Commit
•
999a3ba
1
Parent(s):
fd2cf70
huggingartists
Browse files- README.md +3 -3
- config.json +1 -1
- evaluation.txt +1 -0
- flax_model.msgpack +1 -1
- optimizer.pt +2 -2
- pytorch_model.bin +1 -1
- rng_state.pth +1 -1
- scheduler.pt +1 -1
- tokenizer_config.json +1 -1
- trainer_state.json +2585 -39
- training_args.bin +1 -1
README.md
CHANGED
|
@@ -45,15 +45,15 @@ from datasets import load_dataset
|
|
| 45 |
dataset = load_dataset("huggingartists/the-notorious-big")
|
| 46 |
```
|
| 47 |
|
| 48 |
-
[Explore the data](https://wandb.ai/huggingartists/huggingartists/runs/
|
| 49 |
|
| 50 |
## Training procedure
|
| 51 |
|
| 52 |
The model is based on a pre-trained [GPT-2](https://huggingface.co/gpt2) which is fine-tuned on The Notorious B.I.G.'s lyrics.
|
| 53 |
|
| 54 |
-
Hyperparameters and metrics are recorded in the [W&B training run](https://wandb.ai/huggingartists/huggingartists/runs/
|
| 55 |
|
| 56 |
-
At the end of training, [the final model](https://wandb.ai/huggingartists/huggingartists/runs/
|
| 57 |
|
| 58 |
## How to use
|
| 59 |
|
|
|
|
| 45 |
dataset = load_dataset("huggingartists/the-notorious-big")
|
| 46 |
```
|
| 47 |
|
| 48 |
+
[Explore the data](https://wandb.ai/huggingartists/huggingartists/runs/wkvasju4/artifacts), which is tracked with [W&B artifacts](https://docs.wandb.com/artifacts) at every step of the pipeline.
|
| 49 |
|
| 50 |
## Training procedure
|
| 51 |
|
| 52 |
The model is based on a pre-trained [GPT-2](https://huggingface.co/gpt2) which is fine-tuned on The Notorious B.I.G.'s lyrics.
|
| 53 |
|
| 54 |
+
Hyperparameters and metrics are recorded in the [W&B training run](https://wandb.ai/huggingartists/huggingartists/runs/1coezuy2) for full transparency and reproducibility.
|
| 55 |
|
| 56 |
+
At the end of training, [the final model](https://wandb.ai/huggingartists/huggingartists/runs/1coezuy2/artifacts) is logged and versioned.
|
| 57 |
|
| 58 |
## How to use
|
| 59 |
|
config.json
CHANGED
|
@@ -1,5 +1,5 @@
|
|
| 1 |
{
|
| 2 |
-
"_name_or_path": "
|
| 3 |
"activation_function": "gelu_new",
|
| 4 |
"architectures": [
|
| 5 |
"GPT2LMHeadModel"
|
|
|
|
| 1 |
{
|
| 2 |
+
"_name_or_path": "huggingartists/the-notorious-big",
|
| 3 |
"activation_function": "gelu_new",
|
| 4 |
"architectures": [
|
| 5 |
"GPT2LMHeadModel"
|
evaluation.txt
ADDED
|
@@ -0,0 +1 @@
|
|
|
|
|
|
|
| 1 |
+
{"eval_loss": 3.0452535152435303, "eval_runtime": 3.4776, "eval_samples_per_second": 76.202, "eval_steps_per_second": 9.777, "epoch": 61.0}
|
flax_model.msgpack
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 497764120
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:2465ef196026eef5b790a5a4ddf9c2899d7e0a09586d4f94854a01658a511e7c
|
| 3 |
size 497764120
|
optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
-
size
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:708bbfcc789f8a002f3b68b3c8dd1964c3a830aa3fef114e20958bcad9b97704
|
| 3 |
+
size 995604017
|
pytorch_model.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 510403817
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:9c17befbf5590275aa98389480e7799f838c9cb2763aad57e987e56f562e6cea
|
| 3 |
size 510403817
|
rng_state.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 14503
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:d906afdc1bae6abc1abecf8a957fa8e5f75aed97e2544d3a271b2818cc32823d
|
| 3 |
size 14503
|
scheduler.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 623
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:9d5c4487b77cf248b6e417ea9868d0e074182b34cf36e950b3440538936109e0
|
| 3 |
size 623
|
tokenizer_config.json
CHANGED
|
@@ -1 +1 @@
|
|
| 1 |
-
{"unk_token": "<|endoftext|>", "bos_token": "<|endoftext|>", "eos_token": "<|endoftext|>", "add_prefix_space": false, "model_max_length": 1024, "special_tokens_map_file": null, "name_or_path": "
|
|
|
|
| 1 |
+
{"unk_token": "<|endoftext|>", "bos_token": "<|endoftext|>", "eos_token": "<|endoftext|>", "add_prefix_space": false, "model_max_length": 1024, "special_tokens_map_file": null, "name_or_path": "huggingartists/the-notorious-big", "tokenizer_class": "GPT2Tokenizer"}
|
trainer_state.json
CHANGED
|
@@ -1,78 +1,2624 @@
|
|
| 1 |
{
|
| 2 |
-
"best_metric": 3.
|
| 3 |
-
"best_model_checkpoint": "output/the-notorious-big/checkpoint-
|
| 4 |
-
"epoch":
|
| 5 |
-
"global_step":
|
| 6 |
"is_hyper_param_search": false,
|
| 7 |
"is_local_process_zero": true,
|
| 8 |
"is_world_process_zero": true,
|
| 9 |
"log_history": [
|
| 10 |
{
|
| 11 |
-
"epoch": 0.
|
| 12 |
-
"learning_rate":
|
| 13 |
-
"loss": 3.
|
| 14 |
"step": 5
|
| 15 |
},
|
| 16 |
{
|
| 17 |
-
"epoch": 0.
|
| 18 |
-
"learning_rate":
|
| 19 |
-
"loss": 3.
|
| 20 |
"step": 10
|
| 21 |
},
|
| 22 |
{
|
| 23 |
-
"epoch": 0.
|
| 24 |
-
"learning_rate":
|
| 25 |
-
"loss":
|
| 26 |
"step": 15
|
| 27 |
},
|
| 28 |
{
|
| 29 |
-
"epoch": 0.
|
| 30 |
-
"learning_rate":
|
| 31 |
-
"loss":
|
| 32 |
"step": 20
|
| 33 |
},
|
| 34 |
{
|
| 35 |
-
"epoch": 0.
|
| 36 |
-
"learning_rate":
|
| 37 |
-
"loss":
|
| 38 |
"step": 25
|
| 39 |
},
|
| 40 |
{
|
| 41 |
-
"epoch": 0.
|
| 42 |
-
"learning_rate":
|
| 43 |
-
"loss":
|
| 44 |
"step": 30
|
| 45 |
},
|
| 46 |
{
|
| 47 |
-
"epoch": 0.
|
| 48 |
-
"learning_rate":
|
| 49 |
-
"loss":
|
| 50 |
"step": 35
|
| 51 |
},
|
| 52 |
{
|
| 53 |
-
"epoch": 0.
|
| 54 |
-
"learning_rate":
|
| 55 |
-
"loss":
|
| 56 |
"step": 40
|
| 57 |
},
|
| 58 |
{
|
| 59 |
-
"epoch": 0.
|
| 60 |
-
"learning_rate": 1.
|
| 61 |
-
"loss":
|
| 62 |
"step": 45
|
| 63 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 64 |
{
|
| 65 |
"epoch": 1.0,
|
| 66 |
-
"eval_loss": 3.
|
| 67 |
-
"eval_runtime":
|
| 68 |
-
"eval_samples_per_second": 76.
|
| 69 |
-
"eval_steps_per_second": 9.
|
| 70 |
-
"step":
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 71 |
}
|
| 72 |
],
|
| 73 |
-
"max_steps":
|
| 74 |
-
"num_train_epochs":
|
| 75 |
-
"total_flos":
|
| 76 |
"trial_name": null,
|
| 77 |
"trial_params": null
|
| 78 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"best_metric": 3.0452535152435303,
|
| 3 |
+
"best_model_checkpoint": "output/the-notorious-big/checkpoint-2101",
|
| 4 |
+
"epoch": 11.0,
|
| 5 |
+
"global_step": 2101,
|
| 6 |
"is_hyper_param_search": false,
|
| 7 |
"is_local_process_zero": true,
|
| 8 |
"is_world_process_zero": true,
|
| 9 |
"log_history": [
|
| 10 |
{
|
| 11 |
+
"epoch": 0.03,
|
| 12 |
+
"learning_rate": 2.318581390145182e-07,
|
| 13 |
+
"loss": 3.6069,
|
| 14 |
"step": 5
|
| 15 |
},
|
| 16 |
{
|
| 17 |
+
"epoch": 0.05,
|
| 18 |
+
"learning_rate": 9.258652616957599e-07,
|
| 19 |
+
"loss": 3.6646,
|
| 20 |
"step": 10
|
| 21 |
},
|
| 22 |
{
|
| 23 |
+
"epoch": 0.08,
|
| 24 |
+
"learning_rate": 2.0773300794160956e-06,
|
| 25 |
+
"loss": 3.5544,
|
| 26 |
"step": 15
|
| 27 |
},
|
| 28 |
{
|
| 29 |
+
"epoch": 0.1,
|
| 30 |
+
"learning_rate": 3.6784690210450094e-06,
|
| 31 |
+
"loss": 3.8578,
|
| 32 |
"step": 20
|
| 33 |
},
|
| 34 |
{
|
| 35 |
+
"epoch": 0.13,
|
| 36 |
+
"learning_rate": 5.718458847652147e-06,
|
| 37 |
+
"loss": 3.5273,
|
| 38 |
"step": 25
|
| 39 |
},
|
| 40 |
{
|
| 41 |
+
"epoch": 0.16,
|
| 42 |
+
"learning_rate": 8.183509814491012e-06,
|
| 43 |
+
"loss": 3.6649,
|
| 44 |
"step": 30
|
| 45 |
},
|
| 46 |
{
|
| 47 |
+
"epoch": 0.18,
|
| 48 |
+
"learning_rate": 1.1056958885708937e-05,
|
| 49 |
+
"loss": 3.4757,
|
| 50 |
"step": 35
|
| 51 |
},
|
| 52 |
{
|
| 53 |
+
"epoch": 0.21,
|
| 54 |
+
"learning_rate": 1.4319382371678895e-05,
|
| 55 |
+
"loss": 3.4231,
|
| 56 |
"step": 40
|
| 57 |
},
|
| 58 |
{
|
| 59 |
+
"epoch": 0.24,
|
| 60 |
+
"learning_rate": 1.7948727227557336e-05,
|
| 61 |
+
"loss": 3.4855,
|
| 62 |
"step": 45
|
| 63 |
},
|
| 64 |
+
{
|
| 65 |
+
"epoch": 0.26,
|
| 66 |
+
"learning_rate": 2.1920460125527185e-05,
|
| 67 |
+
"loss": 3.7272,
|
| 68 |
+
"step": 50
|
| 69 |
+
},
|
| 70 |
+
{
|
| 71 |
+
"epoch": 0.29,
|
| 72 |
+
"learning_rate": 2.6207733293040745e-05,
|
| 73 |
+
"loss": 3.6966,
|
| 74 |
+
"step": 55
|
| 75 |
+
},
|
| 76 |
+
{
|
| 77 |
+
"epoch": 0.31,
|
| 78 |
+
"learning_rate": 3.078156599604364e-05,
|
| 79 |
+
"loss": 3.5331,
|
| 80 |
+
"step": 60
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.34,
|
| 84 |
+
"learning_rate": 3.561104044040511e-05,
|
| 85 |
+
"loss": 3.5794,
|
| 86 |
+
"step": 65
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.37,
|
| 90 |
+
"learning_rate": 4.0663510767318035e-05,
|
| 91 |
+
"loss": 3.5518,
|
| 92 |
+
"step": 70
|
| 93 |
+
},
|
| 94 |
+
{
|
| 95 |
+
"epoch": 0.39,
|
| 96 |
+
"learning_rate": 4.590482372991862e-05,
|
| 97 |
+
"loss": 3.5769,
|
| 98 |
+
"step": 75
|
| 99 |
+
},
|
| 100 |
+
{
|
| 101 |
+
"epoch": 0.42,
|
| 102 |
+
"learning_rate": 5.12995495594168e-05,
|
| 103 |
+
"loss": 3.7063,
|
| 104 |
+
"step": 80
|
| 105 |
+
},
|
| 106 |
+
{
|
| 107 |
+
"epoch": 0.45,
|
| 108 |
+
"learning_rate": 5.681122146014723e-05,
|
| 109 |
+
"loss": 3.5721,
|
| 110 |
+
"step": 85
|
| 111 |
+
},
|
| 112 |
+
{
|
| 113 |
+
"epoch": 0.47,
|
| 114 |
+
"learning_rate": 6.240258211462417e-05,
|
| 115 |
+
"loss": 3.45,
|
| 116 |
+
"step": 90
|
| 117 |
+
},
|
| 118 |
+
{
|
| 119 |
+
"epoch": 0.5,
|
| 120 |
+
"learning_rate": 6.803583553229638e-05,
|
| 121 |
+
"loss": 3.5444,
|
| 122 |
+
"step": 95
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 0.52,
|
| 126 |
+
"learning_rate": 7.36729025395774e-05,
|
| 127 |
+
"loss": 3.5872,
|
| 128 |
+
"step": 100
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.55,
|
| 132 |
+
"learning_rate": 7.927567818411118e-05,
|
| 133 |
+
"loss": 3.5725,
|
| 134 |
+
"step": 105
|
| 135 |
+
},
|
| 136 |
+
{
|
| 137 |
+
"epoch": 0.58,
|
| 138 |
+
"learning_rate": 8.480628931329552e-05,
|
| 139 |
+
"loss": 3.5443,
|
| 140 |
+
"step": 110
|
| 141 |
+
},
|
| 142 |
+
{
|
| 143 |
+
"epoch": 0.6,
|
| 144 |
+
"learning_rate": 9.022735058590523e-05,
|
| 145 |
+
"loss": 3.4685,
|
| 146 |
+
"step": 115
