julien-c commited on
Commit
291f0fa
1 Parent(s): 349f52e

Migrate model card from transformers-repo

Browse files

Read announcement at https://discuss.huggingface.co/t/announcement-all-model-cards-will-be-migrated-to-hf-co-model-repos/2755
Original file history: https://github.com/huggingface/transformers/commits/master/model_cards/ahotrod/albert_xxlargev1_squad2_512/README.md

Files changed (1) hide show
  1. README.md +60 -0
README.md ADDED
@@ -0,0 +1,60 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ## Albert xxlarge version 1 language model fine-tuned on SQuAD2.0
2
+
3
+ ### (updated 30Sept2020) with the following results:
4
+
5
+ ```
6
+ exact: 86.11134506864315
7
+ f1: 89.35371214945009
8
+ total': 11873
9
+ HasAns_exact': 83.56950067476383
10
+ HasAns_f1': 90.06353312254078
11
+ HasAns_total': 5928
12
+ NoAns_exact': 88.64592094196804
13
+ NoAns_f1': 88.64592094196804
14
+ NoAns_total': 5945
15
+ best_exact': 86.11134506864315
16
+ best_exact_thresh': 0.0
17
+ best_f1': 89.35371214944985
18
+ best_f1_thresh': 0.0
19
+ ```
20
+
21
+ ### from script:
22
+
23
+ ```
24
+ python ${EXAMPLES}/run_squad.py \
25
+ --model_type albert \
26
+ --model_name_or_path albert-xxlarge-v1 \
27
+ --do_train \
28
+ --do_eval \
29
+ --train_file ${SQUAD}/train-v2.0.json \
30
+ --predict_file ${SQUAD}/dev-v2.0.json \
31
+ --version_2_with_negative \
32
+ --do_lower_case \
33
+ --num_train_epochs 3 \
34
+ --max_steps 8144 \
35
+ --warmup_steps 814 \
36
+ --learning_rate 3e-5 \
37
+ --max_seq_length 512 \
38
+ --doc_stride 128 \
39
+ --per_gpu_train_batch_size 6 \
40
+ --gradient_accumulation_steps 8 \
41
+ --per_gpu_eval_batch_size 48 \
42
+ --fp16 \
43
+ --fp16_opt_level O1 \
44
+ --threads 12 \
45
+ --logging_steps 50 \
46
+ --save_steps 3000 \
47
+ --overwrite_output_dir \
48
+ --output_dir ${MODEL_PATH}
49
+ ```
50
+
51
+ ### using the following software & system:
52
+
53
+ ```
54
+ Transformers: 3.1.0
55
+ PyTorch: 1.6.0
56
+ TensorFlow: 2.3.1
57
+ Python: 3.8.1
58
+ OS: Linux-5.4.0-48-generic-x86_64-with-glibc2.10
59
+ CPU/GPU: Intel i9-9900K / NVIDIA Titan RTX 24GB
60
+ ```