import ecco ERRO

[nikipi]

Hey,

I tried pip install ecco on google colab then import ecco

but I have the error below:

RuntimeError Traceback (most recent call last) init.pxd in numpy.import_array()

RuntimeError: module compiled against API version 0x10 but this version of numpy is 0xf . Check the section C-API incompatibility at the Troubleshooting ImportError section at https://numpy.org/devdocs/user/troubleshooting-importerror.html#c-api-incompatibility for indications on how to solve this problem .

During handling of the above exception, another exception occurred:

ImportError Traceback (most recent call last) in <cell line: 1>() ----> 1 import ecco 2 lm = ecco.from_pretrained('distilgpt2')

5 frames /usr/local/lib/python3.10/dist-packages/ecco/init.py in 14 15 version = '0.1.2' ---> 16 from ecco.lm import LM 17 from transformers import AutoTokenizer, AutoModelForCausalLM, AutoModel, AutoModelForSeq2SeqLM 18 from typing import Any, Dict, Optional, List

/usr/local/lib/python3.10/dist-packages/ecco/lm.py in 13 from torch.nn import functional as F 14 from ecco.attribution import compute_primary_attributions_scores ---> 15 from ecco.output import OutputSeq 16 from typing import Optional, Any, List, Tuple, Dict, Union 17 from operator import attrgetter

/usr/local/lib/python3.10/dist-packages/ecco/output.py in 9 import torch 10 from torch.nn import functional as F ---> 11 from sklearn import decomposition 12 from typing import Dict, Optional, List, Tuple, Union 13 from ecco.util import strip_tokenizer_prefix, is_partial_token

/usr/local/lib/python3.10/dist-packages/sklearn/init.py in 80 from . import _distributor_init # noqa: F401 81 from . import __check_build # noqa: F401 ---> 82 from .base import clone 83 from .utils._show_versions import show_versions 84

/usr/local/lib/python3.10/dist-packages/sklearn/base.py in 15 from . import version 16 from ._config import get_config ---> 17 from .utils import _IS_32BIT 18 from .utils._tags import ( 19 _DEFAULT_TAGS,

/usr/local/lib/python3.10/dist-packages/sklearn/utils/init.py in 20 from scipy.sparse import issparse 21 ---> 22 from .murmurhash import murmurhash3_32 23 from .class_weight import compute_class_weight, compute_sample_weight 24 from . import _joblib

sklearn/utils/murmurhash.pyx in init sklearn.utils.murmurhash()

init.pxd in numpy.import_array()

ImportError: numpy.core.multiarray failed to import

---

NOTE: If your import is failing due to a missing package, you can manually install dependencies using either !pip or !apt.

To view examples of installing some common dependencies, click the "Open Examples" button below.

[SumitDasTR]

I updated numpy using the command pip install -U numpy

Then tried to run the notebook cell

import ecco

lm = ecco.from_pretrained('t5-small') review="""Denis Villeneuve's Dune looks and sounds amazing -- and once the (admittedly slow-building) story gets you hooked, you'll be on the edge of your seat for the sequel."""

output = lm.generate(f"sst2 sentence: {review}", generate=1, do_sample=False)

But its giving the following error

TypeError Traceback (most recent call last) in <cell line: 8>() 6 of your seat for the sequel.""" 7 ----> 8 output = lm.generate(f"sst2 sentence: {review}", generate=1, do_sample=False)

1 frames /usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py in _prepare_decoder_input_ids_for_generation(self, batch_size, model_input_name, model_kwargs, decoder_start_token_id, bos_token_id, device) 654 if model_kwargs is not None and "decoder_input_ids" in model_kwargs: 655 decoder_input_ids = model_kwargs.pop("decoder_input_ids") --> 656 elif "input_ids" in model_kwargs and model_input_name != "input_ids": 657 decoder_input_ids = model_kwargs.pop("input_ids") 658 else:

TypeError: argument of type 'NoneType' is not iterable

[neelgandhi108]

Fixed !pip uninstall numpy -y !pip install numpy==1.21.5

[LAKan233]

I solved this problem by

pip install -U numpy

and the , the most important thing, you must restart your notebook kernel

[SumitDasTR]

tried the above options both in colab and local jupyter notebook. But giving the same error:

TypeError: argument of type 'NoneType' is not iterable

`` ! pip install ecco

! pip uninstall numpy -y ! pip install numpy==1.21.5

import ecco

lm = ecco.from_pretrained('t5-small') review="""Denis Villeneuve's Dune looks and sounds amazing -- and once the (admittedly slow-building) story gets you hooked, you'll be on the edge of your seat for the sequel."""

output = lm.generate(f"sst2 sentence: {review}", generate=1, do_sample=False) ``

TypeError Traceback (most recent call last) /tmp/ipykernel_912/1601604474.py in 6 of your seat for the sequel.""" 7 ----> 8 output = lm.generate(f"sst2 sentence: {review}", generate=1, do_sample=False)