|
| 147 |
+
},
|
| 148 |
+
{
|
| 149 |
+
"epoch": 0.63,
|
| 150 |
+
"learning_rate": 9.550221718625067e-05,
|
| 151 |
+
"loss": 3.5604,
|
| 152 |
+
"step": 120
|
| 153 |
+
},
|
| 154 |
+
{
|
| 155 |
+
"epoch": 0.65,
|
| 156 |
+
"learning_rate": 0.00010059523253259703,
|
| 157 |
+
"loss": 3.5,
|
| 158 |
+
"step": 125
|
| 159 |
+
},
|
| 160 |
+
{
|
| 161 |
+
"epoch": 0.68,
|
| 162 |
+
"learning_rate": 0.00010547196930540564,
|
| 163 |
+
"loss": 3.4513,
|
| 164 |
+
"step": 130
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 0.71,
|
| 168 |
+
"learning_rate": 0.00011009946216611652,
|
| 169 |
+
"loss": 3.6782,
|
| 170 |
+
"step": 135
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.73,
|
| 174 |
+
"learning_rate": 0.00011444643059335986,
|
| 175 |
+
"loss": 3.7351,
|
| 176 |
+
"step": 140
|
| 177 |
+
},
|
| 178 |
+
{
|
| 179 |
+
"epoch": 0.76,
|
| 180 |
+
"learning_rate": 0.00011848349033028726,
|
| 181 |
+
"loss": 3.4672,
|
| 182 |
+
"step": 145
|
| 183 |
+
},
|
| 184 |
+
{
|
| 185 |
+
"epoch": 0.79,
|
| 186 |
+
"learning_rate": 0.0001221833520137007,
|
| 187 |
+
"loss": 3.4939,
|
| 188 |
+
"step": 150
|
| 189 |
+
},
|
| 190 |
+
{
|
| 191 |
+
"epoch": 0.81,
|
| 192 |
+
"learning_rate": 0.00012552100564230202,
|
| 193 |
+
"loss": 3.3836,
|
| 194 |
+
"step": 155
|
| 195 |
+
},
|
| 196 |
+
{
|
| 197 |
+
"epoch": 0.84,
|
| 198 |
+
"learning_rate": 0.00012847388963711152,
|
| 199 |
+
"loss": 3.5585,
|
| 200 |
+
"step": 160
|
| 201 |
+
},
|
| 202 |
+
{
|
| 203 |
+
"epoch": 0.86,
|
| 204 |
+
"learning_rate": 0.00013102204335125564,
|
| 205 |
+
"loss": 3.6885,
|
| 206 |
+
"step": 165
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 0.89,
|
| 210 |
+
"learning_rate": 0.00013314824199820138,
|
| 211 |
+
"loss": 3.4539,
|
| 212 |
+
"step": 170
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.92,
|
| 216 |
+
"learning_rate": 0.00013483811308636105,
|
| 217 |
+
"loss": 3.44,
|
| 218 |
+
"step": 175
|
| 219 |
+
},
|
| 220 |
+
{
|
| 221 |
+
"epoch": 0.94,
|
| 222 |
+
"learning_rate": 0.00013608023357300362,
|
| 223 |
+
"loss": 3.2787,
|
| 224 |
+
"step": 180
|
| 225 |
+
},
|
| 226 |
+
{
|
| 227 |
+
"epoch": 0.97,
|
| 228 |
+
"learning_rate": 0.00013686620708073944,
|
| 229 |
+
"loss": 3.4844,
|
| 230 |
+
"step": 185
|
| 231 |
+
},
|
| 232 |
+
{
|
| 233 |
+
"epoch": 0.99,
|
| 234 |
+
"learning_rate": 0.0001371907206546175,
|
| 235 |
+
"loss": 3.4819,
|
| 236 |
+
"step": 190
|
| 237 |
+
},
|
| 238 |
{
|
| 239 |
"epoch": 1.0,
|
| 240 |
+
"eval_loss": 3.4188125133514404,
|
| 241 |
+
"eval_runtime": 3.4458,
|
| 242 |
+
"eval_samples_per_second": 76.904,
|
| 243 |
+
"eval_steps_per_second": 9.867,
|
| 244 |
+
"step": 191
|
| 245 |
+
},
|
| 246 |
+
{
|
| 247 |
+
"epoch": 1.02,
|
| 248 |
+
"learning_rate": 0.0001370515806761736,
|
| 249 |
+
"loss": 3.3397,
|
| 250 |
+
"step": 195
|
| 251 |
+
},
|
| 252 |
+
{
|
| 253 |
+
"epoch": 1.05,
|
| 254 |
+
"learning_rate": 0.00013644972769166031,
|
| 255 |
+
"loss": 3.3572,
|
| 256 |
+
"step": 200
|
| 257 |
+
},
|
| 258 |
+
{
|
| 259 |
+
"epoch": 1.07,
|
| 260 |
+
"learning_rate": 0.00013538923005422486,
|
| 261 |
+
"loss": 3.262,
|
| 262 |
+
"step": 205
|
| 263 |
+
},
|
| 264 |
+
{
|
| 265 |
+
"epoch": 1.1,
|
| 266 |
+
"learning_rate": 0.00013387725642301142,
|
| 267 |
+
"loss": 3.1682,
|
| 268 |
+
"step": 210
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 1.13,
|
| 272 |
+
"learning_rate": 0.00013192402730508626,
|
| 273 |
+
"loss": 3.4747,
|
| 274 |
+
"step": 215
|
| 275 |
+
},
|
| 276 |
+
{
|
| 277 |
+
"epoch": 1.15,
|
| 278 |
+
"learning_rate": 0.00012954274596774817,
|
| 279 |
+
"loss": 3.4567,
|
| 280 |
+
"step": 220
|
| 281 |
+
},
|
| 282 |
+
{
|
| 283 |
+
"epoch": 1.18,
|
| 284 |
+
"learning_rate": 0.0001267495091882376,
|
| 285 |
+
"loss": 3.2241,
|
| 286 |
+
"step": 225
|
| 287 |
+
},
|
| 288 |
+
{
|
| 289 |
+
"epoch": 1.2,
|
| 290 |
+
"learning_rate": 0.00012356319844415004,
|
| 291 |
+
"loss": 3.104,
|
| 292 |
+
"step": 230
|
| 293 |
+
},
|
| 294 |
+
{
|
| 295 |
+
"epoch": 1.23,
|
| 296 |
+
"learning_rate": 0.00012000535228007566,
|
| 297 |
+
"loss": 3.0082,
|
| 298 |
+
"step": 235
|
| 299 |
+
},
|
| 300 |
+
{
|
| 301 |
+
"epoch": 1.26,
|
| 302 |
+
"learning_rate": 0.0001161000207132303,
|
| 303 |
+
"loss": 3.2858,
|
| 304 |
+
"step": 240
|
| 305 |
+
},
|
| 306 |
+
{
|
| 307 |
+
"epoch": 1.28,
|
| 308 |
+
"learning_rate": 0.00011187360266225318,
|
| 309 |
+
"loss": 3.2368,
|
| 310 |
+
"step": 245
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 1.31,
|
| 314 |
+
"learning_rate": 0.00010735466749810626,
|
| 315 |
+
"loss": 3.2569,
|
| 316 |
+
"step": 250
|
| 317 |
+
},
|
| 318 |
+
{
|
| 319 |
+
"epoch": 1.34,
|
| 320 |
+
"learning_rate": 0.0001025737619233402,
|
| 321 |
+
"loss": 3.1777,
|
| 322 |
+
"step": 255
|
| 323 |
+
},
|
| 324 |
+
{
|
| 325 |
+
"epoch": 1.36,
|
| 326 |
+
"learning_rate": 9.756320348516688e-05,
|
| 327 |
+
"loss": 3.3335,
|
| 328 |
+
"step": 260
|
| 329 |
+
},
|
| 330 |
+
{
|
| 331 |
+
"epoch": 1.39,
|
| 332 |
+
"learning_rate": 9.23568621181315e-05,
|
| 333 |
+
"loss": 3.0882,
|
| 334 |
+
"step": 265
|
| 335 |
+
},
|
| 336 |
+
{
|
| 337 |
+
"epoch": 1.41,
|
| 338 |
+
"learning_rate": 8.69899311930922e-05,
|
| 339 |
+
"loss": 3.2659,
|
| 340 |
+
"step": 270
|
| 341 |
+
},
|
| 342 |
+
{
|
| 343 |
+
"epoch": 1.44,
|
| 344 |
+
"learning_rate": 8.149868962015093e-05,
|
| 345 |
+
"loss": 3.2573,
|
| 346 |
+
"step": 275
|
| 347 |
+
},
|
| 348 |
+
{
|
| 349 |
+
"epoch": 1.47,
|
| 350 |
+
"learning_rate": 7.592025661365123e-05,
|
| 351 |
+
"loss": 3.1257,
|
| 352 |
+
"step": 280
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 1.49,
|
| 356 |
+
"learning_rate": 7.029234077696063e-05,
|
| 357 |
+
"loss": 3.4355,
|
| 358 |
+
"step": 285
|
| 359 |
+
},
|
| 360 |
+
{
|
| 361 |
+
"epoch": 1.52,
|
| 362 |
+
"learning_rate": 6.465298520315264e-05,
|
| 363 |
+
"loss": 3.089,
|
| 364 |
+
"step": 290
|
| 365 |
+
},
|
| 366 |
+
{
|
| 367 |
+
"epoch": 1.54,
|
| 368 |
+
"learning_rate": 5.9040310314635076e-05,
|
| 369 |
+
"loss": 3.0155,
|
| 370 |
+
"step": 295
|
| 371 |
+
},
|
| 372 |
+
{
|
| 373 |
+
"epoch": 1.57,
|
| 374 |
+
"learning_rate": 5.349225618005358e-05,
|
| 375 |
+
"loss": 3.098,
|
| 376 |
+
"step": 300
|
| 377 |
+
},
|
| 378 |
+
{
|
| 379 |
+
"epoch": 1.6,
|
| 380 |
+
"learning_rate": 4.804632605033502e-05,
|
| 381 |
+
"loss": 3.0511,
|
| 382 |
+
"step": 305
|
| 383 |
+
},
|
| 384 |
+
{
|
| 385 |
+
"epoch": 1.62,
|
| 386 |
+
"learning_rate": 4.273933284749334e-05,
|
| 387 |
+
"loss": 3.254,
|
| 388 |
+
"step": 310
|
| 389 |
+
},
|
| 390 |
+
{
|
| 391 |
+
"epoch": 1.65,
|
| 392 |
+
"learning_rate": 3.760715031986035e-05,
|
| 393 |
+
"loss": 3.4357,
|
| 394 |
+
"step": 315
|
| 395 |
+
},
|
| 396 |
+
{
|
| 397 |
+
"epoch": 1.68,
|
| 398 |
+
"learning_rate": 3.268447054585981e-05,
|
| 399 |
+
"loss": 3.1706,
|
| 400 |
+
"step": 320
|
| 401 |
+
},
|
| 402 |
+
{
|
| 403 |
+
"epoch": 1.7,
|
| 404 |
+
"learning_rate": 2.8004569425531536e-05,
|
| 405 |
+
"loss": 3.2949,
|
| 406 |
+
"step": 325
|
| 407 |
+
},
|
| 408 |
+
{
|
| 409 |
+
"epoch": 1.73,
|
| 410 |
+
"learning_rate": 2.359908174501344e-05,
|
| 411 |
+
"loss": 3.0581,
|
| 412 |
+
"step": 330
|
| 413 |
+
},
|
| 414 |
+
{
|
| 415 |
+
"epoch": 1.75,
|
| 416 |
+
"learning_rate": 1.949778733448209e-05,
|
| 417 |
+
"loss": 2.9815,
|
| 418 |
+
"step": 335
|
| 419 |
+
},
|
| 420 |
+
{
|
| 421 |
+
"epoch": 1.78,
|
| 422 |
+
"learning_rate": 1.572840976506218e-05,
|
| 423 |
+
"loss": 3.1285,
|
| 424 |
+
"step": 340
|
| 425 |
+
},
|
| 426 |
+
{
|
| 427 |
+
"epoch": 1.81,
|
| 428 |
+
"learning_rate": 1.2316428945455985e-05,
|
| 429 |
+
"loss": 3.2872,
|
| 430 |
+
"step": 345
|
| 431 |
+
},
|
| 432 |
+
{
|
| 433 |
+
"epoch": 1.83,
|
| 434 |
+
"learning_rate": 9.284908885085886e-06,
|
| 435 |
+
"loss": 3.3075,
|
| 436 |
+
"step": 350
|
| 437 |
+
},
|
| 438 |
+
{
|
| 439 |
+
"epoch": 1.86,
|
| 440 |
+
"learning_rate": 6.654341788021345e-06,
|
| 441 |
+
"loss": 3.1916,
|
| 442 |
+
"step": 355
|
| 443 |
+
},
|
| 444 |
+
{
|
| 445 |
+
"epoch": 1.88,
|
| 446 |
+
"learning_rate": 4.442509531570712e-06,
|
| 447 |
+
"loss": 3.2904,
|
| 448 |
+
"step": 360
|
| 449 |
+
},
|
| 450 |
+
{
|
| 451 |
+
"epoch": 1.91,
|
| 452 |
+
"learning_rate": 2.6643634659029754e-06,
|
| 453 |
+
"loss": 3.1763,
|
| 454 |
+
"step": 365
|
| 455 |
+
},
|
| 456 |
+
{
|
| 457 |
+
"epoch": 1.94,
|
| 458 |
+
"learning_rate": 1.3319233472187158e-06,
|
| 459 |
+
"loss": 3.2474,
|
| 460 |
+
"step": 370
|
| 461 |
+
},
|
| 462 |
+
{
|
| 463 |
+
"epoch": 1.96,
|
| 464 |
+
"learning_rate": 4.541960876539865e-07,
|
| 465 |
+
"loss": 3.295,
|
| 466 |
+
"step": 375
|
| 467 |
+
},
|
| 468 |
+
{
|
| 469 |
+
"epoch": 1.99,
|
| 470 |
+
"learning_rate": 3.711487114362266e-08,
|
| 471 |
+
"loss": 2.9591,
|
| 472 |
+
"step": 380
|
| 473 |
+
},
|
| 474 |
+
{
|
| 475 |
+
"epoch": 2.0,
|
| 476 |
+
"eval_loss": 3.276902914047241,
|
| 477 |
+
"eval_runtime": 3.4584,
|
| 478 |
+
"eval_samples_per_second": 76.624,
|
| 479 |
+
"eval_steps_per_second": 9.831,
|
| 480 |
+
"step": 382
|
| 481 |
+
},
|
| 482 |
+
{
|
| 483 |
+
"epoch": 2.02,
|
| 484 |
+
"learning_rate": 8.349904680360514e-08,
|
| 485 |
+
"loss": 3.0232,
|
| 486 |
+
"step": 385
|
| 487 |
+
},
|
| 488 |
+
{
|
| 489 |
+
"epoch": 2.04,
|
| 490 |
+
"learning_rate": 5.930350709415567e-07,
|
| 491 |
+
"loss": 3.027,
|
| 492 |
+
"step": 390
|
| 493 |
+
},
|
| 494 |
+
{
|
| 495 |
+
"epoch": 2.07,
|
| 496 |
+
"learning_rate": 1.5622786265220133e-06,
|
| 497 |
+
"loss": 3.0191,
|
| 498 |
+
"step": 395
|
| 499 |
+
},
|
| 500 |
+
{
|
| 501 |
+
"epoch": 2.09,
|
| 502 |
+
"learning_rate": 2.9846779057592496e-06,
|
| 503 |
+
"loss": 2.7702,
|
| 504 |
+
"step": 400
|
| 505 |
+
},
|
| 506 |
+
{
|
| 507 |
+
"epoch": 2.12,
|
| 508 |
+
"learning_rate": 4.850617898454477e-06,
|
| 509 |
+
"loss": 2.9301,
|
| 510 |
+
"step": 405
|
