sad
alkinun
AtAndDev
AI & ML interests
LLMs, Alignment, Merging, Unsloth, DPO, SFT, ORPO, SPIN..
Recent Activity
Organizations
reacted to
FlameF0X's
post with 👀😔
about 5 hours ago
Post
1061
I am very sad to say that the budget in creating of SnowflakeCore-G1 1b and 7b MoE models ran out and I can't pre-train them anymore.
-
- 2 replies
reacted to
sweatSmile's
post with ❤️🚀
2 days ago
Post
2605
Teaching a 7B Model to Be Just the Right Amount of Snark
Ever wondered if a language model could get sarcasm? I fine-tuned Mistral-7B using LoRA and 4-bit quantisation—on just ~720 hand-picked sarcastic prompt–response pairs from Reddit, Twitter, and real-life conversations.
The challenge? Keeping it sarcastic but still helpful.
LoRA rank 16 to avoid overfitting
4-bit NF4 quantization to fit on limited GPU memory
10 carefully monitored epochs so it didn’t turn into a full-time comedian
Result: a model that understands “Oh great, another meeting” exactly as you mean it.
Read the full journey, tech details, and lessons learned on my blog:
Fine-Tuning Mistral-7B for Sarcasm with LoRA and 4-Bit Quantisation
Try the model here on Hugging Face: sweatSmile/Mistral-7B-Instruct-v0.1-Sarcasm.
Ever wondered if a language model could get sarcasm? I fine-tuned Mistral-7B using LoRA and 4-bit quantisation—on just ~720 hand-picked sarcastic prompt–response pairs from Reddit, Twitter, and real-life conversations.
The challenge? Keeping it sarcastic but still helpful.
LoRA rank 16 to avoid overfitting
4-bit NF4 quantization to fit on limited GPU memory
10 carefully monitored epochs so it didn’t turn into a full-time comedian
Result: a model that understands “Oh great, another meeting” exactly as you mean it.
Read the full journey, tech details, and lessons learned on my blog:
Fine-Tuning Mistral-7B for Sarcasm with LoRA and 4-Bit Quantisation
Try the model here on Hugging Face: sweatSmile/Mistral-7B-Instruct-v0.1-Sarcasm.
Interesting! Did you have a chance to try k2, glm 4.5 or sonnet 4?
reacted to
dhruv3006's
post with 👀
2 days ago
Post
1840
GPT 5 for Computer Use agents.
Same tasks, same grounding model we just swapped GPT 4o with GPT 5 as the thinking model.
Left = 4o, right = 5.
Watch GPT 5 pull away.
Reasoning model: OpenAI GPT-5
Grounding model: Salesforce GTA1-7B
Action space: CUA Cloud Instances (macOS/Linux/Windows)
The task is: "Navigate to {random_url} and play the game until you reach a score of 5/5”....each task is set up by having claude generate a random app from a predefined list of prompts (multiple choice trivia, form filling, or color matching)"
Try it yourself here : https://github.com/trycua/cua
Docs : https://docs.trycua.com/docs/agent-sdk/supported-agents/composed-agents
Same tasks, same grounding model we just swapped GPT 4o with GPT 5 as the thinking model.
Left = 4o, right = 5.
Watch GPT 5 pull away.
Reasoning model: OpenAI GPT-5
Grounding model: Salesforce GTA1-7B
Action space: CUA Cloud Instances (macOS/Linux/Windows)
The task is: "Navigate to {random_url} and play the game until you reach a score of 5/5”....each task is set up by having claude generate a random app from a predefined list of prompts (multiple choice trivia, form filling, or color matching)"
Try it yourself here : https://github.com/trycua/cua
Docs : https://docs.trycua.com/docs/agent-sdk/supported-agents/composed-agents
-
- 1 reply
reacted to
sweatSmile's
post with ❤️
3 days ago
Post
2782
Qwen3 is the latest version of the Qwen language models. It's smarter, faster, and now understands 119 languages instead of just 29.
It can do both deep reasoning and quick answers using a single model, depending on what you need.
The models range in size from small (0.6B) to huge (235B), with smart ways to save compute.
