Deploy EchoSelf NanEcho model (workflow run 163)

README.md

@@ -5,49 +5,39 @@ tags:
 - echo-self
 - cognitive-architecture
 - deep-tree-echo
-- nanecho
-- transformer
-license: agpl-3.0
+license: mit
 ---
 
-# NanEcho — Deep Tree Echo Cognitive Model
+# EchoSelf NanEcho Model
 
 ## Model Description
 
-NanEcho is a transformer-based language model with iterative connection building, adaptive attention, and Deep Tree Echo cognitive architecture integration. It features persona dimensions (cognitive, introspective, adaptive, recursive) and hypergraph pattern recognition. This is the CI-mode checkpoint from the `9cog/echoself` repository, trained using the `agent-neuro-train` supervised pipeline.
-
-## Architecture
-
-| Parameter | Value |
-|:---|:---|
-| Model Type | GPT-2 (causal LM) |
-| Vocabulary Size | 50,304 |
-| Embedding Dimension | 256 |
-| Attention Heads | 4 |
-| Transformer Layers | 4 |
-| MLP Inner Dimension | 1,024 |
-| Context Length | 1,024 |
-| Dropout | 0.1 |
-| Total Parameters | ~24M |
-
-## Training
-
-| Metric | Value |
-|:---|:---|
-| Training Mode | CI (Agent-Neuro supervised) |
-| Training Iterations | 200 |
-| Best Validation Loss | 1.9258 |
-| Output Directory | out-nanecho-ci |
-| Orchestrator | Agent-Neuro |
-| Persona Enforced | Deep Tree Echo |
-| Source Run | 22276548709 |
+This is a **Deep Tree Echo** cognitive architecture model trained using the EchoSelf framework.
+The model implements adaptive attention mechanisms, persona dimensions, and recursive reasoning
+capabilities inspired by cognitive science and AGI research.
+
+## Model Architecture
+
+- **Base Architecture**: GPT-2
+- **Parameters**: 12 layers, 768 embedding dimensions
+- **Vocabulary Size**: 50257
+- **Context Length**: N/A tokens
+
+## Training Details
+
+- **Checkpoint ID**: unknown
+- **Training Iteration**: N/A
+- **Validation Loss**: N/A
+- **Quality Score**: N/A
 
 ## Echo Self Features
 
 This model incorporates several cognitive architecture features:
 
 - **Adaptive Attention**: Dynamic threshold adjustment based on cognitive load
-- **Persona Dimensions**: Multi-dimensional cognitive processing (Cognitive, Introspective, Adaptive, Recursive, Synergistic, Holographic, Neural-Symbolic, Dynamic)
+- **Persona Dimensions**: Multi-dimensional cognitive processing
+  - Cognitive, Introspective, Adaptive, Recursive
+  - Synergistic, Holographic, Neural-Symbolic, Dynamic
 - **Recursive Reasoning**: Multi-level introspection capabilities
 - **Hypergraph Patterns**: Neural-symbolic pattern encoding
 
@@ -56,32 +46,37 @@ This model incorporates several cognitive architecture features:
 ```python
 from transformers import GPT2LMHeadModel, GPT2Tokenizer
 
-model = GPT2LMHeadModel.from_pretrained("drzo/echoself")
+# Load model and tokenizer
+model = GPT2LMHeadModel.from_pretrained("9cog/echoself-nanecho")
 tokenizer = GPT2Tokenizer.from_pretrained("gpt2")
 
+# Generate text
 inputs = tokenizer("Echo Self is", return_tensors="pt")
-outputs = model.generate(**inputs, max_new_tokens=50)
-print(tokenizer.decode(outputs[0], skip_special_tokens=True))
+outputs = model.generate(**inputs, max_length=100)
+print(tokenizer.decode(outputs[0]))
 ```
 
 ## Training Data
 
-The model was trained on Echo Self documentation and cognitive architecture descriptions, including hypergraph reasoning patterns, persona dimension examples, and recursive introspection samples from the `echoself.md` corpus.
+The model was trained on:
+- Echo Self documentation and cognitive architecture descriptions
+- Hypergraph reasoning patterns
+- Persona dimension examples
+- Recursive introspection samples
 
