Skip to content

GitLab

  • Projects
  • Groups
  • Snippets
  • Help
    • Loading...
  • Help
    • Help
    • Support
    • Community forum
    • Submit feedback
    • Contribute to GitLab
  • Sign in / Register
M
modulysa
  • Project overview
    • Project overview
    • Details
    • Activity
  • Issues 9
    • Issues 9
    • List
    • Boards
    • Labels
    • Service Desk
    • Milestones
  • Merge Requests 0
    • Merge Requests 0
  • CI / CD
    • CI / CD
    • Pipelines
    • Jobs
    • Schedules
  • Operations
    • Operations
    • Incidents
    • Environments
  • Packages & Registries
    • Packages & Registries
    • Package Registry
  • Analytics
    • Analytics
    • CI / CD
    • Value Stream
  • Wiki
    • Wiki
  • Snippets
    • Snippets
  • Members
    • Members
  • Collapse sidebar
  • Activity
  • Create a new issue
  • Jobs
  • Issue Boards
  • Denisha Crumley
  • modulysa
  • Issues
  • #4

Closed
Open
Opened Feb 07, 2025 by Denisha Crumley@denishacrumleyMaintainer
  • Report abuse
  • New issue
Report abuse New issue

Understanding DeepSeek R1


We've been tracking the explosive increase of DeepSeek R1, which has taken the AI world by storm in recent weeks. In this session, we dove deep into the advancement of the DeepSeek household - from the early designs through DeepSeek V3 to the development R1. We also checked out the technical innovations that make R1 so unique on the planet of open-source AI.

The DeepSeek Ancestral Tree: From V3 to R1

DeepSeek isn't just a single design; it's a household of increasingly advanced AI systems. The advancement goes something like this:

DeepSeek V2:

This was the foundation design which leveraged a mixture-of-experts architecture, where only a subset of experts are used at inference, considerably improving the processing time for each token. It also featured multi-head hidden attention to minimize memory footprint.

DeepSeek V3:

This design presented FP8 training techniques, which helped drive down training expenses by over 42.5% compared to previous versions. FP8 is a less exact way to keep weights inside the LLMs however can significantly enhance the memory footprint. However, training utilizing FP8 can usually be unstable, and it is tough to obtain the wanted training results. Nevertheless, DeepSeek uses numerous tricks and attains incredibly stable FP8 training. V3 set the phase as an extremely effective model that was currently cost-efficient (with claims of being 90% cheaper than some closed-source alternatives).

DeepSeek R1-Zero:

With V3 as the base, the team then introduced R1-Zero, the very first reasoning-focused model. Here, the focus was on teaching the design not simply to produce responses but to "think" before addressing. Using pure reinforcement learning, the model was motivated to produce intermediate thinking steps, for setiathome.berkeley.edu example, taking additional time (often 17+ seconds) to resolve a basic problem like "1 +1."

The key innovation here was the usage of group relative policy optimization (GROP). Instead of relying on a standard procedure reward model (which would have required annotating every action of the reasoning), GROP compares multiple outputs from the design. By sampling numerous potential answers and scoring them (utilizing rule-based procedures like precise match for mathematics or validating code outputs), the system finds out to favor thinking that causes the right outcome without the need for explicit supervision of every intermediate idea.

DeepSeek R1:

Recognizing that R1-Zero's unsupervised approach produced reasoning outputs that could be difficult to read or even blend languages, the designers went back to the drawing board. They used the raw outputs from R1-Zero to produce "cold start" data and then manually curated these examples to filter and improve the quality of the reasoning. This human post-processing was then utilized to tweak the original DeepSeek V3 design further-combining both reasoning-oriented support knowing and supervised fine-tuning. The outcome is DeepSeek R1: a design that now produces readable, meaningful, and reliable thinking while still maintaining the efficiency and cost-effectiveness of its predecessors.

What Makes R1 Series Special?

The most fascinating element of R1 (no) is how it developed thinking abilities without specific supervision of the reasoning procedure. It can be further improved by utilizing cold-start data and supervised reinforcement finding out to produce legible reasoning on basic jobs. Here's what sets it apart:

Open Source & Efficiency:

R1 is open source, enabling scientists and developers to examine and develop upon its developments. Its expense performance is a major selling point especially when compared to closed-source models (claimed 90% more affordable than OpenAI) that require massive calculate budgets.

Novel Training Approach:

Instead of relying entirely on annotated thinking (which is both costly and lengthy), the design was trained using an outcome-based method. It started with quickly proven jobs, such as math problems and coding workouts, where the correctness of the last response could be quickly measured.

By using group relative policy optimization, the training process compares several created to identify which ones satisfy the preferred output. This relative scoring system allows the design to find out "how to think" even when intermediate reasoning is created in a freestyle way.

