Roman Rizzi 4f1a3effe0
REFACTOR: Migrate Vllm/TGI-served models to the OpenAI format. (#588)
Both endpoints provide OpenAI-compatible servers. The only difference is that Vllm doesn't support passing tools as a separate parameter. Even if the tool param is supported, it ultimately relies on the model's ability to handle native functions, which is not the case with the models we have today.

As a part of this change, we are dropping support for StableBeluga/Llama2 models. They don't have a chat_template, meaning the new API can translate them.

These changes let us remove some of our existing dialects and are a first step in our plan to support any LLM by defining them as data-driven concepts.

 I rewrote the "translate" method to use a template method and extracted the tool support strategies into its classes to simplify the code.

Finally, these changes bring support for Ollama when running in dev mode. It only works with Mistral for now, but it will change soon..
2024-05-07 10:02:16 -03:00

157 lines
5.2 KiB
Ruby

# frozen_string_literal: true
RSpec.describe DiscourseAi::Completions::Llm do
subject(:llm) do
described_class.new(
DiscourseAi::Completions::Dialects::Mistral,
canned_response,
"hugging_face:Upstage-Llama-2-*-instruct-v2",
gateway: canned_response,
)
end
fab!(:user)
describe ".proxy" do
it "raises an exception when we can't proxy the model" do
fake_model = "unknown:unknown_v2"
expect { described_class.proxy(fake_model) }.to(
raise_error(DiscourseAi::Completions::Llm::UNKNOWN_MODEL),
)
end
end
describe "AiApiAuditLog" do
it "is able to keep track of post and topic id" do
prompt =
DiscourseAi::Completions::Prompt.new(
"You are fake",
messages: [{ type: :user, content: "fake orders" }],
topic_id: 123,
post_id: 1,
)
result = <<~TEXT
data: {"id":"chatcmpl-8xoPOYRmiuBANTmGqdCGVk4ZA3Orz","object":"chat.completion.chunk","created":1709265814,"model":"gpt-4-0125-preview","system_fingerprint":"fp_70b2088885","choices":[{"index":0,"delta":{"role":"assistant","content":""},"logprobs":null,"finish_reason":null}]}
data: {"id":"chatcmpl-8xoPOYRmiuBANTmGqdCGVk4ZA3Orz","object":"chat.completion.chunk","created":1709265814,"model":"gpt-4-0125-preview","system_fingerprint":"fp_70b2088885","choices":[{"index":0,"delta":{"content":"Hello"},"logprobs":null,"finish_reason":null}]}
data: [DONE]
TEXT
WebMock.stub_request(:post, "https://api.openai.com/v1/chat/completions").to_return(
status: 200,
body: result,
)
result = +""
described_class
.proxy("open_ai:gpt-3.5-turbo")
.generate(prompt, user: user) { |partial| result << partial }
expect(result).to eq("Hello")
log = AiApiAuditLog.order("id desc").first
expect(log.topic_id).to eq(123)
expect(log.post_id).to eq(1)
end
end
describe "#generate with fake model" do
before do
DiscourseAi::Completions::Endpoints::Fake.delays = []
DiscourseAi::Completions::Endpoints::Fake.chunk_count = 10
end
let(:llm) { described_class.proxy("fake:fake") }
let(:prompt) do
DiscourseAi::Completions::Prompt.new(
"You are fake",
messages: [{ type: :user, content: "fake orders" }],
)
end
it "can generate a response" do
response = llm.generate(prompt, user: user)
expect(response).to be_present
end
it "can generate content via a block" do
partials = []
response = llm.generate(prompt, user: user) { |partial| partials << partial }
expect(partials.length).to eq(10)
expect(response).to eq(DiscourseAi::Completions::Endpoints::Fake.fake_content)
expect(partials.join).to eq(response)
end
end
describe "#generate with various style prompts" do
let :canned_response do
DiscourseAi::Completions::Endpoints::CannedResponse.new(["world"])
end
it "can generate a response to a simple string" do
response = llm.generate("hello", user: user)
expect(response).to eq("world")
end
it "can generate a response from an array" do
response =
llm.generate(
[{ type: :system, content: "you are a bot" }, { type: :user, content: "hello" }],
user: user,
)
expect(response).to eq("world")
end
end
describe "#generate" do
let(:prompt) do
system_insts = (<<~TEXT).strip
I want you to act as a title generator for written pieces. I will provide you with a text,
and you will generate five attention-grabbing titles. Please keep the title concise and under 20 words,
and ensure that the meaning is maintained. Replies will utilize the language type of the topic.
TEXT
DiscourseAi::Completions::Prompt
.new(system_insts)
.tap { |a_prompt| a_prompt.push(type: :user, content: (<<~TEXT).strip) }
Here is the text, inside <input></input> XML tags:
<input>
To perfect his horror, Caesar, surrounded at the base of the statue by the impatient daggers of his friends,
discovers among the faces and blades that of Marcus Brutus, his protege, perhaps his son, and he no longer
defends himself, but instead exclaims: 'You too, my son!' Shakespeare and Quevedo capture the pathetic cry.
</input>
TEXT
end
let(:canned_response) do
DiscourseAi::Completions::Endpoints::CannedResponse.new(
[
"<ai>The solitary horse.,The horse etched in gold.,A horse's infinite journey.,A horse lost in time.,A horse's last ride.</ai>",
],
)
end
context "when getting the full response" do
it "processes the prompt and return the response" do
llm_response = llm.generate(prompt, user: user)
expect(llm_response).to eq(canned_response.responses[0])
end
end
context "when getting a streamed response" do
it "processes the prompt and call the given block with the partial response" do
llm_response = +""
llm.generate(prompt, user: user) { |partial, cancel_fn| llm_response << partial }
expect(llm_response).to eq(canned_response.responses[0])
end
end
end
end