Главная
Study mode:
on
1
Sampling and Verification
2
Training Compute vs Test Time Compute
3
Part 1 Recap: Sampling and Chain of Thought
4
Video Overview: Parallel Sampling and Filtering with Verifiers
5
How to sample multiple answers in parallel
6
Verifier Methods
7
Improving verifiers with fine-tuning or prompt optimisation
8
Output verifiers versus process verifiers
9
Majority Voting and Monte Carlo MCTS
10
Notebook Setup - Trelis.com/advanced-inference
11
Installation of vLLM with guided decoding
12
Loading Llama 3.2 1B as opposed to 3B in Part 1
13
Baseline Single-shot approach
14
Parallel sampling approach Pass@n / perfect verifier
15
Parallel sampling with a voting verifier using vLLM guided decoding
16
Prompt optimisation for verifiers
17
Parallel sampling with a scoring verifier 1-10
18
Parallel sampling with a binary true/false scoring verifier
19
Llama 3.2 1B Results
20
Literature Review Let’s Verify Step by Step, Large Language Monkeys, Are more LLM calls all you need? Tree of Thought
21
Resources
Description:
Dive into an advanced tutorial on test time compute and verifiers in large language models. Explore parallel sampling techniques, various verifier methods, and their implementation using vLLM with guided decoding. Learn how to optimize prompts for verifiers and compare different approaches like voting, scoring, and binary verifiers. Analyze results using Llama 3.2 1B model and review recent literature on verification methods in LLMs. Gain practical insights through hands-on demonstrations and access to comprehensive resources for further learning and experimentation.

Test Time Compute: Verifiers and Parallel Sampling - Part 2

Trelis Research
Add to list
#Computer Science #Artificial Intelligence #Natural Language Processing (NLP) #LLM (Large Language Model) #LLaMA (Large Language Model Meta AI) #Mathematics #Statistics & Probability #Sampling #Monte Carlo Tree Search #Machine Learning #vLLM
0:00 / 0:00
1 of 21

Sampling and Verification