Commit
•
b937976
1
Parent(s):
44b1319
End of training
Browse files- README.md +124 -0
- all_results.json +15 -0
- datasets/all_binary_and_xe_ey_fae_counterfactual/adapter_config.json +41 -0
- datasets/all_binary_and_xe_ey_fae_counterfactual/head_config.json +15 -0
- datasets/all_binary_and_xe_ey_fae_counterfactual/pytorch_adapter.bin +3 -0
- datasets/all_binary_and_xe_ey_fae_counterfactual/pytorch_model_head.bin +3 -0
- eval_results.json +10 -0
- train_results.json +8 -0
- trainer_state.json +780 -0
README.md
ADDED
|
@@ -0,0 +1,124 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
license: apache-2.0
|
| 3 |
+
base_model: facebook/bart-base
|
| 4 |
+
tags:
|
| 5 |
+
- generated_from_trainer
|
| 6 |
+
datasets:
|
| 7 |
+
- datasets/all_binary_and_xe_ey_fae_counterfactual
|
| 8 |
+
metrics:
|
| 9 |
+
- accuracy
|
| 10 |
+
model-index:
|
| 11 |
+
- name: bart-adapter-finetuned-xe_ey_fae
|
| 12 |
+
results:
|
| 13 |
+
- task:
|
| 14 |
+
name: Masked Language Modeling
|
| 15 |
+
type: fill-mask
|
| 16 |
+
dataset:
|
| 17 |
+
name: datasets/all_binary_and_xe_ey_fae_counterfactual
|
| 18 |
+
type: datasets/all_binary_and_xe_ey_fae_counterfactual
|
| 19 |
+
metrics:
|
| 20 |
+
- name: Accuracy
|
| 21 |
+
type: accuracy
|
| 22 |
+
value: 0.3096946377787028
|
| 23 |
+
---
|
| 24 |
+
|
| 25 |
+
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
| 26 |
+
should probably proofread and complete it, then remove this comment. -->
|
| 27 |
+
|
| 28 |
+
# bart-adapter-finetuned-xe_ey_fae
|
| 29 |
+
|
| 30 |
+
This model is a fine-tuned version of [facebook/bart-base](https://huggingface.co/facebook/bart-base) on the datasets/all_binary_and_xe_ey_fae_counterfactual dataset.
|
| 31 |
+
It achieves the following results on the evaluation set:
|
| 32 |
+
- Loss: 4.2302
|
| 33 |
+
- Accuracy: 0.3097
|
| 34 |
+
|
| 35 |
+
## Model description
|
| 36 |
+
|
| 37 |
+
More information needed
|
| 38 |
+
|
| 39 |
+
## Intended uses & limitations
|
| 40 |
+
|
| 41 |
+
More information needed
|
| 42 |
+
|
| 43 |
+
## Training and evaluation data
|
| 44 |
+
|
| 45 |
+
More information needed
|
| 46 |
+
|
| 47 |
+
## Training procedure
|
| 48 |
+
|
| 49 |
+
### Training hyperparameters
|
| 50 |
+
|
| 51 |
+
The following hyperparameters were used during training:
|
| 52 |
+
- learning_rate: 1e-05
|
| 53 |
+
- train_batch_size: 8
|
| 54 |
+
- eval_batch_size: 8
|
| 55 |
+
- seed: 100
|
| 56 |
+
- gradient_accumulation_steps: 2
|
| 57 |
+
- total_train_batch_size: 16
|
| 58 |
+
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
| 59 |
+
- lr_scheduler_type: linear
|
| 60 |
+
- num_epochs: 3.0
|
| 61 |
+
- mixed_precision_training: Native AMP
|
| 62 |
+
|
| 63 |
+
### Training results
|
| 64 |
+
|
| 65 |
+
| Training Loss | Epoch | Step | Validation Loss | Accuracy |
|
| 66 |
+
|:-------------:|:-----:|:-----:|:---------------:|:--------:|
|
| 67 |
+
| 7.6974 | 0.06 | 500 | 6.7246 | 0.0649 |
|
| 68 |
+
| 6.8017 | 0.12 | 1000 | 6.4067 | 0.0762 |
|
| 69 |
+
| 6.5894 | 0.18 | 1500 | 6.2661 | 0.0821 |
|
| 70 |
+
| 6.443 | 0.24 | 2000 | 6.1350 | 0.0905 |
|
| 71 |
+
| 6.3245 | 0.3 | 2500 | 6.0024 | 0.1008 |
|
| 72 |
+
| 6.2208 | 0.35 | 3000 | 5.8518 | 0.1145 |
|
| 73 |
+
| 6.097 | 0.41 | 3500 | 5.6588 | 0.1330 |
|
| 74 |
+
| 5.9862 | 0.47 | 4000 | 5.4641 | 0.1543 |
|
| 75 |
+
| 5.8742 | 0.53 | 4500 | 5.3200 | 0.1707 |
|
| 76 |
+
| 5.7716 | 0.59 | 5000 | 5.2044 | 0.1840 |
|
| 77 |
+
| 5.6952 | 0.65 | 5500 | 5.1154 | 0.1952 |
|
| 78 |
+
| 5.6209 | 0.71 | 6000 | 5.0428 | 0.2044 |
|
| 79 |
+
| 5.5752 | 0.77 | 6500 | 4.9711 | 0.2136 |
|
| 80 |
+
| 5.5091 | 0.83 | 7000 | 4.9078 | 0.2212 |
|
| 81 |
+
| 5.4657 | 0.89 | 7500 | 4.8495 | 0.2287 |
|
| 82 |
+
| 5.4245 | 0.95 | 8000 | 4.8012 | 0.2360 |
|
| 83 |
+
| 5.3813 | 1.0 | 8500 | 4.7563 | 0.2409 |
|
| 84 |
+
| 5.3501 | 1.06 | 9000 | 4.7166 | 0.2464 |
|
| 85 |
+
| 5.3098 | 1.12 | 9500 | 4.6838 | 0.2501 |
|
| 86 |
+
| 5.2856 | 1.18 | 10000 | 4.6515 | 0.2551 |
|
| 87 |
+
| 5.2549 | 1.24 | 10500 | 4.6121 | 0.2602 |
|
| 88 |
+
| 5.2217 | 1.3 | 11000 | 4.5841 | 0.2637 |
|
| 89 |
+
| 5.1997 | 1.36 | 11500 | 4.5588 | 0.2674 |
|
| 90 |
+
| 5.1844 | 1.42 | 12000 | 4.5309 | 0.2708 |
|
| 91 |
+
| 5.1491 | 1.48 | 12500 | 4.4999 | 0.2748 |
|
| 92 |
+
