Commit
•
0e73214
1
Parent(s):
41eda7c
huggingartists
Browse files- README.md +4 -4
- config.json +1 -1
- evaluation.txt +1 -1
- flax_model.msgpack +1 -1
- optimizer.pt +1 -1
- pytorch_model.bin +1 -1
- rng_state.pth +1 -1
- scheduler.pt +1 -1
- trainer_state.json +1207 -7
- training_args.bin +1 -1
README.md
CHANGED
|
@@ -14,7 +14,7 @@ widget:
|
|
| 14 |
<div class="inline-flex flex-col" style="line-height: 1.5;">
|
| 15 |
<div class="flex">
|
| 16 |
<div
|
| 17 |
-
style="display:DISPLAY_1; margin-left: auto; margin-right: auto; width: 92px; height:92px; border-radius: 50%; background-size: cover; background-image: url('https://images.genius.com/
|
| 18 |
</div>
|
| 19 |
</div>
|
| 20 |
<div style="text-align: center; margin-top: 3px; font-size: 16px; font-weight: 800">🤖 HuggingArtists Model 🤖</div>
|
|
@@ -45,15 +45,15 @@ from datasets import load_dataset
|
|
| 45 |
dataset = load_dataset("huggingartists/drake")
|
| 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 Drake'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 |
|
|
|
|
| 14 |
<div class="inline-flex flex-col" style="line-height: 1.5;">
|
| 15 |
<div class="flex">
|
| 16 |
<div
|
| 17 |
+
style="display:DISPLAY_1; margin-left: auto; margin-right: auto; width: 92px; height:92px; border-radius: 50%; background-size: cover; background-image: url('https://images.genius.com/631b206379b60df5e1da90e84d35fdbe.1000x1000x1.png')">
|
| 18 |
</div>
|
| 19 |
</div>
|
| 20 |
<div style="text-align: center; margin-top: 3px; font-size: 16px; font-weight: 800">🤖 HuggingArtists Model 🤖</div>
|
|
|
|
| 45 |
dataset = load_dataset("huggingartists/drake")
|
| 46 |
```
|
| 47 |
|
| 48 |
+
[Explore the data](https://wandb.ai/huggingartists/huggingartists/runs/1aksz894/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 Drake's lyrics.
|
| 53 |
|
| 54 |
+
Hyperparameters and metrics are recorded in the [W&B training run](https://wandb.ai/huggingartists/huggingartists/runs/3rx3sf99) for full transparency and reproducibility.
|
| 55 |
|
| 56 |
+
At the end of training, [the final model](https://wandb.ai/huggingartists/huggingartists/runs/3rx3sf99/artifacts) is logged and versioned.
|
| 57 |
|
| 58 |
## How to use
|
| 59 |
|
config.json
CHANGED
|
@@ -37,7 +37,7 @@
|
|
| 37 |
}
|
| 38 |
},
|
| 39 |
"torch_dtype": "float32",
|
| 40 |
-
"transformers_version": "4.20.
|
| 41 |
"use_cache": true,
|
| 42 |
"vocab_size": 50257
|
| 43 |
}
|
|
|
|
| 37 |
}
|
| 38 |
},
|
| 39 |
"torch_dtype": "float32",
|
| 40 |
+
"transformers_version": "4.20.1",
|
| 41 |
"use_cache": true,
|
| 42 |
"vocab_size": 50257
|
| 43 |
}
|
evaluation.txt
CHANGED
|
@@ -1 +1 @@
|
|
| 1 |
-
{"eval_loss":
|
|
|
|
| 1 |
+
{"eval_loss": 2.985200881958008, "eval_runtime": 3.2437, "eval_samples_per_second": 143.353, "eval_steps_per_second": 18.189, "epoch": 12.0}
|
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 sha256:57a832a766a6062bfcfcaea51111be32f951d3ccc6a07c44a1c70945134909a5
|
| 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:fc861593254336ef6f7349effb09e8d7559406594cf64d7e16e4803c44118418
|
| 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 510396521
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:a0b2b81b7c939397be91b9a719857e4d86a375656547375710d1aeb2afa5ee16
|
| 3 |
size 510396521
|
rng_state.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 14567
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f37b1fe21c3237da257fc4dad062a40dd556ee8402c38632ba55aeb299ee646a
|
| 3 |
size 14567
|
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:fcfeaed1730d77e101fd5ed4b37bb1cb254cb585fa58085f4a7ca52a63e8aab8
|
| 3 |
size 623
|
trainer_state.json
CHANGED
|
@@ -1,8 +1,8 @@
|
|
| 1 |
{
|
| 2 |
-
"best_metric":
|
| 3 |
-
"best_model_checkpoint": "output/drake/checkpoint-
|
| 4 |
-
"epoch":
|
| 5 |
-
"global_step":
|
| 6 |
"is_hyper_param_search": false,
|
| 7 |
"is_local_process_zero": true,
|
| 8 |
"is_world_process_zero": true,
|
|
@@ -806,11 +806,1211 @@
|
|
| 806 |
"eval_samples_per_second": 45.611,
|
| 807 |
"eval_steps_per_second": 5.777,
|
| 808 |
"step": 660
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 809 |
}
|
| 810 |
],
|
| 811 |
-
"max_steps":
|
| 812 |
-
"num_train_epochs":
|
| 813 |
-
"total_flos":
|
| 814 |
"trial_name": null,
|
| 815 |
"trial_params": null
|
| 816 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"best_metric": 2.985200881958008,
|
| 3 |
+
"best_model_checkpoint": "output/drake/checkpoint-1640",
|
| 4 |
+
"epoch": 5.0,
|
| 5 |
+
"global_step": 1640,
|
| 6 |
"is_hyper_param_search": false,
|
| 7 |
"is_local_process_zero": true,
|
| 8 |
"is_world_process_zero": true,
|
|
|
|
| 806 |
"eval_samples_per_second": 45.611,
|
| 807 |
"eval_steps_per_second": 5.777,
|
| 808 |
"step": 660
|
| 809 |
+
},
|
| 810 |
+
{
|
| 811 |
+
"epoch": 2.03,
|
| 812 |
