freeze tokenizer version for py3.8

[causten]

The python package https://pypi.org/project/tokenizers dropped prebuilt binaries for python3.8 in the latest version. Freezing our Bert example's requirements file to a known working version

[migraphx-bot]

Test	Batch	Rate new fac5cd	Rate old 810b3f	Diff	Compare
torchvision-resnet50	64	3,236.93	3,227.48	0.29%	✅
torchvision-resnet50_fp16	64	6,877.99	6,935.80	-0.83%	✅
torchvision-densenet121	32	2,437.22	2,443.57	-0.26%	✅
torchvision-densenet121_fp16	32	4,191.35	4,190.37	0.02%	✅
torchvision-inceptionv3	32	1,614.12	1,618.09	-0.25%	✅
torchvision-inceptionv3_fp16	32	2,687.40	2,703.46	-0.59%	✅
cadene-inceptionv4	16	750.80	754.54	-0.50%	✅
cadene-resnext64x4	16	809.45	799.94	1.19%	✅
slim-mobilenet	64	6,663.67	7,038.75	-5.33%	🔴
slim-nasnetalarge	64	199.01	203.74	-2.32%	✅
slim-resnet50v2	64	3,428.62	3,418.15	0.31%	✅
bert-mrpc-onnx	8	1,144.47	1,130.84	1.20%	✅
bert-mrpc-tf	1	482.98	483.04	-0.01%	✅
pytorch-examples-wlang-gru	1	514.35	471.20	9.16%	🔆
pytorch-examples-wlang-lstm	1	441.89	446.88	-1.12%	✅
torchvision-resnet50_1	1	794.21	802.10	-0.98%	✅
cadene-dpn92_1	1	430.19	426.83	0.79%	✅
cadene-resnext101_1	1	390.46	383.89	1.71%	✅
onnx-taau-downsample	1	373.53	372.62	0.25%	✅
dlrm-criteoterabyte	1	31.81	32.54	-2.23%	✅
dlrm-criteoterabyte_fp16	1	51.08	51.07	0.01%	✅
agentmodel	1	8,603.11	8,676.04	-0.84%	✅
unet_fp16	2	57.97	57.90	0.13%	✅
resnet50v1_fp16	1	1,026.15	1,041.98	-1.52%	✅
resnet50v1_int8	1	785.10	779.61	0.70%	✅
bert_base_cased_fp16	64	1,172.26	1,172.53	-0.02%	✅
bert_large_uncased_fp16	32	362.59	362.45	0.04%	✅
bert_large_fp16	1	200.38	200.21	0.09%	✅
distilgpt2_fp16	16	2,216.47	2,216.14	0.01%	✅
yolov5s	1	524.12	518.19	1.14%	✅
tinyllama	1	43.61	43.59	0.04%	✅
vicuna-fastchat	1	174.34	174.63	-0.17%	✅
whisper-tiny-encoder	1	411.92	406.05	1.45%	✅
whisper-tiny-decoder	1	409.27	398.24	2.77%	✅

This build is not recommended to merge 🔴

[migraphx-bot]

     ✅ bert-mrpc-onnx: PASSED: MIGraphX meets tolerance

     ✅ bert-mrpc-tf: PASSED: MIGraphX meets tolerance

     ✅ pytorch-examples-wlang-gru: PASSED: MIGraphX meets tolerance

     ✅ pytorch-examples-wlang-lstm: PASSED: MIGraphX meets tolerance

     ✅ torchvision-resnet50_1: PASSED: MIGraphX meets tolerance

     ✅ cadene-dpn92_1: PASSED: MIGraphX meets tolerance

     ✅ cadene-resnext101_1: PASSED: MIGraphX meets tolerance

     ✅ dlrm-criteoterabyte: PASSED: MIGraphX meets tolerance

     ✅ agentmodel: PASSED: MIGraphX meets tolerance

     ✅ unet: PASSED: MIGraphX meets tolerance

     ✅ resnet50v1: PASSED: MIGraphX meets tolerance

     ✅ bert_base_cased_fp16: PASSED: MIGraphX meets tolerance

🔴bert_large_uncased_fp16: FAILED: MIGraphX is not within tolerance - check verbose output

     ✅ bert_large: PASSED: MIGraphX meets tolerance

     ✅ yolov5s: PASSED: MIGraphX meets tolerance

     ✅ tinyllama: PASSED: MIGraphX meets tolerance

     ✅ vicuna-fastchat: PASSED: MIGraphX meets tolerance

     ✅ whisper-tiny-encoder: PASSED: MIGraphX meets tolerance

     ✅ whisper-tiny-decoder: PASSED: MIGraphX meets tolerance

     ✅ distilgpt2_fp16: PASSED: MIGraphX meets tolerance

[codecov]

Codecov Report

All modified and coverable lines are covered by tests ✅

Project coverage is 92.29%. Comparing base (810b3f2) to head (fac5cd5).
Report is 5 commits behind head on develop.

Additional details and impacted files

@@           Coverage Diff            @@
##           develop    #3779   +/-   ##
========================================
  Coverage    92.29%   92.29%           
========================================
  Files          519      519           
  Lines        22217    22217           
========================================
  Hits         20505    20505           
  Misses        1712     1712           

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.