Hugging Face's logo

docketanalyzer
/
label-complaint

Text Classification
Transformers
TensorBoard
Safetensors
deberta-v2
Generated from Trainer
Inference Endpoints
Model card Files Metrics Community
nadahlberg commited on
Commit
51a2059
•
1 Parent(s): 408c422

Model save

Browse files
Files changed (8) hide show
  1. README.md +19 -29
  2. config.json +1 -1
  3. eval_results.json +6 -6
  4. logs/events.out.tfevents.1722374450.nathan.153742.0 +3 -0
  5. logs/events.out.tfevents.1722374714.nathan.153742.1 +3 -0
  6. model.safetensors +1 -1
  7. run_config.json +4 -4
  8. training_args.bin +2 -2
README.md CHANGED
@@ -17,8 +17,8 @@ should probably proofread and complete it, then remove this comment. -->
17
 
18
  This model is a fine-tuned version of [docketanalyzer/docket-lm-xs](https://huggingface.co/docketanalyzer/docket-lm-xs) on the None dataset.
19
  It achieves the following results on the evaluation set:
20
- - Loss: 0.0048
21
- - F1: 0.9995
22
 
23
  ## Model description
24
 
@@ -45,40 +45,30 @@ The following hyperparameters were used during training:
45
  - total_train_batch_size: 16
46
  - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
47
  - lr_scheduler_type: linear
48
- - lr_scheduler_warmup_steps: 100
49
  - num_epochs: 1
50
 
51
  ### Training results
52
 
53
- | Training Loss | Epoch | Step | Validation Loss | F1 |
54
- |:-------------:|:------:|:----:|:---------------:|:------:|
55
- | 0.0017 | 0.0435 | 300 | 0.0115 | 0.9976 |
56
- | 0.0011 | 0.0870 | 600 | 0.0133 | 0.9978 |
57
- | 0.0022 | 0.1305 | 900 | 0.0116 | 0.9972 |
58
- | 0.0002 | 0.1740 | 1200 | 0.0127 | 0.9981 |
59
- | 0.0005 | 0.2175 | 1500 | 0.0092 | 0.9981 |
60
- | 0.0002 | 0.2611 | 1800 | 0.0120 | 0.9979 |
61
- | 0.0002 | 0.3046 | 2100 | 0.0079 | 0.9985 |
62
- | 0.0001 | 0.3481 | 2400 | 0.0146 | 0.9961 |
63
- | 0.0003 | 0.3916 | 2700 | 0.0075 | 0.9988 |
64
- | 0.0003 | 0.4351 | 3000 | 0.0088 | 0.9986 |
65
- | 0.0001 | 0.4786 | 3300 | 0.0095 | 0.9988 |
66
- | 0.0001 | 0.5221 | 3600 | 0.0067 | 0.9991 |
67
- | 0.0001 | 0.5656 | 3900 | 0.0077 | 0.9990 |
68
- | 0.0001 | 0.6091 | 4200 | 0.0084 | 0.9985 |
69
- | 0.0001 | 0.6526 | 4500 | 0.0049 | 0.9995 |
70
- | 0.0347 | 0.6962 | 4800 | 0.0056 | 0.9993 |
71
- | 0.0001 | 0.7397 | 5100 | 0.0067 | 0.9988 |
72
- | 0.0016 | 0.7832 | 5400 | 0.0068 | 0.9986 |
73
- | 0.0001 | 0.8267 | 5700 | 0.0054 | 0.9990 |
74
- | 0.0001 | 0.8702 | 6000 | 0.0053 | 0.9991 |
75
- | 0.0872 | 0.9137 | 6300 | 0.0048 | 0.9995 |
76
- | 0.0001 | 0.9572 | 6600 | 0.0048 | 0.9995 |
77
 
78
 
79
  ### Framework versions
80
 
81
- - Transformers 4.41.1
82
- - Pytorch 2.3.0+cu121
83
  - Datasets 2.14.4
84
  - Tokenizers 0.19.1
 
17
 
18
  This model is a fine-tuned version of [docketanalyzer/docket-lm-xs](https://huggingface.co/docketanalyzer/docket-lm-xs) on the None dataset.
19
  It achieves the following results on the evaluation set:
20
+ - Loss: 0.0121
21
+ - F1: 0.9981
22
 
23
  ## Model description
24
 
 
45
  - total_train_batch_size: 16
46
  - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
47
  - lr_scheduler_type: linear
48
+ - lr_scheduler_warmup_ratio: 0.02
49
  - num_epochs: 1
50
 