+
"learning_rate": 0.0001369452804670866,
|
| 813 |
+
"loss": 2.8801,
|
| 814 |
+
"step": 665
|
| 815 |
+
},
|
| 816 |
+
{
|
| 817 |
+
"epoch": 2.04,
|
| 818 |
+
"learning_rate": 0.00013658418337916058,
|
| 819 |
+
"loss": 3.0411,
|
| 820 |
+
"step": 670
|
| 821 |
+
},
|
| 822 |
+
{
|
| 823 |
+
"epoch": 2.06,
|
| 824 |
+
"learning_rate": 0.00013606719676147636,
|
| 825 |
+
"loss": 3.1263,
|
| 826 |
+
"step": 675
|
| 827 |
+
},
|
| 828 |
+
{
|
| 829 |
+
"epoch": 2.07,
|
| 830 |
+
"learning_rate": 0.00013539550607801564,
|
| 831 |
+
"loss": 3.2868,
|
| 832 |
+
"step": 680
|
| 833 |
+
},
|
| 834 |
+
{
|
| 835 |
+
"epoch": 2.09,
|
| 836 |
+
"learning_rate": 0.00013457065153327282,
|
| 837 |
+
"loss": 3.0776,
|
| 838 |
+
"step": 685
|
| 839 |
+
},
|
| 840 |
+
{
|
| 841 |
+
"epoch": 2.1,
|
| 842 |
+
"learning_rate": 0.00013359452454052534,
|
| 843 |
+
"loss": 3.1467,
|
| 844 |
+
"step": 690
|
| 845 |
+
},
|
| 846 |
+
{
|
| 847 |
+
"epoch": 2.12,
|
| 848 |
+
"learning_rate": 0.00013246936338477295,
|
| 849 |
+
"loss": 2.9318,
|
| 850 |
+
"step": 695
|
| 851 |
+
},
|
| 852 |
+
{
|
| 853 |
+
"epoch": 2.13,
|
| 854 |
+
"learning_rate": 0.00013119774809029134,
|
| 855 |
+
"loss": 3.2317,
|
| 856 |
+
"step": 700
|
| 857 |
+
},
|
| 858 |
+
{
|
| 859 |
+
"epoch": 2.15,
|
| 860 |
+
"learning_rate": 0.00012978259450456783,
|
| 861 |
+
"loss": 2.9684,
|
| 862 |
+
"step": 705
|
| 863 |
+
},
|
| 864 |
+
{
|
| 865 |
+
"epoch": 2.16,
|
| 866 |
+
"learning_rate": 0.0001282271476121864,
|
| 867 |
+
"loss": 3.2684,
|
| 868 |
+
"step": 710
|
| 869 |
+
},
|
| 870 |
+
{
|
| 871 |
+
"epoch": 2.18,
|
| 872 |
+
"learning_rate": 0.0001265349740939919,
|
| 873 |
+
"loss": 3.1091,
|
| 874 |
+
"step": 715
|
| 875 |
+
},
|
| 876 |
+
{
|
| 877 |
+
"epoch": 2.2,
|
| 878 |
+
"learning_rate": 0.00012470995414859683,
|
| 879 |
+
"loss": 3.0846,
|
| 880 |
+
"step": 720
|
| 881 |
+
},
|
| 882 |
+
{
|
| 883 |
+
"epoch": 2.21,
|
| 884 |
+
"learning_rate": 0.00012275627259498339,
|
| 885 |
+
"loss": 3.2676,
|
| 886 |
+
"step": 725
|
| 887 |
+
},
|
| 888 |
+
{
|
| 889 |
+
"epoch": 2.23,
|
| 890 |
+
"learning_rate": 0.00012067840927660262,
|
| 891 |
+
"loss": 3.1157,
|
| 892 |
+
"step": 730
|
| 893 |
+
},
|
| 894 |
+
{
|
| 895 |
+
"epoch": 2.24,
|
| 896 |
+
"learning_rate": 0.00011848112878897519,
|
| 897 |
+
"loss": 3.2578,
|
| 898 |
+
"step": 735
|
| 899 |
+
},
|
| 900 |
+
{
|
| 901 |
+
"epoch": 2.26,
|
| 902 |
+
"learning_rate": 0.00011616946955434783,
|
| 903 |
+
"loss": 3.1545,
|
| 904 |
+
"step": 740
|
| 905 |
+
},
|
| 906 |
+
{
|
| 907 |
+
"epoch": 2.27,
|
| 908 |
+
"learning_rate": 0.00011374873226845785,
|
| 909 |
+
"loss": 3.0167,
|
| 910 |
+
"step": 745
|
| 911 |
+
},
|
| 912 |
+
{
|
| 913 |
+
"epoch": 2.29,
|
| 914 |
+
"learning_rate": 0.00011122446774589795,
|
| 915 |
+
"loss": 3.3236,
|
| 916 |
+
"step": 750
|
| 917 |
+
},
|
| 918 |
+
{
|
| 919 |
+
"epoch": 2.3,
|
| 920 |
+
"learning_rate": 0.00010860246419195128,
|
| 921 |
+
"loss": 3.0855,
|
| 922 |
+
"step": 755
|
| 923 |
+
},
|
| 924 |
+
{
|
| 925 |
+
"epoch": 2.32,
|
| 926 |
+
"learning_rate": 0.00010588873393008382,
|
| 927 |
+
"loss": 3.1622,
|
| 928 |
+
"step": 760
|
| 929 |
+
},
|
| 930 |
+
{
|
| 931 |
+
"epoch": 2.33,
|
| 932 |
+
"learning_rate": 0.00010308949961552774,
|
| 933 |
+
"loss": 3.1558,
|
| 934 |
+
"step": 765
|
| 935 |
+
},
|
| 936 |
+
{
|
| 937 |
+
"epoch": 2.35,
|
| 938 |
+
"learning_rate": 0.0001002111799665686,
|
| 939 |
+
"loss": 3.1256,
|
| 940 |
+
"step": 770
|
| 941 |
+
},
|
| 942 |
+
{
|
| 943 |
+
"epoch": 2.36,
|
| 944 |
+
"learning_rate": 9.726037504625446e-05,
|
| 945 |
+
"loss": 3.0624,
|
| 946 |
+
"step": 775
|
| 947 |
+
},
|
| 948 |
+
{
|
| 949 |
+
"epoch": 2.38,
|
| 950 |
+
"learning_rate": 9.424385112827677e-05,
|
| 951 |
+
"loss": 3.127,
|
| 952 |
+
"step": 780
|
| 953 |
+
},
|
| 954 |
+
{
|
| 955 |
+
"epoch": 2.39,
|
| 956 |
+
"learning_rate": 9.1168525181726e-05,
|
| 957 |
+
"loss": 3.2459,
|
| 958 |
+
"step": 785
|
| 959 |
+
},
|
| 960 |
+
{
|
| 961 |
+
"epoch": 2.41,
|
| 962 |
+
"learning_rate": 8.80414490102987e-05,
|
| 963 |
+
"loss": 3.2356,
|
| 964 |
+
"step": 790
|
| 965 |
+
},
|
| 966 |
+
{
|
| 967 |
+
"epoch": 2.42,
|
| 968 |
+
"learning_rate": 8.486979308232529e-05,
|
| 969 |
+
"loss": 3.1039,
|
| 970 |
+
"step": 795
|
| 971 |
+
},
|
| 972 |
+
{
|
| 973 |
+
"epoch": 2.44,
|
| 974 |
+
"learning_rate": 8.16608300886963e-05,
|
| 975 |
+
"loss": 3.1498,
|
| 976 |
+
"step": 800
|
| 977 |
+
},
|
| 978 |
+
{
|
| 979 |
+
"epoch": 2.45,
|
| 980 |
+
"learning_rate": 7.842191826639015e-05,
|
| 981 |
+
"loss": 3.2395,
|
| 982 |
+
"step": 805
|
| 983 |
+
},
|
| 984 |
+
{
|
| 985 |
+
"epoch": 2.47,
|
| 986 |
+
"learning_rate": 7.516048452584106e-05,
|
| 987 |
+
"loss": 3.0791,
|
| 988 |
+
"step": 810
|
| 989 |
+
},
|
| 990 |
+
{
|
| 991 |
+
"epoch": 2.48,
|