| 511 |
+
},
|
| 512 |
+
{
|
| 513 |
+
"epoch": 2.15,
|
| 514 |
+
"learning_rate": 7.147485386700417e-06,
|
| 515 |
+
"loss": 2.9283,
|
| 516 |
+
"step": 410
|
| 517 |
+
},
|
| 518 |
+
{
|
| 519 |
+
"epoch": 2.17,
|
| 520 |
+
"learning_rate": 9.859754206608232e-06,
|
| 521 |
+
"loss": 3.039,
|
| 522 |
+
"step": 415
|
| 523 |
+
},
|
| 524 |
+
{
|
| 525 |
+
"epoch": 2.2,
|
| 526 |
+
"learning_rate": 1.2969090200711697e-05,
|
| 527 |
+
"loss": 3.1938,
|
| 528 |
+
"step": 420
|
| 529 |
+
},
|
| 530 |
+
{
|
| 531 |
+
"epoch": 2.23,
|
| 532 |
+
"learning_rate": 1.645447515160081e-05,
|
| 533 |
+
"loss": 2.8778,
|
| 534 |
+
"step": 425
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 2.25,
|
| 538 |
+
"learning_rate": 2.0292348859028838e-05,
|
| 539 |
+
"loss": 3.1301,
|
| 540 |
+
"step": 430
|
| 541 |
+
},
|
| 542 |
+
{
|
| 543 |
+
"epoch": 2.28,
|
| 544 |
+
"learning_rate": 2.4456768400092677e-05,
|
| 545 |
+
"loss": 3.2,
|
| 546 |
+
"step": 435
|
| 547 |
+
},
|
| 548 |
+
{
|
| 549 |
+
"epoch": 2.3,
|
| 550 |
+
"learning_rate": 2.891958349593333e-05,
|
| 551 |
+
"loss": 3.2341,
|
| 552 |
+
"step": 440
|
| 553 |
+
},
|
| 554 |
+
{
|
| 555 |
+
"epoch": 2.33,
|
| 556 |
+
"learning_rate": 3.3650626799527644e-05,
|
| 557 |
+
"loss": 3.0918,
|
| 558 |
+
"step": 445
|
| 559 |
+
},
|
| 560 |
+
{
|
| 561 |
+
"epoch": 2.36,
|
| 562 |
+
"learning_rate": 3.861791781828188e-05,
|
| 563 |
+
"loss": 3.0801,
|
| 564 |
+
"step": 450
|
| 565 |
+
},
|
| 566 |
+
{
|
| 567 |
+
"epoch": 2.38,
|
| 568 |
+
"learning_rate": 4.378787909296712e-05,
|
| 569 |
+
"loss": 3.1323,
|
| 570 |
+
"step": 455
|
| 571 |
+
},
|
| 572 |
+
{
|
| 573 |
+
"epoch": 2.41,
|
| 574 |
+
"learning_rate": 4.91255631716905e-05,
|
| 575 |
+
"loss": 2.9001,
|
| 576 |
+
"step": 460
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 2.43,
|
| 580 |
+
"learning_rate": 5.4594888844622835e-05,
|
| 581 |
+
"loss": 2.9404,
|
| 582 |
+
"step": 465
|
| 583 |
+
},
|
| 584 |
+
{
|
| 585 |
+
"epoch": 2.46,
|
| 586 |
+
"learning_rate": 6.0158885042604066e-05,
|
| 587 |
+
"loss": 2.8547,
|
| 588 |
+
"step": 470
|
| 589 |
+
},
|
| 590 |
+
{
|
| 591 |
+
"epoch": 2.49,
|
| 592 |
+
"learning_rate": 6.577994075094185e-05,
|
| 593 |
+
"loss": 3.023,
|
| 594 |
+
"step": 475
|
| 595 |
+
},
|
| 596 |
+
{
|
| 597 |
+
"epoch": 2.51,
|
| 598 |
+
"learning_rate": 7.14200592490581e-05,
|
| 599 |
+
"loss": 2.8448,
|
| 600 |
+
"step": 480
|
| 601 |
+
},
|
| 602 |
+
{
|
| 603 |
+
"epoch": 2.54,
|
| 604 |
+
"learning_rate": 7.704111495739588e-05,
|
| 605 |
+
"loss": 2.933,
|
| 606 |
+
"step": 485
|
| 607 |
+
},
|
| 608 |
+
{
|
| 609 |
+
"epoch": 2.57,
|
| 610 |
+
"learning_rate": 8.260511115537712e-05,
|
| 611 |
+
"loss": 2.9129,
|
| 612 |
+
"step": 490
|
| 613 |
+
},
|
| 614 |
+
{
|
| 615 |
+
"epoch": 2.59,
|
| 616 |
+
"learning_rate": 8.807443682830946e-05,
|
| 617 |
+
"loss": 2.9149,
|
| 618 |
+
"step": 495
|
| 619 |
+
},
|
| 620 |
+
{
|
| 621 |
+
"epoch": 2.62,
|
| 622 |
+
"learning_rate": 9.341212090703282e-05,
|
| 623 |
+
"loss": 3.051,
|
| 624 |
+
"step": 500
|
| 625 |
+
},
|
| 626 |
+
{
|
| 627 |
+
"epoch": 2.64,
|
| 628 |
+
"learning_rate": 9.858208218171807e-05,
|
| 629 |
+
"loss": 2.9097,
|
| 630 |
+
"step": 505
|
| 631 |
+
},
|
| 632 |
+
{
|
| 633 |
+
"epoch": 2.67,
|
| 634 |
+
"learning_rate": 0.00010354937320047231,
|
| 635 |
+
"loss": 3.0169,
|
| 636 |
+
"step": 510
|
| 637 |
+
},
|
| 638 |
+
{
|
| 639 |
+
"epoch": 2.7,
|
| 640 |
+
"learning_rate": 0.00010828041650406663,
|
| 641 |
+
"loss": 2.9528,
|
| 642 |
+
"step": 515
|
| 643 |
+
},
|
| 644 |
+
{
|
| 645 |
+
"epoch": 2.72,
|
| 646 |
+
"learning_rate": 0.00011274323159990728,
|
| 647 |
+
"loss": 3.1967,
|
| 648 |
+
"step": 520
|
| 649 |
+
},
|
| 650 |
+
{
|
| 651 |
+
"epoch": 2.75,
|
| 652 |
+
"learning_rate": 0.00011690765114097114,
|
| 653 |
+
"loss": 3.1819,
|
| 654 |
+
"step": 525
|
| 655 |
+
},
|
| 656 |
+
{
|
| 657 |
+
"epoch": 2.77,
|
| 658 |
+
"learning_rate": 0.00012074552484839915,
|
| 659 |
+
"loss": 2.9504,
|
| 660 |
+
"step": 530
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"epoch": 2.8,
|
| 664 |
+
"learning_rate": 0.00012423090979928828,
|
| 665 |
+
"loss": 2.9404,
|
| 666 |
+
"step": 535
|
| 667 |
+
},
|
| 668 |
+
{
|
| 669 |
+
"epoch": 2.83,
|
| 670 |
+
"learning_rate": 0.00012734024579339175,
|
| 671 |
+
"loss": 3.0818,
|
| 672 |
+
"step": 540
|
| 673 |
+
},
|
| 674 |
+
{
|
| 675 |
+
"epoch": 2.85,
|
| 676 |
+
"learning_rate": 0.00013005251461329956,
|
| 677 |
+
"loss": 3.14,
|
| 678 |
+
"step": 545
|
| 679 |
+
},
|
| 680 |
+
{
|
| 681 |
+
"epoch": 2.88,
|
| 682 |
+
"learning_rate": 0.0001323493821015455,
|
| 683 |
+
"loss": 2.9756,
|
| 684 |
+
"step": 550
|
| 685 |
+
},
|
| 686 |
+
{
|
| 687 |
+
"epoch": 2.91,
|
| 688 |
+
"learning_rate": 0.00013421532209424074,
|
| 689 |
+
"loss": 3.1237,
|
| 690 |
+
"step": 555
|
| 691 |
+
},
|
| 692 |
+
{
|
| 693 |
+
"epoch": 2.93,
|
| 694 |
+
"learning_rate": 0.000135637721373478,
|
| 695 |
+
"loss": 2.7926,
|
| 696 |
+
"step": 560
|
| 697 |
+
},
|
| 698 |
+
{
|
| 699 |
+
"epoch": 2.96,
|
| 700 |
+
"learning_rate": 0.00013660696492905842,
|
| 701 |
+
"loss": 3.0619,
|
| 702 |
+
"step": 565
|
| 703 |
+
},
|
| 704 |
+
{
|
| 705 |
+
"epoch": 2.98,
|
| 706 |
+
"learning_rate": 0.00013711650095319638,
|
| 707 |
+
"loss": 3.0946,
|
| 708 |
+
"step": 570
|
| 709 |
+
},
|
| 710 |
+
{
|
| 711 |
+
"epoch": 3.0,
|
| 712 |
+
"eval_loss": 3.266069173812866,
|
| 713 |
+
"eval_runtime": 3.4464,
|
| 714 |
+
"eval_samples_per_second": 76.891,
|
| 715 |
+
"eval_steps_per_second": 9.865,
|
| 716 |
+
"step": 573
|
| 717 |
+
},
|
| 718 |
+
{
|
| 719 |
+
"epoch": 3.01,
|
| 720 |
+
"learning_rate": 0.00013716288512885638,
|
| 721 |
+
"loss": 3.1661,
|
| 722 |
+
"step": 575
|
| 723 |
+
},
|
| 724 |
+
{
|
| 725 |
+
"epoch": 3.04,
|
| 726 |
+
"learning_rate": 0.000136745803912346,
|
| 727 |
+
"loss": 2.6779,
|
| 728 |
+
"step": 580
|
| 729 |
+
},
|
| 730 |
+
{
|
| 731 |
+
"epoch": 3.06,
|
| 732 |
+
"learning_rate": 0.00013586807665278127,
|
| 733 |
+
"loss": 2.6821,
|
| 734 |
+
"step": 585
|
| 735 |
+
},
|
| 736 |
+
{
|
| 737 |
+
"epoch": 3.09,
|
| 738 |
+
"learning_rate": 0.00013453563653409702,
|
| 739 |
+
"loss": 2.7079,
|
| 740 |
+
"step": 590
|
| 741 |
+
},
|
| 742 |
+
{
|
| 743 |
+
"epoch": 3.12,
|
| 744 |
+
"learning_rate": 0.00013275749046842927,
|
| 745 |
+
"loss": 2.7825,
|
| 746 |
+
"step": 595
|
| 747 |
+
},
|
| 748 |
+
{
|
| 749 |
+
"epoch": 3.14,
|
| 750 |
+
"learning_rate": 0.0001305456582119787,
|
| 751 |
+
"loss": 2.8057,
|
| 752 |
+
"step": 600
|
| 753 |
+
},
|
| 754 |
+
{
|
| 755 |
+
"epoch": 3.17,
|
| 756 |
+
"learning_rate": 0.00012791509111491415,
|
| 757 |
+
"loss": 2.9291,
|
| 758 |
+
"step": 605
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"epoch": 3.19,
|
| 762 |
+
"learning_rate": 0.00012488357105454404,
|
| 763 |
+
"loss": 2.9353,
|
| 764 |
+
"step": 610
|
| 765 |
+
},
|
| 766 |
+
{
|
| 767 |
+
"epoch": 3.22,
|
| 768 |
+
"learning_rate": 0.00012147159023493787,
|
| 769 |
+
"loss": 2.8316,
|
| 770 |
+
"step": 615
|
| 771 |
+
},
|
| 772 |
+
{
|
| 773 |
+
"epoch": 3.25,
|
| 774 |
+
"learning_rate": 0.00011770221266551796,
|
| 775 |
+
"loss": 2.9357,
|
| 776 |
+
"step": 620
|
| 777 |
+
},
|
| 778 |
+
{
|
| 779 |
+
"epoch": 3.27,
|
| 780 |
+
"learning_rate": 0.00011360091825498661,
|
| 781 |
+
"loss": 2.9269,
|
| 782 |
+
"step": 625
|
| 783 |
+
},
|
| 784 |
+
{
|
| 785 |
+
"epoch": 3.3,
|
| 786 |
+
"learning_rate": 0.00010919543057446852,
|
| 787 |
+
"loss": 2.8717,
|
| 788 |
+
"step": 630
|
| 789 |
+
},
|
| 790 |
+
{
|
| 791 |
+
"epoch": 3.32,
|
| 792 |
+
"learning_rate": 0.00010451552945414024,
|
| 793 |
+
"loss": 2.8045,
|
| 794 |
+
"step": 635
|
| 795 |
+
},
|
| 796 |
+
{
|
| 797 |
+
"epoch": 3.35,
|
| 798 |
+
"learning_rate": 9.959284968013971e-05,
|
| 799 |
+
"loss": 2.9829,
|
| 800 |
+
"step": 640
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 3.38,
|
| 804 |
+
"learning_rate": 9.446066715250671e-05,
|
| 805 |
+
"loss": 2.7457,
|
| 806 |
+
"step": 645
|
| 807 |
+
},
|
| 808 |
+
{
|
| 809 |
+
"epoch": 3.4,
|
| 810 |
+
"learning_rate": 8.915367394966503e-05,
|
| 811 |
+
"loss": 2.7768,
|
| 812 |
+
"step": 650
|
| 813 |
+
},
|
| 814 |
+
{
|
| 815 |
+
"epoch": 3.43,
|
| 816 |
+
"learning_rate": 8.370774381994648e-05,
|
| 817 |
+
"loss": 2.5404,
|
| 818 |
+
"step": 655
|
| 819 |
+
},
|
| 820 |
+
{
|
| 821 |
+
"epoch": 3.46,
|
| 822 |
+
"learning_rate": 7.815968968536493e-05,
|
| 823 |
+
"loss": 2.6558,
|
| 824 |
+
"step": 660
|
| 825 |
+
},
|
| 826 |
+
{
|
| 827 |
+
"epoch": 3.48,
|
| 828 |
+
"learning_rate": 7.254701479684736e-05,
|
| 829 |
+
"loss": 2.8658,
|
| 830 |
+
"step": 665
|
| 831 |
+
},
|
| 832 |
+
{
|
| 833 |
+
"epoch": 3.51,
|
| 834 |
+
"learning_rate": 6.690765922303943e-05,
|
| 835 |
+
"loss": 3.0811,
|
| 836 |
+
"step": 670
|
| 837 |
+
},
|
| 838 |
+
{
|
| 839 |
+
"epoch": 3.53,
|
| 840 |
+
"learning_rate": 6.127974338634884e-05,
|
| 841 |
+
"loss": 2.7066,
|
| 842 |
+
"step": 675
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"epoch": 3.56,
|
| 846 |
+
"learning_rate": 5.570131037984919e-05,
|
| 847 |
+
"loss": 2.8037,
|
| 848 |
+
"step": 680
|
| 849 |
+
},
|
| 850 |
+
{
|
| 851 |
+
"epoch": 3.59,
|
| 852 |
+
"learning_rate": 5.0210068806907975e-05,
|
| 853 |
+
"loss": 2.5833,
|
| 854 |
+
"step": 685
|
| 855 |
+
},
|
| 856 |
+
{
|
| 857 |
+
"epoch": 3.61,
|
| 858 |
+
"learning_rate": 4.484313788186866e-05,
|
| 859 |
+
"loss": 2.8664,
|
| 860 |
+
"step": 690
|
| 861 |
+
},
|
| 862 |
+
{
|
| 863 |
+
"epoch": 3.64,
|
| 864 |
+
"learning_rate": 3.963679651483327e-05,
|
| 865 |
+
"loss": 2.7661,
|
| 866 |
+
"step": 695
|
| 867 |
+
},
|
| 868 |
+
{
|
| 869 |
+
"epoch": 3.66,
|
| 870 |
+
"learning_rate": 3.462623807665996e-05,
|
| 871 |
+
"loss": 2.6497,
|
| 872 |
+
"step": 700
|
| 873 |
+
},
|
| 874 |
+
{
|
| 875 |
+
"epoch": 3.69,
|
| 876 |
