learning_rate": 0.0002825454967842195,
|
| 547 |
+
"loss": 5.4414,
|
| 548 |
+
"step": 380
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 0.48580441640378547,
|
| 552 |
+
"grad_norm": 1.1283481084112224,
|
| 553 |
+
"learning_rate": 0.0002817633492856699,
|
| 554 |
+
"loss": 5.3356,
|
| 555 |
+
"step": 385
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 0.4921135646687697,
|
| 559 |
+
"grad_norm": 0.7935578168454879,
|
| 560 |
+
"learning_rate": 0.00028096519145823196,
|
| 561 |
+
"loss": 5.2286,
|
| 562 |
+
"step": 390
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"epoch": 0.49842271293375395,
|
| 566 |
+
"grad_norm": 0.843423834337354,
|
| 567 |
+
"learning_rate": 0.0002801511202846354,
|
| 568 |
+
"loss": 5.174,
|
| 569 |
+
"step": 395
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 0.5047318611987381,
|
| 573 |
+
"grad_norm": 0.9784432577887341,
|
| 574 |
+
"learning_rate": 0.00027932123468121263,
|
| 575 |
+
"loss": 5.2025,
|
| 576 |
+
"step": 400
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 0.5110410094637224,
|
| 580 |
+
"grad_norm": 0.8989990967041858,
|
| 581 |
+
"learning_rate": 0.00027847563548587885,
|
| 582 |
+
"loss": 5.0731,
|
| 583 |
+
"step": 405
|
| 584 |
+
},
|
| 585 |
+
{
|
| 586 |
+
"epoch": 0.5173501577287066,
|
| 587 |
+
"grad_norm": 0.9346344837877385,
|
| 588 |
+
"learning_rate": 0.00027761442544587967,
|
| 589 |
+
"loss": 5.0384,
|
| 590 |
+
"step": 410
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 0.5236593059936908,
|
| 594 |
+
"grad_norm": 0.49545785442898166,
|
| 595 |
+
"learning_rate": 0.0002767377092053067,
|
| 596 |
+
"loss": 4.9484,
|
| 597 |
+
"step": 415
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 0.5299684542586751,
|
| 601 |
+
"grad_norm": 0.880285674560389,
|
| 602 |
+
"learning_rate": 0.00027584559329238217,
|
| 603 |
+
"loss": 4.8691,
|
| 604 |
+
"step": 420
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"epoch": 0.5362776025236593,
|
| 608 |
+
"grad_norm": 0.4501963143777921,
|
| 609 |
+
"learning_rate": 0.00027493818610651487,
|
| 610 |
+
"loss": 4.9105,
|
| 611 |
+
"step": 425
|
| 612 |
+
},
|
| 613 |
+
{
|
| 614 |
+
"epoch": 0.5425867507886435,
|
| 615 |
+
"grad_norm": 0.7932997521608476,
|
| 616 |
+
"learning_rate": 0.000274015597905129,
|
| 617 |
+
"loss": 4.7957,
|
| 618 |
+
"step": 430
|
| 619 |
+
},
|
| 620 |
+
{
|
| 621 |
+
"epoch": 0.5488958990536278,
|
| 622 |
+
"grad_norm": 0.734961203169978,
|
| 623 |
+
"learning_rate": 0.00027307794079026664,
|
| 624 |
+
"loss": 4.732,
|
| 625 |
+
"step": 435
|
| 626 |
+
},
|
| 627 |
+
{
|
| 628 |
+
"epoch": 0.555205047318612,
|
| 629 |
+
"grad_norm": 0.5743582344243288,
|
| 630 |
+
"learning_rate": 0.00027212532869496666,
|
| 631 |
+
"loss": 4.628,
|
| 632 |
+
"step": 440
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 0.5615141955835962,
|
| 636 |
+
"grad_norm": 0.6968894990278519,
|
| 637 |
+
"learning_rate": 0.0002711578773694208,
|
| 638 |
+
"loss": 4.7015,
|
| 639 |
+
"step": 445
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 0.5678233438485805,
|
| 643 |
+
"grad_norm": 0.8348614901245119,
|
| 644 |
+
"learning_rate": 0.0002701757043669091,
|
| 645 |
+
"loss": 4.5554,
|
| 646 |
+
"step": 450
|
| 647 |
+
},
|
| 648 |
+
{
|
| 649 |
+
"epoch": 0.5741324921135647,
|
| 650 |
+
"grad_norm": 0.6331658554162226,
|
| 651 |
+
"learning_rate": 0.0002691789290295164,
|
| 652 |
+
"loss": 4.572,
|
| 653 |
+
"step": 455
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 0.580441640378549,
|
| 657 |
+
"grad_norm": 0.6736093987404245,
|
| 658 |
+
"learning_rate": 0.00026816767247363107,
|
| 659 |
+
"loss": 4.6003,
|
| 660 |
+
"step": 460
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"epoch": 0.5867507886435331,
|
| 664 |
+
"grad_norm": 0.8278673778650902,
|
| 665 |
+
"learning_rate": 0.0002671420575752283,
|
| 666 |
+
"loss": 4.5388,
|
| 667 |
+
"step": 465
|
| 668 |
+
},
|
| 669 |
+
{
|
| 670 |
+
"epoch": 0.5930599369085173,
|
| 671 |
+
"grad_norm": 0.5694315474012022,
|
| 672 |
+
"learning_rate": 0.0002661022089549399,
|
| 673 |
+
"loss": 4.4578,
|
| 674 |
+
"step": 470
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 0.5993690851735016,
|
| 678 |
+
"grad_norm": 0.9886088754097552,
|
| 679 |
+
"learning_rate": 0.0002650482529629117,
|
| 680 |
+
"loss": 4.4192,
|
| 681 |
+
"step": 475
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 0.6056782334384858,
|
| 685 |
+
"grad_norm": 0.6252530671125945,
|
| 686 |
+
"learning_rate": 0.0002639803176634513,
|
| 687 |
+
"loss": 4.4356,
|
| 688 |
+
"step": 480
|
| 689 |
+
},
|
| 690 |
+
{
|
| 691 |
+
"epoch": 0.61198738170347,
|
| 692 |
+
"grad_norm": 0.5002790444273935,
|
| 693 |
+
"learning_rate": 0.0002628985328194665,
|
| 694 |
+
"loss": 4.3081,
|
| 695 |
+
"step": 485
|
| 696 |
+
},
|
| 697 |
+
{
|
| 698 |
+
"epoch": 0.6182965299684543,
|
| 699 |
+
"grad_norm": 0.3986249645458119,
|
| 700 |
+
"learning_rate": 0.00026180302987669884,
|
| 701 |
+
"loss": 4.2495,
|
| 702 |
+
"step": 490
|
| 703 |
+
},
|
| 704 |
+
{
|
| 705 |
+
"epoch": 0.6246056782334385,
|
| 706 |
+
"grad_norm": 0.4984003034249134,
|
| 707 |
+
"learning_rate": 0.00026069394194775145,
|
| 708 |
+
"loss": 4.2671,
|
| 709 |
+
"step": 495
|
| 710 |
+
},
|
| 711 |
+
{
|
| 712 |
+
"epoch": 0.6309148264984227,
|
| 713 |
+
"grad_norm": 0.4111986995411334,
|
| 714 |
+
"learning_rate": 0.0002595714037959146,
|
| 715 |
+
"loss": 4.1762,
|
| 716 |
+
"step": 500
|
| 717 |
+
},
|
| 718 |
+
{
|
| 719 |
+
"epoch": 0.637223974763407,
|
| 720 |
+
"grad_norm": 0.7227035077011171,
|
| 721 |
+
"learning_rate": 0.0002584355518187914,
|
| 722 |
+
"loss": 4.1127,
|
| 723 |
+
"step": 505
|
| 724 |
+
},
|
| 725 |
+
{
|
| 726 |
+
"epoch": 0.6435331230283912,
|
| 727 |
+
"grad_norm": 0.474740993649981,
|
| 728 |
+
"learning_rate": 0.0002572865240317238,
|
| 729 |
+
"loss": 4.0635,
|
| 730 |
+
"step": 510
|
| 731 |
+
},
|
| 732 |
+
{
|
| 733 |
+
"epoch": 0.6498422712933754,
|
| 734 |
+
"grad_norm": 0.4999129381732728,
|
| 735 |
+
"learning_rate": 0.00025612446005102304,
|
| 736 |
+
"loss": 4.1065,
|
| 737 |
+
"step": 515
|
| 738 |
+
},
|
| 739 |
+
{
|
| 740 |
+
"epoch": 0.6561514195583596,
|
| 741 |
+
"grad_norm": 0.5327723288519454,
|
| 742 |
+
"learning_rate": 0.0002549495010770048,
|
| 743 |
+
"loss": 4.0074,
|
| 744 |
+
"step": 520
|
| 745 |
+
},
|
| 746 |
+
{
|
| 747 |
+
"epoch": 0.6624605678233438,
|
| 748 |
+
"grad_norm": 0.5992536156066217,
|
| 749 |
+
"learning_rate": 0.00025376178987683237,
|
| 750 |
+
"loss": 4.0433,
|
| 751 |
+
"step": 525
|
| 752 |
+
},
|
| 753 |
+
{
|
| 754 |
+
"epoch": 0.668769716088328,
|
| 755 |
+
"grad_norm": 0.5067116669423483,
|
| 756 |
+
"learning_rate": 0.0002525614707671693,
|
| 757 |
+
"loss": 4.0681,
|
| 758 |
+
"step": 530
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"epoch": 0.6750788643533123,
|
| 762 |
+
"grad_norm": 0.5663917542488102,
|
| 763 |
+
