End of training
Browse files- README.md +8 -17
- config.json +72 -3
- generation_config.json +1 -2
- model.safetensors +1 -1
- runs/Jun04_10-32-09_70c6696ca1f2/events.out.tfevents.1717497134.70c6696ca1f2.34.0 +3 -0
- runs/Jun04_10-34-18_70c6696ca1f2/events.out.tfevents.1717497259.70c6696ca1f2.34.1 +3 -0
- runs/Jun04_10-34-18_70c6696ca1f2/events.out.tfevents.1717497732.70c6696ca1f2.34.2 +3 -0
- tokenizer.json +0 -0
- training_args.bin +1 -1
README.md
CHANGED
@@ -1,5 +1,5 @@
|
|
1 |
---
|
2 |
-
base_model: google/pegasus-
|
3 |
tags:
|
4 |
- generated_from_trainer
|
5 |
metrics:
|
@@ -14,14 +14,14 @@ should probably proofread and complete it, then remove this comment. -->
|
|
14 |
|
15 |
# pegasus-samsum
|
16 |
|
17 |
-
This model is a fine-tuned version of [google/pegasus-
|
18 |
It achieves the following results on the evaluation set:
|
19 |
-
- Loss:
|
20 |
-
- Rouge1:
|
21 |
-
- Rouge2:
|
22 |
-
- Rougel:
|
23 |
-
- Rougelsum:
|
24 |
-
- Gen Len:
|
25 |
|
26 |
## Model description
|
27 |
|
@@ -53,15 +53,6 @@ The following hyperparameters were used during training:
|
|
53 |
|
54 |
### Training results
|
55 |
|
56 |
-
| Training Loss | Epoch | Step | Validation Loss | Rouge1 | Rouge2 | Rougel | Rougelsum | Gen Len |
|
57 |
-
|:-------------:|:------:|:----:|:---------------:|:-------:|:-------:|:-------:|:---------:|:--------:|
|
58 |
-
| 6.0132 | 0.1314 | 100 | 5.7494 | 39.0961 | 12.9642 | 29.8168 | 35.3077 | 127.2582 |
|
59 |
-
| 5.8274 | 0.2628 | 200 | 5.5325 | 41.7845 | 14.2872 | 31.299 | 37.5468 | 127.2582 |
|
60 |
-
| 5.8027 | 0.3942 | 300 | 5.4423 | 42.4827 | 14.6782 | 31.6679 | 38.2157 | 127.2582 |
|
61 |
-
| 5.5682 | 0.5256 | 400 | 5.3866 | 43.3037 | 15.0392 | 32.0844 | 38.7795 | 127.2582 |
|
62 |
-
| 5.396 | 0.6570 | 500 | 5.3446 | 43.5705 | 15.2971 | 32.2981 | 38.998 | 127.2582 |
|
63 |
-
| 5.5759 | 0.7884 | 600 | 5.3106 | 43.8978 | 15.4979 | 32.6086 | 39.4542 | 127.2582 |
|
64 |
-
| 5.4803 | 0.9198 | 700 | 5.3030 | 43.8476 | 15.4851 | 32.5764 | 39.3579 | 127.2582 |
|
65 |
|
66 |
|
67 |
### Framework versions
|
|
|
1 |
---
|
2 |
+
base_model: google/pegasus-large
|
3 |
tags:
|
4 |
- generated_from_trainer
|
5 |
metrics:
|
|
|
14 |
|
15 |
# pegasus-samsum
|
16 |
|
17 |
+
This model is a fine-tuned version of [google/pegasus-large](https://huggingface.co/google/pegasus-large) on an unknown dataset.