| 5.1244 | 1.54 | 13000 | 4.4783 | 0.2780 |
|
| 93 |
+
| 5.1047 | 1.6 | 13500 | 4.4561 | 0.2812 |
|
| 94 |
+
| 5.0917 | 1.66 | 14000 | 4.4409 | 0.2826 |
|
| 95 |
+
| 5.0631 | 1.71 | 14500 | 4.4198 | 0.2851 |
|
| 96 |
+
| 5.0537 | 1.77 | 15000 | 4.4003 | 0.2881 |
|
| 97 |
+
| 5.0339 | 1.83 | 15500 | 4.3855 | 0.2899 |
|
| 98 |
+
| 5.0235 | 1.89 | 16000 | 4.3650 | 0.2921 |
|
| 99 |
+
| 5.0074 | 1.95 | 16500 | 4.3496 | 0.2942 |
|
| 100 |
+
| 4.9927 | 2.01 | 17000 | 4.3361 | 0.2965 |
|
| 101 |
+
| 4.9797 | 2.07 | 17500 | 4.3203 | 0.2981 |
|
| 102 |
+
| 4.9725 | 2.13 | 18000 | 4.3118 | 0.2995 |
|
| 103 |
+
| 4.9552 | 2.19 | 18500 | 4.2977 | 0.3012 |
|
| 104 |
+
| 4.956 | 2.25 | 19000 | 4.2894 | 0.3019 |
|
| 105 |
+
| 4.9427 | 2.31 | 19500 | 4.2781 | 0.3036 |
|
| 106 |
+
| 4.9337 | 2.36 | 20000 | 4.2773 | 0.3038 |
|
| 107 |
+
| 4.9333 | 2.42 | 20500 | 4.2624 | 0.3056 |
|
| 108 |
+
| 4.9173 | 2.48 | 21000 | 4.2643 | 0.3059 |
|
| 109 |
+
| 4.915 | 2.54 | 21500 | 4.2537 | 0.3069 |
|
| 110 |
+
| 4.9092 | 2.6 | 22000 | 4.2457 | 0.3084 |
|
| 111 |
+
| 4.9043 | 2.66 | 22500 | 4.2456 | 0.3081 |
|
| 112 |
+
| 4.9014 | 2.72 | 23000 | 4.2424 | 0.3087 |
|
| 113 |
+
| 4.8889 | 2.78 | 23500 | 4.2347 | 0.3104 |
|
| 114 |
+
| 4.8898 | 2.84 | 24000 | 4.2340 | 0.3095 |
|
| 115 |
+
| 4.8814 | 2.9 | 24500 | 4.2297 | 0.3100 |
|
| 116 |
+
| 4.8804 | 2.96 | 25000 | 4.2290 | 0.3095 |
|
| 117 |
+
|
| 118 |
+
|
| 119 |
+
### Framework versions
|
| 120 |
+
|
| 121 |
+
- Transformers 4.36.2
|
| 122 |
+
- Pytorch 2.2.0+cu121
|
| 123 |
+
- Datasets 2.17.0
|
| 124 |
+
- Tokenizers 0.15.2
|
all_results.json
ADDED
|
@@ -0,0 +1,15 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 3.0,
|
| 3 |
+
"eval_accuracy": 0.3096946377787028,
|
| 4 |
+
"eval_loss": 4.230222225189209,
|
| 5 |
+
"eval_runtime": 94.6692,
|
| 6 |
+
"eval_samples": 16928,
|
| 7 |
+
"eval_samples_per_second": 178.812,
|
| 8 |
+
"eval_steps_per_second": 22.352,
|
| 9 |
+
"perplexity": 68.73250457066099,
|
| 10 |
+
"train_loss": 5.366477503620405,
|
| 11 |
+
"train_runtime": 9239.2947,
|
| 12 |
+
"train_samples": 135339,
|
| 13 |
+
"train_samples_per_second": 43.945,
|
| 14 |
+
"train_steps_per_second": 2.747
|
| 15 |
+
}
|
datasets/all_binary_and_xe_ey_fae_counterfactual/adapter_config.json
ADDED
|
@@ -0,0 +1,41 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"config": {
|
| 3 |
+
"adapter_residual_before_ln": false,
|
| 4 |
+
"cross_adapter": false,
|
| 5 |
+
"factorized_phm_W": true,
|
| 6 |
+
"factorized_phm_rule": false,
|
| 7 |
+
"hypercomplex_nonlinearity": "glorot-uniform",
|
| 8 |
+
"init_weights": "bert",
|
| 9 |
+
"inv_adapter": null,
|
| 10 |
+
"inv_adapter_reduction_factor": null,
|
| 11 |
+
"is_parallel": false,
|
| 12 |
+
"learn_phm": true,
|
| 13 |
+
"leave_out": [],
|
| 14 |
+
"ln_after": false,
|
| 15 |
+
"ln_before": false,
|
| 16 |
+
"mh_adapter": false,
|
| 17 |
+
"non_linearity": "relu",
|
| 18 |
+
"original_ln_after": true,
|
| 19 |
+
"original_ln_before": true,
|
| 20 |
+
"output_adapter": true,
|
| 21 |
+
"phm_bias": true,
|
| 22 |
+
"phm_c_init": "normal",
|
| 23 |
+
"phm_dim": 4,
|
| 24 |
+
"phm_init_range": 0.0001,
|
| 25 |
+
"phm_layer": false,
|
| 26 |
+
"phm_rank": 1,
|
| 27 |
+
"reduction_factor": 16,
|
| 28 |
+
"residual_before_ln": true,
|
| 29 |
+
"scaling": 1.0,
|
| 30 |
+
"shared_W_phm": false,
|
| 31 |
+
"shared_phm_rule": true,
|
| 32 |
+
"use_gating": false
|
| 33 |
+
},
|
| 34 |
+
"config_id": "9076f36a74755ac4",
|
| 35 |
+
"hidden_size": 768,
|
| 36 |
+
"model_class": "BartForConditionalGeneration",
|
| 37 |
+
"model_name": "facebook/bart-base",
|
| 38 |
+
"model_type": "bart",
|
| 39 |
+
"name": "datasets/all_binary_and_xe_ey_fae_counterfactual",
|
| 40 |
+
"version": "0.1.2"
|
| 41 |
+
}
|
datasets/all_binary_and_xe_ey_fae_counterfactual/head_config.json
ADDED
|
@@ -0,0 +1,15 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"config": null,
|
| 3 |
+
"hidden_size": 768,
|
| 4 |
+
"label2id": {
|
| 5 |
+
"LABEL_0": 0,
|
| 6 |
+
"LABEL_1": 1,
|
| 7 |
+
"LABEL_2": 2
|
| 8 |
+
},
|
| 9 |
+
"model_class": "BartForConditionalGeneration",
|
| 10 |
+
"model_name": "facebook/bart-base",
|
| 11 |
+
"model_type": "bart",
|
| 12 |
+
"name": null,
|
| 13 |
+
"num_labels": 3,
|
| 14 |
+
"version": "0.1.2"
