Upload with huggingface_hub
Browse filesThis view is limited to 50 files because it contains too many changes.
See raw diff
- README.md +29 -3
- best_ckpt.pth +3 -0
- checkpoint-1421/added_tokens.json +3 -0
- checkpoint-1421/config.json +35 -0
- checkpoint-1421/optimizer.pt +3 -0
- checkpoint-1421/pytorch_model.bin +3 -0
- checkpoint-1421/rng_state.pth +3 -0
- checkpoint-1421/scaler.pt +3 -0
- checkpoint-1421/scheduler.pt +3 -0
- checkpoint-1421/special_tokens_map.json +9 -0
- checkpoint-1421/spm.model +3 -0
- checkpoint-1421/tokenizer.json +0 -0
- checkpoint-1421/tokenizer_config.json +16 -0
- checkpoint-1421/trainer_state.json +56 -0
- checkpoint-1421/training_args.bin +3 -0
- checkpoint-2842/added_tokens.json +3 -0
- checkpoint-2842/config.json +35 -0
- checkpoint-2842/optimizer.pt +3 -0
- checkpoint-2842/pytorch_model.bin +3 -0
- checkpoint-2842/rng_state.pth +3 -0
- checkpoint-2842/scaler.pt +3 -0
- checkpoint-2842/scheduler.pt +3 -0
- checkpoint-2842/special_tokens_map.json +9 -0
- checkpoint-2842/spm.model +3 -0
- checkpoint-2842/tokenizer.json +0 -0
- checkpoint-2842/tokenizer_config.json +16 -0
- checkpoint-2842/trainer_state.json +102 -0
- checkpoint-2842/training_args.bin +3 -0
- checkpoint-4263/added_tokens.json +3 -0
- checkpoint-4263/config.json +35 -0
- checkpoint-4263/optimizer.pt +3 -0
- checkpoint-4263/pytorch_model.bin +3 -0
- checkpoint-4263/rng_state.pth +3 -0
- checkpoint-4263/scaler.pt +3 -0
- checkpoint-4263/scheduler.pt +3 -0
- checkpoint-4263/special_tokens_map.json +9 -0
- checkpoint-4263/spm.model +3 -0
- checkpoint-4263/tokenizer.json +0 -0
- checkpoint-4263/tokenizer_config.json +16 -0
- checkpoint-4263/trainer_state.json +148 -0
- checkpoint-4263/training_args.bin +3 -0
- checkpoint-5684/added_tokens.json +3 -0
- checkpoint-5684/config.json +35 -0
- checkpoint-5684/optimizer.pt +3 -0
- checkpoint-5684/pytorch_model.bin +3 -0
- checkpoint-5684/rng_state.pth +3 -0
- checkpoint-5684/scaler.pt +3 -0
- checkpoint-5684/scheduler.pt +3 -0
- checkpoint-5684/special_tokens_map.json +9 -0
- checkpoint-5684/spm.model +3 -0
README.md
CHANGED
@@ -1,3 +1,29 @@
|
|
1 |
-
|
2 |
-
|
3 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
A fine-tuned model based on the **DeBERTaV3** model of Microsoft and fine-tuned on **Glue QQP**, which detects the linguistical similarities between two questions and whether they are similar questions or dissimilar.
