Podcasts about Doms

Rīgas Doms šovasar kļūs par vienu no nozīmīgākajiem Eiropas Ērģeļmūzikas centriem, jo jau 9. jūlijā aizsāksies 38. Starptautiskais Ērģeļmūzikas festivāls "Rīgas Doms 2025". Par veselu mēnesi ilgstošo festivālu, jaundarbiem, izcilu mākslinieku piesaisti – saruna ar Rīgas Doma mūzikas direktoru Aigaru Reini un vienu no jaundarbu komponistēm Madaru Pētersoni.

What if everything you've believed about soreness is wrong? How surprised would you be if the best strength training left you feeling energized, not exhausted? In this episode, I shed light on muscle soreness. Most people think a workout is only good if you're sore afterwards. Not true! You'll learn why soreness isn't a sign of success but actually a big problem, especially if you have an autoimmune condition, and which exercises contribute more to soreness.Muscle soreness is one of three things. First, soreness could be from doing a new exercise, not being used to specific movements, or doing too much. Two, too much time under tension in the eccentric loading phase of exercises. You're either doing too many reps or too many eccentric dominant moves in general. Or third, improper nutrition around workouts. When looking at impressive workouts on social media, for example, it's important to remember that the impressive workouts aren't what those people did to get there. They worked toward that goal. Some workouts make you more sore than others. There is a significant link between muscle soreness, systemic inflammation, and the power of exercise selection. Less is more! Your workout doesn't need to make you incredibly sore to have been a great one.DOMS, or delayed onset muscle soreness, is actually related to inflammation. Muscle soreness is “damage,” but remember, not all inflammation and damage is bad. It's when there is too much inflammation that it becomes a problem. Women who struggle with autoimmune issues want to avoid workouts that cause excessive soreness, because they're already susceptible to being sore, and they're already managing the inflammation related to their condition.Taking all of this into consideration, wanting a program for myself not to over-fatigue my body, I created Strength Without Stress. It's a collection of exercises that are concentric-based, where maximal load is when the muscle is in a shortened position, allowing you to lift more without soreness. Working smarter, and not harder, is the key to striking the right balance in your workouts. Emphasis is determined by the time spent under tension, and if the targeted muscle is in a lengthened, stretched, or shortened position. Once I started masterminding my workouts so that I didn't do too many eccentric exercises, where the muscle is in a lengthened position at maximal load and force, everything changed. A concentric contraction is when a muscle shortens as it produces force. It shortens to overcome the weight load to cause movement. A great example of this is a bicep curl. In general, concentric contractions don't cause soreness. An eccentric contraction is the opposite. It's when the muscle is lengthening as it produces force, and these are the special exercises that cause soreness. A great example here is a squat where you're lowering down to the ground and you're resisting the load that you're holding, and then you're returning to a standing position.So the exercises that are eccentric, dominant, and therefore more taxing to the muscle, and therefore tend to cause soreness, are barbell back squats, lunges of any kind, Bulgarian split squats, Romanian deadlift, good mornings, bench press or a bicep chest press, chest fly, overhead tricep extension, supine hamstring curl, and push ups. My Strength Without Stress program combines all of my favorite concentric exercises with just the bare minimum eccentric exercises. Remember, soreness isn't the goal. Finding the perfect collection of exercises where you can actually increase your weight loads consistently over time, so that you experience intense moments during a set is. Strength training should make you feel great the next day, not knock you down!If you want access to my 4-week Strength Without Stress program for...

Struggling with downhill running? Want to avoid trashed quads and gain free speed on the trails? This episode teaches you exactly how to master downhill running with proven techniques, strength exercises, and mental strategies to help you descend faster and safer, whether you're training for an ultramarathon, trail marathon, or mountain race.

What does it mean to actually be dominant or submissive in sex? Because you told us heaps of people are abusing their preference to be “dominant”, and you're ending up physically, sexually and emotionally hurt. We explore dom/sub dynamics, hearing your stories from how it can be the best sex of your life when done right, but traumatic when done wrong. SHOW NOTES:1800 Respect - 1800 737 732 LifeLine - 13 11 14More info on powerplay and sex: https://www.abc.net.au/triplej/programs/the-sunday-hook-up/power-dynamics-in-sex-top-bottom-dominant-submissive/11996290DM us your thoughts, questions, topics, or to just vent at @triplejthehookup on IG or email us: thehookup@abc.net.auThe Hook Up is an ABC podcast, produced by triple j. It is recorded on the lands of the Wurundjeri people of the Kulin nation. We pay our respects to elders past and present. We acknowledge Aboriginal and Torres Strait Islander peoples as the First Australians and Traditional Custodians of the land where we live, work, and learn.

Hinter dem Bau des Kölner Doms steht eine Geschichte von mittelalterlichem Einfallsreichtum, harter Arbeit und jahrzehntelanger, später jahrhundertelanger, Hingabe. Erfahrt mehr über die Methoden, die Arbeiter und die Herausforderungen, die eines der größten gotischen Bauwerke Europas geformt haben.

怪獸訓練：https://www.facebook.com/monstertraining 小巨蛋/古亭基地：https://www.facebook.com/monstertrainingacademy 竹北基地：https://www.facebook.com/MonsterTrainingHC/ 民權西基地：https://www.facebook.com/MonsterTrainingMQW 民權西基地IG：https://www.instagram.com/monster.training_enerbarbell/ Powered by Firstory Hosting

This week we are joined by return guest Scott Ketcham from the Sisters Act Podcast. He recently just won Best Overall at the Canada Open Games Workshop event with his newly revamped Bringers of Flame Adepta Sororitas! We cover his strategy and his tactics in detail.Then in part 2 for our patrons we cover his 8 games, including 2 of his losses which we go into detail about how he could have played better and actually won in spite of poor rolls. You really wont want to miss this one. Subscribe on patreon.com/aow40kVahlTriumph10 BSS 10 Doms (melta)10 Novis 3 Cass (1BC/2AC)3 Immo 3 Nundams 3x5 Seraphim

Delayed Onset Muscle Soreness (DOMS) is often seen as a badge of honour after a tough workout—but does it really mean you're building muscle? In this episode, Andy dives deep into the science of muscle damage, mechanical tension, and the myths surrounding soreness as a proxy for progress. Learn why feeling broken after training may be doing more harm than good, and what you should really focus on to drive hypertrophy.Chapters:00:16 – What is DOMS and why does it occur?01:30 – Eccentric loading and novelty: Why certain exercises make you sore03:40 – High-rep training, repeated contractions, and muscle damage05:00 – DOMS ≠ Muscle Growth06:58 – The true driver of hypertrophy: Mechanical tension09:20 – Volume, workload, and progressive overload11:47 – Muscle damage vs. muscle remodeling14:10 – Why excessive soreness is counterproductive16:24 – Lifestyle factors that affect soreness18:45 – How to actually measure training progress20:42 – Smart programming: Avoiding unnecessary sorenessLooking for a results-driven training program that avoids the soreness trap and maximises hypertrophy?

Got a question or feedback? Send me a text!You crushed your workout—but two days later, you're hobbling down the stairs like you're 90. That deep, aching muscle soreness? That's DOMS—Delayed Onset Muscle Soreness—and it's not the sign of a great workout you think it is.In this episode, we unpack:What actually causes DOMS (spoiler: it's not lactic acid)Why eccentric movements trigger more sorenessThe most common mistakes that make DOMS worseHow to prevent that can't-move-my-legs feelingSmart recovery tactics that support your body like a well-oiled pit crewI'll also explain why soreness isn't the goal—progress is—and how you can build durability without burning out. Whether you're just getting back into training or you're tired of feeling wrecked after every session, this one's for you.

This week on The Tax Factor Rehana Earle and Ele Theochari look at possible U-turns on the non-dom regime, with Rachel Reeves reportedly reconsidering inheritance tax on global assets amid City concerns. They also discuss the High Court ruling that allows VAT on private school fees, rising tax receipts including a 14% hike in IHT, and a delayed Tax Freedom Day as fiscal drag bites. Plus, we revisit Making Tax Digital, with new guidance ahead of the rollout is HMRC going to be helpful to tax payers and agents around the new tax reporting framework?See omnystudio.com/listener for privacy information.

