Hugging Face's logo

muhtasham
/
web-scale-ner

Token Classification
Transformers
Safetensors
English
longformer
Generated from Trainer
Inference Endpoints
Model card Files Community
muhtasham commited on
Commit
a1264d6
•
1 Parent(s): 2c92033

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +24 -24
README.md CHANGED
@@ -16,27 +16,27 @@ language:
16
  - en
17
  ---
18
 
19
- {
20
- "eval/recall": 0.8099026827985271,
21
- "train_runtime": 3761.6609,
22
- "_timestamp": 1723859671.093963,
23
- "eval/steps_per_second": 0.304,
24
- "total_flos": 55842753586544640,
25
- "train/loss": 0.3471,
26
- "train_loss": 0.7508978596440068,
27
- "train/epoch": 5,
28
- "eval/accuracy": 0.8099026827985271,
29
- "train/grad_norm": 5.0118021965026855,
30
- "train/learning_rate": 1.8801187394248965e-8,
31
- "eval/samples_per_second": 23.205,
32
- "eval/f1": 0.8131537247003896,
33
- "_wandb.runtime": 3822,
34
- "_runtime": 3821.1701335906982,
35
- "eval/loss": 0.5318660736083984,
36
- "eval/runtime": 46.067,
37
- "eval/precision": 0.820569791570792,
38
- "train/global_step": 270,
39
- "train_steps_per_second": 0.072,
40
- "_step": 30,
41
- "train_samples_per_second": 5.681
42
- }
 
16
  - en
17
  ---
18
 
19
+ **Training and Evaluation Metrics**
20
+
21
+ | Metric | Value |
22
+ |---------------------------|--------------------------------|
23
+ | **Evaluation Recall** | 0.8099 |
24
+ | **Training Runtime** | 3761.66 seconds |
25
+ | **Evaluation Steps/Sec** | 0.304 |
26
+ | **Total FLOPs** | 55,842,753,586,544,640 |
27
+ | **Training Loss** | 0.3471 |
28
+ | **Training Loss (Epoch)** | 0.7509 |
29
+ | **Training Epochs** | 5 |
30
+ | **Evaluation Accuracy** | 0.8099 |
31
+ | **Gradient Norm** | 5.0118 |
32
+ | **Learning Rate** | 1.88e-8 |
33
+ | **Evaluation Samples/Sec** | 23.205 |
34
+ | **Evaluation F1 Score** | 0.8132 |
35
+ | **Wandb Runtime** | 3822 seconds |
36
+ | **Runtime** | 3821.17 seconds |
37
+ | **Evaluation Loss** | 0.5319 |
38
+ | **Evaluation Runtime** | 46.067 seconds |
39
+ | **Evaluation Precision** | 0.8206 |
40
+ | **Global Training Steps** | 270 |
41
+ | **Training Steps/Sec** | 0.072 |
42
+ | **Training Samples/Sec** | 5.681 |