|
2 |
+
|
3 |
+
## Model Hyperparameters
|
4 |
+
|
5 |
+
```python
|
6 |
+
epoch=4
|
7 |
+
per_device_train_batch_size=32
|
8 |
+
per_device_eval_batch_size=16
|
9 |
+
lr=2e-5
|
10 |
+
weight_decay=1e-2
|
11 |
+
gradient_checkpointing=True
|
12 |
+
gradient_accumulation_steps=8
|
13 |
+
```
|
14 |
+
## Model Performance
|
15 |
+
|
16 |
+
```JSON
|
17 |
+
{"Training Loss": 0.132400,
|
18 |
+
"Validation Loss": 0.217410,
|
19 |
+
"Validation Accuracy": 0.917969
|
20 |
+
}
|
21 |
+
```
|
22 |
+
|
23 |
+
## Model Dependencies
|
24 |
+
|
25 |
+
```JSON
|
26 |
+
{"Main Model": "microsoft/deberta-v3-base",
|
27 |
+
"Dataset": "SetFit/qqp"
|
28 |
+
}
|
29 |
+
```
|
best_ckpt.pth
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:3df78bb0e84782f642a9b1bef8de67169c388b3693ee48b1d9220349dfeed756
|
3 |
+
size 737782699
|
checkpoint-1421/added_tokens.json
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"[MASK]": 128000
|
3 |
+
}
|
checkpoint-1421/config.json
ADDED
@@ -0,0 +1,35 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"_name_or_path": "microsoft/deberta-v3-base",
|
3 |
+
"architectures": [
|
4 |
+
"DebertaV2ForSequenceClassification"
|
5 |
+
],
|
6 |
+
"attention_probs_dropout_prob": 0.1,
|
7 |
+
"hidden_act": "gelu",
|
8 |
+
"hidden_dropout_prob": 0.1,
|
9 |
+
"hidden_size": 768,
|
10 |
+
"initializer_range": 0.02,
|
11 |
+
"intermediate_size": 3072,
|
12 |
+
"layer_norm_eps": 1e-07,
|
13 |
+
"max_position_embeddings": 512,
|
14 |
+
"max_relative_positions": -1,
|
15 |
+
"model_type": "deberta-v2",
|
16 |
+
"norm_rel_ebd": "layer_norm",
|
17 |
+
"num_attention_heads": 12,
|
18 |
+
"num_hidden_layers": 12,
|
19 |
+
"pad_token_id": 0,
|
20 |
+
"pooler_dropout": 0,
|
21 |
+
"pooler_hidden_act": "gelu",
|
22 |
+
"pooler_hidden_size": 768,
|
23 |
+
"pos_att_type": [
|
24 |
+
"p2c",
|
25 |
+
"c2p"
|
26 |
+
],
|
27 |
+
"position_biased_input": false,
|
28 |
+
"position_buckets": 256,
|
29 |
+
"relative_attention": true,
|
30 |
+
"share_att_key": true,
|
31 |
+
"torch_dtype": "float32",
|
32 |
+
"transformers_version": "4.24.0",
|
33 |
+
"type_vocab_size": 0,
|
34 |
+
"vocab_size": 128100
|
35 |
+
}
|
checkpoint-1421/optimizer.pt
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:fd1a50d4937b9f21324e5ac8f11fa07c5b240587076d0a49af86568b810846d1
|
3 |
+
size 1475556165
|
checkpoint-1421/pytorch_model.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:1287312a4a834efa99b26fe3cb9326cd4719a10068875db89b53dab2f4526f72
|
3 |
+
size 737766955
|
checkpoint-1421/rng_state.pth
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:64c71667ddea107636a065cbc131d35d75bf32d9f87fa554b4fba9e2d87200d5
|
3 |
+
size 14503
|
checkpoint-1421/scaler.pt
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:7afb333ba9bc50bd3301ebb72540795fdde8e265f3b435eef901a1d50a5a8c03
|
3 |
+
size 559
|
checkpoint-1421/scheduler.pt
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:552ec4a0a5613f610ace5350aff6aaf73695ecc08b302f92f760425d2a7a335f
|
3 |
+
size 559
|
checkpoint-1421/special_tokens_map.json
ADDED
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"bos_token": "[CLS]",
|
3 |
+
"cls_token": "[CLS]",
|
4 |
+
"eos_token": "[SEP]",
|
5 |
+
"mask_token": "[MASK]",
|
6 |
+
"pad_token": "[PAD]",
|
7 |
+
"sep_token": "[SEP]",
|
8 |
+
"unk_token": "[UNK]"
|
9 |
+
}
|
checkpoint-1421/spm.model
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:c679fbf93643d19aab7ee10c0b99e460bdbc02fedf34b92b05af343b4af586fd
|
3 |
+
size 2464616
|
checkpoint-1421/tokenizer.json
ADDED
The diff for this file is too large to render.
