cohere-ai/notebooks

Code examples and jupyter notebooks for the Cohere Platform

notebooks's Issues

Issue with Cohere's multilingual model for creating embeddings of CSV

I trying to create embeddings of CSV file of size around 137 MB which has both numerical and text column (total of 6). using the following code
from langchain.document_loaders.csv_loader import CSVLoader loader = CSVLoader(file_path=path, encoding="utf-8") data = loader.load() embeddings = CohereEmbeddings(model="multilingual-22-12", cohere_api_key= cohere_api_key) doc_result = embeddings.embed_documents([data])

above gives the following error

`TypeError Traceback (most recent call last)
in <cell line: 1>()
----> 1 doc_result = embeddings.embed_documents([data])

15 frames
/usr/lib/python3.9/json/encoder.py in default(self, o)
177
178 """
--> 179 raise TypeError(f'Object of type {o.class.name} '
180 f'is not JSON serializable')
181

TypeError: Object of type Document is not JSON serializable`

I tried a lot to figuring out the solution but seems like it hard. Please help me in this regard, I will be very thankful

AttributeError: 'Generations' object has no attribute 'text'

We were trying to run the summarization notebook and got this error.

Can you advise?

Entity Extraction with Generative Models > Notebook comment at odds with actual results

In the "Entity Extraction with Generative Models" workbook, the results are correct. "Pixar's Luca" is not the result mentioned in the notebook comments.

"It also called the second example "Pixar's Luca" instead of "Luca". But maybe we'll let this one slide."

Actual notebook Results at line 1:

OMP: Info #276: omp_set_nested routine deprecated, please use omp_set_max_active_levels instead.

When I run the semantic search notebook

http://localhost:8888/notebooks/Basic_Semantic_Search.ipynb

then I receive the following Info

OMP: Info #276: omp_set_nested routine deprecated, please use omp_set_max_active_levels instead.

Does this depend on my Python3 installation or does this depend on the Cohere code?

how to implement multi-step tool use "without" Langchain?

trying multi-step tool use notebook, it seems multi-step tool use depends on LangChain's Agent ability.

want to know how to implement with Cohere's Python SDK or API, not with Langchain.

Tokenizer tutorial

It would be great to have a jupyter notebook on tokenizer, converting tokens into embeddings, among some other common usecases like multi-class classification (multiple classifications per one input sequence like assigning if a word is verb, etc...).

https://docs.cohere.ai/bpe-tokens-wiki
https://docs.cohere.ai/tokenize-reference

Thanks

Minimal instructions to run the notebooks

I think it would be good to have some minimal instructions in the README to explain how to run the notebooks, e.g.

1. Install Jupyter (https://jupyter.org/install)
2. Clone this repository (git clone https://github.com/cohere-ai/notebooks.git)
3. Change to the directory "notebooks", where the notebooks files are located (ipynb)
4. Run jupyter (jupyter notebook) in the "notebooks" directory
5. Select in your browser the notebook you want to run

API

i sign in to https://app.cohere.io/
free trail and I tried to use the API given in the dashboard but it doesn't work in any of the notebook
it give me the CohereError: invalid api token
so how I suppose to get an API in order to test or apply the code