NRM SOTA Upload: Safetensors + Tokenizer + Config (Step 25000)

README.md

@@ -0,0 +1,63 @@
+---
+license: apache-2.0
+tags:
+  - reasoning
+  - recursive
+  - arc-agi
+  - nvyra-x
+---
+
+# NRM: Nvyra Recursive Reasoning Model
+
+**Developed by Nvyra X** — Fact-Checking and Disinformation Detection Service
+
+## Model Description
+
+NRM (Nvyra Recursive Reasoning Model) is a state-of-the-art reasoning architecture that combines:
+
+- **Mixture of Recursions (MoR)** - Weight-tied transformer blocks applied recursively
+- **Multi-Head Latent Attention (MLA)** - 10× KV cache reduction (DeepSeek-V3)
+- **ConvSwiGLU** - Enhanced nonlinearity from URM paper
+- **Aux-Loss-Free MoE** - Bias-based expert load balancing
+- **PonderNet** - Adaptive computation time
+- **Multi-Token Prediction** - 4-ahead planning
+
+## Training
+
+- **Budget**: $115 ($25 Nebius + $90 Modal)
+- **Hardware**: H200 NVLink GPUs
+- **Framework**: PyTorch 2.9.1, Flash Attention 3, CUDA 12.8
+- **Dataset**: 300K+ reasoning examples (Sudoku, ARC, Logic, Object Tracking)
+
+## Performance
+
+- **Validation Accuracy**: 99.71%
+- **Validation Loss**: 0.0216
+- **Step**: 25000
+
+## Usage
+
+```python
+# This model uses a custom architecture - see repository for full code
+from safetensors.torch import load_file
+weights = load_file("model.safetensors")
+```
+
+## Citation
+
+If you use this model, please cite:
+
+```
+@misc{nrm2025,
+  title={NRM: Nvyra Recursive Reasoning Model},
+  author={Nvyra X Research Team},
+  year={2025},
+  url={https://huggingface.co/Feargal/nvyra-x-reasoning-please}
+}
+```
+
+## References
+
+- [Universal Reasoning Model (URM)](https://arxiv.org/abs/2512.14693)
+- [DeepSeek-V3](https://arxiv.org/abs/2401.02954)
+- [PonderNet](https://arxiv.org/abs/2107.05407)

config.json

@@ -0,0 +1,25 @@
+{
+  "architectures": [
+    "NRM"
+  ],
+  "model_type": "nrm",
+  "dim": 2048,
+  "n_layers": 1,
+  "n_heads": 16,
+  "n_mem_tokens": 64,
+  "vocab_size": 32000,
+  "inner_loops": 8,
+  "outer_loops": 16,
+  "truncation_loops": 2,
+  "moe_experts": 8,
+  "experts_per_token": 2,
+  "num_shared_experts": 2,
+  "use_mla": true,
+  "kv_latent_dim": 512,
+  "rope_head_dim": 64,
+  "rope_base": 10000.0,
+  "mtp_num_heads": 4,
+  "use_conv_swiglu": true,
+  "p_exit": 0.1,
+  "tokenizer_class": "LlamaTokenizer"
+}

ema.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5da0881eac75bcb16cdaf4d0e7bcb0d2daef169e8c62205342cfa4358a52098d
+size 1174807974

model.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:95f6351281a8961cfb644d3b710c300baa771d34e28454966bd6068ba0eaec2b
+size 1174813193

model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0237eeabfb31162b3583576e3f758a16924b966adc1e85e2f2a4d43cc3d45d66
+size 1830160676

optimizer_0.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1af4d526dff07747e2000bf4c8dfe7f651418ace17c2c0464e1a25bf533947ea
+size 2349076825

special_tokens_map.json

@@ -0,0 +1,23 @@
+{
+  "bos_token": {
+    "content": "<s>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "eos_token": {
+    "content": "</s>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "unk_token": {
+    "content": "<unk>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  }
+}

tokenizer.model

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:dadfd56d766715c61d2ef780a525ab43b8e6da4de6865bda3d95fdef5e134055
+size 493443

tokenizer_config.json

@@ -0,0 +1,44 @@
+{
+  "add_bos_token": true,
+  "add_eos_token": false,
+  "add_prefix_space": null,
+  "added_tokens_decoder": {
+    "0": {
+      "content": "<unk>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "1": {
+      "content": "<s>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "2": {
+      "content": "</s>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    }
+  },
+  "additional_special_tokens": [],
+  "bos_token": "<s>",
+  "clean_up_tokenization_spaces": false,
+  "eos_token": "</s>",
+  "extra_special_tokens": {},
+  "legacy": false,
+  "model_max_length": 1000000000000000019884624838656,
+  "pad_token": null,
+  "sp_model_kwargs": {},
+  "spaces_between_special_tokens": false,
+  "tokenizer_class": "LlamaTokenizer",
+  "unk_token": "<unk>",
+  "use_default_system_prompt": false
+}

train_state.json

@@ -0,0 +1 @@
+{"step": 25000, "val_acc": 0.9971077647478286, "val_loss": 0.021596687794636502}