See raw diff
|
|
checkpoint-1421/tokenizer_config.json
ADDED
@@ -0,0 +1,16 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"bos_token": "[CLS]",
|
3 |
+
"cls_token": "[CLS]",
|
4 |
+
"do_lower_case": false,
|
5 |
+
"eos_token": "[SEP]",
|
6 |
+
"mask_token": "[MASK]",
|
7 |
+
"name_or_path": "microsoft/deberta-v3-base",
|
8 |
+
"pad_token": "[PAD]",
|
9 |
+
"sep_token": "[SEP]",
|
10 |
+
"sp_model_kwargs": {},
|
11 |
+
"special_tokens_map_file": null,
|
12 |
+
"split_by_punct": false,
|
13 |
+
"tokenizer_class": "DebertaV2Tokenizer",
|
14 |
+
"unk_token": "[UNK]",
|
15 |
+
"vocab_type": "spm"
|
16 |
+
}
|
checkpoint-1421/trainer_state.json
ADDED
@@ -0,0 +1,56 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"best_metric": 0.22764553129673004,
|
3 |
+
"best_model_checkpoint": "deberta-v3-base-funetuned-qqa/checkpoint-1421",
|
4 |
+
"epoch": 0.9997361709612171,
|
5 |
+
"global_step": 1421,
|
6 |
+
"is_hyper_param_search": false,
|
7 |
+
"is_local_process_zero": true,
|
8 |
+
"is_world_process_zero": true,
|
9 |
+
"log_history": [
|
10 |
+
{
|
11 |
+
"epoch": 0.18,
|
12 |
+
"learning_rate": 1.912033779028853e-05,
|
13 |
+
"loss": 0.3671,
|
14 |
+
"step": 250
|
15 |
+
},
|
16 |
+
{
|
17 |
+
"epoch": 0.35,
|
18 |
+
"learning_rate": 1.824067558057706e-05,
|
19 |
+
"loss": 0.2758,
|
20 |
+
"step": 500
|
21 |
+
},
|
22 |
+
{
|
23 |
+
"epoch": 0.53,
|
24 |
+
"learning_rate": 1.736101337086559e-05,
|
25 |
+
"loss": 0.2616,
|
26 |
+
"step": 750
|
27 |
+
},
|
28 |
+
{
|
29 |
+
"epoch": 0.7,
|
30 |
+
"learning_rate": 1.6481351161154117e-05,
|
31 |
+
"loss": 0.2469,
|
32 |
+
"step": 1000
|
33 |
+
},
|
34 |
+
{
|
35 |
+
"epoch": 0.88,
|
36 |
+
"learning_rate": 1.5601688951442647e-05,
|
37 |
+
"loss": 0.235,
|
38 |
+
"step": 1250
|
39 |
+
},
|
40 |
+
{
|
41 |
+
"epoch": 1.0,
|
42 |
+
"eval_accuracy": 0.90576171875,
|
43 |
+
"eval_binary_crossentropy_loss": 9.43606185913086,
|
44 |
+
"eval_loss": 0.22764553129673004,
|
45 |
+
"eval_runtime": 148.7776,
|
46 |
+
"eval_samples_per_second": 271.748,
|
47 |
+
"eval_steps_per_second": 16.985,
|
48 |
+
"step": 1421
|
49 |
+
}
|
50 |
+
],
|
51 |
+
"max_steps": 5684,
|
52 |
+
"num_train_epochs": 4,
|
53 |
+
"total_flos": 2.393340547233485e+16,
|
54 |
+
"trial_name": null,
|
55 |
+
"trial_params": null
|
56 |
+
}
|
checkpoint-1421/training_args.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:824f49e09d1e24e24b56eb55553854b0550b51f9807cc4bcc80575e87eb1efe4
|
3 |
+
size 3439
|
checkpoint-2842/added_tokens.json
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"[MASK]": 128000
|
3 |
+
}
|
checkpoint-2842/config.json
ADDED
@@ -0,0 +1,35 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"_name_or_path": "microsoft/deberta-v3-base",
|
3 |
+
"architectures": [
|
4 |
+
"DebertaV2ForSequenceClassification"
|
5 |
+
],
|
6 |
+
"attention_probs_dropout_prob": 0.1,
|
7 |
+
"hidden_act": "gelu",
|
8 |
+
"hidden_dropout_prob": 0.1,
|
9 |
+
"hidden_size": 768,
|
10 |
+
"initializer_range": 0.02,
|
11 |
+
"intermediate_size": 3072,
|
12 |
+
"layer_norm_eps": 1e-07,
|
13 |
+
"max_position_embeddings": 512,
|
14 |
+
"max_relative_positions": -1,
|
15 |
+
"model_type": "deberta-v2",
|
16 |
+
