Update fastai with conda doesn't update to the latest version

I think I had the same issue a few days ago.

I remember trying conda install -c fastai fastai=1.0.38

If that doesn’t work, try reinstalling: conda remove fastai && conda install -c fastai fastai

2 Likes

I have tried and got the error:

PackagesNotFoundError: The following packages are not available from current channels:

  - fastai=1.0.38
  - pytorch[version='>=1.0']

Current channels:

  - https://conda.anaconda.org/fastai/linux-64
  - https://conda.anaconda.org/fastai/noarch
  - https://repo.anaconda.com/pkgs/main/linux-64
  - https://repo.anaconda.com/pkgs/main/noarch
  - https://repo.anaconda.com/pkgs/free/linux-64
  - https://repo.anaconda.com/pkgs/free/noarch
  - https://repo.anaconda.com/pkgs/r/linux-64
  - https://repo.anaconda.com/pkgs/r/noarch
  - https://repo.anaconda.com/pkgs/pro/linux-64
  - https://repo.anaconda.com/pkgs/pro/noarch

To search for alternate channels that may provide the conda package you're
looking for, navigate to

    https://anaconda.org

and use the search bar at the top of the page.
1 Like

Have you looked at this thread?

The thing is simple :slight_smile:

1-The lastest version released in pip is v1.0.39

2-The lastest version released in conda is v1.0.39

https://anaconda.org/fastai/fastai/files

Ideally you can just compile with pip and install it.

inside the main git repository (in case of linux):

git pull && \
rm -rf ./dist && \
python3 setup.py bdist_wheel && \
pip install -U ./dist/fastai*.whl

Hi,

I have exactly the same problem as steffenix. On GCP - when I use:
sudo /opt/anaconda3/bin/conda install -c fastai fastai
I get an install of fastai 1.0.34.

I’ve tried rubyrhod’s suggestion to remove the fastai package and re-install and get the same packages not found error reported:

PackagesNotFoundError: The following packages are not available from current channels:

  • fastai=1.0.38
  • pytorch[version=’>=1.0’] . (etc)

I’ve gone back through the forum GCP setup instructions and created a clean instance and then updated, following the instructions - same problem: fastai v 1.0.34.

I don’t want to use PIP as a solution as suggested by willismar - I’d much rather use the conda package manager if there’s a way as that is what I’ve generated familiarity with at this point.

Any help would be very welcome - thanks! Right now, if anyone follows the instructions to set up fastai as in the v3 course GCP server setup they’ll run in to the same problem I think.

1 Like

Ok I figured out for you guys…

Fastai now depends of a package outside its “own” repository and outside the “default” repository too, so if not passed will not be found either:

conda install fastai  -c fastai -c pytorch
13 Likes

Thanks very much willismar.

That works. It installed most recent fastai as you said. On GCP exactly this:

sudo /opt/anaconda3/bin/conda install fastai -c fastai -c pytorch

Many thanks for taking the time to figure this out - really appreciate it.

Joe

2 Likes

I was having the same issue as well but used the following to update to the latest version of fastai on GCP:

sudo /opt/anaconda3/bin/conda update fastai -c fastai -c pytorch

1 Like

Hello,
(my previous post on this subject but in another thread)
Neither conda install fastai -c fastai -c pytorch, not conda update fastai -c fastai -c pytorch allows me to update fastai from 1.0.38 (current version in my environment, but not the most recent one). I’m using Anaconda Prompt and Windows 10. Any suggestion?

Hi @pierreguillou,

I really don’t know exact what the problem on your computer… I just tried to simulate in my linux. The thing is conda is equal in Windows or Linux.
The only other thing I can think of, is about you having a python version where pytorch aren’t compiled yet on the repository. Here I am using Python 3.6 on Linux.

Fastai is version free:
https://anaconda.org/fastai/fastai/files?version=1.0.41

PyTorch 1.0 is dependent in cuda version and python version
https://anaconda.org/pytorch/pytorch/files?version=1.0.0

Those are the windows version available on pytorch channel for anaconda/conda

580.8 MB	 | win-64/pytorch-1.0.0-py3.6_cuda80_cudnn7_1.tar.bz2
654.6 MB	 | win-64/pytorch-1.0.0-py3.5_cuda100_cudnn7_1.tar.bz2
654.5 MB	 | win-64/pytorch-1.0.0-py3.6_cuda100_cudnn7_1.tar.bz2
654.7 MB	 | win-64/pytorch-1.0.0-py3.7_cuda100_cudnn7_1.tar.bz2
605.7 MB	 | win-64/pytorch-1.0.0-py3.7_cuda90_cudnn7_1.tar.bz2
580.7 MB	 | win-64/pytorch-1.0.0-py3.5_cuda80_cudnn7_1.tar.bz2
581.1 MB	 | win-64/pytorch-1.0.0-py3.7_cuda80_cudnn7_1.tar.bz2
605.6 MB	 | win-64/pytorch-1.0.0-py3.6_cuda90_cudnn7_1.tar.bz2
606.0 MB	 | win-64/pytorch-1.0.0-py3.5_cuda90_cudnn7_1.tar.bz2

Anything outside those version you your conda will not find pytorch and so will not install your updated fastai.

Here I just removed pytorch and installed the fastai 1.0.34 to simulate your machine:

1-I tried to find pytorch and fastai on my system after that:

$ conda list | grep -E "fastai|pytorch"



dataclasses               0.6                        py_0    fastai
fastai                    1.0.34                     py_1    fastai
fastprogress              0.1.18                     py_0    fastai

2-Then I just tried to install with the commnand I described in earlier posts, and as you can see the plan shows the correct package is about to download:

$ conda install fastai -c pytorch -c fastai



## Package Plan ##

  environment location: /home/ubuntu/anaconda/envs/projects

  added / updated specs: 
    - fastai


The following packages will be downloaded:

    package                    |            build
    ---------------------------|-----------------
    traitlets-4.3.2            |           py36_0         133 KB
    ipython-7.2.0              |   py36h39e3cac_0         1.0 MB
    pytorch-1.0.0              |py3.6_cuda9.0.176_cudnn7.4.1_1       498.6 MB  pytorch
    bleach-3.1.0               |           py36_0         226 KB
    terminado-0.8.1            |           py36_1          21 KB
    nb_conda-2.2.1             |           py36_0          33 KB
    nbformat-4.4.0             |           py36_0         141 KB
    notebook-5.7.4             |           py36_0         7.2 MB
    ptyprocess-0.6.0           |           py36_0          23 KB
    pyzmq-17.1.2               |   py36h14c3975_0         454 KB
    beautifulsoup4-4.7.1       |           py36_1         143 KB
    soupsieve-1.7.1            |           py36_0          50 KB
    fastai-1.0.41              |                1         127 KB  fastai
    pandocfilters-1.4.2        |           py36_1          13 KB
    nb_conda_kernels-2.2.0     |           py36_0          32 KB
    jinja2-2.10                |           py36_0         184 KB
    regex-2018.01.10           |py36h14c3975_1000         338 KB  fastai
    markupsafe-1.1.0           |   py36h7b6447c_0          26 KB
    zeromq-4.2.5               |       hf484d3e_1         567 KB
    spacy-2.0.18               |py36hf484d3e_1000        51.8 MB  fastai
    prompt_toolkit-2.0.7       |           py36_0         482 KB
    testpath-0.4.2             |           py36_0          91 KB
    jedi-0.13.2                |           py36_0         228 KB
    wcwidth-0.1.7              |           py36_0          25 KB
    packaging-18.0             |           py36_0          33 KB
    ninja-1.8.2                |   py36h6bb024c_1         1.3 MB
    libsodium-1.0.16           |       h1bed415_0         302 KB
    pexpect-4.6.0              |           py36_0          77 KB
    jupyter_core-4.4.0         |           py36_0          63 KB
    mistune-0.8.4              |   py36h7b6447c_0          54 KB
    webencodings-0.5.1         |           py36_1          19 KB
    torchvision-0.2.1          |             py_2          37 KB  pytorch
    pygments-2.3.1             |           py36_0         1.4 MB
    nbconvert-5.3.1            |           py36_0         406 KB
    parso-0.3.1                |           py36_0         114 KB
    pandoc-2.2.3.2             |                0        20.8 MB
    send2trash-1.5.0           |           py36_0          16 KB
    gmp-6.1.2                  |       h6c8ec71_1         744 KB
    decorator-4.3.0            |           py36_0          15 KB
    ipython_genutils-0.2.0     |           py36_0          39 KB
    pickleshare-0.7.5          |           py36_0          13 KB
    prometheus_client-0.5.0    |           py36_0          67 KB
    jupyter_client-5.2.4       |           py36_0         128 KB
    ipykernel-5.1.0            |   py36h39e3cac_0         156 KB
    entrypoints-0.3            |           py36_0          12 KB
    nvidia-ml-py3-7.352.0      |             py_0          22 KB  fastai
    backcall-0.1.0             |           py36_0          19 KB
    jsonschema-2.6.0           |           py36_0          62 KB
    ------------------------------------------------------------
                                           Total:       587.7 MB

The following NEW packages will be INSTALLED:

    backcall:          0.1.0-py36_0                                
    beautifulsoup4:    4.7.1-py36_1                                
    bleach:            3.1.0-py36_0                                
    decorator:         4.3.0-py36_0                                
    entrypoints:       0.3-py36_0                                  
    gmp:               6.1.2-h6c8ec71_1                            
    ipykernel:         5.1.0-py36h39e3cac_0                        
    ipython:           7.2.0-py36h39e3cac_0                        
    ipython_genutils:  0.2.0-py36_0                                
    jedi:              0.13.2-py36_0                               
    jinja2:            2.10-py36_0                                 
    jsonschema:        2.6.0-py36_0                                
    jupyter_client:    5.2.4-py36_0                                
    jupyter_core:      4.4.0-py36_0                                
    libsodium:         1.0.16-h1bed415_0                           
    markupsafe:        1.1.0-py36h7b6447c_0                        
    mistune:           0.8.4-py36h7b6447c_0                        
    nb_conda:          2.2.1-py36_0                                
    nb_conda_kernels:  2.2.0-py36_0                                
    nbconvert:         5.3.1-py36_0                                
    nbformat:          4.4.0-py36_0                                
    ninja:             1.8.2-py36h6bb024c_1                        
    notebook:          5.7.4-py36_0                                
    nvidia-ml-py3:     7.352.0-py_0                         fastai 
    packaging:         18.0-py36_0                                 
    pandoc:            2.2.3.2-0                                   
    pandocfilters:     1.4.2-py36_1                                
    parso:             0.3.1-py36_0                                
    pexpect:           4.6.0-py36_0                                
    pickleshare:       0.7.5-py36_0                                
    prometheus_client: 0.5.0-py36_0                                
    prompt_toolkit:    2.0.7-py36_0                                
    ptyprocess:        0.6.0-py36_0                                
    pygments:          2.3.1-py36_0                                
    pytorch:           1.0.0-py3.6_cuda9.0.176_cudnn7.4.1_1 pytorch
    pyzmq:             17.1.2-py36h14c3975_0                       
    send2trash:        1.5.0-py36_0                                
    soupsieve:         1.7.1-py36_0                                
    terminado:         0.8.1-py36_1                                
    testpath:          0.4.2-py36_0                                
    torchvision:       0.2.1-py_2                           pytorch
    traitlets:         4.3.2-py36_0                                
    wcwidth:           0.1.7-py36_0                                
    webencodings:      0.5.1-py36_1                                
    zeromq:            4.2.5-hf484d3e_1                            

The following packages will be UPDATED:

    fastai:            1.0.34-py_1                          fastai  --> 1.0.41-1                     fastai
    spacy:             2.0.16-py36h962f231_0                        --> 2.0.18-py36hf484d3e_1000     fastai

The following packages will be DOWNGRADED:

    regex:             2018.08.29-py36h7b6447c_0                    --> 2018.01.10-py36h14c3975_1000 fastai

Thanks very much Willismar for your step-by-step guide. I followed it (see below) but did not work on my computer :frowning:
My configuration: Windows 10, Anaconda Prompt, Fastai v1 installed through the conda procedure, python version 3.6.8 (build: h9f7ef89_0).

1. Activation of my fastai v1 environment.

(base) C:\Users(…)>activate fastai_v1

2. Getting the list and version of my packages with fastai and pytorch.

(fastai_v1) C:\Users(…)>conda list | grep -E “fastai|pytorch”

packages in environment at C:\Users(…)\Anaconda3\envs\fastai_v1:

dataclasses               0.6                        py_0    fastai
fastai                    1.0.38                     py_1    fastai
fastprogress              0.1.18                     py_0    fastai
nvidia-ml-py3             7.352.0                    py_0    fastai
pytorch                   1.0.0                     py3.6_cuda90_cudnn7_1    pytorch
torchvision               0.2.1                      py_2    pytorch

3. Trying update fastai version.

(fastai_v1) C:\Users(…)>conda install fastai -c pytorch -c fastai

Solving environment: done
All requested packages already installed.

4. Checking my fastai version.

(fastai_v1) C:\Users(…)>conda list fastai
packages in environment at C:\Users(…)\Anaconda3\envs\fastai_v1:

Name                    Version                   Build  Channel
fastai                   1.0.38                   py_1    fastai

Let me ask you some dumb questions…

  • Do you have other python installed on your machine ? If yes what is it’s folder and what do you have in environment variables ?
  • Have you ever installed some package using pip ?
  • What is the output of pip freeze ?
  • What is the output of conda list ?
  • What is the folder of anaconda is installed ?
  • What is the folder of python packages ?

