Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Tongyi DeepResearch – open-source 30B MoE Model that rivals OpenAI DeepResearch

tongyi-agent.github.io

Post reply on HN
Print view View original

111–120 of 156 posts

Re: Tongyi DeepResearch – open-source 30B MoE Model that rivals OpenAI DeepResearch

#111

Post by radarsat1 »

Imustaskforhelp wrote: post #99
zurfer wrote: post #8

It makes me wonder if we'll see an explosion of purpose trained LLMs because we hit diminishing returns on invest with pre training or if it takes a couple of months to fold these advantages back into the frontier models. Given the size of frontier models I would assume that they can incorporate many specializations and the most lasting thing here is the training environment. But there is probably already some tradeo…

> But there is probably already some tradeoff, as GPT 3.5 was awesome at chess and current models don't seem trained extensively on chess anymore. Wow, I am so curious, can you provide me the source I am so interested in a chess LLM's benchmark as someone who occasionally plays chess. I have thought about creating things like these but it would be very interesting to find the best model at chess which isn't stockfish…

Just search for "chess LLM leaderboard" there are already several. Also check https://www.reddit.com/r/llmchess/ although admittedly it doesn't get a lot of traffic.

Re: Tongyi DeepResearch – open-source 30B MoE Model that rivals OpenAI DeepResearch

#112

Post by pixelmelt »

Imustaskforhelp wrote: post #100
miki123211 wrote: post #27

Earlier quoted context omitted.

It was GPT-3 I think. As far as I remember, it's post-training that kills chess ability for some reason (GPT-3 wasn't post-trained).

This is so interesting, I am curious as to why, can you (or anyone) please provide any resources or insightful comments about it, they would really help a ton out here, thanks!

Gpt3 was trained on completion data so it likely saw lots of raw chess games layed out in whatever standard format moves are listed in, while 3.5 was post trained on instruct data (talking back and forth) which would have needed to explicitly include those chess games as conversational training data for it to retain as much as it would otherwise

Re: Tongyi DeepResearch – open-source 30B MoE Model that rivals OpenAI DeepResearch

#113

Post by larodi »

nextos wrote: post #98
criemen wrote: post #80

Earlier quoted context omitted.

> or if it takes a couple of months to fold these advantages back into the frontier models. Right now, I believe we're seeing that the big general-purpose models outperform approximately everything else. Special-purpose models (essentially: fine tunes) of smaller models make sense when you want to solve a specific task at lower cost/lower latency, and you transfer some/most of the abilities in that domain from a bigg…

The advantage of small purpose-specific models is that they might be much more robust i.e., unlikely to generate wrong sequences for your particular domain. That is at least my experience working on this topic during 2025. And, obviously, smaller models mean you may deploy them on cheaper hardware, latency is reduced, energy consumption is lower, etc. In some domains like robotics, these two advantages might be very…

I second this. Smaller models indeed may be much better positioned for fine-tuning for the very reason you point out - less noise to begin with.

Re: Tongyi DeepResearch – open-source 30B MoE Model that rivals OpenAI DeepResearch

#114

Post by almaight »

zurfer wrote: post #8

It makes me wonder if we'll see an explosion of purpose trained LLMs because we hit diminishing returns on invest with pre training or if it takes a couple of months to fold these advantages back into the frontier models. Given the size of frontier models I would assume that they can incorporate many specializations and the most lasting thing here is the training environment. But there is probably already some tradeo…

https://seed-tars.com/game-tars

Re: Tongyi DeepResearch – open-source 30B MoE Model that rivals OpenAI DeepResearch

#115

Post by almaight »

almaight wrote: post #114
zurfer wrote: post #8

It makes me wonder if we'll see an explosion of purpose trained LLMs because we hit diminishing returns on invest with pre training or if it takes a couple of months to fold these advantages back into the frontier models. Given the size of frontier models I would assume that they can incorporate many specializations and the most lasting thing here is the training environment. But there is probably already some tradeo…

https://seed-tars.com/game-tars

Video games have long served as a crucial proving ground for artificial intelligence. Like the real world, they offer rich, dynamic environments with responsive, real-time settings and complex challenges that push the boundaries of AI capabilities. The history of AI in gaming is marked by landmark achievements, from mastering classic board games to achieving superhuman performance in complex strategy titles. However, the next frontier lies beyond mastering individual, known environments.

To meet this challenge, we introduce Game-TARS: a next-generation generalist game agent designed to master complex video games and interactive digital environments using human-like perception, reasoning, and action. Unlike traditional game bots or modular AI frameworks, Game-TARS integrates all core faculties—visual perception, strategic reasoning, action grounding, and long-term memory—within a single, powerful vision-language model (VLM). This unified approach enables true end-to-end autonomous gameplay, allowing the agent to learn and succeed in any game without game-specific code, scripted behaviors, or manual rules.