| 992 |
+
"learning_rate": 7.18840074208375e-05,
|
| 993 |
+
"loss": 2.9722,
|
| 994 |
+
"step": 815
|
| 995 |
+
},
|
| 996 |
+
{
|
| 997 |
+
"epoch": 2.5,
|
| 998 |
+
"learning_rate": 6.860000000000001e-05,
|
| 999 |
+
"loss": 3.0485,
|
| 1000 |
+
"step": 820
|
| 1001 |
+
},
|
| 1002 |
+
{
|
| 1003 |
+
"epoch": 2.52,
|
| 1004 |
+
"learning_rate": 6.531599257916255e-05,
|
| 1005 |
+
"loss": 3.26,
|
| 1006 |
+
"step": 825
|
| 1007 |
+
},
|
| 1008 |
+
{
|
| 1009 |
+
"epoch": 2.53,
|
| 1010 |
+
"learning_rate": 6.203951547415898e-05,
|
| 1011 |
+
"loss": 3.0569,
|
| 1012 |
+
"step": 830
|
| 1013 |
+
},
|
| 1014 |
+
{
|
| 1015 |
+
"epoch": 2.55,
|
| 1016 |
+
"learning_rate": 5.877808173360989e-05,
|
| 1017 |
+
"loss": 3.0239,
|
| 1018 |
+
"step": 835
|
| 1019 |
+
},
|
| 1020 |
+
{
|
| 1021 |
+
"epoch": 2.56,
|
| 1022 |
+
"learning_rate": 5.553916991130374e-05,
|
| 1023 |
+
"loss": 3.1443,
|
| 1024 |
+
"step": 840
|
| 1025 |
+
},
|
| 1026 |
+
{
|
| 1027 |
+
"epoch": 2.58,
|
| 1028 |
+
"learning_rate": 5.2330206917674755e-05,
|
| 1029 |
+
"loss": 3.1212,
|
| 1030 |
+
"step": 845
|
| 1031 |
+
},
|
| 1032 |
+
{
|
| 1033 |
+
"epoch": 2.59,
|
| 1034 |
+
"learning_rate": 4.915855098970134e-05,
|
| 1035 |
+
"loss": 2.8549,
|
| 1036 |
+
"step": 850
|
| 1037 |
+
},
|
| 1038 |
+
{
|
| 1039 |
+
"epoch": 2.61,
|
| 1040 |
+
"learning_rate": 4.603147481827398e-05,
|
| 1041 |
+
"loss": 2.9992,
|
| 1042 |
+
"step": 855
|
| 1043 |
+
},
|
| 1044 |
+
{
|
| 1045 |
+
"epoch": 2.62,
|
| 1046 |
+
"learning_rate": 4.2956148871723224e-05,
|
| 1047 |
+
"loss": 2.8552,
|
| 1048 |
+
"step": 860
|
| 1049 |
+
},
|
| 1050 |
+
{
|
| 1051 |
+
"epoch": 2.64,
|
| 1052 |
+
"learning_rate": 3.9939624953745524e-05,
|
| 1053 |
+
"loss": 3.0432,
|
| 1054 |
+
"step": 865
|
| 1055 |
+
},
|
| 1056 |
+
{
|
| 1057 |
+
"epoch": 2.65,
|
| 1058 |
+
"learning_rate": 3.698882003343138e-05,
|
| 1059 |
+
"loss": 3.028,
|
| 1060 |
+
"step": 870
|
| 1061 |
+
},
|
| 1062 |
+
{
|
| 1063 |
+
"epoch": 2.67,
|
| 1064 |
+
"learning_rate": 3.411050038447229e-05,
|
| 1065 |
+
"loss": 3.1615,
|
| 1066 |
+
"step": 875
|
| 1067 |
+
},
|
| 1068 |
+
{
|
| 1069 |
+
"epoch": 2.68,
|
| 1070 |
+
"learning_rate": 3.131126606991618e-05,
|
| 1071 |
+
"loss": 3.1442,
|
| 1072 |
+
"step": 880
|
| 1073 |
+
},
|
| 1074 |
+
{
|
| 1075 |
+
"epoch": 2.7,
|
| 1076 |
+
"learning_rate": 2.8597535808048707e-05,
|
| 1077 |
+
"loss": 3.0388,
|
| 1078 |
+
"step": 885
|
| 1079 |
+
},
|
| 1080 |
+
{
|
| 1081 |
+
"epoch": 2.71,
|
| 1082 |
+
"learning_rate": 2.5975532254102027e-05,
|
| 1083 |
+
"loss": 2.9324,
|
| 1084 |
+
"step": 890
|
| 1085 |
+
},
|
| 1086 |
+
{
|
| 1087 |
+
"epoch": 2.73,
|
| 1088 |
+
"learning_rate": 2.3451267731542143e-05,
|
| 1089 |
+
"loss": 2.9219,
|
| 1090 |
+
"step": 895
|
| 1091 |
+
},
|
| 1092 |
+
{
|
| 1093 |
+
"epoch": 2.74,
|
| 1094 |
+
"learning_rate": 2.1030530445652233e-05,
|
| 1095 |
+
"loss": 2.8783,
|
| 1096 |
+
"step": 900
|
| 1097 |
+
},
|
| 1098 |
+
{
|
| 1099 |
+
"epoch": 2.76,
|
| 1100 |
+
"learning_rate": 1.871887121102488e-05,
|
| 1101 |
+
"loss": 3.2507,
|
| 1102 |
+
"step": 905
|
| 1103 |
+
},
|
| 1104 |
+
{
|
| 1105 |
+
"epoch": 2.77,
|
| 1106 |
+
"learning_rate": 1.6521590723397454e-05,
|
| 1107 |
+
"loss": 2.969,
|
| 1108 |
+
"step": 910
|
| 1109 |
+
},
|
| 1110 |
+
{
|
| 1111 |
+
"epoch": 2.79,
|
| 1112 |
+
"learning_rate": 1.4443727405016684e-05,
|
| 1113 |
+
"loss": 2.9703,
|
| 1114 |
+
"step": 915
|
| 1115 |
+
},
|
| 1116 |
+
{
|
| 1117 |
+
"epoch": 2.8,
|
| 1118 |
+
"learning_rate": 1.249004585140324e-05,
|
| 1119 |
+
"loss": 3.105,
|
| 1120 |
+
"step": 920
|
| 1121 |
+
},
|
| 1122 |
+
{
|
| 1123 |
+
"epoch": 2.82,
|
| 1124 |
+
"learning_rate": 1.0665025906008154e-05,
|
| 1125 |
+
"loss": 3.1641,
|
| 1126 |
+
"step": 925
|
| 1127 |
+
},
|
| 1128 |
+
{
|
| 1129 |
+
"epoch": 2.84,
|
| 1130 |
+
"learning_rate": 8.97285238781366e-06,
|
| 1131 |
+
"loss": 2.9782,
|
| 1132 |
+
"step": 930
|
| 1133 |
+
},
|
| 1134 |
+
{
|
| 1135 |
+
"epoch": 2.85,
|
| 1136 |
+
"learning_rate": 7.417405495432199e-06,
|
| 1137 |
+
"loss": 3.0392,
|
| 1138 |
+
"step": 935
|
| 1139 |
+
},
|
| 1140 |
+
{
|
| 1141 |
+
"epoch": 2.87,
|
| 1142 |
+
"learning_rate": 6.002251909708709e-06,
|
| 1143 |
+
"loss": 2.962,
|
| 1144 |
+
"step": 940
|
| 1145 |
+
},
|
| 1146 |
+
{
|
| 1147 |
+
"epoch": 2.88,
|
| 1148 |
+
"learning_rate": 4.730636615227076e-06,
|
| 1149 |
+
"loss": 3.3658,
|
| 1150 |
+
"step": 945
|
| 1151 |
+
},
|
| 1152 |
+
{
|
| 1153 |
+
"epoch": 2.9,
|
| 1154 |
+
"learning_rate": 3.605475459474696e-06,
|
| 1155 |
+
"loss": 2.9995,
|
| 1156 |
+
"step": 950
|
| 1157 |
+
},
|
| 1158 |
+
{
|
| 1159 |
+
"epoch": 2.91,
|
| 1160 |
+
"learning_rate": 2.6293484667271903e-06,
|
| 1161 |
+
"loss": 3.1411,
|
| 1162 |
+
"step": 955
|
| 1163 |
+
},
|
| 1164 |
+
{
|
| 1165 |
+
"epoch": 2.93,
|
