Commit
•
082acb6
1
Parent(s):
fc618da
huggingartists
Browse files- README.md +3 -3
- evaluation.txt +1 -1
- flax_model.msgpack +1 -1
- optimizer.pt +1 -1
- pytorch_model.bin +1 -1
- rng_state.pth +2 -2
- scheduler.pt +1 -1
- trainer_state.json +1013 -7
- training_args.bin +1 -1
README.md
CHANGED
|
@@ -45,15 +45,15 @@ from datasets import load_dataset
|
|
| 45 |
dataset = load_dataset("huggingartists/morgenshtern")
|
| 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 MORGENSHTERN'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/morgenshtern")
|
| 46 |
```
|
| 47 |
|
| 48 |
+
[Explore the data](https://wandb.ai/huggingartists/huggingartists/runs/cy7nthpy/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 MORGENSHTERN's lyrics.
|
| 53 |
|
| 54 |
+
Hyperparameters and metrics are recorded in the [W&B training run](https://wandb.ai/huggingartists/huggingartists/runs/2hgvstnt) for full transparency and reproducibility.
|
| 55 |
|
| 56 |
+
At the end of training, [the final model](https://wandb.ai/huggingartists/huggingartists/runs/2hgvstnt/artifacts) is logged and versioned.
|
| 57 |
|
| 58 |
## How to use
|
| 59 |
|
evaluation.txt
CHANGED
|
@@ -1 +1 @@
|
|
| 1 |
-
{"eval_loss": 1.
|
|
|
|
| 1 |
+
{"eval_loss": 1.3747466802597046, "eval_runtime": 7.026, "eval_samples_per_second": 21.207, "eval_steps_per_second": 2.704, "epoch": 13.0}
|
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:b7a0ed22dbf6a9bd7a962b4206529c5b815a5266affe74c8fbab8d7c67c678b8
|
| 3 |
size 497764120
|
optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 995604017
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ae486a04cf9c310967320cdf13a21303624d9c5b8c37c7e267c7dde9079eb63b
|
| 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 sha256:92d329a8e2a96548eb477ceb62baefe02d077414bc0c397d42965b651a2fa187
|
| 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
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f877e4bcb76e9ecc9bb481c6d8b7ee9505646a43bb655295a3a04f05d40ce503
|
| 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 sha256:1b0aee656c44c094a4ce15e92ad37f3cff7d7b2a11b7b8ea601ce5c5f3e729f2
|
| 3 |
size 623
|
trainer_state.json
CHANGED
|
@@ -1,8 +1,8 @@
|
|
| 1 |
{
|
| 2 |
-
"best_metric": 1.
|
| 3 |
-
"best_model_checkpoint": "output/morgenshtern/checkpoint-
|
| 4 |
-
"epoch":
|
| 5 |
-
"global_step":
|
| 6 |
"is_hyper_param_search": false,
|
| 7 |
"is_local_process_zero": true,
|
| 8 |
"is_world_process_zero": true,
|
|
@@ -402,11 +402,1017 @@
|
|
| 402 |
"eval_samples_per_second": 22.99,
|
| 403 |
"eval_steps_per_second": 3.013,
|
| 404 |
"step": 294
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 405 |
}
|
| 406 |
],
|
| 407 |
-
"max_steps":
|
| 408 |
-
"num_train_epochs":
|
| 409 |
-
"total_flos":
|
| 410 |
"trial_name": null,
|
| 411 |
"trial_params": null
|
| 412 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"best_metric": 1.3747466802597046,
|
| 3 |
+
"best_model_checkpoint": "output/morgenshtern/checkpoint-1078",
|
| 4 |
+
"epoch": 11.0,
|
| 5 |
+
"global_step": 1078,
|
| 6 |
"is_hyper_param_search": false,
|
| 7 |
"is_local_process_zero": true,
|
| 8 |
"is_world_process_zero": true,
|
|
|
|
| 402 |
"eval_samples_per_second": 22.99,
|
| 403 |
"eval_steps_per_second": 3.013,
|
| 404 |
"step": 294
|
| 405 |
+
},
|
| 406 |
+
{
|
| 407 |
+
"epoch": 3.01,
|
| 408 |
+
"learning_rate": 3.52455686328105e-08,
|
| 409 |
+
"loss": 1.4805,
|
| 410 |
+
"step": 295
|
| 411 |
+
},
|
| 412 |
+
{
|
| 413 |
+
"epoch": 3.06,
|
| 414 |
+
"learning_rate": 1.2650418304129032e-06,
|
| 415 |
+
"loss": 1.6327,
|
| 416 |
+
"step": 300
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 3.11,
|
| 420 |
+
"learning_rate": 4.221066247386418e-06,
|
| 421 |
+
"loss": 1.7416,
|
| 422 |
+
"step": 305
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 3.16,
|
| 426 |
+
"learning_rate": 8.827536897135471e-06,
|
| 427 |
+
"loss": 1.6078,
|
| 428 |
+
