Model save
Browse files- README.md +72 -0
- all_results.json +7 -0
- generation_config.json +7 -0
- train_results.json +7 -0
- trainer_state.json +2417 -0
README.md
ADDED
|
@@ -0,0 +1,72 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
license: apache-2.0
|
| 3 |
+
base_model: DatPySci/pythia-1b-sft-full
|
| 4 |
+
tags:
|
| 5 |
+
- generated_from_trainer
|
| 6 |
+
model-index:
|
| 7 |
+
- name: pythia-1b-kto-iter0-epoch1
|
| 8 |
+
results: []
|
| 9 |
+
---
|
| 10 |
+
|
| 11 |
+
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
| 12 |
+
should probably proofread and complete it, then remove this comment. -->
|
| 13 |
+
|
| 14 |
+
# pythia-1b-kto-iter0-epoch1
|
| 15 |
+
|
| 16 |
+
This model is a fine-tuned version of [DatPySci/pythia-1b-sft-full](https://huggingface.co/DatPySci/pythia-1b-sft-full) on an unknown dataset.
|
| 17 |
+
It achieves the following results on the evaluation set:
|
| 18 |
+
- Loss: 0.3355
|
| 19 |
+
- Rewards/real: -0.0044
|
| 20 |
+
- Rewards/generated: -1.1080
|
| 21 |
+
- Rewards/accuracies: 0.9600
|
| 22 |
+
- Rewards/margins: 1.1036
|
| 23 |
+
- Logps/generated: -571.6245
|
| 24 |
+
- Logps/real: -468.8171
|
| 25 |
+
- Logits/generated: 0.2235
|
| 26 |
+
- Logits/real: -0.2848
|
| 27 |
+
|
| 28 |
+
## Model description
|
| 29 |
+
|
| 30 |
+
More information needed
|
| 31 |
+
|
| 32 |
+
## Intended uses & limitations
|
| 33 |
+
|
| 34 |
+
More information needed
|
| 35 |
+
|
| 36 |
+
## Training and evaluation data
|
| 37 |
+
|
| 38 |
+
More information needed
|
| 39 |
+
|
| 40 |
+
## Training procedure
|
| 41 |
+
|
| 42 |
+
### Training hyperparameters
|
| 43 |
+
|
| 44 |
+
The following hyperparameters were used during training:
|
| 45 |
+
- learning_rate: 5e-07
|
| 46 |
+
- train_batch_size: 4
|
| 47 |
+
- eval_batch_size: 4
|
| 48 |
+
- seed: 42
|
| 49 |
+
- distributed_type: multi-GPU
|
| 50 |
+
- num_devices: 2
|
| 51 |
+
- gradient_accumulation_steps: 8
|
| 52 |
+
- total_train_batch_size: 64
|
| 53 |
+
- total_eval_batch_size: 8
|
| 54 |
+
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
| 55 |
+
- lr_scheduler_type: linear
|
| 56 |
+
- lr_scheduler_warmup_ratio: 0.1
|
| 57 |
+
- num_epochs: 1
|
| 58 |
+
|
| 59 |
+
### Training results
|
| 60 |
+
|
| 61 |
+
| Training Loss | Epoch | Step | Validation Loss | Rewards/real | Rewards/generated | Rewards/accuracies | Rewards/margins | Logps/generated | Logps/real | Logits/generated | Logits/real |
|
| 62 |
+
|:-------------:|:-----:|:----:|:---------------:|:------------:|:-----------------:|:------------------:|:---------------:|:---------------:|:----------:|:----------------:|:-----------:|
|
| 63 |
+
| 0.4016 | 0.38 | 300 | 0.3914 | 0.0270 | -0.8143 | 0.9320 | 0.8413 | -568.6872 | -468.5030 | 0.2573 | -0.2517 |
|
| 64 |
+
| 0.3451 | 0.77 | 600 | 0.3355 | -0.0044 | -1.1080 | 0.9600 | 1.1036 | -571.6245 | -468.8171 | 0.2235 | -0.2848 |
|
| 65 |
+
|
| 66 |
+
|
| 67 |
+
### Framework versions
|
| 68 |
+
|
| 69 |
+
- Transformers 4.38.1
|
| 70 |
+
- Pytorch 2.2.1
|
| 71 |
+
- Datasets 2.17.1
|
| 72 |
+
- Tokenizers 0.15.2
|
all_results.json
ADDED
|
@@ -0,0 +1,7 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 1.0,
|
| 3 |
+
"train_loss": 0.41590739681687156,
|
| 4 |
+
"train_runtime": 9143.8407,
|
| 5 |
+
"train_samples_per_second": 5.468,
|
| 6 |
+
"train_steps_per_second": 0.085
|
| 7 |
+
}
|
generation_config.json
ADDED
|
@@ -0,0 +1,7 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_from_model_config": true,
|
| 3 |
+
"bos_token_id": 0,
|
| 4 |
+
"eos_token_id": 0,
|
| 5 |
+
"transformers_version": "4.38.1",
|
| 6 |
+
"use_cache": false
|
| 7 |
+
}
|
train_results.json
ADDED
|
@@ -0,0 +1,7 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 1.0,
|
| 3 |
+
"train_loss": 0.41590739681687156,
|
| 4 |
+
"train_runtime": 9143.8407,
|
| 5 |
+
"train_samples_per_second": 5.468,
|
| 6 |
+
"train_steps_per_second": 0.085
|
| 7 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,2417 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 0.99968,
|
| 5 |
+
"eval_steps": 300,
|
| 6 |
+
"global_step": 781,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.0,
|
| 13 |
+
"grad_norm": 51.5,
|
| 14 |
+
"learning_rate": 6.329113924050633e-09,
|
| 15 |
+
"logits/generated": 0.1862616091966629,
|
| 16 |
+
"logits/real": -0.24196800589561462,
|
| 17 |
+
"logps/generated": -538.626953125,
|
| 18 |
+
"logps/real": -411.1020202636719,
|
| 19 |
+
"loss": 0.6931,
|
| 20 |
+
"rewards/accuracies": 0.0,
|
| 21 |
+
"rewards/generated": 0.0,
|
| 22 |
+
"rewards/margins": 0.0,
|
| 23 |
+
"rewards/real": 0.0,
|
| 24 |
+
"step": 1
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.01,
|
| 28 |
+
"grad_norm": 56.0,
|
| 29 |
+
"learning_rate": 3.1645569620253166e-08,
|
| 30 |
+
"logits/generated": 0.348865807056427,
|
| 31 |
+
"logits/real": -0.1247345358133316,
|
| 32 |
+
"logps/generated": -534.8592529296875,
|
| 33 |
+
"logps/real": -401.95263671875,
|
| 34 |
+
"loss": 0.7116,
|
| 35 |
+
"rewards/accuracies": 0.34375,
|
| 36 |
+
"rewards/generated": 0.018089786171913147,
|
| 37 |
+
"rewards/margins": -0.02859283983707428,
|
| 38 |
+
"rewards/real": -0.010503048077225685,
|
| 39 |
+
"step": 5
|
| 40 |
+
},
|
| 41 |
+
{
|
| 42 |
+
"epoch": 0.01,
|
| 43 |
+
"grad_norm": 52.25,
|
| 44 |
+
"learning_rate": 6.329113924050633e-08,
|
| 45 |
+
"logits/generated": 0.37161511182785034,
|
| 46 |
+
"logits/real": -0.12119238078594208,
|
| 47 |
+
"logps/generated": -555.2481689453125,
|
| 48 |
+
"logps/real": -418.7288513183594,
|
| 49 |
+
"loss": 0.707,
|
| 50 |
+
"rewards/accuracies": 0.5249999761581421,
|
| 51 |
+
"rewards/generated": -0.016772981733083725,
|
| 52 |
+
"rewards/margins": -0.006616026163101196,
|
| 53 |
+
"rewards/real": -0.02338900789618492,
|
| 54 |
+
"step": 10
|
| 55 |
+
},
|
| 56 |
+
{
|
| 57 |
+
"epoch": 0.02,
|
| 58 |
+
"grad_norm": 50.75,
|
| 59 |
+
"learning_rate": 9.49367088607595e-08,
|
| 60 |
+
"logits/generated": 0.3857704699039459,
|
| 61 |
+
"logits/real": -0.10262864828109741,
|
| 62 |
+
"logps/generated": -564.0440673828125,
|
| 63 |
+
"logps/real": -434.94561767578125,
|
| 64 |
+
"loss": 0.6843,
|
| 65 |
+
"rewards/accuracies": 0.550000011920929,
|
| 66 |
+
"rewards/generated": -0.02160583809018135,
|
| 67 |
+
"rewards/margins": 0.07228925079107285,
|
| 68 |
+
"rewards/real": 0.05068342760205269,
|
| 69 |
+
"step": 15
|
| 70 |
+
},
|
| 71 |
+
{
|
| 72 |
+
"epoch": 0.03,
|
| 73 |
+
"grad_norm": 52.5,
|
| 74 |
+
"learning_rate": 1.2658227848101266e-07,
|
| 75 |
+
"logits/generated": 0.3662206530570984,
|
| 76 |
+
"logits/real": -0.06378354132175446,
|
| 77 |
+
"logps/generated": -558.4364013671875,
|
| 78 |
+
"logps/real": -439.30816650390625,
|
| 79 |
+
"loss": 0.7284,
|
| 80 |
+
"rewards/accuracies": 0.4375,
|
| 81 |
+
"rewards/generated": 0.002506026765331626,
|
| 82 |
+
"rewards/margins": -0.03106103278696537,
|
| 83 |
+
"rewards/real": -0.028555002063512802,
|
| 84 |
+
"step": 20
|
| 85 |
+
},
|
| 86 |
+
{
|
| 87 |
+
"epoch": 0.03,
|
| 88 |
+
"grad_norm": 53.25,
|
| 89 |
+
"learning_rate": 1.5822784810126582e-07,
|
| 90 |
+
"logits/generated": 0.39896726608276367,
|
| 91 |
+
"logits/real": -0.12075452506542206,
|
| 92 |
+
"logps/generated": -565.0833740234375,
|
| 93 |
+
"logps/real": -439.34344482421875,
|
| 94 |
+
"loss": 0.708,
|
| 95 |
+
"rewards/accuracies": 0.48750001192092896,
|
| 96 |
+
"rewards/generated": 0.013953039422631264,
|
| 97 |
+
"rewards/margins": -0.00905968714505434,
|
| 98 |
+
"rewards/real": 0.004893349949270487,
|
| 99 |
+
"step": 25
|
| 100 |
+
},
|
| 101 |
+
{
|
| 102 |
+
"epoch": 0.04,
|
| 103 |
+
"grad_norm": 54.75,
|
| 104 |
+
"learning_rate": 1.89873417721519e-07,
|
| 105 |
+
"logits/generated": 0.36376339197158813,
|
| 106 |
+
"logits/real": -0.15110069513320923,
|
| 107 |
+
"logps/generated": -563.3927612304688,
|
| 108 |
+
"logps/real": -444.236083984375,
|
| 109 |
+
"loss": 0.6823,
|
| 110 |
+
"rewards/accuracies": 0.5874999761581421,
|
| 111 |
+
"rewards/generated": -0.03245529532432556,
|
| 112 |
+
"rewards/margins": 0.041769616305828094,
|
| 113 |
+
"rewards/real": 0.009314320981502533,
|
| 114 |
+
"step": 30
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.04,
|
| 118 |
+
"grad_norm": 53.0,
|
| 119 |
+
"learning_rate": 2.2151898734177212e-07,
|
| 120 |
+
"logits/generated": 0.3546258807182312,
|
| 121 |
+
"logits/real": -0.08531169593334198,
|
| 122 |
+
"logps/generated": -567.1099853515625,
|
| 123 |
+
"logps/real": -426.37530517578125,
|
| 124 |
+
"loss": 0.6792,
|
| 125 |
+
"rewards/accuracies": 0.5249999761581421,
|
| 126 |
+
"rewards/generated": -0.04005850851535797,
|
| 127 |
+
"rewards/margins": 0.062456123530864716,
|
| 128 |
+
"rewards/real": 0.022397616878151894,
|
| 129 |
+
"step": 35
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 0.05,
|
| 133 |
+
"grad_norm": 54.0,
|
| 134 |
+
"learning_rate": 2.5316455696202533e-07,
|
| 135 |
+
"logits/generated": 0.3955411911010742,
|
| 136 |
+
"logits/real": -0.09925423562526703,
|
| 137 |
+
"logps/generated": -584.2412719726562,
|
| 138 |
+
"logps/real": -481.74151611328125,
|
| 139 |
+
"loss": 0.7088,
|
| 140 |
+
"rewards/accuracies": 0.518750011920929,
|
| 141 |
+
"rewards/generated": 0.002044306369498372,
|
| 142 |
+
"rewards/margins": 0.014470143243670464,
|
| 143 |
+
"rewards/real": 0.016514450311660767,
|
| 144 |
+
"step": 40
|
| 145 |
+
},
|
| 146 |
+
{
|
| 147 |
+
"epoch": 0.06,
|
| 148 |
+
"grad_norm": 49.75,
|
| 149 |
+
"learning_rate": 2.848101265822785e-07,
|
| 150 |
+
"logits/generated": 0.3811967372894287,
|
| 151 |
+
"logits/real": -0.10687746107578278,
|
| 152 |
+
"logps/generated": -589.9839477539062,
|
| 153 |
+
"logps/real": -475.3309020996094,
|
| 154 |
+
"loss": 0.6826,
|
| 155 |
+
"rewards/accuracies": 0.518750011920929,
|
| 156 |
+
"rewards/generated": -0.01990634948015213,
|
| 157 |
+
"rewards/margins": 0.01696391962468624,
|
| 158 |
+
"rewards/real": -0.0029424328822642565,
|
| 159 |
+
"step": 45
|
| 160 |
+
},
|
| 161 |
+
{
|
| 162 |
+
"epoch": 0.06,
|
| 163 |
+
"grad_norm": 48.0,
|
| 164 |
+
"learning_rate": 3.1645569620253163e-07,
|
| 165 |
+
"logits/generated": 0.36268311738967896,
|
| 166 |
+
"logits/real": -0.13508598506450653,
|
| 167 |
+
"logps/generated": -559.43505859375,
|
| 168 |
+
"logps/real": -434.68145751953125,
|
| 169 |
+
"loss": 0.6677,
|
| 170 |
+
"rewards/accuracies": 0.581250011920929,
|
| 171 |
+
"rewards/generated": -0.050643421709537506,
|
| 172 |
+
"rewards/margins": 0.07995937764644623,
|
| 173 |
+
"rewards/real": 0.029315968975424767,
|
| 174 |
+
"step": 50
|
| 175 |
+
},
|
| 176 |
+
{
|
| 177 |
+
"epoch": 0.07,
|
| 178 |
+
"grad_norm": 54.25,
|
| 179 |
+
"learning_rate": 3.481012658227848e-07,
|
| 180 |
+
"logits/generated": 0.3685925602912903,
|
| 181 |
+
"logits/real": -0.11418122053146362,
|
| 182 |
+
"logps/generated": -591.776611328125,
|
| 183 |
+
"logps/real": -450.0094299316406,
|
| 184 |
+
"loss": 0.6876,
|
| 185 |
+
"rewards/accuracies": 0.5687500238418579,
|
| 186 |
+
"rewards/generated": -0.06491654366254807,
|
| 187 |
+
"rewards/margins": 0.06822647154331207,
|
| 188 |
+
"rewards/real": 0.0033099246211349964,
|
| 189 |
+
"step": 55
|
| 190 |
+
},
|
| 191 |
+
{
|
| 192 |
+
"epoch": 0.08,
|
| 193 |
+
"grad_norm": 52.0,
|
| 194 |
+
"learning_rate": 3.79746835443038e-07,
|
| 195 |
+
"logits/generated": 0.39532652497291565,
|
| 196 |
+
"logits/real": -0.08878536522388458,
|
| 197 |
+
"logps/generated": -570.0611572265625,
|
| 198 |
+
"logps/real": -456.876708984375,
|
| 199 |
+
"loss": 0.6667,
|
| 200 |
+
"rewards/accuracies": 0.606249988079071,
|
| 201 |
+
"rewards/generated": -0.08224920183420181,
|
| 202 |
+
"rewards/margins": 0.05731937289237976,
|
| 203 |
+
"rewards/real": -0.02492983639240265,
|
| 204 |
+
"step": 60
|
| 205 |
+
},
|
| 206 |
+
{
|
| 207 |
+
"epoch": 0.08,
|
| 208 |
+
"grad_norm": 48.0,
|
| 209 |
+
"learning_rate": 4.1139240506329117e-07,
|
| 210 |
+
"logits/generated": 0.3644171953201294,
|
| 211 |
+
"logits/real": -0.11488844454288483,
|
| 212 |
+
"logps/generated": -542.10498046875,
|
| 213 |
+
"logps/real": -441.2933044433594,
|
| 214 |
+
"loss": 0.6619,
|
| 215 |
+
"rewards/accuracies": 0.581250011920929,
|
| 216 |
+
"rewards/generated": -0.07368786633014679,
|
| 217 |
+
"rewards/margins": 0.10115363448858261,
|
| 218 |
+
"rewards/real": 0.02746577560901642,
|
| 219 |
+
"step": 65
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.09,
|
| 223 |
+
"grad_norm": 48.25,
|
| 224 |
+
"learning_rate": 4.4303797468354424e-07,
|
| 225 |
+
"logits/generated": 0.3911452889442444,
|
| 226 |
+
"logits/real": -0.15676851570606232,
|
| 227 |
+
"logps/generated": -570.861083984375,
|
| 228 |
+
