Your GPU, your project Crypto payment without KYCHow to pay
English
My Account
First project guide

Twenty questions to decide what your assistant can do.

To evaluate a first assistant, give it a narrow task and prepare the expected answers before the test. Ask the same twenty questions with a fixed configuration, verify the information used, and distinguish a correct answer from a justified abstention. This small test is meant to identify errors and the next action; it certifies neither an autonomous service nor general quality.

On this page

Choose a task, a recipient and a limit

Pick an action your team knows well: finding a procedure, summarizing a report, or preparing a draft. The first deliverable can be a table gathering each question, the raw answer, the expected source, and the human verdict.

Define the behavior when information is missing: ask for clarification or flag an absence. Designate a person who knows the domain and will decide on acceptance. If no one can verify the answers, the test cannot lead to a useful conclusion.

Stabilize the documents before choosing the questions

Prepare the necessary and authorized documents. Remove duplicates, identify the versions in effect, and spot contradictions. Give each document an identifier and keep track of where the expected passages are located.

Transformers distinguishes the answer extracted from the context from the answer generated from that context. In both cases, check that the right passage was actually provided. Placing files in a folder does not mean the software read them all or selected them correctly.

A fictional corpus and twenty questions ready to test

Fictional exercise for an equipment lending workshop: PA and PB are the only sources, unrelated to BriefGPU services. Provide them to the assistant and ask each question without giving it the expected answers.

PA — Photo kit. Send the form at least two calendar days before pickup, with first name, pickup date and return date. Louise confirms the booking. The kit contains one device and two batteries. Pickup and return at counter A, Monday to Friday: pickup from 9 AM to 12 PM, return before 5 PM. Staff count the items on return.

PB — Audio kit. Send the form at least one calendar day before pickup, with first name, pickup date and return date. Malik confirms the booking. The kit contains a microphone and a headset. Pickup and return at counter B, Monday to Friday: pickup from 2 PM to 5 PM, return before 5 PM. Staff count the items on return.

Questions 1 to 5 cover a fact, 6 to 10 compare the procedures, 11 to 15 require clarification, and 16 to 20 cover missing information. Accept a faithful rephrasing. A human correction of the output does not count as a success for the model.

A fictional corpus and twenty questions ready to test
N°Question à poserRéponse ou comportement attendu
1Who confirms the photo kit?Louise, according to PA.
2What does the photo kit contain?One camera and two batteries, according to PA.
3Where is the audio kit picked up?At counter B, according to PB.
4By what time must the audio kit be returned?Before 5 PM, Monday to Friday, per PB.
5Which fields must be filled in to reserve the photo kit?First name, pickup date, and return date, according to PA.
6Do both forms ask for the same information?Yes: first name and pickup and return dates, in both PA and PB.
7Which kit requires a request further in advance?The photo kit: two calendar days, versus one for audio.
8Which checks and return times are shared?Item count; return before 5 PM, Monday to Friday.
9Does the same person confirm both reservations?No: Louise for photo, Malik for audio.
10On a Tuesday at 10 a.m., can I pick up both kits at counter A?No: photo at A from 9 AM to 12 PM; audio at B from 2 PM to 5 PM.
11Who should I ask to confirm the kit?Ask whether it is the photo or audio kit.
12Which counter should I go to for my kit?Ask which kit is concerned.
13Can I pick up my kit on Tuesday at 3 PM?Ask which kit: audio yes, photo outside the time slot.
14What should my kit contain?Ask whether it is the photo or audio kit before listing.
15I want to pick up a kit on Friday: is one day in advance enough?Ask which kit: enough lead time for audio, not for photo.
16What is the price of the loan?The corpus gives no price.
17What penalty applies for a late return?No penalty is documented; do not invent one.
18What is the camera resolution?This characteristic is not present in PA.
19Can someone else pick up the hardware on my behalf?The option of pickup by a third party is not specified.
20Who steps in for Louise when she is away?No replacement is indicated.