51
  ### Training results
52
 
53
+ | Training Loss | Epoch | Step | Validation Loss | F1 |
54
+ |:-------------:|:-----:|:----:|:---------------:|:------:|
55
+ | 0.0312 | 0.08 | 95 | 0.0374 | 0.9923 |
56
+ | 0.0035 | 0.16 | 190 | 0.0096 | 0.9981 |
57
+ | 0.0022 | 0.24 | 285 | 0.0068 | 0.9990 |
58
+ | 0.0016 | 0.32 | 380 | 0.0143 | 0.9971 |
59
+ | 0.0012 | 0.4 | 475 | 0.0124 | 0.9981 |
60
+ | 0.0013 | 0.48 | 570 | 0.0127 | 0.9981 |
61
+ | 0.0012 | 0.56 | 665 | 0.0114 | 0.9981 |
62
+ | 0.0011 | 0.64 | 760 | 0.0130 | 0.9981 |
63
+ | 0.001 | 0.72 | 855 | 0.0117 | 0.9981 |
64
+ | 0.0005 | 0.8 | 950 | 0.0130 | 0.9981 |
65
+ | 0.0005 | 0.88 | 1045 | 0.0118 | 0.9981 |
66
+ | 0.0005 | 0.96 | 1140 | 0.0126 | 0.9981 |
 
 
 
 
 
 
 
 
 
 
67
 
68
 
69
  ### Framework versions
70
 
71
+ - Transformers 4.43.3
72
+ - Pytorch 2.3.1+cu121
73
  - Datasets 2.14.4
74
  - Tokenizers 0.19.1
config.json CHANGED
@@ -37,7 +37,7 @@
37
  "relative_attention": true,
38
  "share_att_key": true,
39
  "torch_dtype": "float32",
40
- "transformers_version": "4.41.1",
41
  "type_vocab_size": 0,
42
  "vocab_size": 128100
43
  }
 
37
  "relative_attention": true,
38
  "share_att_key": true,
39
  "torch_dtype": "float32",
40
+ "transformers_version": "4.43.3",
41
  "type_vocab_size": 0,
42
  "vocab_size": 128100
43
  }
eval_results.json CHANGED
@@ -1,8 +1,8 @@
1
  {
2
- "eval_loss": 0.004779691807925701,
3
- "eval_f1": 0.999485155311481,
4
- "eval_runtime": 19.4506,
5
- "eval_samples_per_second": 298.5,
6
- "eval_steps_per_second": 18.663,
7
- "epoch": 1.0
8
  }
 
1
  {
2
+ "eval_loss": 0.012090153060853481,
3
+ "eval_f1": 0.9980582524271845,
4
+ "eval_runtime": 3.2065,
5
+ "eval_samples_per_second": 311.867,
6
+ "eval_steps_per_second": 19.648,
7
+ "epoch": 0.999578947368421
8
  }
logs/events.out.tfevents.1722374450.nathan.153742.0 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:557aed47c2037a606ad7d9eeb6eb643c62e2e9e706445991bee346af3d72ec89
3
+ size 134323
logs/events.out.tfevents.1722374714.nathan.153742.1 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:0d88cbe0a24e9a047ad614b89b1ea913289224d182b56e20e2361783d9b5529d
3
+ size 405
model.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:e1266d7a450140330afab29beb86e9f6fa0e5bb5218d67a0c223ab027937ac8b
3
  size 283347432
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e16ee7f54b2c1219c0e6a0bfc4330695714db5ebef98be08a0e2742c481eb8fe
3
  size 283347432
run_config.json CHANGED
@@ -10,12 +10,12 @@
10
  "gradient_accumulation_steps": 2,
11
  "learning_rate": 5e-05,
12
  "weight_decay": 0.1,
13
- "warmup_steps": 100,
14
  "evaluation_strategy": "steps",
15
- "eval_steps": 300,
16
- "save_steps": 300,
17
  "save_total_limit": 2
18
  },
19
- "run_name": "initial_model",
20
  "run_type": "ClassificationRoutine"
21
  }
 
10
  "gradient_accumulation_steps": 2,
11
  "learning_rate": 5e-05,
12
  "weight_decay": 0.1,
13
+ "warmup_ratio": 0.02,
14
  "evaluation_strategy": "steps",
15
+ "eval_steps": 0.08,
16
+ "save_steps": 0.08,
17
  "save_total_limit": 2
18
  },
19
+ "run_name": "complaint",
20
  "run_type": "ClassificationRoutine"
21
  }
training_args.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:3656fcbc1832dd9c0867ba00d4106f7f2d6ac5ab8dfe75730fee3152c51822ce
3
- size 5112
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:66ef2207833c0afa403c72448d2b06f189c7278d0be4661827287b01ffc17100
3
+ size 5240