"norm_rel_ebd": "layer_norm",
|
17 |
+
"num_attention_heads": 12,
|
18 |
+
"num_hidden_layers": 12,
|
19 |
+
"pad_token_id": 0,
|
20 |
+
"pooler_dropout": 0,
|
21 |
+
"pooler_hidden_act": "gelu",
|
22 |
+
"pooler_hidden_size": 768,
|
23 |
+
"pos_att_type": [
|
24 |
+
"p2c",
|
25 |
+
"c2p"
|
26 |
+
],
|
27 |
+
"position_biased_input": false,
|
28 |
+
"position_buckets": 256,
|
29 |
+
"relative_attention": true,
|
30 |
+
"share_att_key": true,
|
31 |
+
"torch_dtype": "float32",
|
32 |
+
"transformers_version": "4.24.0",
|
33 |
+
"type_vocab_size": 0,
|
34 |
+
"vocab_size": 128100
|
35 |
+
}
|
checkpoint-2842/optimizer.pt
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:f280be97c0952ef3e97d8901189e366db805a8c61dcc7d4305ef8fdb28f4a0b9
|
3 |
+
size 1475556165
|
checkpoint-2842/pytorch_model.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:71a2f3358a96959d1fd5bb62b1bf065d36f1246b511749c6e91ae8bb366462cc
|
3 |
+
size 737766955
|
checkpoint-2842/rng_state.pth
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:df17ed2df95769819f66e6ed49342c944e9cc81ada6efd65f6937d5b6b0dc237
|
3 |
+
size 14503
|
checkpoint-2842/scaler.pt
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:4122359f72635429fa624445f1661c497be1cc014af92fc0d1f749759c6b58b1
|
3 |
+
size 559
|
checkpoint-2842/scheduler.pt
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:ca91d8012253dd5f50cc48273ec54d2f325ad0c0852a076e1324fe9ebd684b3c
|
3 |
+
size 559
|
checkpoint-2842/special_tokens_map.json
ADDED
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"bos_token": "[CLS]",
|
3 |
+
"cls_token": "[CLS]",
|
4 |
+
"eos_token": "[SEP]",
|
5 |
+
"mask_token": "[MASK]",
|
6 |
+
"pad_token": "[PAD]",
|
7 |
+
"sep_token": "[SEP]",
|
8 |
+
"unk_token": "[UNK]"
|
9 |
+
}
|
checkpoint-2842/spm.model
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:c679fbf93643d19aab7ee10c0b99e460bdbc02fedf34b92b05af343b4af586fd
|
3 |
+
size 2464616
|
checkpoint-2842/tokenizer.json
ADDED
The diff for this file is too large to render.
See raw diff
|
|
checkpoint-2842/tokenizer_config.json
ADDED
@@ -0,0 +1,16 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"bos_token": "[CLS]",
|
3 |
+
"cls_token": "[CLS]",
|
4 |
+
"do_lower_case": false,
|
5 |
+
"eos_token": "[SEP]",
|
6 |
+
"mask_token": "[MASK]",
|
7 |
+
"name_or_path": "microsoft/deberta-v3-base",
|
8 |
+
"pad_token": "[PAD]",
|
9 |
+
"sep_token": "[SEP]",
|
10 |
+
"sp_model_kwargs": {},
|
11 |
+
"special_tokens_map_file": null,
|
12 |
+
"split_by_punct": false,
|
13 |
+
"tokenizer_class": "DebertaV2Tokenizer",
|
14 |
+
"unk_token": "[UNK]",
|
15 |
+
"vocab_type": "spm"
|
16 |
+
}
|
checkpoint-2842/trainer_state.json
ADDED
@@ -0,0 +1,102 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"best_metric": 0.20898549258708954,
|
3 |
+
"best_model_checkpoint": "deberta-v3-base-funetuned-qqa/checkpoint-2842",
|
4 |
+
"epoch": 1.9997361709612171,
|
5 |
+
"global_step": 2842,
|
6 |
+
"is_hyper_param_search": false,
|
7 |
+
"is_local_process_zero": true,
|
8 |
+
"is_world_process_zero": true,
|
9 |
+
"log_history": [
|
10 |
+
{
|
11 |
+
"epoch": 0.18,
|
12 |
+
"learning_rate": 1.912033779028853e-05,
|
13 |
+
"loss": 0.3671,
|
14 |
+
"step": 250