| 1166 |
+
"learning_rate": 1.8044939219843706e-06,
|
| 1167 |
+
"loss": 3.2262,
|
| 1168 |
+
"step": 960
|
| 1169 |
+
},
|
| 1170 |
+
{
|
| 1171 |
+
"epoch": 2.94,
|
| 1172 |
+
"learning_rate": 1.132803238523659e-06,
|
| 1173 |
+
"loss": 2.995,
|
| 1174 |
+
"step": 965
|
| 1175 |
+
},
|
| 1176 |
+
{
|
| 1177 |
+
"epoch": 2.96,
|
| 1178 |
+
"learning_rate": 6.158166208394294e-07,
|
| 1179 |
+
"loss": 3.1868,
|
| 1180 |
+
"step": 970
|
| 1181 |
+
},
|
| 1182 |
+
{
|
| 1183 |
+
"epoch": 2.97,
|
| 1184 |
+
"learning_rate": 2.547195329134365e-07,
|
| 1185 |
+
"loss": 3.0908,
|
| 1186 |
+
"step": 975
|
| 1187 |
+
},
|
| 1188 |
+
{
|
| 1189 |
+
"epoch": 2.99,
|
| 1190 |
+
"learning_rate": 5.033997991443411e-08,
|
| 1191 |
+
"loss": 3.0958,
|
| 1192 |
+
"step": 980
|
| 1193 |
+
},
|
| 1194 |
+
{
|
| 1195 |
+
"epoch": 3.0,
|
| 1196 |
+
"eval_loss": 3.012192964553833,
|
| 1197 |
+
"eval_runtime": 3.2423,
|
| 1198 |
+
"eval_samples_per_second": 143.415,
|
| 1199 |
+
"eval_steps_per_second": 18.197,
|
| 1200 |
+
"step": 984
|
| 1201 |
+
},
|
| 1202 |
+
{
|
| 1203 |
+
"epoch": 3.0,
|
| 1204 |
+
"learning_rate": 3.1466095605587442e-09,
|
| 1205 |
+
"loss": 3.2577,
|
| 1206 |
+
"step": 985
|
| 1207 |
+
},
|
| 1208 |
+
{
|
| 1209 |
+
"epoch": 3.02,
|
| 1210 |
+
"learning_rate": 1.1324763749765112e-07,
|
| 1211 |
+
"loss": 2.9737,
|
| 1212 |
+
"step": 990
|
| 1213 |
+
},
|
| 1214 |
+
{
|
| 1215 |
+
"epoch": 3.03,
|
| 1216 |
+
"learning_rate": 3.803905991579816e-07,
|
| 1217 |
+
"loss": 2.7524,
|
| 1218 |
+
"step": 995
|
| 1219 |
+
},
|
| 1220 |
+
{
|
| 1221 |
+
"epoch": 3.05,
|
| 1222 |
+
"learning_rate": 8.03962928668091e-07,
|
| 1223 |
+
"loss": 2.9894,
|
| 1224 |
+
"step": 1000
|
| 1225 |
+
},
|
| 1226 |
+
{
|
| 1227 |
+
"epoch": 3.06,
|
| 1228 |
+
"learning_rate": 1.382993363478614e-06,
|
| 1229 |
+
"loss": 2.8537,
|
| 1230 |
+
"step": 1005
|
| 1231 |
+
},
|
| 1232 |
+
{
|
| 1233 |
+
"epoch": 3.08,
|
| 1234 |
+
"learning_rate": 2.116154171494932e-06,
|
| 1235 |
+
"loss": 2.9448,
|
| 1236 |
+
"step": 1010
|
| 1237 |
+
},
|
| 1238 |
+
{
|
| 1239 |
+
"epoch": 3.09,
|
| 1240 |
+
"learning_rate": 3.001764195602037e-06,
|
| 1241 |
+
"loss": 2.8933,
|
| 1242 |
+
"step": 1015
|
| 1243 |
+
},
|
| 1244 |
+
{
|
| 1245 |
+
"epoch": 3.11,
|
| 1246 |
+
"learning_rate": 4.037792708602087e-06,
|
| 1247 |
+
"loss": 2.7567,
|
| 1248 |
+
"step": 1020
|
| 1249 |
+
},
|
| 1250 |
+
{
|
| 1251 |
+
"epoch": 3.12,
|
| 1252 |
+
"learning_rate": 5.22186406972573e-06,
|
| 1253 |
+
"loss": 2.8581,
|
| 1254 |
+
"step": 1025
|
| 1255 |
+
},
|
| 1256 |
+
{
|
| 1257 |
+
"epoch": 3.14,
|
| 1258 |
+
"learning_rate": 6.55126317203911e-06,
|
| 1259 |
+
"loss": 3.0563,
|
| 1260 |
+
"step": 1030
|
| 1261 |
+
},
|
| 1262 |
+
{
|
| 1263 |
+
"epoch": 3.16,
|
| 1264 |
+
"learning_rate": 8.022941668255849e-06,
|
| 1265 |
+
"loss": 3.1461,
|
| 1266 |
+
"step": 1035
|
| 1267 |
+
},
|
| 1268 |
+
{
|
| 1269 |
+
"epoch": 3.17,
|
| 1270 |
+
"learning_rate": 9.633524960678029e-06,
|
| 1271 |
+
"loss": 2.7869,
|
| 1272 |
+
"step": 1040
|
| 1273 |
+
},
|
| 1274 |
+
{
|
| 1275 |
+
"epoch": 3.19,
|
| 1276 |
+
"learning_rate": 1.1379319939237678e-05,
|
| 1277 |
+
"loss": 2.9909,
|
| 1278 |
+
"step": 1045
|
| 1279 |
+
},
|
| 1280 |
+
{
|
| 1281 |
+
"epoch": 3.2,
|
| 1282 |
+
"learning_rate": 1.3256323449896148e-05,
|
| 1283 |
+
"loss": 2.7969,
|
| 1284 |
+
"step": 1050
|
| 1285 |
+
},
|
| 1286 |
+
{
|
| 1287 |
+
"epoch": 3.22,
|
| 1288 |
+
"learning_rate": 1.526023147398194e-05,
|
| 1289 |
+
"loss": 3.0613,
|
| 1290 |
+
"step": 1055
|
| 1291 |
+
},
|
| 1292 |
+
{
|
| 1293 |
+
"epoch": 3.23,
|
| 1294 |
+
"learning_rate": 1.7386448997419335e-05,
|
| 1295 |
+
"loss": 2.9952,
|
| 1296 |
+
"step": 1060
|
| 1297 |
+
},
|
| 1298 |
+
{
|
| 1299 |
+
"epoch": 3.25,
|
| 1300 |
+
"learning_rate": 1.9630100547217162e-05,
|
| 1301 |
+
"loss": 2.8488,
|
| 1302 |
+
"step": 1065
|
| 1303 |
+
},
|
| 1304 |
+
{
|
| 1305 |
+
"epoch": 3.26,
|
| 1306 |
+
"learning_rate": 2.198604137105723e-05,
|
| 1307 |
+
"loss": 2.9286,
|
| 1308 |
+
"step": 1070
|
| 1309 |
+
},
|
| 1310 |
+
{
|
| 1311 |
+
"epoch": 3.28,
|
| 1312 |
+
"learning_rate": 2.444886923434745e-05,
|
| 1313 |
+
"loss": 2.9456,
|
| 1314 |
+
"step": 1075
|
| 1315 |
+
},
|
| 1316 |
+
{
|
| 1317 |
+
"epoch": 3.29,
|
| 1318 |
+
"learning_rate": 2.7012936807688628e-05,
|
| 1319 |
+
"loss": 2.9093,
|
| 1320 |
+
"step": 1080
|
| 1321 |
+
},
|
| 1322 |
+
{
|
| 1323 |
+
"epoch": 3.31,
|
| 1324 |
+
"learning_rate": 2.9672364616349706e-05,
|
| 1325 |
+
"loss": 2.6638,
|
| 1326 |
+
"step": 1085
|
| 1327 |
+
},
|
| 1328 |
+
{
|
| 1329 |
+
"epoch": 3.32,
|
| 1330 |
+
"learning_rate": 3.242105452205932e-05,
|
| 1331 |
+
"loss": 2.8781,
|
| 1332 |
+
"step": 1090
|
| 1333 |
+
},
|
| 1334 |
+
{
|
| 1335 |
+
"epoch": 3.34,
|
| 1336 |
+
"learning_rate": 3.5252703706197704e-05,