[SOLVED]

Hi @pierreguillou

Ok I got some time to install a virtual machine and simulate your problem here…

1.Creating the environment

Just creating the environment using python 3.6.8 or 3.7.2 there is not difference here…

(base) C:\Users\Willismar>conda create -n py372 python=3.7.2
Solving environment: done

## Package Plan ##

  environment location: C:\Users\Willismar\Anaconda3\envs\py372

  added / updated specs:
    - python=3.7.2


The following packages will be downloaded:

    package                    |            build
    ---------------------------|-----------------
    python-3.7.2               |       h8c8aaf0_0        17.7 MB

The following NEW packages will be INSTALLED:

    ca-certificates: 2018.12.5-0
    certifi:         2018.11.29-py37_0
    openssl:         1.1.1a-he774522_0
    pip:             18.1-py37_0
    python:          3.7.2-h8c8aaf0_0
    setuptools:      40.6.3-py37_0
    sqlite:          3.26.0-he774522_0
    vc:              14.1-h0510ff6_4
    vs2015_runtime:  14.15.26706-h3a45250_0
    wheel:           0.32.3-py37_0
    wincertstore:    0.2-py37_0

Proceed ([y]/n)? y


Downloading and Extracting Packages
python-3.7.2         | 17.7 MB   | #################################### | 100%
Preparing transaction: done
Verifying transaction: done
Executing transaction: done
#
# To activate this environment, use
#
#     $ conda activate py372
#
# To deactivate an active environment, use
#
#     $ conda deactivate

2.Activating the environment

(base) C:\Users\Willismar>conda activate py372
(py372) C:\Users\Willismar>

3.Installing only fastai (1.0.34)

(py372) C:\Users\Willismar>conda install fastai -c fastai
Solving environment: done

## Package Plan ##

  environment location: C:\Users\Willismar\Anaconda3\envs\py372

  added / updated specs:
    - fastai


The following packages will be downloaded:

    package                    |            build
    ---------------------------|-----------------
    numpy-1.15.4               |   py37ha559c80_0          36 KB
    numpy-base-1.15.4          |   py37h8128ebf_0         3.9 MB
    plac-0.9.6                 |           py37_0          36 KB
    thinc-6.12.0               |   py37hb854c30_0         1.4 MB
    numexpr-2.6.8              |   py37h9ef55f4_0         129 KB
    fastai-1.0.34              |             py_1         110 KB  fastai
    scipy-1.1.0                |   py37h4f6bf74_1        13.5 MB
    pyreadline-2.1             |           py37_1         141 KB
    cymem-2.0.2                |   py37h74a9793_0          35 KB
    pyparsing-2.3.1            |           py37_0         105 KB
    pytz-2018.9                |           py37_0         264 KB
    mkl_fft-1.0.6              |   py37hdbbee80_0         120 KB
    typing-3.6.4               |           py37_0          44 KB
    dill-0.2.8.2               |           py37_0         134 KB
    preshed-2.0.1              |   py37h33f27b4_0          70 KB
    pillow-5.4.1               |   py37hdc69c19_0         686 KB
    mkl_random-1.0.1           |   py37h77b88f5_1         267 KB
    spacy-2.0.16               |   py37hcce6980_0        46.9 MB
    msgpack-numpy-0.4.3.2      |           py37_0          14 KB
    regex-2018.08.29           |   py37hfa6e2cd_0         330 KB
    ujson-1.35                 |   py37hfa6e2cd_0          26 KB
    murmurhash-1.0.1           |   py37h33f27b4_0          20 KB
    ------------------------------------------------------------
                                           Total:        68.2 MB

The following NEW packages will be INSTALLED:

    asn1crypto:      0.24.0-py37_0
    blas:            1.0-mkl
    bottleneck:      1.2.1-py37h452e1ab_1
    cffi:            1.11.5-py37h74b6da3_1
    chardet:         3.0.4-py37_1
    cryptography:    2.4.2-py37h7a1dbc1_0
    cycler:          0.10.0-py37_0
    cymem:           2.0.2-py37h74a9793_0
    cytoolz:         0.9.0.1-py37hfa6e2cd_1
    dataclasses:     0.6-py_0                  fastai
    dill:            0.2.8.2-py37_0
    fastai:          1.0.34-py_1               fastai
    fastprogress:    0.1.18-py_0               fastai
    freetype:        2.9.1-ha9979f8_1
    icc_rt:          2019.0.0-h0cc432a_1
    icu:             58.2-ha66f8fd_1
    idna:            2.8-py37_0
    intel-openmp:    2019.1-144
    jpeg:            9b-hb83a4c4_2
    kiwisolver:      1.0.1-py37h6538335_0
    libpng:          1.6.36-h2a8f88b_0
    libtiff:         4.0.10-h2929a5b_1001
    matplotlib:      3.0.2-py37hc8f65d3_0
    mkl:             2018.0.3-1
    mkl_fft:         1.0.6-py37hdbbee80_0
    mkl_random:      1.0.1-py37h77b88f5_1
    msgpack-numpy:   0.4.3.2-py37_0
    msgpack-python:  0.5.6-py37he980bc4_1
    murmurhash:      1.0.1-py37h33f27b4_0
    numexpr:         2.6.8-py37h9ef55f4_0
    numpy:           1.15.4-py37ha559c80_0
    numpy-base:      1.15.4-py37h8128ebf_0
    olefile:         0.46-py37_0
    pandas:          0.23.4-py37h830ac7b_0
    pillow:          5.4.1-py37hdc69c19_0
    plac:            0.9.6-py37_0
    preshed:         2.0.1-py37h33f27b4_0
    pycparser:       2.19-py37_0
    pyopenssl:       18.0.0-py37_0
    pyparsing:       2.3.1-py37_0
    pyqt:            5.9.2-py37h6538335_2
    pyreadline:      2.1-py37_1
    pysocks:         1.6.8-py37_0
    python-dateutil: 2.7.5-py37_0
    pytz:            2018.9-py37_0
    pyyaml:          3.13-py37hfa6e2cd_0
    qt:              5.9.7-vc14h73c81de_0
    regex:           2018.08.29-py37hfa6e2cd_0
    requests:        2.21.0-py37_0
    scipy:           1.1.0-py37h4f6bf74_1
    sip:             4.19.8-py37h6538335_0
    six:             1.12.0-py37_0
    spacy:           2.0.16-py37hcce6980_0
    thinc:           6.12.0-py37hb854c30_0
    tk:              8.6.8-hfa6e2cd_0
    toolz:           0.9.0-py37_0
    tornado:         5.1.1-py37hfa6e2cd_0
    tqdm:            4.29.1-py_0
    typing:          3.6.4-py37_0
    ujson:           1.35-py37hfa6e2cd_0
    urllib3:         1.24.1-py37_0
    win_inet_pton:   1.0.1-py37_1
    wrapt:           1.10.11-py37hfa6e2cd_2
    yaml:            0.1.7-hc54c509_2
    zlib:            1.2.11-h62dcd97_3