"logps/real": -487.00848388671875,
|
| 229 |
+
"loss": 0.6486,
|
| 230 |
+
"rewards/accuracies": 0.637499988079071,
|
| 231 |
+
"rewards/generated": -0.1312028169631958,
|
| 232 |
+
"rewards/margins": 0.1361098289489746,
|
| 233 |
+
"rewards/real": 0.004907005000859499,
|
| 234 |
+
"step": 70
|
| 235 |
+
},
|
| 236 |
+
{
|
| 237 |
+
"epoch": 0.1,
|
| 238 |
+
"grad_norm": 51.25,
|
| 239 |
+
"learning_rate": 4.746835443037974e-07,
|
| 240 |
+
"logits/generated": 0.38353046774864197,
|
| 241 |
+
"logits/real": -0.09031276404857635,
|
| 242 |
+
"logps/generated": -571.59765625,
|
| 243 |
+
"logps/real": -419.74285888671875,
|
| 244 |
+
"loss": 0.6689,
|
| 245 |
+
"rewards/accuracies": 0.5562499761581421,
|
| 246 |
+
"rewards/generated": -0.09446991980075836,
|
| 247 |
+
"rewards/margins": 0.06767363846302032,
|
| 248 |
+
"rewards/real": -0.026796285063028336,
|
| 249 |
+
"step": 75
|
| 250 |
+
},
|
| 251 |
+
{
|
| 252 |
+
"epoch": 0.1,
|
| 253 |
+
"grad_norm": 51.25,
|
| 254 |
+
"learning_rate": 4.992877492877492e-07,
|
| 255 |
+
"logits/generated": 0.3525523245334625,
|
| 256 |
+
"logits/real": -0.14858296513557434,
|
| 257 |
+
"logps/generated": -547.0704345703125,
|
| 258 |
+
"logps/real": -442.40228271484375,
|
| 259 |
+
"loss": 0.638,
|
| 260 |
+
"rewards/accuracies": 0.643750011920929,
|
| 261 |
+
"rewards/generated": -0.11326056718826294,
|
| 262 |
+
"rewards/margins": 0.12944021821022034,
|
| 263 |
+
"rewards/real": 0.016179624944925308,
|
| 264 |
+
"step": 80
|
| 265 |
+
},
|
| 266 |
+
{
|
| 267 |
+
"epoch": 0.11,
|
| 268 |
+
"grad_norm": 48.0,
|
| 269 |
+
"learning_rate": 4.957264957264958e-07,
|
| 270 |
+
"logits/generated": 0.360097736120224,
|
| 271 |
+
"logits/real": -0.18511034548282623,
|
| 272 |
+
"logps/generated": -535.1893310546875,
|
| 273 |
+
"logps/real": -456.3688049316406,
|
| 274 |
+
"loss": 0.611,
|
| 275 |
+
"rewards/accuracies": 0.737500011920929,
|
| 276 |
+
"rewards/generated": -0.18863049149513245,
|
| 277 |
+
"rewards/margins": 0.2453247308731079,
|
| 278 |
+
"rewards/real": 0.05669426918029785,
|
| 279 |
+
"step": 85
|
| 280 |
+
},
|
| 281 |
+
{
|
| 282 |
+
"epoch": 0.12,
|
| 283 |
+
"grad_norm": 51.0,
|
| 284 |
+
"learning_rate": 4.921652421652421e-07,
|
| 285 |
+
"logits/generated": 0.3094131052494049,
|
| 286 |
+
"logits/real": -0.13698112964630127,
|
| 287 |
+
"logps/generated": -578.9557495117188,
|
| 288 |
+
"logps/real": -435.10626220703125,
|
| 289 |
+
"loss": 0.6183,
|
| 290 |
+
"rewards/accuracies": 0.706250011920929,
|
| 291 |
+
"rewards/generated": -0.16466441750526428,
|
| 292 |
+
"rewards/margins": 0.2127438485622406,
|
| 293 |
+
"rewards/real": 0.04807942733168602,
|
| 294 |
+
"step": 90
|
| 295 |
+
},
|
| 296 |
+
{
|
| 297 |
+
"epoch": 0.12,
|
| 298 |
+
"grad_norm": 49.5,
|
| 299 |
+
"learning_rate": 4.886039886039886e-07,
|
| 300 |
+
"logits/generated": 0.3409915864467621,
|
| 301 |
+
"logits/real": -0.15773525834083557,
|
| 302 |
+
"logps/generated": -562.3658447265625,
|
| 303 |
+
"logps/real": -472.12261962890625,
|
| 304 |
+
"loss": 0.5971,
|
| 305 |
+
"rewards/accuracies": 0.7124999761581421,
|
| 306 |
+
"rewards/generated": -0.2226313352584839,
|
| 307 |
+
"rewards/margins": 0.24577657878398895,
|
| 308 |
+
"rewards/real": 0.023145252838730812,
|
| 309 |
+
"step": 95
|
| 310 |
+
},
|
| 311 |
+
{
|
| 312 |
+
"epoch": 0.13,
|
| 313 |
+
"grad_norm": 46.5,
|
| 314 |
+
"learning_rate": 4.850427350427351e-07,
|
| 315 |
+
"logits/generated": 0.3600820302963257,
|
| 316 |
+
"logits/real": -0.22068698704242706,
|
| 317 |
+
"logps/generated": -518.3560791015625,
|
| 318 |
+
"logps/real": -505.414794921875,
|
| 319 |
+
"loss": 0.5968,
|
| 320 |
+
"rewards/accuracies": 0.762499988079071,
|
| 321 |
+
"rewards/generated": -0.22140724956989288,
|
| 322 |
+
"rewards/margins": 0.2667246460914612,
|
| 323 |
+
"rewards/real": 0.045317377895116806,
|
| 324 |
+
"step": 100
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 0.13,
|
| 328 |
+
"grad_norm": 47.0,
|
| 329 |
+
"learning_rate": 4.814814814814814e-07,
|
| 330 |
+
"logits/generated": 0.3369310796260834,
|
| 331 |
+
"logits/real": -0.16511984169483185,
|
| 332 |
+
"logps/generated": -597.7069091796875,
|
| 333 |
+
"logps/real": -469.93951416015625,
|
| 334 |
+
"loss": 0.5653,
|
| 335 |
+
"rewards/accuracies": 0.7875000238418579,
|
| 336 |
+
"rewards/generated": -0.24688784778118134,
|
| 337 |
+
"rewards/margins": 0.31749141216278076,
|
| 338 |
+
"rewards/real": 0.07060358673334122,
|
| 339 |
+
"step": 105
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 0.14,
|
| 343 |
+
"grad_norm": 49.75,
|
| 344 |
+
"learning_rate": 4.779202279202279e-07,
|
| 345 |
+
"logits/generated": 0.31372249126434326,
|
| 346 |
+
"logits/real": -0.13822853565216064,
|
| 347 |
+
"logps/generated": -559.5579223632812,
|
| 348 |
+
"logps/real": -436.1619567871094,
|
| 349 |
+
"loss": 0.5843,
|
| 350 |
+
"rewards/accuracies": 0.768750011920929,
|
| 351 |
+
"rewards/generated": -0.24767298996448517,
|
| 352 |
+
"rewards/margins": 0.2614055275917053,
|
| 353 |
+
"rewards/real": 0.013732570223510265,
|
| 354 |
+
"step": 110
|
| 355 |
+
},
|
| 356 |
+
{
|
| 357 |
+
"epoch": 0.15,
|
| 358 |
+
"grad_norm": 44.75,
|
| 359 |
+
"learning_rate": 4.743589743589743e-07,
|
| 360 |
+
"logits/generated": 0.3527405858039856,
|
| 361 |
+
"logits/real": -0.16191214323043823,
|
| 362 |
+
"logps/generated": -540.7269287109375,
|
| 363 |
+
"logps/real": -481.5582580566406,
|
| 364 |
+
"loss": 0.5474,
|
| 365 |
+
"rewards/accuracies": 0.824999988079071,
|
| 366 |
+
"rewards/generated": -0.3221510648727417,
|
| 367 |
+
"rewards/margins": 0.33944302797317505,
|
| 368 |
+
"rewards/real": 0.017291929572820663,
|
| 369 |
+
"step": 115
|
| 370 |
+
},
|
| 371 |
+
{
|
| 372 |
+
"epoch": 0.15,
|
| 373 |
+
"grad_norm": 44.0,
|
| 374 |
+
"learning_rate": 4.707977207977208e-07,
|
| 375 |
+
"logits/generated": 0.31554654240608215,
|
| 376 |
+
"logits/real": -0.14937585592269897,
|
| 377 |
+
"logps/generated": -533.3251953125,
|
| 378 |
+
"logps/real": -425.991455078125,
|
| 379 |
+
"loss": 0.5488,
|
| 380 |
+
"rewards/accuracies": 0.831250011920929,
|
| 381 |
+
"rewards/generated": -0.350756973028183,
|
| 382 |
+
"rewards/margins": 0.37362727522850037,
|
| 383 |
+
"rewards/real": 0.022870313376188278,
|
| 384 |
+
"step": 120
|
| 385 |
+
},
|
| 386 |
+
{
|
| 387 |
+
"epoch": 0.16,
|
| 388 |
+
"grad_norm": 43.25,
|
| 389 |
+
"learning_rate": 4.672364672364672e-07,
|
| 390 |
+
"logits/generated": 0.31335240602493286,
|
| 391 |
+
"logits/real": -0.16108378767967224,
|
| 392 |
+
"logps/generated": -556.67919921875,
|
| 393 |
+
"logps/real": -451.10296630859375,
|
| 394 |
+
"loss": 0.532,
|
| 395 |
+
"rewards/accuracies": 0.856249988079071,
|
| 396 |
+
"rewards/generated": -0.3505321741104126,
|
| 397 |
+
"rewards/margins": 0.4209725260734558,
|
| 398 |
+
"rewards/real": 0.07044035941362381,
|
| 399 |
+
"step": 125
|
| 400 |
+
},
|
| 401 |
+
{
|
| 402 |
+
"epoch": 0.17,
|
| 403 |
+
"grad_norm": 45.75,
|
| 404 |
+
"learning_rate": 4.6367521367521367e-07,
|
| 405 |
+
"logits/generated": 0.32702523469924927,
|
| 406 |
+
"logits/real": -0.16390064358711243,
|
| 407 |
+
"logps/generated": -558.4036254882812,
|
| 408 |
+
"logps/real": -435.8067932128906,
|
| 409 |
+
"loss": 0.5325,
|
| 410 |
+
"rewards/accuracies": 0.8687499761581421,
|
| 411 |
+
"rewards/generated": -0.3638128936290741,
|
| 412 |
+
"rewards/margins": 0.40913742780685425,
|
| 413 |
+
"rewards/real": 0.045324526727199554,
|
| 414 |
+
"step": 130
|
| 415 |
+
},
|
| 416 |
+
{
|
| 417 |
+
"epoch": 0.17,
|
| 418 |
+
"grad_norm": 44.0,
|
| 419 |
+
"learning_rate": 4.601139601139601e-07,
|
| 420 |
+
"logits/generated": 0.28192949295043945,
|
| 421 |
+
"logits/real": -0.23023930191993713,
|
| 422 |
+
"logps/generated": -578.7821655273438,
|
| 423 |
+
"logps/real": -447.467529296875,
|
| 424 |
+
"loss": 0.532,
|
| 425 |
+
"rewards/accuracies": 0.856249988079071,
|
| 426 |
+
"rewards/generated": -0.37364333868026733,
|
| 427 |
+
"rewards/margins": 0.4050399363040924,
|
| 428 |
+
"rewards/real": 0.03139660134911537,
|
| 429 |
+
"step": 135
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 0.18,
|
| 433 |
+
"grad_norm": 45.25,
|
| 434 |
+
"learning_rate": 4.5655270655270654e-07,
|
| 435 |
+
"logits/generated": 0.3297092914581299,
|
| 436 |
+
"logits/real": -0.24259880185127258,
|
| 437 |
+
"logps/generated": -557.1586303710938,
|
| 438 |
+
"logps/real": -533.3529663085938,
|
| 439 |
+
"loss": 0.5193,
|
| 440 |
+
"rewards/accuracies": 0.84375,
|
| 441 |
+
"rewards/generated": -0.3680626153945923,
|
| 442 |
+
"rewards/margins": 0.42688456177711487,
|
| 443 |
+
"rewards/real": 0.058821968734264374,
|
| 444 |
+
"step": 140
|
| 445 |
+
},
|
| 446 |
+
{
|
| 447 |
+
"epoch": 0.19,
|
| 448 |
+
"grad_norm": 41.25,
|
| 449 |
+
"learning_rate": 4.5299145299145297e-07,
|
| 450 |
+
"logits/generated": 0.2855086922645569,
|
| 451 |
+
"logits/real": -0.16449978947639465,
|
| 452 |
+
"logps/generated": -531.9002685546875,
|
| 453 |
+
"logps/real": -427.7098693847656,
|
| 454 |
+
"loss": 0.5324,
|
| 455 |
+
"rewards/accuracies": 0.856249988079071,
|
| 456 |
+
"rewards/generated": -0.4032559394836426,
|
| 457 |
+
"rewards/margins": 0.4310643672943115,
|
| 458 |
+
"rewards/real": 0.027808407321572304,
|
| 459 |
+
"step": 145
|
| 460 |
+
},
|
| 461 |
+
{
|
| 462 |
+
"epoch": 0.19,
|
| 463 |
+
"grad_norm": 47.25,
|
| 464 |
+
"learning_rate": 4.494301994301994e-07,
|
| 465 |
+
"logits/generated": 0.26499122381210327,
|
| 466 |
+
"logits/real": -0.17654789984226227,
|
| 467 |
+
"logps/generated": -544.5510864257812,
|
| 468 |
+
"logps/real": -443.2236328125,
|
| 469 |
+
"loss": 0.5146,
|
| 470 |
+
"rewards/accuracies": 0.8374999761581421,
|
| 471 |
+
"rewards/generated": -0.46017971634864807,
|
| 472 |
+
"rewards/margins": 0.47648102045059204,
|
| 473 |
+
"rewards/real": 0.016301285475492477,
|
| 474 |
+
"step": 150
|
| 475 |
+
},
|
| 476 |
+
{
|
| 477 |
+
"epoch": 0.2,
|
| 478 |
+
"grad_norm": 43.75,
|
| 479 |
+
"learning_rate": 4.4586894586894584e-07,
|
| 480 |
+
"logits/generated": 0.32447490096092224,
|
| 481 |
+
"logits/real": -0.19060441851615906,
|
| 482 |
+
"logps/generated": -565.9913330078125,
|
| 483 |
+
"logps/real": -485.25762939453125,
|
| 484 |
+
"loss": 0.5035,
|
| 485 |
+
"rewards/accuracies": 0.8374999761581421,
|
| 486 |
+
"rewards/generated": -0.44145339727401733,
|
| 487 |
+
"rewards/margins": 0.4792710244655609,
|
| 488 |
+
"rewards/real": 0.03781764581799507,
|
| 489 |
+
"step": 155
|
| 490 |
+
},
|
| 491 |
+
{
|
| 492 |
+
"epoch": 0.2,
|
| 493 |
+
"grad_norm": 43.75,
|
| 494 |
+
"learning_rate": 4.423076923076923e-07,
|
| 495 |
+
"logits/generated": 0.3035694360733032,
|
| 496 |
+
"logits/real": -0.22318044304847717,
|
| 497 |
+
"logps/generated": -566.83251953125,
|
| 498 |
+
"logps/real": -444.7403259277344,
|
| 499 |
+
"loss": 0.5005,
|
| 500 |
+
"rewards/accuracies": 0.862500011920929,
|
| 501 |
+
"rewards/generated": -0.4472742974758148,
|
| 502 |
+
"rewards/margins": 0.5138620138168335,
|
| 503 |
+
"rewards/real": 0.06658776849508286,
|
| 504 |
+
"step": 160
|
| 505 |
+
},
|
| 506 |
+
{
|
| 507 |
+
"epoch": 0.21,
|
| 508 |
+
"grad_norm": 41.25,
|
| 509 |
+
"learning_rate": 4.3874643874643876e-07,
|
| 510 |
+
"logits/generated": 0.3123244643211365,
|
| 511 |
+
"logits/real": -0.1916072815656662,
|
| 512 |
+
"logps/generated": -558.3717041015625,
|
| 513 |
+
"logps/real": -460.4476623535156,
|
| 514 |
+
"loss": 0.4883,
|
| 515 |
+
"rewards/accuracies": 0.800000011920929,
|
| 516 |
+
"rewards/generated": -0.4848392605781555,
|
| 517 |
+
"rewards/margins": 0.4970267713069916,
|
| 518 |
+
"rewards/real": 0.012187558226287365,
|
| 519 |
+
"step": 165
|
| 520 |
+
},
|
| 521 |
+
{
|
| 522 |
+
"epoch": 0.22,
|
| 523 |
+
"grad_norm": 41.25,
|
| 524 |
+
"learning_rate": 4.3518518518518514e-07,
|
| 525 |
+
"logits/generated": 0.250892698764801,
|
| 526 |
+
"logits/real": -0.18243929743766785,
|
| 527 |
+
"logps/generated": -536.6226196289062,
|
| 528 |
+
"logps/real": -416.77984619140625,
|
| 529 |
+
"loss": 0.496,
|
| 530 |
+
"rewards/accuracies": 0.8687499761581421,
|
| 531 |
+
"rewards/generated": -0.4702886939048767,
|
| 532 |
+
"rewards/margins": 0.5208121538162231,
|
| 533 |
+
"rewards/real": 0.05052344128489494,
|
| 534 |
+
"step": 170
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 0.22,
|
| 538 |
+
"grad_norm": 41.0,
|
| 539 |
+
"learning_rate": 4.3162393162393163e-07,
|
| 540 |
+
"logits/generated": 0.29632076621055603,
|
| 541 |
+
"logits/real": -0.2431728094816208,
|
| 542 |
+
"logps/generated": -538.0128173828125,
|
| 543 |
+
"logps/real": -478.095703125,
|
| 544 |
+
"loss": 0.4654,
|
| 545 |