|
| 1337 |
+
"loss": 2.8149,
|
| 1338 |
+
"step": 1095
|
| 1339 |
+
},
|
| 1340 |
+
{
|
| 1341 |
+
"epoch": 3.35,
|
| 1342 |
+
"learning_rate": 3.816081912232647e-05,
|
| 1343 |
+
"loss": 2.9128,
|
| 1344 |
+
"step": 1100
|
| 1345 |
+
},
|
| 1346 |
+
{
|
| 1347 |
+
"epoch": 3.37,
|
| 1348 |
+
"learning_rate": 4.113873238491593e-05,
|
| 1349 |
+
"loss": 2.9003,
|
| 1350 |
+
"step": 1105
|
| 1351 |
+
},
|
| 1352 |
+
{
|
| 1353 |
+
"epoch": 3.38,
|
| 1354 |
+
"learning_rate": 4.4179615060128864e-05,
|
| 1355 |
+
"loss": 3.0516,
|
| 1356 |
+
"step": 1110
|
| 1357 |
+
},
|
| 1358 |
+
{
|
| 1359 |
+
"epoch": 3.4,
|
| 1360 |
+
"learning_rate": 4.727649432360055e-05,
|
| 1361 |
+
"loss": 2.9956,
|
| 1362 |
+
"step": 1115
|
| 1363 |
+
},
|
| 1364 |
+
{
|
| 1365 |
+
"epoch": 3.41,
|
| 1366 |
+
"learning_rate": 5.042226894930894e-05,
|
| 1367 |
+
"loss": 2.9518,
|
| 1368 |
+
"step": 1120
|
| 1369 |
+
},
|
| 1370 |
+
{
|
| 1371 |
+
"epoch": 3.43,
|
| 1372 |
+
"learning_rate": 5.3609725592873736e-05,
|
| 1373 |
+
"loss": 3.0925,
|
| 1374 |
+
"step": 1125
|
| 1375 |
+
},
|
| 1376 |
+
{
|
| 1377 |
+
"epoch": 3.45,
|
| 1378 |
+
"learning_rate": 5.68315553319459e-05,
|
| 1379 |
+
"loss": 2.9208,
|
| 1380 |
+
"step": 1130
|
| 1381 |
+
},
|
| 1382 |
+
{
|
| 1383 |
+
"epoch": 3.46,
|
| 1384 |
+
"learning_rate": 6.00803704257593e-05,
|
| 1385 |
+
"loss": 2.9595,
|
| 1386 |
+
"step": 1135
|
| 1387 |
+
},
|
| 1388 |
+
{
|
| 1389 |
+
"epoch": 3.48,
|
| 1390 |
+
"learning_rate": 6.334872125541636e-05,
|
| 1391 |
+
"loss": 2.7912,
|
| 1392 |
+
"step": 1140
|
| 1393 |
+
},
|
| 1394 |
+
{
|
| 1395 |
+
"epoch": 3.49,
|
| 1396 |
+
"learning_rate": 6.662911340606064e-05,
|
| 1397 |
+
"loss": 3.1123,
|
| 1398 |
+
"step": 1145
|
| 1399 |
+
},
|
| 1400 |
+
{
|
| 1401 |
+
"epoch": 3.51,
|
| 1402 |
+
"learning_rate": 6.991402485176834e-05,
|
| 1403 |
+
"loss": 3.0401,
|
| 1404 |
+
"step": 1150
|
| 1405 |
+
},
|
| 1406 |
+
{
|
| 1407 |
+
"epoch": 3.52,
|
| 1408 |
+
"learning_rate": 7.319592320375272e-05,
|
| 1409 |
+
"loss": 3.007,
|
| 1410 |
+
"step": 1155
|
| 1411 |
+
},
|
| 1412 |
+
{
|
| 1413 |
+
"epoch": 3.54,
|
| 1414 |
+
"learning_rate": 7.646728298233026e-05,
|
| 1415 |
+
"loss": 3.0709,
|
| 1416 |
+
"step": 1160
|
| 1417 |
+
},
|
| 1418 |
+
{
|
| 1419 |
+
"epoch": 3.55,
|
| 1420 |
+
"learning_rate": 7.97206028730452e-05,
|
| 1421 |
+
"loss": 2.8136,
|
| 1422 |
+
"step": 1165
|
| 1423 |
+
},
|
| 1424 |
+
{
|
| 1425 |
+
"epoch": 3.57,
|
| 1426 |
+
"learning_rate": 8.294842292738145e-05,
|
| 1427 |
+
"loss": 3.0615,
|
| 1428 |
+
"step": 1170
|
| 1429 |
+
},
|
| 1430 |
+
{
|
| 1431 |
+
"epoch": 3.58,
|
| 1432 |
+
"learning_rate": 8.6143341668622e-05,
|
| 1433 |
+
"loss": 2.8914,
|
| 1434 |
+
"step": 1175
|
| 1435 |
+
},
|
| 1436 |
+
{
|
| 1437 |
+
"epoch": 3.6,
|
| 1438 |
+
"learning_rate": 8.92980330636302e-05,
|
| 1439 |
+
"loss": 3.0546,
|
| 1440 |
+
"step": 1180
|
| 1441 |
+
},
|
| 1442 |
+
{
|
| 1443 |
+
"epoch": 3.61,
|
| 1444 |
+
"learning_rate": 9.240526332163816e-05,
|
| 1445 |
+
"loss": 2.7764,
|
| 1446 |
+
"step": 1185
|
| 1447 |
+
},
|
| 1448 |
+
{
|
| 1449 |
+
"epoch": 3.63,
|
| 1450 |
+
"learning_rate": 9.545790748151966e-05,
|
| 1451 |
+
"loss": 2.9967,
|
| 1452 |
+
"step": 1190
|
| 1453 |
+
},
|
| 1454 |
+
{
|
| 1455 |
+
"epoch": 3.64,
|
| 1456 |
+
"learning_rate": 9.844896574951464e-05,
|
| 1457 |
+
"loss": 2.8943,
|
| 1458 |
+
"step": 1195
|
| 1459 |
+
},
|
| 1460 |
+
{
|
| 1461 |
+
"epoch": 3.66,
|
| 1462 |
+
"learning_rate": 0.00010137157954994128,
|
| 1463 |
+
"loss": 2.7756,
|
| 1464 |
+
"step": 1200
|
| 1465 |
+
},
|
| 1466 |
+
{
|
| 1467 |
+
"epoch": 3.67,
|
| 1468 |
+
"learning_rate": 0.00010421904725209109,
|
| 1469 |
+
"loss": 2.6902,
|
| 1470 |
+
"step": 1205
|
| 1471 |
+
},
|
| 1472 |
+
{
|
| 1473 |
+
"epoch": 3.69,
|
| 1474 |
+
"learning_rate": 0.0001069848395372458,
|
| 1475 |
+
"loss": 3.0339,
|
| 1476 |
+
"step": 1210
|
| 1477 |
+
},
|
| 1478 |
+
{
|
| 1479 |
+
"epoch": 3.7,
|
| 1480 |
+
"learning_rate": 0.00010966261437057745,
|
| 1481 |
+
"loss": 2.9684,
|
| 1482 |
+
"step": 1215
|
| 1483 |
+
},
|
| 1484 |
+
{
|
| 1485 |
+
"epoch": 3.72,
|
| 1486 |
+
"learning_rate": 0.00011224623154360197,
|
| 1487 |
+
"loss": 2.993,
|
| 1488 |
+
"step": 1220
|
| 1489 |
+
},
|
| 1490 |
+
{
|
| 1491 |
+
"epoch": 3.73,
|
| 1492 |
+
"learning_rate": 0.0001147297667538396,
|
| 1493 |
+
"loss": 2.8925,
|
| 1494 |
+
"step": 1225
|
| 1495 |
+
},
|
| 1496 |
+
{
|
| 1497 |
+
"epoch": 3.75,
|
| 1498 |
+
"learning_rate": 0.0001171075251893971,
|
| 1499 |
+
"loss": 3.0326,
|
| 1500 |
+
"step": 1230
|
| 1501 |
+
},
|
| 1502 |
+
{
|
| 1503 |
+
"epoch": 3.77,
|
| 1504 |
+
"learning_rate": 0.00011937405458732195,
|
| 1505 |
+
"loss": 3.1186,
|
| 1506 |
+
"step": 1235
|
| 1507 |
+
},
|
| 1508 |
+
{
|
| 1509 |
+
"epoch": 3.78,
|
| 1510 |