It's trained on 36 trillion tokens and fine-tuned in four steps to boost performance.
Qwen3 performs as well as or better than many top models, including some from big companies.
It’s fully open-source under licence. Amazing!!!
https://github.com/QwenLM/Qwen3/blob/main/Qwen3_Technical_Report.pdf
It can do both deep reasoning and quick answers using a single model, depending on what you need.
The models range in size from small (0.6B) to huge (235B), with smart ways to save compute.
It's trained on 36 trillion tokens and fine-tuned in four steps to boost performance.
Qwen3 performs as well as or better than many top models, including some from big companies.
It’s fully open-source under licence. Amazing!!!
https://github.com/QwenLM/Qwen3/blob/main/Qwen3_Technical_Report.pdf
reacted to
ImranzamanML's
post with ❤️
4 days ago
Post
3472
Finaly OpenAI is open to share open-source models after GPT2-2019.
gpt-oss-120b
gpt-oss-20b
openai/gpt-oss-120b
#AI #GPT #LLM #Openai
gpt-oss-120b
gpt-oss-20b
openai/gpt-oss-120b
#AI #GPT #LLM #Openai
- 1 reply
reacted to
georgewritescode's
post with 🚀
4 days ago
Post
2737
Announcing Artificial Analysis Long Context Reasoning (AA-LCR), a new benchmark to evaluate long context performance through testing reasoning capabilities across multiple long documents (~100k tokens)
The focus of AA-LCR is to replicate real knowledge work and reasoning tasks, testing capability critical to modern AI applications spanning document analysis, codebase understanding, and complex multi-step workflows.
AA-LCR is 100 hard text-based questions that require reasoning across multiple real-world documents that represent ~100k input tokens. Questions are designed so answers cannot be directly found but must be reasoned from multiple information sources, with human testing verifying that each question requires genuine inference rather than retrieval.
Key takeaways:
➤ Today’s leading models achieve ~70% accuracy: the top three places go to OpenAI o3 (69%), xAI Grok 4 (68%) and Qwen3 235B 2507 Thinking (67%)
➤👀 We also already have gpt-oss results! 120B performs close to o4-mini (high), in-line with OpenAI claims regarding model performance. We will be following up shortly with a Intelligence Index for the models.
➤ 100 hard text-based questions spanning 7 categories of documents (Company Reports, Industry Reports, Government Consultations, Academia, Legal, Marketing Materials and Survey Reports)
➤ ~100k tokens of input per question, requiring models to support a minimum 128K context window to score on this benchmark
➤ ~3M total unique input tokens spanning ~230 documents to run the benchmark (output tokens typically vary by model)
We’re adding AA-LCR to the Artificial Analysis Intelligence Index, and taking the version number to v2.2. Artificial Analysis Intelligence Index v2.2 now includes: MMLU-Pro, GPQA Diamond, AIME 2025, IFBench, LiveCodeBench, SciCode and AA-LCR.
Link to dataset: ArtificialAnalysis/AA-LCR
The focus of AA-LCR is to replicate real knowledge work and reasoning tasks, testing capability critical to modern AI applications spanning document analysis, codebase understanding, and complex multi-step workflows.
AA-LCR is 100 hard text-based questions that require reasoning across multiple real-world documents that represent ~100k input tokens. Questions are designed so answers cannot be directly found but must be reasoned from multiple information sources, with human testing verifying that each question requires genuine inference rather than retrieval.
Key takeaways:
➤ Today’s leading models achieve ~70% accuracy: the top three places go to OpenAI o3 (69%), xAI Grok 4 (68%) and Qwen3 235B 2507 Thinking (67%)
➤👀 We also already have gpt-oss results! 120B performs close to o4-mini (high), in-line with OpenAI claims regarding model performance. We will be following up shortly with a Intelligence Index for the models.