+
"rewards/accuracies": 0.8999999761581421,
|
| 546 |
+
"rewards/generated": -0.5014994144439697,
|
| 547 |
+
"rewards/margins": 0.5601893663406372,
|
| 548 |
+
"rewards/real": 0.0586898997426033,
|
| 549 |
+
"step": 175
|
| 550 |
+
},
|
| 551 |
+
{
|
| 552 |
+
"epoch": 0.23,
|
| 553 |
+
"grad_norm": 44.5,
|
| 554 |
+
"learning_rate": 4.2806267806267807e-07,
|
| 555 |
+
"logits/generated": 0.277912437915802,
|
| 556 |
+
"logits/real": -0.2574039101600647,
|
| 557 |
+
"logps/generated": -561.8548583984375,
|
| 558 |
+
"logps/real": -444.5439453125,
|
| 559 |
+
"loss": 0.4828,
|
| 560 |
+
"rewards/accuracies": 0.875,
|
| 561 |
+
"rewards/generated": -0.5042586922645569,
|
| 562 |
+
"rewards/margins": 0.5849383473396301,
|
| 563 |
+
"rewards/real": 0.08067961782217026,
|
| 564 |
+
"step": 180
|
| 565 |
+
},
|
| 566 |
+
{
|
| 567 |
+
"epoch": 0.24,
|
| 568 |
+
"grad_norm": 43.25,
|
| 569 |
+
"learning_rate": 4.245014245014245e-07,
|
| 570 |
+
"logits/generated": 0.33910220861434937,
|
| 571 |
+
"logits/real": -0.11496976763010025,
|
| 572 |
+
"logps/generated": -599.3956298828125,
|
| 573 |
+
"logps/real": -421.74334716796875,
|
| 574 |
+
"loss": 0.4619,
|
| 575 |
+
"rewards/accuracies": 0.8999999761581421,
|
| 576 |
+
"rewards/generated": -0.572574257850647,
|
| 577 |
+
"rewards/margins": 0.6046732664108276,
|
| 578 |
+
"rewards/real": 0.032099105417728424,
|
| 579 |
+
"step": 185
|
| 580 |
+
},
|
| 581 |
+
{
|
| 582 |
+
"epoch": 0.24,
|
| 583 |
+
"grad_norm": 43.5,
|
| 584 |
+
"learning_rate": 4.2094017094017093e-07,
|
| 585 |
+
"logits/generated": 0.32159385085105896,
|
| 586 |
+
"logits/real": -0.17387528717517853,
|
| 587 |
+
"logps/generated": -586.5929565429688,
|
| 588 |
+
"logps/real": -426.36444091796875,
|
| 589 |
+
"loss": 0.483,
|
| 590 |
+
"rewards/accuracies": 0.875,
|
| 591 |
+
"rewards/generated": -0.5183109641075134,
|
| 592 |
+
"rewards/margins": 0.5664752721786499,
|
| 593 |
+
"rewards/real": 0.0481642410159111,
|
| 594 |
+
"step": 190
|
| 595 |
+
},
|
| 596 |
+
{
|
| 597 |
+
"epoch": 0.25,
|
| 598 |
+
"grad_norm": 39.5,
|
| 599 |
+
"learning_rate": 4.173789173789173e-07,
|
| 600 |
+
"logits/generated": 0.26057395339012146,
|
| 601 |
+
"logits/real": -0.21449565887451172,
|
| 602 |
+
"logps/generated": -527.934326171875,
|
| 603 |
+
"logps/real": -452.62567138671875,
|
| 604 |
+
"loss": 0.465,
|
| 605 |
+
"rewards/accuracies": 0.893750011920929,
|
| 606 |
+
"rewards/generated": -0.53789222240448,
|
| 607 |
+
"rewards/margins": 0.5843125581741333,
|
| 608 |
+
"rewards/real": 0.04642019420862198,
|
| 609 |
+
"step": 195
|
| 610 |
+
},
|
| 611 |
+
{
|
| 612 |
+
"epoch": 0.26,
|
| 613 |
+
"grad_norm": 39.5,
|
| 614 |
+
"learning_rate": 4.138176638176638e-07,
|
| 615 |
+
"logits/generated": 0.3313707411289215,
|
| 616 |
+
"logits/real": -0.19944889843463898,
|
| 617 |
+
"logps/generated": -580.3385620117188,
|
| 618 |
+
"logps/real": -472.051513671875,
|
| 619 |
+
"loss": 0.4567,
|
| 620 |
+
"rewards/accuracies": 0.862500011920929,
|
| 621 |
+
"rewards/generated": -0.5696550607681274,
|
| 622 |
+
"rewards/margins": 0.5835729837417603,
|
| 623 |
+
"rewards/real": 0.013917927630245686,
|
| 624 |
+
"step": 200
|
| 625 |
+
},
|
| 626 |
+
{
|
| 627 |
+
"epoch": 0.26,
|
| 628 |
+
"grad_norm": 41.0,
|
| 629 |
+
"learning_rate": 4.1025641025641024e-07,
|
| 630 |
+
"logits/generated": 0.3330136835575104,
|
| 631 |
+
"logits/real": -0.1847701370716095,
|
| 632 |
+
"logps/generated": -576.3521728515625,
|
| 633 |
+
"logps/real": -470.685302734375,
|
| 634 |
+
"loss": 0.4635,
|
| 635 |
+
"rewards/accuracies": 0.8687499761581421,
|
| 636 |
+
"rewards/generated": -0.5276187062263489,
|
| 637 |
+
"rewards/margins": 0.5738595724105835,
|
| 638 |
+
"rewards/real": 0.04624096304178238,
|
| 639 |
+
"step": 205
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 0.27,
|
| 643 |
+
"grad_norm": 42.25,
|
| 644 |
+
"learning_rate": 4.0669515669515667e-07,
|
| 645 |
+
"logits/generated": 0.25724920630455017,
|
| 646 |
+
"logits/real": -0.2856508791446686,
|
| 647 |
+
"logps/generated": -558.7901611328125,
|
| 648 |
+
"logps/real": -467.52008056640625,
|
| 649 |
+
"loss": 0.4516,
|
| 650 |
+
"rewards/accuracies": 0.90625,
|
| 651 |
+
"rewards/generated": -0.5960500836372375,
|
| 652 |
+
"rewards/margins": 0.6805692911148071,
|
| 653 |
+
"rewards/real": 0.08451925963163376,
|
| 654 |
+
"step": 210
|
| 655 |
+
},
|
| 656 |
+
{
|
| 657 |
+
"epoch": 0.28,
|
| 658 |
+
"grad_norm": 40.25,
|
| 659 |
+
"learning_rate": 4.031339031339031e-07,
|
| 660 |
+
"logits/generated": 0.28837472200393677,
|
| 661 |
+
"logits/real": -0.19708415865898132,
|
| 662 |
+
"logps/generated": -549.651123046875,
|
| 663 |
+
"logps/real": -457.85919189453125,
|
| 664 |
+
"loss": 0.4424,
|
| 665 |
+
"rewards/accuracies": 0.9125000238418579,
|
| 666 |
+
"rewards/generated": -0.6616542935371399,
|
| 667 |
+
"rewards/margins": 0.6647108793258667,
|
| 668 |
+
"rewards/real": 0.0030566223431378603,
|
| 669 |
+
"step": 215
|
| 670 |
+
},
|
| 671 |
+
{
|
| 672 |
+
"epoch": 0.28,
|
| 673 |
+
"grad_norm": 39.5,
|
| 674 |
+
"learning_rate": 3.995726495726496e-07,
|
| 675 |
+
"logits/generated": 0.30279049277305603,
|
| 676 |
+
"logits/real": -0.18387384712696075,
|
| 677 |
+
"logps/generated": -541.563720703125,
|
| 678 |
+
"logps/real": -444.663818359375,
|
| 679 |
+
"loss": 0.4583,
|
| 680 |
+
"rewards/accuracies": 0.887499988079071,
|
| 681 |
+
"rewards/generated": -0.5965785384178162,
|
| 682 |
+
"rewards/margins": 0.6331891417503357,
|
| 683 |
+
"rewards/real": 0.03661050274968147,
|
| 684 |
+
"step": 220
|
| 685 |
+
},
|
| 686 |
+
{
|
| 687 |
+
"epoch": 0.29,
|
| 688 |
+
"grad_norm": 40.5,
|
| 689 |
+
"learning_rate": 3.9601139601139597e-07,
|
| 690 |
+
"logits/generated": 0.25506392121315,
|
| 691 |
+
"logits/real": -0.23414012789726257,
|
| 692 |
+
"logps/generated": -574.35595703125,
|
| 693 |
+
"logps/real": -450.7875061035156,
|
| 694 |
+
"loss": 0.4459,
|
| 695 |
+
"rewards/accuracies": 0.9375,
|
| 696 |
+
"rewards/generated": -0.6340950131416321,
|
| 697 |
+
"rewards/margins": 0.6694762110710144,
|
| 698 |
+
"rewards/real": 0.0353812575340271,
|
| 699 |
+
"step": 225
|
| 700 |
+
},
|
| 701 |
+
{
|
| 702 |
+
"epoch": 0.29,
|
| 703 |
+
"grad_norm": 37.75,
|
| 704 |
+
"learning_rate": 3.924501424501424e-07,
|
| 705 |
+
"logits/generated": 0.25567466020584106,
|
| 706 |
+
"logits/real": -0.20427341759204865,
|
| 707 |
+
"logps/generated": -546.7042236328125,
|
| 708 |
+
"logps/real": -421.8016662597656,
|
| 709 |
+
"loss": 0.4303,
|
| 710 |
+
"rewards/accuracies": 0.90625,
|
| 711 |
+
"rewards/generated": -0.5935453176498413,
|
| 712 |
+
"rewards/margins": 0.685859203338623,
|
| 713 |
+
"rewards/real": 0.09231381118297577,
|
| 714 |
+
"step": 230
|
| 715 |
+
},
|
| 716 |
+
{
|
| 717 |
+
"epoch": 0.3,
|
| 718 |
+
"grad_norm": 41.75,
|
| 719 |
+
"learning_rate": 3.888888888888889e-07,
|
| 720 |
+
"logits/generated": 0.3079773783683777,
|
| 721 |
+
"logits/real": -0.21253804862499237,
|
| 722 |
+
"logps/generated": -598.437744140625,
|
| 723 |
+
"logps/real": -443.98492431640625,
|
| 724 |
+
"loss": 0.4276,
|
| 725 |
+
"rewards/accuracies": 0.918749988079071,
|
| 726 |
+
"rewards/generated": -0.722990870475769,
|
| 727 |
+
"rewards/margins": 0.7524241209030151,
|
| 728 |
+
"rewards/real": 0.029433395713567734,
|
| 729 |
+
"step": 235
|
| 730 |
+
},
|
| 731 |
+
{
|
| 732 |
+
"epoch": 0.31,
|
| 733 |
+
"grad_norm": 37.5,
|
| 734 |
+
"learning_rate": 3.853276353276353e-07,
|
| 735 |
+
"logits/generated": 0.30521565675735474,
|
| 736 |
+
"logits/real": -0.1906270682811737,
|
| 737 |
+
"logps/generated": -541.72607421875,
|
| 738 |
+
"logps/real": -424.0105895996094,
|
| 739 |
+
"loss": 0.4269,
|
| 740 |
+
"rewards/accuracies": 0.8812500238418579,
|
| 741 |
+
"rewards/generated": -0.667791485786438,
|
| 742 |
+
"rewards/margins": 0.7342725992202759,
|
| 743 |
+
"rewards/real": 0.06648121774196625,
|
| 744 |
+
"step": 240
|
| 745 |
+
},
|
| 746 |
+
{
|
| 747 |
+
"epoch": 0.31,
|
| 748 |
+
"grad_norm": 36.75,
|
| 749 |
+
"learning_rate": 3.8176638176638176e-07,
|
| 750 |
+
"logits/generated": 0.3006458282470703,
|
| 751 |
+
"logits/real": -0.2351672649383545,
|
| 752 |
+
"logps/generated": -559.7643432617188,
|
| 753 |
+
"logps/real": -504.17987060546875,
|
| 754 |
+
"loss": 0.4507,
|
| 755 |
+
"rewards/accuracies": 0.90625,
|
| 756 |
+
"rewards/generated": -0.6496528387069702,
|
| 757 |
+
"rewards/margins": 0.67330002784729,
|
| 758 |
+
"rewards/real": 0.023647133260965347,
|
| 759 |
+
"step": 245
|
| 760 |
+
},
|
| 761 |
+
{
|
| 762 |
+
"epoch": 0.32,
|
| 763 |
+
"grad_norm": 37.0,
|
| 764 |
+
"learning_rate": 3.782051282051282e-07,
|
| 765 |
+
"logits/generated": 0.2616468071937561,
|
| 766 |
+
"logits/real": -0.2550817131996155,
|
| 767 |
+
"logps/generated": -556.2052001953125,
|
| 768 |
+
"logps/real": -438.43743896484375,
|
| 769 |
+
"loss": 0.4118,
|
| 770 |
+
"rewards/accuracies": 0.9375,
|
| 771 |
+
"rewards/generated": -0.7023671865463257,
|
| 772 |
+
"rewards/margins": 0.7656707763671875,
|
| 773 |
+
"rewards/real": 0.06330356001853943,
|
| 774 |
+
"step": 250
|
| 775 |
+
},
|
| 776 |
+
{
|
| 777 |
+
"epoch": 0.33,
|
| 778 |
+
"grad_norm": 37.0,
|
| 779 |
+
"learning_rate": 3.7464387464387463e-07,
|
| 780 |
+
"logits/generated": 0.3042835593223572,
|
| 781 |
+
"logits/real": -0.22142143547534943,
|
| 782 |
+
"logps/generated": -538.9241333007812,
|
| 783 |
+
"logps/real": -484.5042419433594,
|
| 784 |
+
"loss": 0.4347,
|
| 785 |
+
"rewards/accuracies": 0.918749988079071,
|
| 786 |
+
"rewards/generated": -0.7196381092071533,
|
| 787 |
+
"rewards/margins": 0.7357234358787537,
|
| 788 |
+
"rewards/real": 0.016085291281342506,
|
| 789 |
+
"step": 255
|
| 790 |
+
},
|
| 791 |
+
{
|
| 792 |
+
"epoch": 0.33,
|
| 793 |
+
"grad_norm": 38.75,
|
| 794 |
+
"learning_rate": 3.7108262108262107e-07,
|
| 795 |
+
"logits/generated": 0.28268715739250183,
|
| 796 |
+
"logits/real": -0.2414659708738327,
|
| 797 |
+
"logps/generated": -552.683837890625,
|
| 798 |
+
"logps/real": -439.206298828125,
|
| 799 |
+
"loss": 0.4091,
|
| 800 |
+
"rewards/accuracies": 0.9624999761581421,
|
| 801 |
+
"rewards/generated": -0.738116443157196,
|
| 802 |
+
"rewards/margins": 0.7844719886779785,
|
| 803 |
+
"rewards/real": 0.04635554179549217,
|
| 804 |
+
"step": 260
|
| 805 |
+
},
|
| 806 |
+
{
|
| 807 |
+
"epoch": 0.34,
|
| 808 |
+
"grad_norm": 36.5,
|
| 809 |
+
"learning_rate": 3.6752136752136755e-07,
|
| 810 |
+
"logits/generated": 0.2755090296268463,
|
| 811 |
+
"logits/real": -0.2653561234474182,
|
| 812 |
+
"logps/generated": -575.4991455078125,
|
| 813 |
+
"logps/real": -458.960693359375,
|
| 814 |
+
"loss": 0.4082,
|
| 815 |
+
"rewards/accuracies": 0.9312499761581421,
|
| 816 |
+
"rewards/generated": -0.7462833523750305,
|
| 817 |
+
"rewards/margins": 0.7978280782699585,
|
| 818 |
+
"rewards/real": 0.05154471844434738,
|
| 819 |
+
"step": 265
|
| 820 |
+
},
|
| 821 |
+
{
|
| 822 |
+
"epoch": 0.35,
|
| 823 |
+
"grad_norm": 40.25,
|
| 824 |
+
"learning_rate": 3.6396011396011393e-07,
|
| 825 |
+
"logits/generated": 0.3112264573574066,
|
| 826 |
+
"logits/real": -0.2405599057674408,
|
| 827 |
+
"logps/generated": -537.7279052734375,
|
| 828 |
+
"logps/real": -463.67718505859375,
|
| 829 |
+
"loss": 0.4195,
|
| 830 |
+
"rewards/accuracies": 0.918749988079071,
|
| 831 |
+
"rewards/generated": -0.7078372240066528,
|
| 832 |
+
"rewards/margins": 0.7539125084877014,
|
| 833 |
+
"rewards/real": 0.04607529193162918,
|
| 834 |
+
"step": 270
|
| 835 |
+
},
|
| 836 |
+
{
|
| 837 |
+
"epoch": 0.35,
|
| 838 |
+
"grad_norm": 36.75,
|
| 839 |
+
"learning_rate": 3.6039886039886037e-07,
|
| 840 |
+
"logits/generated": 0.2953110635280609,
|
| 841 |
+
"logits/real": -0.21746914088726044,
|
| 842 |
+
"logps/generated": -562.6829833984375,
|
| 843 |
+
"logps/real": -432.52105712890625,
|
| 844 |
+
"loss": 0.3822,
|
| 845 |
+
"rewards/accuracies": 0.949999988079071,
|
| 846 |
+
"rewards/generated": -0.7816409468650818,
|
| 847 |
+
"rewards/margins": 0.8443856239318848,
|
| 848 |
+
"rewards/real": 0.06274472177028656,
|
| 849 |
+
"step": 275
|
| 850 |
+
},
|
| 851 |
+
{
|
| 852 |
+
"epoch": 0.36,
|
| 853 |
+
"grad_norm": 36.75,
|
| 854 |
+
"learning_rate": 3.5683760683760686e-07,
|
| 855 |
+
"logits/generated": 0.29402461647987366,
|
| 856 |
+
"logits/real": -0.16522033512592316,
|
| 857 |
+
"logps/generated": -568.8770751953125,
|
| 858 |
+
"logps/real": -452.81866455078125,
|
| 859 |
+
"loss": 0.3823,
|
| 860 |
+
"rewards/accuracies": 0.9437500238418579,
|
| 861 |
+
"rewards/generated": -0.8258803486824036,
|
| 862 |
+
"rewards/margins": 0.9027735590934753,
|
| 863 |
+