"step": 310
|
| 429 |
+
},
|
| 430 |
+
{
|
| 431 |
+
"epoch": 3.21,
|
| 432 |
+
"learning_rate": 1.496636030269314e-05,
|
| 433 |
+
"loss": 1.5727,
|
| 434 |
+
"step": 315
|
| 435 |
+
},
|
| 436 |
+
{
|
| 437 |
+
"epoch": 3.27,
|
| 438 |
+
"learning_rate": 2.2480158928073662e-05,
|
| 439 |
+
"loss": 1.6548,
|
| 440 |
+
"step": 320
|
| 441 |
+
},
|
| 442 |
+
{
|
| 443 |
+
"epoch": 3.32,
|
| 444 |
+
"learning_rate": 3.117630577695637e-05,
|
| 445 |
+
"loss": 1.596,
|
| 446 |
+
"step": 325
|
| 447 |
+
},
|
| 448 |
+
{
|
| 449 |
+
"epoch": 3.37,
|
| 450 |
+
"learning_rate": 4.08318626618038e-05,
|
| 451 |
+
"loss": 1.4367,
|
| 452 |
+
"step": 330
|
| 453 |
+
},
|
| 454 |
+
{
|
| 455 |
+
"epoch": 3.42,
|
| 456 |
+
"learning_rate": 5.119929554380771e-05,
|
| 457 |
+
"loss": 1.621,
|
| 458 |
+
"step": 335
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 3.47,
|
| 462 |
+
"learning_rate": 6.201282042273297e-05,
|
| 463 |
+
"loss": 1.7111,
|
| 464 |
+
"step": 340
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 3.52,
|
| 468 |
+
"learning_rate": 7.299521709067686e-05,
|
| 469 |
+
"loss": 1.6959,
|
| 470 |
+
"step": 345
|
| 471 |
+
},
|
| 472 |
+
{
|
| 473 |
+
"epoch": 3.57,
|
| 474 |
+
"learning_rate": 8.386493606940314e-05,
|
| 475 |
+
"loss": 1.7452,
|
| 476 |
+
"step": 350
|
| 477 |
+
},
|
| 478 |
+
{
|
| 479 |
+
"epoch": 3.62,
|
| 480 |
+
"learning_rate": 9.434331653472505e-05,
|
| 481 |
+
"loss": 1.6352,
|
| 482 |
+
"step": 355
|
| 483 |
+
},
|
| 484 |
+
{
|
| 485 |
+
"epoch": 3.67,
|
| 486 |
+
"learning_rate": 0.00010416173018610202,
|
| 487 |
+
"loss": 1.5285,
|
| 488 |
+
"step": 360
|
| 489 |
+
},
|
| 490 |
+
{
|
| 491 |
+
"epoch": 3.72,
|
| 492 |
+
"learning_rate": 0.00011306846791811431,
|
| 493 |
+
"loss": 1.7379,
|
| 494 |
+
"step": 365
|
| 495 |
+
},
|
| 496 |
+
{
|
| 497 |
+
"epoch": 3.78,
|
| 498 |
+
"learning_rate": 0.00012083519274412256,
|
| 499 |
+
"loss": 1.7212,
|
| 500 |
+
"step": 370
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 3.83,
|
| 504 |
+
"learning_rate": 0.0001272627935421667,
|
| 505 |
+
"loss": 1.6984,
|
| 506 |
+
"step": 375
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 3.88,
|
| 510 |
+
"learning_rate": 0.00013218648955393709,
|
| 511 |
+
"loss": 1.5563,
|
| 512 |
+
"step": 380
|
| 513 |
+
},
|
| 514 |
+
{
|
| 515 |
+
"epoch": 3.93,
|
| 516 |
+
"learning_rate": 0.0001354800547756731,
|
| 517 |
+
"loss": 1.6133,
|
| 518 |
+
"step": 385
|
| 519 |
+
},
|
| 520 |
+
{
|
| 521 |
+
"epoch": 3.98,
|
| 522 |
+
"learning_rate": 0.00013705905394267309,
|
| 523 |
+
"loss": 1.6524,
|
| 524 |
+
"step": 390
|
| 525 |
+
},
|
| 526 |
+
{
|
| 527 |
+
"epoch": 4.0,
|
| 528 |
+
"eval_loss": 1.4850682020187378,
|
| 529 |
+
"eval_runtime": 6.9944,
|
| 530 |
+
"eval_samples_per_second": 21.303,
|
| 531 |
+
"eval_steps_per_second": 2.716,
|
| 532 |
+
"step": 392
|
| 533 |
+
},
|
| 534 |
+
{
|
| 535 |
+
"epoch": 4.03,
|
| 536 |
+
"learning_rate": 0.000136883007148315,
|
| 537 |
+
"loss": 1.6487,
|
| 538 |
+
"step": 395
|
| 539 |
+
},
|
| 540 |
+
{
|
| 541 |
+
"epoch": 4.08,
|
| 542 |
+
"learning_rate": 0.00013495642760447742,
|
| 543 |
+
"loss": 1.5179,
|
| 544 |
+
"step": 400
|
| 545 |
+
},
|
| 546 |
+
{
|
| 547 |
+
"epoch": 4.13,
|
| 548 |
+
"learning_rate": 0.00013132870593888477,
|
| 549 |
+
"loss": 1.3698,
|
| 550 |
+
"step": 405
|
| 551 |
+
},
|
| 552 |
+
{
|
| 553 |
+
"epoch": 4.18,
|
| 554 |
+
"learning_rate": 0.00012609284399558033,
|
| 555 |
+
"loss": 1.6439,
|
| 556 |
+
"step": 410
|
| 557 |
+
},
|
| 558 |
+
{
|
| 559 |
+
"epoch": 4.23,
|
| 560 |
+
"learning_rate": 0.00011938307059936668,
|
| 561 |
+
"loss": 1.5471,
|
| 562 |
+
"step": 415
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"epoch": 4.29,
|
| 566 |
+
"learning_rate": 0.00011137140040750914,
|
| 567 |
+
"loss": 1.6368,
|
| 568 |
+
"step": 420
|
| 569 |
+
},
|
| 570 |
+
{
|
| 571 |
+
"epoch": 4.34,
|
| 572 |
+
"learning_rate": 0.00010226322406747016,