筋トレやスクワットを頑張ったけど、「太ももが痛くてつらい。」そんなときどうすればいいのか？筋肉痛を和らげる方法と、翌日以降の正しいケア方法をわかりやすく解説します。冷やす？温める？ストレッチは？NG行動は？など、よくある質問にも回答。監修は、自身も高い美意識を持ち身体を鍛えている、美容外科専門医でSOグレイスクリニック院長・近藤 惣一郎先生です。今日からできる実践的な太ももの筋肉痛を治す方法を紹介します。太ももの筋肉痛、なぜ起こる？そもそも筋肉痛はなぜ起こるのか？そのメカニズムを知ることで、適切な治し方へとつなげていきましょう。筋肉痛は筋繊維の「小さな傷」が原因筋肉痛が起こるメカニズムは、実はまだ完全には解明されていません。現段階では、筋トレなどの運動によって筋肉は通常より強い力を発揮しようとして、筋繊維（筋肉を構成する細い繊維）の一部が微細に損傷し、目には見えない「筋肉の小さな傷」が生じます。その後、身体がこの傷を修復する過程で炎症が起こり、痛みや張り、重だるさといった筋肉痛の症状が現れる、とされています。筋肉痛には「即時」と「遅れて」の2種類がある●即時性筋肉痛（運動直後に出る）運動中〜直後に感じる筋肉の張りや痛みで、軽い違和感程度でおさまることもあれば、フルマラソンや激しい運動の直後には、かなり強い痛みや脱力感が出ることもあります。これは筋肉への急激な負荷や、長時間にわたる繰り返し動作によって、筋繊維が部分的に強く損傷していることが原因です。●遅発性筋肉痛（DOMS）運動した数日後に出る筋肉痛のことで、運動から24〜48時間後にピークを迎える痛みを指します。これは筋繊維の損傷に伴う炎症反応が原因とされています。スクワットやランジなど太ももを大きく使う運動では、特に強く出やすいのが特徴です。運動直後の筋肉痛の対処法運動を終えた直後に痛みや違和感を感じたときは、まず以下のような応急処置を行いましょう。

Heute ist im Imtim.fm-Podcast Contessa Juliette bei Jenny zu Gast. Sie ist Unternehmerin, Autorin (ihre Autobiographie heißt „Der Engel mit der Peitsche“) und seit langem auch professionelle Domina. Schon in ihren 20ern hat sie ihre Neigung zur dominanten Rolle in BDSM-Spielen entdeckt. Als sie später – Mitte der 90er Jahre – als Single auf SM-Partys unterwegs war, begann sie bald ihre Karriere als professionelle Domina. Contessa Juliette beschreibt sich selbst als streng, aber liebevoll. Sie ist konsequent und zeigt gerne, wo es lang geht: Entscheiden legt ihr im Blut. Auch wenn sie keine Kinder hat, so sieht sie doch etwas mütterliches in sich und bringt das auch in die Sessions mit ihren Klienten ein. Außerdem bietet Contessa Juliette Ausbildungen zur Domina bzw. zum Dominus an. Wichtig ist ihr dabei, dass sie ihren Schüler:innen das bekannt Konzept SSC („safe, sense, consensual“) vermittelt, das in der BDSM-Szene als sehr wichtig gilt. Auch sollen sich Dom bzw. Domina und Sub vor wie nach der Session auf Augenhöhe begegnen. Sie meint, dass es gerade für Dominas bzw. Doms wichtig ist, das „Handwerkszeug“ erst einmal zu lernen, da bei einer Session sonst einiges schief gehen kann. Nicht nur angehende Profis besuchen ihre Workshops, sondern mitunter auch Paare, die neues ausprobieren möchten. Manchmal entdecken Paare erst nach vielen Jahren des Zusammenlebens und des gemeinsamen Sex, dass sie eigentlich eine Dom bzw. Domina/Sub-Beziehung leben möchten. Auch hier hilft sie gerne mit ihrem reichen Erfahrungsschatz. Im Interview erzählt Contessa Juliette auch von ihrem privaten (Sex-)Leben, von mistunter verstörenden Erlebnissen, welche Voraussetzungen man benötigt um Dom bzw. Domina zu werden und noch vieles mehr. Sie sagt: „Wenn man nackt ist, ist die ganze Hülle weg und der Mensch ist auf sein eigentliches Wesen reduziert.“ Viel Spaß mit dieser „strengen Folge“ des Podcast und den intimen Einblicken in das Leben einer Domina! Alle Infos zum Podcast und alle weiteren Folgen findest du auf Intim.fm

Inspired by a post on FetLife and real conversations with kinksters, this episode explores why both Dominants and submissives need clear, affirming rights in BDSM. We walk through the Submissive's Bill of Rights and the Dominant's Bill of Rights—not as rules, but as tools to support autonomy, respect, and consent in power exchange.Because no matter your role, your needs, boundaries, and voice matter.Visit us at pinkkinkpodcast.comAffiliates - SireDonLeather.com (use code PINKKINK to save 10% on your order) Obedienceapp.com/pinkkink for a 20% discountVisit https://linktr.ee/pinkkinkpodcast for links to our Patreon, Pink Kink Boutique, Pink Kink Institute, social media accounts and more!

Welcome to another episode of the Assemble Performance Podcast! In this solo episode, I dive deep into delayed onset muscle soreness (DOMS), explaining what it is, how to mitigate it, and its impact on your training. Inspired by questions from my recent seminar 'Strength Training for Runners,' I break down why DOMS occurs and offer actionable strategies to manage it effectively. Whether you are a runner, strength athlete, or just someone looking to stay healthy, this episode is packed with valuable insights. Don't forget to subscribe to stay updated on future episodes!Contact Me IG: https://www.instagram.com/justinsjones/ Email: justin@assembleperformance.com Website: https://assembleperformance.com/ Youtube: https://www.youtube.com/@justinjonesfitness

Discover all of the podcasts in our network, search for specific episodes, get the Optimal Living Daily workbook, and learn more at: OLDPodcast.com. Episode 3005: Eric Leija breaks down the science behind delayed onset muscle soreness (DOMS) and shares effective strategies to speed up recovery, including foam rolling, Epsom salt baths, and active recovery techniques. By understanding how muscle repair works, listeners can minimize post-workout pain and optimize their fitness gains without unnecessary downtime. Read along with the original article(s) here: https://www.ericleija.com/how-to-help-sore-muscles/ Quotes to ponder: "Foam rolling can be thought of as 'self-massage' for those of us who don't have sport massage therapists handy." "Magnesium plays a role in almost every crucial process, including cellular energy management and muscular contraction." "Not moving can actually create more soreness (and for longer) in the long run, as lack of movement stiffens your muscles and reduces circulation." Learn more about your ad choices. Visit megaphone.fm/adchoices

Discover all of the podcasts in our network, search for specific episodes, get the Optimal Living Daily workbook, and learn more at: OLDPodcast.com. Episode 3005: Eric Leija breaks down the science behind delayed onset muscle soreness (DOMS) and shares effective strategies to speed up recovery, including foam rolling, Epsom salt baths, and active recovery techniques. By understanding how muscle repair works, listeners can minimize post-workout pain and optimize their fitness gains without unnecessary downtime. Read along with the original article(s) here: https://www.ericleija.com/how-to-help-sore-muscles/ Quotes to ponder: "Foam rolling can be thought of as 'self-massage' for those of us who don't have sport massage therapists handy." "Magnesium plays a role in almost every crucial process, including cellular energy management and muscular contraction." "Not moving can actually create more soreness (and for longer) in the long run, as lack of movement stiffens your muscles and reduces circulation." Learn more about your ad choices. Visit megaphone.fm/adchoices

A recent randomized, double-blind, crossover trial published in Nutrients investigated the effects of short-term beetroot juice (BJ) supplementation on resistance-trained males. Eleven healthy men (average age 21.3±1.9 years) consumed either beetroot juice, providing 450 mg of nitrate per day, or a nitrate-free placebo. The study found that BJ supplementation significantly enhanced high-intensity resistance performance. Specifically, participants completed more repetitions at 80% of their one-repetition maximum (1RM) in both bench press and back squat exercises when taking beetroot juice compared to the placebo. Peak movement velocity also improved across all tested intensities (60-80% 1RM) with BJ supplementation. Furthermore, beetroot juice led to higher muscle oxygenation (SmO2 ) particularly at 70-80% 1RM and this effect was even more pronounced after 72 hours of supplementation. The study also noted reduced cardiovascular strain, evidenced by lower peak heart rates during exercise with BJ. Post-exercise recovery favored beetroot juice, with participants showing faster recovery in jump performance (both countermovement and squat jumps) at 24 hours and experiencing reduced upper-limb delayed-onset muscle soreness (DOMS) at 24-48 hours. These findings suggest that short-term BJ supplementation can be a practical strategy for physically active individuals looking to optimize training performance and accelerate recovery.Disclaimers:"This information is for educational purposes only and should not be interpreted as medical advice.""The study discussed was conducted on healthy, physically active young men. Further research may be needed to confirm these findings in other populations, including women, older adults, or individuals with medical conditions.""Always consult with a qualified healthcare professional before making any changes to your diet, supplement regimen, or treatment plan, especially if you have a medical condition or are taking medications.""This channel does not provide medical advice."#BeetrootJuice #NitrateSupplementation #ResistanceTraining #ExercisePerformance #MuscleRecoverySalem A, Ammar A, Kerkeni M, Boujelbane MA, Uyar AM, Köbel LM, Selvaraj S, Zare R, Heinrich KM, Jahrami H, et al. Short-Term Beetroot Juice Supplementation Enhances Strength, Reduces Fatigue, and Promotes Recovery in Physically Active Individuals: A Randomized, Double-Blind, Crossover Trial. Nutrients. 2025; 17(10):1720. https://doi.org/10.3390/nu17101720

