Finalize Bertose IAR resolver release names

README.md

@@ -10,29 +10,29 @@ tags:
 - pytorch
 ---
 
-# Bertose IAR Ambiguity Resolver
+# Bertose IAR Resolver
 
-Draft private release for Bertose ambiguity-resolution inference.
-
-This repository contains the contrastive Bertose checkpoint used to score ambiguous WURCS BPE tokens and support iterative ambiguity resolution.
+This repository contains the contrastively refined Bertose checkpoint used for iterative ambiguity resolution (IAR) over ambiguous WURCS BPE tokens.
 
 ## Files
 
-- `checkpoints/best_v51_contrastive_model.pt` - contrastive ambiguity-resolution checkpoint.
+- `checkpoints/bertose_iar_resolver.pt` - Bertose IAR checkpoint.
 - `vocab/bpe_vocabulary.json` - WURCS BPE vocabulary.
-- `vocab/bpe_ambiguity_tokens.json` - ambiguous BPE token map used by the resolver, including common WURCS uncertainty-marker tokens.
-- `src/multimodal_glycan_bert_v3.py` - model definition.
-- `src/glycan_bert.py` - base BERT layers used by the multimodal model.
+- `vocab/bpe_ambiguity_tokens.json` - ambiguous-token map used by the resolver.
+- `src/bertose_model.py` - Bertose model definition.
+- `src/bertose_layers.py` - Transformer layers used by Bertose.
 - `src/wurcs_bpe_tokenizer.py` - WURCS BPE tokenizer.
 
-## Expected Input
+## Input
 
-Single glycan or batch CSV with WURCS strings.
+Provide one WURCS glycan string or a CSV batch with a WURCS column. The resolver is intended for glycans that already contain uncertainty markers in WURCS form.
 
 ## Output
 
 Token-level ambiguity-resolution predictions with confidence scores. The companion notebook writes both summary and detail CSVs for batch runs.
 
-## Draft Notes
+## Scope
+
+The resolver provides model-backed token updates and confidence values for ambiguous positions. It does not claim to reconstruct a final canonical WURCS string by itself, and it does not perform IUPAC-condensed to WURCS conversion.
 
-This release does not claim to reconstruct final canonical WURCS strings by itself. It provides model-backed token-level updates and confidence values for ambiguous positions.
+The final license and citation text should be set before public release.