|
| 573 |
+
"loss": 1.5678,
|
| 574 |
+
"step": 425
|
| 575 |
+
},
|
| 576 |
+
{
|
| 577 |
+
"epoch": 4.39,
|
| 578 |
+
"learning_rate": 9.229204273330182e-05,
|
| 579 |
+
"loss": 1.6766,
|
| 580 |
+
"step": 430
|
| 581 |
+
},
|
| 582 |
+
{
|
| 583 |
+
"epoch": 4.44,
|
| 584 |
+
"learning_rate": 8.171348192891424e-05,
|
| 585 |
+
"loss": 1.6611,
|
| 586 |
+
"step": 435
|
| 587 |
+
},
|
| 588 |
+
{
|
| 589 |
+
"epoch": 4.49,
|
| 590 |
+
"learning_rate": 7.079873822141565e-05,
|
| 591 |
+
"loss": 1.6183,
|
| 592 |
+
"step": 440
|
| 593 |
+
},
|
| 594 |
+
{
|
| 595 |
+
"epoch": 4.54,
|
| 596 |
+
"learning_rate": 5.982762670844296e-05,
|
| 597 |
+
"loss": 1.5438,
|
| 598 |
+
"step": 445
|
| 599 |
+
},
|
| 600 |
+
{
|
| 601 |
+
"epoch": 4.59,
|
| 602 |
+
"learning_rate": 4.908140755711112e-05,
|
| 603 |
+
"loss": 1.6215,
|
| 604 |
+
"step": 450
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"epoch": 4.64,
|
| 608 |
+
"learning_rate": 3.883557549653544e-05,
|
| 609 |
+
"loss": 1.5207,
|
| 610 |
+
"step": 455
|
| 611 |
+
},
|
| 612 |
+
{
|
| 613 |
+
"epoch": 4.69,
|
| 614 |
+
"learning_rate": 2.9352797115619177e-05,
|
| 615 |
+
"loss": 1.4362,
|
| 616 |
+
"step": 460
|
| 617 |
+
},
|
| 618 |
+
{
|
| 619 |
+
"epoch": 4.74,
|
| 620 |
+
"learning_rate": 2.0876177028600835e-05,
|
| 621 |
+
"loss": 1.5917,
|
| 622 |
+
"step": 465
|
| 623 |
+
},
|
| 624 |
+
{
|
| 625 |
+
"epoch": 4.8,
|
| 626 |
+
"learning_rate": 1.3623025539858168e-05,
|
| 627 |
+
"loss": 1.4801,
|
| 628 |
+
"step": 470
|
| 629 |
+
},
|
| 630 |
+
{
|
| 631 |
+
"epoch": 4.85,
|
| 632 |
+
"learning_rate": 7.779287582812185e-06,
|
| 633 |
+
"loss": 1.4835,
|
| 634 |
+
"step": 475
|
| 635 |
+
},
|
| 636 |
+
{
|
| 637 |
+
"epoch": 4.9,
|
| 638 |
+
"learning_rate": 3.494775755068154e-06,
|
| 639 |
+
"loss": 1.5351,
|
| 640 |
+
"step": 480
|
| 641 |
+
},
|
| 642 |
+
{
|
| 643 |
+
"epoch": 4.95,
|
| 644 |
+
"learning_rate": 8.793296577687332e-07,
|
| 645 |
+
"loss": 1.4665,
|
| 646 |
+
"step": 485
|
| 647 |
+
},
|
| 648 |
+
{
|
| 649 |
+
"epoch": 5.0,
|
| 650 |
+
"learning_rate": 0.0,
|
| 651 |
+
"loss": 1.5628,
|
| 652 |
+
"step": 490
|
| 653 |
+
},
|
| 654 |
+
{
|
| 655 |
+
"epoch": 5.0,
|
| 656 |
+
"eval_loss": 1.417140007019043,
|
| 657 |
+
"eval_runtime": 6.9215,
|
| 658 |
+
"eval_samples_per_second": 21.527,
|
| 659 |
+
"eval_steps_per_second": 2.745,
|
| 660 |
+
"step": 490
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"epoch": 5.05,
|
| 664 |
+
"learning_rate": 8.79329657768718e-07,
|
| 665 |
+
"loss": 1.3775,
|
| 666 |
+
"step": 495
|
| 667 |
+
},
|
| 668 |
+
{
|
| 669 |
+
"epoch": 5.1,
|
| 670 |
+
"learning_rate": 3.4947757550681237e-06,
|
| 671 |
+
"loss": 1.5294,
|
| 672 |
+
"step": 500
|
| 673 |
+
},
|
| 674 |
+
{
|
| 675 |
+
"epoch": 5.15,
|
| 676 |
+
"learning_rate": 7.7792875828122e-06,
|
| 677 |
+
"loss": 1.436,
|
| 678 |
+
"step": 505
|
| 679 |
+
},
|
| 680 |
+
{
|
| 681 |
+
"epoch": 5.2,
|
| 682 |
+
"learning_rate": 1.3623025539858192e-05,
|
| 683 |
+
"loss": 1.5282,
|
| 684 |
+
"step": 510
|
| 685 |
+
},
|
| 686 |
+
{
|
| 687 |
+
"epoch": 5.26,
|
| 688 |
+
"learning_rate": 2.087617702860069e-05,
|
| 689 |
+
"loss": 1.5277,
|
| 690 |
+
"step": 515
|
| 691 |
+
},
|
| 692 |
+
{
|
| 693 |
+
"epoch": 5.31,
|
| 694 |
+
"learning_rate": 2.9352797115619008e-05,
|
| 695 |
+
"loss": 1.6218,
|
| 696 |
+
"step": 520
|
| 697 |
+
},
|
| 698 |
+
{
|
| 699 |
+
"epoch": 5.36,
|
| 700 |
+
"learning_rate": 3.8835575496535365e-05,
|
| 701 |
+
"loss": 1.4628,
|
| 702 |
+
"step": 525
|
| 703 |
+
},
|
| 704 |
+
{
|
| 705 |
+
"epoch": 5.41,
|
| 706 |
+
"learning_rate": 4.9081407557111025e-05,
|
| 707 |
+
"loss": 1.4702,
|
| 708 |
+
"step": 530
|
| 709 |
+
},
|
| 710 |
+
{
|
| 711 |
+
"epoch": 5.46,
|
| 712 |
+
"learning_rate": 5.9827626708442996e-05,
|
| 713 |
+
"loss": 1.3757,
|
| 714 |
+
"step": 535
|
| 715 |
+
},
|
| 716 |
+
{
|
| 717 |
+
"epoch": 5.51,
|
| 718 |
+
"learning_rate": 7.079873822141567e-05,
