nvidia/NVIDIA-NemotronLabs-VoiceChat-11B
11B • Updated • 2.3k • 406
👋 Open to Work
Fredy Rivera
FredyRivera-dev
AI & ML interests
None yet
Recent Activity
liked a model 18 minutes ago
nvidia/NVIDIA-NemotronLabs-VoiceChat-11B liked a dataset about 1 hour ago
bigcode/starcoderdata reacted to theirpost with 🚀 about 3 hours ago
I've written a technical blog post about how we create a multimodal model:
Kairos: a multimodal model built with LFM2.5-2.6B as the LLM, MoonViT-3D (the vision tower of Kimi-K2.6) as the vision encoder, and a custom projector.
The original plan was LLaVA's approach, two stages: first align the projector with the LLM frozen, and then train the projector + LLM together. The first stage worked in terms of loss (ablation with +3.7 nats in favor of the image), but in free generation the image shifted the logits without changing the argmax: the model received the image and ignored it. That's why we jumped directly to early fusion, with a reasoning dataset.
For that, we created Kairos-Multimodal-Reasoning: 116,357 examples with explicit reasoning traces, generated through distillation (60,041 from LLaVA-CC3M, 2,295 from WebSight, and 54,021 from Zebra-CoT), with GPT 5.6 Luna, Inkling, Qwen 3.6 27B, and Qwen 3.7 Plus as teachers.
The training, in two phases:
1. Projector through backbone with 80k image-text pairs (Kairos-Proj-80k).
2. Projector + LoRA (r=16) with 30k examples from the reasoning dataset (Kairos-Alig-30k).
Everything is open source:
- Full blog post with the process: https://aquiles-ai.vercel.app/blog/kairos-a-multimodal-model
- Implementation: https://github.com/Aquiles-ai/Kairos
To be honest: the checkpoints are not a competent model, they are experimental artifacts. But they validated the approach and precisely defined what the next iteration needs.
https://huggingface.co/collections/Aquiles-ai/kairos
https://huggingface.co/Aquiles-ai/MoonViT-3D
https://huggingface.co/LiquidAI/LFM2.5-2.6BOrganizations
liked a
model 18 minutes ago
liked a
dataset about 1 hour ago
reacted to their post with 🚀 about 3 hours ago
Post
17
I've written a technical blog post about how we create a multimodal model:
Kairos: a multimodal model built with LFM2.5-2.6B as the LLM, MoonViT-3D (the vision tower of Kimi-K2.6) as the vision encoder, and a custom projector.
The original plan was LLaVA's approach, two stages: first align the projector with the LLM frozen, and then train the projector + LLM together. The first stage worked in terms of loss (ablation with +3.7 nats in favor of the image), but in free generation the image shifted the logits without changing the argmax: the model received the image and ignored it. That's why we jumped directly to early fusion, with a reasoning dataset.
For that, we created Kairos-Multimodal-Reasoning: 116,357 examples with explicit reasoning traces, generated through distillation (60,041 from LLaVA-CC3M, 2,295 from WebSight, and 54,021 from Zebra-CoT), with GPT 5.6 Luna, Inkling, Qwen 3.6 27B, and Qwen 3.7 Plus as teachers.
The training, in two phases:
1. Projector through backbone with 80k image-text pairs (Kairos-Proj-80k).
2. Projector + LoRA (r=16) with 30k examples from the reasoning dataset (Kairos-Alig-30k).
Everything is open source:
- Full blog post with the process: https://aquiles-ai.vercel.app/blog/kairos-a-multimodal-model
- Implementation: https://github.com/Aquiles-ai/Kairos
To be honest: the checkpoints are not a competent model, they are experimental artifacts. But they validated the approach and precisely defined what the next iteration needs.
https://huggingface.co/collections/Aquiles-ai/kairos
Aquiles-ai/MoonViT-3D
LiquidAI/LFM2.5-2.6B
Kairos: a multimodal model built with LFM2.5-2.6B as the LLM, MoonViT-3D (the vision tower of Kimi-K2.6) as the vision encoder, and a custom projector.
The original plan was LLaVA's approach, two stages: first align the projector with the LLM frozen, and then train the projector + LLM together. The first stage worked in terms of loss (ablation with +3.7 nats in favor of the image), but in free generation the image shifted the logits without changing the argmax: the model received the image and ignored it. That's why we jumped directly to early fusion, with a reasoning dataset.
For that, we created Kairos-Multimodal-Reasoning: 116,357 examples with explicit reasoning traces, generated through distillation (60,041 from LLaVA-CC3M, 2,295 from WebSight, and 54,021 from Zebra-CoT), with GPT 5.6 Luna, Inkling, Qwen 3.6 27B, and Qwen 3.7 Plus as teachers.
The training, in two phases:
1. Projector through backbone with 80k image-text pairs (Kairos-Proj-80k).
2. Projector + LoRA (r=16) with 30k examples from the reasoning dataset (Kairos-Alig-30k).
Everything is open source:
- Full blog post with the process: https://aquiles-ai.vercel.app/blog/kairos-a-multimodal-model
- Implementation: https://github.com/Aquiles-ai/Kairos
To be honest: the checkpoints are not a competent model, they are experimental artifacts. But they validated the approach and precisely defined what the next iteration needs.
https://huggingface.co/collections/Aquiles-ai/kairos
Aquiles-ai/MoonViT-3D
LiquidAI/LFM2.5-2.6B
posted an update about 3 hours ago
Post
17
I've written a technical blog post about how we create a multimodal model:
Kairos: a multimodal model built with LFM2.5-2.6B as the LLM, MoonViT-3D (the vision tower of Kimi-K2.6) as the vision encoder, and a custom projector.