|
| 15 |
+
}
|
datasets/all_binary_and_xe_ey_fae_counterfactual/pytorch_adapter.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:fd537533b0c653bcf2b6e451ff533e956883139d3cee87e3f2b00be4c8b6c9dd
|
| 3 |
+
size 3598054
|
datasets/all_binary_and_xe_ey_fae_counterfactual/pytorch_model_head.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:bbca71e86a3e2c371d8bf759d9ee58c7ad16e830743ef402deeebf64852dd8eb
|
| 3 |
+
size 154616666
|
eval_results.json
ADDED
|
@@ -0,0 +1,10 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 3.0,
|
| 3 |
+
"eval_accuracy": 0.3096946377787028,
|
| 4 |
+
"eval_loss": 4.230222225189209,
|
| 5 |
+
"eval_runtime": 94.6692,
|
| 6 |
+
"eval_samples": 16928,
|
| 7 |
+
"eval_samples_per_second": 178.812,
|
| 8 |
+
"eval_steps_per_second": 22.352,
|
| 9 |
+
"perplexity": 68.73250457066099
|
| 10 |
+
}
|
train_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 3.0,
|
| 3 |
+
"train_loss": 5.366477503620405,
|
| 4 |
+
"train_runtime": 9239.2947,
|
| 5 |
+
"train_samples": 135339,
|
| 6 |
+
"train_samples_per_second": 43.945,
|
| 7 |
+
"train_steps_per_second": 2.747
|
| 8 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,780 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 4.228977203369141,
|
| 3 |
+
"best_model_checkpoint": "finetuning/output/bart-adapter-finetuned_xe_ey_fae/checkpoint-25000",
|
| 4 |
+
"epoch": 3.0,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 25377,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.06,
|
| 13 |
+
"learning_rate": 9.804547424833511e-06,
|
| 14 |
+
"loss": 7.6974,
|
| 15 |
+
"step": 500
|
| 16 |
+
},
|
| 17 |
+
{
|
| 18 |
+
"epoch": 0.06,
|
| 19 |
+
"eval_accuracy": 0.06490582319808594,
|
| 20 |
+
"eval_loss": 6.724620819091797,
|
| 21 |
+
"eval_runtime": 94.6343,
|
| 22 |
+
"eval_samples_per_second": 178.878,
|
| 23 |
+
"eval_steps_per_second": 22.36,
|
| 24 |
+
"step": 500
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.12,
|
| 28 |
+
"learning_rate": 9.607912676833355e-06,
|
| 29 |
+
"loss": 6.8017,
|
| 30 |
+
"step": 1000
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.12,
|
| 34 |
+
"eval_accuracy": 0.07624003920813896,
|
| 35 |
+
"eval_loss": 6.406736373901367,
|
| 36 |
+
"eval_runtime": 94.6629,
|
| 37 |
+
"eval_samples_per_second": 178.824,
|
| 38 |
+
"eval_steps_per_second": 22.353,
|
| 39 |
+
"step": 1000
|
| 40 |
+
},
|
| 41 |
+
{
|
| 42 |
+
"epoch": 0.18,
|
| 43 |
+
"learning_rate": 9.410883871221972e-06,
|
| 44 |
+
"loss": 6.5894,
|
| 45 |
+
"step": 1500
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.18,
|
| 49 |
+
"eval_accuracy": 0.08208813335328147,
|
| 50 |
+
"eval_loss": 6.266135215759277,
|
| 51 |
+
"eval_runtime": 94.5601,
|
| 52 |
+
"eval_samples_per_second": 179.019,
|
| 53 |
+
"eval_steps_per_second": 22.377,
|
| 54 |
+
"step": 1500
|
| 55 |
+
},
|
| 56 |
+
{
|
| 57 |
+
"epoch": 0.24,
|
| 58 |
+
"learning_rate": 9.213855065610593e-06,
|
| 59 |
+
"loss": 6.443,
|
| 60 |
+
"step": 2000
|
| 61 |
+
},
|
| 62 |
+
{
|
| 63 |
+
"epoch": 0.24,
|
| 64 |
+
"eval_accuracy": 0.09051518335536303,
|
| 65 |
+
"eval_loss": 6.134974002838135,
|
| 66 |
+
"eval_runtime": 94.5786,
|
| 67 |
+
"eval_samples_per_second": 178.983,
|
| 68 |
+
"eval_steps_per_second": 22.373,
|
| 69 |
+
"step": 2000
|
| 70 |
+
},
|
| 71 |
+
{
|
| 72 |
+
"epoch": 0.3,
|
| 73 |
+
"learning_rate": 9.016826259999214e-06,
|
| 74 |
+
"loss": 6.3245,
|
| 75 |
+
"step": 2500
|
| 76 |
+
},
|
| 77 |
+
{
|
| 78 |
+
"epoch": 0.3,
|
| 79 |
+
"eval_accuracy": 0.10075202490707738,
|
| 80 |
+
"eval_loss": 6.002374172210693,
|
| 81 |
+
"eval_runtime": 94.5343,
|
| 82 |
+
"eval_samples_per_second": 179.067,
|
| 83 |
+
"eval_steps_per_second": 22.383,
|
| 84 |
+
"step": 2500
|
| 85 |
+
},
|
| 86 |
+
{
|
| 87 |
+
"epoch": 0.35,
|
| 88 |
+
"learning_rate": 8.819797454387831e-06,
|
| 89 |
+
"loss": 6.2208,
|
| 90 |
+
"step": 3000
|
| 91 |
+
},
|
| 92 |
+
{
|
| 93 |
+
"epoch": 0.35,
|
| 94 |
+
"eval_accuracy": 0.11453583223288173,
|
| 95 |
+
"eval_loss": 5.851789474487305,
|
| 96 |
+
"eval_runtime": 94.517,
|
| 97 |
+
"eval_samples_per_second": 179.1,
|
| 98 |
+
"eval_steps_per_second": 22.388,
|
| 99 |
+
"step": 3000
|
| 100 |
+
},
|
| 101 |
+
{
|
| 102 |
+
"epoch": 0.41,
|
| 103 |
+