"learning_rate": 0.0002513486895966435,
|
| 764 |
+
"loss": 3.907,
|
| 765 |
+
"step": 535
|
| 766 |
+
},
|
| 767 |
+
{
|
| 768 |
+
"epoch": 0.6813880126182965,
|
| 769 |
+
"grad_norm": 0.5497458747380433,
|
| 770 |
+
"learning_rate": 0.00025012359372812584,
|
| 771 |
+
"loss": 3.9216,
|
| 772 |
+
"step": 540
|
| 773 |
+
},
|
| 774 |
+
{
|
| 775 |
+
"epoch": 0.6876971608832808,
|
| 776 |
+
"grad_norm": 0.4578173313217405,
|
| 777 |
+
"learning_rate": 0.00024888633202082395,
|
| 778 |
+
"loss": 3.9205,
|
| 779 |
+
"step": 545
|
| 780 |
+
},
|
| 781 |
+
{
|
| 782 |
+
"epoch": 0.694006309148265,
|
| 783 |
+
"grad_norm": 0.5282347165265402,
|
| 784 |
+
"learning_rate": 0.0002476370548121948,
|
| 785 |
+
"loss": 3.9088,
|
| 786 |
+
"step": 550
|
| 787 |
+
},
|
| 788 |
+
{
|
| 789 |
+
"epoch": 0.7003154574132492,
|
| 790 |
+
"grad_norm": 0.3645655751956306,
|
| 791 |
+
"learning_rate": 0.00024637591389967746,
|
| 792 |
+
"loss": 3.8664,
|
| 793 |
+
"step": 555
|
| 794 |
+
},
|
| 795 |
+
{
|
| 796 |
+
"epoch": 0.7066246056782335,
|
| 797 |
+
"grad_norm": 0.6415047301966991,
|
| 798 |
+
"learning_rate": 0.0002451030625222484,
|
| 799 |
+
"loss": 3.8176,
|
| 800 |
+
"step": 560
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 0.7129337539432177,
|
| 804 |
+
"grad_norm": 0.4932401471387324,
|
| 805 |
+
"learning_rate": 0.00024381865534180172,
|
| 806 |
+
"loss": 3.835,
|
| 807 |
+
"step": 565
|
| 808 |
+
},
|
| 809 |
+
{
|
| 810 |
+
"epoch": 0.7192429022082019,
|
| 811 |
+
"grad_norm": 0.5082137835809225,
|
| 812 |
+
"learning_rate": 0.00024252284842435663,
|
| 813 |
+
"loss": 3.846,
|
| 814 |
+
"step": 570
|
| 815 |
+
},
|
| 816 |
+
{
|
| 817 |
+
"epoch": 0.7255520504731862,
|
| 818 |
+
"grad_norm": 0.42580999929000063,
|
| 819 |
+
"learning_rate": 0.0002412157992210938,
|
| 820 |
+
"loss": 3.7607,
|
| 821 |
+
"step": 575
|
| 822 |
+
},
|
| 823 |
+
{
|
| 824 |
+
"epoch": 0.7318611987381703,
|
| 825 |
+
"grad_norm": 0.4163261990969965,
|
| 826 |
+
"learning_rate": 0.00023989766654922403,
|
| 827 |
+
"loss": 3.7311,
|
| 828 |
+
"step": 580
|
| 829 |
+
},
|
| 830 |
+
{
|
| 831 |
+
"epoch": 0.7381703470031545,
|
| 832 |
+
"grad_norm": 0.4419000587445661,
|
| 833 |
+
"learning_rate": 0.00023856861057269053,
|
| 834 |
+
"loss": 3.6884,
|
| 835 |
+
"step": 585
|
| 836 |
+
},
|
| 837 |
+
{
|
| 838 |
+
"epoch": 0.7444794952681388,
|
| 839 |
+
"grad_norm": 0.49473363544940896,
|
| 840 |
+
"learning_rate": 0.00023722879278270783,
|
| 841 |
+
"loss": 3.7119,
|
| 842 |
+
"step": 590
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"epoch": 0.750788643533123,
|
| 846 |
+
"grad_norm": 0.4750134734227391,
|
| 847 |
+
"learning_rate": 0.00023587837597813894,
|
| 848 |
+
"loss": 3.6924,
|
| 849 |
+
"step": 595
|
| 850 |
+
},
|
| 851 |
+
{
|
| 852 |
+
"epoch": 0.7570977917981072,
|
| 853 |
+
"grad_norm": 0.5352899791253288,
|
| 854 |
+
"learning_rate": 0.00023451752424571447,
|
| 855 |
+
"loss": 3.6226,
|
| 856 |
+
"step": 600
|
| 857 |
+
},
|
| 858 |
+
{
|
| 859 |
+
"epoch": 0.7634069400630915,
|
| 860 |
+
"grad_norm": 0.3658077318179394,
|
| 861 |
+
"learning_rate": 0.00023314640294009418,
|
| 862 |
+
"loss": 3.588,
|
| 863 |
+
"step": 605
|
| 864 |
+
},
|
| 865 |
+
{
|
| 866 |
+
"epoch": 0.7697160883280757,
|
| 867 |
+
"grad_norm": 0.3257182602541565,
|
| 868 |
+
"learning_rate": 0.00023176517866377544,
|
| 869 |
+
"loss": 3.6478,
|
| 870 |
+
"step": 610
|
| 871 |
+
},
|
| 872 |
+
{
|
| 873 |
+
"epoch": 0.7760252365930599,
|
| 874 |
+
"grad_norm": 0.4987935132988358,
|
| 875 |
+
"learning_rate": 0.00023037401924684946,
|
| 876 |
+
"loss": 3.5799,
|
| 877 |
+
"step": 615
|
| 878 |
+
},
|
| 879 |
+
{
|
| 880 |
+
"epoch": 0.7823343848580442,
|
| 881 |
+
"grad_norm": 0.5046893060693622,
|
| 882 |
+
"learning_rate": 0.00022897309372660871,
|
| 883 |
+
"loss": 3.5487,
|
| 884 |
+
"step": 620
|
| 885 |
+
},
|
| 886 |
+
{
|
| 887 |
+
"epoch": 0.7886435331230284,
|
| 888 |
+
"grad_norm": 0.627872277058514,
|
| 889 |
+
"learning_rate": 0.0002275625723270073,
|
| 890 |
+
"loss": 3.5755,
|
| 891 |
+
"step": 625
|
| 892 |
+
},
|
| 893 |
+
{
|
| 894 |
+
"epoch": 0.7949526813880127,
|
| 895 |
+
"grad_norm": 0.44145140848187653,
|
| 896 |
+
"learning_rate": 0.00022614262643797753,
|
| 897 |
+
"loss": 3.4622,
|
| 898 |
+
"step": 630
|
| 899 |
+
},
|
| 900 |
+
{
|
| 901 |
+
"epoch": 0.8012618296529969,
|
| 902 |
+
"grad_norm": 0.5980504632653905,
|
| 903 |
+
"learning_rate": 0.0002247134285946045,
|
| 904 |
+
"loss": 3.4863,
|
| 905 |
+
"step": 635
|
| 906 |
+
},
|
| 907 |
+
{
|
| 908 |
+
"epoch": 0.807570977917981,
|
| 909 |
+
"grad_norm": 0.4223181490796927,
|
| 910 |
+
"learning_rate": 0.0002232751524561618,
|
| 911 |
+
"loss": 3.5603,
|
| 912 |
+
"step": 640
|
| 913 |
+
},
|
| 914 |
+
{
|
| 915 |
+
"epoch": 0.8138801261829653,
|
| 916 |
+
"grad_norm": 0.4852780910046954,
|
| 917 |
+
"learning_rate": 0.00022182797278501038,
|
| 918 |
+
"loss": 3.5067,
|
| 919 |
+
"step": 645
|
| 920 |
+
},
|
| 921 |
+
{
|
| 922 |
+
"epoch": 0.8201892744479495,
|
| 923 |
+
"grad_norm": 0.38419619171006636,
|
| 924 |
+
"learning_rate": 0.00022037206542536357,
|
| 925 |
+
"loss": 3.5057,
|
| 926 |
+
"step": 650
|
| 927 |
+
},
|
| 928 |
+
{
|
| 929 |
+
"epoch": 0.8264984227129337,
|
| 930 |
+
"grad_norm": 0.38454756093279036,
|
| 931 |
+
"learning_rate": 0.00021890760728192045,
|
| 932 |
+
"loss": 3.4804,
|
| 933 |
+
"step": 655
|
| 934 |
+
},
|
| 935 |
+
{
|
| 936 |
+
"epoch": 0.832807570977918,
|
| 937 |
+
"grad_norm": 0.413509840686649,
|
| 938 |
+
"learning_rate": 0.00021743477629837055,
|
| 939 |
+
"loss": 3.3881,
|
| 940 |
+
"step": 660
|
| 941 |
+
},
|
| 942 |
+
{
|
| 943 |
+
"epoch": 0.8391167192429022,
|
| 944 |
+
"grad_norm": 0.5414526458354519,
|
| 945 |
+
"learning_rate": 0.00021595375143577226,
|
| 946 |
+
"loss": 3.3393,
|
| 947 |
+
"step": 665
|
| 948 |
+
},
|
| 949 |
+
{
|
| 950 |
+
"epoch": 0.8454258675078864,
|
| 951 |
+
"grad_norm": 0.40594225639177905,
|
| 952 |
+
"learning_rate": 0.0002144647126508076,
|
| 953 |
+
"loss": 3.3492,
|
| 954 |
+
"step": 670
|
| 955 |
+
},
|
| 956 |
+
{
|
| 957 |
+
"epoch": 0.8517350157728707,
|
| 958 |
+
"grad_norm": 0.4817436157511412,
|
| 959 |
+
"learning_rate": 0.00021296784087391585,
|
| 960 |
+
"loss": 3.402,
|
| 961 |
+
"step": 675
|
| 962 |
+
},
|
| 963 |
+
{
|
| 964 |
+
"epoch": 0.8580441640378549,
|
| 965 |
+
"grad_norm": 0.44607482695070316,
|
| 966 |
+
"learning_rate": 0.00021146331798730918,
|
| 967 |
+
"loss": 3.2939,
|
| 968 |
+
"step": 680
|
| 969 |
+
},
|
| 970 |
+
{
|
| 971 |
+
"epoch": 0.8643533123028391,
|
| 972 |
+
"grad_norm": 0.4038582978718023,
|
| 973 |
+
"learning_rate": 0.00020995132680287247,
|
| 974 |
+
"loss": 3.4217,
|
| 975 |
+
"step": 685
|
| 976 |
+
},
|
| 977 |
+
{
|
| 978 |
+
"epoch": 0.8706624605678234,
|
| 979 |
+
"grad_norm": 0.3908221120987279,
|
| 980 |
+
"learning_rate": 0.00020843205103995003,
|
| 981 |
+
"loss": 3.2817,
|
| 982 |
+
"step": 690
|
