Update README.md

README.md

@@ -8,8 +8,6 @@ base_model: convaiinnovations/laya
 
 > Source model card: `Falconsai/laya-v795` @ `main`, carried verbatim below. Its licence is the repository's. The Model Surgeon record follows it.
 
-> Source model card: `convaiinnovations/laya` @ `main`, carried verbatim below. Its licence is the repository's. The Model Surgeon record follows it.
-
 # Laya
 
 **Multilingual, non-autoregressive System 1 decision model.** Give it a **state** (text, email, ticket, or JSON) and **typed questions**; it returns typed answers with mathematically calibrated probabilities in a single forward pass (~33 ms) across 100+ languages. Trained with reinforcement learning against strictly proper scoring rules (**RLCD**), so reporting honest probabilities is the only way to maximise reward. It never generates text, so there is nothing to parse and nothing to hallucinate.
@@ -22,7 +20,7 @@ base_model: convaiinnovations/laya
 
 | Checkpoint | Backbone Encoder | Params | Context | Best at |
 |---|---|---|---|---|
-| **`Falconsai/laya-v795`** (this repo root) | ModernBERT-large | 421M | 512 | English text, guardrails, email triage |
+| **`Falconsai/laya-v796`** (this repo root) | ModernBERT-large | 421M | 512 | English text, guardrails, email triage |
 
 ## What's new in laya 0.3.11
 
@@ -109,7 +107,7 @@ Every result carries full routing metadata explaining why the choice was made:
 res_hi["routing"]
 # {
 #   'model': 'multilingual',
-#   'repo': 'Falconsai/laya-v795/multilingual',
+#   'repo': 'Falconsai/laya-v796/multilingual',
 #   'reason': 'non-Latin script (devanagari, 100% of letters); the English checkpoint cannot read it'
 # }
 ```
@@ -181,9 +179,9 @@ If you only need a single checkpoint for a dedicated pipeline:
 import laya
 
 # 1. Load from the repo root or subfolders (downloads only the requested weights)
-agent = laya.load("Falconsai/laya-v795")                           # English root (~808 MB)
-agent_ml = laya.load("Falconsai/laya-v795", subfolder="multilingual") # 100+ languages (~647 MB)
-agent_td = laya.load("Falconsai/laya-v795a", subfolder="typed-decisions")
+agent = laya.load("Falconsai/laya-v796")                           # English root (~808 MB)
+agent_ml = laya.load("Falconsai/laya-v796", subfolder="multilingual") # 100+ languages (~647 MB)
+agent_td = laya.load("Falconsai/laya-v796", subfolder="typed-decisions")
 
 # 2. Run all questions in ONE single forward pass (~35 ms on GPU)
 result = agent.predict(state, questions)
@@ -327,7 +325,7 @@ Apache 2.0 · Convai Innovations
 
 ---
 
-# Model Card — Falconsai/laya-v795/
+# Model Card — Falconsai/laya-v796/
 
 **[View in Model Surgeon](https://surgeon.falcons.ai/?hub=Falconsai/laya-v796)**
 
@@ -340,7 +338,7 @@ the Surgeon's public verifier or with the bundled `verify_attestation.py`).
 - Source format: `safetensors` · Intended task: not declared
 - `config.json`: synthesized from the anatomy (no source config.json)
 - Source license: apache-2.0
-- Lineage chain: 1 surgery (no prior attestation reachable) · Falconsai/laya-v795
+- Lineage chain: 1 surgery (no prior attestation reachable) · Falconsai/laya-v796
 - Post-surgery totals: 421,293,830 parameters ·
   206 tensors
 - Compute estimate: 46.507164 GFLOPs (comparison
@@ -353,7 +351,7 @@ the Surgeon's public verifier or with the bundled `verify_attestation.py`).
 - Quantized tensors (F32→F16): 0
 
 ## Surgery Log (ordered)
-1. **load** — hub:Falconsai/laya-v795/model.safetensors (842.6 MB, safetensors)
+1. **load** — hub:Falconsai/laya-v796/model.safetensors (842.6 MB, safetensors)
 
 ## Validation
 - Tissue imaging: not run
@@ -372,9 +370,9 @@ This is evidence, not legal advice.
 
 ---
 
-# Model Card — Falconsai/laya-v795/
+# Model Card — Falconsai/laya-v796/
 
-**[View in Model Surgeon](https://surgeon.falcons.ai/?hub=Falconsai/laya-v795)**
+**[View in Model Surgeon](https://surgeon.falcons.ai/?hub=Falconsai/laya-v796)**
 
 This card is generated from the surgical record itself; the package's
 `lineage.intoto.jsonl` is the signed source of truth (verify it free at
@@ -385,20 +383,20 @@ the Surgeon's public verifier or with the bundled `verify_attestation.py`).
 - Source format: `safetensors` · Intended task: not declared
 - `config.json`: the repo's config.json, edited
 - Source license: apache-2.0
-- Lineage chain: 2 surgeries — prior signed by `ed25519:b2eeebe30a12` (FALCONS.AI Model Surgeon V7.95; 1 earlier operation(s) carried) · Falconsai/laya-v795
+- Lineage chain: 2 surgeries — prior signed by `ed25519:b2eeebe30a12` (FALCONS.AI Model Surgeon V7.95; 1 earlier operation(s) carried) · Falconsai/laya-v796
 - Post-surgery totals: 421,293,938 parameters ·
   207 tensors
 - Compute estimate: 46.507164 GFLOPs (comparison
   metric, not a measurement)
 
 ## Provenance & operations
-- Parents: Falconsai/laya-v795/model_edited.safetensors
+- Parents: Falconsai/laya-v796/model_edited.safetensors
 - Operations performed: load×1
 - Weight merges recorded: 0
 - Quantized tensors (F32→F16): 0
 
 ## Surgery Log (ordered)
-1. **load** — hub:Falconsai/laya-v795/model_edited.safetensors (842.6 MB, safetensors)
+1. **load** — hub:Falconsai/laya-v796/model_edited.safetensors (842.6 MB, safetensors)
 
 ## Validation
 - Tissue imaging: not run