Lilly is a sissy crossdresser and he called in to talk all about it. Tune in to hear all the details including how and why he started being into panties and how that led to fully dressing, how he got caught dressing by his mom and how she reacted, how he told some friends in high school and how they reacted and the difference between the guys' vs. the girls reactions, how and when he realized he's a crossdresser and not trans, the relationship he currently has with a girl who treats him as her sissy, how and when she copped to being into his crossdressing and his sissy side,  their first hook up and how it quickly progressed to pegging and more, the tasks she makes him do for her, the sissy he met who he now Doms and what he enjoys about that relationship, what they do together online, how often he dresses up now that he lives alone, how he likes to dress and how passable he is, what he's looking for in his future plus a whole lot more. **To see anonymous a pics of LILLY'S STASH plus see pics my other female guests + get TONS of exclusive CROSSDRESSER and or HOSER episodes and content + gain access to my PRIVATE Discord channel where people get super XX naughty + hear all the anonymous confessions, join my Patreon Crossdresser tier it's only $8 a month for Hosers and $10 a month for crossdressers (you get also get Hoser episodes too!) and you can cancel at any time. You can sign up here: ⁠⁠https://www.patreon.com/StrictlyAnonymousPodcast⁠⁠ MY BOOK IS NOW OUT FOR PRE-ORDER!!!! Strictly Anonymous Confessions: Secret Sex Lives of Total Strangers. A bunch of short, super sexy, TRUE stories. GET YOUR COPY NOW: ⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠https://amzn.to/4i7hBCd⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠  To join SDC and get a FREE Trial! click here: ⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠https://www.sdc.com/?ref=37712⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠ or go to SDC.com and use my code 37712  Want to be on the show? Email me at ⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠strictlyanonymouspodcast@gmail.com⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠ or go to ⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠http://www.strictlyanonymouspodcast.com⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠ and click on "Be on the Show" Have something quick you want to confess while remaining anonymous? Call the CONFESSIONS hotline at 347-420-3579. You can call 24/7. All voices are changed.  Sponsors:  ⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠https://viia.co/STRICTLYANON⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠ Try VIIA and use code STRICTLYANON for great SEX and sleep ⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠https://www.dipseastories.com/strictlyanon⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠ Hear the hottest stories on Dipsea and get a 30-day FREE trial PLUS 25% off your subscription ⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠https://butterwellness.com/⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠ Use the code “STRICTLY” at checkout for 20% off your entire order ⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠https://bluechew.com⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠ Get your first month of the new Blewchew Max FREE! use code: STRICTLYANON ⁠⁠⁠⁠⁠⁠⁠⁠https://beducate.me/pd2512-anonymous⁠⁠⁠⁠⁠⁠⁠⁠⁠ ⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠ Use code anonymous to get an additional 10% off the campaign's current discount - that's 60% off ⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠https://shamelesscare.sjv.io/xLQ3Jv⁠ ⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠Get $10 off Shameless Care's female viagra cream, just click on the link and use code: Strictly   Follow me! Instagram ⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠https://www.instagram.com/strictanonymous/⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠ Twitter ⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠https://twitter.com/strictanonymous?lang=en⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠ Website: ⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠http://www.strictlyanonymouspodcast.com/⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠ Everything else ⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠https://linktr.ee/Strictlyanonymouspodcas⁠⁠⁠ Learn more about your ad choices. Visit megaphone.fm/adchoices

Today Lina answers one of life's age old questions, one that's debated frequently in a public forum much like the God question or origins of the universe — should Doms be subs before they Dom? Together we'll weigh the benefits of gaining that empathetic perspective while also balancing the hot take that nobody should do anything sexually they don't want to do. Also — a bonus discussion of Lina in a room with power advocating for sexuality creates online! Thanks to you! Become a Patreon member to gain access to all the Ask A Sub benefits including our discord server, archive of premium audio and written posts, as well as our new podcast within a podcast, OTK with Lina and Mr. Dune. Submit questions for this podcast by going to memo.fm/askasub and recording a voice memo. Subscribe to the subby substack here. See the paid post archive here. Get 20% off your order at http://www.momotaroapotheca.com with code LINADUNE Twitter | @Lina.Dune | @askasub2.0 CREDITS Created, Hosted, Produced and Edited by Lina Dune With Additional Support from Mr. Dune Artwork by Kayleigh Denner Music by Dan Molad

On 6 April 2025, the longstanding rules for UK-resident, non-UK domiciled individuals – non-doms – were abolished and replaced with a new regime for foreign income and gains (FIG) that is based on residence. The abolition of domicile as a concept is one of the biggest changes in taxation for many years. In this episode of The Tax Track, we'll get behind the headlines and consider the pros and cons of the new rules.LinksEconomy explainers: what is the non-dom regime? (valid up to 5/4/25) https://www.icaew.com/insights/tax-news/2024/sep-2024/non-dom-regime-explainedResidence and FIG Regime Manual https://www.gov.uk/hmrc-internal-manuals/residence-and-fig-regime-manual PanellistsStephen Relf, Technical Manager, Tax, ICAEWKatherine Ford, Technical Manager, Tax, ICAEWAdelle Greenwood, Technical Manager, Tax, ICAEWProducer Ed AdamsEpisode first published: 30 May 2025Podcast recorded: 19 May 2025

Die letzte Folge gehört immer unseren treuen Hörerinnen und Hörern. Diesmal gab es viele Fragen zu den Auswirkungen der neuen US-Politik, aber auch zum Koalitionsvertrag. Und auch das Thema Kirchen ist diesmal dabei

On the latest episode of After Hours Sessions D Rock welcomes back the amazing DommeKat Katarina Pierce!!! The DommeKat is back and she is hyping up her latest Surrender Wrestling event!! On May 30th Surrender Wrestling and Assphyxiation will be happening and it plans on being an amazing event!! Katarina talks to D Rock about the upcoming event and all the exciting wrestlers and Doms who will be there. Katarina and D Rock also chat about what qualifies a great Ass Smother and talk about great booty :) !! All this and more on this episode of After Hours Sessions !! Also check out more info about the upcoming Surrender Wrestling event and everything else The DommeKat has going on. You can go to Twitter/X and follow @TheDommeKat You can also follow her on her website thedominakatarina.com . Check out Surrender Wrestling on May 30th!!

Everything you need to know about muscle soreness. Hosts Amy Hudson and Dr. James Fisher talk about why soreness is NOT indicative of the quality of a workout, what causes – and what doesn't cause – it, the difference between muscle soreness and joint pain, and what you can do to reduce that soreness. If you're someone who exercises on a regular basis, you can't miss this one! Today's episode, which is a part of a series that looks at indicators of the quality of exercising, focuses on soreness. Dr. James Fisher kicks things off by explaining why soreness – or actually delayed onset muscle soreness (DOMS in short) – occurs. How a workout was and how novel it was are factors that contribute to DOMS. Dr. Fisher and Amy Hudson debunk the myths regarding the role that a high degree of lactic acid plays when it comes to exercising and muscle soreness, as well as the idea of torn muscle fibers. Calcium plays a key role in muscular contraction: it initiates the process and interaction between actin and myosin muscle fibers. Dr. Fisher touches upon the reason why some people may use compression garments that compress muscles, and why muscle inflammation is actually a good thing. You know that feeling in the muscle right after a set or a workout? “The pump”, as it's often referred to, is a product of an influx of blood to the muscle and a product of muscular contraction and energy production. Dr. Fisher and Amy discuss muscle pain – and why the pain itself isn't actually in the muscle but in the connective tissues and fascia surrounding it. Feeling sore after a workout? Going back and doing light exercise can help you alleviate it. Dr. Fisher and Amy talk about the so-called repeated bout effect and why it's a strong reason for you not skipping too many workouts… Remember: if you feel sore after a workout, that's a natural process of your body's remodeling and rebuilding. Not feeling sore? That doesn't mean that you didn't have a good workout! As Dr. Fisher puts it: “People will feel sore when exercise is a novel stimulus, but the soreness will alleviate over time as a part of the repeated bout effect.” Keep in mind the muscle soreness you feel may actually be joint soreness. Dr. Fisher and Amy share some recommendations for working out a muscle group when you feel some muscle soreness. A healthy diet (with an adequate protein intake to support the repair process) and sleep are key assets in alleviating muscle soreness. Some people overlook the importance of sleep, which is a big marker for our body's ability to rest, recover, and rebuild.​​ When feeling muscle soreness, getting on a stationary bike or going for a swim, instead of going for a run, can help with the recovery process. Making progressive overload a key "philosophy" of your workout sessions will help you decrease muscle soreness. A final, important reminder from Amy and Dr. Fisher: “Soreness is not indicative of the quality of our workout. It's not a badge of honor.”   Mentioned in This Episode: The Exercise Coach - Get 2 Free Sessions! Submit your questions at StrengthChangesEverything.com     This podcast and blog are provided to you for entertainment and informational purposes only. By accessing either, you agree that neither constitute medical advice nor should they be substituted for professional medical advice or care. Use of this podcast or blog to treat any medical condition is strictly prohibited. Consult your physician for any medical condition you may be having. In no event will any podcast or blog hosts, guests, or contributors, Exercise Coach USA, LLC, Gymbot LLC, any subsidiaries or affiliates of same, or any of their respective directors, officers, employees, or agents, be responsible for any injury, loss, or damage to you or others due to any podcast or blog content.

In dieser Folge werfen wir einen Blick auf das Leben und Vermächtnis von Albertus Magnus, dem berühmten mittelalterlichen Gelehrten, der die intellektuelle Geschichte Kölns prägte. Erfahre, wie er Glauben mit Vernunft verband, zur Naturwissenschaft beitrug und Teil der Legende um den Bau des Kölner Doms wurde. War Albertus wirklich hinter dem bekanntesten Bauwerk der Stadt?

#084 Is it still dominance if it's gentle?