+
"learning_rate": 2.9845332501893884e-05,
|
| 877 |
+
"loss": 2.6264,
|
| 878 |
+
"step": 705
|
| 879 |
+
},
|
| 880 |
+
{
|
| 881 |
+
"epoch": 3.72,
|
| 882 |
+
"learning_rate": 2.5326397337746965e-05,
|
| 883 |
+
"loss": 2.7237,
|
| 884 |
+
"step": 710
|
| 885 |
+
},
|
| 886 |
+
{
|
| 887 |
+
"epoch": 3.74,
|
| 888 |
+
"learning_rate": 2.109997928676983e-05,
|
| 889 |
+
"loss": 2.8732,
|
| 890 |
+
"step": 715
|
| 891 |
+
},
|
| 892 |
+
{
|
| 893 |
+
"epoch": 3.77,
|
| 894 |
+
"learning_rate": 1.719464771992433e-05,
|
| 895 |
+
"loss": 2.9159,
|
| 896 |
+
"step": 720
|
| 897 |
+
},
|
| 898 |
+
{
|
| 899 |
+
"epoch": 3.8,
|
| 900 |
+
"learning_rate": 1.3636801555849966e-05,
|
| 901 |
+
"loss": 2.9253,
|
| 902 |
+
"step": 725
|
| 903 |
+
},
|
| 904 |
+
{
|
| 905 |
+
"epoch": 3.82,
|
| 906 |
+
"learning_rate": 1.0450490811762374e-05,
|
| 907 |
+
"loss": 2.6746,
|
| 908 |
+
"step": 730
|
| 909 |
+
},
|
| 910 |
+
{
|
| 911 |
+
"epoch": 3.85,
|
| 912 |
+
"learning_rate": 7.657254032251816e-06,
|
| 913 |
+
"loss": 2.7104,
|
| 914 |
+
"step": 735
|
| 915 |
+
},
|
| 916 |
+
{
|
| 917 |
+
"epoch": 3.87,
|
| 918 |
+
"learning_rate": 5.275972694913755e-06,
|
| 919 |
+
"loss": 2.8526,
|
| 920 |
+
"step": 740
|
| 921 |
+
},
|
| 922 |
+
{
|
| 923 |
+
"epoch": 3.9,
|
| 924 |
+
"learning_rate": 3.3227435769885726e-06,
|
| 925 |
+
"loss": 2.8344,
|
| 926 |
+
"step": 745
|
| 927 |
+
},
|
| 928 |
+
{
|
| 929 |
+
"epoch": 3.93,
|
| 930 |
+
"learning_rate": 1.8107699457751103e-06,
|
| 931 |
+
"loss": 2.7425,
|
| 932 |
+
"step": 750
|
| 933 |
+
},
|
| 934 |
+
{
|
| 935 |
+
"epoch": 3.95,
|
| 936 |
+
"learning_rate": 7.502723083396683e-07,
|
| 937 |
+
"loss": 2.8643,
|
| 938 |
+
"step": 755
|
| 939 |
+
},
|
| 940 |
+
{
|
| 941 |
+
"epoch": 3.98,
|
| 942 |
+
"learning_rate": 1.4841932382638991e-07,
|
| 943 |
+
"loss": 2.8953,
|
| 944 |
+
"step": 760
|
| 945 |
+
},
|
| 946 |
+
{
|
| 947 |
+
"epoch": 4.0,
|
| 948 |
+
"eval_loss": 3.1601226329803467,
|
| 949 |
+
"eval_runtime": 3.449,
|
| 950 |
+
"eval_samples_per_second": 76.834,
|
| 951 |
+
"eval_steps_per_second": 9.858,
|
| 952 |
+
"step": 764
|
| 953 |
+
},
|
| 954 |
+
{
|
| 955 |
+
"epoch": 4.01,
|
| 956 |
+
"learning_rate": 9.279345382488647e-09,
|
| 957 |
+
"loss": 2.6508,
|
| 958 |
+
"step": 765
|
| 959 |
+
},
|
| 960 |
+
{
|
| 961 |
+
"epoch": 4.03,
|
| 962 |
+
"learning_rate": 3.3379291926057076e-07,
|
| 963 |
+
"loss": 2.4879,
|
| 964 |
+
"step": 770
|
| 965 |
+
},
|
| 966 |
+
{
|
| 967 |
+
"epoch": 4.06,
|
| 968 |
+
"learning_rate": 1.119766426996385e-06,
|
| 969 |
+
"loss": 2.5634,
|
| 970 |
+
"step": 775
|
| 971 |
+
},
|
| 972 |
+
{
|
| 973 |
+
"epoch": 4.08,
|
| 974 |
+
"learning_rate": 2.361886913638944e-06,
|
| 975 |
+
"loss": 2.6256,
|
| 976 |
+
"step": 780
|
| 977 |
+
},
|
| 978 |
+
{
|
| 979 |
+
"epoch": 4.11,
|
| 980 |
+
"learning_rate": 4.051758001798611e-06,
|
| 981 |
+
"loss": 2.7326,
|
| 982 |
+
"step": 785
|
| 983 |
+
},
|
| 984 |
+
{
|
| 985 |
+
"epoch": 4.14,
|
| 986 |
+
"learning_rate": 6.1779566487443804e-06,
|
| 987 |
+
"loss": 2.6579,
|
| 988 |
+
"step": 790
|
| 989 |
+
},
|
| 990 |
+
{
|
| 991 |
+
"epoch": 4.16,
|
| 992 |
+
"learning_rate": 8.726110362888466e-06,
|
| 993 |
+
"loss": 2.5027,
|
| 994 |
+
"step": 795
|
| 995 |
+
},
|
| 996 |
+
{
|
| 997 |
+
"epoch": 4.19,
|
| 998 |
+
"learning_rate": 1.167899435769802e-05,
|
| 999 |
+
"loss": 2.9039,
|
| 1000 |
+
"step": 800
|
| 1001 |
+
},
|
| 1002 |
+
{
|
| 1003 |
+
"epoch": 4.21,
|
| 1004 |
+
"learning_rate": 1.5016647986299231e-05,
|
| 1005 |
+
"loss": 2.4959,
|
| 1006 |
+
"step": 805
|
| 1007 |
+
},
|
| 1008 |
+
{
|
| 1009 |
+
"epoch": 4.24,
|
| 1010 |
+
"learning_rate": 1.8716509669712747e-05,
|
| 1011 |
+
"loss": 2.5862,
|
| 1012 |
+
"step": 810
|
| 1013 |
+
},
|
| 1014 |
+
{
|
| 1015 |
+
"epoch": 4.27,
|
| 1016 |
+
"learning_rate": 2.2753569406640054e-05,
|
| 1017 |
+
"loss": 2.5506,
|
| 1018 |
+
"step": 815
|
| 1019 |
+
},
|
| 1020 |
+
{
|
| 1021 |
+
"epoch": 4.29,
|
| 1022 |
+
"learning_rate": 2.7100537833883422e-05,
|
| 1023 |
+
"loss": 2.3108,
|
| 1024 |
+
"step": 820
|
| 1025 |
+
},
|
| 1026 |
+
{
|
| 1027 |
+
"epoch": 4.32,
|
| 1028 |
+
"learning_rate": 3.172803069459421e-05,
|
| 1029 |
+
"loss": 2.536,
|
| 1030 |
+
"step": 825
|
| 1031 |
+
},
|
| 1032 |
+
{
|
| 1033 |
+
"epoch": 4.35,
|
| 1034 |
+
"learning_rate": 3.660476746740293e-05,
|
| 1035 |
+
"loss": 2.5062,
|
| 1036 |
+
"step": 830
|
| 1037 |
+
},
|
| 1038 |
+
{
|
| 1039 |
+
"epoch": 4.37,
|
| 1040 |
+
"learning_rate": 4.169778281374917e-05,
|
| 1041 |
+
"loss": 2.4466,
|
| 1042 |
+
"step": 835
|
| 1043 |
+
},
|
| 1044 |
+
{
|
| 1045 |
+
"epoch": 4.4,
|
| 1046 |
+
"learning_rate": 4.697264941409472e-05,
|
| 1047 |
+
"loss": 2.4376,
|
| 1048 |
+
"step": 840
|
| 1049 |
+
},
|
| 1050 |
+
{
|
| 1051 |
+
"epoch": 4.42,
|
| 1052 |
+
"learning_rate": 5.2393710686704254e-05,
|
| 1053 |
+
"loss": 2.8146,
|
| 1054 |
+
"step": 845
|
| 1055 |
+
},
|
| 1056 |
+
{
|
| 1057 |
+
"epoch": 4.45,
|
| 1058 |
+
"learning_rate": 5.79243218158887e-05,
|
| 1059 |
+
"loss": 2.6586,
|
| 1060 |
+
"step": 850
|
| 1061 |
+
},
|
| 1062 |
+
{
|
| 1063 |
+
"epoch": 4.48,
|
| 1064 |
+
"learning_rate": 6.35270974604226e-05,
|
| 1065 |
+
"loss": 2.5297,
|
| 1066 |
+
"step": 855
|
| 1067 |
+
},
|
| 1068 |
+
{
|
| 1069 |
+
"epoch": 4.5,
|
| 1070 |
+
"learning_rate": 6.916416446770344e-05,
|
| 1071 |
+
"loss": 2.33,
|
| 1072 |
+
"step": 860
|
| 1073 |
+
},
|
| 1074 |
+
{
|
| 1075 |
+
"epoch": 4.53,
|
| 1076 |
+
"learning_rate": 7.47974178853758e-05,
|
| 1077 |
+
"loss": 2.5292,
|
| 1078 |
+
"step": 865
|
| 1079 |
+
},
|
| 1080 |
+
{
|
| 1081 |
+
"epoch": 4.55,
|
| 1082 |
+
"learning_rate": 8.038877853985258e-05,
|
| 1083 |
+
"loss": 2.5357,
|
| 1084 |
+
"step": 870
|
| 1085 |
+
},
|
| 1086 |
+
{
|
| 1087 |
+
"epoch": 4.58,
|
| 1088 |
+
"learning_rate": 8.590045044058314e-05,
|
| 1089 |
+
"loss": 2.6811,
|
| 1090 |
+
"step": 875
|
| 1091 |
+
},
|
| 1092 |
+
{
|
| 1093 |
+
"epoch": 4.61,
|
| 1094 |
+
"learning_rate": 9.129517627008144e-05,
|
| 1095 |
+
"loss": 2.5603,
|
| 1096 |
+
"step": 880
|
| 1097 |
+
},
|
| 1098 |
+
{
|
| 1099 |
+
"epoch": 4.63,
|
| 1100 |
+
"learning_rate": 9.653648923268214e-05,
|
| 1101 |
+
"loss": 2.716,
|
| 1102 |
+
"step": 885
|
| 1103 |
+
},
|
| 1104 |
+
{
|
| 1105 |
+
"epoch": 4.66,
|
| 1106 |
+
"learning_rate": 0.00010158895955959489,
|
| 1107 |
+
"loss": 2.7213,
|
| 1108 |
+
"step": 890
|
| 1109 |
+
},
|
| 1110 |
+
{
|
| 1111 |
+
"epoch": 4.69,
|
| 1112 |
+
"learning_rate": 0.00010641843400395645,
|
| 1113 |
+
"loss": 2.4893,
|
| 1114 |
+
"step": 895
|
| 1115 |
+
},
|
| 1116 |
+
{
|
| 1117 |
+
"epoch": 4.71,
|
| 1118 |
+
"learning_rate": 0.0001109922667069592,
|
| 1119 |
+
"loss": 2.6099,
|
| 1120 |
+
"step": 900
|
| 1121 |
+
},
|
| 1122 |
+
{
|
| 1123 |
+
"epoch": 4.74,
|
| 1124 |
+
"learning_rate": 0.00011527953987447286,
|
| 1125 |
+
"loss": 2.7473,
|
| 1126 |
+
"step": 905
|
| 1127 |
+
},
|
| 1128 |
+
{
|
| 1129 |
+
"epoch": 4.76,
|
| 1130 |
+
"learning_rate": 0.00011925127277244263,
|
| 1131 |
+
"loss": 2.7703,
|
| 1132 |
+
"step": 910
|
| 1133 |
+
},
|
| 1134 |
+
{
|
| 1135 |
+
"epoch": 4.79,
|
| 1136 |
+
"learning_rate": 0.00012288061762832114,
|
| 1137 |
+
"loss": 2.75,
|
| 1138 |
+
"step": 915
|
| 1139 |
+
},
|
| 1140 |
+
{
|
| 1141 |
+
"epoch": 4.82,
|
| 1142 |
+
"learning_rate": 0.00012614304111429104,
|
| 1143 |
+
"loss": 2.7698,
|
| 1144 |
+
"step": 920
|
| 1145 |
+
},
|
| 1146 |
+
{
|
| 1147 |
+
"epoch": 4.84,
|
| 1148 |
+
"learning_rate": 0.000129016490185509,
|
| 1149 |
+
"loss": 2.7411,
|
| 1150 |
+
"step": 925
|
| 1151 |
+
},
|
| 1152 |
+
{
|
| 1153 |
+
"epoch": 4.87,
|
| 1154 |
+
"learning_rate": 0.0001314815411523478,
|
| 1155 |
+
"loss": 2.6514,
|
| 1156 |
+
"step": 930
|
| 1157 |
+
},
|
| 1158 |
+
{
|
| 1159 |
+
"epoch": 4.9,
|
| 1160 |
+
"learning_rate": 0.000133521530978955,
|
| 1161 |
+
"loss": 2.7788,
|
| 1162 |
+
"step": 935
|
| 1163 |
+
},
|
| 1164 |
+
{
|
| 1165 |
+
"epoch": 4.92,
|
| 1166 |
+
"learning_rate": 0.00013512266992058388,
|
| 1167 |
+
"loss": 2.5737,
|
| 1168 |
+
"step": 940
|
| 1169 |
+
},
|
| 1170 |
+
{
|
| 1171 |
+
"epoch": 4.95,
|
| 1172 |
+
"learning_rate": 0.00013627413473830423,
|
| 1173 |
+
"loss": 2.3355,
|
| 1174 |
+
"step": 945
|
| 1175 |
+
},
|
| 1176 |
+
{
|
| 1177 |
+
"epoch": 4.97,
|
| 1178 |
+
"learning_rate": 0.00013696814186098548,
|
| 1179 |
+
"loss": 2.7536,
|
| 1180 |
+
"step": 950
|
| 1181 |
+
},
|
| 1182 |
+
{
|
| 1183 |
+
"epoch": 5.0,
|
| 1184 |
+
"learning_rate": 0.0001372,
|
| 1185 |
+
"loss": 2.8008,
|
| 1186 |
+
"step": 955
|
| 1187 |
+
},
|
| 1188 |
+
{
|
| 1189 |
+
"epoch": 5.0,
|
| 1190 |
+
"eval_loss": 3.157210350036621,
|
| 1191 |
+
"eval_runtime": 3.4457,
|
| 1192 |
+
"eval_samples_per_second": 76.906,
|
| 1193 |
+
"eval_steps_per_second": 9.867,
|
| 1194 |
+
"step": 955
|
| 1195 |
+
},
|
| 1196 |
+
{
|
| 1197 |
+
"epoch": 5.03,
|
| 1198 |
+
"learning_rate": 0.00013696814186098548,
|
| 1199 |
+
"loss": 2.3276,
|
| 1200 |
+
"step": 960
|
| 1201 |
+
},
|
| 1202 |
+
{
|
| 1203 |
+
"epoch": 5.05,
|
| 1204 |
+
"learning_rate": 0.00013627413473830426,
|
| 1205 |
+
"loss": 2.5804,
|
| 1206 |
+
"step": 965
|
| 1207 |
+
},
|
| 1208 |
+
{
|
| 1209 |
+
"epoch": 5.08,
|
| 1210 |
+
"learning_rate": 0.0001351226699205839,
|
| 1211 |
+
"loss": 2.554,
|
| 1212 |
+
"step": 970
|
| 1213 |
+
},
|
| 1214 |
+
{
|
| 1215 |
+
"epoch": 5.1,
|
| 1216 |
+
"learning_rate": 0.00013352153097895503,
|
| 1217 |
+
"loss": 2.5025,
|
| 1218 |
+
"step": 975
|
| 1219 |
+
},
|
| 1220 |
+
{
|
| 1221 |
+
"epoch": 5.13,
|
| 1222 |
+
"learning_rate": 0.00013148154115234786,
|
| 1223 |
+
"loss": 2.4069,
|
| 1224 |
+
"step": 980
|
| 1225 |
+
},
|
| 1226 |
+
{
|
| 1227 |
+
"epoch": 5.16,
|
| 1228 |
+
"learning_rate": 0.00012901649018550904,
|
| 1229 |
+
"loss": 2.6704,
|
| 1230 |
+
"step": 985
|
| 1231 |
+
},
|
| 1232 |
+
{