"rewards/real": 0.07689327746629715,
|
| 864 |
+
"step": 280
|
| 865 |
+
},
|
| 866 |
+
{
|
| 867 |
+
"epoch": 0.36,
|
| 868 |
+
"grad_norm": 37.5,
|
| 869 |
+
"learning_rate": 3.5327635327635324e-07,
|
| 870 |
+
"logits/generated": 0.2609436810016632,
|
| 871 |
+
"logits/real": -0.23396773636341095,
|
| 872 |
+
"logps/generated": -557.7428588867188,
|
| 873 |
+
"logps/real": -437.74493408203125,
|
| 874 |
+
"loss": 0.4063,
|
| 875 |
+
"rewards/accuracies": 0.90625,
|
| 876 |
+
"rewards/generated": -0.762947678565979,
|
| 877 |
+
"rewards/margins": 0.7624271512031555,
|
| 878 |
+
"rewards/real": -0.0005204956978559494,
|
| 879 |
+
"step": 285
|
| 880 |
+
},
|
| 881 |
+
{
|
| 882 |
+
"epoch": 0.37,
|
| 883 |
+
"grad_norm": 39.0,
|
| 884 |
+
"learning_rate": 3.497150997150997e-07,
|
| 885 |
+
"logits/generated": 0.29128846526145935,
|
| 886 |
+
"logits/real": -0.29488933086395264,
|
| 887 |
+
"logps/generated": -569.3885498046875,
|
| 888 |
+
"logps/real": -493.6559143066406,
|
| 889 |
+
"loss": 0.3907,
|
| 890 |
+
"rewards/accuracies": 0.9437500238418579,
|
| 891 |
+
"rewards/generated": -0.8542326092720032,
|
| 892 |
+
"rewards/margins": 0.8633776903152466,
|
| 893 |
+
"rewards/real": 0.009145094081759453,
|
| 894 |
+
"step": 290
|
| 895 |
+
},
|
| 896 |
+
{
|
| 897 |
+
"epoch": 0.38,
|
| 898 |
+
"grad_norm": 40.0,
|
| 899 |
+
"learning_rate": 3.461538461538461e-07,
|
| 900 |
+
"logits/generated": 0.2510134279727936,
|
| 901 |
+
"logits/real": -0.23930807411670685,
|
| 902 |
+
"logps/generated": -571.7144775390625,
|
| 903 |
+
"logps/real": -420.68890380859375,
|
| 904 |
+
"loss": 0.3903,
|
| 905 |
+
"rewards/accuracies": 0.9437500238418579,
|
| 906 |
+
"rewards/generated": -0.8466684222221375,
|
| 907 |
+
"rewards/margins": 0.8632675409317017,
|
| 908 |
+
"rewards/real": 0.016599006950855255,
|
| 909 |
+
"step": 295
|
| 910 |
+
},
|
| 911 |
+
{
|
| 912 |
+
"epoch": 0.38,
|
| 913 |
+
"grad_norm": 34.5,
|
| 914 |
+
"learning_rate": 3.425925925925926e-07,
|
| 915 |
+
"logits/generated": 0.24713429808616638,
|
| 916 |
+
"logits/real": -0.2522990107536316,
|
| 917 |
+
"logps/generated": -579.431396484375,
|
| 918 |
+
"logps/real": -446.91668701171875,
|
| 919 |
+
"loss": 0.4016,
|
| 920 |
+
"rewards/accuracies": 0.9125000238418579,
|
| 921 |
+
"rewards/generated": -0.852130115032196,
|
| 922 |
+
"rewards/margins": 0.845871090888977,
|
| 923 |
+
"rewards/real": -0.006259086541831493,
|
| 924 |
+
"step": 300
|
| 925 |
+
},
|
| 926 |
+
{
|
| 927 |
+
"epoch": 0.38,
|
| 928 |
+
"eval_logits/generated": 0.25726401805877686,
|
| 929 |
+
"eval_logits/real": -0.2516762316226959,
|
| 930 |
+
"eval_logps/generated": -568.6871948242188,
|
| 931 |
+
"eval_logps/real": -468.50299072265625,
|
| 932 |
+
"eval_loss": 0.3913811147212982,
|
| 933 |
+
"eval_rewards/accuracies": 0.9319999814033508,
|
| 934 |
+
"eval_rewards/generated": -0.8142948746681213,
|
| 935 |
+
"eval_rewards/margins": 0.8413105010986328,
|
| 936 |
+
"eval_rewards/real": 0.027015656232833862,
|
| 937 |
+
"eval_runtime": 262.9656,
|
| 938 |
+
"eval_samples_per_second": 7.606,
|
| 939 |
+
"eval_steps_per_second": 0.951,
|
| 940 |
+
"step": 300
|
| 941 |
+
},
|
| 942 |
+
{
|
| 943 |
+
"epoch": 0.39,
|
| 944 |
+
"grad_norm": 34.0,
|
| 945 |
+
"learning_rate": 3.3903133903133903e-07,
|
| 946 |
+
"logits/generated": 0.2464553862810135,
|
| 947 |
+
"logits/real": -0.24754126369953156,
|
| 948 |
+
"logps/generated": -557.3573608398438,
|
| 949 |
+
"logps/real": -447.5314025878906,
|
| 950 |
+
"loss": 0.3659,
|
| 951 |
+
"rewards/accuracies": 0.9624999761581421,
|
| 952 |
+
"rewards/generated": -0.8741347193717957,
|
| 953 |
+
"rewards/margins": 0.9578303098678589,
|
| 954 |
+
"rewards/real": 0.08369561284780502,
|
| 955 |
+
"step": 305
|
| 956 |
+
},
|
| 957 |
+
{
|
| 958 |
+
"epoch": 0.4,
|
| 959 |
+
"grad_norm": 36.0,
|
| 960 |
+
"learning_rate": 3.354700854700854e-07,
|
| 961 |
+
"logits/generated": 0.2558444142341614,
|
| 962 |
+
"logits/real": -0.2749106287956238,
|
| 963 |
+
"logps/generated": -581.9246826171875,
|
| 964 |
+
"logps/real": -470.64117431640625,
|
| 965 |
+
"loss": 0.3883,
|
| 966 |
+
"rewards/accuracies": 0.918749988079071,
|
| 967 |
+
"rewards/generated": -0.870457649230957,
|
| 968 |
+
"rewards/margins": 0.9120520353317261,
|
| 969 |
+
"rewards/real": 0.04159443825483322,
|
| 970 |
+
"step": 310
|
| 971 |
+
},
|
| 972 |
+
{
|
| 973 |
+
"epoch": 0.4,
|
| 974 |
+
"grad_norm": 40.0,
|
| 975 |
+
"learning_rate": 3.319088319088319e-07,
|
| 976 |
+
"logits/generated": 0.2201302945613861,
|
| 977 |
+
"logits/real": -0.2533838152885437,
|
| 978 |
+
"logps/generated": -551.3731689453125,
|
| 979 |
+
"logps/real": -479.00115966796875,
|
| 980 |
+
"loss": 0.3958,
|
| 981 |
+
"rewards/accuracies": 0.956250011920929,
|
| 982 |
+
"rewards/generated": -0.7957965135574341,
|
| 983 |
+
"rewards/margins": 0.8228376507759094,
|
| 984 |
+
"rewards/real": 0.027041062712669373,
|
| 985 |
+
"step": 315
|
| 986 |
+
},
|
| 987 |
+
{
|
| 988 |
+
"epoch": 0.41,
|
| 989 |
+
"grad_norm": 38.75,
|
| 990 |
+
"learning_rate": 3.2834757834757833e-07,
|
| 991 |
+
"logits/generated": 0.2794730067253113,
|
| 992 |
+
"logits/real": -0.1945553719997406,
|
| 993 |
+
"logps/generated": -570.4600830078125,
|
| 994 |
+
"logps/real": -423.16229248046875,
|
| 995 |
+
"loss": 0.3755,
|
| 996 |
+
"rewards/accuracies": 0.9375,
|
| 997 |
+
"rewards/generated": -0.8808556795120239,
|
| 998 |
+
"rewards/margins": 0.8986592292785645,
|
| 999 |
+
"rewards/real": 0.01780361495912075,
|
| 1000 |
+
"step": 320
|
| 1001 |
+
},
|
| 1002 |
+
{
|
| 1003 |
+
"epoch": 0.42,
|
| 1004 |
+
"grad_norm": 36.25,
|
| 1005 |
+
"learning_rate": 3.2478632478632476e-07,
|
| 1006 |
+
"logits/generated": 0.28926241397857666,
|
| 1007 |
+
"logits/real": -0.2784452438354492,
|
| 1008 |
+
"logps/generated": -584.1411743164062,
|
| 1009 |
+
"logps/real": -480.142578125,
|
| 1010 |
+
"loss": 0.3778,
|
| 1011 |
+
"rewards/accuracies": 0.9437500238418579,
|
| 1012 |
+
"rewards/generated": -0.8571932911872864,
|
| 1013 |
+
"rewards/margins": 0.9160116910934448,
|
| 1014 |
+
"rewards/real": 0.058818407356739044,
|
| 1015 |
+
"step": 325
|
| 1016 |
+
},
|
| 1017 |
+
{
|
| 1018 |
+
"epoch": 0.42,
|
| 1019 |
+
"grad_norm": 34.75,
|
| 1020 |
+
"learning_rate": 3.212250712250712e-07,
|
| 1021 |
+
"logits/generated": 0.25457730889320374,
|
| 1022 |
+
"logits/real": -0.2175012081861496,
|
| 1023 |
+
"logps/generated": -571.3527221679688,
|
| 1024 |
+
"logps/real": -437.7066345214844,
|
| 1025 |
+
"loss": 0.3903,
|
| 1026 |
+
"rewards/accuracies": 0.9125000238418579,
|
| 1027 |
+
"rewards/generated": -0.8724948167800903,
|
| 1028 |
+
"rewards/margins": 0.9130793809890747,
|
| 1029 |
+
"rewards/real": 0.04058459773659706,
|
| 1030 |
+
"step": 330
|
| 1031 |
+
},
|
| 1032 |
+
{
|
| 1033 |
+
"epoch": 0.43,
|
| 1034 |
+
"grad_norm": 35.5,
|
| 1035 |
+
"learning_rate": 3.176638176638177e-07,
|
| 1036 |
+
"logits/generated": 0.2662949562072754,
|
| 1037 |
+
"logits/real": -0.2381666898727417,
|
| 1038 |
+
"logps/generated": -601.4517211914062,
|
| 1039 |
+
"logps/real": -483.5282287597656,
|
| 1040 |
+
"loss": 0.3865,
|
| 1041 |
+
"rewards/accuracies": 0.9375,
|
| 1042 |
+
"rewards/generated": -0.8890978693962097,
|
| 1043 |
+
"rewards/margins": 0.8633550405502319,
|
| 1044 |
+
"rewards/real": -0.02574281021952629,
|
| 1045 |
+
"step": 335
|
| 1046 |
+
},
|
| 1047 |
+
{
|
| 1048 |
+
"epoch": 0.44,
|
| 1049 |
+
"grad_norm": 35.25,
|
| 1050 |
+
"learning_rate": 3.1410256410256407e-07,
|
| 1051 |
+
"logits/generated": 0.25792551040649414,
|
| 1052 |
+
"logits/real": -0.28233546018600464,
|
| 1053 |
+
"logps/generated": -575.9357299804688,
|
| 1054 |
+
"logps/real": -461.888916015625,
|
| 1055 |
+
"loss": 0.3606,
|
| 1056 |
+
"rewards/accuracies": 0.956250011920929,
|
| 1057 |
+
"rewards/generated": -0.9487969279289246,
|
| 1058 |
+
"rewards/margins": 1.0083829164505005,
|
| 1059 |
+
"rewards/real": 0.05958594009280205,
|
| 1060 |
+
"step": 340
|
| 1061 |
+
},
|
| 1062 |
+
{
|
| 1063 |
+
"epoch": 0.44,
|
| 1064 |
+
"grad_norm": 35.5,
|
| 1065 |
+
"learning_rate": 3.1054131054131055e-07,
|
| 1066 |
+
"logits/generated": 0.24235877394676208,
|
| 1067 |
+
"logits/real": -0.19490960240364075,
|
| 1068 |
+
"logps/generated": -574.0829467773438,
|
| 1069 |
+
"logps/real": -401.38031005859375,
|
| 1070 |
+
"loss": 0.3805,
|
| 1071 |
+
"rewards/accuracies": 0.949999988079071,
|
| 1072 |
+
"rewards/generated": -0.8425151705741882,
|
| 1073 |
+
"rewards/margins": 0.8939135670661926,
|
| 1074 |
+
"rewards/real": 0.0513983853161335,
|
| 1075 |
+
"step": 345
|
| 1076 |
+
},
|
| 1077 |
+
{
|
| 1078 |
+
"epoch": 0.45,
|
| 1079 |
+
"grad_norm": 33.25,
|
| 1080 |
+
"learning_rate": 3.06980056980057e-07,
|
| 1081 |
+
"logits/generated": 0.2665616571903229,
|
| 1082 |
+
"logits/real": -0.24914798140525818,
|
| 1083 |
+
"logps/generated": -547.2540283203125,
|
| 1084 |
+
"logps/real": -449.8600158691406,
|
| 1085 |
+
"loss": 0.3741,
|
| 1086 |
+
"rewards/accuracies": 0.9312499761581421,
|
| 1087 |
+
"rewards/generated": -0.8794205784797668,
|
| 1088 |
+
"rewards/margins": 0.8983514904975891,
|
| 1089 |
+
"rewards/real": 0.01893080770969391,
|
| 1090 |
+
"step": 350
|
| 1091 |
+
},
|
| 1092 |
+
{
|
| 1093 |
+
"epoch": 0.45,
|
| 1094 |
+
"grad_norm": 38.0,
|
| 1095 |
+
"learning_rate": 3.0341880341880337e-07,
|
| 1096 |
+
"logits/generated": 0.2690719962120056,
|
| 1097 |
+
"logits/real": -0.26707571744918823,
|
| 1098 |
+
"logps/generated": -573.3060913085938,
|
| 1099 |
+
"logps/real": -484.48492431640625,
|
| 1100 |
+
"loss": 0.3759,
|
| 1101 |
+
"rewards/accuracies": 0.925000011920929,
|
| 1102 |
+
"rewards/generated": -0.9073358774185181,
|
| 1103 |
+
"rewards/margins": 0.927741527557373,
|
| 1104 |
+
"rewards/real": 0.020405706018209457,
|
| 1105 |
+
"step": 355
|
| 1106 |
+
},
|
| 1107 |
+
{
|
| 1108 |
+
"epoch": 0.46,
|
| 1109 |
+
"grad_norm": 35.0,
|
| 1110 |
+
"learning_rate": 2.9985754985754986e-07,
|
| 1111 |
+
"logits/generated": 0.23270320892333984,
|
| 1112 |
+
"logits/real": -0.2126883715391159,
|
| 1113 |
+
"logps/generated": -566.2892456054688,
|
| 1114 |
+
"logps/real": -420.14471435546875,
|
| 1115 |
+
"loss": 0.373,
|
| 1116 |
+
"rewards/accuracies": 0.9437500238418579,
|
| 1117 |
+
"rewards/generated": -0.9121645092964172,
|
| 1118 |
+
"rewards/margins": 0.9226058721542358,
|
| 1119 |
+
"rewards/real": 0.01044147927314043,
|
| 1120 |
+
"step": 360
|
| 1121 |
+
},
|
| 1122 |
+
{
|
| 1123 |
+
"epoch": 0.47,
|
| 1124 |
+
"grad_norm": 33.0,
|
| 1125 |
+
"learning_rate": 2.962962962962963e-07,
|
| 1126 |
+
"logits/generated": 0.24159935116767883,
|
| 1127 |
+
"logits/real": -0.26122623682022095,
|
| 1128 |
+
"logps/generated": -570.7769775390625,
|
| 1129 |
+
"logps/real": -483.9549865722656,
|
| 1130 |
+
"loss": 0.3691,
|
| 1131 |
+
"rewards/accuracies": 0.918749988079071,
|
| 1132 |
+
"rewards/generated": -0.9642108082771301,
|
| 1133 |
+
"rewards/margins": 0.9902675747871399,
|
| 1134 |
+
"rewards/real": 0.0260569266974926,
|
| 1135 |
+
"step": 365
|
| 1136 |
+
},
|
| 1137 |
+
{
|
| 1138 |
+
"epoch": 0.47,
|
| 1139 |
+
"grad_norm": 36.0,
|
| 1140 |
+
"learning_rate": 2.927350427350427e-07,
|
| 1141 |
+
"logits/generated": 0.23782770335674286,
|
| 1142 |
+
"logits/real": -0.2670837640762329,
|
| 1143 |
+
"logps/generated": -573.4016723632812,
|
| 1144 |
+
"logps/real": -469.85491943359375,
|
| 1145 |
+
"loss": 0.3758,
|
| 1146 |
+
"rewards/accuracies": 0.9312499761581421,
|
| 1147 |
+
"rewards/generated": -0.9776171445846558,
|
| 1148 |
+
"rewards/margins": 0.9925469160079956,
|
| 1149 |
+
"rewards/real": 0.01492965966463089,
|
| 1150 |
+
"step": 370
|
| 1151 |
+
},
|
| 1152 |
+
{
|
| 1153 |
+
"epoch": 0.48,
|
| 1154 |
+
"grad_norm": 34.0,
|
| 1155 |
+
"learning_rate": 2.8917378917378916e-07,
|
| 1156 |
+
"logits/generated": 0.2320900857448578,
|
| 1157 |
+
"logits/real": -0.18077997863292694,
|
| 1158 |
+
"logps/generated": -559.6778564453125,
|
| 1159 |
+
"logps/real": -410.2744140625,
|
| 1160 |
+
"loss": 0.3682,
|
| 1161 |
+
"rewards/accuracies": 0.949999988079071,
|
| 1162 |
+
"rewards/generated": -0.9255533218383789,
|
| 1163 |
+
"rewards/margins": 0.9920336008071899,
|
| 1164 |
+
"rewards/real": 0.06648023426532745,
|
| 1165 |
+
"step": 375
|
| 1166 |
+
},
|
| 1167 |
+
{
|
| 1168 |
+
"epoch": 0.49,
|
| 1169 |
+
"grad_norm": 37.25,
|
| 1170 |
+
"learning_rate": 2.8561253561253565e-07,
|
| 1171 |
+
"logits/generated": 0.21389861404895782,
|
| 1172 |
+
"logits/real": -0.26197221875190735,
|
| 1173 |
+
"logps/generated": -571.7706909179688,
|
| 1174 |
+
"logps/real": -451.16253662109375,
|
| 1175 |
+
"loss": 0.3799,
|