In this new monthly series Lina is answering quick text-based q's from paid members and bringing them to the public feed! This month we're talking about vetting questions to identify inexperienced Doms and subs, what to do about Dom4Dom chaos, and how making prints out in the sun is a metaphor for growth (isn't everything around here??) Become a Patreon member to gain access to all the Ask A Sub benefits including our discord server, archive of premium audio and written posts, as well as our new podcast within a podcast, OTK with Lina and Mr. Dune. Submit questions for this podcast by going to memo.fm/askasub and recording a voice memo. Subscribe to the subby substack here. See the paid post archive here. Get 20% off your order at http://www.momotaroapotheca.com with code LINADUNE Twitter | @Lina.Dune | @askasub2.0 CREDITS Created, Hosted, Produced and Edited by Lina Dune With Additional Support from Mr. Dune Artwork by Kayleigh Denner Music by Dan Molad

We get geeky and nerdy with our boy River Cruz

Watch this episode on YouTube: https://youtu.be/_bQujuepbksGrab the book here: Enterthedom.comAndy's Coaching deal - $2000 off the usual price + an extra 4 WEEKS of coaching! Just mention this deal when you apply here: https://kyil-extra.com/coaching(LIMITED TIME ONLY)

Understanding Muscle Soreness in Older Adults In this episode, Dr. Jeffrey Guild, physical therapist and specialist in vestibular and geriatric care, dives into the important topic of delayed onset muscle soreness (DOMS) in the older population.If you're an older adult starting strength training, or if you've experienced muscle soreness and want to know what's normal, what's not, and how to build strength safely — this episode is for you!We break down:✅ What delayed onset muscle soreness really is (and what it isn't)✅ Why a certain level of soreness is normal and even essential for building strength✅ The critical difference between appropriate soreness and dangerous overtraining✅ How eccentric (lowering) movements in strength training promote muscle gains✅ Special considerations for strength training over the age of 60, 70, 80, and even 90+Learn how to optimize your strength training while staying safe, mobile, and independent as you age.

In dieser Folge tauchen wir ein in die spannende Geschichte von Meister Gerhard, dem ersten Baumeister des Kölner Doms im 13. Jahrhundert. Erfahre, wie Gerhards Ehrgeiz ihn den Legenden nach zu einem riskanten Pakt mit dem Teufel brachte, wie er mit einem cleveren Trick seine Seele rettete und welche Geheimnisse sein Leben und Werk bis heute umgeben.

Activist and writer Michelle Tea joined me this month to vulnerably discuss the places magic intersects with lineage, life, and sex. We transitioned into a bit of a slumber party tee-hee-ing about the archetypes of the tarot and how many of the famous figures of the major arcana could be Doms or subs. It's fun! Get in here! To get the full version of this episode Join Patreon at $11/month or substack at $8/month. Subscribe to the subby substack here. See the paid post archive here. Submit questions for this podcast as voice memos to podcast@askasub.com Go here for information on how to record a voice memo Get 20% off your order at http://www.momotaroapotheca.com with code LINADUNE Twitter | @Lina.Dune | @askasub2.0 CREDITS Created, Hosted, Produced and Edited by Lina Dune With Additional Support from Mr. Dune Artwork by Kayleigh Denner Music by Dan Molad

“Virtual power plant” might be trending—but Eversource's Tilak Subrahmanian says it misses the mark.In this RE+ Northeast PowerUp Live stage interview, Nico sits down with Tilak for a refreshingly candid utility perspective on how distributed energy resources (DERs) are being integrated, managed, and monetized at scale. Spoiler: it's not just about demand response anymore.From managing EV charging loads to influencing ISO capacity calculations, Tilak breaks down how Eversource is designing a distributed grid built for policy, performance, and people. If you've ever wondered how utilities actually think about DERs, this is the episode for you.What you'll learn:

Zu groß zum Dienen, zu klein zum Widerstehen Liebes Tagebuch! Nachdem wir uns dazu entschlossen haben, für euch da draußen unsere eigene verschrobene Art von Crusader-Kings-Koop zu spielen und aufzunehmen, ruht das Szepter Neapels und des (hoffentlich bald) vereinigten Südens in Folge 4 wieder in Doms herrschaftlicher Hand. Während der der jugendliche Herzog in Süditalien sein Reich stärkt und erweitert, schmiedet das Schicksal auf der anderen Seite des Meeres, in Griechenland, seine eigenen Pläne. Bürgerkriege, Glaubenszwist und ein überaus ehrgeiziger Kaiser werfen ihre langen Schatten über Neapel. Dom & Jochen P.S. Bock auf Staffel 2? Hier Abonnent werden und direkt weiterhören: https://www.gamespodcast.de/abo

This week I'm joined by long time Smoke Buddy Princess Dandy comes back to the Smoker's Lounge and she discuss her teaching and monitoring couples and individuals in the BDSM world. We unpack that and the conversation on her lifestyle Doms treat Pro Doms. We discuss the mental aspects of that life . She talks about her doing more parties and creating a community and her drifting away from shooting boy/girl hardcore scenes. This leads of to Good Sex on camera VS Good Sex off camera. We discuss her doing more fetish content and how do someone figure out " What Type Of Dominate Are You" and more Want More Content? 2 ways to get it 1. Subscribe my Savage Smoke Sessions on Spotify ( $4.99 a month) ⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠https://podcasters.spotify.com/pod/show/smokethisova/subscribe⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠ 2. Become A Premium Smoker Subscribe to the Premium Smoke Room On Loyalfans ⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠https://www.loyalfans.com/PremiumSmokeRoom⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠ Want More Content. Become a Premium Smoker ⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠for 5 Premium Podcasts , Special Events and More $25.99 a month Sponsored By Hottest Adult Mag Online ⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠https://eroticismmagazine.com/⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠ ⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠Hottest Adult Film Company ⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠blusherotica.com/videos⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠ Best Scented Candles On The Net ⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠thekinkycandleco.etsy.com⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠ Queen Of Law ⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠https://www.instagram.com/dividenqueen/⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠ Porn/ Music/ Social Media ⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠https://allmylinks.com/pornrapstar⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠ Get The Merch: ⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠https://www.bonfire.com/store/s-t-o-merch-store/⁠ ⁠⁠⁠⁠Guest: Princess Dandy https://x.com/princessdandyx dandysplayroom.com

Chief and Moineau are back, answering more of your burning BDSM and relationship questions!Ever wondered what to do if impact play isn't your thing? Or whether all Doms need a title? They break down alternatives for brats and subs who aren't into spanking, and debate the role of labels in the Dom/sub dynamic. Plus, what's the best way to meet a unicorn?Moineau shares an update on mastering the cowgirl position and Chief reveals his latest quest for the ultimate male sex toy - this time, an octopus-like vibrating stroker.Support the show--------------Enjoyed the episode? Take your sex life to new heights and turn your fantasies into reality with my book and courses. Your dream dynamic is just a step away! The BDSM Bundle: Get my two best selling digital courses, plus my ebook as a bonus (https://courses.kinkyevents.co.uk/bundles/bdsm-bundle).Sensational Scenes. Learn how to heighten arousal, induce subspace, and have mind-blowing sexual experiences. Pick up a copy of my ebook today (https://kinkyevents.co.uk/book).The Art of Submission. Become a confident submissive and make your kinky fantasies a reality with our online video course for subs (https://kinkyevents.co.uk/aos).Sinful Shibari: Learn rope bondage for couples with our online video course (https://kinkyevents.co.uk/sinful-shibari).Freebies: Free ebook, sex menu template, and email content ...

Andreas Treichl, Chairman der Erste Foundation undVorstandsmitglied im Verein „Unser Stephansdom“ über seine persönliche erste Erinnerung an den Stephansdom und die Wichtigkeit des Doms für Wien und Österreich.

Der Dompfarrer von St. Stephan ist Vorstandsmitglied des Vereins Unser Stephansdom. Er spricht über die Wichtigkeit von Spenden zur Erhaltung des Doms.

En este episodio nos metemos al mundo de la dominación con una guía básica para empezar con el pie derecho. Hablamos con Derlina, una dominatrix profesional, sobre cómo dominar con responsabilidad, placer y mucha conciencia. Exploramos:

If you feel as stiff as a board and it sucks to walk down the stairs, we need to talk. In this episode we're tackling the topic of delayed onset muscle soreness. How to prevent it and what to do if you're experiencing it. HINT: managing and preventing DOMS may require tweaks to your training and nutrition. Tune in! RPE Training: https://youtu.be/yZ1YdfyRX8o CAITLIN LINKS: 1. Join Caitlin's Newsletter LAUREN LINKS: 1. Get on Lauren's newsletter for a chance at FREE COACHING! 2. Sign up for the Fast Track Movement Course Additional Links: Submit your episode ideas to us: https://docs.google.com/forms/d/e/1FAIpQLSdcmb5M_M4ZyaDqT26cYTKVNNVhyvSW-y-7nCsCps4wAgCAUw/viewform Send Lauren an email and get a sticker for submitting a podcast review! https://www.goodsprayclimbing.com/contact Support us on Patreon: https://www.patreon.com/averageclimberpodcast Learn more about Plug Tone Audio: www.plugtoneaudio.com

In this episode, we are continuing last week's Q&A!! Have you ever wondered how long you're supposed to be sore after a workout—or better yet, how to be less sore? We're tackling DOMS (delayed onset muscle soreness): how long it should last, what's actually happening in your muscles, and when it might be a sign of something more serious. Then we dive into insulin resistance—what it is, how it develops, and how to catch it early before it turns into something bigger. You'll learn how to tell if your workouts are helping or hurting, and which simple habits can make a real difference in blood sugar regulation. If you've ever been sore for days or confused about prediabetes, this one's for you.To hear the rest of this episode, sign up for the Foodie Fam!For weekly notes on our episodes, sign up for the Food We Need to Talk newsletter!Check out our book!Chat with us on IG!Be friends with Juna on IG and Tiktok! Learn about your ad choices: dovetail.prx.org/ad-choices

This is thee pod that talks what they love and TFP loves them some WWE man soap!!! In today's episode the guys talk Smackdown, RAW and NXT, plus we get into a round table of topics!!! Has Lex Luger been forgiven by the WWE Universe? Is Rey Fenix the mystery luchador? Does Cody need help at Wrestlemania? All this and so much more!!! CHEERS!!!JERKING THE CURTAINROUND TABLE OF TOPICSNEWSA new HOF announcement, Lex Luger……has he been forgiven?WWE has removed freaken from Seth Rollins name…….do we care???Travis Scott catching heat for slap on Cody…..will WWE spin this story and is it one we would want???Stone Cold will be in Vegas for WMSmackdown Drew takes his anger out on Priest Randy kicks off the show Give us more Chelsea and Tiffany Aldis doing trigonometry with the tag teams…..we're just as confused Yesssss boy

Is pain really a necessary part of progress in fitness? Kevin dives into the origins of the "no pain, no gain" mantra, explores the difference between DOMS and chronic injuries, and discusses why normalizing pain in training can be a dangerous mindset. He also answers a listener's question about staying motivated in the gym as we age. 