Proceed ([y]/n)? y


Downloading and Extracting Packages
numpy-1.15.4         | 36 KB     | #################################### | 100%
numpy-base-1.15.4    | 3.9 MB    | #################################### | 100%
plac-0.9.6           | 36 KB     | #################################### | 100%
thinc-6.12.0         | 1.4 MB    | #################################### | 100%
numexpr-2.6.8        | 129 KB    | #################################### | 100%
fastai-1.0.34        | 110 KB    | #################################### | 100%
scipy-1.1.0          | 13.5 MB   | #################################### | 100%
pyreadline-2.1       | 141 KB    | #################################### | 100%
cymem-2.0.2          | 35 KB     | #################################### | 100%
pyparsing-2.3.1      | 105 KB    | #################################### | 100%
pytz-2018.9          | 264 KB    | #################################### | 100%
mkl_fft-1.0.6        | 120 KB    | #################################### | 100%
typing-3.6.4         | 44 KB     | #################################### | 100%
dill-0.2.8.2         | 134 KB    | #################################### | 100%
preshed-2.0.1        | 70 KB     | #################################### | 100%
pillow-5.4.1         | 686 KB    | #################################### | 100%
mkl_random-1.0.1     | 267 KB    | #################################### | 100%
spacy-2.0.16         | 46.9 MB   | #################################### | 100%
msgpack-numpy-0.4.3. | 14 KB     | #################################### | 100%
regex-2018.08.29     | 330 KB    | #################################### | 100%
ujson-1.35           | 26 KB     | #################################### | 100%
murmurhash-1.0.1     | 20 KB     | #################################### | 100%
Preparing transaction: done
Verifying transaction: done
Executing transaction: done

(py372) C:\Users\Willismar>

4.Updating fastai to 1.0.38

(py372) C:\Users\Willismar>conda install -c fastai -c pytorch fastai
Solving environment: done

## Package Plan ##

  environment location: C:\Users\Willismar\Anaconda3\envs\py372

  added / updated specs:
    - fastai


The following packages will be downloaded:

    package                    |            build
    ---------------------------|-----------------
    ninja-1.8.2                |   py37he980bc4_1         109 KB
    pytorch-1.0.0              |py3.7_cuda90_cudnn7_1       605.7 MB  pytorch
    ------------------------------------------------------------
                                           Total:       605.8 MB

The following NEW packages will be INSTALLED:

    ninja:         1.8.2-py37he980bc4_1
    nvidia-ml-py3: 7.352.0-py_0                fastai
    packaging:     18.0-py37_0
    pytorch:       1.0.0-py3.7_cuda90_cudnn7_1 pytorch
    torchvision:   0.2.1-py_2                  pytorch

The following packages will be UPDATED:

    fastai:        1.0.34-py_1                 fastai  --> 1.0.38-py_1 fastai

Proceed ([y]/n)? y


Downloading and Extracting Packages
ninja-1.8.2          | 109 KB    | #################################### | 100%
pytorch-1.0.0        | 605.7 MB  | #################################### | 100%
Preparing transaction: done
Verifying transaction: done
Executing transaction: done

(py372) C:\Users\Willismar>

5.Trying to update to fastai=1.0.41

I can find the version on repository

(py372) C:\Users\Willismar>conda search fastai=1.0.41 -c fastai
Loading channels: done
# Name                  Version           Build  Channel
fastai                   1.0.41               1  fastai

Trying to install I can see that it shows the problem

(py372) C:\Users\Willismar>conda install fastai=1.0.41 -c fastai
Solving environment: failed

PackagesNotFoundError: The following packages are not available from current cha
nnels:

  - fastai=1.0.41
  - spacy[version='>=2.0.18']

Current channels:

  - https://conda.anaconda.org/fastai/win-64
  - https://conda.anaconda.org/fastai/noarch
  - https://repo.anaconda.com/pkgs/main/win-64
  - https://repo.anaconda.com/pkgs/main/noarch
  - https://repo.anaconda.com/pkgs/free/win-64
  - https://repo.anaconda.com/pkgs/free/noarch
  - https://repo.anaconda.com/pkgs/r/win-64
  - https://repo.anaconda.com/pkgs/r/noarch
  - https://repo.anaconda.com/pkgs/pro/win-64
  - https://repo.anaconda.com/pkgs/pro/noarch
  - https://repo.anaconda.com/pkgs/msys2/win-64
  - https://repo.anaconda.com/pkgs/msys2/noarch

To search for alternate channels that may provide the conda package you're
looking for, navigate to

    https://anaconda.org

and use the search bar at the top of the page.



(py372) C:\Users\Willismar>

Conclusion

The problem is :slight_smile:, because of this, the lack of the release version of spacy 2.0.18

PackagesNotFoundError: The following packages are not available from current cha
nnels:

  - fastai=1.0.41
  - spacy[version='>=2.0.18']

Let’s try to find it on the default channel:

(py372) C:\Users\Willismar>conda search spacy
Loading channels: done
# Name                  Version           Build  Channel
spacy                     1.9.0 py27_vc9h1d62622_1  pkgs/main
spacy                     1.9.0 py35_vc14hcdb467c_1  pkgs/main
spacy                     1.9.0 py36_vc14hffcad6b_1  pkgs/main
spacy                     2.0.5  py27hc56fc5f_0  pkgs/main
spacy                     2.0.5  py35h6538335_0  pkgs/main
spacy                     2.0.5  py36h6538335_0  pkgs/main
spacy                     2.0.9  py27h504ef13_0  pkgs/main
spacy                     2.0.9  py35h8300f20_0  pkgs/main
spacy                     2.0.9  py36h8300f20_0  pkgs/main
spacy                    2.0.10  py27h39f3610_0  pkgs/main
spacy                    2.0.10  py35h830ac7b_0  pkgs/main
spacy                    2.0.10  py36h830ac7b_0  pkgs/main
spacy                    2.0.11  py27h504ef13_0  pkgs/main
spacy                    2.0.11  py27h504ef13_2  pkgs/main
spacy                    2.0.11  py35h8300f20_0  pkgs/main
spacy                    2.0.11  py35h8300f20_2  pkgs/main
spacy                    2.0.11  py36h8300f20_0  pkgs/main
spacy                    2.0.11  py36h8300f20_2  pkgs/main
spacy                    2.0.11  py37h8300f20_2  pkgs/main
spacy                    2.0.12  py27h504ef13_0  pkgs/main
spacy                    2.0.12  py35h8300f20_0  pkgs/main
spacy                    2.0.12  py36h8300f20_0  pkgs/main
spacy                    2.0.12  py37h8300f20_0  pkgs/main
spacy                    2.0.16  py27h504ef13_0  pkgs/main
spacy                    2.0.16  py36hcce6980_0  pkgs/main
spacy                    2.0.16  py37hcce6980_0  pkgs/main

(py372) C:\Users\Willismar>

Let’s try to find it on the spacy channel.
We can see that the version v2.0.18 also isn’t released there yet:

(py372) C:\Users\Willismar>conda search spacy -c spacy
Loading channels: done
# Name                  Version           Build  Channel
spacy                   0.100.5          py27_0  spacy
spacy                   0.100.5          py35_0  spacy
spacy                   0.100.6          py27_0  spacy
spacy                   0.100.6          py35_0  spacy
spacy                   0.100.7      py27_vc9_0  spacy
spacy                   0.100.7     py35_vc14_0  spacy
spacy                   0.101.0      py27_vc9_0  spacy
spacy                   0.101.0     py35_vc14_0  spacy

But take a look on this other channel conda-forge:

(py372) C:\Users\Willismar>conda search spacy -c conda-forge
Loading channels: done
# Name                  Version           Build  Channel
spacy                     1.1.0 np111py27_vc9_1  conda-forge
spacy                     1.1.0 np111py35_vc14_0  conda-forge
spacy                     1.1.0 np111py35_vc14_1  conda-forge
spacy                     1.1.0 np111py36_vc14_0  conda-forge
spacy                     1.1.0 np111py36_vc14_1  conda-forge
spacy                     1.1.0 np112py27_vc9_1  conda-forge
spacy                     1.1.0 np112py35_vc14_0  conda-forge
spacy                     1.1.0 np112py35_vc14_1  conda-forge
spacy                     1.1.0 np112py36_vc14_0  conda-forge
spacy                     1.1.0 np112py36_vc14_1  conda-forge
spacy                     1.2.0 np111py27_vc9_1  conda-forge
spacy                     1.2.0 np111py35_vc14_1  conda-forge
spacy                     1.2.0 np111py36_vc14_1  conda-forge
spacy                     1.2.0 np112py27_vc9_1  conda-forge
spacy                     1.2.0 np112py35_vc14_1  conda-forge
spacy                     1.2.0 np112py36_vc14_1  conda-forge
spacy                     1.3.0 np111py27_vc9_0  conda-forge
spacy                     1.3.0 np111py35_vc14_0  conda-forge
spacy                     1.3.0 np111py36_vc14_0  conda-forge
spacy                     1.3.0 np112py27_vc9_0  conda-forge
spacy                     1.3.0 np112py35_vc14_0  conda-forge
spacy                     1.3.0 np112py36_vc14_0  conda-forge
spacy                     1.4.0 np111py27_vc9_0  conda-forge
spacy                     1.4.0 np111py35_vc14_0  conda-forge
spacy                     1.4.0 np111py36_vc14_0  conda-forge
spacy                     1.4.0 np112py27_vc9_0  conda-forge
spacy                     1.4.0 np112py35_vc14_0  conda-forge
spacy                     1.4.0 np112py36_vc14_0  conda-forge
spacy                     1.5.1 np111py27_vc9_0  conda-forge
spacy                     1.5.1 np111py35_vc14_0  conda-forge
spacy                     1.5.1 np111py36_vc14_0  conda-forge
spacy                     1.5.1 np112py27_vc9_0  conda-forge
spacy                     1.5.1 np112py35_vc14_0  conda-forge
spacy                     1.5.1 np112py36_vc14_0  conda-forge
spacy                     1.6.0 np111py27_vc9_0  conda-forge
spacy                     1.6.0 np111py35_vc14_0  conda-forge
spacy                     1.6.0 np111py36_vc14_0  conda-forge
spacy                     1.6.0 np112py27_vc9_0  conda-forge
spacy                     1.6.0 np112py35_vc14_0  conda-forge
spacy                     1.6.0 np112py36_vc14_0  conda-forge
spacy                     1.7.2 np111py27_vc9_0  conda-forge
spacy                     1.7.2 np111py35_vc14_0  conda-forge
spacy                     1.7.2 np111py36_vc14_0  conda-forge
spacy                     1.7.2 np112py27_vc9_0  conda-forge
spacy                     1.7.2 np112py35_vc14_0  conda-forge
spacy                     1.7.2 np112py36_vc14_0  conda-forge
spacy                     1.7.3 np111py27_vc9_0  conda-forge
spacy                     1.7.3 np111py35_vc14_0  conda-forge
spacy                     1.7.3 np111py36_vc14_0  conda-forge
spacy                     1.7.3 np112py27_vc9_0  conda-forge
spacy                     1.7.3 np112py35_vc14_0  conda-forge
spacy                     1.7.3 np112py36_vc14_0  conda-forge
spacy                     1.8.1 np111py27_vc9_0  conda-forge
spacy                     1.8.1 np111py35_vc14_0  conda-forge
spacy                     1.8.1 np111py36_vc14_0  conda-forge
spacy                     1.8.1 np112py27_vc9_0  conda-forge
spacy                     1.8.1 np112py35_vc14_0  conda-forge
spacy                     1.8.1 np112py36_vc14_0  conda-forge
spacy                     1.8.2 np111py27_vc9_0  conda-forge
spacy                     1.8.2 np111py35_vc14_0  conda-forge
spacy                     1.8.2 np111py36_vc14_0  conda-forge
spacy                     1.8.2 np112py27_vc9_0  conda-forge
spacy                     1.8.2 np112py35_vc14_0  conda-forge
spacy                     1.8.2 np112py36_vc14_0  conda-forge
spacy                     1.9.0 np111py27_vc9_0  conda-forge
spacy                     1.9.0 np111py27_vc9_1  conda-forge
spacy                     1.9.0 np111py35_vc14_0  conda-forge
spacy                     1.9.0 np111py35_vc14_1  conda-forge
spacy                     1.9.0 np111py36_vc14_0  conda-forge
spacy                     1.9.0 np111py36_vc14_1  conda-forge
spacy                     1.9.0 np112py27_vc9_0  conda-forge
spacy                     1.9.0 np112py27_vc9_1  conda-forge
spacy                     1.9.0 np112py35_vc14_0  conda-forge
spacy                     1.9.0 np112py35_vc14_1  conda-forge
spacy                     1.9.0 np112py36_vc14_0  conda-forge
spacy                     1.9.0 np112py36_vc14_1  conda-forge
spacy                     1.9.0 py27_vc9h1d62622_1  pkgs/main
spacy                     1.9.0 py35_vc14hcdb467c_1  pkgs/main
spacy                     1.9.0 py36_vc14hffcad6b_1  pkgs/main
spacy                     2.0.2      py27_vc9_0  conda-forge
spacy                     2.0.2     py35_vc14_0  conda-forge
spacy                     2.0.2     py36_vc14_0  conda-forge
spacy                     2.0.3      py27_vc9_0  conda-forge
spacy                     2.0.3     py35_vc14_0  conda-forge
spacy                     2.0.3     py36_vc14_0  conda-forge
spacy                     2.0.4      py27_vc9_0  conda-forge
spacy                     2.0.4     py35_vc14_0  conda-forge
spacy                     2.0.4     py36_vc14_0  conda-forge
spacy                     2.0.5  py27hc56fc5f_0  pkgs/main
spacy                     2.0.5  py35h6538335_0  pkgs/main
spacy                     2.0.5  py36h6538335_0  pkgs/main
spacy                     2.0.6      py27_vc9_2  conda-forge
spacy                     2.0.6     py35_vc14_2  conda-forge
spacy                     2.0.6     py36_vc14_2  conda-forge
spacy                     2.0.7      py27_vc9_2  conda-forge
spacy                     2.0.7     py35_vc14_2  conda-forge
spacy                     2.0.7     py36_vc14_2  conda-forge
spacy                     2.0.8      py27_vc9_0  conda-forge
spacy                     2.0.8     py35_vc14_0  conda-forge
spacy                     2.0.8     py36_vc14_0  conda-forge
spacy                     2.0.9      py27_vc9_0  conda-forge
spacy                     2.0.9  py27h504ef13_0  pkgs/main
spacy                     2.0.9     py35_vc14_0  conda-forge
spacy                     2.0.9  py35h8300f20_0  pkgs/main
spacy                     2.0.9     py36_vc14_0  conda-forge
spacy                     2.0.9  py36h8300f20_0  pkgs/main
spacy                    2.0.10      py27_vc9_0  conda-forge
spacy                    2.0.10  py27h39f3610_0  pkgs/main
spacy                    2.0.10     py35_vc14_0  conda-forge
spacy                    2.0.10  py35h830ac7b_0  pkgs/main
spacy                    2.0.10     py36_vc14_0  conda-forge
spacy                    2.0.10  py36h830ac7b_0  pkgs/main
spacy                    2.0.11      py27_vc9_0  conda-forge
spacy                    2.0.11  py27h39f3610_1  conda-forge
spacy                    2.0.11  py27h504ef13_0  pkgs/main
spacy                    2.0.11  py27h504ef13_2  pkgs/main
spacy                    2.0.11     py35_vc14_0  conda-forge
spacy                    2.0.11  py35h8300f20_0  pkgs/main
spacy                    2.0.11  py35h8300f20_2  pkgs/main
spacy                    2.0.11  py35h830ac7b_1  conda-forge
spacy                    2.0.11     py36_vc14_0  conda-forge
spacy                    2.0.11  py36h8300f20_0  pkgs/main
spacy                    2.0.11  py36h8300f20_2  pkgs/main
spacy                    2.0.11  py36h830ac7b_1  conda-forge
spacy                    2.0.11  py37h8300f20_2  pkgs/main
spacy                    2.0.12  py27h39f3610_0  conda-forge
spacy                    2.0.12  py27h504ef13_0  pkgs/main
spacy                    2.0.12  py35h8300f20_0  pkgs/main
spacy                    2.0.12  py35h830ac7b_0  conda-forge
spacy                    2.0.12  py36h8300f20_0  pkgs/main
spacy                    2.0.12  py36h830ac7b_0  conda-forge
spacy                    2.0.12  py37h8300f20_0  pkgs/main
spacy                    2.0.13  py27h39f3610_0  conda-forge
spacy                    2.0.13  py36h830ac7b_0  conda-forge
spacy                    2.0.16  py27h504ef13_0  pkgs/main
spacy                    2.0.16 py27hc56fc5f_1000  conda-forge
spacy                    2.0.16 py27hc56fc5f_1001  conda-forge
spacy                    2.0.16 py36h6538335_1000  conda-forge
spacy                    2.0.16 py36h6538335_1001  conda-forge
spacy                    2.0.16  py36hcce6980_0  pkgs/main
spacy                    2.0.16 py37h6538335_1000  conda-forge
spacy                    2.0.16 py37h6538335_1001  conda-forge
spacy                    2.0.16  py37hcce6980_0  pkgs/main
spacy                    2.0.17 py27hc56fc5f_1001  conda-forge
spacy                    2.0.17 py36h6538335_1001  conda-forge
spacy                    2.0.17 py37h6538335_1001  conda-forge
spacy                    2.0.18 py27hc56fc5f_1000  conda-forge
spacy                    2.0.18 py36h6538335_1000  conda-forge
spacy                    2.0.18 py37h6538335_1000  conda-forge

