Add support for Phi-1 and Phi 1.5

[arnavgarg1]

In Transformers 4.36, the transformers library added support for Phi-1 and Phi-1.5 models from Microsoft. However, there are two caveats with using this model:

1. The original models, microsofot/phi-1 and micosoft/phi-1.5 don't work out of the box since they require remote_code to be trusted because the tensor operations are implemented through einops instead of PyTorch. Instead, in the original PR that adds support for Phi based models, they add two models that are supported (https://github.com/huggingface/transformers/pull/26170/files#diff-88cb36bfb13c1dc5f52bb952b74697a1c79e286a1a57e4ed3f20ecd5e9f8749bR25):

• susnato/phi-1_dev
• susnato/phi-1_5_dev

This is the recommendation from official phi model docs on huggingface as well: https://huggingface.co/docs/transformers/main/model_doc/phi

from transformers import PhiForCausalLM, AutoTokenizer

# define the model and tokenizer.
model = PhiForCausalLM.from_pretrained("susnato/phi-1_5_dev")
tokenizer = AutoTokenizer.from_pretrained("susnato/phi-1_5_dev")

My understanding is that someone from the huggingface team has converted the official weights into huggingface compatible weights under the two new mappings. I've filed an issue here to understand what the expected behavior is supposed to be: huggingface/transformers#28049

2. Both susnato/phi-1_dev and susnato/phi-1_5_dev don't support to device_map auto model load kwarg that we set when we load models in quantized state, say when initializing the model using 4 bit quantization. However, it seems that the model weights get correctly loaded onto the right device depending on the quantization kwargs anyway, so we just skip using this load kwarg for phi based models.

Closes #3630

[github-actions]

Unit Test Results

  6 files  ±0    6 suites  ±0   14m 15s ⏱️ -13s
12 tests ±0    9 ✔️ ±0    3 💤 ±0  0 ❌ ±0 
60 runs  ±0  42 ✔️ ±0  18 💤 ±0  0 ❌ ±0 

Results for commit 83514f6. ± Comparison against base commit bccfb4e.