We're back with a special episode featuring Elden Ring expert (and gaming aficionado) Doms Roundtable! Both Ed and Dom attended an exclusive Nightreign preview event recently, and they're here to share everything they learned. Plus, we dive into the art of creating in-depth video guides for challenging games, chat about Final Fantasy 7, and discuss the most exciting upcoming releases of the year!   ▼ Swapping Joysticks ▼ ● All previous episodes available at http://swappingjoysticks.com    ▼ Dom's links ▼ ● Follow on YouTube - http://youtube.com/domsroundtable ● Follow on Twitch - https://www.twitch.tv/domsroundtable ● Follow on BlueSky - https://bsky.app/profile/domsroundtable.bsky.social   ▼ Ben's links ▼ ● Follow on Twitch - http://twitch.tv/biggusbennus ● Follow on Bluesky - https://bsky.app/profile/biggusbennus.bsky.social ● Follow on Instagram - http://instagram.com/biggusbennus ● Join the Discord - http://discord.gg/biggusbennus   ▼ Ed's links ▼ ● Follow on Twitch - http://twitch.tv/ed_nights ● Follow on Bluesky - https://bsky.app/profile/ednights.bsky.social 

Joe Smith returns and Dom has his latest audition! Join our Patreon - https://www.patreon.com/sloppyjoespodcast We have merch! https://sloppyjoespod.com YOU NEED OUR HELP? Send in your advice question to: SloppyJoesPodcast@gmail.com Find us: twitter.com/SloppyJoesPod tiktok.com/@sloppyjoespodcast https://www.instagram.com/thesloppyjoespod/?hl=en Presenters: Joe Smith: https://twitter.com/Joesmith93 Joe McGrath: https://twitter.com/RadioJoeM Dom:  @DommyBW   Ethan James: https://twitter.com/EthanJamesMedia #sloppyjoes #comedy #food #foodreview

In December 2024, Sunrise announced the newest Gundam TV series coming in 2025 titled Mobile Suit Gundam GQuuuuuux.  Created by the minds behind hit anime Neon Genesis Evangelion and FLCL, what universal secret is GQuuuuuux hiding?  Just how many u's are in GQuuuuuux?  How do you even pronounce GQuuuuuux for that matter?  Who gave the White Base a green paint wash and go-fast stripes?   Brian fears he might be the old man yelling at clouds while Isaac demands those Zaku hips must be lying.  Will there be sparkles and Gundam guitars?  What about the Doms?  We'll find out in 2025!

SUBSCRIBE TO THE BNC CHANNEL: https://bit.ly/45Pspyl   Ad Free & Bonus Episodes: https://bit.ly/3OZxwpr NEW MERCH: https://shoptmgstudios.com  The Kelly Clarkson collection chairs have seen it all, from Voldemort's Megan knees to Brooke's DOMs diagnosis and beyond. We even studied abroad in Philly with Aunt Suesue! We've had an incredible year thanks to you and we can't wait for more adventures in 2025!   Join our Facebook page: https://www.facebook.com/groups/5356639204457124/  Check out https://rocketrx.com and use code BANDC40 for 40% off your first order.  Give the gift of Good Wipes this holiday season! Grab them at Target, Walmart, Kroger, and your local grocery store.  Stay hydrated through the holidays and get 20% off your first order of Liquid I.V. at https://liquidiv.com with code BANDC at checkout. B+C IG: https://www.instagram.com/bncmap/ B+C Twitter: https://twitter.com/bncmap TMG Studios YouTube: https://www.youtube.com/tinymeatgang TMG Studios IG: https://www.instagram.com/realtmgstudios/ TMG Studios Twitter: https://twitter.com/realtmgstudios BROOKE https://www.instagram.com/brookeaverick https://twitter.com/ladyefron https://www.tiktok.com/@ladyefron  CONNOR https://www.instagram.com/fibula/ https://twitter.com/fibulaa https://www.tiktok.com/@fibulaa Hosted by Brooke Averick & Connor Wood, Created by TMG Studios, Brooke Averick & Connor Wood, and Produced by TMG Studios, Brooke Averick & Connor Wood. 0:00 Thanks you guys!!! 0:50 Intro 1:08 Sweaty pits (Ep 102) 2:15 Brookes scares Connor (Ep 104) 2:59 New AI scams (Ep 105) 4:57 Who farted? (Ep 106) 8:50 Puhh— (Ep 109) 10:38 rocketRx 11:38 FMK (Ep 110) 12:13 Underground railroad simulation (Ep 111) 13:33 Aunt Suesue! (Ep 112) 15:45 Connor leaving the Zoom (Ep 114) 1647 No? (Ep 116) 17:20 Dood (Ep 119) 18:00 But Officer… (Ep 120) 19:47 Brooke's DOMs (Ep 123) 22:47 goodwipes 24:16 Fibula's feminist takes (Ep 124) 26:36 Redownloading Hinge (Ep 124) 28:51 Kat and Connor get pulled over (Ep 125) 37:12 Brooke's bottle smash (Ep 125) 39:12 Liquid IV 41:04 Just say his name (Ep 126) 41:39 Cuck queen (Ep 126) 44:39 Brooke's DMs (Ep 128) 46:47 Connor's spider propaganda (Ep 129) 52:21 Hi my love! (Ep 130) 53:45 Chinese ribs (Ep 132) 54:53 Tristan's language (Ep 134) 57:36 Lying vs Laying in bars (Ep 137) 59:01 Brooke's DOMs pt 2 (Ep 138) 1:02:25 George Pitt (Ep 142)  1:04:18 Brooke's stress dream (Ep 143) 1:06:03 Connor's mesh shorts (Ep 143)