+
"learning_rate": 0.00012152415773578527,
|
| 1511 |
+
"loss": 2.9294,
|
| 1512 |
+
"step": 1240
|
| 1513 |
+
},
|
| 1514 |
+
{
|
| 1515 |
+
"epoch": 3.8,
|
| 1516 |
+
"learning_rate": 0.00012355290439142633,
|
| 1517 |
+
"loss": 3.0367,
|
| 1518 |
+
"step": 1245
|
| 1519 |
+
},
|
| 1520 |
+
{
|
| 1521 |
+
"epoch": 3.81,
|
| 1522 |
+
"learning_rate": 0.0001254556425845306,
|
| 1523 |
+
"loss": 2.6842,
|
| 1524 |
+
"step": 1250
|
| 1525 |
+
},
|
| 1526 |
+
{
|
| 1527 |
+
"epoch": 3.83,
|
| 1528 |
+
"learning_rate": 0.00012722800928611938,
|
| 1529 |
+
"loss": 2.807,
|
| 1530 |
+
"step": 1255
|
| 1531 |
+
},
|
| 1532 |
+
{
|
| 1533 |
+
"epoch": 3.84,
|
| 1534 |
+
"learning_rate": 0.0001288659404124906,
|
| 1535 |
+
"loss": 2.874,
|
| 1536 |
+
"step": 1260
|
| 1537 |
+
},
|
| 1538 |
+
{
|
| 1539 |
+
"epoch": 3.86,
|
| 1540 |
+
"learning_rate": 0.00013036568014427024,
|
| 1541 |
+
"loss": 2.9999,
|
| 1542 |
+
"step": 1265
|
| 1543 |
+
},
|
| 1544 |
+
{
|
| 1545 |
+
"epoch": 3.87,
|
| 1546 |
+
"learning_rate": 0.00013172378953860524,
|
| 1547 |
+
"loss": 3.0116,
|
| 1548 |
+
"step": 1270
|
| 1549 |
+
},
|
| 1550 |
+
{
|
| 1551 |
+
"epoch": 3.89,
|
| 1552 |
+
"learning_rate": 0.00013293715441475062,
|
| 1553 |
+
"loss": 3.1574,
|
| 1554 |
+
"step": 1275
|
| 1555 |
+
},
|
| 1556 |
+
{
|
| 1557 |
+
"epoch": 3.9,
|
| 1558 |
+
"learning_rate": 0.00013400299249496822,
|
| 1559 |
+
"loss": 2.6983,
|
| 1560 |
+
"step": 1280
|
| 1561 |
+
},
|
| 1562 |
+
{
|
| 1563 |
+
"epoch": 3.92,
|
| 1564 |
+
"learning_rate": 0.0001349188597843629,
|
| 1565 |
+
"loss": 2.8115,
|
| 1566 |
+
"step": 1285
|
| 1567 |
+
},
|
| 1568 |
+
{
|
| 1569 |
+
"epoch": 3.93,
|
| 1570 |
+
"learning_rate": 0.00013568265617502779,
|
| 1571 |
+
"loss": 2.7446,
|
| 1572 |
+
"step": 1290
|
| 1573 |
+
},
|
| 1574 |
+
{
|
| 1575 |
+
"epoch": 3.95,
|
| 1576 |
+
"learning_rate": 0.0001362926302616468,
|
| 1577 |
+
"loss": 2.8728,
|
| 1578 |
+
"step": 1295
|
| 1579 |
+
},
|
| 1580 |
+
{
|
| 1581 |
+
"epoch": 3.96,
|
| 1582 |
+
"learning_rate": 0.0001367473833575137,
|
| 1583 |
+
"loss": 3.134,
|
| 1584 |
+
"step": 1300
|
| 1585 |
+
},
|
| 1586 |
+
{
|
| 1587 |
+
"epoch": 3.98,
|
| 1588 |
+
"learning_rate": 0.0001370458727017576,
|
| 1589 |
+
"loss": 2.8436,
|
| 1590 |
+
"step": 1305
|
| 1591 |
+
},
|
| 1592 |
+
{
|
| 1593 |
+
"epoch": 3.99,
|
| 1594 |
+
"learning_rate": 0.0001371874138504211,
|
| 1595 |
+
"loss": 2.901,
|
| 1596 |
+
"step": 1310
|
| 1597 |
+
},
|
| 1598 |
+
{
|
| 1599 |
+
"epoch": 4.0,
|
| 1600 |
+
"eval_loss": 3.0180249214172363,
|
| 1601 |
+
"eval_runtime": 3.2449,
|
| 1602 |
+
"eval_samples_per_second": 143.301,
|
| 1603 |
+
"eval_steps_per_second": 18.182,
|
| 1604 |
+
"step": 1312
|
| 1605 |
+
},
|
| 1606 |
+
{
|
| 1607 |
+
"epoch": 4.01,
|
| 1608 |
+
"learning_rate": 0.0001371716822459084,
|
| 1609 |
+
"loss": 2.7412,
|
| 1610 |
+
"step": 1315
|
| 1611 |
+
},
|
| 1612 |
+
{
|
| 1613 |
+
"epoch": 4.02,
|
| 1614 |
+
"learning_rate": 0.00013699871396120457,
|
| 1615 |
+
"loss": 2.5891,
|
| 1616 |
+
"step": 1320
|
| 1617 |
+
},
|
| 1618 |
+
{
|
| 1619 |
+
"epoch": 4.04,
|
| 1620 |
+
"learning_rate": 0.00013666890561715917,
|
| 1621 |
+
"loss": 3.1228,
|
| 1622 |
+
"step": 1325
|
| 1623 |
+
},
|
| 1624 |
+
{
|
| 1625 |
+
"epoch": 4.05,
|
| 1626 |
+
"learning_rate": 0.00013618301347302405,
|
| 1627 |
+
"loss": 2.9228,
|
| 1628 |
+
"step": 1330
|
| 1629 |
+
},
|
| 1630 |
+
{
|
| 1631 |
+
"epoch": 4.07,
|
| 1632 |
+
"learning_rate": 0.000135542151692331,
|
| 1633 |
+
"loss": 2.6346,
|
| 1634 |
+
"step": 1335
|
| 1635 |
+
},
|
| 1636 |
+
{
|
| 1637 |
+
"epoch": 4.09,
|
| 1638 |
+
"learning_rate": 0.0001347477897880855,
|
| 1639 |
+
"loss": 2.7065,
|
| 1640 |
+
"step": 1340
|
| 1641 |
+
},
|
| 1642 |
+
{
|
| 1643 |
+
"epoch": 4.1,
|
| 1644 |
+
"learning_rate": 0.00013380174925313403,
|
| 1645 |
+
"loss": 2.7945,
|
| 1646 |
+
"step": 1345
|
| 1647 |
+
},
|
| 1648 |
+
{
|
| 1649 |
+
"epoch": 4.12,
|
| 1650 |
+
"learning_rate": 0.00013270619938343336,
|
| 1651 |
+
"loss": 2.5638,
|
| 1652 |
+
"step": 1350
|
| 1653 |
+
},
|
| 1654 |
+
{
|
| 1655 |
+
"epoch": 4.13,
|
| 1656 |
+
"learning_rate": 0.00013146365230379744,
|
| 1657 |
+
"loss": 2.8108,
|
| 1658 |
+
"step": 1355
|
| 1659 |
+
},
|
| 1660 |
+
{
|
| 1661 |
+
"epoch": 4.15,
|
| 1662 |
+
"learning_rate": 0.0001300769572075284,
|
| 1663 |
+
"loss": 2.8114,
|
| 1664 |
+
"step": 1360
|
| 1665 |
+
},
|
| 1666 |
+
{
|
| 1667 |
+
"epoch": 4.16,
|
| 1668 |
+
"learning_rate": 0.00012854929382314167,
|
| 1669 |
+
"loss": 2.7208,
|
| 1670 |
+
"step": 1365
|
| 1671 |
+
},
|
| 1672 |
+
{
|
| 1673 |
+
"epoch": 4.18,
|
| 1674 |
+
"learning_rate": 0.0001268841651231636,
|
| 1675 |
+
"loss": 2.8124,
|
| 1676 |
+
"step": 1370
|
| 1677 |
+
},
|
| 1678 |
+
{
|
| 1679 |
+
"epoch": 4.19,
|
| 1680 |
+