SHA256SUMS

@@ -1,34 +1,10 @@
-684888c0ebb17f374298b65ee2807526c066094c701bcc7ebbe1c1095f494fc1  ./.cache/huggingface/.gitignore
-e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855  ./.cache/huggingface/upload/.gitattributes.lock
-3098e38608a2c2375ac1f78d4c4f52680796f4ff9c0dbaad6b4f0b110fbc7fc3  ./.cache/huggingface/upload/.gitattributes.metadata
-e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855  ./.cache/huggingface/upload/README.md.lock
-ecc75cccadd48cf2cc8d22daec846b6a760f492162ca145c4cfef3536dafcc2a  ./.cache/huggingface/upload/README.md.metadata
-e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855  ./.cache/huggingface/upload/SHA256SUMS.lock
-aa2c2e921401dba265bdd190a662861cffd8ff05eaf6ae45a96a25385bd6c5e4  ./.cache/huggingface/upload/SHA256SUMS.metadata
-e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855  ./.cache/huggingface/upload/checkpoints/best_v51_contrastive_model.pt.lock
-0bc5904fe02b6a64df35829729c29d40f0c0a795d586b10d844fbee91e6fa0e7  ./.cache/huggingface/upload/checkpoints/best_v51_contrastive_model.pt.metadata
-e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855  ./.cache/huggingface/upload/config.json.lock
-9370200adedd2172ffd8459528e7fd47c5913bf9e791f5b731b0e16121ca3ebf  ./.cache/huggingface/upload/config.json.metadata
-e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855  ./.cache/huggingface/upload/requirements.txt.lock
-fef169fb7e8af9c14c21240bb9034cd567bd18dc327ab39423d68ba3b2ee413a  ./.cache/huggingface/upload/requirements.txt.metadata
-e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855  ./.cache/huggingface/upload/src/multimodal_glycan_bert_v3.py.lock
-65dcbe6e66d8bba618e4d22209bd2e83b73b5de767b892c1bbd43db1c9326f42  ./.cache/huggingface/upload/src/multimodal_glycan_bert_v3.py.metadata
-e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855  ./.cache/huggingface/upload/src/wurcs_bpe_tokenizer.py.lock
-28ca0e31a94c80afc124627b62a574125270a5f269bdff012fd36b465578dc82  ./.cache/huggingface/upload/src/wurcs_bpe_tokenizer.py.metadata
-e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855  ./.cache/huggingface/upload/vocab/bpe_ambiguity_tokens.json.lock
-eb200fe67e613751c0571950e9a7f22f9f44fde0f85b73a40d392189a203f465  ./.cache/huggingface/upload/vocab/bpe_ambiguity_tokens.json.metadata
-e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855  ./.cache/huggingface/upload/vocab/bpe_vocabulary.json.lock
-c00560217b399adfb341aacc38053299c7d4b33b4229e89e68275cd454bb7f5b  ./.cache/huggingface/upload/vocab/bpe_vocabulary.json.metadata
 622368f62c23e97e9137c277eaadcc93ee3901cbb420b591422bb1c2e19689a5  ./.gitattributes
-4d7dc814c23492bd75c154e35436fff108221285956106deeb117e561a6c560b  ./README.md
-ae468f4e8c06dc0c3848138a474dc43249aa6d14dfd0df8f58d68fcaad371152  ./checkpoints/best_v51_contrastive_model.pt
-daf55c190fece0678064e41697a9545592beb1285f8aa74e595b933b9d37b4c2  ./config.json
+06de01e215e940f2633d47561e541b6574fe2c4366ffec9ef3014eef66fa6b6b  ./README.md
+ae468f4e8c06dc0c3848138a474dc43249aa6d14dfd0df8f58d68fcaad371152  ./checkpoints/bertose_iar_resolver.pt
+99dc07803204bf4d8e9bc6f7d04e18204e9dbc6582ef3582904a80371cff09a8  ./config.json
 6a56e6f73b8f874470ecde6e538f3f5029ae23aa6c10559817d1c2a8b59b7c0f  ./requirements.txt
-789fde2ce01f83a5bb363aee29fe33809e2a7015c47c1915655c208d8beec496  ./src/__pycache__/glycan_bert.cpython-312.pyc
-9a0d7855e244b3a1ff369eba4da5303d528f067d1092fefd5a93c9db164de000  ./src/__pycache__/multimodal_glycan_bert_v3.cpython-312.pyc
-62259d1fe3d8736e57cadf8ce5a8bf24a7b73368d4d653c2e0d56ac94b94fe76  ./src/__pycache__/wurcs_bpe_tokenizer.cpython-312.pyc
-b69f14c9976951325e3a0a4e8107a16126e67d410e966650f513f1f538a732bb  ./src/glycan_bert.py
-0d9ce16bf90242f38621d64cd974ea5679bff4c2013bea8d7bffe1b8dd120794  ./src/multimodal_glycan_bert_v3.py
+b69f14c9976951325e3a0a4e8107a16126e67d410e966650f513f1f538a732bb  ./src/bertose_layers.py
+f247a6c09132a61cb649acfe022b269b5b94c37a5069fcb62045f3340b96b191  ./src/bertose_model.py
 0bc54399362945601bcfd403441fc80968d173200dd0561f57568b2053a94839  ./src/wurcs_bpe_tokenizer.py
 ae6ab1ee4f2be992099ee5766de073954c74ccb005c490179cc70418c587c5b7  ./vocab/bpe_ambiguity_tokens.json
 6a572afdf53f1494ab96c896876b824ca7ea749777352606aa9f96bf270ceecc  ./vocab/bpe_vocabulary.json

config.json

@@ -1,7 +1,7 @@
 {
   "model_family": "Bertose",
-  "release_name": "bertose-iar-ambiguity-resolver",
-  "checkpoint": "checkpoints/best_v51_contrastive_model.pt",
+  "release_name": "bertose-iar-resolver",
+  "checkpoint": "checkpoints/bertose_iar_resolver.pt",
   "vocabulary": "vocab/bpe_vocabulary.json",
   "ambiguity_tokens": "vocab/bpe_ambiguity_tokens.json",
   "embedding_dim": 768,

src/{multimodal_glycan_bert_v3.py → bertose_model.py}

@@ -15,9 +15,9 @@ from typing import Dict, Optional, Tuple
 import math
 
 try:
-    from .glycan_bert import GlycanBERTConfig, GlycanBERTEmbeddings, GlycanBERTLayer
+    from .bertose_layers import GlycanBERTConfig, GlycanBERTEmbeddings, GlycanBERTLayer
 except ImportError:
-    from glycan_bert import GlycanBERTConfig, GlycanBERTEmbeddings, GlycanBERTLayer
+    from bertose_layers import GlycanBERTConfig, GlycanBERTEmbeddings, GlycanBERTLayer
 
 
 class ConvGlycanBERTEmbeddings(nn.Module):