|
| 1233 |
+
"epoch": 5.18,
|
| 1234 |
+
"learning_rate": 0.0001261430411142911,
|
| 1235 |
+
"loss": 2.2362,
|
| 1236 |
+
"step": 990
|
| 1237 |
+
},
|
| 1238 |
+
{
|
| 1239 |
+
"epoch": 5.21,
|
| 1240 |
+
"learning_rate": 0.00012288061762832122,
|
| 1241 |
+
"loss": 2.3459,
|
| 1242 |
+
"step": 995
|
| 1243 |
+
},
|
| 1244 |
+
{
|
| 1245 |
+
"epoch": 5.24,
|
| 1246 |
+
"learning_rate": 0.0001192512727724427,
|
| 1247 |
+
"loss": 2.2415,
|
| 1248 |
+
"step": 1000
|
| 1249 |
+
},
|
| 1250 |
+
{
|
| 1251 |
+
"epoch": 5.26,
|
| 1252 |
+
"learning_rate": 0.00011527953987447295,
|
| 1253 |
+
"loss": 2.5082,
|
| 1254 |
+
"step": 1005
|
| 1255 |
+
},
|
| 1256 |
+
{
|
| 1257 |
+
"epoch": 5.29,
|
| 1258 |
+
"learning_rate": 0.00011099226670695929,
|
| 1259 |
+
"loss": 2.3534,
|
| 1260 |
+
"step": 1010
|
| 1261 |
+
},
|
| 1262 |
+
{
|
| 1263 |
+
"epoch": 5.31,
|
| 1264 |
+
"learning_rate": 0.00010641843400395655,
|
| 1265 |
+
"loss": 2.5017,
|
| 1266 |
+
"step": 1015
|
| 1267 |
+
},
|
| 1268 |
+
{
|
| 1269 |
+
"epoch": 5.34,
|
| 1270 |
+
"learning_rate": 0.00010158895955959498,
|
| 1271 |
+
"loss": 2.4172,
|
| 1272 |
+
"step": 1020
|
| 1273 |
+
},
|
| 1274 |
+
{
|
| 1275 |
+
"epoch": 5.37,
|
| 1276 |
+
"learning_rate": 9.653648923268222e-05,
|
| 1277 |
+
"loss": 2.5386,
|
| 1278 |
+
"step": 1025
|
| 1279 |
+
},
|
| 1280 |
+
{
|
| 1281 |
+
"epoch": 5.39,
|
| 1282 |
+
"learning_rate": 9.129517627008154e-05,
|
| 1283 |
+
"loss": 2.4429,
|
| 1284 |
+
"step": 1030
|
| 1285 |
+
},
|
| 1286 |
+
{
|
| 1287 |
+
"epoch": 5.42,
|
| 1288 |
+
"learning_rate": 8.590045044058324e-05,
|
| 1289 |
+
"loss": 2.3742,
|
| 1290 |
+
"step": 1035
|
| 1291 |
+
},
|
| 1292 |
+
{
|
| 1293 |
+
"epoch": 5.45,
|
| 1294 |
+
"learning_rate": 8.038877853985269e-05,
|
| 1295 |
+
"loss": 2.4793,
|
| 1296 |
+
"step": 1040
|
| 1297 |
+
},
|
| 1298 |
+
{
|
| 1299 |
+
"epoch": 5.47,
|
| 1300 |
+
"learning_rate": 7.479741788537588e-05,
|
| 1301 |
+
"loss": 2.6054,
|
| 1302 |
+
"step": 1045
|
| 1303 |
+
},
|
| 1304 |
+
{
|
| 1305 |
+
"epoch": 5.5,
|
| 1306 |
+
"learning_rate": 6.916416446770354e-05,
|
| 1307 |
+
"loss": 2.4472,
|
| 1308 |
+
"step": 1050
|
| 1309 |
+
},
|
| 1310 |
+
{
|
| 1311 |
+
"epoch": 5.52,
|
| 1312 |
+
"learning_rate": 6.35270974604227e-05,
|
| 1313 |
+
"loss": 2.5238,
|
| 1314 |
+
"step": 1055
|
| 1315 |
+
},
|
| 1316 |
+
{
|
| 1317 |
+
"epoch": 5.55,
|
| 1318 |
+
"learning_rate": 5.792432181588881e-05,
|
| 1319 |
+
"loss": 2.2441,
|
| 1320 |
+
"step": 1060
|
| 1321 |
+
},
|
| 1322 |
+
{
|
| 1323 |
+
"epoch": 5.58,
|
| 1324 |
+
"learning_rate": 5.239371068670435e-05,
|
| 1325 |
+
"loss": 2.5048,
|
| 1326 |
+
"step": 1065
|
| 1327 |
+
},
|
| 1328 |
+
{
|
| 1329 |
+
"epoch": 5.6,
|
| 1330 |
+
"learning_rate": 4.6972649414094817e-05,
|
| 1331 |
+
"loss": 2.4761,
|
| 1332 |
+
"step": 1070
|
| 1333 |
+
},
|
| 1334 |
+
{
|
| 1335 |
+
"epoch": 5.63,
|
| 1336 |
+
"learning_rate": 4.1697782813749265e-05,
|
| 1337 |
+
"loss": 2.5068,
|
| 1338 |
+
"step": 1075
|
| 1339 |
+
},
|
| 1340 |
+
{
|
| 1341 |
+
"epoch": 5.65,
|
| 1342 |
+
"learning_rate": 3.6604767467403014e-05,
|
| 1343 |
+
"loss": 2.4981,
|
| 1344 |
+
"step": 1080
|
| 1345 |
+
},
|
| 1346 |
+
{
|
| 1347 |
+
"epoch": 5.68,
|
| 1348 |
+
"learning_rate": 3.17280306945943e-05,
|
| 1349 |
+
"loss": 2.3925,
|
| 1350 |
+
"step": 1085
|
| 1351 |
+
},
|
| 1352 |
+
{
|
| 1353 |
+
"epoch": 5.71,
|
| 1354 |
+
"learning_rate": 2.7100537833883503e-05,
|
| 1355 |
+
"loss": 2.271,
|
| 1356 |
+
"step": 1090
|
| 1357 |
+
},
|
| 1358 |
+
{
|
| 1359 |
+
"epoch": 5.73,
|
| 1360 |
+
"learning_rate": 2.275356940664013e-05,
|
| 1361 |
+
"loss": 2.5484,
|
| 1362 |
+
"step": 1095
|
| 1363 |
+
},
|
| 1364 |
+
{
|
| 1365 |
+
"epoch": 5.76,
|
| 1366 |
+
"learning_rate": 1.8716509669712814e-05,
|
| 1367 |
+
"loss": 2.3203,
|
| 1368 |
+
"step": 1100
|
| 1369 |
+
},
|
| 1370 |
+
{
|
| 1371 |
+
"epoch": 5.79,
|
| 1372 |
+
"learning_rate": 1.50166479862993e-05,
|
| 1373 |
+
"loss": 2.4399,
|
| 1374 |
+
"step": 1105
|
| 1375 |
+
},
|
| 1376 |
+
{
|
| 1377 |
+
"epoch": 5.81,
|
| 1378 |
+
"learning_rate": 1.1678994357698073e-05,
|
| 1379 |
+
"loss": 2.5271,
|
| 1380 |
+
"step": 1110
|
| 1381 |
+
},
|
| 1382 |
+
{
|
| 1383 |
+
"epoch": 5.84,
|
| 1384 |
+
"learning_rate": 8.726110362888518e-06,
|
| 1385 |
+
"loss": 2.535,
|
| 1386 |
+
"step": 1115
|
| 1387 |
+
},
|
| 1388 |
+
{
|
| 1389 |
+
"epoch": 5.86,
|
| 1390 |
+
"learning_rate": 6.177956648744426e-06,
|
| 1391 |
+
"loss": 2.3464,
|
| 1392 |
+
"step": 1120
|
| 1393 |
+
},
|
| 1394 |
+
{
|
| 1395 |
+
"epoch": 5.89,
|
| 1396 |
+
"learning_rate": 4.0517580017986414e-06,
|
| 1397 |
+
"loss": 2.2768,
|
| 1398 |
+
"step": 1125
|
| 1399 |
+
},
|
| 1400 |
+
{
|
| 1401 |
+
"epoch": 5.92,
|
| 1402 |
+
"learning_rate": 2.3618869136389365e-06,
|
| 1403 |
+
"loss": 2.5173,
|
| 1404 |
+
"step": 1130
|
| 1405 |
+
},
|
| 1406 |
+
{
|
| 1407 |
+
"epoch": 5.94,
|
| 1408 |
+
"learning_rate": 1.1197664269964003e-06,
|
| 1409 |
+
"loss": 2.4667,
|
| 1410 |
+
"step": 1135
|
| 1411 |
+
},
|
| 1412 |
+
{
|
| 1413 |
+
"epoch": 5.97,
|
| 1414 |
+
"learning_rate": 3.3379291926057076e-07,
|
| 1415 |
+
"loss": 2.4227,
|
| 1416 |
+
"step": 1140
|
| 1417 |
+
},
|
| 1418 |
+
{
|
| 1419 |
+
"epoch": 5.99,
|
| 1420 |
+
"learning_rate": 9.279345382488647e-09,
|
| 1421 |
+
"loss": 2.5067,
|
| 1422 |
+
"step": 1145
|
| 1423 |
+
},
|
| 1424 |
+
{
|
| 1425 |
+
"epoch": 6.0,
|
| 1426 |
+
"eval_loss": 3.07564377784729,
|
| 1427 |
+
"eval_runtime": 3.4511,
|
| 1428 |
+
"eval_samples_per_second": 76.787,
|
| 1429 |
+
"eval_steps_per_second": 9.852,
|
| 1430 |
+
"step": 1146
|
| 1431 |
+
},
|
| 1432 |
+
{
|
| 1433 |
+
"epoch": 6.02,
|
| 1434 |
+
"learning_rate": 1.4841932382638991e-07,
|
| 1435 |
+
"loss": 2.2549,
|
| 1436 |
+
"step": 1150
|
| 1437 |
+
},
|
| 1438 |
+
{
|
| 1439 |
+
"epoch": 6.05,
|
| 1440 |
+
"learning_rate": 7.502723083396531e-07,
|
| 1441 |
+
"loss": 2.2228,
|
| 1442 |
+
"step": 1155
|
| 1443 |
+
},
|
| 1444 |
+
{
|
| 1445 |
+
"epoch": 6.07,
|
| 1446 |
+
"learning_rate": 1.8107699457751103e-06,
|
| 1447 |
+
"loss": 2.1061,
|
| 1448 |
+
"step": 1160
|
| 1449 |
+
},
|
| 1450 |
+
{
|
| 1451 |
+
"epoch": 6.1,
|
| 1452 |
+
"learning_rate": 3.322743576988504e-06,
|
| 1453 |
+
"loss": 2.1815,
|
| 1454 |
+
"step": 1165
|
| 1455 |
+
},
|
| 1456 |
+
{
|
| 1457 |
+
"epoch": 6.13,
|
| 1458 |
+
"learning_rate": 5.275972694913717e-06,
|
| 1459 |
+
"loss": 2.4782,
|
| 1460 |
+
"step": 1170
|
| 1461 |
+
},
|
| 1462 |
+
{
|
| 1463 |
+
"epoch": 6.15,
|
| 1464 |
+
"learning_rate": 7.657254032251716e-06,
|
| 1465 |
+
"loss": 2.1363,
|
| 1466 |
+
"step": 1175
|
| 1467 |
+
},
|
| 1468 |
+
{
|
| 1469 |
+
"epoch": 6.18,
|
| 1470 |
+
"learning_rate": 1.045049081176232e-05,
|
| 1471 |
+
"loss": 2.2368,
|
| 1472 |
+
"step": 1180
|
| 1473 |
+
},
|
| 1474 |
+
{
|
| 1475 |
+
"epoch": 6.2,
|
| 1476 |
+
"learning_rate": 1.3636801555849831e-05,
|
| 1477 |
+
"loss": 2.2992,
|
| 1478 |
+
"step": 1185
|
| 1479 |
+
},
|
| 1480 |
+
{
|
| 1481 |
+
"epoch": 6.23,
|
| 1482 |
+
"learning_rate": 1.719464771992426e-05,
|
| 1483 |
+
"loss": 2.3285,
|
| 1484 |
+
"step": 1190
|
| 1485 |
+
},
|
| 1486 |
+
{
|
| 1487 |
+
"epoch": 6.26,
|
| 1488 |
+
"learning_rate": 2.1099979286769666e-05,
|
| 1489 |
+
"loss": 2.2711,
|
| 1490 |
+
"step": 1195
|
| 1491 |
+
},
|
| 1492 |
+
{
|
| 1493 |
+
"epoch": 6.28,
|
| 1494 |
+
"learning_rate": 2.5326397337746883e-05,
|
| 1495 |
+
"loss": 2.3187,
|
| 1496 |
+
"step": 1200
|
| 1497 |
+
},
|
| 1498 |
+
{
|
| 1499 |
+
"epoch": 6.31,
|
| 1500 |
+
"learning_rate": 2.9845332501893694e-05,
|
| 1501 |
+
"loss": 2.0419,
|
| 1502 |
+
"step": 1205
|
| 1503 |
+
},
|
| 1504 |
+
{
|
| 1505 |
+
"epoch": 6.34,
|
| 1506 |
+
"learning_rate": 3.462623807665987e-05,
|
| 1507 |
+
"loss": 2.0977,
|
| 1508 |
+
"step": 1210
|
| 1509 |
+
},
|
| 1510 |
+
{
|
| 1511 |
+
"epoch": 6.36,
|
| 1512 |
+
"learning_rate": 3.9636796514833065e-05,
|
| 1513 |
+
"loss": 2.2842,
|
| 1514 |
+
"step": 1215
|
| 1515 |
+
},
|
| 1516 |
+
{
|
| 1517 |
+
"epoch": 6.39,
|
| 1518 |
+
"learning_rate": 4.484313788186857e-05,
|
| 1519 |
+
"loss": 2.3898,
|
| 1520 |
+
"step": 1220
|
| 1521 |
+
},
|
| 1522 |
+
{
|
| 1523 |
+
"epoch": 6.41,
|
| 1524 |
+
"learning_rate": 5.021006880690775e-05,
|
| 1525 |
+
"loss": 2.3513,
|
| 1526 |
+
"step": 1225
|
| 1527 |
+
},
|
| 1528 |
+
{
|
| 1529 |
+
"epoch": 6.44,
|
| 1530 |
+
"learning_rate": 5.570131037984909e-05,
|
| 1531 |
+
"loss": 1.947,
|
| 1532 |
+
"step": 1230
|
| 1533 |
+
},
|
| 1534 |
+
{
|
| 1535 |
+
"epoch": 6.47,
|
| 1536 |
+
"learning_rate": 6.127974338634885e-05,
|
| 1537 |
+
"loss": 2.2545,
|
| 1538 |
+
"step": 1235
|
| 1539 |
+
},
|
| 1540 |
+
{
|
| 1541 |
+
"epoch": 6.49,
|
| 1542 |
+
"learning_rate": 6.690765922303933e-05,
|
| 1543 |
+
"loss": 2.07,
|
| 1544 |
+
"step": 1240
|
| 1545 |
+
},
|
| 1546 |
+
{
|
| 1547 |
+
"epoch": 6.52,
|
| 1548 |
+
"learning_rate": 7.254701479684739e-05,
|
| 1549 |
+
"loss": 2.1936,
|
| 1550 |
+
"step": 1245
|
| 1551 |
+
},
|
| 1552 |
+
{
|
| 1553 |
+
"epoch": 6.54,
|
| 1554 |
+
"learning_rate": 7.815968968536482e-05,
|
| 1555 |
+
"loss": 2.4149,
|
| 1556 |
+
"step": 1250
|
| 1557 |
+
},
|
| 1558 |
+
{
|
| 1559 |
+
"epoch": 6.57,
|
| 1560 |
+
"learning_rate": 8.37077438199465e-05,
|
| 1561 |
+
"loss": 2.3692,
|
| 1562 |
+
"step": 1255
|
| 1563 |
+
},
|
| 1564 |
+
{
|
| 1565 |
+
"epoch": 6.6,
|
| 1566 |
+
"learning_rate": 8.915367394966493e-05,
|
| 1567 |
+
"loss": 2.3514,
|
| 1568 |
+
"step": 1260
|
| 1569 |
+
},
|
| 1570 |
+
{
|
| 1571 |
+
"epoch": 6.62,
|
| 1572 |
+
"learning_rate": 9.446066715250662e-05,
|
| 1573 |
+
"loss": 1.9397,
|
| 1574 |
+
"step": 1265
|
| 1575 |
+
},
|
| 1576 |
+
{
|
| 1577 |
+
"epoch": 6.65,
|
| 1578 |
+
"learning_rate": 9.959284968013953e-05,
|
| 1579 |
+
"loss": 2.3535,
|
| 1580 |
+
"step": 1270
|
| 1581 |
+
},
|
| 1582 |
+
{
|
| 1583 |
+
"epoch": 6.68,
|