"learning_rate": 8.622768648776452e-06,
|
| 104 |
+
"loss": 6.097,
|
| 105 |
+
"step": 3500
|
| 106 |
+
},
|
| 107 |
+
{
|
| 108 |
+
"epoch": 0.41,
|
| 109 |
+
"eval_accuracy": 0.13295434177816914,
|
| 110 |
+
"eval_loss": 5.658839225769043,
|
| 111 |
+
"eval_runtime": 94.5475,
|
| 112 |
+
"eval_samples_per_second": 179.042,
|
| 113 |
+
"eval_steps_per_second": 22.38,
|
| 114 |
+
"step": 3500
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.47,
|
| 118 |
+
"learning_rate": 8.425739843165071e-06,
|
| 119 |
+
"loss": 5.9862,
|
| 120 |
+
"step": 4000
|
| 121 |
+
},
|
| 122 |
+
{
|
| 123 |
+
"epoch": 0.47,
|
| 124 |
+
"eval_accuracy": 0.15434364158992747,
|
| 125 |
+
"eval_loss": 5.464097023010254,
|
| 126 |
+
"eval_runtime": 94.6204,
|
| 127 |
+
"eval_samples_per_second": 178.904,
|
| 128 |
+
"eval_steps_per_second": 22.363,
|
| 129 |
+
"step": 4000
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 0.53,
|
| 133 |
+
"learning_rate": 8.22871103755369e-06,
|
| 134 |
+
"loss": 5.8742,
|
| 135 |
+
"step": 4500
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.53,
|
| 139 |
+
"eval_accuracy": 0.17065464547534417,
|
| 140 |
+
"eval_loss": 5.319990158081055,
|
| 141 |
+
"eval_runtime": 94.6537,
|
| 142 |
+
"eval_samples_per_second": 178.841,
|
| 143 |
+
"eval_steps_per_second": 22.355,
|
| 144 |
+
"step": 4500
|
| 145 |
+
},
|
| 146 |
+
{
|
| 147 |
+
"epoch": 0.59,
|
| 148 |
+
"learning_rate": 8.03168223194231e-06,
|
| 149 |
+
"loss": 5.7716,
|
| 150 |
+
"step": 5000
|
| 151 |
+
},
|
| 152 |
+
{
|
| 153 |
+
"epoch": 0.59,
|
| 154 |
+
"eval_accuracy": 0.1840178168169571,
|
| 155 |
+
"eval_loss": 5.204446315765381,
|
| 156 |
+
"eval_runtime": 94.6335,
|
| 157 |
+
"eval_samples_per_second": 178.879,
|
| 158 |
+
"eval_steps_per_second": 22.36,
|
| 159 |
+
"step": 5000
|
| 160 |
+
},
|
| 161 |
+
{
|
| 162 |
+
"epoch": 0.65,
|
| 163 |
+
"learning_rate": 7.83465342633093e-06,
|
| 164 |
+
"loss": 5.6952,
|
| 165 |
+
"step": 5500
|
| 166 |
+
},
|
| 167 |
+
{
|
| 168 |
+
"epoch": 0.65,
|
| 169 |
+
"eval_accuracy": 0.19522825907475083,
|
| 170 |
+
"eval_loss": 5.11543607711792,
|
| 171 |
+
"eval_runtime": 94.6161,
|
| 172 |
+
"eval_samples_per_second": 178.913,
|
| 173 |
+
"eval_steps_per_second": 22.364,
|
| 174 |
+
"step": 5500
|
| 175 |
+
},
|
| 176 |
+
{
|
| 177 |
+
"epoch": 0.71,
|
| 178 |
+
"learning_rate": 7.63762462071955e-06,
|
| 179 |
+
"loss": 5.6209,
|
| 180 |
+
"step": 6000
|
| 181 |
+
},
|
| 182 |
+
{
|
| 183 |
+
"epoch": 0.71,
|
| 184 |
+
"eval_accuracy": 0.20442877856808384,
|
| 185 |
+
"eval_loss": 5.042799949645996,
|
| 186 |
+
"eval_runtime": 94.6498,
|
| 187 |
+
"eval_samples_per_second": 178.849,
|
| 188 |
+
"eval_steps_per_second": 22.356,
|
| 189 |
+
"step": 6000
|
| 190 |
+
},
|
| 191 |
+
{
|
| 192 |
+
"epoch": 0.77,
|
| 193 |
+
"learning_rate": 7.440595815108169e-06,
|
| 194 |
+
"loss": 5.5752,
|
| 195 |
+
"step": 6500
|
| 196 |
+
},
|
| 197 |
+
{
|
| 198 |
+
"epoch": 0.77,
|
| 199 |
+
"eval_accuracy": 0.21356098820506358,
|
| 200 |
+
"eval_loss": 4.97105598449707,
|
| 201 |
+
"eval_runtime": 94.647,
|
| 202 |
+
"eval_samples_per_second": 178.854,
|
| 203 |
+
"eval_steps_per_second": 22.357,
|
| 204 |
+
"step": 6500
|
| 205 |
+
},
|
| 206 |
+
{
|
| 207 |
+
"epoch": 0.83,
|
| 208 |
+
"learning_rate": 7.2435670094967895e-06,
|
| 209 |
+
"loss": 5.5091,
|
| 210 |
+
"step": 7000
|
| 211 |
+
},
|
| 212 |
+
{
|
| 213 |
+
"epoch": 0.83,
|
| 214 |
+
"eval_accuracy": 0.2211917808858752,
|
| 215 |
+
"eval_loss": 4.907817363739014,
|
| 216 |
+
"eval_runtime": 94.6132,
|
| 217 |
+
"eval_samples_per_second": 178.918,
|
| 218 |
+
"eval_steps_per_second": 22.365,
|
| 219 |
+
"step": 7000
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.89,
|
| 223 |
+
"learning_rate": 7.0465382038854095e-06,
|
| 224 |
+
"loss": 5.4657,
|
| 225 |
+
"step": 7500
|
| 226 |
+
},
|
| 227 |
+
{
|
| 228 |
+
"epoch": 0.89,
|
| 229 |
+
"eval_accuracy": 0.22866674452100122,
|
| 230 |
+
"eval_loss": 4.849499225616455,
|
| 231 |
+
"eval_runtime": 94.6751,
|
| 232 |
+
"eval_samples_per_second": 178.801,
|
| 233 |
+
"eval_steps_per_second": 22.35,
|
| 234 |
+
"step": 7500
|
| 235 |
+
},
|
| 236 |
+
{
|
| 237 |
+
"epoch": 0.95,
|
| 238 |
+
"learning_rate": 6.849509398274028e-06,
|
| 239 |
+