➤ 100 hard text-based questions spanning 7 categories of documents (Company Reports, Industry Reports, Government Consultations, Academia, Legal, Marketing Materials and Survey Reports)
➤ ~100k tokens of input per question, requiring models to support a minimum 128K context window to score on this benchmark
➤ ~3M total unique input tokens spanning ~230 documents to run the benchmark (output tokens typically vary by model)
We’re adding AA-LCR to the Artificial Analysis Intelligence Index, and taking the version number to v2.2. Artificial Analysis Intelligence Index v2.2 now includes: MMLU-Pro, GPQA Diamond, AIME 2025, IFBench, LiveCodeBench, SciCode and AA-LCR.
Link to dataset: ArtificialAnalysis/AA-LCR
reacted to
fdaudens's
post with ❤️
5 days ago
Post
2516
Well, it took just 2 hours for
openai/gpt-oss-120b to hit #1 on Hugging Face. Don’t remember seeing anything rise that fast!
- 1 reply
replied to
danielhanchen's
post
5 days ago
I wish OAI delivered as much as you..
reacted to
danielhanchen's
post with 🚀🔥❤️
5 days ago
Post
4050
Run OpenAI's new gpt-oss models locally with Unsloth GGUFs! 🔥🦥
20b GGUF: unsloth/gpt-oss-20b-GGUF
120b GGUF: unsloth/gpt-oss-120b-GGUF
Model will run on 14GB RAM for 20b and 66GB for 120b.
20b GGUF: unsloth/gpt-oss-20b-GGUF
120b GGUF: unsloth/gpt-oss-120b-GGUF
Model will run on 14GB RAM for 20b and 66GB for 120b.
-
- 2 replies
Fr tho? Many ppl say its just benchmaxxed.
reacted to
YerbaPage's
post with 🔥
9 days ago
Post
2864
Latest work on SWE-Bench 🐛
Our two new papers from the SJTU & Huawei: Powered by DeepSeek-V3, we've achieved a new SOTA on the SWE-Bench benchmark!
We introduce two innovative approaches:
⚔️ SWE-Debate: AI agents compete and "debate" to generate the best code fix.
🧠 SWE-Exp: An AI agent learns from past repair "experience" to solve new issues more efficiently.
👇 Explore the future of software development:
SWE-Debate
📄 Paper: https://arxiv.org/abs/2507.23348
💻 Code: https://github.com/YerbaPage/SWE-Debate
SWE-Exp
📄 Paper: https://arxiv.org/abs/2507.23361
💻 Code: https://github.com/YerbaPage/SWE-Exp
Our two new papers from the SJTU & Huawei: Powered by DeepSeek-V3, we've achieved a new SOTA on the SWE-Bench benchmark!
We introduce two innovative approaches:
⚔️ SWE-Debate: AI agents compete and "debate" to generate the best code fix.
🧠 SWE-Exp: An AI agent learns from past repair "experience" to solve new issues more efficiently.
👇 Explore the future of software development:
SWE-Debate
📄 Paper: https://arxiv.org/abs/2507.23348
💻 Code: https://github.com/YerbaPage/SWE-Debate
SWE-Exp
📄 Paper: https://arxiv.org/abs/2507.23361
💻 Code: https://github.com/YerbaPage/SWE-Exp
reacted to
AdinaY's
post with 🔥
9 days ago
Post
1597
Qwen team did it again!!
They just released Qwen3-Coder-30B-A3B-Instruct on the hub🔥
Qwen/Qwen3-Coder-30B-A3B-Instruct
✨ Apache 2.0
✨30B total / 3.3B active (128 experts, 8 top-k)
✨ Native 256K context, extendable to 1M via Yarn
✨ Built for Agentic Coding
They just released Qwen3-Coder-30B-A3B-Instruct on the hub🔥
Qwen/Qwen3-Coder-30B-A3B-Instruct
✨ Apache 2.0
✨30B total / 3.3B active (128 experts, 8 top-k)
✨ Native 256K context, extendable to 1M via Yarn
✨ Built for Agentic Coding
Cool! But ll4 is a bit of a low bar ngl
reacted to
FlameF0X's
post with 👍
15 days ago
Post
2940
The development of SnowflakeCore-G1-7B-MoE. I can't say when it would be publish yet because it's big and it requires a lot of computational power.
-
- 1 reply