Mario Kostelac

I got a little bit confused when reading recommendation for package management solution and project structure. Package management section recommends pipenv (thanks for that :)), but then project structure recommends...

redshift.unload_to_files doesn't escape sql query

5

**Describe the bug** As the code shows in https://github.com/awslabs/aws-data-wrangler/blob/ad06fab6e967618d930595bbebaa1ae0ab906a87/awswrangler/redshift.py#L957, the library is not escaping the query while creating a parent `UNLOAD` query. **Environment** N/A. **To Reproduce** ```py In [5]: wr.redshift.unload_to_files(...

Store.save raises if document id provided in _id field

4

`Store.save(document)` with a document serializing `_id` field raises because ElasticSearch does not allow `_id` being serialized as part of the document body. While investigating the fix, I've found interesting differences...

first take on per-test qunit splitting

3

This is an example of how we could support per-test test splitting for qunit. I would like to get feedback on the approach and whether it has some future so...

Hi guys, do you have plan to add any support for setting limits on procs monit runs? I am interested in setting RLIMIT_NOFILE with http://linux.die.net/man/2/setrlimit. Need any help with that?

I've recently spent several hours debugging a problem that was caused by an empty hash definition ```ruby required do hash :hash_name end ``` Such definition discards any hash content being...

OpenAI often sets rate limits per model, per organization so it makes sense to allow tracking token usage information per that granularity. Most consumers won't need that granularity, but for...

It's not possible to detect which tool returned a result, if return_direct=True

2

Since https://github.com/hwchase17/langchain/pull/997 the last action (with return_direct=True) is not contained in the `response["intermediate_steps"]` list. That makes it impossible (or very hard and non-ergonomic at least) to detect which tool actually...

Given that different models have very different latencies and pricings, it's benefitial to pass the information about the model that generated the response. Such information allows implementing custom callback managers...

I've found that some recent langchain upgrade broke our detection of "agent exceeded max iterations error", returned in https://github.com/hwchase17/langchain/blob/bee59b4689fe23dce1450bde1a5d96b0aa52ee61/langchain/agents/agent.py#L92-L96. The reason it broke is that we had to detect the...