|
15 |
+
},
|
16 |
+
{
|
17 |
+
"epoch": 0.35,
|
18 |
+
"learning_rate": 1.824067558057706e-05,
|
19 |
+
"loss": 0.2758,
|
20 |
+
"step": 500
|
21 |
+
},
|
22 |
+
{
|
23 |
+
"epoch": 0.53,
|
24 |
+
"learning_rate": 1.736101337086559e-05,
|
25 |
+
"loss": 0.2616,
|
26 |
+
"step": 750
|
27 |
+
},
|
28 |
+
{
|
29 |
+
"epoch": 0.7,
|
30 |
+
"learning_rate": 1.6481351161154117e-05,
|
31 |
+
"loss": 0.2469,
|
32 |
+
"step": 1000
|
33 |
+
},
|
34 |
+
{
|
35 |
+
"epoch": 0.88,
|
36 |
+
"learning_rate": 1.5601688951442647e-05,
|
37 |
+
"loss": 0.235,
|
38 |
+
"step": 1250
|
39 |
+
},
|
40 |
+
{
|
41 |
+
"epoch": 1.0,
|
42 |
+
"eval_accuracy": 0.90576171875,
|
43 |
+
"eval_binary_crossentropy_loss": 9.43606185913086,
|
44 |
+
"eval_loss": 0.22764553129673004,
|
45 |
+
"eval_runtime": 148.7776,
|
46 |
+
"eval_samples_per_second": 271.748,
|
47 |
+
"eval_steps_per_second": 16.985,
|
48 |
+
"step": 1421
|
49 |
+
},
|
50 |
+
{
|
51 |
+
"epoch": 1.06,
|
52 |
+
"learning_rate": 1.4722026741731177e-05,
|
53 |
+
"loss": 0.2228,
|
54 |
+
"step": 1500
|
55 |
+
},
|
56 |
+
{
|
57 |
+
"epoch": 1.23,
|
58 |
+
"learning_rate": 1.3842364532019705e-05,
|
59 |
+
"loss": 0.1956,
|
60 |
+
"step": 1750
|
61 |
+
},
|
62 |
+
{
|
63 |
+
"epoch": 1.41,
|
64 |
+
"learning_rate": 1.2962702322308235e-05,
|
65 |
+
"loss": 0.1994,
|
66 |
+
"step": 2000
|
67 |
+
},
|
68 |
+
{
|
69 |
+
"epoch": 1.58,
|
70 |
+
"learning_rate": 1.2083040112596764e-05,
|
71 |
+
"loss": 0.1957,
|
72 |
+
"step": 2250
|
73 |
+
},
|
74 |
+
{
|
75 |
+
"epoch": 1.76,
|
76 |
+
"learning_rate": 1.1203377902885292e-05,
|
77 |
+
"loss": 0.1893,
|
78 |
+
"step": 2500
|
79 |
+
},
|
80 |
+
{
|
81 |
+
"epoch": 1.94,
|
82 |
+
"learning_rate": 1.0323715693173822e-05,
|
83 |
+
"loss": 0.1908,
|
84 |
+
"step": 2750
|
85 |
+
},
|
86 |
+
{
|
87 |
+
"epoch": 2.0,
|
88 |
+
"eval_accuracy": 0.91357421875,
|
89 |
+
"eval_binary_crossentropy_loss": 8.624783515930176,
|
90 |
+
"eval_loss": 0.20898549258708954,
|
91 |
+
"eval_runtime": 148.8373,
|
92 |
+
"eval_samples_per_second": 271.639,
|
93 |
+
"eval_steps_per_second": 16.978,
|
94 |
+
"step": 2842
|
95 |
+
}
|
96 |
+
],
|
97 |
+
"max_steps": 5684,
|
98 |
+
"num_train_epochs": 4,
|
99 |
+
"total_flos": 4.78668109446697e+16,
|
100 |
+
"trial_name": null,
|
101 |
+
"trial_params": null
|
102 |
+
}
|
checkpoint-2842/training_args.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:824f49e09d1e24e24b56eb55553854b0550b51f9807cc4bcc80575e87eb1efe4
|
3 |
+
size 3439
|
checkpoint-4263/added_tokens.json
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"[MASK]": 128000
|
3 |
+
}
|
checkpoint-4263/config.json
ADDED
@@ -0,0 +1,35 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"_name_or_path": "microsoft/deberta-v3-base",
|
3 |
+
"architectures": [
|
4 |
+
"DebertaV2ForSequenceClassification"
|
5 |
+
],
|
6 |
+
"attention_probs_dropout_prob": 0.1,
|
7 |
+
"hidden_act": "gelu",
|
8 |
+
"hidden_dropout_prob": 0.1,
|
9 |
+
"hidden_size": 768,
|
10 |
+
"initializer_range": 0.02,
|
11 |
+
"intermediate_size": 3072,
|
12 |
+
"layer_norm_eps": 1e-07,
|
13 |
+
"max_position_embeddings": 512,
|
14 |
+
"max_relative_positions": -1,
|
15 |
+
"model_type": "deberta-v2",
|