| 1176 |
+
"rewards/accuracies": 0.90625,
|
| 1177 |
+
"rewards/generated": -0.9532166719436646,
|
| 1178 |
+
"rewards/margins": 0.9206304550170898,
|
| 1179 |
+
"rewards/real": -0.03258631005883217,
|
| 1180 |
+
"step": 380
|
| 1181 |
+
},
|
| 1182 |
+
{
|
| 1183 |
+
"epoch": 0.49,
|
| 1184 |
+
"grad_norm": 36.0,
|
| 1185 |
+
"learning_rate": 2.8205128205128203e-07,
|
| 1186 |
+
"logits/generated": 0.22054708003997803,
|
| 1187 |
+
"logits/real": -0.25541600584983826,
|
| 1188 |
+
"logps/generated": -549.7349853515625,
|
| 1189 |
+
"logps/real": -407.8997497558594,
|
| 1190 |
+
"loss": 0.3712,
|
| 1191 |
+
"rewards/accuracies": 0.9312499761581421,
|
| 1192 |
+
"rewards/generated": -0.875189483165741,
|
| 1193 |
+
"rewards/margins": 0.9158505201339722,
|
| 1194 |
+
"rewards/real": 0.04066096618771553,
|
| 1195 |
+
"step": 385
|
| 1196 |
+
},
|
| 1197 |
+
{
|
| 1198 |
+
"epoch": 0.5,
|
| 1199 |
+
"grad_norm": 35.25,
|
| 1200 |
+
"learning_rate": 2.7849002849002846e-07,
|
| 1201 |
+
"logits/generated": 0.24272923171520233,
|
| 1202 |
+
"logits/real": -0.2981414198875427,
|
| 1203 |
+
"logps/generated": -567.8425903320312,
|
| 1204 |
+
"logps/real": -499.34588623046875,
|
| 1205 |
+
"loss": 0.3782,
|
| 1206 |
+
"rewards/accuracies": 0.918749988079071,
|
| 1207 |
+
"rewards/generated": -0.8826544880867004,
|
| 1208 |
+
"rewards/margins": 0.8991689682006836,
|
| 1209 |
+
"rewards/real": 0.01651454158127308,
|
| 1210 |
+
"step": 390
|
| 1211 |
+
},
|
| 1212 |
+
{
|
| 1213 |
+
"epoch": 0.51,
|
| 1214 |
+
"grad_norm": 36.25,
|
| 1215 |
+
"learning_rate": 2.749287749287749e-07,
|
| 1216 |
+
"logits/generated": 0.25423914194107056,
|
| 1217 |
+
"logits/real": -0.27515119314193726,
|
| 1218 |
+
"logps/generated": -561.5721435546875,
|
| 1219 |
+
"logps/real": -458.6827087402344,
|
| 1220 |
+
"loss": 0.3479,
|
| 1221 |
+
"rewards/accuracies": 0.9624999761581421,
|
| 1222 |
+
"rewards/generated": -0.976852536201477,
|
| 1223 |
+
"rewards/margins": 1.0118058919906616,
|
| 1224 |
+
"rewards/real": 0.034953318536281586,
|
| 1225 |
+
"step": 395
|
| 1226 |
+
},
|
| 1227 |
+
{
|
| 1228 |
+
"epoch": 0.51,
|
| 1229 |
+
"grad_norm": 31.125,
|
| 1230 |
+
"learning_rate": 2.7136752136752133e-07,
|
| 1231 |
+
"logits/generated": 0.23657219111919403,
|
| 1232 |
+
"logits/real": -0.28398722410202026,
|
| 1233 |
+
"logps/generated": -593.6157836914062,
|
| 1234 |
+
"logps/real": -420.7015686035156,
|
| 1235 |
+
"loss": 0.3569,
|
| 1236 |
+
"rewards/accuracies": 0.9375,
|
| 1237 |
+
"rewards/generated": -0.9960700869560242,
|
| 1238 |
+
"rewards/margins": 1.0824501514434814,
|
| 1239 |
+
"rewards/real": 0.08638016134500504,
|
| 1240 |
+
"step": 400
|
| 1241 |
+
},
|
| 1242 |
+
{
|
| 1243 |
+
"epoch": 0.52,
|
| 1244 |
+
"grad_norm": 34.75,
|
| 1245 |
+
"learning_rate": 2.678062678062678e-07,
|
| 1246 |
+
"logits/generated": 0.24148711562156677,
|
| 1247 |
+
"logits/real": -0.21437129378318787,
|
| 1248 |
+
"logps/generated": -577.3068237304688,
|
| 1249 |
+
"logps/real": -450.6416015625,
|
| 1250 |
+
"loss": 0.3691,
|
| 1251 |
+
"rewards/accuracies": 0.9312499761581421,
|
| 1252 |
+
"rewards/generated": -0.9539799690246582,
|
| 1253 |
+
"rewards/margins": 1.0042264461517334,
|
| 1254 |
+
"rewards/real": 0.05024648830294609,
|
| 1255 |
+
"step": 405
|
| 1256 |
+
},
|
| 1257 |
+
{
|
| 1258 |
+
"epoch": 0.52,
|
| 1259 |
+
"grad_norm": 34.0,
|
| 1260 |
+
"learning_rate": 2.642450142450142e-07,
|
| 1261 |
+
"logits/generated": 0.27527815103530884,
|
| 1262 |
+
"logits/real": -0.2802174687385559,
|
| 1263 |
+
"logps/generated": -584.6751708984375,
|
| 1264 |
+
"logps/real": -474.64678955078125,
|
| 1265 |
+
"loss": 0.3432,
|
| 1266 |
+
"rewards/accuracies": 0.9437500238418579,
|
| 1267 |
+
"rewards/generated": -1.0197203159332275,
|
| 1268 |
+
"rewards/margins": 1.0799630880355835,
|
| 1269 |
+
"rewards/real": 0.06024279445409775,
|
| 1270 |
+
"step": 410
|
| 1271 |
+
},
|
| 1272 |
+
{
|
| 1273 |
+
"epoch": 0.53,
|
| 1274 |
+
"grad_norm": 34.25,
|
| 1275 |
+
"learning_rate": 2.606837606837607e-07,
|
| 1276 |
+
"logits/generated": 0.24102333188056946,
|
| 1277 |
+
"logits/real": -0.2790641188621521,
|
| 1278 |
+
"logps/generated": -580.9298706054688,
|
| 1279 |
+
"logps/real": -423.9723205566406,
|
| 1280 |
+
"loss": 0.3673,
|
| 1281 |
+
"rewards/accuracies": 0.9125000238418579,
|
| 1282 |
+
"rewards/generated": -0.9787546396255493,
|
| 1283 |
+
"rewards/margins": 1.0065668821334839,
|
| 1284 |
+
"rewards/real": 0.027812320739030838,
|
| 1285 |
+
"step": 415
|
| 1286 |
+
},
|
| 1287 |
+
{
|
| 1288 |
+
"epoch": 0.54,
|
| 1289 |
+
"grad_norm": 34.75,
|
| 1290 |
+
"learning_rate": 2.571225071225071e-07,
|
| 1291 |
+
"logits/generated": 0.2540944218635559,
|
| 1292 |
+
"logits/real": -0.19810980558395386,
|
| 1293 |
+
"logps/generated": -562.8670654296875,
|
| 1294 |
+
"logps/real": -426.3075256347656,
|
| 1295 |
+
"loss": 0.35,
|
| 1296 |
+
"rewards/accuracies": 0.949999988079071,
|
| 1297 |
+
"rewards/generated": -1.0469673871994019,
|
| 1298 |
+
"rewards/margins": 1.0949853658676147,
|
| 1299 |
+
"rewards/real": 0.04801792651414871,
|
| 1300 |
+
"step": 420
|
| 1301 |
+
},
|
| 1302 |
+
{
|
| 1303 |
+
"epoch": 0.54,
|
| 1304 |
+
"grad_norm": 33.0,
|
| 1305 |
+
"learning_rate": 2.5356125356125355e-07,
|
| 1306 |
+
"logits/generated": 0.20098896324634552,
|
| 1307 |
+
"logits/real": -0.24582453072071075,
|
| 1308 |
+
"logps/generated": -573.8992309570312,
|
| 1309 |
+
"logps/real": -420.2386779785156,
|
| 1310 |
+
"loss": 0.3456,
|
| 1311 |
+
"rewards/accuracies": 0.9437500238418579,
|
| 1312 |
+
"rewards/generated": -1.0094668865203857,
|
| 1313 |
+
"rewards/margins": 1.0302053689956665,
|
| 1314 |
+
"rewards/real": 0.020738471299409866,
|
| 1315 |
+
"step": 425
|
| 1316 |
+
},
|
| 1317 |
+
{
|
| 1318 |
+
"epoch": 0.55,
|
| 1319 |
+
"grad_norm": 36.25,
|
| 1320 |
+
"learning_rate": 2.5e-07,
|
| 1321 |
+
"logits/generated": 0.2519596219062805,
|
| 1322 |
+
"logits/real": -0.27877140045166016,
|
| 1323 |
+
"logps/generated": -565.7482299804688,
|
| 1324 |
+
"logps/real": -534.8162841796875,
|
| 1325 |
+
"loss": 0.3519,
|
| 1326 |
+
"rewards/accuracies": 0.9125000238418579,
|
| 1327 |
+
"rewards/generated": -0.9428899884223938,
|
| 1328 |
+
"rewards/margins": 0.976216197013855,
|
| 1329 |
+
"rewards/real": 0.033326275646686554,
|
| 1330 |
+
"step": 430
|
| 1331 |
+
},
|
| 1332 |
+
{
|
| 1333 |
+
"epoch": 0.56,
|
| 1334 |
+
"grad_norm": 36.75,
|
| 1335 |
+
"learning_rate": 2.464387464387464e-07,
|
| 1336 |
+
"logits/generated": 0.23958304524421692,
|
| 1337 |
+
"logits/real": -0.3024999499320984,
|
| 1338 |
+
"logps/generated": -558.3704223632812,
|
| 1339 |
+
"logps/real": -470.7825622558594,
|
| 1340 |
+
"loss": 0.3492,
|
| 1341 |
+
"rewards/accuracies": 0.9437500238418579,
|
| 1342 |
+
"rewards/generated": -1.0124276876449585,
|
| 1343 |
+
"rewards/margins": 1.0718011856079102,
|
| 1344 |
+
"rewards/real": 0.059373389929533005,
|
| 1345 |
+
"step": 435
|
| 1346 |
+
},
|
| 1347 |
+
{
|
| 1348 |
+
"epoch": 0.56,
|
| 1349 |
+
"grad_norm": 30.75,
|
| 1350 |
+
"learning_rate": 2.4287749287749286e-07,
|
| 1351 |
+
"logits/generated": 0.2602802515029907,
|
| 1352 |
+
"logits/real": -0.28142985701560974,
|
| 1353 |
+
"logps/generated": -619.5274047851562,
|
| 1354 |
+
"logps/real": -502.71832275390625,
|
| 1355 |
+
"loss": 0.3466,
|
| 1356 |
+
"rewards/accuracies": 0.9750000238418579,
|
| 1357 |
+
"rewards/generated": -1.0051062107086182,
|
| 1358 |
+
"rewards/margins": 1.0426944494247437,
|
| 1359 |
+
"rewards/real": 0.037588391453027725,
|
| 1360 |
+
"step": 440
|
| 1361 |
+
},
|
| 1362 |
+
{
|
| 1363 |
+
"epoch": 0.57,
|
| 1364 |
+
"grad_norm": 33.25,
|
| 1365 |
+
"learning_rate": 2.393162393162393e-07,
|
| 1366 |
+
"logits/generated": 0.21535196900367737,
|
| 1367 |
+
"logits/real": -0.2805056869983673,
|
| 1368 |
+
"logps/generated": -587.7852172851562,
|
| 1369 |
+
"logps/real": -429.0276794433594,
|
| 1370 |
+
"loss": 0.3489,
|
| 1371 |
+
"rewards/accuracies": 0.949999988079071,
|
| 1372 |
+
"rewards/generated": -0.9763972163200378,
|
| 1373 |
+
"rewards/margins": 0.9719440340995789,
|
| 1374 |
+
"rewards/real": -0.004453244619071484,
|
| 1375 |
+
"step": 445
|
| 1376 |
+
},
|
| 1377 |
+
{
|
| 1378 |
+
"epoch": 0.58,
|
| 1379 |
+
"grad_norm": 33.75,
|
| 1380 |
+
"learning_rate": 2.3575498575498575e-07,
|
| 1381 |
+
"logits/generated": 0.22773954272270203,
|
| 1382 |
+
"logits/real": -0.2382933795452118,
|
| 1383 |
+
"logps/generated": -591.6124877929688,
|
| 1384 |
+
"logps/real": -456.0254821777344,
|
| 1385 |
+
"loss": 0.3504,
|
| 1386 |
+
"rewards/accuracies": 0.956250011920929,
|
| 1387 |
+
"rewards/generated": -1.0408380031585693,
|
| 1388 |
+
"rewards/margins": 1.030290961265564,
|
| 1389 |
+
"rewards/real": -0.010547079145908356,
|
| 1390 |
+
"step": 450
|
| 1391 |
+
},
|
| 1392 |
+
{
|
| 1393 |
+
"epoch": 0.58,
|
| 1394 |
+
"grad_norm": 35.25,
|
| 1395 |
+
"learning_rate": 2.3219373219373216e-07,
|
| 1396 |
+
"logits/generated": 0.22827064990997314,
|
| 1397 |
+
"logits/real": -0.3192758560180664,
|
| 1398 |
+
"logps/generated": -563.3525390625,
|
| 1399 |
+
"logps/real": -482.61541748046875,
|
| 1400 |
+
"loss": 0.3454,
|
| 1401 |
+
"rewards/accuracies": 0.9312499761581421,
|
| 1402 |
+
"rewards/generated": -0.9959697723388672,
|
| 1403 |
+
"rewards/margins": 1.0202291011810303,
|
| 1404 |
+
"rewards/real": 0.024259375408291817,
|
| 1405 |
+
"step": 455
|
| 1406 |
+
},
|
| 1407 |
+
{
|
| 1408 |
+
"epoch": 0.59,
|
| 1409 |
+
"grad_norm": 36.25,
|
| 1410 |
+
"learning_rate": 2.2863247863247862e-07,
|
| 1411 |
+
"logits/generated": 0.2440454512834549,
|
| 1412 |
+
"logits/real": -0.22086623311042786,
|
| 1413 |
+
"logps/generated": -557.6519165039062,
|
| 1414 |
+
"logps/real": -446.55230712890625,
|
| 1415 |
+
"loss": 0.3376,
|
| 1416 |
+
"rewards/accuracies": 0.9624999761581421,
|
| 1417 |
+
"rewards/generated": -1.0488998889923096,
|
| 1418 |
+
"rewards/margins": 1.0662249326705933,
|
| 1419 |
+
"rewards/real": 0.017325039952993393,
|
| 1420 |
+
"step": 460
|
| 1421 |
+
},
|
| 1422 |
+
{
|
| 1423 |
+
"epoch": 0.6,
|
| 1424 |
+
"grad_norm": 34.5,
|
| 1425 |
+
"learning_rate": 2.2507122507122505e-07,
|
| 1426 |
+
"logits/generated": 0.21471574902534485,
|
| 1427 |
+
"logits/real": -0.2694615423679352,
|
| 1428 |
+
"logps/generated": -556.4073486328125,
|
| 1429 |
+
"logps/real": -466.55120849609375,
|
| 1430 |
+
"loss": 0.3528,
|
| 1431 |
+
"rewards/accuracies": 0.90625,
|
| 1432 |
+
"rewards/generated": -1.063711404800415,
|
| 1433 |
+
"rewards/margins": 1.067757487297058,
|
| 1434 |
+
"rewards/real": 0.004046155605465174,
|
| 1435 |
+
"step": 465
|
| 1436 |
+
},
|
| 1437 |
+
{
|
| 1438 |
+
"epoch": 0.6,
|
| 1439 |
+
"grad_norm": 35.75,
|
| 1440 |
+
"learning_rate": 2.215099715099715e-07,
|
| 1441 |
+
"logits/generated": 0.223112553358078,
|
| 1442 |
+
"logits/real": -0.2675575315952301,
|
| 1443 |
+
"logps/generated": -554.9754028320312,
|
| 1444 |
+
"logps/real": -409.7915954589844,
|
| 1445 |
+
"loss": 0.3554,
|
| 1446 |
+
"rewards/accuracies": 0.9312499761581421,
|
| 1447 |
+
"rewards/generated": -0.9735461473464966,
|
| 1448 |
+
"rewards/margins": 0.9401170015335083,
|
| 1449 |
+
"rewards/real": -0.03342916816473007,
|
| 1450 |
+
"step": 470
|
| 1451 |
+
},
|
| 1452 |
+
{
|
| 1453 |
+
"epoch": 0.61,
|
| 1454 |
+
"grad_norm": 32.75,
|
| 1455 |
+
"learning_rate": 2.1794871794871795e-07,
|
| 1456 |
+
"logits/generated": 0.24918599426746368,
|
| 1457 |
+
"logits/real": -0.28682878613471985,
|
| 1458 |
+
"logps/generated": -571.9830322265625,
|
| 1459 |
+
"logps/real": -467.03125,
|
| 1460 |
+
"loss": 0.3409,
|
| 1461 |
+
"rewards/accuracies": 0.9624999761581421,
|
| 1462 |
+
"rewards/generated": -1.0399566888809204,
|
| 1463 |
+
"rewards/margins": 1.0961642265319824,
|
| 1464 |
+
"rewards/real": 0.056207604706287384,
|
| 1465 |
+
"step": 475
|
| 1466 |
+
},
|
| 1467 |
+
{
|
| 1468 |
+
"epoch": 0.61,
|
| 1469 |
+
"grad_norm": 30.875,
|
| 1470 |
+
"learning_rate": 2.1438746438746438e-07,
|
| 1471 |
+
"logits/generated": 0.2370508462190628,
|
| 1472 |
+
"logits/real": -0.285112202167511,
|
| 1473 |
+
"logps/generated": -598.0405883789062,
|
| 1474 |
+
"logps/real": -468.107421875,
|
| 1475 |
+
"loss": 0.3299,
|
| 1476 |
+
"rewards/accuracies": 0.956250011920929,
|
| 1477 |
+
"rewards/generated": -1.1196238994598389,
|
| 1478 |
+
"rewards/margins": 1.1285395622253418,
|
| 1479 |
+
"rewards/real": 0.008915687911212444,
|
| 1480 |
+
"step": 480
|
| 1481 |
+
},
|
| 1482 |
+
{
|
| 1483 |
+
"epoch": 0.62,
|
| 1484 |
+
"grad_norm": 33.25,
|
| 1485 |
+
"learning_rate": 2.1082621082621082e-07,
|
| 1486 |
+
"logits/generated": 0.21018759906291962,