| 1584 |
+
"learning_rate": 0.00010451552945414016,
|
| 1585 |
+
"loss": 2.111,
|
| 1586 |
+
"step": 1275
|
| 1587 |
+
},
|
| 1588 |
+
{
|
| 1589 |
+
"epoch": 6.7,
|
| 1590 |
+
"learning_rate": 0.00010919543057446833,
|
| 1591 |
+
"loss": 2.2016,
|
| 1592 |
+
"step": 1280
|
| 1593 |
+
},
|
| 1594 |
+
{
|
| 1595 |
+
"epoch": 6.73,
|
| 1596 |
+
"learning_rate": 0.00011360091825498652,
|
| 1597 |
+
"loss": 2.4254,
|
| 1598 |
+
"step": 1285
|
| 1599 |
+
},
|
| 1600 |
+
{
|
| 1601 |
+
"epoch": 6.75,
|
| 1602 |
+
"learning_rate": 0.00011770221266551781,
|
| 1603 |
+
"loss": 2.4963,
|
| 1604 |
+
"step": 1290
|
| 1605 |
+
},
|
| 1606 |
+
{
|
| 1607 |
+
"epoch": 6.78,
|
| 1608 |
+
"learning_rate": 0.0001214715902349378,
|
| 1609 |
+
"loss": 2.3937,
|
| 1610 |
+
"step": 1295
|
| 1611 |
+
},
|
| 1612 |
+
{
|
| 1613 |
+
"epoch": 6.81,
|
| 1614 |
+
"learning_rate": 0.00012488357105454407,
|
| 1615 |
+
"loss": 2.2234,
|
| 1616 |
+
"step": 1300
|
| 1617 |
+
},
|
| 1618 |
+
{
|
| 1619 |
+
"epoch": 6.83,
|
| 1620 |
+
"learning_rate": 0.0001279150911149141,
|
| 1621 |
+
"loss": 2.3214,
|
| 1622 |
+
"step": 1305
|
| 1623 |
+
},
|
| 1624 |
+
{
|
| 1625 |
+
"epoch": 6.86,
|
| 1626 |
+
"learning_rate": 0.0001305456582119787,
|
| 1627 |
+
"loss": 2.3266,
|
| 1628 |
+
"step": 1310
|
| 1629 |
+
},
|
| 1630 |
+
{
|
| 1631 |
+
"epoch": 6.88,
|
| 1632 |
+
"learning_rate": 0.00013275749046842922,
|
| 1633 |
+
"loss": 2.3065,
|
| 1634 |
+
"step": 1315
|
| 1635 |
+
},
|
| 1636 |
+
{
|
| 1637 |
+
"epoch": 6.91,
|
| 1638 |
+
"learning_rate": 0.00013453563653409702,
|
| 1639 |
+
"loss": 2.4139,
|
| 1640 |
+
"step": 1320
|
| 1641 |
+
},
|
| 1642 |
+
{
|
| 1643 |
+
"epoch": 6.94,
|
| 1644 |
+
"learning_rate": 0.00013586807665278124,
|
| 1645 |
+
"loss": 2.38,
|
| 1646 |
+
"step": 1325
|
| 1647 |
+
},
|
| 1648 |
+
{
|
| 1649 |
+
"epoch": 6.96,
|
| 1650 |
+
"learning_rate": 0.000136745803912346,
|
| 1651 |
+
"loss": 2.3551,
|
| 1652 |
+
"step": 1330
|
| 1653 |
+
},
|
| 1654 |
+
{
|
| 1655 |
+
"epoch": 6.99,
|
| 1656 |
+
"learning_rate": 0.00013716288512885636,
|
| 1657 |
+
"loss": 2.3378,
|
| 1658 |
+
"step": 1335
|
| 1659 |
+
},
|
| 1660 |
+
{
|
| 1661 |
+
"epoch": 7.0,
|
| 1662 |
+
"eval_loss": 3.0805811882019043,
|
| 1663 |
+
"eval_runtime": 3.449,
|
| 1664 |
+
"eval_samples_per_second": 76.834,
|
| 1665 |
+
"eval_steps_per_second": 9.858,
|
| 1666 |
+
"step": 1337
|
| 1667 |
+
},
|
| 1668 |
+
{
|
| 1669 |
+
"epoch": 7.02,
|
| 1670 |
+
"learning_rate": 0.0001371165009531964,
|
| 1671 |
+
"loss": 2.2552,
|
| 1672 |
+
"step": 1340
|
| 1673 |
+
},
|
| 1674 |
+
{
|
| 1675 |
+
"epoch": 7.04,
|
| 1676 |
+
"learning_rate": 0.00013660696492905845,
|
| 1677 |
+
"loss": 2.1842,
|
| 1678 |
+
"step": 1345
|
| 1679 |
+
},
|
| 1680 |
+
{
|
| 1681 |
+
"epoch": 7.07,
|
| 1682 |
+
"learning_rate": 0.00013563772137347802,
|
| 1683 |
+
"loss": 2.097,
|
| 1684 |
+
"step": 1350
|
| 1685 |
+
},
|
| 1686 |
+
{
|
| 1687 |
+
"epoch": 7.09,
|
| 1688 |
+
"learning_rate": 0.00013421532209424082,
|
| 1689 |
+
"loss": 1.7536,
|
| 1690 |
+
"step": 1355
|
| 1691 |
+
},
|
| 1692 |
+
{
|
| 1693 |
+
"epoch": 7.12,
|
| 1694 |
+
"learning_rate": 0.00013234938210154558,
|
| 1695 |
+
"loss": 2.2293,
|
| 1696 |
+
"step": 1360
|
| 1697 |
+
},
|
| 1698 |
+
{
|
| 1699 |
+
"epoch": 7.15,
|
| 1700 |
+
"learning_rate": 0.0001300525146132997,
|
| 1701 |
+
"loss": 2.0082,
|
| 1702 |
+
"step": 1365
|
| 1703 |
+
},
|
| 1704 |
+
{
|
| 1705 |
+
"epoch": 7.17,
|
| 1706 |
+
"learning_rate": 0.00012734024579339186,
|
| 1707 |
+
"loss": 1.951,
|
| 1708 |
+
"step": 1370
|
| 1709 |
+
},
|
| 1710 |
+
{
|
| 1711 |
+
"epoch": 7.2,
|
| 1712 |
+
"learning_rate": 0.00012423090979928834,
|
| 1713 |
+
"loss": 2.3563,
|
| 1714 |
+
"step": 1375
|
| 1715 |
+
},
|
| 1716 |
+
{
|
| 1717 |
+
"epoch": 7.23,
|
| 1718 |
+
"learning_rate": 0.00012074552484839929,
|
| 1719 |
+
"loss": 2.0015,
|
| 1720 |
+
"step": 1380
|
| 1721 |
+
},
|
| 1722 |
+
{
|
| 1723 |
+
"epoch": 7.25,
|
| 1724 |
+
"learning_rate": 0.00011690765114097119,
|
| 1725 |
+
"loss": 2.3068,
|
| 1726 |
+
"step": 1385
|
| 1727 |
+
},
|
| 1728 |
+
{
|
| 1729 |
+
"epoch": 7.28,
|
| 1730 |
+
"learning_rate": 0.00011274323159990744,
|
| 1731 |
+
"loss": 2.2242,
|
| 1732 |
+
"step": 1390
|
| 1733 |
+
},
|
| 1734 |
+
{
|
| 1735 |
+
"epoch": 7.3,
|
| 1736 |
+
"learning_rate": 0.0001082804165040668,
|
| 1737 |
+
"loss": 2.1933,
|
| 1738 |
+
"step": 1395
|
| 1739 |
+
},
|
| 1740 |
+
{
|
| 1741 |
+
"epoch": 7.33,
|
| 1742 |
+
"learning_rate": 0.00010354937320047261,
|
| 1743 |
+
"loss": 2.0466,
|
| 1744 |
+
"step": 1400
|
| 1745 |
+
},
|
| 1746 |
+
{
|
| 1747 |
+
"epoch": 7.36,
|
| 1748 |
+
"learning_rate": 9.858208218171827e-05,
|
| 1749 |
+
"loss": 2.1267,
|
| 1750 |
+
"step": 1405
|
| 1751 |
+
},
|
| 1752 |
+
{
|
| 1753 |
+
"epoch": 7.38,
|
| 1754 |
+
"learning_rate": 9.341212090703316e-05,
|
| 1755 |
+
"loss": 2.08,
|
| 1756 |
+
"step": 1410
|
| 1757 |
+
},
|
| 1758 |
+
{
|
| 1759 |
+
"epoch": 7.41,
|
| 1760 |
+
"learning_rate": 8.807443682830967e-05,
|
| 1761 |
+
"loss": 2.1487,
|
| 1762 |
+
"step": 1415
|
| 1763 |
+
},
|
| 1764 |
+
{
|
| 1765 |
+
"epoch": 7.43,
|
| 1766 |
+
"learning_rate": 8.260511115537746e-05,
|
| 1767 |
+
"loss": 2.0919,
|
| 1768 |
+
"step": 1420
|
| 1769 |
+
},
|
| 1770 |
+
{
|
| 1771 |
+
"epoch": 7.46,
|
| 1772 |
+
"learning_rate": 7.70411149573961e-05,
|
| 1773 |
+
"loss": 2.2303,
|
| 1774 |
+
"step": 1425
|
| 1775 |
+
},
|
| 1776 |
+
{
|
| 1777 |
+
"epoch": 7.49,
|
| 1778 |
+
"learning_rate": 7.142005924905844e-05,
|
| 1779 |
+
"loss": 2.2703,
|
| 1780 |
+
"step": 1430
|
| 1781 |
+
},
|
| 1782 |
+
{
|
| 1783 |
+
"epoch": 7.51,
|
| 1784 |
+
"learning_rate": 6.57799407509417e-05,
|
| 1785 |
+
"loss": 2.1199,
|
| 1786 |
+
"step": 1435
|
| 1787 |
+
},
|
| 1788 |
+
{
|
| 1789 |
+
"epoch": 7.54,
|
| 1790 |
+
"learning_rate": 6.015888504260404e-05,
|
| 1791 |
+
"loss": 2.1918,
|
| 1792 |
+
"step": 1440
|
| 1793 |
+
},
|
| 1794 |
+
{
|
| 1795 |
+
"epoch": 7.57,
|
| 1796 |
+
"learning_rate": 5.4594888844622686e-05,
|
| 1797 |
+
"loss": 2.1859,
|
| 1798 |
+
"step": 1445
|
| 1799 |
+
},
|
| 1800 |
+
{
|
| 1801 |
+
"epoch": 7.59,
|
| 1802 |
+
"learning_rate": 4.912556317169047e-05,
|
| 1803 |
+
"loss": 2.2244,
|
| 1804 |
+
"step": 1450
|
| 1805 |
+
},
|
| 1806 |
+
{
|
| 1807 |
+
"epoch": 7.62,
|
| 1808 |
+
"learning_rate": 4.378787909296698e-05,
|
| 1809 |
+
"loss": 2.0414,
|
| 1810 |
+
"step": 1455
|
| 1811 |
+
},
|
| 1812 |
+
{
|
| 1813 |
+
"epoch": 7.64,
|
| 1814 |
+
"learning_rate": 3.861791781828185e-05,
|
| 1815 |
+
"loss": 2.2903,
|
| 1816 |
+
"step": 1460
|
| 1817 |
+
},
|
| 1818 |
+
{
|
| 1819 |
+
"epoch": 7.67,
|
| 1820 |
+
"learning_rate": 3.365062679952752e-05,
|
| 1821 |
+
"loss": 2.1773,
|
| 1822 |
+
"step": 1465
|
| 1823 |
+
},
|
| 1824 |
+
{
|
| 1825 |
+
"epoch": 7.7,
|
| 1826 |
+
"learning_rate": 2.8919583495933312e-05,
|
| 1827 |
+
"loss": 2.0957,
|
| 1828 |
+
"step": 1470
|
| 1829 |
+
},
|
| 1830 |
+
{
|
| 1831 |
+
"epoch": 7.72,
|
| 1832 |
+
"learning_rate": 2.4456768400092664e-05,
|
| 1833 |
+
"loss": 2.1373,
|
| 1834 |
+
"step": 1475
|
| 1835 |
+
},
|
| 1836 |
+
{
|
| 1837 |
+
"epoch": 7.75,
|
| 1838 |
+
"learning_rate": 2.029234885902891e-05,
|
| 1839 |
+
"loss": 2.2073,
|
| 1840 |
+
"step": 1480
|
| 1841 |
+
},
|
| 1842 |
+
{
|
| 1843 |
+
"epoch": 7.77,
|
| 1844 |
+
"learning_rate": 1.6454475151600793e-05,
|
| 1845 |
+
"loss": 2.0692,
|
| 1846 |
+
"step": 1485
|
| 1847 |
+
},
|
| 1848 |
+
{
|
| 1849 |
+
"epoch": 7.8,
|
| 1850 |
+
"learning_rate": 1.2969090200711758e-05,
|
| 1851 |
+
"loss": 2.1887,
|
| 1852 |
+
"step": 1490
|
| 1853 |
+
},
|
| 1854 |
+
{
|
| 1855 |
+
"epoch": 7.83,
|
| 1856 |
+
"learning_rate": 9.859754206608216e-06,
|
| 1857 |
+
"loss": 2.0833,
|
| 1858 |
+
"step": 1495
|
| 1859 |
+
},
|
| 1860 |
+
{
|
| 1861 |
+
"epoch": 7.85,
|
| 1862 |
+
"learning_rate": 7.147485386700356e-06,
|
| 1863 |
+
"loss": 2.3646,
|
| 1864 |
+
"step": 1500
|
| 1865 |
+
},
|
| 1866 |
+
{
|
| 1867 |
+
"epoch": 7.88,
|
| 1868 |
+
"learning_rate": 4.850617898454462e-06,
|
| 1869 |
+
"loss": 2.2123,
|
| 1870 |
+
"step": 1505
|
| 1871 |
+
},
|
| 1872 |
+
{
|
| 1873 |
+
"epoch": 7.91,
|
| 1874 |
+
"learning_rate": 2.9846779057592115e-06,
|
| 1875 |
+
"loss": 2.0273,
|
| 1876 |
+
"step": 1510
|
| 1877 |
+
},
|
| 1878 |
+
{
|
| 1879 |
+
"epoch": 7.93,
|
| 1880 |
+
"learning_rate": 1.5622786265220057e-06,
|
| 1881 |
+
"loss": 1.9322,
|
| 1882 |
+
"step": 1515
|
| 1883 |
+
},
|
| 1884 |
+
{
|
| 1885 |
+
"epoch": 7.96,
|
| 1886 |
+
"learning_rate": 5.93035070941549e-07,
|
| 1887 |
+
"loss": 2.0401,
|
| 1888 |
+
"step": 1520
|
| 1889 |
+
},
|
| 1890 |
+
{
|
| 1891 |
+
"epoch": 7.98,
|
| 1892 |
+
"learning_rate": 8.349904680360514e-08,
|
| 1893 |
+
"loss": 2.0603,
|
| 1894 |
+
"step": 1525
|
| 1895 |
+
},
|
| 1896 |
+
{
|
| 1897 |
+
"epoch": 8.0,
|
| 1898 |
+
"eval_loss": 3.059950828552246,
|
| 1899 |
+
"eval_runtime": 3.4514,
|
| 1900 |
+
"eval_samples_per_second": 76.78,
|
| 1901 |
+
"eval_steps_per_second": 9.851,
|
| 1902 |
+
"step": 1528
|
| 1903 |
+
},
|
| 1904 |
+
{
|
| 1905 |
+
"epoch": 8.01,
|
| 1906 |
+
"learning_rate": 3.711487114362266e-08,
|
| 1907 |
+
"loss": 2.0764,
|
| 1908 |
+
"step": 1530
|
| 1909 |
+
},
|
| 1910 |
+
{
|
| 1911 |
+
"epoch": 8.04,
|
| 1912 |
+
"learning_rate": 4.5419608765397886e-07,
|
| 1913 |
+
"loss": 1.9304,
|
| 1914 |
+
"step": 1535
|
| 1915 |
+
},
|
| 1916 |
+
{
|
| 1917 |
+
"epoch": 8.06,
|
| 1918 |
+
"learning_rate": 1.3319233472187234e-06,
|
| 1919 |
+
"loss": 1.8933,
|
| 1920 |
+
"step": 1540
|
| 1921 |
+
},
|
| 1922 |
+
{
|
| 1923 |
+
"epoch": 8.09,
|
| 1924 |
+
"learning_rate": 2.664363465902945e-06,
|
| 1925 |
+
"loss": 2.2206,
|
| 1926 |
+
"step": 1545
|
| 1927 |
+
},
|
| 1928 |
+
{
|
| 1929 |
+
"epoch": 8.12,
|
| 1930 |
+
"learning_rate": 4.44250953157072e-06,
|
| 1931 |
+
"loss": 2.0486,
|
| 1932 |
+
"step": 1550