"loss": 5.4245,
|
| 240 |
+
"step": 8000
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 0.95,
|
| 244 |
+
"eval_accuracy": 0.2359544531397667,
|
| 245 |
+
"eval_loss": 4.801201343536377,
|
| 246 |
+
"eval_runtime": 94.5944,
|
| 247 |
+
"eval_samples_per_second": 178.954,
|
| 248 |
+
"eval_steps_per_second": 22.369,
|
| 249 |
+
"step": 8000
|
| 250 |
+
},
|
| 251 |
+
{
|
| 252 |
+
"epoch": 1.0,
|
| 253 |
+
"learning_rate": 6.652480592662648e-06,
|
| 254 |
+
"loss": 5.3813,
|
| 255 |
+
"step": 8500
|
| 256 |
+
},
|
| 257 |
+
{
|
| 258 |
+
"epoch": 1.0,
|
| 259 |
+
"eval_accuracy": 0.24091048075194954,
|
| 260 |
+
"eval_loss": 4.756309986114502,
|
| 261 |
+
"eval_runtime": 94.6708,
|
| 262 |
+
"eval_samples_per_second": 178.809,
|
| 263 |
+
"eval_steps_per_second": 22.351,
|
| 264 |
+
"step": 8500
|
| 265 |
+
},
|
| 266 |
+
{
|
| 267 |
+
"epoch": 1.06,
|
| 268 |
+
"learning_rate": 6.4554517870512675e-06,
|
| 269 |
+
"loss": 5.3501,
|
| 270 |
+
"step": 9000
|
| 271 |
+
},
|
| 272 |
+
{
|
| 273 |
+
"epoch": 1.06,
|
| 274 |
+
"eval_accuracy": 0.24637192061403781,
|
| 275 |
+
"eval_loss": 4.716611862182617,
|
| 276 |
+
"eval_runtime": 94.6539,
|
| 277 |
+
"eval_samples_per_second": 178.841,
|
| 278 |
+
"eval_steps_per_second": 22.355,
|
| 279 |
+
"step": 9000
|
| 280 |
+
},
|
| 281 |
+
{
|
| 282 |
+
"epoch": 1.12,
|
| 283 |
+
"learning_rate": 6.2584229814398866e-06,
|
| 284 |
+
"loss": 5.3098,
|
| 285 |
+
"step": 9500
|
| 286 |
+
},
|
| 287 |
+
{
|
| 288 |
+
"epoch": 1.12,
|
| 289 |
+
"eval_accuracy": 0.2501098972917417,
|
| 290 |
+
"eval_loss": 4.683783054351807,
|
| 291 |
+
"eval_runtime": 94.7201,
|
| 292 |
+
"eval_samples_per_second": 178.716,
|
| 293 |
+
"eval_steps_per_second": 22.339,
|
| 294 |
+
"step": 9500
|
| 295 |
+
},
|
| 296 |
+
{
|
| 297 |
+
"epoch": 1.18,
|
| 298 |
+
"learning_rate": 6.0613941758285065e-06,
|
| 299 |
+
"loss": 5.2856,
|
| 300 |
+
"step": 10000
|
| 301 |
+
},
|
| 302 |
+
{
|
| 303 |
+
"epoch": 1.18,
|
| 304 |
+
"eval_accuracy": 0.2551260313174862,
|
| 305 |
+
"eval_loss": 4.651512622833252,
|
| 306 |
+
"eval_runtime": 94.7148,
|
| 307 |
+
"eval_samples_per_second": 178.726,
|
| 308 |
+
"eval_steps_per_second": 22.341,
|
| 309 |
+
"step": 10000
|
| 310 |
+
},
|
| 311 |
+
{
|
| 312 |
+
"epoch": 1.24,
|
| 313 |
+
"learning_rate": 5.864365370217126e-06,
|
| 314 |
+
"loss": 5.2549,
|
| 315 |
+
"step": 10500
|
| 316 |
+
},
|
| 317 |
+
{
|
| 318 |
+
"epoch": 1.24,
|
| 319 |
+
"eval_accuracy": 0.2601890994584574,
|
| 320 |
+
"eval_loss": 4.612128257751465,
|
| 321 |
+
"eval_runtime": 94.6453,
|
| 322 |
+
"eval_samples_per_second": 178.857,
|
| 323 |
+
"eval_steps_per_second": 22.357,
|
| 324 |
+
"step": 10500
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 1.3,
|
| 328 |
+
"learning_rate": 5.6673365646057455e-06,
|
| 329 |
+
"loss": 5.2217,
|
| 330 |
+
"step": 11000
|
| 331 |
+
},
|
| 332 |
+
{
|
| 333 |
+
"epoch": 1.3,
|
| 334 |
+
"eval_accuracy": 0.2636951256551434,
|
| 335 |
+
"eval_loss": 4.5841240882873535,
|
| 336 |
+
"eval_runtime": 94.6678,
|
| 337 |
+
"eval_samples_per_second": 178.815,
|
| 338 |
+
"eval_steps_per_second": 22.352,
|
| 339 |
+
"step": 11000
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 1.36,
|
| 343 |
+
"learning_rate": 5.470701816605588e-06,
|
| 344 |
+
"loss": 5.1997,
|
| 345 |
+
"step": 11500
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 1.36,
|
| 349 |
+
"eval_accuracy": 0.267359922045739,
|
| 350 |
+
"eval_loss": 4.558795928955078,
|
| 351 |
+
"eval_runtime": 94.6574,
|
| 352 |
+
"eval_samples_per_second": 178.834,
|
| 353 |
+
"eval_steps_per_second": 22.354,
|
| 354 |
+
"step": 11500
|
| 355 |
+
},
|
| 356 |
+
{
|
| 357 |
+
"epoch": 1.42,
|
| 358 |
+
"learning_rate": 5.273673010994208e-06,
|
| 359 |
+
"loss": 5.1844,
|
| 360 |
+
"step": 12000
|
| 361 |
+
},
|
| 362 |
+
{
|
| 363 |
+
"epoch": 1.42,
|
| 364 |
+
"eval_accuracy": 0.27077985002194555,
|
| 365 |
+
"eval_loss": 4.530928611755371,
|
| 366 |
+
"eval_runtime": 94.6472,
|
| 367 |
+
"eval_samples_per_second": 178.854,
|
| 368 |
+
"eval_steps_per_second": 22.357,
|
| 369 |
+
"step": 12000
|
| 370 |
+
},
|
| 371 |
+
{
|
| 372 |
+
"epoch": 1.48,
|
| 373 |
+
"learning_rate": 5.076644205382827e-06,
|
| 374 |
+
"loss": 5.1491,
|
| 375 |
+