Happy holidays! We'll be sharing snippets from Latent Space LIVE! through the break bringing you the best of 2024! We want to express our deepest appreciation to event sponsors AWS, Daylight Computer, Thoth.ai, StrongCompute, Notable Capital, and most of all all our LS supporters who helped fund the gorgeous venue and A/V production!For NeurIPS last year we did our standard conference podcast coverage interviewing selected papers (that we have now also done for ICLR and ICML), however we felt that we could be doing more to help AI Engineers 1) get more industry-relevant content, and 2) recap 2024 year in review from experts. As a result, we organized the first Latent Space LIVE!, our first in person miniconference, at NeurIPS 2024 in Vancouver. Today, we're proud to share Loubna's highly anticipated talk (slides here)!Synthetic DataWe called out the Synthetic Data debate at last year's NeurIPS, and no surprise that 2024 was dominated by the rise of synthetic data everywhere:* Apple's Rephrasing the Web, Microsoft's Phi 2-4 and Orca/AgentInstruct, Tencent's Billion Persona dataset, DCLM, and HuggingFace's FineWeb-Edu, and Loubna's own Cosmopedia extended the ideas of synthetic textbook and agent generation to improve raw web scrape dataset quality* This year we also talked to the IDEFICS/OBELICS team at HuggingFace who released WebSight this year, the first work on code-vs-images synthetic data.* We called Llama 3.1 the Synthetic Data Model for its extensive use (and documentation!) of synthetic data in its pipeline, as well as its permissive license. * Nemotron CC and Nemotron-4-340B also made a big splash this year for how they used 20k items of human data to synthesize over 98% of the data used for SFT/PFT.* Cohere introduced Multilingual Arbitrage: Optimizing Data Pools to Accelerate Multilingual Progress observing gains of up to 56.5% improvement in win rates comparing multiple teachers vs the single best teacher model* In post training, AI2's Tülu3 (discussed by Luca in our Open Models talk) and Loubna's Smol Talk were also notable open releases this year.This comes in the face of a lot of scrutiny and criticism, with Scale AI as one of the leading voices publishing AI models collapse when trained on recursively generated data in Nature magazine bringing mainstream concerns to the potential downsides of poor quality syndata:Part of the concerns we highlighted last year on low-background tokens are coming to bear: ChatGPT contaminated data is spiking in every possible metric:But perhaps, if Sakana's AI Scientist pans out this year, we will have mostly-AI AI researchers publishing AI research anyway so do we really care as long as the ideas can be verified to be correct?Smol ModelsMeta surprised many folks this year by not just aggressively updating Llama 3 and adding multimodality, but also adding a new series of “small” 1B and 3B “on device” models this year, even working on quantized numerics collaborations with Qualcomm, Mediatek, and Arm. It is near unbelievable that a 1B model today can qualitatively match a 13B model of last year:and the minimum size to hit a given MMLU bar has come down roughly 10x in the last year. We have been tracking this proxied by Lmsys Elo and inference price:The key reads this year are:* MobileLLM: Optimizing Sub-billion Parameter Language Models for On-Device Use Cases* Apple Intelligence Foundation Language Models* Hymba: A Hybrid-head Architecture for Small Language Models* Loubna's SmolLM and SmolLM2: a family of state-of-the-art small models with 135M, 360M, and 1.7B parameters on the pareto efficiency frontier.* and Moondream, which we already covered in the 2024 in Vision talkFull Talk on YouTubeplease like and subscribe!Timestamps* [00:00:05] Loubna Intro* [00:00:33] The Rise of Synthetic Data Everywhere* [00:02:57] Model Collapse* [00:05:14] Phi, FineWeb, Cosmopedia - Synthetic Textbooks* [00:12:36] DCLM, Nemotron-CC* [00:13:28] Post Training - AI2 Tulu, Smol Talk, Cohere Multilingual Arbitrage* [00:16:17] Smol Models* [00:18:24] On Device Models* [00:22:45] Smol Vision Models* [00:25:14] What's NextTranscript2024 in Synthetic Data and Smol Models[00:00:00] ​[00:00:05] Loubna Intro[00:00:05] Speaker: ​I'm very happy to be here. Thank you for the invitation. So I'm going to be talking about synthetic data in 2024. And then I'm going to be talking about small on device models. So I think the most interesting thing about synthetic data this year is that like now we have it everywhere in the large language models pipeline.[00:00:33] The Rise of Synthetic Data Everywhere[00:00:33] Speaker: I think initially, synthetic data was mainly used just for post training, because naturally that's the part where we needed human annotators. And then after that, we realized that we don't really have good benchmarks to [00:01:00] measure if models follow instructions well, if they are creative enough, or if they are chatty enough, so we also started using LLMs as judges.[00:01:08] Speaker: Thank you. And I think this year and towards the end of last year, we also went to the pre training parts and we started generating synthetic data for pre training to kind of replace some parts of the web. And the motivation behind that is that you have a lot of control over synthetic data. You can control your prompt and basically also the kind of data that you generate.[00:01:28] Speaker: So instead of just trying to filter the web, you could try to get the LLM to generate what you think the best web pages could look like and then train your models on that. So this is how we went from not having synthetic data at all in the LLM pipeline to having it everywhere. And so the cool thing is like today you can train an LLM with like an entirely synthetic pipeline.[00:01:49] Speaker: For example, you can use our Cosmopedia datasets and you can train a 1B model on like 150 billion tokens that are 100 percent synthetic. And those are also of good quality. And then you can [00:02:00] instruction tune the model on a synthetic SFT dataset. You can also do DPO on a synthetic dataset. And then to evaluate if the model is good, you can use.[00:02:07] Speaker: A benchmark that uses LLMs as a judge, for example, MTBench or AlpacaEvil. So I think this is like a really mind blowing because like just a few years ago, we wouldn't think this is possible. And I think there's a lot of concerns about model collapse, and I'm going to talk about that later. But we'll see that like, if we use synthetic data properly and we curate it carefully, that shouldn't happen.[00:02:29] Speaker: And the reason synthetic data is very popular right now is that we have really strong models, both open and closed. It is really cheap and fast to use compared to human annotations, which cost a lot and take a lot of time. And also for open models right now, we have some really good inference frameworks.[00:02:47] Speaker: So if you have enough GPUs, it's really easy to spawn these GPUs and generate like a lot of synthetic data. Some examples are VLM, TGI, and TensorRT.[00:02:57] Model Collapse[00:02:57] Speaker: Now let's talk about the elephant in the room, model [00:03:00] collapse. Is this the end? If you look at the media and all of like, for example, some papers in nature, it's really scary because there's a lot of synthetic data out there in the web.[00:03:09] Speaker: And naturally we train on the web. So we're going to be training a lot of synthetic data. And if model collapse is going to happen, we should really try to take that seriously. And the other issue is that, as I said, we think, a lot of people think the web is polluted because there's a lot of synthetic data.[00:03:24] Speaker: And for example, when we're building fine web datasets here at Guillerm and Hinek, we're interested in like, how much synthetic data is there in the web? So there isn't really a method to properly measure the amount of synthetic data or to save a webpage synthetic or not. But one thing we can do is to try to look for like proxy words, for example, expressions like as a large language model or words like delve that we know are actually generated by chat GPT.[00:03:49] Speaker: We could try to measure the amount of these words in our data system and compare them to the previous years. For example, here, we measured like a, these words ratio in different dumps of common crawl. [00:04:00] And we can see that like the ratio really increased after chat GPT's release. So if we were to say that synthetic data amount didn't change, you would expect this ratio to stay constant, which is not the case.[00:04:11] Speaker: So there's a lot of synthetic data probably on the web, but does this really make models worse? So what we did is we trained different models on these different dumps. And we then computed their performance on popular, like, NLP benchmarks, and then we computed the aggregated score. And surprisingly, you can see that the latest DOMs are actually even better than the DOMs that are before.[00:04:31] Speaker: So if there's some synthetic data there, at least it did not make the model's worse. Yeah, which is really encouraging. So personally, I wouldn't say the web is positive with Synthetic Data. Maybe it's even making it more rich. And the issue with like model collapse is that, for example, those studies, they were done at like a small scale, and you would ask the model to complete, for example, a Wikipedia paragraph, and then you would train it on these new generations, and you would do that every day.[00:04:56] Speaker: iteratively. I think if you do that approach, it's normal to [00:05:00] observe this kind of behavior because the quality is going to be worse because the model is already small. And then if you train it just on its generations, you shouldn't expect it to become better. But what we're really doing here is that we take a model that is very large and we try to distill its knowledge into a model that is smaller.[00:05:14] Phi, FineWeb, Cosmopedia - Synthetic Textbooks[00:05:14] Speaker: And in this way, you can expect to get like a better performance for your small model. And using synthetic data for pre-training has become really popular. After the textbooks are all you need papers where Microsoft basically trained a series of small models on textbooks that were using a large LLM.[00:05:32] Speaker: And then they found that these models were actually better than models that are much larger. So this was really interesting. It was like first of its time, but it was also met with a lot of skepticism, which is a good thing in research. It pushes you to question things because the dataset that they trained on was not public, so people were not really sure if these models are really good or maybe there's just some data contamination.[00:05:55] Speaker: So it was really hard to check if you just have the weights of the models. [00:06:00] And as Hugging Face, because we like open source, we tried to reproduce what they did. So this is our Cosmopedia dataset. We basically tried to follow a similar approach to what they documented in the paper. And we created a synthetic dataset of textbooks and blog posts and stories that had almost 30 billion tokens.[00:06:16] Speaker: And we tried to train some models on that. And we found that like the key ingredient to getting a good data set that is synthetic is trying as much as possible to keep it diverse. Because if you just throw the same prompts as your model, like generate like a textbook about linear algebra, and even if you change the temperature, the textbooks are going to look alike.[00:06:35] Speaker: So there's no way you could scale to like millions of samples. And the way you do that is by creating prompts that have some seeds that make them diverse. In our case, the prompt, we would ask the model to generate a textbook, but make it related to an extract from a webpage. And also we try to frame it within, to stay within topic.[00:06:55] Speaker: For example, here, we put like an extract about cardiovascular bioimaging, [00:07:00] and then we ask the model to generate a textbook related to medicine that is also related to this webpage. And this is a really nice approach because there's so many webpages out there. So you can. Be sure that your generation is not going to be diverse when you change the seed example.[00:07:16] Speaker: One thing that's challenging with this is that you want the seed samples to be related to your topics. So we use like a search tool to try to go all of fine web datasets. And then we also do a lot of experiments with the type of generations we want the model to generate. For example, we ask it for textbooks for middle school students or textbook for college.[00:07:40] Speaker: And we found that like some generation styles help on some specific benchmarks, while others help on other benchmarks. For example, college textbooks are really good for MMLU, while middle school textbooks are good for benchmarks like OpenBookQA and Pico. This is like a sample from like our search tool.[00:07:56] Speaker: For example, you have a top category, which is a topic, and then you have some [00:08:00] subtopics, and then you have the topic hits, which are basically the web pages in fine web does belong to these topics. And here you can see the comparison between Cosmopedia. We had two versions V1 and V2 in blue and red, and you can see the comparison to fine web, and as you can see throughout the training training on Cosmopedia was consistently better.[00:08:20] Speaker: So we managed to get a data set that was actually good to train these models on. It's of course so much smaller than FineWeb, it's only 30 billion tokens, but that's the scale that Microsoft data sets was, so we kind of managed to reproduce a bit what they did. And the data set is public, so everyone can go there, check if everything is all right.[00:08:38] Speaker: And now this is a recent paper from NVIDIA, Neumatron CC. They took things a bit further, and they generated not a few billion tokens, but 1. 9 trillion tokens, which is huge. And we can see later how they did that. It's more of, like, rephrasing the web. So we can see today that there's, like, some really huge synthetic datasets out there, and they're public, so, [00:09:00] like, you can try to filter them even further if you want to get, like, more high quality corpses.[00:09:04] Speaker: So for this, rephrasing the web this approach was suggested in this paper by Pratyush, where basically in this paper, they take some samples from C4 datasets, and then they use an LLM to rewrite these samples into a better format. For example, they ask an LLM to rewrite the sample into a Wikipedia passage or into a Q& A page.[00:09:25] Speaker: And the interesting thing in this approach is that you can use a model that is Small because it doesn't, rewriting doesn't require knowledge. It's just rewriting a page into a different style. So the model doesn't need to have like knowledge that is like extensive of what is rewriting compared to just asking a model to generate a new textbook and not giving it like ground truth.[00:09:45] Speaker: So here they rewrite some samples from C4 into Q& A, into Wikipedia, and they find that doing this works better than training just on C4. And so what they did in Nemo Trans CC is a similar approach. [00:10:00] They rewrite some pages from Common Crawl for two reasons. One is to, like improve Pages that are low quality, so they rewrite them into, for example, Wikipedia page, so they look better.[00:10:11] Speaker: And another reason is to create more diverse datasets. So they have a dataset that they already heavily filtered, and then they take these pages that are already high quality, and they ask the model to rewrite them in Question and Answer format. into like open ended questions or like multi choice questions.[00:10:27] Speaker: So this way they can reuse the same page multiple times without fearing like having multiple duplicates, because it's the same information, but it's going to be written differently. So I think that's also a really interesting approach for like generating synthetic data just by rephrasing the pages that you already have.[00:10:44] Speaker: There's also this approach called Prox where they try to start from a web page and then they generate a program which finds how to write that page to make it better and less noisy. For example, here you can see that there's some leftover metadata in the web page and you don't necessarily want to keep that for training [00:11:00] your model.[00:11:00] Speaker: So So they train a model that can generate programs that can like normalize and remove lines that are extra. So I think this approach is also interesting, but it's maybe less scalable than the approaches that I presented before. So that was it for like rephrasing and generating new textbooks.[00:11:17] Speaker: Another approach that I think is really good and becoming really popular for using synthetic data for pre training is basically building a better classifiers. For filtering the web for example, here we release the data sets called fine web edu. And the way we built it is by taking Llama3 and asking it to rate the educational content of web pages from zero to five.[00:11:39] Speaker: So for example, if a page is like a really good textbook that could be useful in a school setting, it would get a really high score. And if a page is just like an advertisement or promotional material, it would get a lower score. And then after that, we take these synthetic annotations and we train a classifier on them.