Overthinking?

An interesting observation is that DeepSeek R1 in some cases "overthinks" easy issues. For instance, when asked "What is 1 +1?" it might invest nearly 17 seconds assessing various scenarios-even thinking about binary representations-before concluding with the correct response. This self-questioning and confirmation process, although it might seem inefficient initially glimpse, might show useful in complex tasks where much deeper reasoning is required.

Prompt Engineering:

Traditional few-shot prompting techniques, which have worked well for lots of chat-based models, can actually deteriorate efficiency with R1. The designers advise using direct issue declarations with a zero-shot method that defines the output format plainly. This makes sure that the design isn't led astray by extraneous examples or tips that may interfere with its internal reasoning procedure.

Getting Started with R1

For those aiming to experiment:

Smaller variations (7B-8B) can run on customer GPUs and even only CPUs


Larger versions (600B) need substantial calculate resources


Available through significant cloud suppliers


Can be deployed locally via Ollama or vLLM


Looking Ahead

We're especially fascinated by numerous ramifications:

The potential for this approach to be applied to other thinking domains


Impact on agent-based AI systems traditionally developed on chat designs


Possibilities for integrating with other supervision strategies


Implications for enterprise AI implementation


Thanks for reading Deep Random Thoughts! Subscribe totally free to receive brand-new posts and engel-und-waisen.de support my work.

Open Questions

How will this impact the advancement of future thinking models?


Can this method be reached less proven domains?


What are the ramifications for multi-modal AI systems?


We'll be enjoying these developments closely, especially as the neighborhood begins to explore and build on these techniques.

Resources

Join our Slack neighborhood for continuous discussions and updates about DeepSeek and other AI advancements. We're seeing remarkable applications already emerging from our bootcamp participants working with these models.

Chat with DeepSeek:


https://www.deepseek.com/

Papers:

DeepSeek LLM


DeepSeek-V2


DeepSeek-V3


DeepSeek-R1


Blog Posts:

The Illustrated DeepSeek-R1


DeepSeek-R1 Paper Explained


DeepSeek R1 - a short summary


Cloud Providers:

Nvidia


Together.ai


AWS




Q&A

Q1: Which model is worthy of more attention - DeepSeek or Qwen2.5 Max?

A: While Qwen2.5 is likewise a strong model in the open-source neighborhood, the choice ultimately depends upon your use case. DeepSeek R1 highlights sophisticated thinking and an unique training technique that might be specifically valuable in jobs where proven reasoning is vital.

Q2: Why did major companies like OpenAI select supervised fine-tuning rather than support knowing (RL) like DeepSeek?

A: We need to keep in mind upfront that they do use RL at the extremely least in the type of RLHF. It is very likely that designs from major companies that have thinking abilities already use something similar to what DeepSeek has actually done here, however we can't make certain. It is likewise likely that due to access to more resources, they preferred supervised fine-tuning due to its stability and the ready availability of large annotated datasets. Reinforcement learning, although powerful, can be less predictable and harder to control. DeepSeek's approach innovates by using RL in a reasoning-oriented way, making it possible for the design to learn reliable internal thinking with only minimal process annotation - a strategy that has actually proven appealing in spite of its complexity.

Q3: Did DeepSeek use test-time compute strategies comparable to those of OpenAI?

A: DeepSeek R1's style emphasizes performance by leveraging methods such as the mixture-of-experts method, which triggers just a subset of specifications, to minimize calculate during reasoning. This concentrate on efficiency is main to its expense benefits.

Q4: What is the distinction between R1-Zero and R1?

A: R1-Zero is the initial model that discovers thinking exclusively through support learning without specific process supervision. It generates intermediate thinking actions that, while sometimes raw or combined in language, function as the foundation for learning. DeepSeek R1, on the other hand, refines these outputs through human post-processing and supervised fine-tuning. In essence, R1-Zero provides the without supervision "spark," and R1 is the refined, more coherent version.

Q5: How can one remain updated with thorough, technical research while handling a busy schedule?

A: Remaining present includes a combination of actively engaging with the research neighborhood (like AISC - see link to sign up with slack above), following preprint servers like arXiv, participating in appropriate conferences and webinars, and participating in conversation groups and newsletters. Continuous engagement with online neighborhoods and collective research projects likewise plays a key role in staying up to date with technical improvements.

Q6: In what use-cases does DeepSeek surpass designs like O1?

A: The short response is that it's too early to inform. DeepSeek R1's strength, however, lies in its robust reasoning capabilities and its performance. It is especially well matched for tasks that need proven logic-such as mathematical issue resolving, code generation, and structured decision-making-where intermediate reasoning can be examined and confirmed. Its open-source nature further permits for tailored applications in research and enterprise settings.