|
| 719 |
+
"loss": 1.3535,
|
| 720 |
+
"step": 540
|
| 721 |
+
},
|
| 722 |
+
{
|
| 723 |
+
"epoch": 5.56,
|
| 724 |
+
"learning_rate": 8.171348192891405e-05,
|
| 725 |
+
"loss": 1.4668,
|
| 726 |
+
"step": 545
|
| 727 |
+
},
|
| 728 |
+
{
|
| 729 |
+
"epoch": 5.61,
|
| 730 |
+
"learning_rate": 9.229204273330163e-05,
|
| 731 |
+
"loss": 1.4463,
|
| 732 |
+
"step": 550
|
| 733 |
+
},
|
| 734 |
+
{
|
| 735 |
+
"epoch": 5.66,
|
| 736 |
+
"learning_rate": 0.00010226322406747008,
|
| 737 |
+
"loss": 1.3922,
|
| 738 |
+
"step": 555
|
| 739 |
+
},
|
| 740 |
+
{
|
| 741 |
+
"epoch": 5.71,
|
| 742 |
+
"learning_rate": 0.00011137140040750908,
|
| 743 |
+
"loss": 1.4607,
|
| 744 |
+
"step": 560
|
| 745 |
+
},
|
| 746 |
+
{
|
| 747 |
+
"epoch": 5.77,
|
| 748 |
+
"learning_rate": 0.00011938307059936662,
|
| 749 |
+
"loss": 1.533,
|
| 750 |
+
"step": 565
|
| 751 |
+
},
|
| 752 |
+
{
|
| 753 |
+
"epoch": 5.82,
|
| 754 |
+
"learning_rate": 0.00012609284399558025,
|
| 755 |
+
"loss": 1.5392,
|
| 756 |
+
"step": 570
|
| 757 |
+
},
|
| 758 |
+
{
|
| 759 |
+
"epoch": 5.87,
|
| 760 |
+
"learning_rate": 0.00013132870593888474,
|
| 761 |
+
"loss": 1.6347,
|
| 762 |
+
"step": 575
|
| 763 |
+
},
|
| 764 |
+
{
|
| 765 |
+
"epoch": 5.92,
|
| 766 |
+
"learning_rate": 0.00013495642760447742,
|
| 767 |
+
"loss": 1.6022,
|
| 768 |
+
"step": 580
|
| 769 |
+
},
|
| 770 |
+
{
|
| 771 |
+
"epoch": 5.97,
|
| 772 |
+
"learning_rate": 0.000136883007148315,
|
| 773 |
+
"loss": 1.4927,
|
| 774 |
+
"step": 585
|
| 775 |
+
},
|
| 776 |
+
{
|
| 777 |
+
"epoch": 6.0,
|
| 778 |
+
"eval_loss": 1.4285378456115723,
|
| 779 |
+
"eval_runtime": 6.9515,
|
| 780 |
+
"eval_samples_per_second": 21.434,
|
| 781 |
+
"eval_steps_per_second": 2.733,
|
| 782 |
+
"step": 588
|
| 783 |
+
},
|
| 784 |
+
{
|
| 785 |
+
"epoch": 6.02,
|
| 786 |
+
"learning_rate": 0.00013705905394267309,
|
| 787 |
+
"loss": 1.415,
|
| 788 |
+
"step": 590
|
| 789 |
+
},
|
| 790 |
+
{
|
| 791 |
+
"epoch": 6.07,
|
| 792 |
+
"learning_rate": 0.00013548005477567314,
|
| 793 |
+
"loss": 1.4139,
|
| 794 |
+
"step": 595
|
| 795 |
+
},
|
| 796 |
+
{
|
| 797 |
+
"epoch": 6.12,
|
| 798 |
+
"learning_rate": 0.00013218648955393714,
|
| 799 |
+
"loss": 1.522,
|
| 800 |
+
"step": 600
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 6.17,
|
| 804 |
+
"learning_rate": 0.00012726279354216682,
|
| 805 |
+
"loss": 1.4313,
|
| 806 |
+
"step": 605
|
| 807 |
+
},
|
| 808 |
+
{
|
| 809 |
+
"epoch": 6.22,
|
| 810 |
+
"learning_rate": 0.0001208351927441227,
|
| 811 |
+
"loss": 1.3831,
|
| 812 |
+
"step": 610
|
| 813 |
+
},
|
| 814 |
+
{
|
| 815 |
+
"epoch": 6.28,
|
| 816 |
+
"learning_rate": 0.00011306846791811419,
|
| 817 |
+
"loss": 1.4225,
|
| 818 |
+
"step": 615
|
| 819 |
+
},
|
| 820 |
+
{
|
| 821 |
+
"epoch": 6.33,
|
| 822 |
+
"learning_rate": 0.0001041617301861021,
|
| 823 |
+
"loss": 1.5253,
|
| 824 |
+
"step": 620
|
| 825 |
+
},
|
| 826 |
+
{
|
| 827 |
+
"epoch": 6.38,
|
| 828 |
+
"learning_rate": 9.434331653472514e-05,
|
| 829 |
+
"loss": 1.3459,
|
| 830 |
+
"step": 625
|
| 831 |
+
},
|
| 832 |
+
{
|
| 833 |
+
"epoch": 6.43,
|
| 834 |
+
"learning_rate": 8.386493606940322e-05,
|
| 835 |
+
"loss": 1.3962,
|
| 836 |
+
"step": 630
|
| 837 |
+
},
|
| 838 |
+
{
|
| 839 |
+
"epoch": 6.48,
|
| 840 |
+
"learning_rate": 7.299521709067695e-05,
|
| 841 |
+
"loss": 1.3858,
|
| 842 |
+
"step": 635
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"epoch": 6.53,
|
| 846 |
+
"learning_rate": 6.201282042273305e-05,
|
| 847 |
+
"loss": 1.4634,
|
| 848 |
+
"step": 640
|
| 849 |
+
},
|
| 850 |
+
{
|
| 851 |
+
"epoch": 6.58,
|
| 852 |
+
"learning_rate": 5.11992955438078e-05,
|
| 853 |
+
"loss": 1.4289,
|
| 854 |
+
"step": 645
|
| 855 |
+
},
|
| 856 |
+
{
|
| 857 |
+
"epoch": 6.63,
|
| 858 |
+
"learning_rate": 4.0831862661803776e-05,
|
| 859 |
+
"loss": 1.4365,
|
| 860 |
+
"step": 650
|
| 861 |
+
},
|
| 862 |
+
{
|
| 863 |
+
"epoch": 6.68,
|