"learning_rate": 0.00012508538929172376,
|
| 1681 |
+
"loss": 2.9135,
|
| 1682 |
+
"step": 1375
|
| 1683 |
+
},
|
| 1684 |
+
{
|
| 1685 |
+
"epoch": 4.21,
|
| 1686 |
+
"learning_rate": 0.0001231570909693571,
|
| 1687 |
+
"loss": 2.5881,
|
| 1688 |
+
"step": 1380
|
| 1689 |
+
},
|
| 1690 |
+
{
|
| 1691 |
+
"epoch": 4.22,
|
| 1692 |
+
"learning_rate": 0.00012110369179509431,
|
| 1693 |
+
"loss": 2.6689,
|
| 1694 |
+
"step": 1385
|
| 1695 |
+
},
|
| 1696 |
+
{
|
| 1697 |
+
"epoch": 4.24,
|
| 1698 |
+
"learning_rate": 0.00011892990026752702,
|
| 1699 |
+
"loss": 2.8313,
|
| 1700 |
+
"step": 1390
|
| 1701 |
+
},
|
| 1702 |
+
{
|
| 1703 |
+
"epoch": 4.25,
|
| 1704 |
+
"learning_rate": 0.0001166407009480955,
|
| 1705 |
+
"loss": 2.9475,
|
| 1706 |
+
"step": 1395
|
| 1707 |
+
},
|
| 1708 |
+
{
|
| 1709 |
+
"epoch": 4.27,
|
| 1710 |
+
"learning_rate": 0.0001142413430313578,
|
| 1711 |
+
"loss": 3.0502,
|
| 1712 |
+
"step": 1400
|
| 1713 |
+
},
|
| 1714 |
+
{
|
| 1715 |
+
"epoch": 4.28,
|
| 1716 |
+
"learning_rate": 0.00011173732830844787,
|
| 1717 |
+
"loss": 2.6531,
|
| 1718 |
+
"step": 1405
|
| 1719 |
+
},
|
| 1720 |
+
{
|
| 1721 |
+
"epoch": 4.3,
|
| 1722 |
+
"learning_rate": 0.00010913439855132291,
|
| 1723 |
+
"loss": 2.9528,
|
| 1724 |
+
"step": 1410
|
| 1725 |
+
},
|
| 1726 |
+
{
|
| 1727 |
+
"epoch": 4.31,
|
| 1728 |
+
"learning_rate": 0.00010643852234672852,
|
| 1729 |
+
"loss": 2.7701,
|
| 1730 |
+
"step": 1415
|
| 1731 |
+
},
|
| 1732 |
+
{
|
| 1733 |
+
"epoch": 4.33,
|
| 1734 |
+
"learning_rate": 0.00010365588141007188,
|
| 1735 |
+
"loss": 2.6251,
|
| 1736 |
+
"step": 1420
|
| 1737 |
+
},
|
| 1738 |
+
{
|
| 1739 |
+
"epoch": 4.34,
|
| 1740 |
+
"learning_rate": 0.00010079285641058387,
|
| 1741 |
+
"loss": 2.9979,
|
| 1742 |
+
"step": 1425
|
| 1743 |
+
},
|
| 1744 |
+
{
|
| 1745 |
+
"epoch": 4.36,
|
| 1746 |
+
"learning_rate": 9.785601234027734e-05,
|
| 1747 |
+
"loss": 2.807,
|
| 1748 |
+
"step": 1430
|
| 1749 |
+
},
|
| 1750 |
+
{
|
| 1751 |
+
"epoch": 4.38,
|
| 1752 |
+
"learning_rate": 9.485208346024522e-05,
|
| 1753 |
+
"loss": 2.8977,
|
| 1754 |
+
"step": 1435
|
| 1755 |
+
},
|
| 1756 |
+
{
|
| 1757 |
+
"epoch": 4.39,
|
| 1758 |
+
"learning_rate": 9.178795785882326e-05,
|
| 1759 |
+
"loss": 2.8219,
|
| 1760 |
+
"step": 1440
|
| 1761 |
+
},
|
| 1762 |
+
{
|
| 1763 |
+
"epoch": 4.41,
|
| 1764 |
+
"learning_rate": 8.867066165702077e-05,
|
| 1765 |
+
"loss": 2.6527,
|
| 1766 |
+
"step": 1445
|
| 1767 |
+
},
|
| 1768 |
+
{
|
| 1769 |
+
"epoch": 4.42,
|
| 1770 |
+
"learning_rate": 8.550734289743871e-05,
|
| 1771 |
+
"loss": 2.8599,
|
| 1772 |
+
"step": 1450
|
| 1773 |
+
},
|
| 1774 |
+
{
|
| 1775 |
+
"epoch": 4.44,
|
| 1776 |
+
"learning_rate": 8.230525515362062e-05,
|
| 1777 |
+
"loss": 2.9224,
|
| 1778 |
+
"step": 1455
|
| 1779 |
+
},
|
| 1780 |
+
{
|
| 1781 |
+
"epoch": 4.45,
|
| 1782 |
+
"learning_rate": 7.907174089741485e-05,
|
| 1783 |
+
"loss": 2.6402,
|
| 1784 |
+
"step": 1460
|
| 1785 |
+
},
|
| 1786 |
+
{
|
| 1787 |
+
"epoch": 4.47,
|
| 1788 |
+
"learning_rate": 7.581421466249422e-05,
|
| 1789 |
+
"loss": 2.658,
|
| 1790 |
+
"step": 1465
|
| 1791 |
+
},
|
| 1792 |
+
{
|
| 1793 |
+
"epoch": 4.48,
|
| 1794 |
+
"learning_rate": 7.254014604263352e-05,
|
| 1795 |
+
"loss": 2.6781,
|
| 1796 |
+
"step": 1470
|
| 1797 |
+
},
|
| 1798 |
+
{
|
| 1799 |
+
"epoch": 4.5,
|
| 1800 |
+
"learning_rate": 6.92570425637328e-05,
|
| 1801 |
+
"loss": 2.8161,
|
| 1802 |
+
"step": 1475
|
| 1803 |
+
},
|
| 1804 |
+
{
|
| 1805 |
+
"epoch": 4.51,
|
| 1806 |
+
"learning_rate": 6.597243246886372e-05,
|
| 1807 |
+
"loss": 2.7003,
|
| 1808 |
+
"step": 1480
|
| 1809 |
+
},
|
| 1810 |
+
{
|
| 1811 |
+
"epoch": 4.53,
|
| 1812 |
+
"learning_rate": 6.269384745580722e-05,
|
| 1813 |
+
"loss": 2.6335,
|
| 1814 |
+
"step": 1485
|
| 1815 |
+
},
|
| 1816 |
+
{
|
| 1817 |
+
"epoch": 4.54,
|
| 1818 |
+
"learning_rate": 5.942880540667308e-05,
|
| 1819 |
+
"loss": 2.9172,
|
| 1820 |
+
"step": 1490
|
| 1821 |
+
},
|
| 1822 |
+
{
|
| 1823 |
+
"epoch": 4.56,
|
| 1824 |
+
"learning_rate": 5.618479314919675e-05,
|
| 1825 |
+
"loss": 2.6901,
|
| 1826 |
+
"step": 1495
|
| 1827 |
+
},
|
| 1828 |
+
{
|
| 1829 |
+
"epoch": 4.57,
|
| 1830 |
+
"learning_rate": 5.2969249289244765e-05,
|
| 1831 |
+
"loss": 3.0955,
|
| 1832 |
+
"step": 1500
|
| 1833 |
+
},
|
| 1834 |
+
{
|
| 1835 |
+
"epoch": 4.59,
|
| 1836 |
+
"learning_rate": 4.978954715389626e-05,
|
| 1837 |
+
"loss": 2.8052,
|
| 1838 |
+
"step": 1505
|
| 1839 |
+
},
|
| 1840 |
+
{
|
| 1841 |
+
"epoch": 4.6,
|
| 1842 |
+
"learning_rate": 4.6652977884208274e-05,
|
| 1843 |
+
"loss": 2.5181,
|
| 1844 |
+
"step": 1510
|
| 1845 |
+
},
|
| 1846 |
+
{
|
| 1847 |
+
"epoch": 4.62,
|
| 1848 |
+
"learning_rate": 4.356673371643594e-05,
|
| 1849 |
+
"loss": 2.8346,
|
| 1850 |
+
"step": 1515
|
| 1851 |
+
},
|
| 1852 |
+
{
|
| 1853 |
+