|
18 |
It achieves the following results on the evaluation set:
|
19 |
+
- Loss: 6.2452
|
20 |
+
- Rouge1: 34.6462
|
21 |
+
- Rouge2: 8.2009
|
22 |
+
- Rougel: 24.8919
|
23 |
+
- Rougelsum: 31.6451
|
24 |
+
- Gen Len: 125.97
|
25 |
|
26 |
## Model description
|
27 |
|
|
|
53 |
|
54 |
### Training results
|
55 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
56 |
|
57 |
|
58 |
### Framework versions
|
config.json
CHANGED
@@ -1,5 +1,5 @@
|
|
1 |
{
|
2 |
-
"_name_or_path": "google/pegasus-
|
3 |
"activation_dropout": 0.1,
|
4 |
"activation_function": "relu",
|
5 |
"add_bias_logits": false,
|
@@ -10,6 +10,7 @@
|
|
10 |
"attention_dropout": 0.1,
|
11 |
"bos_token_id": 0,
|
12 |
"classif_dropout": 0.0,
|
|
|
13 |
"d_model": 1024,
|
14 |
"decoder_attention_heads": 16,
|
15 |
"decoder_ffn_dim": 4096,
|
@@ -23,7 +24,9 @@
|
|
23 |
"encoder_layers": 16,
|
24 |
"eos_token_id": 1,
|
25 |
"extra_pos_embeddings": 1,
|
|
|
26 |
"forced_eos_token_id": 1,
|
|
|
27 |
"id2label": {
|
28 |
"0": "LABEL_0",
|
29 |
"1": "LABEL_1",
|
@@ -37,9 +40,8 @@
|
|
37 |
"LABEL_2": 2
|
38 |
},
|
39 |
"length_penalty": 0.8,
|
40 |
-
"max_length":
|
41 |
"max_position_embeddings": 1024,
|
42 |
-
"min_length": 32,
|
43 |
"model_type": "pegasus",
|
44 |
"normalize_before": true,
|
45 |
"normalize_embedding": false,
|
@@ -48,6 +50,73 @@
|
|
48 |
"pad_token_id": 0,
|
49 |
"scale_embedding": true,
|
50 |
"static_position_embeddings": true,
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
51 |
"torch_dtype": "float32",
|
52 |
"transformers_version": "4.41.1",
|
53 |
"use_cache": true,
|
|
|
1 |
{
|
2 |
+
"_name_or_path": "google/pegasus-large",
|
3 |
"activation_dropout": 0.1,
|
4 |
"activation_function": "relu",
|
5 |
"add_bias_logits": false,
|
|
|
10 |
"attention_dropout": 0.1,
|
11 |
"bos_token_id": 0,
|
12 |
"classif_dropout": 0.0,
|
13 |
+
"classifier_dropout": 0.0,
|
14 |
"d_model": 1024,
|
15 |
"decoder_attention_heads": 16,
|
16 |
"decoder_ffn_dim": 4096,
|
|
|
24 |
"encoder_layers": 16,
|
25 |
"eos_token_id": 1,
|
26 |
"extra_pos_embeddings": 1,
|
27 |
+
"force_bos_token_to_be_generated": false,
|
28 |
"forced_eos_token_id": 1,
|
29 |
+
"gradient_checkpointing": false,
|
30 |
"id2label": {
|
31 |
"0": "LABEL_0",
|
32 |
"1": "LABEL_1",
|
|
|
40 |
"LABEL_2": 2
|
41 |
},
|
42 |
"length_penalty": 0.8,
|
43 |
+
"max_length": 256,
|
44 |
"max_position_embeddings": 1024,
|
|
|
45 |
"model_type": "pegasus",
|
46 |
"normalize_before": true,
|
47 |
"normalize_embedding": false,
|
|
|
50 |
"pad_token_id": 0,
|
51 |
"scale_embedding": true,
|
52 |
"static_position_embeddings": true,
|
53 |
+
"task_specific_params": {
|
54 |
+
"summarization_aeslc": {
|
55 |
+
"length_penalty": 0.6,
|
56 |
+
"max_length": 32,
|
57 |
+
"max_position_embeddings": 512
|
58 |
+
},
|
59 |
+
"summarization_arxiv": {
|
60 |
+
"length_penalty": 0.8,
|
61 |
+
"max_length": 256,
|
62 |
+
"max_position_embeddings": 1024
|
63 |
+
},
|
64 |
+
"summarization_big_patent": {
|
65 |
+
"length_penalty": 0.7,
|
66 |
+
"max_length": 256,
|
67 |
+
"max_position_embeddings": 1024
|
68 |
+
},
|
69 |
+
"summarization_billsum": {
|
70 |
+
"length_penalty": 0.6,
|
71 |
+
"max_length": 256,
|
72 |
+