| 864 |
+
"learning_rate": 3.1176305776956335e-05,
|
| 865 |
+
"loss": 1.4483,
|
| 866 |
+
"step": 655
|
| 867 |
+
},
|
| 868 |
+
{
|
| 869 |
+
"epoch": 6.73,
|
| 870 |
+
"learning_rate": 2.248015892807363e-05,
|
| 871 |
+
"loss": 1.387,
|
| 872 |
+
"step": 660
|
| 873 |
+
},
|
| 874 |
+
{
|
| 875 |
+
"epoch": 6.79,
|
| 876 |
+
"learning_rate": 1.496636030269327e-05,
|
| 877 |
+
"loss": 1.4562,
|
| 878 |
+
"step": 665
|
| 879 |
+
},
|
| 880 |
+
{
|
| 881 |
+
"epoch": 6.84,
|
| 882 |
+
"learning_rate": 8.827536897135571e-06,
|
| 883 |
+
"loss": 1.4136,
|
| 884 |
+
"step": 670
|
| 885 |
+
},
|
| 886 |
+
{
|
| 887 |
+
"epoch": 6.89,
|
| 888 |
+
"learning_rate": 4.221066247386487e-06,
|
| 889 |
+
"loss": 1.3972,
|
| 890 |
+
"step": 675
|
| 891 |
+
},
|
| 892 |
+
{
|
| 893 |
+
"epoch": 6.94,
|
| 894 |
+
"learning_rate": 1.2650418304129413e-06,
|
| 895 |
+
"loss": 1.4283,
|
| 896 |
+
"step": 680
|
| 897 |
+
},
|
| 898 |
+
{
|
| 899 |
+
"epoch": 6.99,
|
| 900 |
+
"learning_rate": 3.5245568632818114e-08,
|
| 901 |
+
"loss": 1.4148,
|
| 902 |
+
"step": 685
|
| 903 |
+
},
|
| 904 |
+
{
|
| 905 |
+
"epoch": 7.0,
|
| 906 |
+
"eval_loss": 1.388408899307251,
|
| 907 |
+
"eval_runtime": 6.9772,
|
| 908 |
+
"eval_samples_per_second": 21.355,
|
| 909 |
+
"eval_steps_per_second": 2.723,
|
| 910 |
+
"step": 686
|
| 911 |
+
},
|
| 912 |
+
{
|
| 913 |
+
"epoch": 7.04,
|
| 914 |
+
"learning_rate": 5.632050517253056e-07,
|
| 915 |
+
"loss": 1.4365,
|
| 916 |
+
"step": 690
|
| 917 |
+
},
|
| 918 |
+
{
|
| 919 |
+
"epoch": 7.09,
|
| 920 |
+
"learning_rate": 2.8353852816850615e-06,
|
| 921 |
+
"loss": 1.2923,
|
| 922 |
+
"step": 695
|
| 923 |
+
},
|
| 924 |
+
{
|
| 925 |
+
"epoch": 7.14,
|
| 926 |
+
"learning_rate": 6.793535661894024e-06,
|
| 927 |
+
"loss": 1.3917,
|
| 928 |
+
"step": 700
|
| 929 |
+
},
|
| 930 |
+
{
|
| 931 |
+
"epoch": 7.19,
|
| 932 |
+
"learning_rate": 1.2336183334648805e-05,
|
| 933 |
+
"loss": 1.4588,
|
| 934 |
+
"step": 705
|
| 935 |
+
},
|
| 936 |
+
{
|
| 937 |
+
"epoch": 7.24,
|
| 938 |
+
"learning_rate": 1.932123458329587e-05,
|
| 939 |
+
"loss": 1.3238,
|
| 940 |
+
"step": 710
|
| 941 |
+
},
|
| 942 |
+
{
|
| 943 |
+
"epoch": 7.3,
|
| 944 |
+
"learning_rate": 2.7569617608302577e-05,
|
| 945 |
+
"loss": 1.3604,
|
| 946 |
+
"step": 715
|
| 947 |
+
},
|
| 948 |
+
{
|
| 949 |
+
"epoch": 7.35,
|
| 950 |
+
"learning_rate": 3.686987328947871e-05,
|
| 951 |
+
"loss": 1.3916,
|
| 952 |
+
"step": 720
|
| 953 |
+
},
|
| 954 |
+
{
|
| 955 |
+
"epoch": 7.4,
|
| 956 |
+
"learning_rate": 4.698357624357965e-05,
|
| 957 |
+
"loss": 1.3839,
|
| 958 |
+
"step": 725
|
| 959 |
+
},
|
| 960 |
+
{
|
| 961 |
+
"epoch": 7.45,
|
| 962 |
+
"learning_rate": 5.7651447200710234e-05,
|
| 963 |
+
"loss": 1.3886,
|
| 964 |
+
"step": 730
|
| 965 |
+
},
|
| 966 |
+
{
|
| 967 |
+
"epoch": 7.5,
|
| 968 |
+
"learning_rate": 6.859999999999982e-05,
|
| 969 |
+
"loss": 1.3475,
|
| 970 |
+
"step": 735
|
| 971 |
+
},
|
| 972 |
+
{
|
| 973 |
+
"epoch": 7.55,
|
| 974 |
+
"learning_rate": 7.954855279928965e-05,
|
| 975 |
+
"loss": 1.3304,
|
| 976 |
+
"step": 740
|
| 977 |
+
},
|
| 978 |
+
{
|
| 979 |
+
"epoch": 7.6,
|
| 980 |
+
"learning_rate": 9.021642375642024e-05,
|
| 981 |
+
"loss": 1.3287,
|
| 982 |
+
"step": 745
|
| 983 |
+
},
|
| 984 |
+
{
|
| 985 |
+
"epoch": 7.65,
|
| 986 |
+
"learning_rate": 0.00010033012671052118,
|
| 987 |
+
"loss": 1.4354,
|
| 988 |
+
"step": 750
|
| 989 |
+
},
|
| 990 |
+
{
|
| 991 |
+
"epoch": 7.7,
|
| 992 |
+
"learning_rate": 0.00010963038239169733,
|
| 993 |
+
"loss": 1.3597,
|
| 994 |
+
"step": 755
|
| 995 |
+
},
|
| 996 |
+
{
|
| 997 |
+
"epoch": 7.76,
|
| 998 |
+
"learning_rate": 0.00011787876541670406,
|
| 999 |
+
"loss": 1.3221,
|
| 1000 |
+
"step": 760
|
| 1001 |
+
},
|
| 1002 |
+
{
|
| 1003 |
+
"epoch": 7.81,
|
| 1004 |
+
"learning_rate": 0.00012486381666535114,
|
| 1005 |
+
"loss": 1.3458,
|
| 1006 |
+
"step": 765
|
| 1007 |
+
},
|
| 1008 |
+
{
|
| 1009 |
