Vedika35
/

File size: 3,034 Bytes
7944312
 
 
 
 
 
 
 
 
7189f14
7944312
7189f14
7944312
 
 
 
 
 
 
 
 
ed560d2
7944312
7189f14
7944312
7189f14
7944312
7189f14
7944312
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
7189f14
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
---
library_name: transformers
license: apache-2.0
license_link: https://huggingface.co/Qwen/Qwen3.5-2B/blob/main/LICENSE
pipeline_tag: image-text-to-text
base_model:
- Qwen/Qwen3.5-2B-Base
---

# Vedika 3.5 flash

<img width="400px" src="https://i.ibb.co/ZR5f1rxF/Blue-and-Black-Minimalist-Brand-Logo-20260423-171504-0000.png">

> [!Note]
> This repository contains model weights and configuration files for the post-trained model in the Hugging Face Transformers format. 
>
> These artifacts are compatible with Hugging Face Transformers, vLLM, SGLang, KTransformers, etc.
> 
> In light of its parameter scale, the intended use cases are prototyping, task-specific fine-tuning, and other research or development purposes.


Over recent months, we have intensified our focus on developing foundation models that deliver exceptional utility and performance. Vedika 3.5 flash represents a significant leap forward, integrating breakthroughs in multimodal learning, architectural efficiency, reinforcement learning scale, and global accessibility to empower developers and enterprises with unprecedented capability and efficiency.

## Vedika 3.5 flash Highlights

Vedika 3.5 flash features the following enhancement:

- **Unified Vision-Language Foundation**: Early fusion training on multimodal tokens achieves cross-generational parity with Vedika and outperforms Vedika 3.5 flash models across reasoning, coding, agents, and visual understanding benchmarks.

- **Efficient Hybrid Architecture**: Gated Delta Networks combined with sparse Mixture-of-Experts deliver high-throughput inference with minimal latency and cost overhead.

- **Scalable RL Generalization**: Reinforcement learning scaled across million-agent environments with progressively complex task distributions for robust real-world adaptability.

- **Global Linguistic Coverage**: Expanded support to 201 languages and dialects, enabling inclusive, worldwide deployment with nuanced cultural and regional understanding.

- **Next-Generation Training Infrastructure**: Near-100% multimodal training efficiency compared to text-only training and asynchronous RL frameworks supporting massive-scale agent scaffolds and environment orchestration.



## Model Overview

- Type: Causal Language Model with Vision Encoder
- Training Stage: Pre-training & Post-training
- Language Model
    - Number of Parameters: 2B
    - Hidden Dimension: 2048
    - Token Embedding: 248320 (Padded)
    - Number of Layers: 24
    - Hidden Layout: 6 × (3 × (Gated DeltaNet → FFN) → 1 × (Gated Attention → FFN))
    - Gated DeltaNet:
        - Number of Linear Attention Heads: 16 for V and 16 for QK
        - Head Dimension: 128
    - Gated Attention:
        - Number of Attention Heads: 8 for Q and 2 for KV
        - Head Dimension: 256
        - Rotary Position Embedding Dimension: 64
    - Feed Forward Network:
        - Intermediate Dimension: 6144
    - LM Output: 248320 (Tied to token embedding)
    - MTP: trained with multi-steps  
- Context Length: 262,144 natively