|
| 1487 |
+
"logits/real": -0.2482105940580368,
|
| 1488 |
+
"logps/generated": -559.3743896484375,
|
| 1489 |
+
"logps/real": -418.2901916503906,
|
| 1490 |
+
"loss": 0.354,
|
| 1491 |
+
"rewards/accuracies": 0.9624999761581421,
|
| 1492 |
+
"rewards/generated": -1.0670310258865356,
|
| 1493 |
+
"rewards/margins": 1.0917372703552246,
|
| 1494 |
+
"rewards/real": 0.02470635250210762,
|
| 1495 |
+
"step": 485
|
| 1496 |
+
},
|
| 1497 |
+
{
|
| 1498 |
+
"epoch": 0.63,
|
| 1499 |
+
"grad_norm": 32.5,
|
| 1500 |
+
"learning_rate": 2.0726495726495728e-07,
|
| 1501 |
+
"logits/generated": 0.2351360023021698,
|
| 1502 |
+
"logits/real": -0.3030211925506592,
|
| 1503 |
+
"logps/generated": -561.6700439453125,
|
| 1504 |
+
"logps/real": -508.479248046875,
|
| 1505 |
+
"loss": 0.3462,
|
| 1506 |
+
"rewards/accuracies": 0.9125000238418579,
|
| 1507 |
+
"rewards/generated": -1.0717344284057617,
|
| 1508 |
+
"rewards/margins": 1.0173569917678833,
|
| 1509 |
+
"rewards/real": -0.05437753349542618,
|
| 1510 |
+
"step": 490
|
| 1511 |
+
},
|
| 1512 |
+
{
|
| 1513 |
+
"epoch": 0.63,
|
| 1514 |
+
"grad_norm": 33.25,
|
| 1515 |
+
"learning_rate": 2.0370370370370369e-07,
|
| 1516 |
+
"logits/generated": 0.21695998311042786,
|
| 1517 |
+
"logits/real": -0.262247234582901,
|
| 1518 |
+
"logps/generated": -560.5780029296875,
|
| 1519 |
+
"logps/real": -447.3353576660156,
|
| 1520 |
+
"loss": 0.346,
|
| 1521 |
+
"rewards/accuracies": 0.949999988079071,
|
| 1522 |
+
"rewards/generated": -1.0143311023712158,
|
| 1523 |
+
"rewards/margins": 1.0173934698104858,
|
| 1524 |
+
"rewards/real": 0.0030622382182627916,
|
| 1525 |
+
"step": 495
|
| 1526 |
+
},
|
| 1527 |
+
{
|
| 1528 |
+
"epoch": 0.64,
|
| 1529 |
+
"grad_norm": 33.5,
|
| 1530 |
+
"learning_rate": 2.0014245014245012e-07,
|
| 1531 |
+
"logits/generated": 0.24932141602039337,
|
| 1532 |
+
"logits/real": -0.28666001558303833,
|
| 1533 |
+
"logps/generated": -548.8845825195312,
|
| 1534 |
+
"logps/real": -473.7200622558594,
|
| 1535 |
+
"loss": 0.359,
|
| 1536 |
+
"rewards/accuracies": 0.925000011920929,
|
| 1537 |
+
"rewards/generated": -0.9833143353462219,
|
| 1538 |
+
"rewards/margins": 0.9953653216362,
|
| 1539 |
+
"rewards/real": 0.012050976976752281,
|
| 1540 |
+
"step": 500
|
| 1541 |
+
},
|
| 1542 |
+
{
|
| 1543 |
+
"epoch": 0.65,
|
| 1544 |
+
"grad_norm": 30.5,
|
| 1545 |
+
"learning_rate": 1.9658119658119656e-07,
|
| 1546 |
+
"logits/generated": 0.2177625447511673,
|
| 1547 |
+
"logits/real": -0.27629366517066956,
|
| 1548 |
+
"logps/generated": -541.3976440429688,
|
| 1549 |
+
"logps/real": -459.02227783203125,
|
| 1550 |
+
"loss": 0.3403,
|
| 1551 |
+
"rewards/accuracies": 0.956250011920929,
|
| 1552 |
+
"rewards/generated": -0.9844304919242859,
|
| 1553 |
+
"rewards/margins": 1.0443334579467773,
|
| 1554 |
+
"rewards/real": 0.0599028654396534,
|
| 1555 |
+
"step": 505
|
| 1556 |
+
},
|
| 1557 |
+
{
|
| 1558 |
+
"epoch": 0.65,
|
| 1559 |
+
"grad_norm": 33.25,
|
| 1560 |
+
"learning_rate": 1.9301994301994302e-07,
|
| 1561 |
+
"logits/generated": 0.22767607867717743,
|
| 1562 |
+
"logits/real": -0.25904038548469543,
|
| 1563 |
+
"logps/generated": -568.2036743164062,
|
| 1564 |
+
"logps/real": -443.6368713378906,
|
| 1565 |
+
"loss": 0.3417,
|
| 1566 |
+
"rewards/accuracies": 0.9437500238418579,
|
| 1567 |
+
"rewards/generated": -1.0040171146392822,
|
| 1568 |
+
"rewards/margins": 1.078296184539795,
|
| 1569 |
+
"rewards/real": 0.07427913695573807,
|
| 1570 |
+
"step": 510
|
| 1571 |
+
},
|
| 1572 |
+
{
|
| 1573 |
+
"epoch": 0.66,
|
| 1574 |
+
"grad_norm": 34.75,
|
| 1575 |
+
"learning_rate": 1.8945868945868945e-07,
|
| 1576 |
+
"logits/generated": 0.2245851755142212,
|
| 1577 |
+
"logits/real": -0.3053443431854248,
|
| 1578 |
+
"logps/generated": -569.513671875,
|
| 1579 |
+
"logps/real": -470.54547119140625,
|
| 1580 |
+
"loss": 0.3494,
|
| 1581 |
+
"rewards/accuracies": 0.9375,
|
| 1582 |
+
"rewards/generated": -1.0781540870666504,
|
| 1583 |
+
"rewards/margins": 1.0763808488845825,
|
| 1584 |
+
"rewards/real": -0.0017731055850163102,
|
| 1585 |
+
"step": 515
|
| 1586 |
+
},
|
| 1587 |
+
{
|
| 1588 |
+
"epoch": 0.67,
|
| 1589 |
+
"grad_norm": 33.25,
|
| 1590 |
+
"learning_rate": 1.8589743589743588e-07,
|
| 1591 |
+
"logits/generated": 0.228353351354599,
|
| 1592 |
+
"logits/real": -0.27608901262283325,
|
| 1593 |
+
"logps/generated": -549.1878662109375,
|
| 1594 |
+
"logps/real": -457.79681396484375,
|
| 1595 |
+
"loss": 0.3254,
|
| 1596 |
+
"rewards/accuracies": 0.9437500238418579,
|
| 1597 |
+
"rewards/generated": -1.0618976354599,
|
| 1598 |
+
"rewards/margins": 1.0646899938583374,
|
| 1599 |
+
"rewards/real": 0.0027922815643250942,
|
| 1600 |
+
"step": 520
|
| 1601 |
+
},
|
| 1602 |
+
{
|
| 1603 |
+
"epoch": 0.67,
|
| 1604 |
+
"grad_norm": 32.5,
|
| 1605 |
+
"learning_rate": 1.8233618233618234e-07,
|
| 1606 |
+
"logits/generated": 0.24573054909706116,
|
| 1607 |
+
"logits/real": -0.24587111175060272,
|
| 1608 |
+
"logps/generated": -527.3595581054688,
|
| 1609 |
+
"logps/real": -461.79351806640625,
|
| 1610 |
+
"loss": 0.3365,
|
| 1611 |
+
"rewards/accuracies": 0.9312499761581421,
|
| 1612 |
+
"rewards/generated": -0.9843952059745789,
|
| 1613 |
+
"rewards/margins": 1.0428409576416016,
|
| 1614 |
+
"rewards/real": 0.05844569206237793,
|
| 1615 |
+
"step": 525
|
| 1616 |
+
},
|
| 1617 |
+
{
|
| 1618 |
+
"epoch": 0.68,
|
| 1619 |
+
"grad_norm": 30.5,
|
| 1620 |
+
"learning_rate": 1.7877492877492878e-07,
|
| 1621 |
+
"logits/generated": 0.2456216812133789,
|
| 1622 |
+
"logits/real": -0.3523867428302765,
|
| 1623 |
+
"logps/generated": -584.2765502929688,
|
| 1624 |
+
"logps/real": -485.2999572753906,
|
| 1625 |
+
"loss": 0.3442,
|
| 1626 |
+
"rewards/accuracies": 0.96875,
|
| 1627 |
+
"rewards/generated": -1.1297935247421265,
|
| 1628 |
+
"rewards/margins": 1.1257268190383911,
|
| 1629 |
+
"rewards/real": -0.0040667857974767685,
|
| 1630 |
+
"step": 530
|
| 1631 |
+
},
|
| 1632 |
+
{
|
| 1633 |
+
"epoch": 0.68,
|
| 1634 |
+
"grad_norm": 32.75,
|
| 1635 |
+
"learning_rate": 1.752136752136752e-07,
|
| 1636 |
+
"logits/generated": 0.24020275473594666,
|
| 1637 |
+
"logits/real": -0.2194860428571701,
|
| 1638 |
+
"logps/generated": -575.4654541015625,
|
| 1639 |
+
"logps/real": -414.576416015625,
|
| 1640 |
+
"loss": 0.3328,
|
| 1641 |
+
"rewards/accuracies": 0.956250011920929,
|
| 1642 |
+
"rewards/generated": -1.1157715320587158,
|
| 1643 |
+
"rewards/margins": 1.0961633920669556,
|
| 1644 |
+
"rewards/real": -0.01960797980427742,
|
| 1645 |
+
"step": 535
|
| 1646 |
+
},
|
| 1647 |
+
{
|
| 1648 |
+
"epoch": 0.69,
|
| 1649 |
+
"grad_norm": 31.75,
|
| 1650 |
+
"learning_rate": 1.7165242165242165e-07,
|
| 1651 |
+
"logits/generated": 0.20955972373485565,
|
| 1652 |
+
"logits/real": -0.25930672883987427,
|
| 1653 |
+
"logps/generated": -591.9100341796875,
|
| 1654 |
+
"logps/real": -403.3531799316406,
|
| 1655 |
+
"loss": 0.3286,
|
| 1656 |
+
"rewards/accuracies": 0.925000011920929,
|
| 1657 |
+
"rewards/generated": -1.076982855796814,
|
| 1658 |
+
"rewards/margins": 1.1210120916366577,
|
| 1659 |
+
"rewards/real": 0.044029366225004196,
|
| 1660 |
+
"step": 540
|
| 1661 |
+
},
|
| 1662 |
+
{
|
| 1663 |
+
"epoch": 0.7,
|
| 1664 |
+
"grad_norm": 31.375,
|
| 1665 |
+
"learning_rate": 1.6809116809116808e-07,
|
| 1666 |
+
"logits/generated": 0.18421868979930878,
|
| 1667 |
+
"logits/real": -0.29251617193222046,
|
| 1668 |
+
"logps/generated": -552.1446533203125,
|
| 1669 |
+
"logps/real": -425.1080627441406,
|
| 1670 |
+
"loss": 0.3362,
|
| 1671 |
+
"rewards/accuracies": 0.925000011920929,
|
| 1672 |
+
"rewards/generated": -0.986895740032196,
|
| 1673 |
+
"rewards/margins": 1.0399030447006226,
|
| 1674 |
+
"rewards/real": 0.05300729721784592,
|
| 1675 |
+
"step": 545
|
| 1676 |
+
},
|
| 1677 |
+
{
|
| 1678 |
+
"epoch": 0.7,
|
| 1679 |
+
"grad_norm": 32.5,
|
| 1680 |
+
"learning_rate": 1.6452991452991452e-07,
|
| 1681 |
+
"logits/generated": 0.22791898250579834,
|
| 1682 |
+
"logits/real": -0.28665608167648315,
|
| 1683 |
+
"logps/generated": -598.5328369140625,
|
| 1684 |
+
"logps/real": -407.1868591308594,
|
| 1685 |
+
"loss": 0.3265,
|
| 1686 |
+
"rewards/accuracies": 0.949999988079071,
|
| 1687 |
+
"rewards/generated": -1.205278754234314,
|
| 1688 |
+
"rewards/margins": 1.2098186016082764,
|
| 1689 |
+
"rewards/real": 0.004539764020591974,
|
| 1690 |
+
"step": 550
|
| 1691 |
+
},
|
| 1692 |
+
{
|
| 1693 |
+
"epoch": 0.71,
|
| 1694 |
+
"grad_norm": 31.625,
|
| 1695 |
+
"learning_rate": 1.6096866096866095e-07,
|
| 1696 |
+
"logits/generated": 0.22963444888591766,
|
| 1697 |
+
"logits/real": -0.24107995629310608,
|
| 1698 |
+
"logps/generated": -566.4512329101562,
|
| 1699 |
+
"logps/real": -450.96978759765625,
|
| 1700 |
+
"loss": 0.3351,
|
| 1701 |
+
"rewards/accuracies": 0.956250011920929,
|
| 1702 |
+
"rewards/generated": -1.081211805343628,
|
| 1703 |
+
"rewards/margins": 1.0905853509902954,
|
| 1704 |
+
"rewards/real": 0.009373527020215988,
|
| 1705 |
+
"step": 555
|
| 1706 |
+
},
|
| 1707 |
+
{
|
| 1708 |
+
"epoch": 0.72,
|
| 1709 |
+
"grad_norm": 34.0,
|
| 1710 |
+
"learning_rate": 1.574074074074074e-07,
|
| 1711 |
+
"logits/generated": 0.24042251706123352,
|
| 1712 |
+
"logits/real": -0.21857628226280212,
|
| 1713 |
+
"logps/generated": -569.6424560546875,
|
| 1714 |
+
"logps/real": -448.65948486328125,
|
| 1715 |
+
"loss": 0.331,
|
| 1716 |
+
"rewards/accuracies": 0.949999988079071,
|
| 1717 |
+
"rewards/generated": -1.0992162227630615,
|
| 1718 |
+
"rewards/margins": 1.157705545425415,
|
| 1719 |
+
"rewards/real": 0.058489274233579636,
|
| 1720 |
+
"step": 560
|
| 1721 |
+
},
|
| 1722 |
+
{
|
| 1723 |
+
"epoch": 0.72,
|
| 1724 |
+
"grad_norm": 33.25,
|
| 1725 |
+
"learning_rate": 1.5384615384615385e-07,
|
| 1726 |
+
"logits/generated": 0.1823839247226715,
|
| 1727 |
+
"logits/real": -0.2623021900653839,
|
| 1728 |
+
"logps/generated": -577.726318359375,
|
| 1729 |
+
"logps/real": -445.66534423828125,
|
| 1730 |
+
"loss": 0.3352,
|
| 1731 |
+
"rewards/accuracies": 0.956250011920929,
|
| 1732 |
+
"rewards/generated": -1.0678300857543945,
|
| 1733 |
+
"rewards/margins": 1.0674610137939453,
|
| 1734 |
+
"rewards/real": -0.000368914392311126,
|
| 1735 |
+
"step": 565
|
| 1736 |
+
},
|
| 1737 |
+
{
|
| 1738 |
+
"epoch": 0.73,
|
| 1739 |
+
"grad_norm": 30.75,
|
| 1740 |
+
"learning_rate": 1.5028490028490028e-07,
|
| 1741 |
+
"logits/generated": 0.2117844521999359,
|
| 1742 |
+
"logits/real": -0.2711123526096344,
|
| 1743 |
+
"logps/generated": -577.0066528320312,
|
| 1744 |
+
"logps/real": -406.86175537109375,
|
| 1745 |
+
"loss": 0.3154,
|
| 1746 |
+
"rewards/accuracies": 0.981249988079071,
|
| 1747 |
+
"rewards/generated": -1.0694202184677124,
|
| 1748 |
+
"rewards/margins": 1.1183080673217773,
|
| 1749 |
+
"rewards/real": 0.04888775572180748,
|
| 1750 |
+
"step": 570
|
| 1751 |
+
},
|
| 1752 |
+
{
|
| 1753 |
+
"epoch": 0.74,
|
| 1754 |
+
"grad_norm": 32.75,
|
| 1755 |
+
"learning_rate": 1.4672364672364671e-07,
|
| 1756 |
+
"logits/generated": 0.2499610185623169,
|
| 1757 |
+
"logits/real": -0.2550189197063446,
|
| 1758 |
+
"logps/generated": -583.8297119140625,
|
| 1759 |
+
"logps/real": -480.6053771972656,
|
| 1760 |
+
"loss": 0.3408,
|
| 1761 |
+
"rewards/accuracies": 0.9437500238418579,
|
| 1762 |
+
"rewards/generated": -1.01523756980896,
|
| 1763 |
+
"rewards/margins": 1.0397374629974365,
|
| 1764 |
+
"rewards/real": 0.024500016123056412,
|
| 1765 |
+
"step": 575
|
| 1766 |
+
},
|
| 1767 |
+
{
|
| 1768 |
+
"epoch": 0.74,
|
| 1769 |
+
"grad_norm": 32.5,
|
| 1770 |
+
"learning_rate": 1.4316239316239315e-07,
|
| 1771 |
+
"logits/generated": 0.1771518439054489,
|
| 1772 |
+
"logits/real": -0.3018808662891388,
|
| 1773 |
+
"logps/generated": -560.4537963867188,
|
| 1774 |
+
"logps/real": -397.22137451171875,
|
| 1775 |
+
"loss": 0.3314,
|
| 1776 |
+
"rewards/accuracies": 0.981249988079071,
|
| 1777 |
+
"rewards/generated": -1.1572097539901733,
|
| 1778 |
+
"rewards/margins": 1.1902085542678833,
|
| 1779 |
+
"rewards/real": 0.03299868851900101,
|
| 1780 |
+
"step": 580
|
| 1781 |
+
},
|
| 1782 |
+
{
|
| 1783 |
+
"epoch": 0.75,
|
| 1784 |
+
"grad_norm": 31.375,
|
| 1785 |
+
"learning_rate": 1.3960113960113958e-07,
|
| 1786 |
+
"logits/generated": 0.192919060587883,
|
| 1787 |
+
"logits/real": -0.31041043996810913,
|
| 1788 |
+
"logps/generated": -532.7413330078125,
|
| 1789 |
+
"logps/real": -462.16693115234375,
|
| 1790 |
+
"loss": 0.3311,
|
| 1791 |
+
"rewards/accuracies": 0.949999988079071,
|
| 1792 |
+
"rewards/generated": -1.1133582592010498,
|
| 1793 |
+
"rewards/margins": 1.1442750692367554,
|
| 1794 |
+