+
"epoch": 7.86,
|
| 1010 |
+
"learning_rate": 0.00013040646433810593,
|
| 1011 |
+
"loss": 1.4032,
|
| 1012 |
+
"step": 770
|
| 1013 |
+
},
|
| 1014 |
+
{
|
| 1015 |
+
"epoch": 7.91,
|
| 1016 |
+
"learning_rate": 0.00013436461471831492,
|
| 1017 |
+
"loss": 1.403,
|
| 1018 |
+
"step": 775
|
| 1019 |
+
},
|
| 1020 |
+
{
|
| 1021 |
+
"epoch": 7.96,
|
| 1022 |
+
"learning_rate": 0.00013663679494827467,
|
| 1023 |
+
"loss": 1.4234,
|
| 1024 |
+
"step": 780
|
| 1025 |
+
},
|
| 1026 |
+
{
|
| 1027 |
+
"epoch": 8.0,
|
| 1028 |
+
"eval_loss": 1.4104630947113037,
|
| 1029 |
+
"eval_runtime": 6.9539,
|
| 1030 |
+
"eval_samples_per_second": 21.427,
|
| 1031 |
+
"eval_steps_per_second": 2.732,
|
| 1032 |
+
"step": 784
|
| 1033 |
+
},
|
| 1034 |
+
{
|
| 1035 |
+
"epoch": 8.01,
|
| 1036 |
+
"learning_rate": 0.0001371647544313672,
|
| 1037 |
+
"loss": 1.1334,
|
| 1038 |
+
"step": 785
|
| 1039 |
+
},
|
| 1040 |
+
{
|
| 1041 |
+
"epoch": 8.06,
|
| 1042 |
+
"learning_rate": 0.0001359349581695871,
|
| 1043 |
+
"loss": 1.2244,
|
| 1044 |
+
"step": 790
|
| 1045 |
+
},
|
| 1046 |
+
{
|
| 1047 |
+
"epoch": 8.11,
|
| 1048 |
+
"learning_rate": 0.00013297893375261365,
|
| 1049 |
+
"loss": 1.361,
|
| 1050 |
+
"step": 795
|
| 1051 |
+
},
|
| 1052 |
+
{
|
| 1053 |
+
"epoch": 8.16,
|
| 1054 |
+
"learning_rate": 0.00012837246310286448,
|
| 1055 |
+
"loss": 1.3544,
|
| 1056 |
+
"step": 800
|
| 1057 |
+
},
|
| 1058 |
+
{
|
| 1059 |
+
"epoch": 8.21,
|
| 1060 |
+
"learning_rate": 0.00012223363969730697,
|
| 1061 |
+
"loss": 1.1642,
|
| 1062 |
+
"step": 805
|
| 1063 |
+
},
|
| 1064 |
+
{
|
| 1065 |
+
"epoch": 8.27,
|
| 1066 |
+
"learning_rate": 0.00011471984107192647,
|
| 1067 |
+
"loss": 1.3799,
|
| 1068 |
+
"step": 810
|
| 1069 |
+
},
|
| 1070 |
+
{
|
| 1071 |
+
"epoch": 8.32,
|
| 1072 |
+
"learning_rate": 0.00010602369422304377,
|
| 1073 |
+
"loss": 1.3705,
|
| 1074 |
+
"step": 815
|
| 1075 |
+
},
|
| 1076 |
+
{
|
| 1077 |
+
"epoch": 8.37,
|
| 1078 |
+
"learning_rate": 9.636813733819635e-05,
|
| 1079 |
+
"loss": 1.2818,
|
| 1080 |
+
"step": 820
|
| 1081 |
+
},
|
| 1082 |
+
{
|
| 1083 |
+
"epoch": 8.42,
|
| 1084 |
+
"learning_rate": 8.600070445619209e-05,
|
| 1085 |
+
"loss": 1.3989,
|
| 1086 |
+
"step": 825
|
| 1087 |
+
},
|
| 1088 |
+
{
|
| 1089 |
+
"epoch": 8.47,
|
| 1090 |
+
"learning_rate": 7.518717957726708e-05,
|
| 1091 |
+
"loss": 1.2671,
|
| 1092 |
+
"step": 830
|
| 1093 |
+
},
|
| 1094 |
+
{
|
| 1095 |
+
"epoch": 8.52,
|
| 1096 |
+
"learning_rate": 6.420478290932294e-05,
|
| 1097 |
+
"loss": 1.359,
|
| 1098 |
+
"step": 835
|
| 1099 |
+
},
|
| 1100 |
+
{
|
| 1101 |
+
"epoch": 8.57,
|
| 1102 |
+
"learning_rate": 5.33350639305969e-05,
|
| 1103 |
+
"loss": 1.3868,
|
| 1104 |
+
"step": 840
|
| 1105 |
+
},
|
| 1106 |
+
{
|
| 1107 |
+
"epoch": 8.62,
|
| 1108 |
+
"learning_rate": 4.285668346527499e-05,
|
| 1109 |
+
"loss": 1.4007,
|
| 1110 |
+
"step": 845
|
| 1111 |
+
},
|
| 1112 |
+
{
|
| 1113 |
+
"epoch": 8.67,
|
| 1114 |
+
"learning_rate": 3.3038269813898015e-05,
|
| 1115 |
+
"loss": 1.351,
|
| 1116 |
+
"step": 850
|
| 1117 |
+
},
|
| 1118 |
+
{
|
| 1119 |
+
"epoch": 8.72,
|
| 1120 |
+
"learning_rate": 2.413153208188573e-05,
|
| 1121 |
+
"loss": 1.317,
|
| 1122 |
+
"step": 855
|
| 1123 |
+
},
|
| 1124 |
+
{
|
| 1125 |
+
"epoch": 8.78,
|
| 1126 |
+
"learning_rate": 1.636480725587754e-05,
|
| 1127 |
+
"loss": 1.2524,
|
| 1128 |
+
"step": 860
|
| 1129 |
+
},
|
| 1130 |
+
{
|
| 1131 |
+
"epoch": 8.83,
|
| 1132 |
+
"learning_rate": 9.937206457833243e-06,
|
| 1133 |
+
"loss": 1.3045,
|
| 1134 |
+
"step": 865
|
| 1135 |
+
},
|
| 1136 |
+
{
|
| 1137 |
+
"epoch": 8.88,
|
| 1138 |
+
"learning_rate": 5.013510446062984e-06,
|
| 1139 |
+
"loss": 1.4154,
|
| 1140 |
+
"step": 870
|
| 1141 |
+
},
|
| 1142 |
+
{
|
| 1143 |
+
"epoch": 8.93,
|
| 1144 |
+
"learning_rate": 1.719945224326892e-06,
|
| 1145 |
+
"loss": 1.2452,
|
| 1146 |
+
"step": 875
|
| 1147 |
+
},
|
| 1148 |
+
{
|
| 1149 |
+
