Nano GPT model

[MarcelWilnicki]

the command:
python run.py --model_name gpt2 -f pytorch --lambada_path /ampere/aml/natural_language_processing/text_generation/nanogpt/lambada_test_plain_text.txt

results in following error:

`Traceback (most recent call last):
File "/ampere/mzd/ampere_model_library/utils/pytorch.py", line 65, in init
self._frozen_script = torch.jit.freeze(torch.jit.script(self._model), preserved_attrs=[func])
File "/usr/local/lib/python3.10/dist-packages/torch/jit/_script.py", line 1284, in script
return torch.jit._recursive.create_script_module(
File "/usr/local/lib/python3.10/dist-packages/torch/jit/_recursive.py", line 480, in create_script_module
return create_script_module_impl(nn_module, concrete_type, stubs_fn)
File "/usr/local/lib/python3.10/dist-packages/torch/jit/_recursive.py", line 542, in create_script_module_impl
script_module = torch.jit.RecursiveScriptModule._construct(cpp_module, init_fn)
File "/usr/local/lib/python3.10/dist-packages/torch/jit/_script.py", line 614, in _construct
init_fn(script_module)
File "/usr/local/lib/python3.10/dist-packages/torch/jit/_recursive.py", line 520, in init_fn
scripted = create_script_module_impl(orig_value, sub_concrete_type, stubs_fn)
File "/usr/local/lib/python3.10/dist-packages/torch/jit/_recursive.py", line 542, in create_script_module_impl
script_module = torch.jit.RecursiveScriptModule._construct(cpp_module, init_fn)
File "/usr/local/lib/python3.10/dist-packages/torch/jit/_script.py", line 614, in _construct
init_fn(script_module)
File "/usr/local/lib/python3.10/dist-packages/torch/jit/_recursive.py", line 520, in init_fn
scripted = create_script_module_impl(orig_value, sub_concrete_type, stubs_fn)
File "/usr/local/lib/python3.10/dist-packages/torch/jit/_recursive.py", line 542, in create_script_module_impl
script_module = torch.jit.RecursiveScriptModule._construct(cpp_module, init_fn)
File "/usr/local/lib/python3.10/dist-packages/torch/jit/_script.py", line 614, in _construct
init_fn(script_module)
File "/usr/local/lib/python3.10/dist-packages/torch/jit/_recursive.py", line 520, in init_fn
scripted = create_script_module_impl(orig_value, sub_concrete_type, stubs_fn)
File "/usr/local/lib/python3.10/dist-packages/torch/jit/_recursive.py", line 542, in create_script_module_impl
script_module = torch.jit.RecursiveScriptModule._construct(cpp_module, init_fn)
File "/usr/local/lib/python3.10/dist-packages/torch/jit/_script.py", line 614, in _construct
init_fn(script_module)
File "/usr/local/lib/python3.10/dist-packages/torch/jit/_recursive.py", line 520, in init_fn
scripted = create_script_module_impl(orig_value, sub_concrete_type, stubs_fn)
File "/usr/local/lib/python3.10/dist-packages/torch/jit/_recursive.py", line 546, in create_script_module_impl
create_methods_and_properties_from_stubs(concrete_type, method_stubs, property_stubs)
File "/usr/local/lib/python3.10/dist-packages/torch/jit/_recursive.py", line 397, in create_methods_and_properties_from_stubs
concrete_type._create_methods_and_properties(property_defs, property_rcbs, method_defs, method_rcbs, method_defaults)
RuntimeError:
Module 'CausalSelfAttention' has no attribute 'bias' :
File "/ampere/aml/natural_language_processing/text_generation/nanogpt/nanoGPT/model.py", line 78
# manual implementation of attention
att = (q @ k.transpose(-2, -1)) * (1.0 / math.sqrt(k.size(-1)))
att = att.masked_fill(self.bias[:,:,:T,:T] == 0, float('-inf'))
~~~~~~~~~ <--- HERE
att = F.softmax(att, dim=-1)
att = self.attn_dropout(att)

During handling of the above exception, another exception occurred:

Traceback (most recent call last):
File "/ampere/aml/natural_language_processing/text_generation/nanogpt/run.py", line 78, in
main()
File "/ampere/aml/natural_language_processing/text_generation/nanogpt/run.py", line 71, in main
run_pytorch_fp32(**vars(args))
File "/ampere/aml/natural_language_processing/text_generation/nanogpt/run.py", line 63, in run_pytorch_fp32
run_pytorch(model_name, batch_size, num_runs, timeout, lambada_path, disable_jit_freeze, **kwargs)
File "/ampere/aml/natural_language_processing/text_generation/nanogpt/run.py", line 57, in run_pytorch
runner = PyTorchRunner(model, disable_jit_freeze=disable_jit_freeze, func="generate")
File "/ampere/mzd/ampere_model_library/utils/pytorch.py", line 69, in init
self._frozen_script = torch.jit.freeze(torch.jit.trace(self._model, example_inputs))
File "/usr/local/lib/python3.10/dist-packages/torch/jit/_trace.py", line 793, in trace
raise RuntimeError("example_kwarg_inputs should be a dict")
RuntimeError: example_kwarg_inputs should be a dict`