Fix a bug in DeepSpeedMLP

[sakogan]

The initialization of data_type in DeepSpeedMLP appears to be incorrect when int8 inference is requested, leading to seemingly wrong text generation.

For instance, running the following command (using inference-test.py from the DeepSpeedExamples repo)

deepspeed --num_gpus 1 inference/huggingface/text-generation/inference-test.py --model bigscience/bloom-7b1 --batch_size 1 --use_kernel --use_meta_tensor --dtype int8

produces the following output:

in=DeepSpeed is a machine learning framework
out=DeepSpeed is a machine learning framework 6 10 tràm kú Tá aixướm de de截útil恩opció 15 Beraz Tá P L Tá 7 80pata terbakar aixjai尼 Tá l'Oficina de al 3'hi aix-Els L Tá T Haut-Commissariat Tá 80juntament de de a 20 10 kú

The proposed fix modifies the change to the data_type initialization introduced in #3425

[loadams]

@sakogan - when you tested this with the DeepSpeedExamples change, did it produce quality output?

[sakogan]

@loadams Not sure what DeepSpeedExamples change you are referring. I tested it with the latest version of the DeepSpeedExamples master branch. With the proposed fix, the output is valid (as well as when using --dtype float16 in that command, with or without the fix)

[loadams]

@loadams Not sure what DeepSpeedExamples change you are referring. I tested it with the latest version of the DeepSpeedExamples master branch. With the proposed fix, the output is valid (as well as when using --dtype float16 in that command, with or without the fix)

@sakogan - apologies, I meant if you tested this change with the error repro you had in DeepSpeedExamples. But sounds like you did, thanks!