"epoch": 8.98,
|
| 1150 |
+
"learning_rate": 1.4094605732693502e-07,
|
| 1151 |
+
"loss": 1.3688,
|
| 1152 |
+
"step": 880
|
| 1153 |
+
},
|
| 1154 |
+
{
|
| 1155 |
+
"epoch": 9.0,
|
| 1156 |
+
"eval_loss": 1.3758981227874756,
|
| 1157 |
+
"eval_runtime": 6.9963,
|
| 1158 |
+
"eval_samples_per_second": 21.297,
|
| 1159 |
+
"eval_steps_per_second": 2.716,
|
| 1160 |
+
"step": 882
|
| 1161 |
+
},
|
| 1162 |
+
{
|
| 1163 |
+
"epoch": 9.03,
|
| 1164 |
+
"learning_rate": 3.169928516849862e-07,
|
| 1165 |
+
"loss": 1.4731,
|
| 1166 |
+
"step": 885
|
| 1167 |
+
},
|
| 1168 |
+
{
|
| 1169 |
+
"epoch": 9.08,
|
| 1170 |
+
"learning_rate": 2.2435723955225417e-06,
|
| 1171 |
+
"loss": 1.217,
|
| 1172 |
+
"step": 890
|
| 1173 |
+
},
|
| 1174 |
+
{
|
| 1175 |
+
"epoch": 9.13,
|
| 1176 |
+
"learning_rate": 5.8712940611152096e-06,
|
| 1177 |
+
"loss": 1.2492,
|
| 1178 |
+
"step": 895
|
| 1179 |
+
},
|
| 1180 |
+
{
|
| 1181 |
+
"epoch": 9.18,
|
| 1182 |
+
"learning_rate": 1.1107156004419803e-05,
|
| 1183 |
+
"loss": 1.2326,
|
| 1184 |
+
"step": 900
|
| 1185 |
+
},
|
| 1186 |
+
{
|
| 1187 |
+
"epoch": 9.23,
|
| 1188 |
+
"learning_rate": 1.7816929400633287e-05,
|
| 1189 |
+
"loss": 1.2444,
|
| 1190 |
+
"step": 905
|
| 1191 |
+
},
|
| 1192 |
+
{
|
| 1193 |
+
"epoch": 9.29,
|
| 1194 |
+
"learning_rate": 2.582859959249101e-05,
|
| 1195 |
+
"loss": 1.2759,
|
| 1196 |
+
"step": 910
|
| 1197 |
+
},
|
| 1198 |
+
{
|
| 1199 |
+
"epoch": 9.34,
|
| 1200 |
+
"learning_rate": 3.493677593252981e-05,
|
| 1201 |
+
"loss": 1.2315,
|
| 1202 |
+
"step": 915
|
| 1203 |
+
},
|
| 1204 |
+
{
|
| 1205 |
+
"epoch": 9.39,
|
| 1206 |
+
"learning_rate": 4.490795726669825e-05,
|
| 1207 |
+
"loss": 1.2353,
|
| 1208 |
+
"step": 920
|
| 1209 |
+
},
|
| 1210 |
+
{
|
| 1211 |
+
"epoch": 9.44,
|
| 1212 |
+
"learning_rate": 5.548651807108583e-05,
|
| 1213 |
+
"loss": 1.2733,
|
| 1214 |
+
"step": 925
|
| 1215 |
+
},
|
| 1216 |
+
{
|
| 1217 |
+
"epoch": 9.49,
|
| 1218 |
+
"learning_rate": 6.640126177858445e-05,
|
| 1219 |
+
"loss": 1.205,
|
| 1220 |
+
"step": 930
|
| 1221 |
+
},
|
| 1222 |
+
{
|
| 1223 |
+
"epoch": 9.54,
|
| 1224 |
+
"learning_rate": 7.737237329155688e-05,
|
| 1225 |
+
"loss": 1.2441,
|
| 1226 |
+
"step": 935
|
| 1227 |
+
},
|
| 1228 |
+
{
|
| 1229 |
+
"epoch": 9.59,
|
| 1230 |
+
"learning_rate": 8.811859244288885e-05,
|
| 1231 |
+
"loss": 1.2469,
|
| 1232 |
+
"step": 940
|
| 1233 |
+
},
|
| 1234 |
+
{
|
| 1235 |
+
"epoch": 9.64,
|
| 1236 |
+
"learning_rate": 9.83644245034643e-05,
|
| 1237 |
+
"loss": 1.2763,
|
| 1238 |
+
"step": 945
|
| 1239 |
+
},
|
| 1240 |
+
{
|
| 1241 |
+
"epoch": 9.69,
|
| 1242 |
+
"learning_rate": 0.00010784720288438088,
|
| 1243 |
+
"loss": 1.1781,
|
| 1244 |
+
"step": 950
|
| 1245 |
+
},
|
| 1246 |
+
{
|
| 1247 |
+
"epoch": 9.74,
|
| 1248 |
+
"learning_rate": 0.00011632382297139905,
|
| 1249 |
+
"loss": 1.2822,
|
| 1250 |
+
"step": 955
|
| 1251 |
+
},
|
| 1252 |
+
{
|
| 1253 |
+
"epoch": 9.8,
|
| 1254 |
+
"learning_rate": 0.00012357697446014173,
|
| 1255 |
+
"loss": 1.2663,
|
| 1256 |
+
"step": 960
|
| 1257 |
+
},
|
| 1258 |
+
{
|
| 1259 |
+
"epoch": 9.85,
|
| 1260 |
+
"learning_rate": 0.00012942071241718773,
|
| 1261 |
+
"loss": 1.3906,
|
| 1262 |
+
"step": 965
|
| 1263 |
+
},
|
| 1264 |
+
{
|
| 1265 |
+
"epoch": 9.9,
|
| 1266 |
+
"learning_rate": 0.00013370522424493184,
|
| 1267 |
+
"loss": 1.3015,
|
| 1268 |
+
"step": 970
|
| 1269 |
+
},
|
| 1270 |
+
{
|
| 1271 |
+
"epoch": 9.95,
|
| 1272 |
+
"learning_rate": 0.00013632067034223124,
|
| 1273 |
+
"loss": 1.2348,
|
| 1274 |
+
"step": 975
|
| 1275 |
+
},
|
| 1276 |
+
{
|
| 1277 |
+
"epoch": 10.0,
|
| 1278 |
+
"learning_rate": 0.0001372,
|
| 1279 |
+
"loss": 1.2793,
|
| 1280 |
+
"step": 980
|
| 1281 |
+
},
|
| 1282 |
+
{
|
| 1283 |
+
"epoch": 10.0,
|
| 1284 |
+
"eval_loss": 1.3976633548736572,
|
| 1285 |
+
"eval_runtime": 7.0041,
|
| 1286 |
+
"eval_samples_per_second": 21.273,
|
| 1287 |
+
"eval_steps_per_second": 2.713,