"max_position_embeddings": 1024
|
73 |
+
},
|
74 |
+
"summarization_cnn_dailymail": {
|
75 |
+
"length_penalty": 0.8,
|
76 |
+
"max_length": 128,
|
77 |
+
"max_position_embeddings": 1024
|
78 |
+
},
|
79 |
+
"summarization_gigaword": {
|
80 |
+
"length_penalty": 0.6,
|
81 |
+
"max_length": 32,
|
82 |
+
"max_position_embeddings": 128
|
83 |
+
},
|
84 |
+
"summarization_large": {
|
85 |
+
"length_penalty": 0.8,
|
86 |
+
"max_length": 256,
|
87 |
+
"max_position_embeddings": 1024
|
88 |
+
},
|
89 |
+
"summarization_multi_news": {
|
90 |
+
"length_penalty": 0.8,
|
91 |
+
"max_length": 256,
|
92 |
+
"max_position_embeddings": 1024
|
93 |
+
},
|
94 |
+
"summarization_newsroom": {
|
95 |
+
"length_penalty": 0.8,
|
96 |
+
"max_length": 128,
|
97 |
+
"max_position_embeddings": 512
|
98 |
+
},
|
99 |
+
"summarization_pubmed": {
|
100 |
+
"length_penalty": 0.8,
|
101 |
+
"max_length": 256,
|
102 |
+
"max_position_embeddings": 1024
|
103 |
+
},
|
104 |
+
"summarization_reddit_tifu": {
|
105 |
+
"length_penalty": 0.6,
|
106 |
+
"max_length": 128,
|
107 |
+
"max_position_embeddings": 512
|
108 |
+
},
|
109 |
+
"summarization_wikihow": {
|
110 |
+
"length_penalty": 0.6,
|
111 |
+
"max_length": 256,
|
112 |
+
"max_position_embeddings": 512
|
113 |
+
},
|
114 |
+
"summarization_xsum": {
|
115 |
+
"length_penalty": 0.8,
|
116 |
+
"max_length": 64,
|
117 |
+
"max_position_embeddings": 512
|
118 |
+
}
|
119 |
+
},
|
120 |
"torch_dtype": "float32",
|
121 |
"transformers_version": "4.41.1",
|
122 |
"use_cache": true,
|
generation_config.json
CHANGED
@@ -5,8 +5,7 @@
|
|
5 |
"eos_token_id": 1,
|
6 |
"forced_eos_token_id": 1,
|
7 |
"length_penalty": 0.8,
|
8 |
-
"max_length":
|
9 |
-
"min_length": 32,
|
10 |
"num_beams": 8,
|
11 |
"pad_token_id": 0,
|
12 |
"transformers_version": "4.41.1"
|
|
|
5 |
"eos_token_id": 1,
|
6 |
"forced_eos_token_id": 1,
|
7 |
"length_penalty": 0.8,
|
8 |
+
"max_length": 256,
|
|
|
9 |
"num_beams": 8,
|
10 |
"pad_token_id": 0,
|
11 |
"transformers_version": "4.41.1"
|
model.safetensors
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
size 2283652852
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:e2e4ee73058cdfaa92d5f197117bda37fdea440f5330078c6a3b109ffd35a7ca
|
3 |
size 2283652852
|
runs/Jun04_10-32-09_70c6696ca1f2/events.out.tfevents.1717497134.70c6696ca1f2.34.0
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:7f0b6b49f3700339f47afd0f62cfd831a851cbeae427d42c7cbacce7ad5e1642
|
3 |
+
size 7262
|
runs/Jun04_10-34-18_70c6696ca1f2/events.out.tfevents.1717497259.70c6696ca1f2.34.1
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:5b95ac13ed74dc4866fd965b542ab8cdb26e99a6bcf553a6014078aa0d4fe64c
|
3 |
+
size 8852
|
runs/Jun04_10-34-18_70c6696ca1f2/events.out.tfevents.1717497732.70c6696ca1f2.34.2
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:e56362f168b75d6f09d2503162279f92f3aa4abef3e750fae8f305a175dc20ea
|
3 |
+
size 603
|
tokenizer.json
CHANGED
The diff for this file is too large to render.
See raw diff
|
|
training_args.bin
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
size 5112
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:f8187be43213b7271b7d487c3ee4f5bc5e9f7a9c46387aecb21ef954e6ea5fe1
|
3 |
size 5112
|