|
| 1933 |
+
},
|
| 1934 |
+
{
|
| 1935 |
+
"epoch": 8.14,
|
| 1936 |
+
"learning_rate": 6.654341788021246e-06,
|
| 1937 |
+
"loss": 1.8876,
|
| 1938 |
+
"step": 1555
|
| 1939 |
+
},
|
| 1940 |
+
{
|
| 1941 |
+
"epoch": 8.17,
|
| 1942 |
+
"learning_rate": 9.28490888508584e-06,
|
| 1943 |
+
"loss": 1.7926,
|
| 1944 |
+
"step": 1560
|
| 1945 |
+
},
|
| 1946 |
+
{
|
| 1947 |
+
"epoch": 8.19,
|
| 1948 |
+
"learning_rate": 1.2316428945455863e-05,
|
| 1949 |
+
"loss": 1.7425,
|
| 1950 |
+
"step": 1565
|
| 1951 |
+
},
|
| 1952 |
+
{
|
| 1953 |
+
"epoch": 8.22,
|
| 1954 |
+
"learning_rate": 1.572840976506211e-05,
|
| 1955 |
+
"loss": 1.9577,
|
| 1956 |
+
"step": 1570
|
| 1957 |
+
},
|
| 1958 |
+
{
|
| 1959 |
+
"epoch": 8.25,
|
| 1960 |
+
"learning_rate": 1.9497787334482106e-05,
|
| 1961 |
+
"loss": 2.0153,
|
| 1962 |
+
"step": 1575
|
| 1963 |
+
},
|
| 1964 |
+
{
|
| 1965 |
+
"epoch": 8.27,
|
| 1966 |
+
"learning_rate": 2.3599081745013365e-05,
|
| 1967 |
+
"loss": 1.9377,
|
| 1968 |
+
"step": 1580
|
| 1969 |
+
},
|
| 1970 |
+
{
|
| 1971 |
+
"epoch": 8.3,
|
| 1972 |
+
"learning_rate": 2.800456942553155e-05,
|
| 1973 |
+
"loss": 2.2322,
|
| 1974 |
+
"step": 1585
|
| 1975 |
+
},
|
| 1976 |
+
{
|
| 1977 |
+
"epoch": 8.32,
|
| 1978 |
+
"learning_rate": 3.268447054585973e-05,
|
| 1979 |
+
"loss": 1.7475,
|
| 1980 |
+
"step": 1590
|
| 1981 |
+
},
|
| 1982 |
+
{
|
| 1983 |
+
"epoch": 8.35,
|
| 1984 |
+
"learning_rate": 3.760715031986036e-05,
|
| 1985 |
+
"loss": 1.8289,
|
| 1986 |
+
"step": 1595
|
| 1987 |
+
},
|
| 1988 |
+
{
|
| 1989 |
+
"epoch": 8.38,
|
| 1990 |
+
"learning_rate": 4.273933284749324e-05,
|
| 1991 |
+
"loss": 1.8663,
|
| 1992 |
+
"step": 1600
|
| 1993 |
+
},
|
| 1994 |
+
{
|
| 1995 |
+
"epoch": 8.4,
|
| 1996 |
+
"learning_rate": 4.804632605033493e-05,
|
| 1997 |
+
"loss": 2.0319,
|
| 1998 |
+
"step": 1605
|
| 1999 |
+
},
|
| 2000 |
+
{
|
| 2001 |
+
"epoch": 8.43,
|
| 2002 |
+
"learning_rate": 5.349225618005337e-05,
|
| 2003 |
+
"loss": 2.1028,
|
| 2004 |
+
"step": 1610
|
| 2005 |
+
},
|
| 2006 |
+
{
|
| 2007 |
+
"epoch": 8.46,
|
| 2008 |
+
"learning_rate": 5.9040310314635035e-05,
|
| 2009 |
+
"loss": 1.9332,
|
| 2010 |
+
"step": 1615
|
| 2011 |
+
},
|
| 2012 |
+
{
|
| 2013 |
+
"epoch": 8.48,
|
| 2014 |
+
"learning_rate": 6.465298520315248e-05,
|
| 2015 |
+
"loss": 1.8289,
|
| 2016 |
+
"step": 1620
|
| 2017 |
+
},
|
| 2018 |
+
{
|
| 2019 |
+
"epoch": 8.51,
|
| 2020 |
+
"learning_rate": 7.029234077696052e-05,
|
| 2021 |
+
"loss": 1.8439,
|
| 2022 |
+
"step": 1625
|
| 2023 |
+
},
|
| 2024 |
+
{
|
| 2025 |
+
"epoch": 8.53,
|
| 2026 |
+
"learning_rate": 7.5920256613651e-05,
|
| 2027 |
+
"loss": 1.7968,
|
| 2028 |
+
"step": 1630
|
| 2029 |
+
},
|
| 2030 |
+
{
|
| 2031 |
+
"epoch": 8.56,
|
| 2032 |
+
"learning_rate": 8.149868962015078e-05,
|
| 2033 |
+
"loss": 2.0554,
|
| 2034 |
+
"step": 1635
|
| 2035 |
+
},
|
| 2036 |
+
{
|
| 2037 |
+
"epoch": 8.59,
|
| 2038 |
+
"learning_rate": 8.698993119309211e-05,
|
| 2039 |
+
"loss": 2.0439,
|
| 2040 |
+
"step": 1640
|
| 2041 |
+
},
|
| 2042 |
+
{
|
| 2043 |
+
"epoch": 8.61,
|
| 2044 |
+
"learning_rate": 9.235686211813131e-05,
|
| 2045 |
+
"loss": 2.1254,
|
| 2046 |
+
"step": 1645
|
| 2047 |
+
},
|
| 2048 |
+
{
|
| 2049 |
+
"epoch": 8.64,
|
| 2050 |
+
"learning_rate": 9.75632034851668e-05,
|
| 2051 |
+
"loss": 1.8997,
|
| 2052 |
+
"step": 1650
|
| 2053 |
+
},
|
| 2054 |
+
{
|
| 2055 |
+
"epoch": 8.66,
|
| 2056 |
+
"learning_rate": 0.00010257376192334,
|
| 2057 |
+
"loss": 1.8282,
|
| 2058 |
+
"step": 1655
|
| 2059 |
+
},
|
| 2060 |
+
{
|
| 2061 |
+
"epoch": 8.69,
|
| 2062 |
+
"learning_rate": 0.00010735466749810619,
|
| 2063 |
+
"loss": 1.977,
|
| 2064 |
+
"step": 1660
|
| 2065 |
+
},
|
| 2066 |
+
{
|
| 2067 |
+
"epoch": 8.72,
|
| 2068 |
+
"learning_rate": 0.000111873602662253,
|
| 2069 |
+
"loss": 1.9521,
|
| 2070 |
+
"step": 1665
|
| 2071 |
+
},
|
| 2072 |
+
{
|
| 2073 |
+
"epoch": 8.74,
|
| 2074 |
+
"learning_rate": 0.00011610002071323023,
|
| 2075 |
+
"loss": 2.227,
|
| 2076 |
+
"step": 1670
|
| 2077 |
+
},
|
| 2078 |
+
{
|
| 2079 |
+
"epoch": 8.77,
|
| 2080 |
+
"learning_rate": 0.00012000535228007549,
|
| 2081 |
+
"loss": 1.9954,
|
| 2082 |
+
"step": 1675
|
| 2083 |
+
},
|
| 2084 |
+
{
|
| 2085 |
+
"epoch": 8.8,
|
| 2086 |
+
"learning_rate": 0.00012356319844414993,
|
| 2087 |
+
"loss": 2.1384,
|
| 2088 |
+
"step": 1680
|
| 2089 |
+
},
|
| 2090 |
+
{
|
| 2091 |
+
"epoch": 8.82,
|
| 2092 |
+
"learning_rate": 0.00012674950918823747,
|
| 2093 |
+
"loss": 1.9971,
|
| 2094 |
+
"step": 1685
|
| 2095 |
+
},
|
| 2096 |
+
{
|
| 2097 |
+
"epoch": 8.85,
|
| 2098 |
+
"learning_rate": 0.0001295427459677481,
|
| 2099 |
+
"loss": 1.9247,
|
| 2100 |
+
"step": 1690
|
| 2101 |
+
},
|
| 2102 |
+
{
|
| 2103 |
+
"epoch": 8.87,
|
| 2104 |
+
"learning_rate": 0.00013192402730508612,
|
| 2105 |
+
"loss": 2.0361,
|
| 2106 |
+
"step": 1695
|
| 2107 |
+
},
|
| 2108 |
+
{
|
| 2109 |
+
"epoch": 8.9,
|
| 2110 |
+
"learning_rate": 0.00013387725642301137,
|
| 2111 |
+
"loss": 1.9686,
|
| 2112 |
+
"step": 1700
|
| 2113 |
+
},
|
| 2114 |
+
{
|
| 2115 |
+
"epoch": 8.93,
|
| 2116 |
+
"learning_rate": 0.0001353892300542248,
|
| 2117 |
+
"loss": 2.1605,
|
| 2118 |
+
"step": 1705
|
| 2119 |
+
},
|
| 2120 |
+
{
|
| 2121 |
+
"epoch": 8.95,
|
| 2122 |
+
"learning_rate": 0.0001364497276916603,
|
| 2123 |
+
"loss": 2.12,
|
| 2124 |
+
"step": 1710
|
| 2125 |
+
},
|
| 2126 |
+
{
|
| 2127 |
+
"epoch": 8.98,
|
| 2128 |
+
"learning_rate": 0.0001370515806761736,
|
| 2129 |
+
"loss": 2.0884,
|
| 2130 |
+
"step": 1715
|
| 2131 |
+
},
|
| 2132 |
+
{
|
| 2133 |
+
"epoch": 9.0,
|
| 2134 |
+
"eval_loss": 3.1394710540771484,
|
| 2135 |
+
"eval_runtime": 3.4497,
|
| 2136 |
+
"eval_samples_per_second": 76.819,
|
| 2137 |
+
"eval_steps_per_second": 9.856,
|
| 2138 |
+
"step": 1719
|
| 2139 |
+
},
|
| 2140 |
+
{
|
| 2141 |
+
"epoch": 9.01,
|
| 2142 |
+
"learning_rate": 0.0001371907206546175,
|
| 2143 |
+
"loss": 2.0551,
|
| 2144 |
+
"step": 1720
|
| 2145 |
+
},
|
| 2146 |
+
{
|
| 2147 |
+
"epoch": 9.03,
|
| 2148 |
+
"learning_rate": 0.00013686620708073944,
|
| 2149 |
+
"loss": 1.8553,
|
| 2150 |
+
"step": 1725
|
| 2151 |
+
},
|
| 2152 |
+
{
|
| 2153 |
+
"epoch": 9.06,
|
| 2154 |
+
"learning_rate": 0.00013608023357300367,
|
| 2155 |
+
"loss": 1.9159,
|
| 2156 |
+
"step": 1730
|
| 2157 |
+
},
|
| 2158 |
+
{
|
| 2159 |
+
"epoch": 9.08,
|
| 2160 |
+
"learning_rate": 0.0001348381130863611,
|
| 2161 |
+
"loss": 1.7198,
|
| 2162 |
+
"step": 1735
|
| 2163 |
+
},
|
| 2164 |
+
{
|
| 2165 |
+
"epoch": 9.11,
|
| 2166 |
+
"learning_rate": 0.0001331482419982015,
|
| 2167 |
+
"loss": 1.8409,
|
| 2168 |
+
"step": 1740
|
| 2169 |
+
},
|
| 2170 |
+
{
|
| 2171 |
+
"epoch": 9.14,
|
| 2172 |
+
"learning_rate": 0.00013102204335125575,
|
| 2173 |
+
"loss": 1.8302,
|
| 2174 |
+
"step": 1745
|
| 2175 |
+
},
|
| 2176 |
+
{
|
| 2177 |
+
"epoch": 9.16,
|
| 2178 |
+
"learning_rate": 0.00012847388963711166,
|
| 2179 |
+
"loss": 1.8527,
|
| 2180 |
+
"step": 1750
|
| 2181 |
+
},
|
| 2182 |
+
{
|
| 2183 |
+
"epoch": 9.19,
|
| 2184 |
+
"learning_rate": 0.00012552100564230202,
|
| 2185 |
+
"loss": 2.1229,
|
| 2186 |
+
"step": 1755
|
| 2187 |
+
},
|
| 2188 |
+
{
|
| 2189 |
+
"epoch": 9.21,
|
| 2190 |
+
"learning_rate": 0.0001221833520137008,
|
| 2191 |
+
"loss": 1.8018,
|
| 2192 |
+
"step": 1760
|
| 2193 |
+
},
|
| 2194 |
+
{
|
| 2195 |
+
"epoch": 9.24,
|
| 2196 |
+
"learning_rate": 0.00011848349033028711,
|
| 2197 |
+
"loss": 2.0999,
|
| 2198 |
+
"step": 1765
|
| 2199 |
+
},
|
| 2200 |
+
{
|
| 2201 |
+
"epoch": 9.27,
|
| 2202 |
+
"learning_rate": 0.00011444643059335979,
|
| 2203 |
+
"loss": 1.8556,
|
| 2204 |
+
"step": 1770
|
| 2205 |
+
},
|
| 2206 |
+
{
|
| 2207 |
+
"epoch": 9.29,
|
| 2208 |
+
"learning_rate": 0.00011009946216611662,
|
| 2209 |
+
"loss": 1.918,
|
| 2210 |
+
"step": 1775
|
| 2211 |
+
},
|
| 2212 |
+
{
|
| 2213 |
+
"epoch": 9.32,
|
| 2214 |
+
"learning_rate": 0.00010547196930540541,
|
| 2215 |
+
"loss": 1.8345,
|
| 2216 |
+
"step": 1780
|
| 2217 |
+
},
|
| 2218 |
+
{
|
| 2219 |
+
"epoch": 9.35,
|
| 2220 |
+
"learning_rate": 0.0001005952325325969,
|
| 2221 |
+
"loss": 1.9419,
|
| 2222 |
+
"step": 1785
|
| 2223 |
+
},
|
| 2224 |
+
{
|
| 2225 |
+
"epoch": 9.37,
|
| 2226 |
+
"learning_rate": 9.550221718625064e-05,
|
| 2227 |
+
"loss": 1.7719,
|
| 2228 |
+
"step": 1790
|
| 2229 |
+
},
|
| 2230 |
+
{
|
| 2231 |
+
"epoch": 9.4,
|
| 2232 |
+
"learning_rate": 9.022735058590532e-05,
|
| 2233 |
+
"loss": 1.6571,
|
| 2234 |
+
"step": 1795
|
| 2235 |
+
},
|
| 2236 |
+
{
|
| 2237 |
+
"epoch": 9.42,
|
| 2238 |
+
"learning_rate": 8.480628931329531e-05,
|
| 2239 |
+
"loss": 1.7121,
|
| 2240 |
+
"step": 1800
|
| 2241 |
+
},
|
| 2242 |
+
{
|
| 2243 |
+
"epoch": 9.45,
|
| 2244 |
+
"learning_rate": 7.92756781841111e-05,
|
| 2245 |
+
"loss": 1.8914,
|
| 2246 |
+
"step": 1805
|
| 2247 |
+
},
|
| 2248 |
+
{
|
| 2249 |
+
"epoch": 9.48,
|
| 2250 |
+
"learning_rate": 7.367290253957744e-05,
|
| 2251 |
+
"loss": 2.0196,
|
| 2252 |
+
"step": 1810
|
| 2253 |
+
},
|
| 2254 |
+
{
|
| 2255 |
+
"epoch": 9.5,
|
| 2256 |
+
"learning_rate": 6.80358355322966e-05,
|
| 2257 |
+
"loss": 1.7844,
|
| 2258 |
+
"step": 1815
|
| 2259 |
+
},
|
| 2260 |
+
{
|
| 2261 |
+
"epoch": 9.53,
|
| 2262 |
+
"learning_rate": 6.240258211462402e-05,
|
| 2263 |
+
"loss": 1.9618,
|
| 2264 |
+
"step": 1820
|
| 2265 |
+
},
|
| 2266 |
+
{
|
| 2267 |
+
"epoch": 9.55,
|
| 2268 |
+
"learning_rate": 5.681122146014721e-05,
|
| 2269 |
+
"loss": 1.9428,
|
| 2270 |
+
"step": 1825
|
| 2271 |
+
},
|
| 2272 |
+
{
|
| 2273 |
+
"epoch": 9.58,
|
| 2274 |
+
"learning_rate": 5.129954955941689e-05,
|
| 2275 |
+
"loss": 1.7754,
|
| 2276 |
+
"step": 1830
|
| 2277 |
+
},
|
| 2278 |
+
{
|
| 2279 |
+
"epoch": 9.61,
|
| 2280 |
+
"learning_rate": 4.590482372991883e-05,
|
| 2281 |
+
"loss": 1.9829,
|
| 2282 |
+
"step": 1835
|
| 2283 |
+
},
|
| 2284 |
+