16 |
+
"norm_rel_ebd": "layer_norm",
|
17 |
+
"num_attention_heads": 12,
|
18 |
+
"num_hidden_layers": 12,
|
19 |
+
"pad_token_id": 0,
|
20 |
+
"pooler_dropout": 0,
|
21 |
+
"pooler_hidden_act": "gelu",
|
22 |
+
"pooler_hidden_size": 768,
|
23 |
+
"pos_att_type": [
|
24 |
+
"p2c",
|
25 |
+
"c2p"
|
26 |
+
],
|
27 |
+
"position_biased_input": false,
|
28 |
+
"position_buckets": 256,
|
29 |
+
"relative_attention": true,
|
30 |
+
"share_att_key": true,
|
31 |
+
"torch_dtype": "float32",
|
32 |
+
"transformers_version": "4.24.0",
|
33 |
+
"type_vocab_size": 0,
|
34 |
+
"vocab_size": 128100
|
35 |
+
}
|
checkpoint-4263/optimizer.pt
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:d6ffd88cf0814c3cff051615a0ca02b7de2d9ecddbbe8426289a4343fd344824
|
3 |
+
size 1475556165
|
checkpoint-4263/pytorch_model.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:7e2d4995cbb293d758e532f6f64f4f7a8b8d99262f71f11e4423ea2b0b871fce
|
3 |
+
size 737766955
|
checkpoint-4263/rng_state.pth
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:a4ec02662cb6e351dc958e648f9561454b4378240f801e2312867fca9b3e8b6c
|
3 |
+
size 14503
|
checkpoint-4263/scaler.pt
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:80e3586cc756d547702508d1cb28650a4f7a15c38a1f8d0f2c449fa8814c3dfb
|
3 |
+
size 559
|
checkpoint-4263/scheduler.pt
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:ed99f924de94e7d256684b4dba0d7474d73ac90b13d10de522a3d324e3ce7e2a
|
3 |
+
size 559
|
checkpoint-4263/special_tokens_map.json
ADDED
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"bos_token": "[CLS]",
|
3 |
+
"cls_token": "[CLS]",
|
4 |
+
"eos_token": "[SEP]",
|
5 |
+
"mask_token": "[MASK]",
|
6 |
+
"pad_token": "[PAD]",
|
7 |
+
"sep_token": "[SEP]",
|
8 |
+
"unk_token": "[UNK]"
|
9 |
+
}
|
checkpoint-4263/spm.model
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:c679fbf93643d19aab7ee10c0b99e460bdbc02fedf34b92b05af343b4af586fd
|
3 |
+
size 2464616
|
checkpoint-4263/tokenizer.json
ADDED
The diff for this file is too large to render.
See raw diff
|
|
checkpoint-4263/tokenizer_config.json
ADDED
@@ -0,0 +1,16 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"bos_token": "[CLS]",
|
3 |
+
"cls_token": "[CLS]",
|
4 |
+
"do_lower_case": false,
|
5 |
+
"eos_token": "[SEP]",
|
6 |
+
"mask_token": "[MASK]",
|
7 |
+
"name_or_path": "microsoft/deberta-v3-base",
|
8 |
+
"pad_token": "[PAD]",
|
9 |
+
"sep_token": "[SEP]",
|
10 |
+
"sp_model_kwargs": {},
|
11 |
+
"special_tokens_map_file": null,
|
12 |
+
"split_by_punct": false,
|
13 |
+
"tokenizer_class": "DebertaV2Tokenizer",
|
14 |
+
"unk_token": "[UNK]",
|
15 |
+
"vocab_type": "spm"
|
16 |
+
}
|
checkpoint-4263/trainer_state.json
ADDED
@@ -0,0 +1,148 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"best_metric": 0.20898549258708954,
|
3 |
+
"best_model_checkpoint": "deberta-v3-base-funetuned-qqa/checkpoint-2842",
|
4 |
+
"epoch": 2.999736170961217,
|
5 |
+
"global_step": 4263,
|
6 |
+
"is_hyper_param_search": false,
|
7 |
+
"is_local_process_zero": true,
|
8 |
+
"is_world_process_zero": true,
|
9 |
+
"log_history": [
|
10 |
+
{
|
11 |
+
"epoch": 0.18,
|
12 |
+
"learning_rate": 1.912033779028853e-05,
|
13 |
+
"loss": 0.3671,
|