| 983 |
+
},
|
| 984 |
+
{
|
| 985 |
+
"epoch": 0.8769716088328076,
|
| 986 |
+
"grad_norm": 0.3565314025366524,
|
| 987 |
+
"learning_rate": 0.0002069056753030223,
|
| 988 |
+
"loss": 3.2629,
|
| 989 |
+
"step": 695
|
| 990 |
+
},
|
| 991 |
+
{
|
| 992 |
+
"epoch": 0.8832807570977917,
|
| 993 |
+
"grad_norm": 0.3263516177004363,
|
| 994 |
+
"learning_rate": 0.00020537238505927488,
|
| 995 |
+
"loss": 3.2523,
|
| 996 |
+
"step": 700
|
| 997 |
+
},
|
| 998 |
+
{
|
| 999 |
+
"epoch": 0.889589905362776,
|
| 1000 |
+
"grad_norm": 0.45573489205548634,
|
| 1001 |
+
"learning_rate": 0.00020383236661606274,
|
| 1002 |
+
"loss": 3.2533,
|
| 1003 |
+
"step": 705
|
| 1004 |
+
},
|
| 1005 |
+
{
|
| 1006 |
+
"epoch": 0.8958990536277602,
|
| 1007 |
+
"grad_norm": 0.40418914957962,
|
| 1008 |
+
"learning_rate": 0.00020228580709827227,
|
| 1009 |
+
"loss": 3.2037,
|
| 1010 |
+
"step": 710
|
| 1011 |
+
},
|
| 1012 |
+
{
|
| 1013 |
+
"epoch": 0.9022082018927445,
|
| 1014 |
+
"grad_norm": 0.35849283249585384,
|
| 1015 |
+
"learning_rate": 0.0002007328944255842,
|
| 1016 |
+
"loss": 3.236,
|
| 1017 |
+
"step": 715
|
| 1018 |
+
},
|
| 1019 |
+
{
|
| 1020 |
+
"epoch": 0.9085173501577287,
|
| 1021 |
+
"grad_norm": 0.356011163797442,
|
| 1022 |
+
"learning_rate": 0.0001991738172896396,
|
| 1023 |
+
"loss": 3.1218,
|
| 1024 |
+
"step": 720
|
| 1025 |
+
},
|
| 1026 |
+
{
|
| 1027 |
+
"epoch": 0.9148264984227129,
|
| 1028 |
+
"grad_norm": 0.4150175125843562,
|
| 1029 |
+
"learning_rate": 0.00019760876513111276,
|
| 1030 |
+
"loss": 3.0525,
|
| 1031 |
+
"step": 725
|
| 1032 |
+
},
|
| 1033 |
+
{
|
| 1034 |
+
"epoch": 0.9211356466876972,
|
| 1035 |
+
"grad_norm": 0.38930594826186016,
|
| 1036 |
+
"learning_rate": 0.00019603792811669193,
|
| 1037 |
+
"loss": 3.0326,
|
| 1038 |
+
"step": 730
|
| 1039 |
+
},
|
| 1040 |
+
{
|
| 1041 |
+
"epoch": 0.9274447949526814,
|
| 1042 |
+
"grad_norm": 0.35818356422199943,
|
| 1043 |
+
"learning_rate": 0.00019446149711597305,
|
| 1044 |
+
"loss": 3.0956,
|
| 1045 |
+
"step": 735
|
| 1046 |
+
},
|
| 1047 |
+
{
|
| 1048 |
+
"epoch": 0.9337539432176656,
|
| 1049 |
+
"grad_norm": 0.27796871555567926,
|
| 1050 |
+
"learning_rate": 0.00019287966367826726,
|
| 1051 |
+
"loss": 3.1284,
|
| 1052 |
+
"step": 740
|
| 1053 |
+
},
|
| 1054 |
+
{
|
| 1055 |
+
"epoch": 0.9400630914826499,
|
| 1056 |
+
"grad_norm": 0.4199411353676195,
|
| 1057 |
+
"learning_rate": 0.00019129262000932613,
|
| 1058 |
+
"loss": 2.9957,
|
| 1059 |
+
"step": 745
|
| 1060 |
+
},
|
| 1061 |
+
{
|
| 1062 |
+
"epoch": 0.9463722397476341,
|
| 1063 |
+
"grad_norm": 0.3851610269896564,
|
| 1064 |
+
"learning_rate": 0.00018970055894798724,
|
| 1065 |
+
"loss": 3.0016,
|
| 1066 |
+
"step": 750
|
| 1067 |
+
},
|
| 1068 |
+
{
|
| 1069 |
+
"epoch": 0.9526813880126183,
|
| 1070 |
+
"grad_norm": 0.3521675389671091,
|
| 1071 |
+
"learning_rate": 0.0001881036739427424,
|
| 1072 |
+
"loss": 2.9738,
|
| 1073 |
+
"step": 755
|
| 1074 |
+
},
|
| 1075 |
+
{
|
| 1076 |
+
"epoch": 0.9589905362776026,
|
| 1077 |
+
"grad_norm": 0.31279860012428956,
|
| 1078 |
+
"learning_rate": 0.0001865021590282323,
|
| 1079 |
+
"loss": 2.9809,
|
| 1080 |
+
"step": 760
|
| 1081 |
+
},
|
| 1082 |
+
{
|
| 1083 |
+
"epoch": 0.9652996845425867,
|
| 1084 |
+
"grad_norm": 0.4363140634705567,
|
| 1085 |
+
"learning_rate": 0.00018489620880166954,
|
| 1086 |
+
"loss": 2.9942,
|
| 1087 |
+
"step": 765
|
| 1088 |
+
},
|
| 1089 |
+
{
|
| 1090 |
+
"epoch": 0.9716088328075709,
|
| 1091 |
+
"grad_norm": 0.32752889613341674,
|
| 1092 |
+
"learning_rate": 0.00018328601839919366,
|
| 1093 |
+
"loss": 2.9172,
|
| 1094 |
+
"step": 770
|
| 1095 |
+
},
|
| 1096 |
+
{
|
| 1097 |
+
"epoch": 0.9779179810725552,
|
| 1098 |
+
"grad_norm": 0.3911479256262791,
|
| 1099 |
+
"learning_rate": 0.0001816717834721603,
|
| 1100 |
+
"loss": 2.9121,
|
| 1101 |
+
"step": 775
|
| 1102 |
+
},
|
| 1103 |
+
{
|
| 1104 |
+
"epoch": 0.9842271293375394,
|
| 1105 |
+
"grad_norm": 0.44340416579755526,
|
| 1106 |
+
"learning_rate": 0.00018005370016336818,
|
| 1107 |
+
"loss": 2.9145,
|
| 1108 |
+
"step": 780
|
| 1109 |
+
},
|
| 1110 |
+
{
|
| 1111 |
+
"epoch": 0.9905362776025236,
|
| 1112 |
+
"grad_norm": 0.2988734755579885,
|
| 1113 |
+
"learning_rate": 0.0001784319650832258,
|
| 1114 |
+
"loss": 2.9399,
|
| 1115 |
+
"step": 785
|
| 1116 |
+
},
|
| 1117 |
+
{
|
| 1118 |
+
"epoch": 0.9968454258675079,
|
| 1119 |
+
"grad_norm": 0.2952218938014115,
|
| 1120 |
+
"learning_rate": 0.00017680677528586204,
|
| 1121 |
+
"loss": 2.9221,
|
| 1122 |
+
"step": 790
|
| 1123 |
+
},
|
| 1124 |
+
{
|
| 1125 |
+
"epoch": 1.0025236593059936,
|
| 1126 |
+
"grad_norm": 0.43852839245902553,
|
| 1127 |
+
"learning_rate": 0.00017517832824518213,
|
| 1128 |
+
"loss": 2.7204,
|
| 1129 |
+
"step": 795
|
| 1130 |
+
},
|
| 1131 |
+
{
|
| 1132 |
+
"epoch": 1.0088328075709778,
|
| 1133 |
+
"grad_norm": 0.3032280667716131,
|
| 1134 |
+
"learning_rate": 0.00017354682183087336,
|
| 1135 |
+
"loss": 2.7808,
|
| 1136 |
+
"step": 800
|
| 1137 |
+
},
|
| 1138 |
+
{
|
| 1139 |
+
"epoch": 1.0151419558359622,
|
| 1140 |
+
"grad_norm": 0.32705069583903734,
|
| 1141 |
+
"learning_rate": 0.00017191245428436173,
|
| 1142 |
+
"loss": 2.8124,
|
| 1143 |
+
"step": 805
|
| 1144 |
+
},
|
| 1145 |
+
{
|
| 1146 |
+
"epoch": 1.0214511041009464,
|
| 1147 |
+
"grad_norm": 0.38031310820668335,
|
| 1148 |
+
"learning_rate": 0.00017027542419472452,
|
| 1149 |
+
"loss": 2.7273,
|
| 1150 |
+
"step": 810
|
| 1151 |
+
},
|
| 1152 |
+
{
|
| 1153 |
+
"epoch": 1.0277602523659306,
|
| 1154 |
+
"grad_norm": 0.34447230271596563,
|
| 1155 |
+
"learning_rate": 0.0001686359304745598,
|
| 1156 |
+
"loss": 2.8214,
|
| 1157 |
+
"step": 815
|
| 1158 |
+
},
|
| 1159 |
+
{
|
| 1160 |
+
"epoch": 1.0340694006309148,
|
| 1161 |
+
"grad_norm": 0.33403746114727506,
|
| 1162 |
+
"learning_rate": 0.0001669941723358169,
|
| 1163 |
+
"loss": 2.6577,
|
| 1164 |
+
"step": 820
|
| 1165 |
+
},
|
| 1166 |
+
{
|
| 1167 |
+
"epoch": 1.040378548895899,
|
| 1168 |
+
"grad_norm": 0.3263259722913048,
|
| 1169 |
+
"learning_rate": 0.00016535034926559074,
|
| 1170 |
+
"loss": 2.7708,
|
| 1171 |
+
"step": 825
|
| 1172 |
+
},
|
| 1173 |
+
{
|
| 1174 |
+
"epoch": 1.0466876971608832,
|
| 1175 |
+
"grad_norm": 0.30783319833386963,
|
| 1176 |
+
"learning_rate": 0.00016370466100188227,
|
| 1177 |
+
"loss": 2.7126,
|
| 1178 |
+
"step": 830
|
| 1179 |
+
},
|
| 1180 |
+
{
|
| 1181 |
+
"epoch": 1.0529968454258676,
|
| 1182 |
+
"grad_norm": 0.2797600927675117,
|
| 1183 |
+
"learning_rate": 0.0001620573075093288,
|
| 1184 |
+
"loss": 2.7099,
|
| 1185 |
+
"step": 835
|
| 1186 |
+
},
|
| 1187 |
+
{
|
| 1188 |
+
"epoch": 1.0593059936908518,
|
| 1189 |
+
"grad_norm": 0.22581901939440369,
|
| 1190 |
+
"learning_rate": 0.00016040848895490658,
|
| 1191 |
+
"loss": 2.5855,
|
| 1192 |
+
"step": 840
|
| 1193 |
+
},
|
| 1194 |
+
{
|
| 1195 |
+
"epoch": 1.065615141955836,
|
| 1196 |
+
"grad_norm": 0.21088436539127395,