With Game-TARS, this work is not about achieving the highest possible score in a single game. Instead, our focus is on building a robust foundation model for both generalist game-playing and broader computer use. We aim to create an agent that can learn to operate in any interactive digital environment it encounters, following instructions just like a human.

Re: Tongyi DeepResearch – open-source 30B MoE Model that rivals OpenAI DeepResearch

#116

Post by vogu66 »

ainch wrote: post #63
tbruckner wrote: post #34

Has anyone found these deep research tools useful? In my experience, they generate really bland reports don't go much further than summarization of what a search engine would return.

The reports are definitely bland, but I find them very helpful for discovering sources. For example, if I'm trying to ask an academic question like "has X been done before," sending something to scour the internet and find me examples to dig into is really helpful - especially since LLMs have some base knowledge which can help with finding the right search terms. It's not doing all the thinking, but those kind of bro…

I do that too, I wonder how much of it is the LLM being helpful and how much of it is the RAG algorithm somehow providing better references to the LLM than a google search can?

Re: Tongyi DeepResearch – open-source 30B MoE Model that rivals OpenAI DeepResearch

#117

Post by Lapel2742 »

aliljet wrote: post #5

Sunday morning, and I find myself wondering how the engineering tinkerer is supposed to best self-host these models? I'd love to load this up on the old 2080ti with 128gb of vram and play, even slowly. I'm curious what the current recommendation on that path looks like. Constraints are the fun part here. I know this isn't the 8x Blackwell Lamborghini, that's the point. :)

> I'd love to load this up on the old 2080ti with 128gb of vram and play, even slowly.

I think you mean ram and no vram. AFAIK this is a 30b moe model with 3b active parameters. Comparable to the Qwen3 MOE model. If you do not expect 60 tps such models should run sufficiently fast.

I run the Qwen3 MOE Model (https://huggingface.co/unsloth/Qwen3-30B-A3B-GGUF/blob/main/...) in 4-bit quantization on an 11 year old i5-6600 (32GB) and a Radeon 6600 with 8GB. According to a quick search your card is faster than that and I get ~12 tps with 16k context on Llama.cpp, which is ok for playing around.

My Radeon (ROCm) specific batch file to start this:

llama-server --ctx-size 16384 --flash-attn on --cache-type-k q8_0 --cache-type-v q8_0 --device ROCm0 -ngl -1 --model /usr/local/share/gguf/Qwen3-30B-A3B-Q4_0.gguf --cache-ram 16384 --cpu-moe --numa distribute --override-tensor "\.ffn_.*_exps\.weight=CPU" --jinja --temp 0.7 --port 8080

Re: Tongyi DeepResearch – open-source 30B MoE Model that rivals OpenAI DeepResearch

#119

Post by zurfer »

deepanwadhwa wrote: post #18
zurfer wrote: post #8

It makes me wonder if we'll see an explosion of purpose trained LLMs because we hit diminishing returns on invest with pre training or if it takes a couple of months to fold these advantages back into the frontier models. Given the size of frontier models I would assume that they can incorporate many specializations and the most lasting thing here is the training environment. But there is probably already some tradeo…

-> GPT 3.5 was awesome at chess I don't agree with this. I did try to play chess with GPT3.5 and it was horrible. Full of hallucinations.

Yeah I was not precise; it was `gpt-3.5-turbo-instruct`, other variants weren't trained on it apparently. https://dynomight.substack.com/p/chess

Re: Tongyi DeepResearch – open-source 30B MoE Model that rivals OpenAI DeepResearch

#120

Post by zurfer »

Imustaskforhelp wrote: post #99
zurfer wrote: post #8

It makes me wonder if we'll see an explosion of purpose trained LLMs because we hit diminishing returns on invest with pre training or if it takes a couple of months to fold these advantages back into the frontier models. Given the size of frontier models I would assume that they can incorporate many specializations and the most lasting thing here is the training environment. But there is probably already some tradeo…

> But there is probably already some tradeoff, as GPT 3.5 was awesome at chess and current models don't seem trained extensively on chess anymore. Wow, I am so curious, can you provide me the source I am so interested in a chess LLM's benchmark as someone who occasionally plays chess. I have thought about creating things like these but it would be very interesting to find the best model at chess which isn't stockfish…

this was the article I had in mind, when writing this: https://dynomight.substack.com/p/chess
Post reply on HN