From f745f61c9971e7dd91b2f123f0eb5e75752881bf Mon Sep 17 00:00:00 2001
From: ahotrod <44321615+ahotrod@users.noreply.github.com>
Date: Thu, 1 Oct 2020 05:50:07 -0700
Subject: [PATCH] Update README.md (#7491)

Model now fine-tuned on Transformers 3.1.0, previous out-of-date model was fine-tuned on Transformers 2.3.0.
---
 .../albert_xxlargev1_squad2_512/README.md     | 93 ++++++++-----------
 1 file changed, 41 insertions(+), 52 deletions(-)

diff --git a/model_cards/ahotrod/albert_xxlargev1_squad2_512/README.md b/model_cards/ahotrod/albert_xxlargev1_squad2_512/README.md
index 61e0c291a5..2f7cf73689 100644
--- a/model_cards/ahotrod/albert_xxlargev1_squad2_512/README.md
+++ b/model_cards/ahotrod/albert_xxlargev1_squad2_512/README.md
@@ -1,71 +1,60 @@
 ## Albert xxlarge version 1 language model fine-tuned on SQuAD2.0
 
-### with the following results:
+###  (updated 30Sept2020) with the following results:
 
 ```
-exact: 85.65653162637918
-f1: 89.260458954177
+exact: 86.11134506864315
+f1: 89.35371214945009
 total': 11873
-HasAns_exact': 82.6417004048583
-HasAns_f1': 89.8598902096736
+HasAns_exact': 83.56950067476383
+HasAns_f1': 90.06353312254078
 HasAns_total': 5928
-NoAns_exact': 88.66274179983179
-NoAns_f1': 88.66274179983179
+NoAns_exact': 88.64592094196804
+NoAns_f1': 88.64592094196804
 NoAns_total': 5945
-best_exact': 85.65653162637918
+best_exact': 86.11134506864315
 best_exact_thresh': 0.0
-best_f1': 89.2604589541768
+best_f1': 89.35371214944985
 best_f1_thresh': 0.0
 ```
 
 ### from script:
 
 ```
-python -m torch.distributed.launch --nproc_per_node=2 ${RUN_SQUAD_DIR}/run_squad.py \
---model_type albert \
---model_name_or_path albert-xxlarge-v1 \
---do_train \
---train_file ${SQUAD_DIR}/train-v2.0.json \
---predict_file ${SQUAD_DIR}/dev-v2.0.json \
---version_2_with_negative \
---num_train_epochs 3 \
---max_steps 8144 \
---warmup_steps 814 \
---do_lower_case \
---learning_rate 3e-5 \
---max_seq_length 512 \
---doc_stride 128 \
---save_steps 2000 \
---per_gpu_train_batch_size 1 \
---gradient_accumulation_steps 24 \
---output_dir ${MODEL_PATH}
-
-CUDA_VISIBLE_DEVICES=0 python ${RUN_SQUAD_DIR}/run_squad.py \
---model_type albert \
---model_name_or_path ${MODEL_PATH} \
---do_eval \
---train_file ${SQUAD_DIR}/train-v2.0.json \
---predict_file ${SQUAD_DIR}/dev-v2.0.json \
---version_2_with_negative \
---do_lower_case \
---max_seq_length 512 \
---per_gpu_eval_batch_size 48 \
---output_dir ${MODEL_PATH}
+python ${EXAMPLES}/run_squad.py \
+  --model_type albert \
+  --model_name_or_path albert-xxlarge-v1 \
+  --do_train \
+  --do_eval \
+  --train_file ${SQUAD}/train-v2.0.json \
+  --predict_file ${SQUAD}/dev-v2.0.json \
+  --version_2_with_negative \
+  --do_lower_case \
+  --num_train_epochs 3 \
+  --max_steps 8144 \
+  --warmup_steps 814 \
+  --learning_rate 3e-5 \
+  --max_seq_length 512 \
+  --doc_stride 128 \
+  --per_gpu_train_batch_size 6 \
+  --gradient_accumulation_steps 8 \
+  --per_gpu_eval_batch_size 48 \
+  --fp16 \
+  --fp16_opt_level O1 \
+  --threads 12 \
+  --logging_steps 50 \
+  --save_steps 3000 \
+  --overwrite_output_dir \
+  --output_dir ${MODEL_PATH}
 ```
 
-### using the following system & software:
+### using the following software & system:
 
 ```
-OS/Platform: Linux-4.15.0-76-generic-x86_64-with-debian-buster-sid
-GPU/CPU: 2 x NVIDIA 1080Ti / Intel i7-8700
-Transformers: 2.3.0
-PyTorch: 1.4.0
-TensorFlow: 2.1.0
-Python: 3.7.6
+Transformers: 3.1.0
+PyTorch: 1.6.0
+TensorFlow: 2.3.1
+Python: 3.8.1
+OS: Linux-5.4.0-48-generic-x86_64-with-glibc2.10
+CPU/GPU: Intel i9-9900K / NVIDIA Titan RTX 24GB
 ```
-
-### Access this albert_xxlargev1_sqd2_512 fine-tuned model with:
-
-```python
-tokenizer = AutoTokenizer.from_pretrained("ahotrod/albert_xxlargev1_squad2_512")
-model = AutoModelForQuestionAnswering.from_pretrained("ahotrod/albert_xxlargev1_squad2_512")