"rewards/real": 0.030916890129446983,
|
| 1795 |
+
"step": 585
|
| 1796 |
+
},
|
| 1797 |
+
{
|
| 1798 |
+
"epoch": 0.76,
|
| 1799 |
+
"grad_norm": 32.75,
|
| 1800 |
+
"learning_rate": 1.3603988603988604e-07,
|
| 1801 |
+
"logits/generated": 0.20842699706554413,
|
| 1802 |
+
"logits/real": -0.30351361632347107,
|
| 1803 |
+
"logps/generated": -582.7373657226562,
|
| 1804 |
+
"logps/real": -480.19073486328125,
|
| 1805 |
+
"loss": 0.3336,
|
| 1806 |
+
"rewards/accuracies": 0.9437500238418579,
|
| 1807 |
+
"rewards/generated": -1.1188085079193115,
|
| 1808 |
+
"rewards/margins": 1.1176929473876953,
|
| 1809 |
+
"rewards/real": -0.0011154465610161424,
|
| 1810 |
+
"step": 590
|
| 1811 |
+
},
|
| 1812 |
+
{
|
| 1813 |
+
"epoch": 0.76,
|
| 1814 |
+
"grad_norm": 31.125,
|
| 1815 |
+
"learning_rate": 1.3247863247863248e-07,
|
| 1816 |
+
"logits/generated": 0.1861368715763092,
|
| 1817 |
+
"logits/real": -0.2930435240268707,
|
| 1818 |
+
"logps/generated": -560.4434814453125,
|
| 1819 |
+
"logps/real": -442.20574951171875,
|
| 1820 |
+
"loss": 0.3037,
|
| 1821 |
+
"rewards/accuracies": 0.9750000238418579,
|
| 1822 |
+
"rewards/generated": -1.1219825744628906,
|
| 1823 |
+
"rewards/margins": 1.1991544961929321,
|
| 1824 |
+
"rewards/real": 0.0771719291806221,
|
| 1825 |
+
"step": 595
|
| 1826 |
+
},
|
| 1827 |
+
{
|
| 1828 |
+
"epoch": 0.77,
|
| 1829 |
+
"grad_norm": 32.5,
|
| 1830 |
+
"learning_rate": 1.289173789173789e-07,
|
| 1831 |
+
"logits/generated": 0.2440781146287918,
|
| 1832 |
+
"logits/real": -0.2519104480743408,
|
| 1833 |
+
"logps/generated": -577.5955810546875,
|
| 1834 |
+
"logps/real": -426.29083251953125,
|
| 1835 |
+
"loss": 0.3451,
|
| 1836 |
+
"rewards/accuracies": 0.956250011920929,
|
| 1837 |
+
"rewards/generated": -1.0716168880462646,
|
| 1838 |
+
"rewards/margins": 1.0959482192993164,
|
| 1839 |
+
"rewards/real": 0.024331340566277504,
|
| 1840 |
+
"step": 600
|
| 1841 |
+
},
|
| 1842 |
+
{
|
| 1843 |
+
"epoch": 0.77,
|
| 1844 |
+
"eval_logits/generated": 0.22350193560123444,
|
| 1845 |
+
"eval_logits/real": -0.28475213050842285,
|
| 1846 |
+
"eval_logps/generated": -571.62451171875,
|
| 1847 |
+
"eval_logps/real": -468.817138671875,
|
| 1848 |
+
"eval_loss": 0.3355465829372406,
|
| 1849 |
+
"eval_rewards/accuracies": 0.9599999785423279,
|
| 1850 |
+
"eval_rewards/generated": -1.108030080795288,
|
| 1851 |
+
"eval_rewards/margins": 1.1036338806152344,
|
| 1852 |
+
"eval_rewards/real": -0.004396026488393545,
|
| 1853 |
+
"eval_runtime": 262.6885,
|
| 1854 |
+
"eval_samples_per_second": 7.614,
|
| 1855 |
+
"eval_steps_per_second": 0.952,
|
| 1856 |
+
"step": 600
|
| 1857 |
+
},
|
| 1858 |
+
{
|
| 1859 |
+
"epoch": 0.77,
|
| 1860 |
+
"grad_norm": 31.25,
|
| 1861 |
+
"learning_rate": 1.2535612535612535e-07,
|
| 1862 |
+
"logits/generated": 0.26258888840675354,
|
| 1863 |
+
"logits/real": -0.27824801206588745,
|
| 1864 |
+
"logps/generated": -587.973876953125,
|
| 1865 |
+
"logps/real": -470.12603759765625,
|
| 1866 |
+
"loss": 0.3066,
|
| 1867 |
+
"rewards/accuracies": 0.9624999761581421,
|
| 1868 |
+
"rewards/generated": -1.2446457147598267,
|
| 1869 |
+
"rewards/margins": 1.2537733316421509,
|
| 1870 |
+
"rewards/real": 0.009127420373260975,
|
| 1871 |
+
"step": 605
|
| 1872 |
+
},
|
| 1873 |
+
{
|
| 1874 |
+
"epoch": 0.78,
|
| 1875 |
+
"grad_norm": 34.0,
|
| 1876 |
+
"learning_rate": 1.2179487179487178e-07,
|
| 1877 |
+
"logits/generated": 0.22428826987743378,
|
| 1878 |
+
"logits/real": -0.24908749759197235,
|
| 1879 |
+
"logps/generated": -574.6671752929688,
|
| 1880 |
+
"logps/real": -423.6082458496094,
|
| 1881 |
+
"loss": 0.3503,
|
| 1882 |
+
"rewards/accuracies": 0.956250011920929,
|
| 1883 |
+
"rewards/generated": -1.131711721420288,
|
| 1884 |
+
"rewards/margins": 1.1664810180664062,
|
| 1885 |
+
"rewards/real": 0.03476935625076294,
|
| 1886 |
+
"step": 610
|
| 1887 |
+
},
|
| 1888 |
+
{
|
| 1889 |
+
"epoch": 0.79,
|
| 1890 |
+
"grad_norm": 36.0,
|
| 1891 |
+
"learning_rate": 1.1823361823361823e-07,
|
| 1892 |
+
"logits/generated": 0.2149450033903122,
|
| 1893 |
+
"logits/real": -0.3522172272205353,
|
| 1894 |
+
"logps/generated": -569.8680419921875,
|
| 1895 |
+
"logps/real": -522.5927734375,
|
| 1896 |
+
"loss": 0.3041,
|
| 1897 |
+
"rewards/accuracies": 0.987500011920929,
|
| 1898 |
+
"rewards/generated": -1.102595567703247,
|
| 1899 |
+
"rewards/margins": 1.1690142154693604,
|
| 1900 |
+
"rewards/real": 0.06641869246959686,
|
| 1901 |
+
"step": 615
|
| 1902 |
+
},
|
| 1903 |
+
{
|
| 1904 |
+
"epoch": 0.79,
|
| 1905 |
+
"grad_norm": 34.0,
|
| 1906 |
+
"learning_rate": 1.1467236467236467e-07,
|
| 1907 |
+
"logits/generated": 0.21769335865974426,
|
| 1908 |
+
"logits/real": -0.3188803493976593,
|
| 1909 |
+
"logps/generated": -525.9594116210938,
|
| 1910 |
+
"logps/real": -459.16900634765625,
|
| 1911 |
+
"loss": 0.3244,
|
| 1912 |
+
"rewards/accuracies": 0.96875,
|
| 1913 |
+
"rewards/generated": -1.1512959003448486,
|
| 1914 |
+
"rewards/margins": 1.1399847269058228,
|
| 1915 |
+
"rewards/real": -0.011311108246445656,
|
| 1916 |
+
"step": 620
|
| 1917 |
+
},
|
| 1918 |
+
{
|
| 1919 |
+
"epoch": 0.8,
|
| 1920 |
+
"grad_norm": 37.25,
|
| 1921 |
+
"learning_rate": 1.111111111111111e-07,
|
| 1922 |
+
"logits/generated": 0.1936269998550415,
|
| 1923 |
+
"logits/real": -0.2803877294063568,
|
| 1924 |
+
"logps/generated": -540.1295166015625,
|
| 1925 |
+
"logps/real": -444.78375244140625,
|
| 1926 |
+
"loss": 0.3322,
|
| 1927 |
+
"rewards/accuracies": 0.9375,
|
| 1928 |
+
"rewards/generated": -1.0809428691864014,
|
| 1929 |
+
"rewards/margins": 1.1129690408706665,
|
| 1930 |
+
"rewards/real": 0.03202588111162186,
|
| 1931 |
+
"step": 625
|
| 1932 |
+
},
|
| 1933 |
+
{
|
| 1934 |
+
"epoch": 0.81,
|
| 1935 |
+
"grad_norm": 29.625,
|
| 1936 |
+
"learning_rate": 1.0754985754985754e-07,
|
| 1937 |
+
"logits/generated": 0.19856727123260498,
|
| 1938 |
+
"logits/real": -0.27393585443496704,
|
| 1939 |
+
"logps/generated": -560.6895751953125,
|
| 1940 |
+
"logps/real": -426.6712951660156,
|
| 1941 |
+
"loss": 0.3194,
|
| 1942 |
+
"rewards/accuracies": 0.9624999761581421,
|
| 1943 |
+
"rewards/generated": -1.1213538646697998,
|
| 1944 |
+
"rewards/margins": 1.142456293106079,
|
| 1945 |
+
"rewards/real": 0.021102434024214745,
|
| 1946 |
+
"step": 630
|
| 1947 |
+
},
|
| 1948 |
+
{
|
| 1949 |
+
"epoch": 0.81,
|
| 1950 |
+
"grad_norm": 31.75,
|
| 1951 |
+
"learning_rate": 1.0398860398860399e-07,
|
| 1952 |
+
"logits/generated": 0.21278016269207,
|
| 1953 |
+
"logits/real": -0.24991516768932343,
|
| 1954 |
+
"logps/generated": -587.5706176757812,
|
| 1955 |
+
"logps/real": -404.28033447265625,
|
| 1956 |
+
"loss": 0.3189,
|
| 1957 |
+
"rewards/accuracies": 0.9437500238418579,
|
| 1958 |
+
"rewards/generated": -1.1712287664413452,
|
| 1959 |
+
"rewards/margins": 1.2088580131530762,
|
| 1960 |
+
"rewards/real": 0.037629302591085434,
|
| 1961 |
+
"step": 635
|
| 1962 |
+
},
|
| 1963 |
+
{
|
| 1964 |
+
"epoch": 0.82,
|
| 1965 |
+
"grad_norm": 31.75,
|
| 1966 |
+
"learning_rate": 1.0042735042735042e-07,
|
| 1967 |
+
"logits/generated": 0.22561617195606232,
|
| 1968 |
+
"logits/real": -0.20534536242485046,
|
| 1969 |
+
"logps/generated": -575.3756713867188,
|
| 1970 |
+
"logps/real": -442.997802734375,
|
| 1971 |
+
"loss": 0.3369,
|
| 1972 |
+
"rewards/accuracies": 0.96875,
|
| 1973 |
+
"rewards/generated": -1.1977014541625977,
|
| 1974 |
+
"rewards/margins": 1.1940991878509521,
|
| 1975 |
+
"rewards/real": -0.003602252807468176,
|
| 1976 |
+
"step": 640
|
| 1977 |
+
},
|
| 1978 |
+
{
|
| 1979 |
+
"epoch": 0.83,
|
| 1980 |
+
"grad_norm": 31.875,
|
| 1981 |
+
"learning_rate": 9.686609686609686e-08,
|
| 1982 |
+
"logits/generated": 0.2502003610134125,
|
| 1983 |
+
"logits/real": -0.2464480847120285,
|
| 1984 |
+
"logps/generated": -574.8699340820312,
|
| 1985 |
+
"logps/real": -462.56134033203125,
|
| 1986 |
+
"loss": 0.3468,
|
| 1987 |
+
"rewards/accuracies": 0.925000011920929,
|
| 1988 |
+
"rewards/generated": -1.0637601613998413,
|
| 1989 |
+
"rewards/margins": 1.0312858819961548,
|
| 1990 |
+
"rewards/real": -0.032474346458911896,
|
| 1991 |
+
"step": 645
|
| 1992 |
+
},
|
| 1993 |
+
{
|
| 1994 |
+
"epoch": 0.83,
|
| 1995 |
+
"grad_norm": 29.875,
|
| 1996 |
+
"learning_rate": 9.33048433048433e-08,
|
| 1997 |
+
"logits/generated": 0.2645814120769501,
|
| 1998 |
+
"logits/real": -0.2936360836029053,
|
| 1999 |
+
"logps/generated": -551.9849243164062,
|
| 2000 |
+
"logps/real": -463.1893005371094,
|
| 2001 |
+
"loss": 0.3202,
|
| 2002 |
+
"rewards/accuracies": 0.9375,
|
| 2003 |
+
"rewards/generated": -1.102562665939331,
|
| 2004 |
+
"rewards/margins": 1.100036382675171,
|
| 2005 |
+
"rewards/real": -0.002526202006265521,
|
| 2006 |
+
"step": 650
|
| 2007 |
+
},
|
| 2008 |
+
{
|
| 2009 |
+
"epoch": 0.84,
|
| 2010 |
+
"grad_norm": 33.25,
|
| 2011 |
+
"learning_rate": 8.974358974358974e-08,
|
| 2012 |
+
"logits/generated": 0.24026402831077576,
|
| 2013 |
+
"logits/real": -0.24332275986671448,
|
| 2014 |
+
"logps/generated": -586.1027221679688,
|
| 2015 |
+
"logps/real": -457.7566833496094,
|
| 2016 |
+
"loss": 0.3439,
|
| 2017 |
+
"rewards/accuracies": 0.925000011920929,
|
| 2018 |
+
"rewards/generated": -1.0856201648712158,
|
| 2019 |
+
"rewards/margins": 1.0498621463775635,
|
| 2020 |
+
"rewards/real": -0.035757891833782196,
|
| 2021 |
+
"step": 655
|
| 2022 |
+
},
|
| 2023 |
+
{
|
| 2024 |
+
"epoch": 0.84,
|
| 2025 |
+
"grad_norm": 32.75,
|
| 2026 |
+
"learning_rate": 8.618233618233619e-08,
|
| 2027 |
+
"logits/generated": 0.22486528754234314,
|
| 2028 |
+
"logits/real": -0.33010414242744446,
|
| 2029 |
+
"logps/generated": -575.5369262695312,
|
| 2030 |
+
"logps/real": -444.61712646484375,
|
| 2031 |
+
"loss": 0.3306,
|
| 2032 |
+
"rewards/accuracies": 0.956250011920929,
|
| 2033 |
+
"rewards/generated": -1.1358319520950317,
|
| 2034 |
+
"rewards/margins": 1.1036947965621948,
|
| 2035 |
+
"rewards/real": -0.03213699907064438,
|
| 2036 |
+
"step": 660
|
| 2037 |
+
},
|
| 2038 |
+
{
|
| 2039 |
+
"epoch": 0.85,
|
| 2040 |
+
"grad_norm": 35.0,
|
| 2041 |
+
"learning_rate": 8.262108262108261e-08,
|
| 2042 |
+
"logits/generated": 0.22603563964366913,
|
| 2043 |
+
"logits/real": -0.2633499503135681,
|
| 2044 |
+
"logps/generated": -588.1068725585938,
|
| 2045 |
+
"logps/real": -409.90362548828125,
|
| 2046 |
+
"loss": 0.3222,
|
| 2047 |
+
"rewards/accuracies": 0.9437500238418579,
|
| 2048 |
+
"rewards/generated": -1.1506733894348145,
|
| 2049 |
+
"rewards/margins": 1.1825590133666992,
|
| 2050 |
+
"rewards/real": 0.03188558295369148,
|
| 2051 |
+
"step": 665
|
| 2052 |
+
},
|
| 2053 |
+
{
|
| 2054 |
+
"epoch": 0.86,
|
| 2055 |
+
"grad_norm": 32.0,
|
| 2056 |
+
"learning_rate": 7.905982905982906e-08,
|
| 2057 |
+
"logits/generated": 0.2575734257698059,
|
| 2058 |
+
"logits/real": -0.26269882917404175,
|
| 2059 |
+
"logps/generated": -613.2880859375,
|
| 2060 |
+
"logps/real": -436.75689697265625,
|
| 2061 |
+
"loss": 0.3172,
|
| 2062 |
+
"rewards/accuracies": 0.981249988079071,
|
| 2063 |
+
"rewards/generated": -1.1868383884429932,
|
| 2064 |
+
"rewards/margins": 1.1978209018707275,
|
| 2065 |
+
"rewards/real": 0.010982497595250607,
|
| 2066 |
+
"step": 670
|
| 2067 |
+
},
|
| 2068 |
+
{
|
| 2069 |
+
"epoch": 0.86,
|
| 2070 |
+
"grad_norm": 36.5,
|
| 2071 |
+
"learning_rate": 7.549857549857549e-08,
|
| 2072 |
+
"logits/generated": 0.21627910435199738,
|
| 2073 |
+
"logits/real": -0.31892699003219604,
|
| 2074 |
+
"logps/generated": -590.7273559570312,
|
| 2075 |
+
"logps/real": -460.1044921875,
|
| 2076 |
+
"loss": 0.3361,
|
| 2077 |
+
"rewards/accuracies": 0.9437500238418579,
|
| 2078 |
+
"rewards/generated": -1.1508667469024658,
|
| 2079 |
+
"rewards/margins": 1.1213773488998413,
|
| 2080 |
+
"rewards/real": -0.029489323496818542,
|
| 2081 |
+
"step": 675
|
| 2082 |
+
},
|
| 2083 |
+
{
|
| 2084 |
+
"epoch": 0.87,
|
| 2085 |
+
"grad_norm": 30.5,
|
| 2086 |
+
"learning_rate": 7.193732193732194e-08,
|
| 2087 |
+
"logits/generated": 0.2537713944911957,
|
| 2088 |
+
"logits/real": -0.2855969965457916,
|
| 2089 |
+
"logps/generated": -577.8758544921875,
|
| 2090 |
+
"logps/real": -475.7315368652344,
|
| 2091 |
+
"loss": 0.3277,
|
| 2092 |
+
"rewards/accuracies": 0.9125000238418579,
|
| 2093 |
+
"rewards/generated": -1.1186813116073608,
|
| 2094 |
+
"rewards/margins": 1.1131179332733154,
|
| 2095 |
+
"rewards/real": -0.005563369486480951,
|
| 2096 |
+
"step": 680
|
| 2097 |
+
},
|
| 2098 |
+
{
|
| 2099 |
+
"epoch": 0.88,
|
| 2100 |
+
"grad_norm": 31.25,
|
| 2101 |
+
"learning_rate": 6.837606837606839e-08,
|
| 2102 |
+
"logits/generated": 0.20594389736652374,
|
| 2103 |
+
"logits/real": -0.31517493724823,
|