|
| 1197 |
+
"learning_rate": 0.000158758405683609,
|
| 1198 |
+
"loss": 2.664,
|
| 1199 |
+
"step": 845
|
| 1200 |
+
},
|
| 1201 |
+
{
|
| 1202 |
+
"epoch": 1.0719242902208201,
|
| 1203 |
+
"grad_norm": 0.2622166632758035,
|
| 1204 |
+
"learning_rate": 0.00015710725819410282,
|
| 1205 |
+
"loss": 2.711,
|
| 1206 |
+
"step": 850
|
| 1207 |
+
},
|
| 1208 |
+
{
|
| 1209 |
+
"epoch": 1.0782334384858043,
|
| 1210 |
+
"grad_norm": 0.3440088918783455,
|
| 1211 |
+
"learning_rate": 0.00015545524711436614,
|
| 1212 |
+
"loss": 2.6512,
|
| 1213 |
+
"step": 855
|
| 1214 |
+
},
|
| 1215 |
+
{
|
| 1216 |
+
"epoch": 1.0845425867507887,
|
| 1217 |
+
"grad_norm": 0.238575905149363,
|
| 1218 |
+
"learning_rate": 0.00015380257317731033,
|
| 1219 |
+
"loss": 2.6456,
|
| 1220 |
+
"step": 860
|
| 1221 |
+
},
|
| 1222 |
+
{
|
| 1223 |
+
"epoch": 1.090851735015773,
|
| 1224 |
+
"grad_norm": 0.20336568509851155,
|
| 1225 |
+
"learning_rate": 0.00015214943719638933,
|
| 1226 |
+
"loss": 2.5136,
|
| 1227 |
+
"step": 865
|
| 1228 |
+
},
|
| 1229 |
+
{
|
| 1230 |
+
"epoch": 1.0971608832807571,
|
| 1231 |
+
"grad_norm": 0.22797587055296945,
|
| 1232 |
+
"learning_rate": 0.00015049604004119927,
|
| 1233 |
+
"loss": 2.5813,
|
| 1234 |
+
"step": 870
|
| 1235 |
+
},
|
| 1236 |
+
{
|
| 1237 |
+
"epoch": 1.1034700315457413,
|
| 1238 |
+
"grad_norm": 0.2985560450175076,
|
| 1239 |
+
"learning_rate": 0.0001488425826130711,
|
| 1240 |
+
"loss": 2.6549,
|
| 1241 |
+
"step": 875
|
| 1242 |
+
},
|
| 1243 |
+
{
|
| 1244 |
+
"epoch": 1.1097791798107255,
|
| 1245 |
+
"grad_norm": 0.25165243516492625,
|
| 1246 |
+
"learning_rate": 0.00014718926582065934,
|
| 1247 |
+
"loss": 2.6242,
|
| 1248 |
+
"step": 880
|
| 1249 |
+
},
|
| 1250 |
+
{
|
| 1251 |
+
"epoch": 1.1160883280757097,
|
| 1252 |
+
"grad_norm": 0.2112774778496676,
|
| 1253 |
+
"learning_rate": 0.0001455362905555301,
|
| 1254 |
+
"loss": 2.4842,
|
| 1255 |
+
"step": 885
|
| 1256 |
+
},
|
| 1257 |
+
{
|
| 1258 |
+
"epoch": 1.122397476340694,
|
| 1259 |
+
"grad_norm": 0.23763188968183924,
|
| 1260 |
+
"learning_rate": 0.0001438838576677514,
|
| 1261 |
+
"loss": 2.6185,
|
| 1262 |
+
"step": 890
|
| 1263 |
+
},
|
| 1264 |
+
{
|
| 1265 |
+
"epoch": 1.1287066246056783,
|
| 1266 |
+
"grad_norm": 0.33160686110509024,
|
| 1267 |
+
"learning_rate": 0.00014223216794148757,
|
| 1268 |
+
"loss": 2.5439,
|
| 1269 |
+
"step": 895
|
| 1270 |
+
},
|
| 1271 |
+
{
|
| 1272 |
+
"epoch": 1.1350157728706625,
|
| 1273 |
+
"grad_norm": 0.23338458338027596,
|
| 1274 |
+
"learning_rate": 0.000140581422070603,
|
| 1275 |
+
"loss": 2.5387,
|
| 1276 |
+
"step": 900
|
| 1277 |
+
},
|
| 1278 |
+
{
|
| 1279 |
+
"epoch": 1.1413249211356467,
|
| 1280 |
+
"grad_norm": 0.22293918427638346,
|
| 1281 |
+
"learning_rate": 0.00013893182063427572,
|
| 1282 |
+
"loss": 2.4752,
|
| 1283 |
+
"step": 905
|
| 1284 |
+
},
|
| 1285 |
+
{
|
| 1286 |
+
"epoch": 1.1476340694006308,
|
| 1287 |
+
"grad_norm": 0.22068649705463098,
|
| 1288 |
+
"learning_rate": 0.00013728356407262566,
|
| 1289 |
+
"loss": 2.6106,
|
| 1290 |
+
"step": 910
|
| 1291 |
+
},
|
| 1292 |
+
{
|
| 1293 |
+
"epoch": 1.1539432176656153,
|
| 1294 |
+
"grad_norm": 0.6151653752381766,
|
| 1295 |
+
"learning_rate": 0.00013563685266235946,
|
| 1296 |
+
"loss": 2.5522,
|
| 1297 |
+
"step": 915
|
| 1298 |
+
},
|
| 1299 |
+
{
|
| 1300 |
+
"epoch": 1.1602523659305994,
|
| 1301 |
+
"grad_norm": 0.44947734899142877,
|
| 1302 |
+
"learning_rate": 0.00013399188649243512,
|
| 1303 |
+
"loss": 2.5496,
|
| 1304 |
+
"step": 920
|
| 1305 |
+
},
|
| 1306 |
+
{
|
| 1307 |
+
"epoch": 1.1665615141955836,
|
| 1308 |
+
"grad_norm": 0.27091483804754285,
|
| 1309 |
+
"learning_rate": 0.00013234886543974958,
|
| 1310 |
+
"loss": 2.4731,
|
| 1311 |
+
"step": 925
|
| 1312 |
+
},
|
| 1313 |
+
{
|
| 1314 |
+
"epoch": 1.1728706624605678,
|
| 1315 |
+
"grad_norm": 0.21787265409661474,
|
| 1316 |
+
"learning_rate": 0.0001307079891448521,
|
| 1317 |
+
"loss": 2.4506,
|
| 1318 |
+
"step": 930
|
| 1319 |
+
},
|
| 1320 |
+
{
|
| 1321 |
+
"epoch": 1.179179810725552,
|
| 1322 |
+
"grad_norm": 0.19406944924431835,
|
| 1323 |
+
"learning_rate": 0.00012906945698768627,
|
| 1324 |
+
"loss": 2.4742,
|
| 1325 |
+
"step": 935
|
| 1326 |
+
},
|
| 1327 |
+
{
|
| 1328 |
+
"epoch": 1.1854889589905362,
|
| 1329 |
+
"grad_norm": 0.21779879193956866,
|
| 1330 |
+
"learning_rate": 0.00012743346806336371,
|
| 1331 |
+
"loss": 2.4805,
|
| 1332 |
+
"step": 940
|
| 1333 |
+
},
|
| 1334 |
+
{
|
| 1335 |
+
"epoch": 1.1917981072555206,
|
| 1336 |
+
"grad_norm": 0.18762009788555759,
|
| 1337 |
+
"learning_rate": 0.00012580022115797236,
|
| 1338 |
+
"loss": 2.4608,
|
| 1339 |
+
"step": 945
|
| 1340 |
+
},
|
| 1341 |
+
{
|
| 1342 |
+
"epoch": 1.1981072555205048,
|
| 1343 |
+
"grad_norm": 0.22961054100875933,
|
| 1344 |
+
"learning_rate": 0.00012416991472442225,
|
| 1345 |
+
"loss": 2.5478,
|
| 1346 |
+
"step": 950
|
| 1347 |
+
},
|
| 1348 |
+
{
|
| 1349 |
+
"epoch": 1.204416403785489,
|
| 1350 |
+
"grad_norm": 0.25514705150325423,
|
| 1351 |
+
"learning_rate": 0.0001225427468583321,
|
| 1352 |
+
"loss": 2.4253,
|
| 1353 |
+
"step": 955
|
| 1354 |
+
},
|
| 1355 |
+
{
|
| 1356 |
+
"epoch": 1.2107255520504732,
|
| 1357 |
+
"grad_norm": 0.2364921491120275,
|
| 1358 |
+
"learning_rate": 0.00012091891527395877,
|
| 1359 |
+
"loss": 2.4299,
|
| 1360 |
+
"step": 960
|
| 1361 |
+
},
|
| 1362 |
+
{
|
| 1363 |
+
"epoch": 1.2170347003154574,
|
| 1364 |
+
"grad_norm": 0.3656632694786032,
|
| 1365 |
+
"learning_rate": 0.00011929861728017357,
|
| 1366 |
+
"loss": 2.3909,
|
| 1367 |
+
"step": 965
|
| 1368 |
+
},
|
| 1369 |
+
{
|
| 1370 |
+
"epoch": 1.2233438485804418,
|
| 1371 |
+
"grad_norm": 0.24777723089373072,
|
| 1372 |
+
"learning_rate": 0.00011768204975648757,
|
| 1373 |
+
"loss": 2.4232,
|
| 1374 |
+
"step": 970
|
| 1375 |
+
},
|
| 1376 |
+
{
|
| 1377 |
+
"epoch": 1.229652996845426,
|
| 1378 |
+
"grad_norm": 0.25021255502737505,
|
| 1379 |
+
"learning_rate": 0.00011606940912912898,
|
| 1380 |
+
"loss": 2.3574,
|
| 1381 |
+
"step": 975
|
| 1382 |
+
},
|
| 1383 |
+
{
|
| 1384 |
+
"epoch": 1.2359621451104101,
|
| 1385 |
+
"grad_norm": 0.2043220994355216,
|
| 1386 |
+
"learning_rate": 0.00011446089134717593,
|
| 1387 |
+
"loss": 2.3422,
|
| 1388 |
+
"step": 980
|
| 1389 |
+
},
|
| 1390 |
+
{
|
| 1391 |
+
"epoch": 1.2422712933753943,
|
| 1392 |
+
"grad_norm": 0.26833493933813163,
|
| 1393 |
+
"learning_rate": 0.00011285669185874695,
|
| 1394 |
+
"loss": 2.3056,
|
| 1395 |
+
"step": 985
|
| 1396 |
+
},
|
| 1397 |
+
{
|
| 1398 |
+
"epoch": 1.2485804416403785,
|
| 1399 |
+
"grad_norm": 0.2531180052762243,
|
| 1400 |
+
"learning_rate": 0.00011125700558725254,
|
| 1401 |
+
"loss": 2.4417,
|
| 1402 |
+
"step": 990
|
| 1403 |
+
},
|
| 1404 |
+
{
|
| 1405 |
+
"epoch": 1.2548895899053627,
|
| 1406 |
+
"grad_norm": 0.23358330954464956,
|
| 1407 |
+
"learning_rate": 0.00010966202690771014,
|
| 1408 |
+
"loss": 2.3947,