The original plan was LLaVA's approach, two stages: first align the projector with the LLM frozen, and then train the projector + LLM together. The first stage worked in terms of loss (ablation with +3.7 nats in favor of the image), but in free generation the image shifted the logits without changing the argmax: the model received the image and ignored it. That's why we jumped directly to early fusion, with a reasoning dataset.
For that, we created Kairos-Multimodal-Reasoning: 116,357 examples with explicit reasoning traces, generated through distillation (60,041 from LLaVA-CC3M, 2,295 from WebSight, and 54,021 from Zebra-CoT), with GPT 5.6 Luna, Inkling, Qwen 3.6 27B, and Qwen 3.7 Plus as teachers.
The training, in two phases:
1. Projector through backbone with 80k image-text pairs (Kairos-Proj-80k).
2. Projector + LoRA (r=16) with 30k examples from the reasoning dataset (Kairos-Alig-30k).
Everything is open source:
- Full blog post with the process: https://aquiles-ai.vercel.app/blog/kairos-a-multimodal-model
- Implementation: https://github.com/Aquiles-ai/Kairos
To be honest: the checkpoints are not a competent model, they are experimental artifacts. But they validated the approach and precisely defined what the next iteration needs.
https://huggingface.co/collections/Aquiles-ai/kairos
Aquiles-ai/MoonViT-3D
LiquidAI/LFM2.5-2.6B
Kairos: a multimodal model built with LFM2.5-2.6B as the LLM, MoonViT-3D (the vision tower of Kimi-K2.6) as the vision encoder, and a custom projector.
The original plan was LLaVA's approach, two stages: first align the projector with the LLM frozen, and then train the projector + LLM together. The first stage worked in terms of loss (ablation with +3.7 nats in favor of the image), but in free generation the image shifted the logits without changing the argmax: the model received the image and ignored it. That's why we jumped directly to early fusion, with a reasoning dataset.
For that, we created Kairos-Multimodal-Reasoning: 116,357 examples with explicit reasoning traces, generated through distillation (60,041 from LLaVA-CC3M, 2,295 from WebSight, and 54,021 from Zebra-CoT), with GPT 5.6 Luna, Inkling, Qwen 3.6 27B, and Qwen 3.7 Plus as teachers.
The training, in two phases:
1. Projector through backbone with 80k image-text pairs (Kairos-Proj-80k).
2. Projector + LoRA (r=16) with 30k examples from the reasoning dataset (Kairos-Alig-30k).
Everything is open source:
- Full blog post with the process: https://aquiles-ai.vercel.app/blog/kairos-a-multimodal-model
- Implementation: https://github.com/Aquiles-ai/Kairos
To be honest: the checkpoints are not a competent model, they are experimental artifacts. But they validated the approach and precisely defined what the next iteration needs.
https://huggingface.co/collections/Aquiles-ai/kairos
Aquiles-ai/MoonViT-3D
LiquidAI/LFM2.5-2.6B
liked a
model about 4 hours ago
reacted to Banaxi-Tech's post with 👀 about 14 hours ago
Post
1864
We're releasing BananaMind 2.1 Unified, a 35M three-tower relay model where the two output towers can only talk to each other through a silent middle tower that has no output head and no loss term.
Tower B trains entirely on indirect gradient. It was never told what to predict. It woke up anyway. Jacobian lens shows it carrying the correct answer ("Paris", "oxygen", "blue") at its deepest layer. Its bridge gates grew 4-47x from init. Feed it from only one side and the representations collapse to junk — it needs both outer towers to become semantic.
The PIQA result is the cleanest demonstration: Tower A alone scores 50.11 (chance). Tower C alone 52.12. Full system 61.75. All physical reasoning lives in the integration. The 35M three-tower beats the 50M single-tower BananaMind 2 Medium on PIQA.
Trained on 38B tokens in ~7.5 hours on 8x RTX PRO 6000. Ships with 7 ablation modes so you can surgically cut the model apart without retraining. Full training logs, J-lens fits, and eval outputs for every mode included.
This is a research model. It is interesting because you can take it apart.
BananaMind/BananaMind-2.1-Unified
Follow us for more:
BananaMind
@Banaxi-Tech
Tower B trains entirely on indirect gradient. It was never told what to predict. It woke up anyway. Jacobian lens shows it carrying the correct answer ("Paris", "oxygen", "blue") at its deepest layer. Its bridge gates grew 4-47x from init. Feed it from only one side and the representations collapse to junk — it needs both outer towers to become semantic.
The PIQA result is the cleanest demonstration: Tower A alone scores 50.11 (chance). Tower C alone 52.12. Full system 61.75. All physical reasoning lives in the integration. The 35M three-tower beats the 50M single-tower BananaMind 2 Medium on PIQA.
Trained on 38B tokens in ~7.5 hours on 8x RTX PRO 6000. Ships with 7 ablation modes so you can surgically cut the model apart without retraining. Full training logs, J-lens fits, and eval outputs for every mode included.
This is a research model. It is interesting because you can take it apart.
BananaMind/BananaMind-2.1-Unified
Follow us for more:
@Banaxi-Tech
- 3 replies
liked 2
models about 18 hours ago
published a
model about 18 hours ago
liked a
model about 18 hours ago
updated 3
models about 22 hours ago
liked 3
models 1 day ago
liked 2
models 2 days ago
liked a
dataset 2 days ago
published a
model 3 days ago