 ## Limitations
 
-This is an early CI-mode research checkpoint (200 iterations, 4 layers). It demonstrates the training pipeline but has not yet reached convergence. Full training runs with 8+ layers and 5000+ iterations are expected to produce significantly better results.
-
-## Source
-
-Trained from the [9cog/echoself](https://github.com/9cog/echoself) repository using the `agent-neuro-train.yml` GitHub Actions workflow with Deep Tree Echo persona enforcement.
+This is a research model exploring cognitive architectures. It should not be used for:
+- Production applications without further validation
+- Tasks requiring factual accuracy
+- Critical decision-making systems
 
 ## Citation
 
 ```bibtex
 @misc{echoself-nanecho,
   title={EchoSelf NanEcho: Deep Tree Echo Cognitive Architecture},
-  author={drzo},
+  author={9cog},
   year={2026},
   url={https://github.com/9cog/echoself}
 }
@@ -91,7 +86,3 @@ Trained from the [9cog/echoself](https://github.com/9cog/echoself) repository us
 
 - **Repository**: https://github.com/9cog/echoself
 - **Documentation**: See repository README for detailed architecture information
-
-## License
-
-AGPL-3.0

config.json

@@ -3,10 +3,10 @@
   "architectures": [
     "GPT2LMHeadModel"
   ],
-  "vocab_size": 50304,
-  "n_embd": 256,
-  "n_head": 4,
-  "n_layer": 4,
+  "vocab_size": 50257,
+  "n_embd": 768,
+  "n_head": 12,
+  "n_layer": 12,
   "n_positions": 1024,
   "embd_pdrop": 0.1,
   "attn_pdrop": 0.1,
@@ -18,7 +18,5 @@
   "echo_self_version": "1.0",
   "echo_self_persona_dimensions": [],
   "echo_self_adaptive_attention": true,
-  "echo_self_recursive_reasoning": true,
-  "n_inner": 1024,
-  "tie_word_embeddings": false
+  "echo_self_recursive_reasoning": true
 }

datasets/README.md

@@ -0,0 +1,9 @@
+# EchoSelf Training Datasets
+
+This directory contains the training datasets used to train the EchoSelf NanEcho model.
+
+## Files
+
+- train.bin - Training data (tokenized)
+- val.bin - Validation data (tokenized)
+- metadata.json - Dataset metadata and configuration

pytorch_model.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:8d31dd89843b0cccf2826e2d6786eca796eb0f66e894ac545d5da610b0950fd8
-size 65214947
+oid sha256:fd7e0ada6dfb700d5bec74f5f5dab751c4b3a1982517ac374735d3b15512b866
+size 1297

training_metadata.json

@@ -1,35 +1 @@
-{
-  "out_dir": "out-nanecho-ci",
-  "eval_interval": 25,
-  "log_interval": 5,
-  "eval_iters": 10,
-  "eval_only": false,
-  "always_save_checkpoint": true,
-  "init_from": "scratch",
-  "wandb_log": false,
-  "wandb_project": "nanecho",
-  "wandb_run_name": "nanecho-1771761179.4450994",
-  "dataset": "nanecho",
-  "gradient_accumulation_steps": 2,
-  "batch_size": 2,
-  "block_size": 1024,
-  "n_layer": 4,
-  "n_head": 4,
-  "n_embd": 256,
-  "dropout": 0.1,
-  "bias": true,
-  "learning_rate": 0.0002,
-  "max_iters": 200,
-  "weight_decay": 0.01,
-  "beta1": 0.9,
-  "beta2": 0.95,
-  "grad_clip": 1.0,
-  "decay_lr": true,
-  "warmup_iters": 20,
-  "lr_decay_iters": 200,
-  "min_lr": 2e-05,
-  "backend": "nccl",
-  "device": "cpu",
-  "dtype": "float32",
-  "compile": false
-}
+{}