{
|
| 2285 |
+
"epoch": 9.63,
|
| 2286 |
+
"learning_rate": 4.066351076731789e-05,
|
| 2287 |
+
"loss": 1.8845,
|
| 2288 |
+
"step": 1840
|
| 2289 |
+
},
|
| 2290 |
+
{
|
| 2291 |
+
"epoch": 9.66,
|
| 2292 |
+
"learning_rate": 3.561104044040515e-05,
|
| 2293 |
+
"loss": 1.9994,
|
| 2294 |
+
"step": 1845
|
| 2295 |
+
},
|
| 2296 |
+
{
|
| 2297 |
+
"epoch": 9.69,
|
| 2298 |
+
"learning_rate": 3.078156599604378e-05,
|
| 2299 |
+
"loss": 1.826,
|
| 2300 |
+
"step": 1850
|
| 2301 |
+
},
|
| 2302 |
+
{
|
| 2303 |
+
"epoch": 9.71,
|
| 2304 |
+
"learning_rate": 2.6207733293040633e-05,
|
| 2305 |
+
"loss": 1.8695,
|
| 2306 |
+
"step": 1855
|
| 2307 |
+
},
|
| 2308 |
+
{
|
| 2309 |
+
"epoch": 9.74,
|
| 2310 |
+
"learning_rate": 2.1920460125527168e-05,
|
| 2311 |
+
"loss": 1.9426,
|
| 2312 |
+
"step": 1860
|
| 2313 |
+
},
|
| 2314 |
+
{
|
| 2315 |
+
"epoch": 9.76,
|
| 2316 |
+
"learning_rate": 1.7948727227557404e-05,
|
| 2317 |
+
"loss": 1.7794,
|
| 2318 |
+
"step": 1865
|
| 2319 |
+
},
|
| 2320 |
+
{
|
| 2321 |
+
"epoch": 9.79,
|
| 2322 |
+
"learning_rate": 1.4319382371679026e-05,
|
| 2323 |
+
"loss": 1.5958,
|
| 2324 |
+
"step": 1870
|
| 2325 |
+
},
|
| 2326 |
+
{
|
| 2327 |
+
"epoch": 9.82,
|
| 2328 |
+
"learning_rate": 1.105695888570886e-05,
|
| 2329 |
+
"loss": 1.9165,
|
| 2330 |
+
"step": 1875
|
| 2331 |
+
},
|
| 2332 |
+
{
|
| 2333 |
+
"epoch": 9.84,
|
| 2334 |
+
"learning_rate": 8.18350981449102e-06,
|
| 2335 |
+
"loss": 1.8373,
|
| 2336 |
+
"step": 1880
|
| 2337 |
+
},
|
| 2338 |
+
{
|
| 2339 |
+
"epoch": 9.87,
|
| 2340 |
+
"learning_rate": 5.7184588476522e-06,
|
| 2341 |
+
"loss": 1.9335,
|
| 2342 |
+
"step": 1885
|
| 2343 |
+
},
|
| 2344 |
+
{
|
| 2345 |
+
"epoch": 9.9,
|
| 2346 |
+
"learning_rate": 3.6784690210450856e-06,
|
| 2347 |
+
"loss": 1.7418,
|
| 2348 |
+
"step": 1890
|
| 2349 |
+
},
|
| 2350 |
+
{
|
| 2351 |
+
"epoch": 9.92,
|
| 2352 |
+
"learning_rate": 2.0773300794160803e-06,
|
| 2353 |
+
"loss": 2.0059,
|
| 2354 |
+
"step": 1895
|
| 2355 |
+
},
|
| 2356 |
+
{
|
| 2357 |
+
"epoch": 9.95,
|
| 2358 |
+
"learning_rate": 9.258652616957751e-07,
|
| 2359 |
+
"loss": 1.7969,
|
| 2360 |
+
"step": 1900
|
| 2361 |
+
},
|
| 2362 |
+
{
|
| 2363 |
+
"epoch": 9.97,
|
| 2364 |
+
"learning_rate": 2.3185813901453346e-07,
|
| 2365 |
+
"loss": 1.9667,
|
| 2366 |
+
"step": 1905
|
| 2367 |
+
},
|
| 2368 |
+
{
|
| 2369 |
+
"epoch": 10.0,
|
| 2370 |
+
"learning_rate": 0.0,
|
| 2371 |
+
"loss": 2.0133,
|
| 2372 |
+
"step": 1910
|
| 2373 |
+
},
|
| 2374 |
+
{
|
| 2375 |
+
"epoch": 10.0,
|
| 2376 |
+
"eval_loss": 3.0807313919067383,
|
| 2377 |
+
"eval_runtime": 3.4521,
|
| 2378 |
+
"eval_samples_per_second": 76.764,
|
| 2379 |
+
"eval_steps_per_second": 9.849,
|
| 2380 |
+
"step": 1910
|
| 2381 |
+
},
|
| 2382 |
+
{
|
| 2383 |
+
"epoch": 10.03,
|
| 2384 |
+
"learning_rate": 2.318581390145182e-07,
|
| 2385 |
+
"loss": 1.8296,
|
| 2386 |
+
"step": 1915
|
| 2387 |
+
},
|
| 2388 |
+
{
|
| 2389 |
+
"epoch": 10.05,
|
| 2390 |
+
"learning_rate": 9.258652616957447e-07,
|
| 2391 |
+
"loss": 1.6133,
|
| 2392 |
+
"step": 1920
|
| 2393 |
+
},
|
| 2394 |
+
{
|
| 2395 |
+
"epoch": 10.08,
|
| 2396 |
+
"learning_rate": 2.0773300794160346e-06,
|
| 2397 |
+
"loss": 1.7176,
|
| 2398 |
+
"step": 1925
|
| 2399 |
+
},
|
| 2400 |
+
{
|
| 2401 |
+
"epoch": 10.1,
|
| 2402 |
+
"learning_rate": 3.6784690210450323e-06,
|
| 2403 |
+
"loss": 1.7257,
|
| 2404 |
+
"step": 1930
|
| 2405 |
+
},
|
| 2406 |
+
{
|
| 2407 |
+
"epoch": 10.13,
|
| 2408 |
+
"learning_rate": 5.718458847652131e-06,
|
| 2409 |
+
"loss": 1.7978,
|
| 2410 |
+
"step": 1935
|
| 2411 |
+
},
|
| 2412 |
+
{
|
| 2413 |
+
"epoch": 10.16,
|
| 2414 |
+
"learning_rate": 8.183509814490936e-06,
|
| 2415 |
+
"loss": 1.798,
|
| 2416 |
+
"step": 1940
|
| 2417 |
+
},
|
| 2418 |
+
{
|
| 2419 |
+
"epoch": 10.18,
|
| 2420 |
+
"learning_rate": 1.1056958885708754e-05,
|
| 2421 |
+
"loss": 1.6063,
|
| 2422 |
+
"step": 1945
|
| 2423 |
+
},
|
| 2424 |
+
{
|
| 2425 |
+
"epoch": 10.21,
|
| 2426 |
+
"learning_rate": 1.4319382371678919e-05,
|
| 2427 |
+
"loss": 1.5478,
|
| 2428 |
+
"step": 1950
|
| 2429 |
+
},
|
| 2430 |
+
{
|
| 2431 |
+
"epoch": 10.24,
|
| 2432 |
+
"learning_rate": 1.7948727227557282e-05,
|
| 2433 |
+
"loss": 1.6348,
|
| 2434 |
+
"step": 1955
|
| 2435 |
+
},
|
| 2436 |
+
{
|
| 2437 |
+
"epoch": 10.26,
|
| 2438 |
+
"learning_rate": 2.1920460125527032e-05,
|
| 2439 |
+
"loss": 1.7918,
|
| 2440 |
+
"step": 1960
|
| 2441 |
+
},
|
| 2442 |
+
{
|
| 2443 |
+
"epoch": 10.29,
|
| 2444 |
+
"learning_rate": 2.6207733293040488e-05,
|
| 2445 |
+
"loss": 1.7877,
|
| 2446 |
+
"step": 1965
|
| 2447 |
+
},
|
| 2448 |
+
{
|
| 2449 |
+
"epoch": 10.31,
|
| 2450 |
+
"learning_rate": 3.078156599604362e-05,
|
| 2451 |
+
"loss": 1.7034,
|
| 2452 |
+
"step": 1970
|
| 2453 |
+
},
|
| 2454 |
+
{
|
| 2455 |
+
"epoch": 10.34,
|
| 2456 |
+
"learning_rate": 3.561104044040498e-05,
|
| 2457 |
+
"loss": 1.6054,
|
| 2458 |
+
"step": 1975
|
| 2459 |
+
},
|
| 2460 |
+
{
|
| 2461 |
+
"epoch": 10.37,
|
| 2462 |
+
"learning_rate": 4.066351076731773e-05,
|
| 2463 |
+
"loss": 1.7727,
|
| 2464 |
+
"step": 1980
|
| 2465 |
+
},
|
| 2466 |
+
{
|
| 2467 |
+
"epoch": 10.39,
|
| 2468 |
+
"learning_rate": 4.5904823729918646e-05,
|
| 2469 |
+
"loss": 1.7093,
|
| 2470 |
+
"step": 1985
|
| 2471 |
+
},
|
| 2472 |
+
{
|
| 2473 |
+
"epoch": 10.42,
|
| 2474 |
+
"learning_rate": 5.1299549559416716e-05,
|
| 2475 |
+
"loss": 1.5078,
|
| 2476 |
+
"step": 1990
|
| 2477 |
+
},
|
| 2478 |
+
{
|
| 2479 |
+
"epoch": 10.45,
|
| 2480 |
+
"learning_rate": 5.681122146014703e-05,
|
| 2481 |
+
"loss": 1.7173,
|
| 2482 |
+
"step": 1995
|
| 2483 |
+
},
|
| 2484 |
+
{
|
| 2485 |
+
"epoch": 10.47,
|
| 2486 |
+
"learning_rate": 6.240258211462383e-05,
|
| 2487 |
+
"loss": 2.0244,
|
| 2488 |
+
"step": 2000
|
| 2489 |
+
},
|
| 2490 |
+
{
|
| 2491 |
+
"epoch": 10.5,
|
| 2492 |
+
"learning_rate": 6.803583553229642e-05,
|
| 2493 |
+
"loss": 1.5563,
|
| 2494 |
+
"step": 2005
|
| 2495 |
+
},
|
| 2496 |
+
{
|
| 2497 |
+
"epoch": 10.52,
|
| 2498 |
+
"learning_rate": 7.367290253957726e-05,
|
| 2499 |
+
"loss": 1.587,
|
| 2500 |
+
"step": 2010
|
| 2501 |
+
},
|
| 2502 |
+
{
|
| 2503 |
+
"epoch": 10.55,
|
| 2504 |
+
"learning_rate": 7.927567818411092e-05,
|
| 2505 |
+
"loss": 1.6739,
|
| 2506 |
+
"step": 2015
|
| 2507 |
+
},
|
| 2508 |
+
{
|
| 2509 |
+
"epoch": 10.58,
|
| 2510 |
+
"learning_rate": 8.480628931329514e-05,
|
| 2511 |
+
"loss": 1.7724,
|
| 2512 |
+
"step": 2020
|
| 2513 |
+
},
|
| 2514 |
+
{
|
| 2515 |
+
"epoch": 10.6,
|
| 2516 |
+
"learning_rate": 9.022735058590513e-05,
|
| 2517 |
+
"loss": 1.7843,
|
| 2518 |
+
"step": 2025
|
| 2519 |
+
},
|
| 2520 |
+
{
|
| 2521 |
+
"epoch": 10.63,
|
| 2522 |
+
"learning_rate": 9.550221718625046e-05,
|
| 2523 |
+
"loss": 1.7604,
|
| 2524 |
+
"step": 2030
|
| 2525 |
+
},
|
| 2526 |
+
{
|
| 2527 |
+
"epoch": 10.65,
|
| 2528 |
+
"learning_rate": 0.00010059523253259673,
|
| 2529 |
+
"loss": 1.8025,
|
| 2530 |
+
"step": 2035
|
| 2531 |
+
},
|
| 2532 |
+
{
|
| 2533 |
+
"epoch": 10.68,
|
| 2534 |
+
"learning_rate": 0.00010547196930540525,
|
| 2535 |
+
"loss": 1.6344,
|
| 2536 |
+
"step": 2040
|
| 2537 |
+
},
|
| 2538 |
+
{
|
| 2539 |
+
"epoch": 10.71,
|
| 2540 |
+
"learning_rate": 0.00011009946216611647,
|
| 2541 |
+
"loss": 1.8952,
|
| 2542 |
+
"step": 2045
|
| 2543 |
+
},
|
| 2544 |
+
{
|
| 2545 |
+
"epoch": 10.73,
|
| 2546 |
+
"learning_rate": 0.00011444643059335966,
|
| 2547 |
+
"loss": 1.5879,
|
| 2548 |
+
"step": 2050
|
| 2549 |
+
},
|
| 2550 |
+
{
|
| 2551 |
+
"epoch": 10.76,
|
| 2552 |
+
"learning_rate": 0.00011848349033028699,
|
| 2553 |
+
"loss": 1.7946,
|
| 2554 |
+
"step": 2055
|
| 2555 |
+
},
|
| 2556 |
+
{
|
| 2557 |
+
"epoch": 10.79,
|
| 2558 |
+
"learning_rate": 0.00012218335201370068,
|
| 2559 |
+
"loss": 1.6615,
|
| 2560 |
+
"step": 2060
|
| 2561 |
+
},
|
| 2562 |
+
{
|
| 2563 |
+
"epoch": 10.81,
|
| 2564 |
+
"learning_rate": 0.0001255210056423019,
|
| 2565 |
+
"loss": 1.6975,
|
| 2566 |
+
"step": 2065
|
| 2567 |
+
},
|
| 2568 |
+
{
|
| 2569 |
+
"epoch": 10.84,
|
| 2570 |
+
"learning_rate": 0.00012847388963711158,
|
| 2571 |
+
"loss": 1.465,
|
| 2572 |
+
"step": 2070
|
| 2573 |
+
},
|
| 2574 |
+
{
|
| 2575 |
+
"epoch": 10.86,
|
| 2576 |
+
"learning_rate": 0.00013102204335125564,
|
| 2577 |
+
"loss": 2.0658,
|
| 2578 |
+
"step": 2075
|
| 2579 |
+
},
|
| 2580 |
+
{
|
| 2581 |
+
"epoch": 10.89,
|
| 2582 |
+
"learning_rate": 0.00013314824199820135,
|
| 2583 |
+
"loss": 1.9283,
|
| 2584 |
+
"step": 2080
|
| 2585 |
+
},
|
| 2586 |
+
{
|
| 2587 |
+
"epoch": 10.92,
|
| 2588 |
+
"learning_rate": 0.00013483811308636113,
|
| 2589 |
+
"loss": 1.8069,
|
| 2590 |
+
"step": 2085
|
| 2591 |
+
},
|
| 2592 |
+
{
|
| 2593 |
+
"epoch": 10.94,
|
| 2594 |
+
"learning_rate": 0.00013608023357300364,
|
| 2595 |
+
"loss": 1.7805,
|
| 2596 |
+
"step": 2090
|
| 2597 |
+
},
|
| 2598 |
+
{
|
| 2599 |
+
"epoch": 10.97,
|
| 2600 |
+
"learning_rate": 0.0001368662070807394,
|
| 2601 |
+
"loss": 1.6908,
|
| 2602 |
+
"step": 2095
|
| 2603 |
+
},
|
| 2604 |
+
{
|
| 2605 |
+
"epoch": 10.99,
|
| 2606 |
+
"learning_rate": 0.0001371907206546175,
|
| 2607 |
+
"loss": 1.8003,
|
| 2608 |
+
"step": 2100
|
| 2609 |
+
},
|
| 2610 |
+
{
|
| 2611 |
+
"epoch": 11.0,
|
| 2612 |
+
"eval_loss": 3.0452535152435303,
|
| 2613 |
+
"eval_runtime": 3.4493,
|
| 2614 |
+
"eval_samples_per_second": 76.827,
|
| 2615 |
+
"eval_steps_per_second": 9.857,
|
| 2616 |
+
"step": 2101
|
| 2617 |
}
|
| 2618 |
],
|
| 2619 |
+
"max_steps": 11651,
|
| 2620 |
+
"num_train_epochs": 61,
|
| 2621 |
+
"total_flos": 2193024024576000.0,
|
| 2622 |
"trial_name": null,
|
| 2623 |
"trial_params": null
|
| 2624 |
}
|
training_args.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 2863
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:0b09f999d242a890a3a5f9b40bbe7f0c8905d471874fb635c2d4739e4fdcbb1b
|
| 3 |
size 2863
|