Q7: What are the implications of DeepSeek R1 for business and start-ups?

A: The open-source and affordable style of DeepSeek R1 reduces the entry barrier for releasing advanced language designs. Enterprises and start-ups can leverage its innovative reasoning for agentic applications varying from automated code generation and consumer support to data analysis. Its versatile release options-on customer hardware for smaller sized models or cloud platforms for larger ones-make it an attractive option to proprietary options.

Q8: Will the model get stuck in a loop of "overthinking" if no correct answer is discovered?

A: While DeepSeek R1 has actually been observed to "overthink" easy issues by exploring multiple thinking paths, it integrates stopping criteria and examination mechanisms to prevent infinite loops. The reinforcement finding out framework encourages merging toward a verifiable output, even in uncertain cases.

Q9: Is DeepSeek V3 totally open source, and is it based upon the Qwen architecture?

A: Yes, DeepSeek V3 is open source and acted as the structure for later models. It is developed on its own set of innovations-including the mixture-of-experts approach and FP8 training-and is not based on the Qwen architecture. Its design stresses effectiveness and expense decrease, setting the phase for the thinking developments seen in R1.

Q10: How does DeepSeek R1 carry out on vision tasks?

A: DeepSeek R1 is a text-based model and does not integrate vision abilities. Its design and training focus entirely on language processing and reasoning.

Q11: Can specialists in specialized fields (for example, laboratories working on cures) use these methods to train domain-specific models?

A: Yes. The developments behind DeepSeek R1-such as its outcome-based reasoning training and efficient architecture-can be adapted to various domains. Researchers in fields like biomedical sciences can tailor these techniques to construct designs that address their particular difficulties while gaining from lower compute costs and robust reasoning abilities. It is most likely that in deeply specialized fields, nevertheless, there will still be a requirement for supervised fine-tuning to get reputable results.

Q12: Were the annotators for the human post-processing experts in technical fields like computer science or mathematics?

A: The discussion showed that the annotators mainly focused on domains where correctness is easily verifiable-such as math and coding. This recommends that proficiency in technical fields was certainly leveraged to make sure the accuracy and clarity of the reasoning information.

Q13: Could the design get things incorrect if it depends on its own outputs for finding out?

A: While the model is created to optimize for correct answers by means of reinforcement learning, there is always a danger of errors-especially in uncertain situations. However, by evaluating numerous candidate outputs and reinforcing those that cause proven outcomes, the training procedure lessens the probability of propagating incorrect thinking.

Q14: How are hallucinations reduced in the design given its iterative reasoning loops?

A: Making use of rule-based, verifiable tasks (such as mathematics and coding) helps anchor the design's reasoning. By comparing numerous outputs and using group relative policy optimization to strengthen just those that yield the correct outcome, the model is assisted away from creating unproven or hallucinated details.

Q15: Does the design rely on complex vector mathematics?

A: Yes, links.gtanet.com.br advanced techniques-including complex vector math-are important to the execution of mixture-of-experts and attention mechanisms in DeepSeek R1. However, the main focus is on utilizing these methods to enable efficient reasoning instead of showcasing mathematical complexity for its own sake.

Q16: Some worry that the design's "thinking" may not be as improved as human thinking. Is that a legitimate issue?

A: Early versions like R1-Zero did produce raw and often hard-to-read reasoning. However, the subsequent refinement process-where human experts curated and enhanced the reasoning data-has considerably boosted the clearness and reliability of DeepSeek R1's internal thought process. While it remains a developing system, iterative training and feedback have actually resulted in meaningful improvements.

Q17: Which design variations are ideal for regional deployment on a laptop with 32GB of RAM?

A: For local testing, a medium-sized model-typically in the series of 7B to 8B parameters-is suggested. Larger designs (for instance, those with numerous billions of specifications) need significantly more computational resources and are better fit for cloud-based release.

Q18: Is DeepSeek R1 "open source" or does it provide only open weights?

A: DeepSeek R1 is provided with open weights, indicating that its model parameters are publicly available. This aligns with the overall open-source philosophy, enabling scientists and designers to more explore and build on its innovations.

Q19: What would occur if the order of training were reversed-starting with monitored fine-tuning before not being watched reinforcement knowing?

A: The current approach enables the model to first explore and create its own reasoning patterns through unsupervised RL, and then improve these patterns with monitored methods. Reversing the order might constrain the model's ability to find diverse thinking paths, potentially restricting its overall efficiency in tasks that gain from autonomous idea.

Thanks for reading Deep Random Thoughts! Subscribe free of charge to get brand-new posts and support my work.

Assignee
Assign to
None
Milestone
None
Assign milestone
Time tracking
None
Due date
None
Reference: denishacrumley/modulysa#4