"step": 12500
|
| 376 |
+
},
|
| 377 |
+
{
|
| 378 |
+
"epoch": 1.48,
|
| 379 |
+
"eval_accuracy": 0.27483096084023484,
|
| 380 |
+
"eval_loss": 4.499913692474365,
|
| 381 |
+
"eval_runtime": 94.5904,
|
| 382 |
+
"eval_samples_per_second": 178.961,
|
| 383 |
+
"eval_steps_per_second": 22.37,
|
| 384 |
+
"step": 12500
|
| 385 |
+
},
|
| 386 |
+
{
|
| 387 |
+
"epoch": 1.54,
|
| 388 |
+
"learning_rate": 4.879615399771447e-06,
|
| 389 |
+
"loss": 5.1244,
|
| 390 |
+
"step": 13000
|
| 391 |
+
},
|
| 392 |
+
{
|
| 393 |
+
"epoch": 1.54,
|
| 394 |
+
"eval_accuracy": 0.27797822970103025,
|
| 395 |
+
"eval_loss": 4.47827672958374,
|
| 396 |
+
"eval_runtime": 94.6192,
|
| 397 |
+
"eval_samples_per_second": 178.907,
|
| 398 |
+
"eval_steps_per_second": 22.363,
|
| 399 |
+
"step": 13000
|
| 400 |
+
},
|
| 401 |
+
{
|
| 402 |
+
"epoch": 1.6,
|
| 403 |
+
"learning_rate": 4.6825865941600665e-06,
|
| 404 |
+
"loss": 5.1047,
|
| 405 |
+
"step": 13500
|
| 406 |
+
},
|
| 407 |
+
{
|
| 408 |
+
"epoch": 1.6,
|
| 409 |
+
"eval_accuracy": 0.28116894803558673,
|
| 410 |
+
"eval_loss": 4.456052780151367,
|
| 411 |
+
"eval_runtime": 94.6508,
|
| 412 |
+
"eval_samples_per_second": 178.847,
|
| 413 |
+
"eval_steps_per_second": 22.356,
|
| 414 |
+
"step": 13500
|
| 415 |
+
},
|
| 416 |
+
{
|
| 417 |
+
"epoch": 1.66,
|
| 418 |
+
"learning_rate": 4.485557788548686e-06,
|
| 419 |
+
"loss": 5.0917,
|
| 420 |
+
"step": 14000
|
| 421 |
+
},
|
| 422 |
+
{
|
| 423 |
+
"epoch": 1.66,
|
| 424 |
+
"eval_accuracy": 0.28257652944074685,
|
| 425 |
+
"eval_loss": 4.440920352935791,
|
| 426 |
+
"eval_runtime": 94.6448,
|
| 427 |
+
"eval_samples_per_second": 178.858,
|
| 428 |
+
"eval_steps_per_second": 22.357,
|
| 429 |
+
"step": 14000
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 1.71,
|
| 433 |
+
"learning_rate": 4.2885289829373055e-06,
|
| 434 |
+
"loss": 5.0631,
|
| 435 |
+
"step": 14500
|
| 436 |
+
},
|
| 437 |
+
{
|
| 438 |
+
"epoch": 1.71,
|
| 439 |
+
"eval_accuracy": 0.2851284480287185,
|
| 440 |
+
"eval_loss": 4.4198174476623535,
|
| 441 |
+
"eval_runtime": 94.6157,
|
| 442 |
+
"eval_samples_per_second": 178.913,
|
| 443 |
+
"eval_steps_per_second": 22.364,
|
| 444 |
+
"step": 14500
|
| 445 |
+
},
|
| 446 |
+
{
|
| 447 |
+
"epoch": 1.77,
|
| 448 |
+
"learning_rate": 4.091500177325925e-06,
|
| 449 |
+
"loss": 5.0537,
|
| 450 |
+
"step": 15000
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"epoch": 1.77,
|
| 454 |
+
"eval_accuracy": 0.2881034378022738,
|
| 455 |
+
"eval_loss": 4.400303840637207,
|
| 456 |
+
"eval_runtime": 94.6637,
|
| 457 |
+
"eval_samples_per_second": 178.823,
|
| 458 |
+
"eval_steps_per_second": 22.353,
|
| 459 |
+
"step": 15000
|
| 460 |
+
},
|
| 461 |
+
{
|
| 462 |
+
"epoch": 1.83,
|
| 463 |
+
"learning_rate": 3.894865429325768e-06,
|
| 464 |
+
"loss": 5.0339,
|
| 465 |
+
"step": 15500
|
| 466 |
+
},
|
| 467 |
+
{
|
| 468 |
+
"epoch": 1.83,
|
| 469 |
+
"eval_accuracy": 0.2898631699682636,
|
| 470 |
+
"eval_loss": 4.385478973388672,
|
| 471 |
+
"eval_runtime": 94.7914,
|
| 472 |
+
"eval_samples_per_second": 178.582,
|
| 473 |
+
"eval_steps_per_second": 22.323,
|
| 474 |
+
"step": 15500
|
| 475 |
+
},
|
| 476 |
+
{
|
| 477 |
+
"epoch": 1.89,
|
| 478 |
+
"learning_rate": 3.6978366237143875e-06,
|
| 479 |
+
"loss": 5.0235,
|
| 480 |
+
"step": 16000
|
| 481 |
+
},
|
| 482 |
+
{
|
| 483 |
+
"epoch": 1.89,
|
| 484 |
+
"eval_accuracy": 0.29213215439190976,
|
| 485 |
+
"eval_loss": 4.3650221824646,
|
| 486 |
+
"eval_runtime": 94.7339,
|
| 487 |
+
"eval_samples_per_second": 178.69,
|
| 488 |
+
"eval_steps_per_second": 22.336,
|
| 489 |
+
"step": 16000
|
| 490 |
+
},
|
| 491 |
+
{
|
| 492 |
+
"epoch": 1.95,
|
| 493 |
+
"learning_rate": 3.500807818103007e-06,
|
| 494 |
+
"loss": 5.0074,
|
| 495 |
+
"step": 16500
|
| 496 |
+
},
|
| 497 |
+
{
|
| 498 |
+
"epoch": 1.95,
|
| 499 |
+
"eval_accuracy": 0.2941743633357399,
|
| 500 |
+
"eval_loss": 4.349618911743164,
|
| 501 |
+
"eval_runtime": 94.7186,
|
| 502 |
+
"eval_samples_per_second": 178.719,
|
| 503 |
+
"eval_steps_per_second": 22.34,
|
| 504 |
+
"step": 16500
|
| 505 |
+
},
|
| 506 |
+
{
|
| 507 |
+
"epoch": 2.01,
|
| 508 |
+
"learning_rate": 3.304173070102849e-06,
|
| 509 |
+
"loss": 4.9927,
|
| 510 |
+
"step": 17000
|
