How does the BERT model (in Tensorflow or Paddle-paddle frameworks) relate to nodes of the underlying neural-net that’s being trained?

The BERT model in frameworks like TensorFlow/Paddle-paddle shows various kinds of computation nodes (like subtract, accumulate, add, mult etc) in a graph like form in 12 layers.

But this graph doesn’t look anything like a neural-network, one that’s typically shown in textbooks (e.g. like this https://en.wikipedia.org/wiki/Artificial_neural_network#/media/File:Colored_neural_network.svg) where each edge has a weight that’s being trained and there is an input layer and output layer.

Instead, when I print out the BERT graph, I can’t figure out how a node in the BERT graph relates to a node in the neural-network that’s being trained.

I have been using the BERT framework models to compile them to a form where we can run the model on a PC/CPU. But I still lack this basic aspect of how BERT relates to neural-net as I don’t see which neural-network topology is being trained (as i’d expect topology/connections between/among various layers/nodes of the neural-net dictate how training of the neural net occurs).

Could someone explain what underlying neural-net is being trained by BERT? How do nodes in the BERT graph relate to neural-net nodes and weights on neural-net edges?

Combining Game Effects and Combining Magical Effects, and how does this relate to damage types?

Combining Game Effects and Combining Magical Effects, and how does this relate to damage types?

Fire Bolts and Fireballs!

I have a question about how to interpret the sections on Combining Game Effects and Combining Magical Effects. I have used fire damage and piercing damage as examples.

Combining Game Effects:

Different game features can affect a target at the same time. But when two or more game features have the same name, only the effects of one of them —the most potent one— apply while the durations of the effects overlap. For example, if a target is ignited by a fire elemental’s Fire Form trait, the ongoing fire damage doesn’t increase if the burning target is subjected to that trait again. Game features include spells, class features, feats, racial traits, monster abilities, and magic items. (DMG p. 252)

Combining Magical Effects:

The effects of different spells add together while the durations of those spells overlap. The effects of the same spell cast multiple times don’t combine, however. Instead, the most potent effect—such as the highest bonus from those castings applies while their durations overlap. For example, if two clerics cast bless on the same target, that character gains the spell’s benefit only once; he or she doesn’t get to roll two bonus dice. (PHB p. 204)

To set the scene: Two Wizards, Pyro and Scorchee, enter a tavern, and they bump ito their arch-enemies, Rangers Legolad and Sir Pokealot – a fight ensues! They roll for initiative. L5 characters.

I’ve had to split it into two scenarios because each presents a nuance as to how to interpret “Combining Game Effects” and “Combining Magical Effects”.

Scenario 1: Both Wizards end up with the same initiative. When it comes to choose their actions: both of them cast Fire Bolt at Legolad and are successful in their ranged spell attacks. Does this mean that Legolad takes: two separate sets of damage of 2d10, or only one, because it is the same spell name?

Scenario 2: Both Wizards end up with the same initiative. One casts Fire Bolt at Sir Pokealot and hits; and the other casts Fireball (to the tavern keeper’s discontent), and he fails his saving throw. Does this mean that Sir Pokealot takes: 2d10 from the Fire Bolt and 8d6 from the Fireball; or only the damage from the spell that caused more damage, because it is the same damage type?

Combining Game Effects and Combining Magical Effects – how does this relate to spells with multiple effects?

Combining Game Effects and Combining Magical Effects – how does this relate to spells with multiple effects?

Here is the information from the Dungeon Master’s Guide and the Player’s Handbook that relates to my question.

Combining Game Effects:

Different game features can affect a target at the same time. But when two or more game features have the same name, only the effects of one of them —the most potent one— apply while the durations of the effects overlap. For example, if a target is ignited by a fire elemental’s Fire Form trait, the ongoing fire damage doesn’t increase if the burning target is subjected to that trait again. Game features include spells, class features, feats, racial traits, monster abilities, and magic items. (DMG p. 252)

Combining Magical Effects:

The effects of different spells add together while the durations of those spells overlap. The effects of the same spell cast multiple times don’t combine, however. Instead, the most potent effect—such as the highest bonus—from those castings applies while their durations overlap. For example, if two clerics cast bless on the same target, that character gains the spell’s benefit only once; he or she doesn’t get to roll two bonus dice. (PHB p. 204)

How does the information above interact with spells that have multiple effects? I have used three spells as examples because each of them represents a nuance, but they could all combine to provide an answer: Blindness/Deafness (PHB p. 219), Bestow Curse (p. 218) and Contagion (p. 227).

For Blindness/Deafness (separate conditions): If this is cast twice, once to cause Blindness and once to cause Deafness, does this mean that the creature is both blind and deaf? Or, because it’s part of the same spell, only one of the effects is present – if so, which one?

For Bestow Curse (various curses): If two separate casters bestow separate curses on the same creature, do both curses remain in place? If so, how does we decide which one is active?

For Contagion (natural diseases): If you cast Contagion multiple times, to cause different natural diseases, would the creature have multiple diseases active, or only one of them?

I appreciate your thoughts on this.

How does explicitly defining proxy workers relate to MaxRequestWorkers in Apache?

I have the exact setup and question which was asked and answered here:

Apache mod_proxy_fcgi: One proxy worker per vhost?

However, I do not fully understand the answer, which suggests that each explicitly defined proxy worker is an “mpm worker”.

How do the 2 built-in forward/reverse proxy workers and the explicitly defined proxy workers relate to the event mpm worker configuration options?

Should each vhost define its own proxy worker by using a unique name?

Security element that relate to breaking protocol

I am working on evaluating an unsecure protocol. I am trying to categorize the vulnerabilites according to some commonly accepted security elements like the ones specified in the CIA triad or the Parkerian hexad.

There are many vulnerabilites that i can relate to the Parkerian hexad, for example participants can fake their id which would relate to the authenticity element of the Parkerian hexad.

Another vulnerability is that the rules of the protocol can easily be broken, for example participants that follow the protocol aims to form groups and one rule is that participant should never be in more than one group at any given time. The rule is broken if malicious participants send join-requests to different groups.

Basically, i am looking for some security element like confidentiality in the CIA triad or authenticity in the Parkerian hexad that would relate to “breaking of protocol-rules”. Are there any security element, for example in the parkerian hexad or other that relates to “breaking of protocol-rules”?

Parkerian hexad

Cia triad

How to “always relate” nodes to group – D8

In OG in D7, there was a field that was set to say this node belongs to group x.

With the Group module in D8, if a user creates a node through the normal /node/add/page link, it will not be linked to a group without later going in to relate that node to the group. I want the users to be adding content that directly belongs to the group and not outside of it. This is for the purpose of content moderation rules inside of the groups.

With the intention of making a pull request, what conventions exist for branch naming in a forked repo that relate to an issue in the upstream repo

Often when contributing to a project where I have push rights, I have the following workflow of:

  1. Make an issue on Github
  2. Branch from the develop or master branch, depending whichever is appropriate with the following naming convention: issue/<issue-number>/two-word-description-of-issue
  3. Make changes and push with:

git push --set-upstream origin issue/<issue-number>/two-word-description-of-issue

  1. Then open a pull request

I’m curious if a similar convention exists for when one fork’s a project but now makes an issue on the upstream repository.

Would one carry out the same workflow above referencing the issue-number in the upstream repo, or reference an issue in origin?

I have often seen other repos where it seems the convention is simply prefix the branch with one’s username, i.e. <username>/two-word-description but I would have thought it would be good to reference to a specific issue perhaps?