~/miniconda3/envs/py3.8/lib/python3.8/site-packages/ecco/lm.py in generate(self, input_str, max_length, temperature, top_k, top_p, do_sample, attribution, generate, beam_size, **generate_kwargs) 202 assert len(input_ids.size()) == 2 # will break otherwise 203 if version.parse(transformers.version) >= version.parse('4.13'): --> 204 decoder_input_ids = self.model._prepare_decoder_input_ids_for_generation(input_ids.shape[0], None, None) 205 else: 206 decoder_input_ids = self.model._prepare_decoder_input_ids_for_generation(input_ids, None, None)

~/miniconda3/envs/py3.8/lib/python3.8/site-packages/transformers/generation/utils.py in _prepare_decoder_input_ids_for_generation(self, batch_size, model_input_name, model_kwargs, decoder_start_token_id, bos_token_id, device) 654 if model_kwargs is not None and "decoder_input_ids" in model_kwargs: 655 decoder_input_ids = model_kwargs.pop("decoder_input_ids") --> 656 elif "input_ids" in model_kwargs and model_input_name != "input_ids": 657 decoder_input_ids = model_kwargs.pop("input_ids") 658 else:

TypeError: argument of type 'NoneType' is not iterable

[Peter-Zhoutuanjie]

SumitDasTR

Same

[BiEchi]

Yeah the pip install numpy==1.21.5 solution does not work for me either. Has anyone solved this problem?

lm = ecco.from_pretrained('valhalla/t5-small-qa-qg-hl')#, gpu=False)
output = lm.generate(text, generate=20, do_sample=True, attribution=['ig', 'grad_x_input'])

gives

---> 14 output = lm.generate(text, generate=20, do_sample=True, attribution=['ig', 'grad_x_input'])
     16 output.primary_attributions(attr_method='ig')

File [/opt/conda/envs/ecco/lib/python3.9/site-packages/ecco/lm.py:204](https://vscode-remote+ssh-002dremote-002bec2-002d34-002d204-002d44-002d163-002ecompute-002d1-002eamazonaws-002ecom.vscode-resource.vscode-cdn.net/opt/conda/envs/ecco/lib/python3.9/site-packages/ecco/lm.py:204), in LM.generate(self, input_str, max_length, temperature, top_k, top_p, do_sample, attribution, generate, beam_size, **generate_kwargs)
    202 assert len(input_ids.size()) == 2 # will break otherwise
    203 if version.parse(transformers.__version__) >= version.parse('4.13'):
--> 204     decoder_input_ids = self.model._prepare_decoder_input_ids_for_generation(input_ids.shape[0], None, None)
    205 else:
    206     decoder_input_ids = self.model._prepare_decoder_input_ids_for_generation(input_ids, None, None)

File [/opt/conda/envs/ecco/lib/python3.9/site-packages/transformers/generation/utils.py:662](https://vscode-remote+ssh-002dremote-002bec2-002d34-002d204-002d44-002d163-002ecompute-002d1-002eamazonaws-002ecom.vscode-resource.vscode-cdn.net/opt/conda/envs/ecco/lib/python3.9/site-packages/transformers/generation/utils.py:662), in GenerationMixin._prepare_decoder_input_ids_for_generation(self, batch_size, model_input_name, model_kwargs, decoder_start_token_id, bos_token_id, device)
    660 if model_kwargs is not None and "decoder_input_ids" in model_kwargs:
    661     decoder_input_ids = model_kwargs.pop("decoder_input_ids")
--> 662 elif "input_ids" in model_kwargs and model_input_name != "input_ids":
    663     decoder_input_ids = model_kwargs.pop("input_ids")
    664 else:

TypeError: argument of type 'NoneType' is not iterable

[BiEchi]

Just fixed the issue. Downgrading your transformers to 4.13.0 solves the problem. This is due to an API update from transformers update.

@SumitDasTR @Peter-Zhoutuanjie cc @jalammar

[BiEchi]

Note that this solution is a quick fix. If you want to use models like OPT, LLAMA, etc., you still need to fix the code from source to make it work.

[martin12333]

Maybe another solution is: Weeks ago, I made the ecco *.ipynb mostly work in Python 3.10, with transformers 4.29.2

Locally. In a conda env with many other programs. I dimly remember having problems with version numbers of dependencies. I had to edit the req file(s) of ecco: as a false-beginner, I edited both setup.py, and requirements.txt

My notes are at https://github.com/martin12333/marti-onedrive/blob/main2/AI/pip----31pip310ecco.e.f8.sh

My files setup.py, and requirements.txt are at https://github.com/martin12333/marti-onedrive/tree/main2/AI/Jalammar

---

EDIT: additional info is at https://github.com/jalammar/ecco/issues/102#issuecomment-1675574099