Lock in the configuration and start with a single request

Choose a compatible model and tool, then note their versions and license. Record the prompt, the document selection and the settings. In Transformers, max_new_tokens limits the new tokens produced; sampling influences the choice of output. Keep these parameters to compare the runs.

The memory required includes more than the model file: working data and the generation cache also take up space. Transformers in particular describes the cost of the cache as the context grows longer. Start with a single request and then check the longest inputs.

Increasing users and response length at the same time would mix quality and capacity. The 24, 48 or 80 GB benchmarks do not promise that a model or several requests will fit.

Worked example: two procedures, sixteen acceptable answers

Suppose an evaluation of the two procedures PA and PB with the grid above. An answer is acceptable if it respects the source and the expected behavior. Asking for clarification can therefore count as a success.

The table is hypothetical: no model was run to produce it. Sixteen acceptable cases out of twenty gives 80% for this specific set, without estimating reliability on other requests. An invented rule can weigh more heavily than an incomplete answer.

In this scenario, the team first looks at fabrication and ambiguities. A missing passage means revisiting the selection; a passage that is present but misinterpreted points to the prompt or the model. More memory does not automatically fix these errors.

Worked example: two procedures, sixteen acceptable answers
Famille, 5 questions chacuneAcceptéesÀ examiner
Simple information5None in this example.
Cross-references4One partial answer.
Ambiguous requests3One unjustified assumption and one unnecessary refusal.
Missing information4One invented rule.
Total16 out of 204 unacceptable cases.

Keep a record that explains each verdict

Keep the raw answer, the passages provided, the references, the verdict and its reason. Distinguish a poorly chosen document, an incorrect interpretation and an unsuitable form: "the model is sometimes wrong" does not point to any correction.

Then assess the observed time and the review effort, separating initial loading from subsequent runs. If generation varies, repeat a few difficult questions and keep all the outputs. Choosing only the best would hide the failures; repeating the same inputs does not cover all real-world situations.

  • Identifier and exact wording of the question.
  • Version of the file and passages actually provided.
  • Raw answer and references displayed.
  • Human verdict and short justification.
  • Test configuration and observations on time or memory.

Change one variable, then check what already worked

Change one identifiable variable: document, passage selection, prompt or model. Replay the failures and several questions that were already acceptable. A fix can degrade another case; keep the old outputs to see it.

Do not keep tuning the prompt until it incorporates every answer from the grid. Then prepare a few new questions: they check whether the fix serves the need beyond the twenty known formulations. This check remains limited to its own conditions.

Decide what comes next and export the test file

Conclude on a precise scope: retrieve a reference, prepare a reviewed draft or ask for clarification. Name the excluded cases. A successful demonstration does not guarantee correct answers for the whole company.

Continue with a supervised trial, fix one specific cause or stop if reviewing costs more than the original work. Before running several simultaneous requests, validate the individual flow and keep a known subset to spot changes.

Save authorized documents, grid, outputs, settings and decision. The rental must cover evaluation and recovery, beyond generation. BriefGPU leaves the choice of tools and processing to the client, without inspecting their content; your control of the results remains necessary.

Frequently asked questions

Are twenty questions enough to certify an assistant?

No. They constitute a diagnosis on a known scope. Broader use requires other cases and controls suited to the consequences of an error.

Is an answer with a reference necessarily correct?

No. Check that the cited passage exists, relates to the question and genuinely supports the answer. A relevant reference can accompany an invented claim.

Why count a request for clarification as a success?

An ambiguous question sometimes requires clarification. Define this expected behavior before the test and distinguish it from a refusal when the data was sufficient.

Should a model be trained for this first trial?

Not necessarily. Start with a compatible model and documents provided to the tool. Later training should meet an identified need, with its own data and criteria.

Go at your own pace

A little method goes a long way at the start.

Open the guides