[00:11:57] Speaker: It's a classifier like a BERT model. [00:12:00] And then we run this classifier on all of FineWeb, which is a 15 trillion tokens dataset. And then we only keep the pages that have like a score that's higher than 3. So for example, in our case, we went from 15 trillion tokens to 3. to just 1. 5 trillion tokens. Those are really highly educational.[00:12:16] Speaker: And as you can see here, a fine web EDU outperforms all the other public web datasets by a larger margin on a couple of benchmarks here, I show the aggregated score and you can see that this approach is really effective for filtering web datasets to get like better corpuses for training your LLMs.[00:12:36] DCLM, Nemotron-CC[00:12:36] Speaker: Others also try to do this approach. There's, for example, the DCLM datasets where they also train the classifier, but not to detect educational content. Instead, they trained it on OpenHermes dataset, which is a dataset for instruction tuning. And also they explain like IAM5 subreddits, and then they also get really high quality dataset which is like very information dense and can help [00:13:00] you train some really good LLMs.[00:13:01] Speaker: And then Nemotron Common Crawl, they also did this approach, but instead of using one classifier, they used an ensemble of classifiers. So they used, for example, the DCLM classifier, and also classifiers like the ones we used in FineWebEducational, and then they combined these two. Scores into a, with an ensemble method to only retain the best high quality pages, and they get a data set that works even better than the ones we develop.[00:13:25] Speaker: So that was it for like synthetic data for pre-training.[00:13:28] Post Training - AI2 Tulu, Smol Talk, Cohere Multilingual Arbitrage[00:13:28] Speaker: Now we can go back to post training. I think there's a lot of interesting post training data sets out there. One that was released recently, the agent instructs by Microsoft where they basically try to target some specific skills. And improve the performance of models on them.[00:13:43] Speaker: For example, here, you can see code, brain teasers, open domain QA, and they managed to get a dataset that outperforms that's when fine tuning Mistral 7b on it, it outperforms the original instruct model that was released by Mistral. And as I said, to get good synthetic data, you really [00:14:00] have to have a framework to make sure that your data is diverse.[00:14:03] Speaker: So for example, for them, they always. And then they see the generations on either source code or raw text documents, and then they rewrite them to make sure they're easier to generate instructions from, and then they use that for their like instruction data generation. There's also the Tool3SFT mixture, which was released recently by Allen AI.[00:14:23] Speaker: It's also really good quality and it covers a wide range of tasks. And the way they make sure that this dataset is diverse is by using personas from the persona hub datasets. Which is basically a data set of like I think over a million personas. And for example, in the tool mixture to generate like a new code snippet, they would give like the model persona, for example, a machine learning researcher interested in neural networks, and then ask it to generate like a coding problem.[00:14:49] Speaker: This way you make sure that your data set is really diverse, and then you can further filter the data sets, for example, using the reward models. We also released a dataset called Smalltalk, [00:15:00] and we also tried to cover the wide range of tasks, and as you can see here, for example, when fine tuning Mistral 7b on the dataset, we also outperformed the original Mistral instructs on a number of benchmarks, notably on mathematics and instruction following with ifevil.[00:15:18] Speaker: Another paper that's really interesting I wanted to mention is this one called Multilingual Data Arbitrage by Cohere. And basically they want to generate a data set for post training that is multilingual. And they have a really interesting problem. It's the fact that there isn't like one model that's really good at all the languages they wanted.[00:15:36] Speaker: So what they do is that like they use not just one teacher model, but multiple teachers. And then they have a router which basically sends the prompts they have to all these models. And then they get the completions and they have a reward model that traces all these generations and only keeps the best one.[00:15:52] Speaker: And this is like arbitrage and finance. So well, I think what's interesting in this, it shows that like synthetic data, it doesn't have to come from a single model. [00:16:00] And because we have so many good models now, you could like pull these models together and get like a dataset that's really high quality and that's diverse and that's covers all your needs.[00:16:12] Speaker: I was supposed to put a meme there, but. Yeah, so that was it for like a synthetic data.[00:16:17] Smol Models[00:16:17] Speaker: Now we can go to see what's happening in the small models field in 2024. I don't know if you know, but like now we have some really good small models. For example, Lama 3. 2 1B is. It matches Lama 2. 13b from, that was released last year on the LMSYS arena, which is basically the default go to leaderboard for evaluating models using human evaluation.[00:16:39] Speaker: And as you can see here, the scores of the models are really close. So I think we've made like hugely forward in terms of small models. Of course, that's one, just one data point, but there's more. For example, if you look at this chart from the Quint 2. 5 blog post, it shows that today we have some really good models that are only like 3 billion parameters [00:17:00] and 4 billion that score really high on MMLU.[00:17:03] Speaker: Which is a really popular benchmark for evaluating models. And you can see here that the red, the blue dots have more than 65 on MMLU. And the grey ones have less. And for example, Llama33b had less. So now we have a 3b model that outperforms a 33b model that was released earlier. So I think now people are starting to realize that like, we shouldn't just scale and scale models, but we should try to make them more efficient.[00:17:33] Speaker: I don't know if you knew, but you can also chat with a 3B plus model on your iPhone. For example, here, this is an app called PocketPal, where you can go and select a model from Hugging Face. It has a large choice. For example, here we loaded the 5. 3. 5, which is 3. 8 billion parameters on this iPhone. And we can chat with this and you can see that even the latency is also acceptable.[00:17:57] Speaker: For example, here, I asked it to give me a joke about [00:18:00] NeurIPS. So let's see what it has to say.[00:18:06] Speaker: Okay, why did the neural network attend NeurIPS? Because it heard there would be a lot of layers and fun and it wanted to train its sense of humor. So not very funny, but at least it can run on device. Yeah, so I think now we have good small models, but we also have like good frameworks and tools to use these small models.[00:18:24] On Device Models[00:18:24] Speaker: So I think we're really close to having like really on edge and on device models that are really good. And I think for a while we've had this narrative. But just training larger models is better. Of course, this is supported by science scaling laws. As you can see here, for example, when we scale the model size, the loss is lower and obviously you get a better model.[00:18:46] Speaker: But and we can see this, for example, in the GPT family of models, how we went from just a hundred million parameters to more than a trillion. parameters. And of course, we all observed the performance improvement when using the latest model. But [00:19:00] one thing that we shouldn't forget is that when we scale the model, we also scale the inference costs and time.[00:19:05] Speaker: And so the largest models were are going to cost so much more. So I think now instead of just building larger models, we should be focusing on building more efficient models. It's no longer a race for the largest models since these models are really expensive to run and they require like a really good infrastructure to do that and they cannot run on, for example, consumer hardware.[00:19:27] Speaker: And when you try to build more efficient models that match larger models, that's when you can really unlock some really interesting on device use cases. And I think a trend that we're noticing now is the trend of training smaller models longer. For example, if you compare how much, how long LLAMA was trained compared to LLAMA3, there is a huge increase in the pre training length.[00:19:50] Speaker: LLAMA was trained on 1 trillion tokens, but LLAMA3 8b was trained on 15 trillion tokens. So Meta managed to get a model that's the same size, but But it performs so much [00:20:00] better by choosing to like spend the sacrifice during training, because as we know, training is a one time cost, but inference is something that's ongoing.[00:20:08] Speaker: If we want to see what are like the small models reads in 2024, I think this mobile LLM paper by Meta is interesting. They try to study different models that are like have the less than 1 billion parameters and find which architecture makes most sense for these models. For example, they find that depth is more important than width.[00:20:29] Speaker: So it's more important to have models that have like more layers than just one. making them more wide. They also find that GQA helps, that tying the embedding helps. So I think it's a nice study overall for models that are just a few hundred million parameters. There's also the Apple intelligence tech report, which is interesting.[00:20:48] Speaker: So for Apple intelligence, they had two models, one that was like on server and another model that was on device. It had 3 billion parameters. And I think the interesting part is that they trained this model using [00:21:00] pruning. And then distillation. And for example, they have this table where they show that, like, using pruning and distillation works much better than training from scratch.[00:21:08] Speaker: And they also have some interesting insights about, like, how they specialize their models on specific tasks, like, for example, summarization and rewriting. There's also this paper by NVIDIA that was released recently. I think you've already had a talk about, like, hybrid models that was all interesting.[00:21:23] Speaker: And this model, they used, like, a hybrid architecture between state space models and transformers. And they managed to train a 1B model that's really performant without needing to train it on a lot of tokens. And regarding our work, we just recently released SmallM2, so it's a series of three models, which are the best in class in each model size.[00:21:46] Speaker: For example, our 1. 7b model outperforms Lama 1b and also Qt 2. 5. And how we managed to train this model is the following. That's where you spent a lot of time trying to curate the pre training datasets. We did a lot of [00:22:00] ablations, trying to find which datasets are good and also how to mix them. We also created some new math and code datasets that we're releasing soon.[00:22:08] Speaker: But you basically really spent a lot of time trying to find what's the best mixture that you can train these models on. And then we spent some time trying to like we also trained these models for very long. For example, small M1 was trained only on 1 trillion tokens, but this model is trained on 11 trillion tokens.[00:22:24] Speaker: And we saw that the performance kept improving. The models didn't really plateau mid training, which I think is really interesting. It shows that you can train such small models for very long and keep getting performance gains. What's interesting about SmallLM2 is that it's fully open. We also released, like the pre training code base, the fine tuning code, the datasets, and also evaluation in this repository.[00:22:45] Smol Vision Models[00:22:45] Speaker: Also there's, like, really interesting small models for text, but also for vision. For example, here you can see SmallVLM, which is a 2B model that's really efficient. It doesn't consume a lot of RAM, and it also has a good performance. There's also Moondream 0. [00:23:00] 5b, which was released recently. It's like the smallest visual language model.[00:23:04] Speaker: And as you can see, there isn't like a big trade off compared to Moondream 2b. So now I showed you that we have some really good small models. We also have the tools to use them, but why should you consider using small models and when? I think, like, small models are really interesting because of the on device feature.[00:23:23] Speaker: Because these models are small and they can run fast, you can basically run them on your laptop, but also on your mobile phone. And this means that your dataset stays locally. You don't have to send your queries to third parties. And this really enhances privacy. That was, for example, one of the big selling points for Apple Intelligence.[00:23:42] Speaker: Also, right now, we really have a lot of work to do. So many frameworks to do on device inference. For example, there's MLX, MLC, Llama, CPP, Transformers, JS. So we have a lot of options and each of them have like great features. So you have so many options for doing that. Small models are also really powerful if you choose to specialize them.[00:24:00][00:24:00] Speaker: For example, here there's a startup called Numind, which took small LM and then they fine tuned it on text extraction datasets. And they managed to get a model that's not very far from models that are much larger. So I think text extraction is like one use case where small models can be really performant and it makes sense to use them instead of just using larger models.[00:24:19] Speaker: You can also chat with these models in browser. For example, here, you can go there, you can load the model, you can even turn off your internet and just start chatting with the model locally. Speaking of text extraction, if you don't want to fine tune the models, there's a really good method of structure generation.[00:24:36] Speaker: We can basically force the models to follow a JSON schema that you defined. For example, here, we try to force the model to follow a schema for extracting key information from GitHub issues. So you can input free text, which is a complaint about a GitHub repository, something not working. And then you can run it there and the model can extract anything that is relevant for your GitHub issue creation.[00:24:58] Speaker: For example, the [00:25:00] priority, for example, here, priority is high, the type of the issue bug, and then a title and the estimation of how long this will take to fix. And you can just like do this in the browser, you can transform your text into a GitHub issue that's properly formatted.[00:25:14] What's Next[00:25:14] Speaker: So what's next for synthetic data and small models?[00:25:18] Speaker: I think that domain specific synthetic data is going to be, it's already important, it's going to be even more important. For example, generating synthetic data for math. I think this really would help improve the reasoning of a lot of models. And a lot of people are doing it, for example, Quint 2. 12 math, everyone's trying to reproduce a one.[00:25:37] Speaker: And so I think for synthetic data, trying to specialize it on some domains is going to be really important. And then for small models, I think specializing them through fine tuning, it's also going to be really important because I think a lot of companies are just trying to use these large models because they are better.[00:25:53] Speaker: But on some tasks, I think you can already get decent performance with small models. So you don't need to Pay like a [00:26:00] cost that's much larger just to make your model better at your task by a few percent. And this is not just for text. And I think it also applies for other modalities like vision and audio.[00:26:11] Speaker: And I think you should also watch out for on device frameworks and applications. For example, like the app I showed, or lama, all these frameworks are becoming really popular and I'm pretty sure that we're gonna get like more of them in 2025. And users really like that. Maybe for other, I should also say hot take.[00:26:28] Speaker: I think that like in AI, we just started like with fine tuning, for example, trying to make BERT work on some specific use cases, and really struggling to do that. And then we had some models that are much larger. So we just switched to like prompt engineering to get the models And I think we're going back to fine tuning where we realize these models are really costly.[00:26:47] Speaker: It's better to use just a small model or try to specialize it. So I think it's a little bit of a cycle and we're going to start to see like more fine tuning and less of just like a prompt engineering the models. So that was my talk. Thank you for following. And if you have [00:27:00] any questions, we can take them now. Get full access to Latent Space at www.latent.space/subscribe