"epoch": 4.63,
|
| 1854 |
+
"learning_rate": 4.0537891490046174e-05,
|
| 1855 |
+
"loss": 2.9267,
|
| 1856 |
+
"step": 1520
|
| 1857 |
+
},
|
| 1858 |
+
{
|
| 1859 |
+
"epoch": 4.65,
|
| 1860 |
+
"learning_rate": 3.757339642033588e-05,
|
| 1861 |
+
"loss": 2.6995,
|
| 1862 |
+
"step": 1525
|
| 1863 |
+
},
|
| 1864 |
+
{
|
| 1865 |
+
"epoch": 4.66,
|
| 1866 |
+
"learning_rate": 3.468004617287139e-05,
|
| 1867 |
+
"loss": 2.8038,
|
| 1868 |
+
"step": 1530
|
| 1869 |
+
},
|
| 1870 |
+
{
|
| 1871 |
+
"epoch": 4.68,
|
| 1872 |
+
"learning_rate": 3.1864475276260315e-05,
|
| 1873 |
+
"loss": 2.7572,
|
| 1874 |
+
"step": 1535
|
| 1875 |
+
},
|
| 1876 |
+
{
|
| 1877 |
+
"epoch": 4.7,
|
| 1878 |
+
"learning_rate": 2.9133139909002586e-05,
|
| 1879 |
+
"loss": 2.8888,
|
| 1880 |
+
"step": 1540
|
| 1881 |
+
},
|
| 1882 |
+
{
|
| 1883 |
+
"epoch": 4.71,
|
| 1884 |
+
"learning_rate": 2.6492303095302796e-05,
|
| 1885 |
+
"loss": 2.5902,
|
| 1886 |
+
"step": 1545
|
| 1887 |
+
},
|
| 1888 |
+
{
|
| 1889 |
+
"epoch": 4.73,
|
| 1890 |
+
"learning_rate": 2.3948020343789084e-05,
|
| 1891 |
+
"loss": 2.6841,
|
| 1892 |
+
"step": 1550
|
| 1893 |
+
},
|
| 1894 |
+
{
|
| 1895 |
+
"epoch": 4.74,
|
| 1896 |
+
"learning_rate": 2.1506125762073123e-05,
|
| 1897 |
+
"loss": 2.6346,
|
| 1898 |
+
"step": 1555
|
| 1899 |
+
},
|
| 1900 |
+
{
|
| 1901 |
+
"epoch": 4.76,
|
| 1902 |
+
"learning_rate": 1.917221867898604e-05,
|
| 1903 |
+
"loss": 2.574,
|
| 1904 |
+
"step": 1560
|
| 1905 |
+
},
|
| 1906 |
+
{
|
| 1907 |
+
"epoch": 4.77,
|
| 1908 |
+
"learning_rate": 1.695165080516989e-05,
|
| 1909 |
+
"loss": 2.742,
|
| 1910 |
+
"step": 1565
|
| 1911 |
+
},
|
| 1912 |
+
{
|
| 1913 |
+
"epoch": 4.79,
|
| 1914 |
+
"learning_rate": 1.4849513961463578e-05,
|
| 1915 |
+
"loss": 2.8388,
|
| 1916 |
+
"step": 1570
|
| 1917 |
+
},
|
| 1918 |
+
{
|
| 1919 |
+
"epoch": 4.8,
|
| 1920 |
+
"learning_rate": 1.2870628403223036e-05,
|
| 1921 |
+
"loss": 2.6697,
|
| 1922 |
+
"step": 1575
|
| 1923 |
+
},
|
| 1924 |
+
{
|
| 1925 |
+
"epoch": 4.82,
|
| 1926 |
+
"learning_rate": 1.101953176734824e-05,
|
| 1927 |
+
"loss": 2.7003,
|
| 1928 |
+
"step": 1580
|
| 1929 |
+
},
|
| 1930 |
+
{
|
| 1931 |
+
"epoch": 4.83,
|
| 1932 |
+
"learning_rate": 9.30046866736211e-06,
|
| 1933 |
+
"loss": 3.0252,
|
| 1934 |
+
"step": 1585
|
| 1935 |
+
},
|
| 1936 |
+
{
|
| 1937 |
+
"epoch": 4.85,
|
| 1938 |
+
"learning_rate": 7.717380960398831e-06,
|
| 1939 |
+
"loss": 2.8193,
|
| 1940 |
+
"step": 1590
|
| 1941 |
+
},
|
| 1942 |
+
{
|
| 1943 |
+
"epoch": 4.86,
|
| 1944 |
+
"learning_rate": 6.273898708422032e-06,
|
| 1945 |
+
"loss": 2.895,
|
| 1946 |
+
"step": 1595
|
| 1947 |
+
},
|
| 1948 |
+
{
|
| 1949 |
+
"epoch": 4.88,
|
| 1950 |
+
"learning_rate": 4.9733318543963394e-06,
|
| 1951 |
+
"loss": 2.7687,
|
| 1952 |
+
"step": 1600
|
| 1953 |
+
},
|
| 1954 |
+
{
|
| 1955 |
+
"epoch": 4.89,
|
| 1956 |
+
"learning_rate": 3.818662632500395e-06,
|
| 1957 |
+
"loss": 2.7358,
|
| 1958 |
+
"step": 1605
|
| 1959 |
+
},
|
| 1960 |
+
{
|
| 1961 |
+
"epoch": 4.91,
|
| 1962 |
+
"learning_rate": 2.812538729785672e-06,
|
| 1963 |
+
"loss": 2.7698,
|
| 1964 |
+
"step": 1610
|
| 1965 |
+
},
|
| 1966 |
+
{
|
| 1967 |
+
"epoch": 4.92,
|
| 1968 |
+
"learning_rate": 1.9572672149594433e-06,
|
| 1969 |
+
"loss": 2.7906,
|
| 1970 |
+
"step": 1615
|
| 1971 |
+
},
|
| 1972 |
+
{
|
| 1973 |
+
"epoch": 4.94,
|
| 1974 |
+
"learning_rate": 1.2548092482157443e-06,
|
| 1975 |
+
"loss": 2.6523,
|
| 1976 |
+
"step": 1620
|
| 1977 |
+
},
|
| 1978 |
+
{
|
| 1979 |
+
"epoch": 4.95,
|
| 1980 |
+
"learning_rate": 7.067755842429393e-07,
|
| 1981 |
+
"loss": 2.7697,
|
| 1982 |
+
"step": 1625
|
| 1983 |
+
},
|
| 1984 |
+
{
|
| 1985 |
+
"epoch": 4.97,
|
| 1986 |
+
"learning_rate": 3.144228787205802e-07,
|
| 1987 |
+
"loss": 2.7642,
|
| 1988 |
+
"step": 1630
|
| 1989 |
+
},
|
| 1990 |
+
{
|
| 1991 |
+
"epoch": 4.98,
|
| 1992 |
+
"learning_rate": 7.865080677495002e-08,
|
| 1993 |
+
"loss": 2.6786,
|
| 1994 |
+
"step": 1635
|
| 1995 |
+
},
|
| 1996 |
+
{
|
| 1997 |
+
"epoch": 5.0,
|
| 1998 |
+
"learning_rate": 0.0,
|
| 1999 |
+
"loss": 2.8629,
|
| 2000 |
+
"step": 1640
|
| 2001 |
+
},
|
| 2002 |
+
{
|
| 2003 |
+
"epoch": 5.0,
|
| 2004 |
+
"eval_loss": 2.985200881958008,
|
| 2005 |
+
"eval_runtime": 3.2565,
|
| 2006 |
+
"eval_samples_per_second": 142.792,
|
| 2007 |
+
"eval_steps_per_second": 18.118,
|
| 2008 |
+
"step": 1640
|
| 2009 |
}
|
| 2010 |
],
|
| 2011 |
+
"max_steps": 3936,
|
| 2012 |
+
"num_train_epochs": 12,
|
| 2013 |
+
"total_flos": 1711070871552000.0,
|
| 2014 |
"trial_name": null,
|
| 2015 |
"trial_params": null
|
| 2016 |
}
|
training_args.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 3311
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:05834366ef55540d4e60793dd81a674c9d97225933a6dddaf453b9a50205534e
|
| 3 |
size 3311
|