| 511 |
+
},
|
| 512 |
+
{
|
| 513 |
+
"epoch": 2.01,
|
| 514 |
+
"eval_accuracy": 0.2964732381378149,
|
| 515 |
+
"eval_loss": 4.336081027984619,
|
| 516 |
+
"eval_runtime": 94.7044,
|
| 517 |
+
"eval_samples_per_second": 178.746,
|
| 518 |
+
"eval_steps_per_second": 22.343,
|
| 519 |
+
"step": 17000
|
| 520 |
+
},
|
| 521 |
+
{
|
| 522 |
+
"epoch": 2.07,
|
| 523 |
+
"learning_rate": 3.107144264491469e-06,
|
| 524 |
+
"loss": 4.9797,
|
| 525 |
+
"step": 17500
|
| 526 |
+
},
|
| 527 |
+
{
|
| 528 |
+
"epoch": 2.07,
|
| 529 |
+
"eval_accuracy": 0.2981180855432632,
|
| 530 |
+
"eval_loss": 4.320304870605469,
|
| 531 |
+
"eval_runtime": 94.752,
|
| 532 |
+
"eval_samples_per_second": 178.656,
|
| 533 |
+
"eval_steps_per_second": 22.332,
|
| 534 |
+
"step": 17500
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 2.13,
|
| 538 |
+
"learning_rate": 2.9101154588800883e-06,
|
| 539 |
+
"loss": 4.9725,
|
| 540 |
+
"step": 18000
|
| 541 |
+
},
|
| 542 |
+
{
|
| 543 |
+
"epoch": 2.13,
|
| 544 |
+
"eval_accuracy": 0.2994866192996549,
|
| 545 |
+
"eval_loss": 4.311823844909668,
|
| 546 |
+
"eval_runtime": 94.7748,
|
| 547 |
+
"eval_samples_per_second": 178.613,
|
| 548 |
+
"eval_steps_per_second": 22.327,
|
| 549 |
+
"step": 18000
|
| 550 |
+
},
|
| 551 |
+
{
|
| 552 |
+
"epoch": 2.19,
|
| 553 |
+
"learning_rate": 2.713086653268708e-06,
|
| 554 |
+
"loss": 4.9552,
|
| 555 |
+
"step": 18500
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 2.19,
|
| 559 |
+
"eval_accuracy": 0.3011751340054841,
|
| 560 |
+
"eval_loss": 4.297677516937256,
|
| 561 |
+
"eval_runtime": 94.6902,
|
| 562 |
+
"eval_samples_per_second": 178.772,
|
| 563 |
+
"eval_steps_per_second": 22.347,
|
| 564 |
+
"step": 18500
|
| 565 |
+
},
|
| 566 |
+
{
|
| 567 |
+
"epoch": 2.25,
|
| 568 |
+
"learning_rate": 2.5160578476573277e-06,
|
| 569 |
+
"loss": 4.956,
|
| 570 |
+
"step": 19000
|
| 571 |
+
},
|
| 572 |
+
{
|
| 573 |
+
"epoch": 2.25,
|
| 574 |
+
"eval_accuracy": 0.3018521891702124,
|
| 575 |
+
"eval_loss": 4.289434909820557,
|
| 576 |
+
"eval_runtime": 94.6916,
|
| 577 |
+
"eval_samples_per_second": 178.77,
|
| 578 |
+
"eval_steps_per_second": 22.346,
|
| 579 |
+
"step": 19000
|
| 580 |
+
},
|
| 581 |
+
{
|
| 582 |
+
"epoch": 2.31,
|
| 583 |
+
"learning_rate": 2.3190290420459476e-06,
|
| 584 |
+
"loss": 4.9427,
|
| 585 |
+
"step": 19500
|
| 586 |
+
},
|
| 587 |
+
{
|
| 588 |
+
"epoch": 2.31,
|
| 589 |
+
"eval_accuracy": 0.30359935941560284,
|
| 590 |
+
"eval_loss": 4.278063774108887,
|
| 591 |
+
"eval_runtime": 94.7241,
|
| 592 |
+
"eval_samples_per_second": 178.709,
|
| 593 |
+
"eval_steps_per_second": 22.339,
|
| 594 |
+
"step": 19500
|
| 595 |
+
},
|
| 596 |
+
{
|
| 597 |
+
"epoch": 2.36,
|
| 598 |
+
"learning_rate": 2.12239429404579e-06,
|
| 599 |
+
"loss": 4.9337,
|
| 600 |
+
"step": 20000
|
| 601 |
+
},
|
| 602 |
+
{
|
| 603 |
+
"epoch": 2.36,
|
| 604 |
+
"eval_accuracy": 0.30382226071056156,
|
| 605 |
+
"eval_loss": 4.277279853820801,
|
| 606 |
+
"eval_runtime": 94.6289,
|
| 607 |
+
"eval_samples_per_second": 178.888,
|
| 608 |
+
"eval_steps_per_second": 22.361,
|
| 609 |
+
"step": 20000
|
| 610 |
+
},
|
| 611 |
+
{
|
| 612 |
+
"epoch": 2.42,
|
| 613 |
+
"learning_rate": 1.9253654884344093e-06,
|
| 614 |
+
"loss": 4.9333,
|
| 615 |
+
"step": 20500
|
| 616 |
+
},
|
| 617 |
+
{
|
| 618 |
+
"epoch": 2.42,
|
| 619 |
+
"eval_accuracy": 0.305628714892609,
|
| 620 |
+
"eval_loss": 4.262350559234619,
|
| 621 |
+
"eval_runtime": 94.5958,
|
| 622 |
+
"eval_samples_per_second": 178.951,
|
| 623 |
+
"eval_steps_per_second": 22.369,
|
| 624 |
+
"step": 20500
|
| 625 |
+
},
|
| 626 |
+
{
|
| 627 |
+
"epoch": 2.48,
|
| 628 |
+
"learning_rate": 1.7283366828230288e-06,
|
| 629 |
+
"loss": 4.9173,
|
| 630 |
+
"step": 21000
|
| 631 |
+
},
|
| 632 |
+
{
|
| 633 |
+
"epoch": 2.48,
|
| 634 |
+
"eval_accuracy": 0.30594861364613246,
|
| 635 |
+
"eval_loss": 4.26426887512207,
|
| 636 |
+
"eval_runtime": 94.6851,
|
| 637 |
+
"eval_samples_per_second": 178.782,
|
| 638 |
+
"eval_steps_per_second": 22.348,
|
| 639 |
+
"step": 21000
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 2.54,
|
| 643 |
+
"learning_rate": 1.5313078772116485e-06,
|
| 644 |
+
"loss": 4.915,
|
| 645 |
+
"step": 21500
|
| 646 |
+
},
|
| 647 |