|
| 1409 |
+
"step": 995
|
| 1410 |
+
},
|
| 1411 |
+
{
|
| 1412 |
+
"epoch": 1.261198738170347,
|
| 1413 |
+
"grad_norm": 0.20113735373995575,
|
| 1414 |
+
"learning_rate": 0.00010807194962312622,
|
| 1415 |
+
"loss": 2.3112,
|
| 1416 |
+
"step": 1000
|
| 1417 |
+
},
|
| 1418 |
+
{
|
| 1419 |
+
"epoch": 1.2675078864353313,
|
| 1420 |
+
"grad_norm": 0.22100568862902675,
|
| 1421 |
+
"learning_rate": 0.0001064869669409475,
|
| 1422 |
+
"loss": 2.2881,
|
| 1423 |
+
"step": 1005
|
| 1424 |
+
},
|
| 1425 |
+
{
|
| 1426 |
+
"epoch": 1.2738170347003155,
|
| 1427 |
+
"grad_norm": 0.1725094492338002,
|
| 1428 |
+
"learning_rate": 0.00010490727144958466,
|
| 1429 |
+
"loss": 2.2907,
|
| 1430 |
+
"step": 1010
|
| 1431 |
+
},
|
| 1432 |
+
{
|
| 1433 |
+
"epoch": 1.2801261829652997,
|
| 1434 |
+
"grad_norm": 0.19894899283580408,
|
| 1435 |
+
"learning_rate": 0.00010333305509501118,
|
| 1436 |
+
"loss": 2.3339,
|
| 1437 |
+
"step": 1015
|
| 1438 |
+
},
|
| 1439 |
+
{
|
| 1440 |
+
"epoch": 1.2864353312302839,
|
| 1441 |
+
"grad_norm": 0.1937895354579373,
|
| 1442 |
+
"learning_rate": 0.0001017645091574407,
|
| 1443 |
+
"loss": 2.2792,
|
| 1444 |
+
"step": 1020
|
| 1445 |
+
},
|
| 1446 |
+
{
|
| 1447 |
+
"epoch": 1.2927444794952683,
|
| 1448 |
+
"grad_norm": 0.1949939225885986,
|
| 1449 |
+
"learning_rate": 0.00010020182422808427,
|
| 1450 |
+
"loss": 2.2787,
|
| 1451 |
+
"step": 1025
|
| 1452 |
+
},
|
| 1453 |
+
{
|
| 1454 |
+
"epoch": 1.2990536277602525,
|
| 1455 |
+
"grad_norm": 0.1929420127844576,
|
| 1456 |
+
"learning_rate": 9.864519018599263e-05,
|
| 1457 |
+
"loss": 2.321,
|
| 1458 |
+
"step": 1030
|
| 1459 |
+
},
|
| 1460 |
+
{
|
| 1461 |
+
"epoch": 1.3053627760252366,
|
| 1462 |
+
"grad_norm": 0.2295428945420844,
|
| 1463 |
+
"learning_rate": 9.709479617498383e-05,
|
| 1464 |
+
"loss": 2.2552,
|
| 1465 |
+
"step": 1035
|
| 1466 |
+
},
|
| 1467 |
+
{
|
| 1468 |
+
"epoch": 1.3116719242902208,
|
| 1469 |
+
"grad_norm": 0.17469777825275942,
|
| 1470 |
+
"learning_rate": 9.555083058066103e-05,
|
| 1471 |
+
"loss": 2.2713,
|
| 1472 |
+
"step": 1040
|
| 1473 |
+
},
|
| 1474 |
+
{
|
| 1475 |
+
"epoch": 1.317981072555205,
|
| 1476 |
+
"grad_norm": 0.1877965729964323,
|
| 1477 |
+
"learning_rate": 9.401348100752191e-05,
|
| 1478 |
+
"loss": 2.2856,
|
| 1479 |
+
"step": 1045
|
| 1480 |
+
},
|
| 1481 |
+
{
|
| 1482 |
+
"epoch": 1.3242902208201892,
|
| 1483 |
+
"grad_norm": 0.1871049094092869,
|
| 1484 |
+
"learning_rate": 9.248293425616325e-05,
|
| 1485 |
+
"loss": 2.3345,
|
| 1486 |
+
"step": 1050
|
| 1487 |
+
},
|
| 1488 |
+
{
|
| 1489 |
+
"epoch": 1.3305993690851734,
|
| 1490 |
+
"grad_norm": 0.17562182730820872,
|
| 1491 |
+
"learning_rate": 9.095937630058319e-05,
|
| 1492 |
+
"loss": 2.32,
|
| 1493 |
+
"step": 1055
|
| 1494 |
+
},
|
| 1495 |
+
{
|
| 1496 |
+
"epoch": 1.3369085173501578,
|
| 1497 |
+
"grad_norm": 0.19721727707111575,
|
| 1498 |
+
"learning_rate": 8.944299226558368e-05,
|
| 1499 |
+
"loss": 2.2763,
|
| 1500 |
+
"step": 1060
|
| 1501 |
+
},
|
| 1502 |
+
{
|
| 1503 |
+
"epoch": 1.343217665615142,
|
| 1504 |
+
"grad_norm": 0.22327603256233047,
|
| 1505 |
+
"learning_rate": 8.793396640427651e-05,
|
| 1506 |
+
"loss": 2.2606,
|
| 1507 |
+
"step": 1065
|
| 1508 |
+
},
|
| 1509 |
+
{
|
| 1510 |
+
"epoch": 1.3495268138801262,
|
| 1511 |
+
"grad_norm": 0.18530477438109358,
|
| 1512 |
+
"learning_rate": 8.643248207569496e-05,
|
| 1513 |
+
"loss": 2.2444,
|
| 1514 |
+
"step": 1070
|
| 1515 |
+
},
|
| 1516 |
+
{
|
| 1517 |
+
"epoch": 1.3558359621451104,
|
| 1518 |
+
"grad_norm": 0.16263455786356146,
|
| 1519 |
+
"learning_rate": 8.493872172251407e-05,
|
| 1520 |
+
"loss": 2.2728,
|
| 1521 |
+
"step": 1075
|
| 1522 |
+
},
|
| 1523 |
+
{
|
| 1524 |
+
"epoch": 1.3621451104100946,
|
| 1525 |
+
"grad_norm": 0.35294377229309676,
|
| 1526 |
+
"learning_rate": 8.345286684888243e-05,
|
| 1527 |
+
"loss": 2.2576,
|
| 1528 |
+
"step": 1080
|
| 1529 |
+
},
|
| 1530 |
+
{
|
| 1531 |
+
"epoch": 1.368454258675079,
|
| 1532 |
+
"grad_norm": 0.2760131126845038,
|
| 1533 |
+
"learning_rate": 8.197509799836805e-05,
|
| 1534 |
+
"loss": 2.2347,
|
| 1535 |
+
"step": 1085
|
| 1536 |
+
},
|
| 1537 |
+
{
|
| 1538 |
+
"epoch": 1.3747634069400632,
|
| 1539 |
+
"grad_norm": 0.23341839967965983,
|
| 1540 |
+
"learning_rate": 8.050559473202077e-05,
|
| 1541 |
+
"loss": 2.2017,
|
| 1542 |
+
"step": 1090
|
| 1543 |
+
},
|
| 1544 |
+
{
|
| 1545 |
+
"epoch": 1.3810725552050473,
|
| 1546 |
+
"grad_norm": 0.19650097566649447,
|
| 1547 |
+
"learning_rate": 7.904453560655388e-05,
|
| 1548 |
+
"loss": 2.2612,
|
| 1549 |
+
"step": 1095
|
| 1550 |
+
},
|
| 1551 |
+
{
|
| 1552 |
+
"epoch": 1.3873817034700315,
|
| 1553 |
+
"grad_norm": 0.16490504881808582,
|
| 1554 |
+
"learning_rate": 7.75920981526484e-05,
|
| 1555 |
+
"loss": 2.286,
|
| 1556 |
+
"step": 1100
|
| 1557 |
+
},
|
| 1558 |
+
{
|
| 1559 |
+
"epoch": 1.3936908517350157,
|
| 1560 |
+
"grad_norm": 0.15798695744734495,
|
| 1561 |
+
"learning_rate": 7.614845885338143e-05,
|
| 1562 |
+
"loss": 2.2467,
|
| 1563 |
+
"step": 1105
|
| 1564 |
+
},
|
| 1565 |
+
{
|
| 1566 |
+
"epoch": 1.4,
|
| 1567 |
+
"grad_norm": 0.1861219611030828,
|
| 1568 |
+
"learning_rate": 7.471379312278193e-05,
|
| 1569 |
+
"loss": 2.1576,
|
| 1570 |
+
"step": 1110
|
| 1571 |
+
},
|
| 1572 |
+
{
|
| 1573 |
+
"epoch": 1.406309148264984,
|
| 1574 |
+
"grad_norm": 0.15862620551898476,
|
| 1575 |
+
"learning_rate": 7.32882752845167e-05,
|
| 1576 |
+
"loss": 2.1794,
|
| 1577 |
+
"step": 1115
|
| 1578 |
+
},
|
| 1579 |
+
{
|
| 1580 |
+
"epoch": 1.4126182965299685,
|
| 1581 |
+
"grad_norm": 0.1788631052832408,
|
| 1582 |
+
"learning_rate": 7.187207855070852e-05,
|
| 1583 |
+
"loss": 2.1762,
|
| 1584 |
+
"step": 1120
|
| 1585 |
+
},
|
| 1586 |
+
{
|
| 1587 |
+
"epoch": 1.4189274447949527,
|
| 1588 |
+
"grad_norm": 0.16537444249451605,
|
| 1589 |
+
"learning_rate": 7.046537500088932e-05,
|
| 1590 |
+
"loss": 2.2638,
|
| 1591 |
+
"step": 1125
|
| 1592 |
+
},
|
| 1593 |
+
{
|
| 1594 |
+
"epoch": 1.4252365930599369,
|
| 1595 |
+
"grad_norm": 0.18092063003029385,
|
| 1596 |
+
"learning_rate": 6.906833556109125e-05,
|
| 1597 |
+
"loss": 2.1259,
|
| 1598 |
+
"step": 1130
|
| 1599 |
+
},
|
| 1600 |
+
{
|
| 1601 |
+
"epoch": 1.431545741324921,
|
| 1602 |
+
"grad_norm": 0.1625918159746515,
|
| 1603 |
+
"learning_rate": 6.768112998307783e-05,
|
| 1604 |
+
"loss": 2.255,
|
| 1605 |
+
"step": 1135
|
| 1606 |
+
},
|
| 1607 |
+
{
|
| 1608 |
+
"epoch": 1.4378548895899055,
|
| 1609 |
+
"grad_norm": 0.15835303177425294,
|
| 1610 |
+
"learning_rate": 6.630392682371759e-05,
|
| 1611 |
+
"loss": 2.057,
|
| 1612 |
+
"step": 1140
|
| 1613 |
+
},
|
| 1614 |
+
{
|
| 1615 |
+
"epoch": 1.4441640378548897,
|
| 1616 |
+
"grad_norm": 0.15396658395549295,
|
| 1617 |
+
"learning_rate": 6.493689342450297e-05,
|
| 1618 |
+
"loss": 2.1481,
|
| 1619 |
+
"step": 1145
|
| 1620 |
+
},
|
| 1621 |
+
{
|
| 1622 |
+
"epoch": 1.4504731861198739,