Ever wondered how to drop the news of some of your less mainstream kinks in a relationship? You know… the really head-turning ones that would shock polite company? The ones that keep you up at night googling really specific gear? The ones that pop up in your head even as you're trying to write the description of your podcast?? Well this week Lina is sharing an actionable formula for introducing your new partner to what you want, and also laying down a new Ask A Sub commandment — no dragging Doms into Dominance. Now back to those kinks… To support the pod and join our thriving ask a sub community of kinky pocket friends, Join Patreon starting at $6/month! Submit questions for this podcast by going to memo.fm/askasub and recording a voice memo. Subscribe to the subby substack here. See the paid post archive here. Get 20% off your order at http://www.momotaroapotheca.com with code LINADUNE Twitter | @Lina.Dune | @askasub2.0 CREDITS Created, Hosted, Produced and Edited by Lina Dune With Additional Support from Mr. Dune Artwork by Kayleigh Denner Music by Dan Molad

We discuss NBR, AFI, IDA, Indie Spirits, Saturns + the Seattle Critics and go category by category starting with Best Picture to discuss the rest of this week's many Oscar Race Checkpoints. The Big Best Picture Conversation - 1:44 - We discuss the Best Picture category in light of how this week's NBR & AFI confirmed strong starts for Wicked, Sing Sing & several other films. On the flipside, we doubt the Gladiator and Dune sequels. Then we breakdown what the Indie Spirit Feature, First Feature, and Cassavettes categories tell us about the year in film at large. Best International Feature - 18:43 - The Girl with the Needle and The Seed of the Sacred Fig are on the rise. Plus, we further discuss whether or not All We Imagine As Light can sustain momentum towards a Picture nom. Best Director & Jon M. Chu's good week - 22:10 The Best Screenplays & the Hard Truths bit lives anew - 25:11 Lead Acting Category Questions including Mikey & Coleman hopes - 28:51 Supporting Categories leads to talk on Kieran's strength and Elle's surprise win - 35:59 Some early tech talk on Cinematography & Editing + More on the Docs with IDA - 43:19 Animated Feature: Flow's surge + reviews of That Christmas & Spellbound - 47:09 A Quick Thing on the Saturn Awards + a review of Fly Me To The Moon - 50:05 A Quick Thing on the Seattle Critics Society - 55:41 WHAT WE'RE WATCHING: The Order - 57:03 The Hustler & The Color of Money - 58:09 The Great Lillian Hall - 1:00:04 A Woman Under The Influence - 1:01:51 OUTRO - Send us your thoughts via social media. A reminder that we're now on Blue Sky. Please rate and review us. Also, please make sure to stay ready for upcoming Oscar Race Checkpoints on the upcoming Golden Globes and Critics Choice Noms + the Oscar Shortlists and then get ready for some fun guests for Oscar Bets and Nosferatu.

Dillon and Dave discuss Dave being named People's World's Sexiest Man, Dave's breakfast anecdote, Dillon having DOMs again, shirtless gym guy has a new accessory, and This Weekend in Fun.  Support us on Patreon and receive weekly episodes for as low $5 per month: www.patreon.com/circlingbackpodcast Watch all of our full episodes on YouTube: www.youtube.com/washedmedia Shop Washed Merch: www.washedmedia.shop (0:00) Fun & Easy Banter  (11:40) Dave is People's World's Sexiest Man (28:50) Dave's Anecdote (42:25) Dillon Has DOMs Again (48:10) Shirtless Gym Guy Back At It (1:02:15) This Weekend in Fun Support This Episode's Sponsors:     •    Tecovas: Get 10% off at https://tecovas.com/crclbk   •   Rocket Money: Cancel your unwanted subscriptions by going to https://rocketmoney.com/circling   •  Twillory: Use code WASHED18 for $18 off your first order of $139 or more at https://twillory.com/   •   Earlybird CBD: Get 20% OFF your first purchase with promo code CB20 at https://earlybirdcbd.com/ If you've purchased before, these codes are brand new so go load up the cart.  Learn more about your ad choices. Visit megaphone.fm/adchoices