(py372) C:\Users\Willismar>

6.Forcing installing Spacy 2.0.18 from (conda-forge)

(py372) C:\Users\Willismar>conda install spacy -c conda-forge
Solving environment: done

## Package Plan ##

  environment location: C:\Users\Willismar\Anaconda3\envs\py372

  added / updated specs:
    - spacy


The following packages will be downloaded:

    package                    |            build
    ---------------------------|-----------------
    ca-certificates-2018.11.29 |       ha4d7672_0         179 KB  conda-forge
    certifi-2018.11.29         |        py37_1000         144 KB  conda-forge
    ------------------------------------------------------------
                                           Total:         323 KB

The following packages will be UPDATED:

    certifi:         2018.11.29-py37_0 --> 2018.11.29-py37_1000  conda-forge

The following packages will be DOWNGRADED:

    ca-certificates: 2018.12.5-0       --> 2018.11.29-ha4d7672_0 conda-forge

Proceed ([y]/n)? y


Downloading and Extracting Packages
ca-certificates-2018 | 179 KB    | #################################### | 100%
certifi-2018.11.29   | 144 KB    | #################################### | 100%
Preparing transaction: done
Verifying transaction: done
Executing transaction: done

(py372) C:\Users\Willismar>

7.Forcing installing fastai v1.0.41

Passing only the fastai channel wont work, because previously only the certificates was updated.
Then it still ask the system to have the correct version

(py372) C:\Users\Willismar>conda install fastai=1.0.41 -c fastai
Solving environment: failed

PackagesNotFoundError: The following packages are not available from current cha
nnels:

  - fastai=1.0.41
  - spacy[version='>=2.0.18']

Current channels:

  - https://conda.anaconda.org/fastai/win-64
  - https://conda.anaconda.org/fastai/noarch
  - https://repo.anaconda.com/pkgs/main/win-64
  - https://repo.anaconda.com/pkgs/main/noarch
  - https://repo.anaconda.com/pkgs/free/win-64
  - https://repo.anaconda.com/pkgs/free/noarch
  - https://repo.anaconda.com/pkgs/r/win-64
  - https://repo.anaconda.com/pkgs/r/noarch
  - https://repo.anaconda.com/pkgs/pro/win-64
  - https://repo.anaconda.com/pkgs/pro/noarch
  - https://repo.anaconda.com/pkgs/msys2/win-64
  - https://repo.anaconda.com/pkgs/msys2/noarch

To search for alternate channels that may provide the conda package you're
looking for, navigate to

    https://anaconda.org

and use the search bar at the top of the page.

Passing the conda-forge channel together make everything agreed.

(py372) C:\Users\Willismar>conda install fastai=1.0.41 -c fastai -c conda-forge
Solving environment: done

## Package Plan ##

  environment location: C:\Users\Willismar\Anaconda3\envs\py372

  added / updated specs:
    - fastai=1.0.41


The following packages will be downloaded:

    package                    |            build
    ---------------------------|-----------------
    spacy-2.0.18               |py37h6538335_1000        50.5 MB  conda-forge
    ipykernel-5.1.0            |py37h39e3cac_1001         170 KB  conda-forge
    markupsafe-1.1.0           |py37hfa6e2cd_1000          27 KB  conda-forge
    terminado-0.8.1            |        py37_1001          22 KB  conda-forge
    ipython_genutils-0.2.0     |             py_1          21 KB  conda-forge
    notebook-5.7.4             |        py37_1000         7.4 MB  conda-forge
    parso-0.3.1                |             py_0          59 KB  conda-forge
    fastai-1.0.41              |                1         127 KB  fastai
    decorator-4.3.0            |             py_0          10 KB  conda-forge
    jinja2-2.10                |             py_1          89 KB  conda-forge
    pywinpty-0.5.5             |        py37_1000          46 KB  conda-forge
    jupyter_client-5.2.4       |             py_0          63 KB  conda-forge
    soupsieve-1.7.1            |        py37_1000          48 KB  conda-forge
    prompt_toolkit-2.0.7       |             py_0         218 KB  conda-forge
    colorama-0.4.1             |             py_0          15 KB  conda-forge
    zeromq-4.2.5               |    he025d50_1006         9.6 MB  conda-forge
    attrs-18.2.0               |             py_0          30 KB  conda-forge
    pyzmq-17.1.2               |py37hf576995_1001         396 KB  conda-forge
    jupyter_core-4.4.0         |             py_0          44 KB  conda-forge
    prometheus_client-0.5.0    |             py_0          34 KB  conda-forge
    nbconvert-5.3.1            |             py_1         320 KB  conda-forge
    winpty-0.4.3               |                4         1.1 MB  conda-forge
    pickleshare-0.7.5          |        py37_1000          12 KB  conda-forge
    libsodium-1.0.16           |    h2fa13f4_1001         582 KB  conda-forge
    mistune-0.8.4              |py37hfa6e2cd_1000          53 KB  conda-forge
    webencodings-0.5.1         |             py_1          12 KB  conda-forge
    traitlets-4.3.2            |        py37_1000         130 KB  conda-forge
    bleach-3.1.0               |             py_0         110 KB  conda-forge
    entrypoints-0.3            |        py37_1000          10 KB  conda-forge
    send2trash-1.5.0           |             py_0          12 KB  conda-forge
    jsonschema-3.0.0a3         |        py37_1000          98 KB  conda-forge
    pandoc-2.5                 |                1        46.0 MB  conda-forge
    regex-2018.01.10           |py37hfa6e2cd_1000         319 KB  conda-forge
    beautifulsoup4-4.7.1       |        py37_1001         140 KB  conda-forge
    wcwidth-0.1.7              |             py_1          17 KB  conda-forge
    pandocfilters-1.4.2        |             py_1           9 KB  conda-forge
    pygments-2.3.1             |             py_0         641 KB  conda-forge
    nb_conda-2.2.1             |           py37_0          33 KB
    testpath-0.4.2             |        py37_1000          90 KB  conda-forge
    nb_conda_kernels-2.2.0     |        py37_1000          31 KB  conda-forge
    jedi-0.13.2                |        py37_1000         228 KB  conda-forge
    ipython-7.2.0              |py37h39e3cac_1000         1.1 MB  conda-forge
    nbformat-4.4.0             |             py_1          91 KB  conda-forge
    pyrsistent-0.14.9          |py37hfa6e2cd_1000          89 KB  conda-forge
    backcall-0.1.0             |             py_0          13 KB  conda-forge
    ------------------------------------------------------------
                                           Total:       120.1 MB

The following NEW packages will be INSTALLED:

    attrs:                   18.2.0-py_0               conda-forge
    backcall:                0.1.0-py_0                conda-forge
    beautifulsoup4:          4.7.1-py37_1001           conda-forge
    bleach:                  3.1.0-py_0                conda-forge
    colorama:                0.4.1-py_0                conda-forge
    decorator:               4.3.0-py_0                conda-forge
    entrypoints:             0.3-py37_1000             conda-forge
    ipykernel:               5.1.0-py37h39e3cac_1001   conda-forge
    ipython:                 7.2.0-py37h39e3cac_1000   conda-forge
    ipython_genutils:        0.2.0-py_1                conda-forge
    jedi:                    0.13.2-py37_1000          conda-forge
    jinja2:                  2.10-py_1                 conda-forge
    jsonschema:              3.0.0a3-py37_1000         conda-forge
    jupyter_client:          5.2.4-py_0                conda-forge
    jupyter_core:            4.4.0-py_0                conda-forge
    libsodium:               1.0.16-h2fa13f4_1001      conda-forge
    m2w64-gcc-libgfortran:   5.3.0-6
    m2w64-gcc-libs:          5.3.0-7
    m2w64-gcc-libs-core:     5.3.0-7
    m2w64-gmp:               6.1.0-2
    m2w64-libwinpthread-git: 5.0.0.4634.697f757-2
    markupsafe:              1.1.0-py37hfa6e2cd_1000   conda-forge
    mistune:                 0.8.4-py37hfa6e2cd_1000   conda-forge
    msys2-conda-epoch:       20160418-1
    nb_conda:                2.2.1-py37_0
    nb_conda_kernels:        2.2.0-py37_1000           conda-forge
    nbconvert:               5.3.1-py_1                conda-forge
    nbformat:                4.4.0-py_1                conda-forge
    notebook:                5.7.4-py37_1000           conda-forge
    pandoc:                  2.5-1                     conda-forge
    pandocfilters:           1.4.2-py_1                conda-forge
    parso:                   0.3.1-py_0                conda-forge
    pickleshare:             0.7.5-py37_1000           conda-forge
    prometheus_client:       0.5.0-py_0                conda-forge
    prompt_toolkit:          2.0.7-py_0                conda-forge
    pygments:                2.3.1-py_0                conda-forge
    pyrsistent:              0.14.9-py37hfa6e2cd_1000  conda-forge
    pywinpty:                0.5.5-py37_1000           conda-forge
    pyzmq:                   17.1.2-py37hf576995_1001  conda-forge
    send2trash:              1.5.0-py_0                conda-forge
    soupsieve:               1.7.1-py37_1000           conda-forge
    terminado:               0.8.1-py37_1001           conda-forge
    testpath:                0.4.2-py37_1000           conda-forge
    traitlets:               4.3.2-py37_1000           conda-forge
    wcwidth:                 0.1.7-py_1                conda-forge
    webencodings:            0.5.1-py_1                conda-forge
    winpty:                  0.4.3-4                   conda-forge
    zeromq:                  4.2.5-he025d50_1006       conda-forge

The following packages will be UPDATED:

    fastai:                  1.0.38-py_1               fastai      --> 1.0.41-1
                    fastai
    spacy:                   2.0.16-py37hcce6980_0                 --> 2.0.18-py
37h6538335_1000     conda-forge

The following packages will be DOWNGRADED:

    regex:                   2018.08.29-py37hfa6e2cd_0             --> 2018.01.1
0-py37hfa6e2cd_1000 conda-forge

Proceed ([y]/n)?