14 |
+
"step": 250
|
15 |
+
},
|
16 |
+
{
|
17 |
+
"epoch": 0.35,
|
18 |
+
"learning_rate": 1.824067558057706e-05,
|
19 |
+
"loss": 0.2758,
|
20 |
+
"step": 500
|
21 |
+
},
|
22 |
+
{
|
23 |
+
"epoch": 0.53,
|
24 |
+
"learning_rate": 1.736101337086559e-05,
|
25 |
+
"loss": 0.2616,
|
26 |
+
"step": 750
|
27 |
+
},
|
28 |
+
{
|
29 |
+
"epoch": 0.7,
|
30 |
+
"learning_rate": 1.6481351161154117e-05,
|
31 |
+
"loss": 0.2469,
|
32 |
+
"step": 1000
|
33 |
+
},
|
34 |
+
{
|
35 |
+
"epoch": 0.88,
|
36 |
+
"learning_rate": 1.5601688951442647e-05,
|
37 |
+
"loss": 0.235,
|
38 |
+
"step": 1250
|
39 |
+
},
|
40 |
+
{
|
41 |
+
"epoch": 1.0,
|
42 |
+
"eval_accuracy": 0.90576171875,
|
43 |
+
"eval_binary_crossentropy_loss": 9.43606185913086,
|
44 |
+
"eval_loss": 0.22764553129673004,
|
45 |
+
"eval_runtime": 148.7776,
|
46 |
+
"eval_samples_per_second": 271.748,
|
47 |
+
"eval_steps_per_second": 16.985,
|
48 |
+
"step": 1421
|
49 |
+
},
|
50 |
+
{
|
51 |
+
"epoch": 1.06,
|
52 |
+
"learning_rate": 1.4722026741731177e-05,
|
53 |
+
"loss": 0.2228,
|
54 |
+
"step": 1500
|
55 |
+
},
|
56 |
+
{
|
57 |
+
"epoch": 1.23,
|
58 |
+
"learning_rate": 1.3842364532019705e-05,
|
59 |
+
"loss": 0.1956,
|
60 |
+
"step": 1750
|
61 |
+
},
|
62 |
+
{
|
63 |
+
"epoch": 1.41,
|
64 |
+
"learning_rate": 1.2962702322308235e-05,
|
65 |
+
"loss": 0.1994,
|
66 |
+
"step": 2000
|
67 |
+
},
|
68 |
+
{
|
69 |
+
"epoch": 1.58,
|
70 |
+
"learning_rate": 1.2083040112596764e-05,
|
71 |
+
"loss": 0.1957,
|
72 |
+
"step": 2250
|
73 |
+
},
|
74 |
+
{
|
75 |
+
"epoch": 1.76,
|
76 |
+
"learning_rate": 1.1203377902885292e-05,
|
77 |
+
"loss": 0.1893,
|
78 |
+
"step": 2500
|
79 |
+
},
|
80 |
+
{
|
81 |
+
"epoch": 1.94,
|
82 |
+
"learning_rate": 1.0323715693173822e-05,
|
83 |
+
"loss": 0.1908,
|
84 |
+
"step": 2750
|
85 |
+
},
|
86 |
+
{
|
87 |
+
"epoch": 2.0,
|
88 |
+
"eval_accuracy": 0.91357421875,
|
89 |
+
"eval_binary_crossentropy_loss": 8.624783515930176,
|
90 |
+
"eval_loss": 0.20898549258708954,
|
91 |
+
"eval_runtime": 148.8373,
|
92 |
+
"eval_samples_per_second": 271.639,
|
93 |
+
"eval_steps_per_second": 16.978,
|
94 |
+
"step": 2842
|
95 |
+
},
|
96 |
+
{
|
97 |
+
"epoch": 2.11,
|
98 |
+
"learning_rate": 9.44405348346235e-06,
|
99 |
+
"loss": 0.1706,
|
100 |
+
"step": 3000
|
101 |
+
},
|
102 |
+
{
|
103 |
+
"epoch": 2.29,
|
104 |
+
"learning_rate": 8.56439127375088e-06,
|
105 |
+
"loss": 0.1574,
|
106 |
+
"step": 3250
|
107 |
+
},
|
108 |
+
{
|
109 |
+
"epoch": 2.46,
|
110 |
+
"learning_rate": 7.68472906403941e-06,
|
111 |
+
"loss": 0.1573,
|
112 |
+
"step": 3500
|
113 |
+
},
|
114 |
+
{
|
115 |
+
"epoch": 2.64,
|
116 |
+
"learning_rate": 6.808585503166785e-06,
|
117 |
+
"loss": 0.1619,
|
118 |
+
"step": 3750
|
119 |
+
},
|
120 |
+
{
|
121 |
+
"epoch": 2.81,
|
122 |
+
"learning_rate": 5.928923293455313e-06,
|
123 |
+
"loss": 0.1574,
|
124 |
+
"step": 4000
|
125 |
+
},
|
126 |
+
{
|
127 |
+
"epoch": 2.99,
|
128 |
+
"learning_rate": 5.049261083743843e-06,
|
129 |
+
"loss": 0.1575,
|
130 |
+
"step": 4250
|
131 |
+
},
|
132 |
+
{
|
133 |
+
"epoch": 3.0,
|
134 |
+
"eval_accuracy": 0.9169921875,
|
135 |
+
"eval_binary_crossentropy_loss": 8.323027610778809,
|
136 |
+
"eval_loss": 0.20993904769420624,