+
{
|
| 648 |
+
"epoch": 2.54,
|
| 649 |
+
"eval_accuracy": 0.3068805715197335,
|
| 650 |
+
"eval_loss": 4.253678321838379,
|
| 651 |
+
"eval_runtime": 94.5821,
|
| 652 |
+
"eval_samples_per_second": 178.977,
|
| 653 |
+
"eval_steps_per_second": 22.372,
|
| 654 |
+
"step": 21500
|
| 655 |
+
},
|
| 656 |
+
{
|
| 657 |
+
"epoch": 2.6,
|
| 658 |
+
"learning_rate": 1.3342790716002682e-06,
|
| 659 |
+
"loss": 4.9092,
|
| 660 |
+
"step": 22000
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"epoch": 2.6,
|
| 664 |
+
"eval_accuracy": 0.30841780910899225,
|
| 665 |
+
"eval_loss": 4.245660781860352,
|
| 666 |
+
"eval_runtime": 94.6548,
|
| 667 |
+
"eval_samples_per_second": 178.839,
|
| 668 |
+
"eval_steps_per_second": 22.355,
|
| 669 |
+
"step": 22000
|
| 670 |
+
},
|
| 671 |
+
{
|
| 672 |
+
"epoch": 2.66,
|
| 673 |
+
"learning_rate": 1.1372502659888877e-06,
|
| 674 |
+
"loss": 4.9043,
|
| 675 |
+
"step": 22500
|
| 676 |
+
},
|
| 677 |
+
{
|
| 678 |
+
"epoch": 2.66,
|
| 679 |
+
"eval_accuracy": 0.3080689795368392,
|
| 680 |
+
"eval_loss": 4.245626926422119,
|
| 681 |
+
"eval_runtime": 94.7104,
|
| 682 |
+
"eval_samples_per_second": 178.734,
|
| 683 |
+
"eval_steps_per_second": 22.342,
|
| 684 |
+
"step": 22500
|
| 685 |
+
},
|
| 686 |
+
{
|
| 687 |
+
"epoch": 2.72,
|
| 688 |
+
"learning_rate": 9.402214603775073e-07,
|
| 689 |
+
"loss": 4.9014,
|
| 690 |
+
"step": 23000
|
| 691 |
+
},
|
| 692 |
+
{
|
| 693 |
+
"epoch": 2.72,
|
| 694 |
+
"eval_accuracy": 0.3087008130169778,
|
| 695 |
+
"eval_loss": 4.24239444732666,
|
| 696 |
+
"eval_runtime": 94.6532,
|
| 697 |
+
"eval_samples_per_second": 178.842,
|
| 698 |
+
"eval_steps_per_second": 22.355,
|
| 699 |
+
"step": 23000
|
| 700 |
+
},
|
| 701 |
+
{
|
| 702 |
+
"epoch": 2.78,
|
| 703 |
+
"learning_rate": 7.431926547661269e-07,
|
| 704 |
+
"loss": 4.8889,
|
| 705 |
+
"step": 23500
|
| 706 |
+
},
|
| 707 |
+
{
|
| 708 |
+
"epoch": 2.78,
|
| 709 |
+
"eval_accuracy": 0.3103718452365669,
|
| 710 |
+
"eval_loss": 4.2347259521484375,
|
| 711 |
+
"eval_runtime": 94.6443,
|
| 712 |
+
"eval_samples_per_second": 178.859,
|
| 713 |
+
"eval_steps_per_second": 22.357,
|
| 714 |
+
"step": 23500
|
| 715 |
+
},
|
| 716 |
+
{
|
| 717 |
+
"epoch": 2.84,
|
| 718 |
+
"learning_rate": 5.461638491547465e-07,
|
| 719 |
+
"loss": 4.8898,
|
| 720 |
+
"step": 24000
|
| 721 |
+
},
|
| 722 |
+
{
|
| 723 |
+
"epoch": 2.84,
|
| 724 |
+
"eval_accuracy": 0.30947218110795954,
|
| 725 |
+
"eval_loss": 4.233951091766357,
|
| 726 |
+
"eval_runtime": 94.5592,
|
| 727 |
+
"eval_samples_per_second": 179.02,
|
| 728 |
+
"eval_steps_per_second": 22.378,
|
| 729 |
+
"step": 24000
|
| 730 |
+
},
|
| 731 |
+
{
|
| 732 |
+
"epoch": 2.9,
|
| 733 |
+
"learning_rate": 3.4952910115458886e-07,
|
| 734 |
+
"loss": 4.8814,
|
| 735 |
+
"step": 24500
|
| 736 |
+
},
|
| 737 |
+
{
|
| 738 |
+
"epoch": 2.9,
|
| 739 |
+
"eval_accuracy": 0.31004912351666813,
|
| 740 |
+
"eval_loss": 4.2296977043151855,
|
| 741 |
+
"eval_runtime": 94.6234,
|
| 742 |
+
"eval_samples_per_second": 178.899,
|
| 743 |
+
"eval_steps_per_second": 22.362,
|
| 744 |
+
"step": 24500
|
| 745 |
+
},
|
| 746 |
+
{
|
| 747 |
+
"epoch": 2.96,
|
| 748 |
+
"learning_rate": 1.5250029554320843e-07,
|
| 749 |
+
"loss": 4.8804,
|
| 750 |
+
"step": 25000
|
| 751 |
+
},
|
| 752 |
+
{
|
| 753 |
+
"epoch": 2.96,
|
| 754 |
+
"eval_accuracy": 0.3095025688930406,
|
| 755 |
+
"eval_loss": 4.228977203369141,
|
| 756 |
+
"eval_runtime": 94.5894,
|
| 757 |
+
"eval_samples_per_second": 178.963,
|
| 758 |
+
"eval_steps_per_second": 22.37,
|
| 759 |
+
"step": 25000
|
| 760 |
+
},
|
| 761 |
+
{
|
| 762 |
+
"epoch": 3.0,
|
| 763 |
+
"step": 25377,
|
| 764 |
+
"total_flos": 1.2489741444494131e+17,
|
| 765 |
+
"train_loss": 5.366477503620405,
|
| 766 |
+
"train_runtime": 9239.2947,
|
| 767 |
+
"train_samples_per_second": 43.945,
|
| 768 |
+
"train_steps_per_second": 2.747
|
| 769 |
+
}
|
| 770 |
+
],
|
| 771 |
+
"logging_steps": 500,
|
| 772 |
+
"max_steps": 25377,
|
| 773 |
+
"num_input_tokens_seen": 0,
|
| 774 |
+
"num_train_epochs": 3,
|
| 775 |
+
"save_steps": 500,
|
| 776 |
+
"total_flos": 1.2489741444494131e+17,
|
| 777 |
+
"train_batch_size": 8,
|
| 778 |
+
"trial_name": null,
|
| 779 |
+
"trial_params": null
|
| 780 |
+
}
|