Proceed ([y]/n)? y


Downloading and Extracting Packages
spacy-2.0.18         | 50.5 MB   | #################################### | 100%
ipykernel-5.1.0      | 170 KB    | #################################### | 100%
markupsafe-1.1.0     | 27 KB     | #################################### | 100%
terminado-0.8.1      | 22 KB     | #################################### | 100%
ipython_genutils-0.2 | 21 KB     | #################################### | 100%
notebook-5.7.4       | 7.4 MB    | #################################### | 100%
parso-0.3.1          | 59 KB     | #################################### | 100%
fastai-1.0.41        | 127 KB    | #################################### | 100%
decorator-4.3.0      | 10 KB     | #################################### | 100%
jinja2-2.10          | 89 KB     | #################################### | 100%
pywinpty-0.5.5       | 46 KB     | #################################### | 100%
jupyter_client-5.2.4 | 63 KB     | #################################### | 100%
soupsieve-1.7.1      | 48 KB     | #################################### | 100%
prompt_toolkit-2.0.7 | 218 KB    | #################################### | 100%
colorama-0.4.1       | 15 KB     | #################################### | 100%
zeromq-4.2.5         | 9.6 MB    | #################################### | 100%
attrs-18.2.0         | 30 KB     | #################################### | 100%
pyzmq-17.1.2         | 396 KB    | #################################### | 100%
jupyter_core-4.4.0   | 44 KB     | #################################### | 100%
prometheus_client-0. | 34 KB     | #################################### | 100%
nbconvert-5.3.1      | 320 KB    | #################################### | 100%
winpty-0.4.3         | 1.1 MB    | #################################### | 100%
pickleshare-0.7.5    | 12 KB     | #################################### | 100%
libsodium-1.0.16     | 582 KB    | #################################### | 100%
mistune-0.8.4        | 53 KB     | #################################### | 100%
webencodings-0.5.1   | 12 KB     | #################################### | 100%
traitlets-4.3.2      | 130 KB    | #################################### | 100%
bleach-3.1.0         | 110 KB    | #################################### | 100%
entrypoints-0.3      | 10 KB     | #################################### | 100%
send2trash-1.5.0     | 12 KB     | #################################### | 100%
jsonschema-3.0.0a3   | 98 KB     | #################################### | 100%
pandoc-2.5           | 46.0 MB   | #################################### | 100%
regex-2018.01.10     | 319 KB    | #################################### | 100%
beautifulsoup4-4.7.1 | 140 KB    | #################################### | 100%
wcwidth-0.1.7        | 17 KB     | #################################### | 100%
pandocfilters-1.4.2  | 9 KB      | #################################### | 100%
pygments-2.3.1       | 641 KB    | #################################### | 100%
nb_conda-2.2.1       | 33 KB     | #################################### | 100%
testpath-0.4.2       | 90 KB     | #################################### | 100%
nb_conda_kernels-2.2 | 31 KB     | #################################### | 100%
jedi-0.13.2          | 228 KB    | #################################### | 100%
ipython-7.2.0        | 1.1 MB    | #################################### | 100%
nbformat-4.4.0       | 91 KB     | #################################### | 100%
pyrsistent-0.14.9    | 89 KB     | #################################### | 100%
backcall-0.1.0       | 13 KB     | #################################### | 100%
Preparing transaction: done
Verifying transaction: done
Executing transaction: / Enabling nb_conda_kernels...
Status: enabled

| Enabling notebook extension nb_conda/main...
      - Validating: ok
Enabling tree extension nb_conda/tree...
      - Validating: ok
Enabling: nb_conda
- Writing config: C:\Users\Willismar\Anaconda3\envs\py372\etc\jupyter
    - Validating...
      nb_conda 2.2.1 ok

done

Results

(py372) C:\Users\Willismar>conda list fastai
# packages in environment at C:\Users\Willismar\Anaconda3\envs\py372:
#
# Name                    Version                   Build  Channel
fastai                    1.0.41                        1    fastai


8 Likes

So to resume everything you can install using this, at this time, because the spacy 2.0.18 is not released in the default channels yet:

conda install fastai -c fastai -c pytorch -c conda-forge

The details of why and how I found the missing link you can see following the previous post.

12 Likes

Just perfect :slight_smile: Thanks a lot Willismar!

1 Like

This is the answer I was looking for!

2 Likes

Tried installing fastai using the suggested command, in my conda virtual environment (in windows 10):

(venv) (base) C:\Users\Samtha\Documents\automation_tool>conda install fastai -c fastai -c pytorch -c conda-forge
Collecting package metadata (current_repodata.json): done
Solving environment: done

# All requested packages already installed.

However when I run my python code in “venv”: from fastai import *
it errors as
ModuleNotFoundError: No module named ‘fastai’

I checked under venv\Lib\site-packages-- I cannot see fastai folder, but see fastcore-1.0.13.dist-info, fastprogress-1.0.0.dist-info
However on typing: “conda list fastai”, it shows up in another path, where I has previously installed it as shown below:

base) (venv) C:\Users\Samtha\Documents\automation_tool>conda list fastai
# packages in environment at C:\Users\Samtha\Anaconda3:
#
# Name                    Version                   Build  Channel
fastai                    1.0.61                        1    fastai

Please suggest how can make fastai work in my virtual environment

Hi …

I am sorry for the time, I just saw your message now…

Well this is a message not because you did something wrong but the Venv wasn’t activated when you installed, then after you activated and tried to use it the module wasn’t there.

First make sure that when you restart your machine, you activate the environment before install any package.

Are you still having this problem or is it already solved?

In case yes ,… describe the entire process you did so I can verify too.

Ok I had some time to see this…

The problem almost like I told you before. conda and venv has it’s own private environments.

Conda (base) environment

(base) C:\Users\willismar>python --version
Python 3.8.3

(base) C:\Users\willismar>where python
C:\Users\willismar\miniconda3\python.exe

(base) C:\Users\willismar>pip --version
pip 20.0.2 from C:\Users\willismar\miniconda3\lib\site-packages\pip (python 3.8)

executing the command pip freeze on base environment is possible to find some packages already installed.

If I create a new environment for venv called venv_myproject

Venv (venv_myproject) environment

(venv_myproject) (base) C:\Users\willismar\venv_myproject>where python
C:\Users\willismar\venv_myproject\Scripts\python.exe
C:\Users\willismar\miniconda3\python.exe

(venv_myproject) (base) C:\Users\willismar\venv_myproject>python --version
Python 3.8.3

(venv_myproject) (base) C:\Users\willismar\venv_myproject>pip --version
pip 19.2.3 from c:\users\willismar\venv_myproject\lib\site-packages\pip (python 3.8)

(venv_myproject) (base) C:\Users\willismar\venv_myproject>

executing the command pip freeze now will not show nothing installed in this environment.

executing the command conda install fastai -c fastai -c pytorch -c conda-forge will install everything in conda environment loaded.
Even after it if you execute pip freeze will not find nothing either. Only if deactivate venv environment you will find everything that conda installed.

So to solve this problem you need to go back to old fashioned pip way to deal with venv inside conda environment.
Actually conda environment is not being useful almost for nothing. Just to select the version of python you are using.

This is your solution to keep using venv inside conda. Install fastai v2 (because all previous torch and tourchvision was removed from wheel)

pip install torch torchvision
pip install fastai

I split into 2 commands because you have opportunity to change from torch for GPU to torch for CPU if needed.

1 Like

Thank you! this exactly the issue conda and venv are separate environments and conda command was not installing the package in my venv.

I used the below command to install fastai:
C:\Users\Samtha\Documents\automation_tool\venv\Scripts\pip install fastai==1.0.61 --no-cache-dir

Here I had to specifically pick the pip path of my venv environment. The simple pip command was also not installing in the correct path for some reason, it kept installing the fastai in C:\Users\Samtha\Anaconda3

1 Like