|
137 |
+
"eval_runtime": 148.6942,
|
138 |
+
"eval_samples_per_second": 271.9,
|
139 |
+
"eval_steps_per_second": 16.995,
|
140 |
+
"step": 4263
|
141 |
+
}
|
142 |
+
],
|
143 |
+
"max_steps": 5684,
|
144 |
+
"num_train_epochs": 4,
|
145 |
+
"total_flos": 7.180021641700454e+16,
|
146 |
+
"trial_name": null,
|
147 |
+
"trial_params": null
|
148 |
+
}
|
checkpoint-4263/training_args.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:824f49e09d1e24e24b56eb55553854b0550b51f9807cc4bcc80575e87eb1efe4
|
3 |
+
size 3439
|
checkpoint-5684/added_tokens.json
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"[MASK]": 128000
|
3 |
+
}
|
checkpoint-5684/config.json
ADDED
@@ -0,0 +1,35 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"_name_or_path": "microsoft/deberta-v3-base",
|
3 |
+
"architectures": [
|
4 |
+
"DebertaV2ForSequenceClassification"
|
5 |
+
],
|
6 |
+
"attention_probs_dropout_prob": 0.1,
|
7 |
+
"hidden_act": "gelu",
|
8 |
+
"hidden_dropout_prob": 0.1,
|
9 |
+
"hidden_size": 768,
|
10 |
+
"initializer_range": 0.02,
|
11 |
+
"intermediate_size": 3072,
|
12 |
+
"layer_norm_eps": 1e-07,
|
13 |
+
"max_position_embeddings": 512,
|
14 |
+
"max_relative_positions": -1,
|
15 |
+
"model_type": "deberta-v2",
|
16 |
+
"norm_rel_ebd": "layer_norm",
|
17 |
+
"num_attention_heads": 12,
|
18 |
+
"num_hidden_layers": 12,
|
19 |
+
"pad_token_id": 0,
|
20 |
+
"pooler_dropout": 0,
|
21 |
+
"pooler_hidden_act": "gelu",
|
22 |
+
"pooler_hidden_size": 768,
|
23 |
+
"pos_att_type": [
|
24 |
+
"p2c",
|
25 |
+
"c2p"
|
26 |
+
],
|
27 |
+
"position_biased_input": false,
|
28 |
+
"position_buckets": 256,
|
29 |
+
"relative_attention": true,
|
30 |
+
"share_att_key": true,
|
31 |
+
"torch_dtype": "float32",
|
32 |
+
"transformers_version": "4.24.0",
|
33 |
+
"type_vocab_size": 0,
|
34 |
+
"vocab_size": 128100
|
35 |
+
}
|
checkpoint-5684/optimizer.pt
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:89b4701c310864ec95f47dd060ab3172cec793bc50583c324ab20d76876eda8b
|
3 |
+
size 1475556165
|
checkpoint-5684/pytorch_model.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:e9844060415fe086b8fc9d401b82589fce254389c80d41ab68a74543f5f3cc16
|
3 |
+
size 737766955
|
checkpoint-5684/rng_state.pth
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:51dd1148f1869e988e323551f902630f4c6058ace117c23d947be1febda42ee8
|
3 |
+
size 14503
|
checkpoint-5684/scaler.pt
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:1c3bf7d39d50349c13985b5d79e8e43f189781b9a084733bba5f6296b8118011
|
3 |
+
size 559
|
checkpoint-5684/scheduler.pt
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:6d7fa27de976f860090af1dc2e9c3ce9327d48a9d735296824b1ac8c15ae2810
|
3 |
+
size 559
|
checkpoint-5684/special_tokens_map.json
ADDED
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"bos_token": "[CLS]",
|
3 |
+
"cls_token": "[CLS]",
|
4 |
+
"eos_token": "[SEP]",
|
5 |
+
"mask_token": "[MASK]",
|
6 |
+
"pad_token": "[PAD]",
|
7 |
+
"sep_token": "[SEP]",
|
8 |
+
"unk_token": "[UNK]"
|
9 |
+
}
|
checkpoint-5684/spm.model
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:c679fbf93643d19aab7ee10c0b99e460bdbc02fedf34b92b05af343b4af586fd
|
3 |
+
size 2464616
|