|
| 1623 |
+
"grad_norm": 0.5759225638946531,
|
| 1624 |
+
"learning_rate": 6.358019589121717e-05,
|
| 1625 |
+
"loss": 2.2069,
|
| 1626 |
+
"step": 1150
|
| 1627 |
+
},
|
| 1628 |
+
{
|
| 1629 |
+
"epoch": 1.456782334384858,
|
| 1630 |
+
"grad_norm": 0.21617372193785292,
|
| 1631 |
+
"learning_rate": 6.223399907375077e-05,
|
| 1632 |
+
"loss": 2.2331,
|
| 1633 |
+
"step": 1155
|
| 1634 |
+
},
|
| 1635 |
+
{
|
| 1636 |
+
"epoch": 1.4630914826498422,
|
| 1637 |
+
"grad_norm": 0.15857547850162454,
|
| 1638 |
+
"learning_rate": 6.089846654607106e-05,
|
| 1639 |
+
"loss": 2.2549,
|
| 1640 |
+
"step": 1160
|
| 1641 |
+
},
|
| 1642 |
+
{
|
| 1643 |
+
"epoch": 1.4694006309148264,
|
| 1644 |
+
"grad_norm": 0.15193268581792235,
|
| 1645 |
+
"learning_rate": 5.9573760586346685e-05,
|
| 1646 |
+
"loss": 2.2273,
|
| 1647 |
+
"step": 1165
|
| 1648 |
+
},
|
| 1649 |
+
{
|
| 1650 |
+
"epoch": 1.4757097791798106,
|
| 1651 |
+
"grad_norm": 0.19274566394887077,
|
| 1652 |
+
"learning_rate": 5.826004215722936e-05,
|
| 1653 |
+
"loss": 2.1593,
|
| 1654 |
+
"step": 1170
|
| 1655 |
+
},
|
| 1656 |
+
{
|
| 1657 |
+
"epoch": 1.482018927444795,
|
| 1658 |
+
"grad_norm": 0.15589014517579985,
|
| 1659 |
+
"learning_rate": 5.6957470886295556e-05,
|
| 1660 |
+
"loss": 2.18,
|
| 1661 |
+
"step": 1175
|
| 1662 |
+
},
|
| 1663 |
+
{
|
| 1664 |
+
"epoch": 1.4883280757097792,
|
| 1665 |
+
"grad_norm": 0.18064354841375416,
|
| 1666 |
+
"learning_rate": 5.566620504665043e-05,
|
| 1667 |
+
"loss": 2.2148,
|
| 1668 |
+
"step": 1180
|
| 1669 |
+
},
|
| 1670 |
+
{
|
| 1671 |
+
"epoch": 1.4946372239747634,
|
| 1672 |
+
"grad_norm": 0.18542702498191943,
|
| 1673 |
+
"learning_rate": 5.4386401537696536e-05,
|
| 1674 |
+
"loss": 2.1082,
|
| 1675 |
+
"step": 1185
|
| 1676 |
+
},
|
| 1677 |
+
{
|
| 1678 |
+
"epoch": 1.5009463722397476,
|
| 1679 |
+
"grad_norm": 0.19491318243915529,
|
| 1680 |
+
"learning_rate": 5.311821586606916e-05,
|
| 1681 |
+
"loss": 2.1317,
|
| 1682 |
+
"step": 1190
|
| 1683 |
+
},
|
| 1684 |
+
{
|
| 1685 |
+
"epoch": 1.507255520504732,
|
| 1686 |
+
"grad_norm": 0.1525482068778186,
|
| 1687 |
+
"learning_rate": 5.1861802126740765e-05,
|
| 1688 |
+
"loss": 2.1283,
|
| 1689 |
+
"step": 1195
|
| 1690 |
+
},
|
| 1691 |
+
{
|
| 1692 |
+
"epoch": 1.5135646687697162,
|
| 1693 |
+
"grad_norm": 0.1579695680187345,
|
| 1694 |
+
"learning_rate": 5.0617312984297544e-05,
|
| 1695 |
+
"loss": 2.1336,
|
| 1696 |
+
"step": 1200
|
| 1697 |
+
},
|
| 1698 |
+
{
|
| 1699 |
+
"epoch": 1.5198738170347004,
|
| 1700 |
+
"grad_norm": 0.14455231450713402,
|
| 1701 |
+
"learning_rate": 4.9384899654389214e-05,
|
| 1702 |
+
"loss": 2.121,
|
| 1703 |
+
"step": 1205
|
| 1704 |
+
},
|
| 1705 |
+
{
|
| 1706 |
+
"epoch": 1.5261829652996846,
|
| 1707 |
+
"grad_norm": 0.1495253349240125,
|
| 1708 |
+
"learning_rate": 4.816471188535503e-05,
|
| 1709 |
+
"loss": 2.0602,
|
| 1710 |
+
"step": 1210
|
| 1711 |
+
},
|
| 1712 |
+
{
|
| 1713 |
+
"epoch": 1.5324921135646687,
|
| 1714 |
+
"grad_norm": 0.4456707210896612,
|
| 1715 |
+
"learning_rate": 4.695689794002831e-05,
|
| 1716 |
+
"loss": 2.2148,
|
| 1717 |
+
"step": 1215
|
| 1718 |
+
},
|
| 1719 |
+
{
|
| 1720 |
+
"epoch": 1.538801261829653,
|
| 1721 |
+
"grad_norm": 0.12806001797338484,
|
| 1722 |
+
"learning_rate": 4.5761604577721176e-05,
|
| 1723 |
+
"loss": 2.0928,
|
| 1724 |
+
"step": 1220
|
| 1725 |
+
},
|
| 1726 |
+
{
|
| 1727 |
+
"epoch": 1.5451104100946371,
|
| 1728 |
+
"grad_norm": 0.16676732955115475,
|
| 1729 |
+
"learning_rate": 4.457897703639208e-05,
|
| 1730 |
+
"loss": 2.0774,
|
| 1731 |
+
"step": 1225
|
| 1732 |
+
},
|
| 1733 |
+
{
|
| 1734 |
+
"epoch": 1.5514195583596213,
|
| 1735 |
+
"grad_norm": 0.1412930517579618,
|
| 1736 |
+
"learning_rate": 4.340915901499813e-05,
|
| 1737 |
+
"loss": 2.1857,
|
| 1738 |
+
"step": 1230
|
| 1739 |
+
},
|
| 1740 |
+
{
|
| 1741 |
+
"epoch": 1.5577287066246057,
|
| 1742 |
+
"grad_norm": 0.16064426543465465,
|
| 1743 |
+
"learning_rate": 4.225229265603471e-05,
|
| 1744 |
+
"loss": 2.0963,
|
| 1745 |
+
"step": 1235
|
| 1746 |
+
},
|
| 1747 |
+
{
|
| 1748 |
+
"epoch": 1.56403785488959,
|
| 1749 |
+
"grad_norm": 0.1430616546259629,
|
| 1750 |
+
"learning_rate": 4.110851852826391e-05,
|
| 1751 |
+
"loss": 2.0729,
|
| 1752 |
+
"step": 1240
|
| 1753 |
+
},
|
| 1754 |
+
{
|
| 1755 |
+
"epoch": 1.570347003154574,
|
| 1756 |
+
"grad_norm": 0.14246744072657969,
|
| 1757 |
+
"learning_rate": 3.9977975609634035e-05,
|
| 1758 |
+
"loss": 2.1033,
|
| 1759 |
+
"step": 1245
|
| 1760 |
+
},
|
| 1761 |
+
{
|
| 1762 |
+
"epoch": 1.5766561514195585,
|
| 1763 |
+
"grad_norm": 0.15017155229777918,
|
| 1764 |
+
"learning_rate": 3.886080127039296e-05,
|
| 1765 |
+
"loss": 2.1265,
|
| 1766 |
+
"step": 1250
|
| 1767 |
+
},
|
| 1768 |
+
{
|
| 1769 |
+
"epoch": 1.5829652996845427,
|
| 1770 |
+
"grad_norm": 0.16087886115690112,
|
| 1771 |
+
"learning_rate": 3.775713125639647e-05,
|
| 1772 |
+
"loss": 2.0803,
|
| 1773 |
+
"step": 1255
|
| 1774 |
+
},
|
| 1775 |
+
{
|
| 1776 |
+
"epoch": 1.5892744479495269,
|
| 1777 |
+
"grad_norm": 0.1648280147102689,
|
| 1778 |
+
"learning_rate": 3.666709967261373e-05,
|
| 1779 |
+
"loss": 2.0021,
|
| 1780 |
+
"step": 1260
|
| 1781 |
+
},
|
| 1782 |
+
{
|
| 1783 |
+
"epoch": 1.595583596214511,
|
| 1784 |
+
"grad_norm": 0.13608315931815615,
|
| 1785 |
+
"learning_rate": 3.55908389668328e-05,
|
| 1786 |
+
"loss": 2.1933,
|
| 1787 |
+
"step": 1265
|
| 1788 |
+
},
|
| 1789 |
+
{
|
| 1790 |
+
"epoch": 1.6018927444794953,
|
| 1791 |
+
"grad_norm": 0.12682001843872154,
|
| 1792 |
+
"learning_rate": 3.452847991356698e-05,
|
| 1793 |
+
"loss": 1.9983,
|
| 1794 |
+
"step": 1270
|
| 1795 |
+
},
|
| 1796 |
+
{
|
| 1797 |
+
"epoch": 1.6082018927444794,
|
| 1798 |
+
"grad_norm": 0.15045660085948862,
|
| 1799 |
+
"learning_rate": 3.3480151598164575e-05,
|
| 1800 |
+
"loss": 2.0043,
|
| 1801 |
+
"step": 1275
|
| 1802 |
+
},
|
| 1803 |
+
{
|
| 1804 |
+
"epoch": 1.6145110410094636,
|
| 1805 |
+
"grad_norm": 0.5262148348424824,
|
| 1806 |
+
"learning_rate": 3.2445981401124035e-05,
|
| 1807 |
+
"loss": 2.0584,
|
| 1808 |
+
"step": 1280
|
| 1809 |
+
},
|
| 1810 |
+
{
|
| 1811 |
+
"epoch": 1.6208201892744478,
|
| 1812 |
+
"grad_norm": 0.18086944931617577,
|
| 1813 |
+
"learning_rate": 3.142609498261623e-05,
|
| 1814 |
+
"loss": 2.1297,
|
| 1815 |
+
"step": 1285
|
| 1816 |
+
},
|
| 1817 |
+
{
|
| 1818 |
+
"epoch": 1.6271293375394322,
|
| 1819 |
+
"grad_norm": 0.14876006201921135,
|
| 1820 |
+
"learning_rate": 3.0420616267215637e-05,
|
| 1821 |
+
"loss": 2.1075,
|
| 1822 |
+
"step": 1290
|
| 1823 |
+
},
|
| 1824 |
+
{
|
| 1825 |
+
"epoch": 1.6334384858044164,
|
| 1826 |
+
"grad_norm": 0.1261445893089227,
|
| 1827 |
+
"learning_rate": 2.9429667428842424e-05,
|
| 1828 |
+
"loss": 2.0413,
|
| 1829 |
+
"step": 1295
|
| 1830 |
+
},
|
| 1831 |
+
{
|
| 1832 |
+
"epoch": 1.6397476340694006,
|
| 1833 |
+
"grad_norm": 0.4002387370006984,
|
| 1834 |
+
