Error while serving model as per documentation,

[jchauhan]

Describe the bug Error while serving a toy model using EasyDel with the following exception.

To Reproduce

Context

• Using Google TPU VM
• Follow the instructions as suggested here https://github.com/erfanzar/EasyDeL/blob/36209097d2a3a92fd5174d9e3ae55b1785c24500/README.md?plain=1#L79
• Use TinyLLM model TinyLlama/TinyLlama-1.1B-Chat-v1.0

python sev-tiny.py
Traceback (most recent call last):
  File "/home/***/research/EasyDeL/sev-tiny.py", line 23, in <module>
    params=model.params,
  File "/home/***/research/EasyDeL/.venv/lib/python3.10/site-packages/transformers/modeling_flax_utils.py", line 271, in params
    raise ValueError(
ValueError: `params` cannot be accessed from model when the model is created with `_do_init=False`. You must call `init_weights` manually and store the params outside of the model and pass it explicitly where needed.

[erfanzar]

You are trying to access the frozen model and that's empty, ill create an example for you in next commit.

[erfanzar]

i have tested all of the examples and all of them are working fine use this as a mindset example

from EasyDel import (
    JAXServer,
    JAXServerConfig
)

server_config = JAXServerConfig(
    top_k=40,
    top_p=0.9,
    temperature=0.8,
    max_length=2048,
    max_new_tokens=4096,
    max_compile_tokens=128
)

model_id = "TinyLlama/TinyLlama-1.1B-intermediate-step-1431k-3T"

server = JAXServer.from_torch_pretrained(
    server_config,
    model_id
)

tkns = 0
for response, tokens_generated in server.process(
    "```python\ndef make_pytorch_linear("
):
    print(response[tkns:], end="")
    tkns += len(response)