| 2104 |
+
"logps/generated": -553.8355712890625,
|
| 2105 |
+
"logps/real": -434.3343811035156,
|
| 2106 |
+
"loss": 0.3286,
|
| 2107 |
+
"rewards/accuracies": 0.9624999761581421,
|
| 2108 |
+
"rewards/generated": -1.1672719717025757,
|
| 2109 |
+
"rewards/margins": 1.2395508289337158,
|
| 2110 |
+
"rewards/real": 0.07227896898984909,
|
| 2111 |
+
"step": 685
|
| 2112 |
+
},
|
| 2113 |
+
{
|
| 2114 |
+
"epoch": 0.88,
|
| 2115 |
+
"grad_norm": 32.0,
|
| 2116 |
+
"learning_rate": 6.481481481481481e-08,
|
| 2117 |
+
"logits/generated": 0.205632284283638,
|
| 2118 |
+
"logits/real": -0.255433589220047,
|
| 2119 |
+
"logps/generated": -570.3898315429688,
|
| 2120 |
+
"logps/real": -464.691650390625,
|
| 2121 |
+
"loss": 0.3308,
|
| 2122 |
+
"rewards/accuracies": 0.956250011920929,
|
| 2123 |
+
"rewards/generated": -1.0862834453582764,
|
| 2124 |
+
"rewards/margins": 1.1157293319702148,
|
| 2125 |
+
"rewards/real": 0.029445823282003403,
|
| 2126 |
+
"step": 690
|
| 2127 |
+
},
|
| 2128 |
+
{
|
| 2129 |
+
"epoch": 0.89,
|
| 2130 |
+
"grad_norm": 33.75,
|
| 2131 |
+
"learning_rate": 6.125356125356125e-08,
|
| 2132 |
+
"logits/generated": 0.23636643588542938,
|
| 2133 |
+
"logits/real": -0.2590656280517578,
|
| 2134 |
+
"logps/generated": -598.99755859375,
|
| 2135 |
+
"logps/real": -467.8699645996094,
|
| 2136 |
+
"loss": 0.3365,
|
| 2137 |
+
"rewards/accuracies": 0.956250011920929,
|
| 2138 |
+
"rewards/generated": -1.104642629623413,
|
| 2139 |
+
"rewards/margins": 1.1386440992355347,
|
| 2140 |
+
"rewards/real": 0.034001342952251434,
|
| 2141 |
+
"step": 695
|
| 2142 |
+
},
|
| 2143 |
+
{
|
| 2144 |
+
"epoch": 0.9,
|
| 2145 |
+
"grad_norm": 33.5,
|
| 2146 |
+
"learning_rate": 5.7692307692307695e-08,
|
| 2147 |
+
"logits/generated": 0.25425681471824646,
|
| 2148 |
+
"logits/real": -0.2653648853302002,
|
| 2149 |
+
"logps/generated": -569.2490844726562,
|
| 2150 |
+
"logps/real": -475.88763427734375,
|
| 2151 |
+
"loss": 0.3354,
|
| 2152 |
+
"rewards/accuracies": 0.956250011920929,
|
| 2153 |
+
"rewards/generated": -1.1180486679077148,
|
| 2154 |
+
"rewards/margins": 1.0991861820220947,
|
| 2155 |
+
"rewards/real": -0.018862556666135788,
|
| 2156 |
+
"step": 700
|
| 2157 |
+
},
|
| 2158 |
+
{
|
| 2159 |
+
"epoch": 0.9,
|
| 2160 |
+
"grad_norm": 30.875,
|
| 2161 |
+
"learning_rate": 5.413105413105413e-08,
|
| 2162 |
+
"logits/generated": 0.22857370972633362,
|
| 2163 |
+
"logits/real": -0.22129984200000763,
|
| 2164 |
+
"logps/generated": -563.016357421875,
|
| 2165 |
+
"logps/real": -403.68597412109375,
|
| 2166 |
+
"loss": 0.3157,
|
| 2167 |
+
"rewards/accuracies": 0.949999988079071,
|
| 2168 |
+
"rewards/generated": -1.0614380836486816,
|
| 2169 |
+
"rewards/margins": 1.1257336139678955,
|
| 2170 |
+
"rewards/real": 0.06429564207792282,
|
| 2171 |
+
"step": 705
|
| 2172 |
+
},
|
| 2173 |
+
{
|
| 2174 |
+
"epoch": 0.91,
|
| 2175 |
+
"grad_norm": 36.75,
|
| 2176 |
+
"learning_rate": 5.056980056980057e-08,
|
| 2177 |
+
"logits/generated": 0.21650013327598572,
|
| 2178 |
+
"logits/real": -0.3323083817958832,
|
| 2179 |
+
"logps/generated": -581.8919677734375,
|
| 2180 |
+
"logps/real": -487.620361328125,
|
| 2181 |
+
"loss": 0.3226,
|
| 2182 |
+
"rewards/accuracies": 0.981249988079071,
|
| 2183 |
+
"rewards/generated": -1.2059048414230347,
|
| 2184 |
+
"rewards/margins": 1.1841565370559692,
|
| 2185 |
+
"rewards/real": -0.021748319268226624,
|
| 2186 |
+
"step": 710
|
| 2187 |
+
},
|
| 2188 |
+
{
|
| 2189 |
+
"epoch": 0.92,
|
| 2190 |
+
"grad_norm": 30.875,
|
| 2191 |
+
"learning_rate": 4.7008547008547005e-08,
|
| 2192 |
+
"logits/generated": 0.2445484846830368,
|
| 2193 |
+
"logits/real": -0.31186121702194214,
|
| 2194 |
+
"logps/generated": -557.0134887695312,
|
| 2195 |
+
"logps/real": -493.17230224609375,
|
| 2196 |
+
"loss": 0.3212,
|
| 2197 |
+
"rewards/accuracies": 0.956250011920929,
|
| 2198 |
+
"rewards/generated": -1.1671897172927856,
|
| 2199 |
+
"rewards/margins": 1.1586859226226807,
|
| 2200 |
+
"rewards/real": -0.00850372202694416,
|
| 2201 |
+
"step": 715
|
| 2202 |
+
},
|
| 2203 |
+
{
|
| 2204 |
+
"epoch": 0.92,
|
| 2205 |
+
"grad_norm": 34.25,
|
| 2206 |
+
"learning_rate": 4.3447293447293445e-08,
|
| 2207 |
+
"logits/generated": 0.21253076195716858,
|
| 2208 |
+
"logits/real": -0.2138686180114746,
|
| 2209 |
+
"logps/generated": -586.7265625,
|
| 2210 |
+
"logps/real": -408.2170104980469,
|
| 2211 |
+
"loss": 0.3419,
|
| 2212 |
+
"rewards/accuracies": 0.9375,
|
| 2213 |
+
"rewards/generated": -1.1229650974273682,
|
| 2214 |
+
"rewards/margins": 1.1265299320220947,
|
| 2215 |
+
"rewards/real": 0.0035647773183882236,
|
| 2216 |
+
"step": 720
|
| 2217 |
+
},
|
| 2218 |
+
{
|
| 2219 |
+
"epoch": 0.93,
|
| 2220 |
+
"grad_norm": 29.75,
|
| 2221 |
+
"learning_rate": 3.9886039886039886e-08,
|
| 2222 |
+
"logits/generated": 0.18125857412815094,
|
| 2223 |
+
"logits/real": -0.32901662588119507,
|
| 2224 |
+
"logps/generated": -537.9093017578125,
|
| 2225 |
+
"logps/real": -448.9685974121094,
|
| 2226 |
+
"loss": 0.3366,
|
| 2227 |
+
"rewards/accuracies": 0.956250011920929,
|
| 2228 |
+
"rewards/generated": -1.0713820457458496,
|
| 2229 |
+
"rewards/margins": 1.105197548866272,
|
| 2230 |
+
"rewards/real": 0.033815376460552216,
|
| 2231 |
+
"step": 725
|
| 2232 |
+
},
|
| 2233 |
+
{
|
| 2234 |
+
"epoch": 0.93,
|
| 2235 |
+
"grad_norm": 32.25,
|
| 2236 |
+
"learning_rate": 3.632478632478633e-08,
|
| 2237 |
+
"logits/generated": 0.21634677052497864,
|
| 2238 |
+
"logits/real": -0.30626875162124634,
|
| 2239 |
+
"logps/generated": -582.5755004882812,
|
| 2240 |
+
"logps/real": -460.1416931152344,
|
| 2241 |
+
"loss": 0.3136,
|
| 2242 |
+
"rewards/accuracies": 0.956250011920929,
|
| 2243 |
+
"rewards/generated": -1.1917552947998047,
|
| 2244 |
+
"rewards/margins": 1.2281749248504639,
|
| 2245 |
+
"rewards/real": 0.03641948103904724,
|
| 2246 |
+
"step": 730
|
| 2247 |
+
},
|
| 2248 |
+
{
|
| 2249 |
+
"epoch": 0.94,
|
| 2250 |
+
"grad_norm": 30.625,
|
| 2251 |
+
"learning_rate": 3.276353276353276e-08,
|
| 2252 |
+
"logits/generated": 0.2181752622127533,
|
| 2253 |
+
"logits/real": -0.34954366087913513,
|
| 2254 |
+
"logps/generated": -566.8768310546875,
|
| 2255 |
+
"logps/real": -490.39385986328125,
|
| 2256 |
+
"loss": 0.3149,
|
| 2257 |
+
"rewards/accuracies": 0.949999988079071,
|
| 2258 |
+
"rewards/generated": -1.2017440795898438,
|
| 2259 |
+
"rewards/margins": 1.2231340408325195,
|
| 2260 |
+
"rewards/real": 0.0213897917419672,
|
| 2261 |
+
"step": 735
|
| 2262 |
+
},
|
| 2263 |
+
{
|
| 2264 |
+
"epoch": 0.95,
|
| 2265 |
+
"grad_norm": 30.75,
|
| 2266 |
+
"learning_rate": 2.92022792022792e-08,
|
| 2267 |
+
"logits/generated": 0.250019907951355,
|
| 2268 |
+
"logits/real": -0.23180410265922546,
|
| 2269 |
+
"logps/generated": -569.7205810546875,
|
| 2270 |
+
"logps/real": -436.3052673339844,
|
| 2271 |
+
"loss": 0.3193,
|
| 2272 |
+
"rewards/accuracies": 0.9437500238418579,
|
| 2273 |
+
"rewards/generated": -1.0622925758361816,
|
| 2274 |
+
"rewards/margins": 1.074755311012268,
|
| 2275 |
+
"rewards/real": 0.012462759390473366,
|
| 2276 |
+
"step": 740
|
| 2277 |
+
},
|
| 2278 |
+
{
|
| 2279 |
+
"epoch": 0.95,
|
| 2280 |
+
"grad_norm": 32.0,
|
| 2281 |
+
"learning_rate": 2.564102564102564e-08,
|
| 2282 |
+
"logits/generated": 0.2107166051864624,
|
| 2283 |
+
"logits/real": -0.28946200013160706,
|
| 2284 |
+
"logps/generated": -562.4656982421875,
|
| 2285 |
+
"logps/real": -423.35076904296875,
|
| 2286 |
+
"loss": 0.3288,
|
| 2287 |
+
"rewards/accuracies": 0.956250011920929,
|
| 2288 |
+
"rewards/generated": -1.1599444150924683,
|
| 2289 |
+
"rewards/margins": 1.1884777545928955,
|
| 2290 |
+
"rewards/real": 0.02853330597281456,
|
| 2291 |
+
"step": 745
|
| 2292 |
+
},
|
| 2293 |
+
{
|
| 2294 |
+
"epoch": 0.96,
|
| 2295 |
+
"grad_norm": 33.25,
|
| 2296 |
+
"learning_rate": 2.2079772079772077e-08,
|
| 2297 |
+
"logits/generated": 0.2310904562473297,
|
| 2298 |
+
"logits/real": -0.2765614092350006,
|
| 2299 |
+
"logps/generated": -604.2635498046875,
|
| 2300 |
+
"logps/real": -450.979248046875,
|
| 2301 |
+
"loss": 0.3074,
|
| 2302 |
+
"rewards/accuracies": 0.9624999761581421,
|
| 2303 |
+
"rewards/generated": -1.2296929359436035,
|
| 2304 |
+
"rewards/margins": 1.2464648485183716,
|
| 2305 |
+
"rewards/real": 0.016771936789155006,
|
| 2306 |
+
"step": 750
|
| 2307 |
+
},
|
| 2308 |
+
{
|
| 2309 |
+
"epoch": 0.97,
|
| 2310 |
+
"grad_norm": 31.125,
|
| 2311 |
+
"learning_rate": 1.8518518518518518e-08,
|
| 2312 |
+
"logits/generated": 0.17430028319358826,
|
| 2313 |
+
"logits/real": -0.33389025926589966,
|
| 2314 |
+
"logps/generated": -549.4244995117188,
|
| 2315 |
+
"logps/real": -464.75653076171875,
|
| 2316 |
+
"loss": 0.3085,
|
| 2317 |
+
"rewards/accuracies": 0.9624999761581421,
|
| 2318 |
+
"rewards/generated": -1.1594291925430298,
|
| 2319 |
+
"rewards/margins": 1.2221739292144775,
|
| 2320 |
+
"rewards/real": 0.0627446323633194,
|
| 2321 |
+
"step": 755
|
| 2322 |
+
},
|
| 2323 |
+
{
|
| 2324 |
+
"epoch": 0.97,
|
| 2325 |
+
"grad_norm": 32.0,
|
| 2326 |
+
"learning_rate": 1.4957264957264956e-08,
|
| 2327 |
+
"logits/generated": 0.24278989434242249,
|
| 2328 |
+
"logits/real": -0.26737385988235474,
|
| 2329 |
+
"logps/generated": -573.1634521484375,
|
| 2330 |
+
"logps/real": -469.896484375,
|
| 2331 |
+
"loss": 0.3263,
|
| 2332 |
+
"rewards/accuracies": 0.949999988079071,
|
| 2333 |
+
"rewards/generated": -1.049228310585022,
|
| 2334 |
+
"rewards/margins": 1.1048952341079712,
|
| 2335 |
+
"rewards/real": 0.055667001754045486,
|
| 2336 |
+
"step": 760
|
| 2337 |
+
},
|
| 2338 |
+
{
|
| 2339 |
+
"epoch": 0.98,
|
| 2340 |
+
"grad_norm": 34.5,
|
| 2341 |
+
"learning_rate": 1.1396011396011397e-08,
|
| 2342 |
+
"logits/generated": 0.1711110770702362,
|
| 2343 |
+
"logits/real": -0.35616591572761536,
|
| 2344 |
+
"logps/generated": -552.7576293945312,
|
| 2345 |
+
"logps/real": -433.0294494628906,
|
| 2346 |
+
"loss": 0.3352,
|
| 2347 |
+
"rewards/accuracies": 0.96875,
|
| 2348 |
+
"rewards/generated": -1.0972100496292114,
|
| 2349 |
+
"rewards/margins": 1.1236265897750854,
|
| 2350 |
+
"rewards/real": 0.02641637995839119,
|
| 2351 |
+
"step": 765
|
| 2352 |
+
},
|
| 2353 |
+
{
|
| 2354 |
+
"epoch": 0.99,
|
| 2355 |
+
"grad_norm": 33.75,
|
| 2356 |
+
"learning_rate": 7.834757834757834e-09,
|
| 2357 |
+
"logits/generated": 0.3013153672218323,
|
| 2358 |
+
"logits/real": -0.22638103365898132,
|
| 2359 |
+
"logps/generated": -591.3074951171875,
|
| 2360 |
+
"logps/real": -471.1341857910156,
|
| 2361 |
+
"loss": 0.3269,
|
| 2362 |
+
"rewards/accuracies": 0.956250011920929,
|
| 2363 |
+
"rewards/generated": -1.1128944158554077,
|
| 2364 |
+
"rewards/margins": 1.088680624961853,
|
| 2365 |
+
"rewards/real": -0.02421378344297409,
|
| 2366 |
+
"step": 770
|
| 2367 |
+
},
|
| 2368 |
+
{
|
| 2369 |
+
"epoch": 0.99,
|
| 2370 |
+
"grad_norm": 31.0,
|
| 2371 |
+
"learning_rate": 4.273504273504274e-09,
|
| 2372 |
+
"logits/generated": 0.2685278058052063,
|
| 2373 |
+
"logits/real": -0.22944375872612,
|
| 2374 |
+
"logps/generated": -570.2601928710938,
|
| 2375 |
+
"logps/real": -418.1946716308594,
|
| 2376 |
+
"loss": 0.3241,
|
| 2377 |
+
"rewards/accuracies": 0.9624999761581421,
|
| 2378 |
+
"rewards/generated": -1.138912320137024,
|
| 2379 |
+
"rewards/margins": 1.1833699941635132,
|
| 2380 |
+
"rewards/real": 0.04445788636803627,
|
| 2381 |
+
"step": 775
|
| 2382 |
+
},
|
| 2383 |
+
{
|
| 2384 |
+
"epoch": 1.0,
|
| 2385 |
+
"grad_norm": 33.25,
|
| 2386 |
+
"learning_rate": 7.122507122507123e-10,
|
| 2387 |
+
"logits/generated": 0.23055453598499298,
|
| 2388 |
+
"logits/real": -0.3259603977203369,
|
| 2389 |
+
"logps/generated": -538.0984497070312,
|
| 2390 |
+
"logps/real": -500.5188903808594,
|
| 2391 |
+
"loss": 0.3448,
|
| 2392 |
+
"rewards/accuracies": 0.925000011920929,
|
| 2393 |
+
"rewards/generated": -1.0663647651672363,
|
| 2394 |
+
"rewards/margins": 1.0742255449295044,
|
| 2395 |
+
"rewards/real": 0.007860781624913216,
|
| 2396 |
+
"step": 780
|
| 2397 |
+
},
|
| 2398 |
+
{
|
| 2399 |
+
"epoch": 1.0,
|
| 2400 |
+
"step": 781,
|
| 2401 |
+
"total_flos": 0.0,
|
| 2402 |
+
"train_loss": 0.41590739681687156,
|
| 2403 |
+
"train_runtime": 9143.8407,
|
| 2404 |
+
"train_samples_per_second": 5.468,
|
| 2405 |
+
"train_steps_per_second": 0.085
|
| 2406 |
+
}
|
| 2407 |
+
],
|
| 2408 |
+
"logging_steps": 5,
|
| 2409 |
+
"max_steps": 781,
|
| 2410 |
+
"num_input_tokens_seen": 0,
|
| 2411 |
+
"num_train_epochs": 1,
|
| 2412 |
+
"save_steps": 500,
|
| 2413 |
+
"total_flos": 0.0,
|
| 2414 |
+
"train_batch_size": 4,
|
| 2415 |
+
"trial_name": null,
|
| 2416 |
+
"trial_params": null
|
| 2417 |
+
}
|