"learning_rate": 2.845336887591746e-05,
|
| 1835 |
+
"loss": 2.0749,
|
| 1836 |
+
"step": 1300
|
| 1837 |
+
},
|
| 1838 |
+
{
|
| 1839 |
+
"epoch": 1.646056782334385,
|
| 1840 |
+
"grad_norm": 0.14748809888793443,
|
| 1841 |
+
"learning_rate": 2.7491839236731682e-05,
|
| 1842 |
+
"loss": 2.0686,
|
| 1843 |
+
"step": 1305
|
| 1844 |
+
},
|
| 1845 |
+
{
|
| 1846 |
+
"epoch": 1.6523659305993692,
|
| 1847 |
+
"grad_norm": 0.1745242948450349,
|
| 1848 |
+
"learning_rate": 2.6545195345031574e-05,
|
| 1849 |
+
"loss": 2.1086,
|
| 1850 |
+
"step": 1310
|
| 1851 |
+
},
|
| 1852 |
+
{
|
| 1853 |
+
"epoch": 1.6586750788643534,
|
| 1854 |
+
"grad_norm": 0.15200439297183374,
|
| 1855 |
+
"learning_rate": 2.5613552225823212e-05,
|
| 1856 |
+
"loss": 2.0192,
|
| 1857 |
+
"step": 1315
|
| 1858 |
+
},
|
| 1859 |
+
{
|
| 1860 |
+
"epoch": 1.6649842271293376,
|
| 1861 |
+
"grad_norm": 0.15361689677227675,
|
| 1862 |
+
"learning_rate": 2.4697023081395652e-05,
|
| 1863 |
+
"loss": 2.0913,
|
| 1864 |
+
"step": 1320
|
| 1865 |
+
},
|
| 1866 |
+
{
|
| 1867 |
+
"epoch": 1.6712933753943218,
|
| 1868 |
+
"grad_norm": 0.14672539202377308,
|
| 1869 |
+
"learning_rate": 2.37957192775658e-05,
|
| 1870 |
+
"loss": 2.066,
|
| 1871 |
+
"step": 1325
|
| 1872 |
+
},
|
| 1873 |
+
{
|
| 1874 |
+
"epoch": 1.677602523659306,
|
| 1875 |
+
"grad_norm": 0.12371192272196913,
|
| 1876 |
+
"learning_rate": 2.2909750330146665e-05,
|
| 1877 |
+
"loss": 2.0496,
|
| 1878 |
+
"step": 1330
|
| 1879 |
+
},
|
| 1880 |
+
{
|
| 1881 |
+
"epoch": 1.6839116719242901,
|
| 1882 |
+
"grad_norm": 0.12950618798301597,
|
| 1883 |
+
"learning_rate": 2.2039223891640406e-05,
|
| 1884 |
+
"loss": 2.0276,
|
| 1885 |
+
"step": 1335
|
| 1886 |
+
},
|
| 1887 |
+
{
|
| 1888 |
+
"epoch": 1.6902208201892743,
|
| 1889 |
+
"grad_norm": 0.12243785970327074,
|
| 1890 |
+
"learning_rate": 2.1184245738157395e-05,
|
| 1891 |
+
"loss": 2.0978,
|
| 1892 |
+
"step": 1340
|
| 1893 |
+
},
|
| 1894 |
+
{
|
| 1895 |
+
"epoch": 1.6965299684542585,
|
| 1896 |
+
"grad_norm": 0.15037055871528202,
|
| 1897 |
+
"learning_rate": 2.034491975656381e-05,
|
| 1898 |
+
"loss": 2.1044,
|
| 1899 |
+
"step": 1345
|
| 1900 |
+
},
|
| 1901 |
+
{
|
| 1902 |
+
"epoch": 1.702839116719243,
|
| 1903 |
+
"grad_norm": 0.20679512982442985,
|
| 1904 |
+
"learning_rate": 1.952134793185843e-05,
|
| 1905 |
+
"loss": 1.9915,
|
| 1906 |
+
"step": 1350
|
| 1907 |
+
},
|
| 1908 |
+
{
|
| 1909 |
+
"epoch": 1.7091482649842271,
|
| 1910 |
+
"grad_norm": 0.1068051803289863,
|
| 1911 |
+
"learning_rate": 1.871363033478046e-05,
|
| 1912 |
+
"loss": 2.0557,
|
| 1913 |
+
"step": 1355
|
| 1914 |
+
},
|
| 1915 |
+
{
|
| 1916 |
+
"epoch": 1.7154574132492113,
|
| 1917 |
+
"grad_norm": 0.13489201228856657,
|
| 1918 |
+
"learning_rate": 1.7921865109650447e-05,
|
| 1919 |
+
"loss": 2.0287,
|
| 1920 |
+
"step": 1360
|
| 1921 |
+
},
|
| 1922 |
+
{
|
| 1923 |
+
"epoch": 1.7217665615141957,
|
| 1924 |
+
"grad_norm": 0.11635989070750272,
|
| 1925 |
+
"learning_rate": 1.7146148462444654e-05,
|
| 1926 |
+
"loss": 2.0908,
|
| 1927 |
+
"step": 1365
|
| 1928 |
+
},
|
| 1929 |
+
{
|
| 1930 |
+
"epoch": 1.72807570977918,
|
| 1931 |
+
"grad_norm": 0.1292414052200366,
|
| 1932 |
+
"learning_rate": 1.6386574649105416e-05,
|
| 1933 |
+
"loss": 2.0663,
|
| 1934 |
+
"step": 1370
|
| 1935 |
+
},
|
| 1936 |
+
{
|
| 1937 |
+
"epoch": 1.734384858044164,
|
| 1938 |
+
"grad_norm": 0.1336092412195213,
|
| 1939 |
+
"learning_rate": 1.5643235964088064e-05,
|
| 1940 |
+
"loss": 2.0834,
|
| 1941 |
+
"step": 1375
|
| 1942 |
+
},
|
| 1943 |
+
{
|
| 1944 |
+
"epoch": 1.7406940063091483,
|
| 1945 |
+
"grad_norm": 0.12457408177697836,
|
| 1946 |
+
"learning_rate": 1.4916222729146688e-05,
|
| 1947 |
+
"loss": 2.0607,
|
| 1948 |
+
"step": 1380
|
| 1949 |
+
},
|
| 1950 |
+
{
|
| 1951 |
+
"epoch": 1.7470031545741325,
|
| 1952 |
+
"grad_norm": 0.12563174174929131,
|
| 1953 |
+
"learning_rate": 1.420562328235904e-05,
|
| 1954 |
+
"loss": 1.9503,
|
| 1955 |
+
"step": 1385
|
| 1956 |
+
},
|
| 1957 |
+
{
|
| 1958 |
+
"epoch": 1.7533123028391167,
|
| 1959 |
+
"grad_norm": 0.12518112489593036,
|
| 1960 |
+
"learning_rate": 1.351152396739284e-05,
|
| 1961 |
+
"loss": 2.0616,
|
| 1962 |
+
"step": 1390
|
| 1963 |
+
},
|
| 1964 |
+
{
|
| 1965 |
+
"epoch": 1.7596214511041008,
|
| 1966 |
+
"grad_norm": 0.10831791097675274,
|
| 1967 |
+
"learning_rate": 1.2834009123014338e-05,
|
| 1968 |
+
"loss": 2.0618,
|
| 1969 |
+
"step": 1395
|
| 1970 |
+
},
|
| 1971 |
+
{
|
| 1972 |
+
"epoch": 1.765930599369085,
|
| 1973 |
+
"grad_norm": 0.29518082944664253,
|
| 1974 |
+
"learning_rate": 1.2173161072840443e-05,
|
| 1975 |
+
"loss": 1.9915,
|
| 1976 |
+
"step": 1400
|
| 1977 |
+
},
|
| 1978 |
+
{
|
| 1979 |
+
"epoch": 1.7722397476340694,
|
| 1980 |
+
"grad_norm": 0.10583438239238828,
|
| 1981 |
+
"learning_rate": 1.1529060115335631e-05,
|
| 1982 |
+
"loss": 2.0576,
|
| 1983 |
+
"step": 1405
|
| 1984 |
+
},
|
| 1985 |
+
{
|
| 1986 |
+
"epoch": 1.7785488958990536,
|
| 1987 |
+
"grad_norm": 0.12799025071341882,
|
| 1988 |
+
"learning_rate": 1.090178451405519e-05,
|
| 1989 |
+
"loss": 2.0389,
|
| 1990 |
+
"step": 1410
|
| 1991 |
+
},
|
| 1992 |
+
{
|
| 1993 |
+
"epoch": 1.7848580441640378,
|
| 1994 |
+
"grad_norm": 0.11694082806514786,
|
| 1995 |
+
"learning_rate": 1.0291410488135438e-05,
|
| 1996 |
+
"loss": 1.9651,
|
| 1997 |
+
"step": 1415
|
| 1998 |
+
},
|
| 1999 |
+
{
|
| 2000 |
+
"epoch": 1.7911671924290222,
|
| 2001 |
+
"grad_norm": 0.11051587412406515,
|
| 2002 |
+
"learning_rate": 9.698012203032468e-06,
|
| 2003 |
+
"loss": 2.0653,
|
| 2004 |
+
"step": 1420
|
| 2005 |
+
},
|
| 2006 |
+
{
|
| 2007 |
+
"epoch": 1.7974763406940064,
|
| 2008 |
+
"grad_norm": 0.16943769742937304,
|
| 2009 |
+
"learning_rate": 9.121661761510507e-06,
|
| 2010 |
+
"loss": 1.9954,
|
| 2011 |
+
"step": 1425
|
| 2012 |
+
},
|
| 2013 |
+
{
|
| 2014 |
+
"epoch": 1.8037854889589906,
|
| 2015 |
+
"grad_norm": 0.20933132837978605,
|
| 2016 |
+
"learning_rate": 8.562429194880821e-06,
|
| 2017 |
+
"loss": 2.112,
|
| 2018 |
+
"step": 1430
|
| 2019 |
+
},
|
| 2020 |
+
{
|
| 2021 |
+
"epoch": 1.8100946372239748,
|
| 2022 |
+
"grad_norm": 0.11236448868998791,
|
| 2023 |
+
"learning_rate": 8.020382454492275e-06,
|
| 2024 |
+
"loss": 2.0302,
|
| 2025 |
+
"step": 1435
|
| 2026 |
+
},
|
| 2027 |
+
{
|
| 2028 |
+
"epoch": 1.816403785488959,
|
| 2029 |
+
"grad_norm": 0.12107899788754,
|
| 2030 |
+
"learning_rate": 7.495587403474701e-06,
|
| 2031 |
+
"loss": 2.0256,
|
| 2032 |
+
"step": 1440
|
| 2033 |
+
},
|
| 2034 |
+
{
|
| 2035 |
+
"epoch": 1.8227129337539432,
|
| 2036 |
+
"grad_norm": 0.1044284090142104,
|
| 2037 |
+
"learning_rate": 6.988107808736104e-06,
|
| 2038 |
+
"loss": 2.0336,
|
| 2039 |
+
"step": 1445
|
| 2040 |
+
},
|
| 2041 |
+
{
|
| 2042 |
+
"epoch": 1.8290220820189274,
|
| 2043 |
+
"grad_norm": 0.12046953009870044,
|
| 2044 |
+
"learning_rate": 6.49800533321434e-06,
|
| 2045 |
+
"loss": 1.982,