|
| 1288 |
+
"step": 980
|
| 1289 |
+
},
|
| 1290 |
+
{
|
| 1291 |
+
"epoch": 10.05,
|
| 1292 |
+
"learning_rate": 0.00013632067034223126,
|
| 1293 |
+
"loss": 1.2203,
|
| 1294 |
+
"step": 985
|
| 1295 |
+
},
|
| 1296 |
+
{
|
| 1297 |
+
"epoch": 10.1,
|
| 1298 |
+
"learning_rate": 0.0001337052242449319,
|
| 1299 |
+
"loss": 1.3016,
|
| 1300 |
+
"step": 990
|
| 1301 |
+
},
|
| 1302 |
+
{
|
| 1303 |
+
"epoch": 10.15,
|
| 1304 |
+
"learning_rate": 0.00012942071241718781,
|
| 1305 |
+
"loss": 1.2078,
|
| 1306 |
+
"step": 995
|
| 1307 |
+
},
|
| 1308 |
+
{
|
| 1309 |
+
"epoch": 10.2,
|
| 1310 |
+
"learning_rate": 0.00012357697446014183,
|
| 1311 |
+
"loss": 1.2408,
|
| 1312 |
+
"step": 1000
|
| 1313 |
+
},
|
| 1314 |
+
{
|
| 1315 |
+
"epoch": 10.26,
|
| 1316 |
+
"learning_rate": 0.00011632382297139899,
|
| 1317 |
+
"loss": 1.2247,
|
| 1318 |
+
"step": 1005
|
| 1319 |
+
},
|
| 1320 |
+
{
|
| 1321 |
+
"epoch": 10.31,
|
| 1322 |
+
"learning_rate": 0.00010784720288438083,
|
| 1323 |
+
"loss": 1.0939,
|
| 1324 |
+
"step": 1010
|
| 1325 |
+
},
|
| 1326 |
+
{
|
| 1327 |
+
"epoch": 10.36,
|
| 1328 |
+
"learning_rate": 9.836442450346467e-05,
|
| 1329 |
+
"loss": 1.1944,
|
| 1330 |
+
"step": 1015
|
| 1331 |
+
},
|
| 1332 |
+
{
|
| 1333 |
+
"epoch": 10.41,
|
| 1334 |
+
"learning_rate": 8.811859244288877e-05,
|
| 1335 |
+
"loss": 1.2171,
|
| 1336 |
+
"step": 1020
|
| 1337 |
+
},
|
| 1338 |
+
{
|
| 1339 |
+
"epoch": 10.46,
|
| 1340 |
+
"learning_rate": 7.737237329155728e-05,
|
| 1341 |
+
"loss": 1.2546,
|
| 1342 |
+
"step": 1025
|
| 1343 |
+
},
|
| 1344 |
+
{
|
| 1345 |
+
"epoch": 10.51,
|
| 1346 |
+
"learning_rate": 6.640126177858486e-05,
|
| 1347 |
+
"loss": 1.1607,
|
| 1348 |
+
"step": 1030
|
| 1349 |
+
},
|
| 1350 |
+
{
|
| 1351 |
+
"epoch": 10.56,
|
| 1352 |
+
"learning_rate": 5.5486518071086e-05,
|
| 1353 |
+
"loss": 1.2547,
|
| 1354 |
+
"step": 1035
|
| 1355 |
+
},
|
| 1356 |
+
{
|
| 1357 |
+
"epoch": 10.61,
|
| 1358 |
+
"learning_rate": 4.4907957266698644e-05,
|
| 1359 |
+
"loss": 1.2295,
|
| 1360 |
+
"step": 1040
|
| 1361 |
+
},
|
| 1362 |
+
{
|
| 1363 |
+
"epoch": 10.66,
|
| 1364 |
+
"learning_rate": 3.493677593252996e-05,
|
| 1365 |
+
"loss": 1.2566,
|
| 1366 |
+
"step": 1045
|
| 1367 |
+
},
|
| 1368 |
+
{
|
| 1369 |
+
"epoch": 10.71,
|
| 1370 |
+
"learning_rate": 2.5828599592491143e-05,
|
| 1371 |
+
"loss": 1.2949,
|
| 1372 |
+
"step": 1050
|
| 1373 |
+
},
|
| 1374 |
+
{
|
| 1375 |
+
"epoch": 10.77,
|
| 1376 |
+
"learning_rate": 1.7816929400633402e-05,
|
| 1377 |
+
"loss": 1.2886,
|
| 1378 |
+
"step": 1055
|
| 1379 |
+
},
|
| 1380 |
+
{
|
| 1381 |
+
"epoch": 10.82,
|
| 1382 |
+
"learning_rate": 1.1107156004419895e-05,
|
| 1383 |
+
"loss": 1.1858,
|
| 1384 |
+
"step": 1060
|
| 1385 |
+
},
|
| 1386 |
+
{
|
| 1387 |
+
"epoch": 10.87,
|
| 1388 |
+
"learning_rate": 5.871294061115278e-06,
|
| 1389 |
+
"loss": 1.2322,
|
| 1390 |
+
"step": 1065
|
| 1391 |
+
},
|
| 1392 |
+
{
|
| 1393 |
+
"epoch": 10.92,
|
| 1394 |
+
"learning_rate": 2.2435723955225265e-06,
|
| 1395 |
+
"loss": 1.0434,
|
| 1396 |
+
"step": 1070
|
| 1397 |
+
},
|
| 1398 |
+
{
|
| 1399 |
+
"epoch": 10.97,
|
| 1400 |
+
"learning_rate": 3.169928516850014e-07,
|
| 1401 |
+
"loss": 1.2505,
|
| 1402 |
+
"step": 1075
|
| 1403 |
+
},
|
| 1404 |
+
{
|
| 1405 |
+
"epoch": 11.0,
|
| 1406 |
+
"eval_loss": 1.3747466802597046,
|
| 1407 |
+
"eval_runtime": 7.0127,
|
| 1408 |
+
"eval_samples_per_second": 21.247,
|
| 1409 |
+
"eval_steps_per_second": 2.709,
|
| 1410 |
+
"step": 1078
|
| 1411 |
}
|
| 1412 |
],
|
| 1413 |
+
"max_steps": 1274,
|
| 1414 |
+
"num_train_epochs": 13,
|
| 1415 |
+
"total_flos": 1118199250944000.0,
|
| 1416 |
"trial_name": null,
|
| 1417 |
"trial_params": null
|
| 1418 |
}
|
training_args.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 2671
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:a2f1578ed0263fa3cde0930e9e148d11708f37717f22b869ef9dc3fb789f5d60
|
| 3 |
size 2671
|