|
| 2046 |
+
"step": 1450
|
| 2047 |
+
},
|
| 2048 |
+
{
|
| 2049 |
+
"epoch": 1.8353312302839115,
|
| 2050 |
+
"grad_norm": 0.10963642369637508,
|
| 2051 |
+
"learning_rate": 6.025339528384493e-06,
|
| 2052 |
+
"loss": 2.0201,
|
| 2053 |
+
"step": 1455
|
| 2054 |
+
},
|
| 2055 |
+
{
|
| 2056 |
+
"epoch": 1.841640378548896,
|
| 2057 |
+
"grad_norm": 0.10565688363765909,
|
| 2058 |
+
"learning_rate": 5.570167827023115e-06,
|
| 2059 |
+
"loss": 2.0158,
|
| 2060 |
+
"step": 1460
|
| 2061 |
+
},
|
| 2062 |
+
{
|
| 2063 |
+
"epoch": 1.8479495268138801,
|
| 2064 |
+
"grad_norm": 0.11025682367925119,
|
| 2065 |
+
"learning_rate": 5.132545536229471e-06,
|
| 2066 |
+
"loss": 2.0276,
|
| 2067 |
+
"step": 1465
|
| 2068 |
+
},
|
| 2069 |
+
{
|
| 2070 |
+
"epoch": 1.8542586750788643,
|
| 2071 |
+
"grad_norm": 0.1662709864261435,
|
| 2072 |
+
"learning_rate": 4.712525830705338e-06,
|
| 2073 |
+
"loss": 2.0273,
|
| 2074 |
+
"step": 1470
|
| 2075 |
+
},
|
| 2076 |
+
{
|
| 2077 |
+
"epoch": 1.8605678233438487,
|
| 2078 |
+
"grad_norm": 0.11419381927710037,
|
| 2079 |
+
"learning_rate": 4.310159746293834e-06,
|
| 2080 |
+
"loss": 1.9885,
|
| 2081 |
+
"step": 1475
|
| 2082 |
+
},
|
| 2083 |
+
{
|
| 2084 |
+
"epoch": 1.866876971608833,
|
| 2085 |
+
"grad_norm": 0.10525019599087308,
|
| 2086 |
+
"learning_rate": 3.925496173778214e-06,
|
| 2087 |
+
"loss": 2.0498,
|
| 2088 |
+
"step": 1480
|
| 2089 |
+
},
|
| 2090 |
+
{
|
| 2091 |
+
"epoch": 1.8731861198738171,
|
| 2092 |
+
"grad_norm": 0.11114684416115388,
|
| 2093 |
+
"learning_rate": 3.558581852941106e-06,
|
| 2094 |
+
"loss": 2.027,
|
| 2095 |
+
"step": 1485
|
| 2096 |
+
},
|
| 2097 |
+
{
|
| 2098 |
+
"epoch": 1.8794952681388013,
|
| 2099 |
+
"grad_norm": 0.13571234830138024,
|
| 2100 |
+
"learning_rate": 3.209461366885335e-06,
|
| 2101 |
+
"loss": 2.0834,
|
| 2102 |
+
"step": 1490
|
| 2103 |
+
},
|
| 2104 |
+
{
|
| 2105 |
+
"epoch": 1.8858044164037855,
|
| 2106 |
+
"grad_norm": 0.24625307840667934,
|
| 2107 |
+
"learning_rate": 2.878177136616694e-06,
|
| 2108 |
+
"loss": 2.0052,
|
| 2109 |
+
"step": 1495
|
| 2110 |
+
},
|
| 2111 |
+
{
|
| 2112 |
+
"epoch": 1.8921135646687697,
|
| 2113 |
+
"grad_norm": 0.11032569559654276,
|
| 2114 |
+
"learning_rate": 2.564769415889445e-06,
|
| 2115 |
+
"loss": 2.0982,
|
| 2116 |
+
"step": 1500
|
| 2117 |
+
},
|
| 2118 |
+
{
|
| 2119 |
+
"epoch": 1.8984227129337539,
|
| 2120 |
+
"grad_norm": 0.1072743202373107,
|
| 2121 |
+
"learning_rate": 2.2692762863151528e-06,
|
| 2122 |
+
"loss": 2.0614,
|
| 2123 |
+
"step": 1505
|
| 2124 |
+
},
|
| 2125 |
+
{
|
| 2126 |
+
"epoch": 1.904731861198738,
|
| 2127 |
+
"grad_norm": 0.1161531648663479,
|
| 2128 |
+
"learning_rate": 1.9917336527354334e-06,
|
| 2129 |
+
"loss": 2.018,
|
| 2130 |
+
"step": 1510
|
| 2131 |
+
},
|
| 2132 |
+
{
|
| 2133 |
+
"epoch": 1.9110410094637222,
|
| 2134 |
+
"grad_norm": 0.09491412647258665,
|
| 2135 |
+
"learning_rate": 1.7321752388592792e-06,
|
| 2136 |
+
"loss": 2.0276,
|
| 2137 |
+
"step": 1515
|
| 2138 |
+
},
|
| 2139 |
+
{
|
| 2140 |
+
"epoch": 1.9173501577287066,
|
| 2141 |
+
"grad_norm": 0.11872568657394318,
|
| 2142 |
+
"learning_rate": 1.4906325831652687e-06,
|
| 2143 |
+
"loss": 2.0551,
|
| 2144 |
+
"step": 1520
|
| 2145 |
+
},
|
| 2146 |
+
{
|
| 2147 |
+
"epoch": 1.9236593059936908,
|
| 2148 |
+
"grad_norm": 0.11084979497518595,
|
| 2149 |
+
"learning_rate": 1.2671350350694498e-06,
|
| 2150 |
+
"loss": 1.9304,
|
| 2151 |
+
"step": 1525
|
| 2152 |
+
},
|
| 2153 |
+
{
|
| 2154 |
+
"epoch": 1.929968454258675,
|
| 2155 |
+
"grad_norm": 0.10258401422045245,
|
| 2156 |
+
"learning_rate": 1.061709751359141e-06,
|
| 2157 |
+
"loss": 2.011,
|
| 2158 |
+
"step": 1530
|
| 2159 |
+
},
|
| 2160 |
+
{
|
| 2161 |
+
"epoch": 1.9362776025236594,
|
| 2162 |
+
"grad_norm": 0.1043719731206757,
|
| 2163 |
+
"learning_rate": 8.743816928931391e-07,
|
| 2164 |
+
"loss": 2.0109,
|
| 2165 |
+
"step": 1535
|
| 2166 |
+
},
|
| 2167 |
+
{
|
| 2168 |
+
"epoch": 1.9425867507886436,
|
| 2169 |
+
"grad_norm": 0.11288186084819649,
|
| 2170 |
+
"learning_rate": 7.051736215687109e-07,
|
| 2171 |
+
"loss": 1.987,
|
| 2172 |
+
"step": 1540
|
| 2173 |
+
},
|
| 2174 |
+
{
|
| 2175 |
+
"epoch": 1.9488958990536278,
|
| 2176 |
+
"grad_norm": 0.10751919505403919,
|
| 2177 |
+
"learning_rate": 5.541060975559731e-07,
|
| 2178 |
+
"loss": 2.0008,
|
| 2179 |
+
"step": 1545
|
| 2180 |
+
},
|
| 2181 |
+
{
|
| 2182 |
+
"epoch": 1.955205047318612,
|
| 2183 |
+
"grad_norm": 0.1420810807632004,
|
| 2184 |
+
"learning_rate": 4.2119747679955626e-07,
|
| 2185 |
+
"loss": 2.0652,
|
| 2186 |
+
"step": 1550
|
| 2187 |
+
},
|
| 2188 |
+
{
|
| 2189 |
+
"epoch": 1.9615141955835962,
|
| 2190 |
+
"grad_norm": 0.09911696220850397,
|
| 2191 |
+
"learning_rate": 3.064639087882237e-07,
|
| 2192 |
+
"loss": 2.0239,
|
| 2193 |
+
"step": 1555
|
| 2194 |
+
},
|
| 2195 |
+
{
|
| 2196 |
+
"epoch": 1.9678233438485804,
|
| 2197 |
+
"grad_norm": 0.10402203428912504,
|
| 2198 |
+
"learning_rate": 2.0991933459260647e-07,
|
| 2199 |
+
"loss": 2.0042,
|
| 2200 |
+
"step": 1560
|
| 2201 |
+
},
|
| 2202 |
+
{
|
| 2203 |
+
"epoch": 1.9741324921135646,
|
| 2204 |
+
"grad_norm": 0.09525079220403034,
|
| 2205 |
+
"learning_rate": 1.315754851712425e-07,
|
| 2206 |
+
"loss": 1.9372,
|
| 2207 |
+
"step": 1565
|
| 2208 |
+
},
|
| 2209 |
+
{
|
| 2210 |
+
"epoch": 1.9804416403785488,
|
| 2211 |
+
"grad_norm": 0.13416921576674462,
|
| 2212 |
+
"learning_rate": 7.144187994513328e-08,
|
| 2213 |
+
"loss": 2.0804,
|
| 2214 |
+
"step": 1570
|
| 2215 |
+
},
|
| 2216 |
+
{
|
| 2217 |
+
"epoch": 1.9867507886435332,
|
| 2218 |
+
"grad_norm": 0.10198260208904694,
|
| 2219 |
+
"learning_rate": 2.9525825641085743e-08,
|
| 2220 |
+
"loss": 2.024,
|
| 2221 |
+
"step": 1575
|
| 2222 |
+
},
|
| 2223 |
+
{
|
| 2224 |
+
"epoch": 1.9930599369085173,
|
| 2225 |
+
"grad_norm": 0.1109028054454439,
|
| 2226 |
+
"learning_rate": 5.83241540387247e-09,
|
| 2227 |
+
"loss": 2.0887,
|
| 2228 |
+
"step": 1580
|
| 2229 |
+
},
|
| 2230 |
+
{
|
| 2231 |
+
"epoch": 1.9981072555205048,
|
| 2232 |
+
"step": 1584,
|
| 2233 |
+
"total_flos": 7.945866149068014e+18,
|
| 2234 |
+
"train_loss": 3.2101790476778542,
|
| 2235 |
+
"train_runtime": 19192.2556,
|
| 2236 |
+
"train_samples_per_second": 1.982,
|
| 2237 |
+
"train_steps_per_second": 0.083
|
| 2238 |
}
|
| 2239 |
],
|
| 2240 |
"logging_steps": 5,
|
| 2241 |
+
"max_steps": 1584,
|
| 2242 |
"num_input_tokens_seen": 0,
|
| 2243 |
"num_train_epochs": 2,
|
| 2244 |
"save_steps": 500,
|
|
|
|
| 2254 |
"attributes": {}
|
| 2255 |
}
|
| 2256 |
},
|
| 2257 |
+
"total_flos": 7.945866149068014e+18,
|
| 2258 |
"train_batch_size": 1,
|
| 2259 |
"trial_name": null,
|
| 2260 |
"trial_params": null
|
training_args.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 18673016
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:81df401c513f9fd518b2367aaa92c8356384a848eb4306ccf496860c837d05b2
|
| 3 |
size 18673016
|